The fog of war: development, conflict narratives, and civilian victimisation in Colombia

Abstract Conflict generates uncertainty and a supply of violent actors willing to contract their services to third parties that makes identifying motivation for civilian victimisation difficult. In this paper, I examine how elites not only have opportunities to capitalise on violence during war for developmental interests, but also take advantage of the fog of war to lump all cases of victimisation under a clear overarching narrative of political contestation rather than repression for economic interests. I build on the literature of developmental violence in conflict to assess two cases of dam construction by the Public Enterprises of Medellin (EPM) in Antioquia, Colombia between 1995 and 2018: Hidroituango and Porce II. I argue that civilians opposing the dams were associated with one of the armed groups in the territory and elites dismissed their victimisation as a product of competition between insurgents and counterinsurgents. This conflict narrative obscures other potential realities and precludes deeper investigation into the ambiguous motivations behind attacks, even as the victims themselves may challenge and present counternarratives of developmental violence.


Introduction
War generates overlapping patterns of violence and motivation for civilian victimisation.On one level there are the clear-cut political aims of armed actors disputing territory with the aim to establish political authority; in civil wars this is the realm of the insurgent vs the state.Civilian victimisation is conceptualised along a logic of territorial competition, where civilians are killed for supporting the wrong group or as innocent bystanders (Steele 2017;De la Calle 2017;Stanton 2015;Kalyvas 2006).Counterinsurgent militias or paramilitary proxies complicate the landscape further and muddy the ability to determine the sequence of events and motivation behind civilian victimisation (Williamson 2023;Carey, Colaresi, and Mitchell 2015).On another level, developmental elites -private and state actors involved in large-scale infrastructure -may also rely on extra-legal violence during conflict to advance their economic interests against civilian challengers (Le Billon and Lujala 2020; Middeldorp and Le Billon 2019;McNeish 2018).Reno argues that this collaboration between elites and conflict actors is often perceived as '[t]he cheapest and most efficacious option for indirect management of disorder' (Reno 2004, 623).
Armed conflict, however, does not simply offer elites opportunities to advance their interests through violence but also offers the discursive capacity to obscure personal or economic incentives in civilian repression.As Tellez (2022, 407) notes, civil wars are 'information-poor environments where uncertainty reigns as to which actors are responsible for what actions' .In Colombia, for instance, a large proportion of wartime deaths are never attributed to a known perpetrator (Echandía and Salas 2008).This heightens the fog of war not just for soldiers, but for civilians and researchers, who try to divine motive from imprecise or muddled accounts of victimisation.As a result, both academic and policy audiences seek simpler narratives of conflicts where 'disputes are conceptualised as having single-issue causal factors' (Olumba et al. 2022(Olumba et al. , 2084)), rather than examine the relationship between these overlapping patterns of violence.
In this paper, I argue that the uncertainty of war provides elites a narrative that associates civilians to armed groups and dismisses other motives for their victimisation, including development interests.I build on scholarship that examines the links between extra-legal violence and developmental elites to show how elites can take advantage of the uncertainty of war in what I call an 'elite inversion' of Kalyvas (2006, 338) 'malicious denunciation' to advance individual interests.In this article, I demonstrate how the framing of a political armed conflict becomes entangled with developmental interests and challenges.I present my findingsbased on interviews with local leaders, project managers, and international observers, and secondary data collected from archival sources -in the displacement and assassination of civilian activists in the construction of the Hidroituango dam between 1998 and 2018 and the assassination of union leaders protesting working conditions in the construction of the Porce II dam in 1997 and 1998.
Through these two cases I present novel empirical contributions and show how overlapping patterns of violence allows elites to use conflict narratives to undermine civilian challengers and dismiss accusations relating violence to development.As Chenoweth, Perkoski, andKang (2017, 1962) observe, states and elites 'rarely use repression in a vacuum' .Uncertainty around civilian victimisation led the private-public corporation Public Enterprises of Medellin (EPM), government officials, the military, and even judicial ruling by the Colombian court systems to frame these events -intentionally or not -as products of the territorial conflict dynamics without further investigation.Conflict provides such a compelling framework to understand violence that even parties without personal incentive to generate these narratives may reproduce them, such as academics or human rights organisations, particularly given the uncertainty surrounding the precise motivation behind attacks.These narratives may be challenged by the victims, but often such challenges bear no legal consequences or worse, appear to invite further repression.

Conflict, opportunity, and elite interests
War provides civilians the opportunity to take advantage of armed groups to falsely accuse other civilians of involvement with opposing groups to settle personal scores or advance personal interests (Barter 2016, 26;Bhatia 2021).Kalyvas terms this 'malicious denunciation' (2006,(338)(339) which provides 'the less powerful and the disadvantaged an opportunity to take out their spite' , to use 'systems hungry for information' for their own interests' (Gellately 2001, 24;Clark 2014;Janssens 2021).
The elite inversion of Kalyvas' 'malicious denunciation' can be deployed to silence or repress challenges to elite interests.Evidently if war offers opportunity to the less powerful, it offers ample opportunity for elites -a group of people 'who have vastly disproportionate control over or access to a resource' (Khan 2012, 362).The resources available to elites mean that these denunciations are not necessarily geared at 'tricking' one armed group to repress difficult civilians (activists, unionists, or un-cooperative land-owners) but rather can also obfuscate motives behind systematic civilian victimisation to broader audiences and avoid investigation even as targeted civilians are clear about who they fear criticising.
In this paper, I examine the elite inversion of this dynamic by focusing on how developmental elites respond to information scarcity and uncertainty.I use the term developmental elites, not merely economic elites, because I include both the private and public actors involved in funding and managing large-scale infrastructural roll-out, such as private contractors, public agencies, officials, and other government branches involved.Developmental elites may rely on colonial or entrepreneurial narratives that justify their interests and supress civilian opposition (Lima and Mafra 2023;Grajales 2020).Civil war not only facilitates material opportunities for elites to divest civilians from desired land in cases as diverse as Colombia, Indonesia, and Myanmar (Gutiérrez-Sanín and Vargas Reina 2016;Lund 2018;O'Connor 2011) or work with armed groups to capture resources such as cases in Central Africa (Berman et al. 2017;Patel 2016), but also a new narrative of justification.
Research has demonstrated that elites may have established agreements with armed groups built through mutual economic interests; counterinsurgent forces frequently target non-violent protestors as 'extremists' or insurgent' to enforce the trajectory of elite development efforts (Dunlap 2019;Brock and Dunlap 2018;Grajales 2013).In the Philippines, for example, anti-dam activists were stigmatised as terrorist supporters and targeted by paramilitary actors (Delina 2020, 6).This stigmatisation happens when a person possesses or is believed to possess 'some attribute or characteristic that conveys a social identity that is devalued in a particular social context' (Crocker, Major, and Steele 1998, 505).It assigns blame to the victims for their supposed association to undesirable actors [i.e.armed groups] and reduces the need for investigation and accountability by state actors, similar to the 'politics of naming' of conflict actors (Bhatia 2005).Blaming the conflict, rather than development, for civilian victimisation draws useful 'political frontiers between "insiders" and "outsiders"' (Howarth, Norval, and Stavrakakis 2000, 4), that retains elites' plausible deniability (Stephens 2017;Maassarani 2005).
The difficulty in investigating violence in ongoing conflict combined with the presence of armed groups in a territory offers a convenient discursive framing for civilian victimisation.This framing often becomes the dominating narrative as elites have better access to wider audiences through official statements, media, or even judicial rulings (Jäger and Maier 2015, 39).Conflict-affected regions like other informal or illicit spaces are often 'subject to mischaracterizations (e.g.marginal, criminal, and chaotic) by powerful actors who stand to gain from certain narratives and thus strategically cultivate ambiguity' that requires further investigation to challenge these dominating narratives (Dev et al. 2022, 666).In this paper I examine how elites capitalise on the uncertainty generated by armed conflict to advance these stigmatising narratives that undermine and discredit civilian opposition and the violence they face by shadowy actors.

Counterinsurgency and development in Colombia
In Colombia the link between private economic interests and the rise of paramilitary groups in the 1980s and 1990s has been well-established (Ballvé 2020(Ballvé , 2013;;Grajales 2013Grajales , 2011)), underscored by historic inequalities and weak property rights that enabled large-scale land appropriation (Kinosian 2012).Molano argues that civilian displacement in Colombia 'is not an 'effect' of conflict but rather a weapon of war itself and a strategy of economic accumulation ' (2000, 42).The paramilitary violence against the Union Patriotica political party in the early 1990s was only in part a reaction to its association of the Armed Revolutionary Forces of Colombia (FARC), but also a systematic destruction of union leaders, civil rights activists, and left-wing politicians (Gomez-Suarez 2007).Indeed, statements by former president Alvaro Uribe on the 'terrorist nature' of anti-development activists delegitimsed their efforts and made them targets for violent attacks by the military and paramilitary groups (Chambers 2013, 126).Yet I argue that elites do not merely label undesirable civilians as terrorists, but rather capitalise on the very murkiness that conflict produces to attribute instances of civilian victimisation to local territorial competition.
There are few high-profile incidents that explicitly tie developmental interests to civilian victimisation.However, those that emerge reveal how elites instrumentalise not only conflict actors but conflict narratives.In 2020, Salvatore Mancuso, a former paramilitary commander, testified that he was sent by the Colombian state and its private contractors to supress dissent to the construction of the Urra dam in the department of Cordoba in 2001.In his testimony, he states that the targeting of the local Embera Indigenous people, especially the killing of leader Kimy Pernía Domicó, was to quash the opposition to the dam while disguising it as a counterinsurgent operation: What happened to the Indigenous leader Kimy Pernía Domicó was a crime of the state.The excuses they were giving, because you remember that the State was censuring you and prohibiting you from entering the area freely, because they said you were supporting subversive armed groups in the area, and that when they were going to build the hydroelectric dam [Urrá], they started the systematic actions by the Colombian state intended to weaken you and end whatever effort to claim or realise your rights.(Florez Arias 2020) In 2001 Colombian officials had stated that Pernía Domicó's disappearance was linked to his support for the FARC (Jaramillo 2011, 147), which Mancuso reinforced in his own testimony in 2007 (Florez Arias 2020).Yet his later testimony, after his sudden extradition in 2008 to the USA by the Colombian government, supported the claims the Embera had been making since 2001 about the dangers of opposing the dam (Jaramillo 2011;Rodríguez Garavito and Orduz Salinas 2012;Leguizamón Castillo 2015).Yet few cases of civilian victimisation in Colombia can rely on first-hand testimony of the perpetrators.Civilians face hostile armed actors and elite narratives that undermine and delegitimise their own experiences and protests.As I show in this paper, the fog of war makes it easy for elites to frame conflict as the dominating impetus for civilian victimisation.

Methods
This analysis revolves around two megadams (over $100 million each) built in Antioquia between 1995 and 2018.Both Hidroituango and Porce II were spearheaded by EPM and constructed in territories historically and currently affected by the Colombian civil war with the presence of multiple armed groups, including paramilitary groups, paramilitary successor groups such as the Clan del Golfo and Caparrapos, and the FARC, and National Liberation Army (ELN).Hidroituango was built on the site of some of the most violent paramilitary incursions of the 1990s, resulting in decades of fruitless investigations into links between then governor (later president) Alvaro Uribe and the counterinsurgent operations of the Bloque Minero of the Colombian United Self-Defence group (AUC).Porce II is less prolific but my own efforts in understanding its similar patterns of civilian displacement by paramilitary actors uncovered significant archival information on the suppression of union leadership during its construction.I do not take these as individual comparative cases given the overlap of timeframe and actors.Instead, these cases are representative of the different ways developmental elites take advantage of the fog of war to stigmatise and discredit civilian opposition to their interests.
This analysis draws on interviews collected with 15 community leaders, activists, and civilians, as well as government officials and functionaries from EPM. Interviews were carried out with informed oral consent of participants that all data should be kept anonymous and safely secured by the researcher as stipulated by the ethics approval granted by the University of Oxford.Citations only include the place and month of interview and a generalised description of their position to avoid identification.The case of Porce II builds on initial interviews with local leaders in the area, but the core of analysis examines eleven letters and attached material sent by the SINTRAPORCE union leaders to various offices in 1997.The letters were preserved by the Colombian government's Virtual Archive of Human Rights and Historical Memory.The analysis, carried out between 2020 and 2022, also drew on additional publications by the National School of Syndicalists and contemporary news articles.This archival data was essential due to difficulty in tracking down first-hand testimony on the assassinations and events leading up to them.This archival data also contributes to understanding the 'real-time' uncertainty in how actors perceived and responded to the existing data at the time (Chenoweth, Perkoski, andKang 2017, 1964).
The very uncertainty I assess in this paper poses a methodological challenge to researching violence.Conflict and civil war databases have issues in disaggregating potential conflict-related violence from other kinds, largely because of the difficulty in distinguishing between material and intellectual authors of such events (Eberwein and Chojnacki 2001).Armed groups with clear political ideologies can operate on multiple levels, either for political (i.e.counterinsurgency) or for economic (supressing challenges to developmental interests) objectives.Equally, the presence of uniformed combatants makes it relatively simple for others to pose as an established actor in the territory by using a uniform, badge, or announcing themselves as such.
Conflict environments therefore generate both uncertainty in data collection, and intentional efforts to mislead or obfuscate investigation (Chojnacki et al. 2012, 391).I therefore draw from Perrera's Bermuda triangulation, which is not focused on 'the verification of a single piece of information, but rather the uncovering of multiple narratives related to the initial information-each with their own additional details which lead to recursive processes of triangulation' which 'reveal important dynamics of conflict' (Perera 2017, 54).In this paper, I make no claim about any specific elite's complicity with the victimisation detailed below.Instead, I illustrate the empirical evidence surrounding instances of civilian victimisation, and the different narratives (elite and victims) employed to explain such events.I show that it is precisely the ambiguity that makes it easy for elites to frame every instance of victimisation as arising from territorial conflict disputes, despite credible justification for further investigation.This approach therefore incorporates the stories and silences -both in person or in archival gaps (Gonzalez-Ocantos and LaPorte 2019) -of counternarratives by victimised civilians to understand who they identify as responsible.Overlapping narratives of conflict and developmental violence are entangled, rather than discrete explanations for victimisation.

Counterinsurgency and Hidroituango
Thousands of civilians fled the municipality of Ituango and neighbouring municipalities in 1996 and 1997 when the Bloque Minero carried out some of the most notorious massacres in recent Colombian history in the veredas [sub-municipal units] of El Aro and the Granja.Between 1998 and 2008 a total of 62 additional massacres were recorded in the twelve municipalities in the area of Hidroituango's influence, leading to a mass exodus of the population (Agencia Prensa Rural 2019).In her analysis of war-time displacement, Steele (2017, 30) classifies the civilian flight from Ituango as displacement by paramilitary actors geared at counterinsurgency objectives, an assessment supported by the Interamerican Court for Human Rights in 2006 that ruled that displacement was part of 'a counterinsurgent plan to terrorise the civilian population of the area to remove whatever support -real or imaginedin favour of the guerrilla' (CIDH 2006, 2).While EPM denied any involvement in the displacement, they did justify the dam's construction as part of a wider counterinsurgent effort; according to a former high-level official, Hidroituango was intended to 'generate dynamics so that the project has a greater value [to the community] than the armed group [FARC]… [who] needed dynamics of exclusion to control' . 1 The arrival of the dam was therefore a balm meant to turn the page on the history of 'conflict and displacement in the region' (EPM 2014, 22) (Figure 1).However, it is not as simple as distinguishing this mass displacement as counterinsurgent rather than economic (Steele 2017, 16).One local inhabitant of El Aro recounted that a paramilitary combatant had told him their objective in the territory: 'they said that we were on land that would be used [for the dam].When we refused to leave, that's when they started shooting' (Parkin Daniels and Ebus 2018). 2 One activist recounts that this practice continued up to the dam's construction started in 2008; armed actors who identified themselves as paramilitary knocked on the doors of civilians who had refused to sell their land to EPM and told them to leave. 3 Accusations of links between elites and armed groups dogged the implementation of the dam, but never resulted in charges, must less convictions.In 2010 the official tender for the construction of the project, managed by EPM, was administered by the governor of Antioquia, Luis Alfredo Ramos.He awarded it to the firms of Camargo Correa and Conconcreto (CCC) Consortium.In 2011 the Colombian Supreme Court opened an investigation into alleged links Ramos had with paramilitary groups (El Espectador 2021), while in 2015, the Conconcreto company was also implicated in financing the AUC by former paramilitary commander Rodrigo Zapata (UNHCHR 2015).Official investigations in 2010 by the Prosecutor of the Peace and Justice Unit, Patricia Hernandez, observed that 'parallel to the [hydroelectric] project's evolution was the paramilitary advancement' , and in her judgement, the incursions were to pacify the territory for the hydroelectric project and the paramilitaries 'did the dirty work' .No legal action came of the investigation, Hernandez was moved to another office, and eventually the investigation appeared abandoned (Noticias Caracol 2011;Torres Ramírez 2018, 436).
Yet civilians continued to challenge the dam's implementation.By 2012, various communities displaced by the dam formed the Rios Vivos (Living River) movement and started organising protests and marches against the construction, citing issues of civilian displacement, environmental concerns, and the fact that the dam would flood an area thought to contain mass graves from the conflict (Lombana 2020).While EPM attempted to win community favour in promising to pave roads and build football pitches (EPM 2014), for many activists it was seen as patronising and insufficient, 'behaving like they were the head of the family who helps expecting the children and wife to be grateful' . 4 Therefore the protests continued, even as shadowy threats emerged in response.In March 2013, when Rios Vivos began organising marches in Ituango, armed fighters calling themselves the AUC targeted organisers of the movement (Sumapaz 2014, 66), leading to their displacement and eventually the assassination of multiple leaders, including Nelson Giraldo who was found dead in the Cauca river, shot multiple times and his throat slit (El Mundo 2013a).A short investigation was quickly abandoned, but the brutal manner of his death send a clear and intimidating message to Rios Vivos about returning to Ituango (Semana 2013).In 2013, teachers who protested the military taking a local school in Ituango as a base to protect the dam were sent threatening letters by individuals claiming to work with the CCC Consortium.The letter warned them they had 24 hours to leave Ituango because the teachers' 'social leadership' was interfering with their 'political efforts' against terrorism in the region (Rios Vivos 2013).The case was never followed up or investigated. 5 A new wave of repression emerged after problems with the dam caused the river to rise rapidly in April 2018, leading to a crisis downstream in Puerto Valdivia and Tarazá.The Association for Pan-miners of Northern Antioquia (ASOCBARENA) established a refugee camp in Ituango for those displaced by the rising water.Three of its leaders were killed between June and August 2018 (Verdad Abierta 2018); according to one leader, those killings were 'to finish with that camp.It was receiving too much attention and the company did not want that much attention' . 6Eventually those in the camp abandoned it out fear of continued reprisal. 7A member of Rios Vivos also explained: 'they would attack us because EPM would pay them to threaten us….They would say to us at Rios Vivos, we will give you only so much time, like 15 days to stop bothering about the project' . 8 The elites involved with Hidroituango had a different narrative to explain the attacks over the years.Military leaders claimed that the FARC had coerced the local community to protest, making them vulnerable to opposition groups (El Mundo 2013b).A former EPM official insisted that 'social confrontations [against the dam] were directed by the FARC' because the complaints were anti-development and critical of the state, though he acknowledged that probably not all protestors were guerrilla. 9A project manager in EPM also claimed that though Rios Vivos had caused a lot of noise, the project was not opposed by the real local community, calling into question the true identity of the organisation's members. 10When asked about the assassination of community leaders protesting the dam in Ituango, the Secretary of Government of Antioquia, Victoria Eugenia Ramirez said that 'the prosecutor must resolve if these were defenders of human rights or people implicated in other circumstances' (Hernández 2018), effectively questioning and unsubtly hinting that the victims had actually been involved with local conflict actors.By contrast, even civilians not involved with and critical of Rios Vivos dismissed the idea of guerrilla involvement due to the FARC's lukewarm opposition to the dam despite public statements against it. 11 State actors appeared to understand explicitly how conflict dynamics could frame civilian victimisation.A local leader with ASOCBARENA recounted a conversation overheard between two military officials in 2017 near one of the camps for miners displaced by the Hidroituango dam: A major told another major that they should 'fill us full of lead' and then say it was the guerrillas.But the other major said no, we cannot shoot them because everyone knows the guerrillas don't get involved with the miners. 12This exchange reveals not only that violence against civilians by armed actors may not be motivated by armed conflict or counterinsurgency but that its perpetrators are aware of the power of discursive framing of violence, as revealed by Mancuso.
The language of the threats is also informative.A miner displaced in 2018 from Tarazá became involved with ASOCBARENA protesting against EPM and began to receive threats from unknown sources, who accused him and fellow leaders of being 'informants for the guerrilla' . 13If this conflict language is meant to stigmatise victims, the victims have a different idea about who is ultimately responsible.One local activist bluntly described why civilians were afraid to challenge Hidroituango: 'No-one wants to be opposed to [the dam] because EPM has a lot of power and they will kill them' . 14 Stigmatisation of civilians reinforces the dominant narrative of violence in the region to broader audiences as a struggle between insurgents and counterinsurgents.Elites do not need to prove this: rather they rely on the incontrovertible presence of conflict actors in the area and the existing uncertainty shrouding specific acts of violence.To illustrate the point, one local civilian in Ituango accused Rios Vivos of being infiltrated by guerrillas and causing violence -in line with EPM's narrative -but then later admitted that the group appeared under threat for fighting for the rights of the local communities.He paused, and observed that 'one never knows if that [threat] is motivated by the subversion, or who is killing these leaders protesting Hidroituango.It's all a bit dark who is the actor behind these assassinations' . 15Overlapping violences obscure actual responsibility, while developmental elites maintain a simpler narrative of insurgency and counterinsurgency to account for civilian victimisation in the territory.

Insurgent revenge and Sintraporce
Mass displacements by counterinsurgent actors also occurred around the construction of Porce II (and later Porce III) dam by EPM and Conconcreto in northeastern Antioquia.The presence of the FARC's 36 th Front and the ELN Heroes and Martyrs of Anorí (Echandía Castilla 2013, 5) justified incursions by the paramilitary group Bloque Metro (Cívico 2009) and led to over 1045 forced displacements in the territories slated for dam construction in the Porce Canyon between 1990and 2006(CEV 2020, 88).Uncertainty pervaded the violence.According to local leaders, 'one was not always sure who was directly responsible for the displacement.Different armed groups would come, and they would confront each other.Always in the areas of influence of the dams' . 16Another civilian leader explained that 'suddenly, they were assassinating people and you did not know who, that was when people turned up who said they were working for EPM [to take a census], and that scared people because they were afraid that data would be used to kill them…as a result many people left' (Cárdenas Delgado 2011) (Figure 2).
Violence associated with the dam did not end with displacement.The assassination of five leaders of SINTRAPORCE (Porce II workers union) Leadership Committee between 1997 and 1998 provides a different case for how developmental elites benefit from extralegal violence and ambiguity produced by armed conflict.On the 25 th of January 1997 SINTRAPORCE raised a complaint on a series of violations of workers' rights by Conconcreto, including issues of payment, safe working conditions, and the presence of medical support onsite (SINTRAPORCE 1997a(SINTRAPORCE , 1997b)).The president of the union, Alberto Jaramillo, was assassinated on the 20 th of March 1997 by hooded armed men wearing ELN armbands.On the 7 th of October, the leadership of SINTRAPORCE received a pamphlet signed by the ELN explicitly threatening their lives and demanding that they leave the territory, which the leadership did, in Medellin by the 17 th of October (SINTRAPORCE 1997h).The letter accuses the leadership of complicity with local paramilitary organisations and warns them 'not to cry when we take you off the buses to apply the full weight of revolutionary justice' (DDHH 1997).
Yet it remains unclear if the ELN were behind the threats, or whether these accusations were legitimate.While the new president Hector Castrillon was kidnapped and threatened in March after Jaramillo's assassination by men dressed as ELN, he also later received threatening voice messages from paramilitary groups in the area (ENS 1998, 35).On the 24 th of October 1997, after the letter was sent, the ELN stopped trucks taking workers to the Porce II campsite in a show of strength.Several workers took the opportunity to ask the combatants about the letter and attacks.The combatants denied sending the letter or responsibility for the assassination of Jaramillo (Juan Diego 1997), adding additional uncertainty about why a leftist armed group would openly intimidate union leaders for corruption one moment and deny it the next.
Further testimony by other workers indicated that the head of security onsite, Jaramillo Panesso, had warned the new union leadership after Alberto Jaramillo's death that 'they should not make a lot of problems for the company because there is a lot of danger with so many guerrilla organisations and paramilitary in the area' (ENS 1998, 35).In 1998, members of the union testified that the 'paramilitaries were using the camp as well as the company vehicles for their own work and that the head of security was an active supporter of Convivir' (ENS 1998, 35-37).The SINTRAPORCE leadership at the time also suspected the company's involvement and reported receiving various anonymous threatening calls to their house that included specific details about family members and contact numbers that the leadership believed was only accessible in the company employee records (SINTRAPORCE 1997h).
From the assassination of Jaramillo until October of 1997 tensions mounted as the new union leadership continued to publish detailed lists of workplace violation and demands (SINTRAPORCE 1997c(SINTRAPORCE , 1997d, 1997e, 1997f, 1997g), 1997e, 1997f, 1997g).In October one of their leaders, Moises Caicedo, was forcibly removed from the worksite by the security guards (SINTRAPORCE 1997h).After the phone calls, harassment, and the pamphlet, the leadership fled to Medellin.There they met with officials from the Council of Medellin and agreed to travel back to the worksite on the 10 th of November to inspect working conditions alongside government functionaries; the Committee even wrote to General Ospina of the 4 th Brigade requesting security for Elkin Clavijo -the new president -and Alberto Niño, another union leader, to travel back to Anori on the 10 th of November (SINTRAPORCE 1997j).In a chilling bit of irony, General Ospina would later be accused of supporting the paramilitary massacres in Ituango two weeks prior (Restrepo 2015).
The plan was to fly to the worksite from Medellin by helicopter with government functionaries but a few days before Clavijo and Niño were informed that there was not enough space and that they alone would need to return by bus (ENS 2000, 36). 17En route, other passengers reported that hooded armed men wearing ELN armbands stopped the bus and asked for Clavijo and Niño by name.Their bodies -showing signs of torturereappeared by the road on the 13 th of November (ENS 2000, 37-38;SINTRAPORCE 1997j).Several weeks later, another member of the displaced leadership, Luis Puerto, was assassinated during his lunch in Medellin by unknown assailants (ENS 2000, 38;SINTRAPORCE 1997i).The remaining members of the Committee moved to Bogota in December to receive protection from the national office of Human Rights in the Ministry of Interior (Ministerio del Interior 1997).However, financial pressure forced Moises Caicedo to return to Medellin in June 1998 to find work, where he was assassinated a few days later (ENS 2000, 38).
The perpetrators remain unknown, yet the end result appears to be in EPM and Conconcreto's interests.In early 1998, after the murder of Clavijo and Niño, the Constitutional Court of Colombia ruled against SINTRAPORCE when members of the Committee tried to sue the Consortium, claiming their involvement in the threats made them responsible for paying for protection and compensation.The judge ruled that the companies had no obligation to pay since: the Consortium of Porce II is not responsible for the acts of violence directed against the workers of SINTRAPORCE…indeed, the plaintiffs did not provide any evidence that would lead to presume the veracity of the accusations levelled against the company, much less, of its alleged complicity with the armed groups that sow terror in that area of the Antioquia department.In short, the cause of the violation of the rights of the guardians does not come from the action of the company but from dark forces, presumably of guerrilla origin [emphasis own].(Tribunal Superior de Antioquia 1998) The new union president Jorge Bustamante following the persecution of the former Leadership Committee was seen by other union members as 'in [the corporation's] pocket' , appointed to further the Porce II Consortium's interests (ENS 2000, 37).Bustamante expressly distanced himself from his predecessors, arguing that 'workers know what is the problem in the region.As long as one doesn't get involved with anyone nothing will happen.No current worker on the project has been threatened or killed' .According to him, 'the problem is that some members of the Committee were not neutral and got involved with some of the actors of the conflict in an area where the guerrilla, paramilitary, narcos, and common criminals operate' (El Tiempo 1998).His explanation of the events reinforced the narrative presented in the Constitutional Court's ruling.
In contrast, other union leaders were more critical of the ruling and its implications of who was behind the attack: Jorge Ortega, the president of the Single Confederation of Workers of Colombia (CUT) with whom SINTRAPORCE was affiliated, was concerned that it allowed corporations to leave workers unprotected in conflict areas, and particularly undermined the safety and position of union leadership (El Tiempo 1998).According to further testimony from other union members of SINTRAPORCE: the Committee never believed that the pamphlet was sent by the ELN.They were more suspicious that the head of security was fishing in turbulent waters and he wanted to make the threat look like it came from the ELN.(ENS 1998, 35-37) Whoever was ultimately responsible, the analysis from the National School of Syndicalists concluded that 'in the middle of this confusion the company has become a biased judge with an explicit interest that the labour organisation find itself fragmented' (ENS 1998, 18).The 'turbulent waters' of conflict conveniently removed one set of union challengers while the lack of concrete evidence and judicial ruling reinforced the spectre of dangerous 'dark forces' and helped undermine future challenges to developmental elites in the area.

Conclusion
What is the value in taking such a granular view in tracing the repression of a handful of civilian activists and union leaders?I argue that these cases show how developmental elites use conflict narratives to delegitimise civilian challengers and dismiss their victimisation.While existing research has demonstrated the material links and shared incentives between elites and armed groups during war (Brosché 2014;Reina 2022;Ballvé 2013;Gutiérrez-Sanín 2019;Biberman 2018), here I argue that war also allows elites to invert 'malicious denunciations' and frame victims in the language of territorial conflict to broader audiences.This narrative is countered by the experiences and perceptions of victimised civilians, but the fog of war and the obvious presence of conflict actors makes claims of developmental violence more difficult to support.
When armed actors force civilians off land, or accuse activists or union leaders of supporting opposing armed groups, elites turn to easy narratives of armed conflict and territorial competition even while the victims themselves offered counter-narratives that implicate developmental elites.Paramilitaries such as the AUC have combined counter-insurgent efforts with broader targeting of activists and union leaders (Gomez-Suarez 2007; Chambers 2013) even without specific involvement by developmental elites.Yet elites advance a narrative that associates the victims with insurgency and conflict rather than their activism and opposition to the dam.Guerrilla groups may also have motivation for targeting civilian activists and union leaders: corruption, association with other armed groups, or theoretically even to silence activists that threaten a source of extortive income in the form of the dam.However, whether or not there are cases of guerrilla groups targeting civilians, the focus of this paper is precisely on how ambiguity around motive permits elites to advance a narrative that denies or dismisses any connection between the violence and development interests.
Violence by conflict actors combined with elite denial produces what Beban, in her analysis of violent landgrabs and state-making in Cambodia, refers to as the cyclical production of uncertainty (2021,215).Civilian victimisation when there are political armed groups already present in a territory warrant less investigation by authorities who frame their existence as explanation enough.Only a handful of actual attacks and fatalities may occur that advance a developmental rather than conflict agenda, lost in the wider miasma of war-time violence and uncertainty, but these attacks send a clear message to civilian challengers of elite interests.The very nature of such dynamics shroud actions in secrecy so that few are unable to concretely identify the authorship of the abuses (Krahmann 2016(Krahmann , 1412)), or unwilling to verbalise if they do know (Global Witness 2012, 8).The civilians who continue to press counternarratives do so bravely -and challenge elite's dominating narrative -but there is little to no actual consequences or even investigation for the elites accused.
Peace processes can reveal the explanatory weaknesses of conflict narratives.Optimism about peace dividends began to fade not long after the historic 2016 peace accord with the FARC.While overall rates of violence dropped, the assassination of social leaders remained alarmingly high.The administration of President Ivan Duque attributed these killings to territorial competition between various remaining armed groups, including dissident FARC (Manetto 2020;Banchon 2020).Yet these killing are difficult to fit into old conflict narratives, even as government officials use the old playbook to lump the killings under 'criminal disputes' , reporting the armed group associated with the attack rather than investigating their motives (COEuropa 2021; Manetto 2021).Grajales (2020) argues that peace in Colombia presents a challenge to elites to adapt old frontier narratives that legitimised legally-dubious land acquisition in times of war as bold investments in dangerous and marginal territories.As one Colombian analyst observed, 'there are no more easy narratives of guerrilla and paramilitaries [to explain assassinations].Now you wonder who the social leaders are, who are the labour unionists, when are they linked to the FARC and when is this just an excuse?' 18 The transition marks not merely a shift in violence, but also the dissolution of convenient narratives.These findings therefore add to the urgent need for peacebuilders to address not merely the political motives for national and local actors in armed conflict, but the economic interests that intersect with violence against civilians (Meger and Sachseder 2020).
The cases examined in this paper and the continued assassination of social leaders challenges us to consider how political armed conflict overlays one logic of violence over others, generating uncertainty about motivation and challenges to peace.The civilians displaced by the Hidroituango and Porce II dam call themselves victims of development rather than conflict (Lombana 2020, 7), 19 inverting Steele's assessment.Yet I argue the civilians displaced, threatened, or killed in these cases are victims of both.Both dams were unquestionably built in war-affected contexts with the presence of revolutionary armed groups and politically-motivated counterinsurgents.But to collapse all violence in these territories under the convenient aegis of conflict ignores how developmental elites may take advantage of both conflict's extra-legal violence and its narrative to undermine and discredit civilian opposition to their interests.Though the study of these overlapping patterns generates incongruities and precludes one clear explanation for civilian repression, it is these very incongruities that scholars and policy-makers alike should take seriously in attempting to disentangle the violences of conflict and development.
11.Many of the locals interviewed also claimed the FARC was more interested in extorting the dam constructors for money than blocking the project (Interview with local inhabitants of Ituango (2,6,8,9), Ituango -November 2018.For more on this co-option see Voyvodic (2022).12. Interview with local leader (5)
, Santa Rosas -December 2018.13.Interview with panminer (3), Medellin -December 2018.14.Interview with local activist (4), Puerto Valdivia -April 2019.15.Interview local inhabitant of Ituango (10), Ituango -November 2018.16.Interview with local leader from Porce Canyon (11), Amalfi -November 2018.Reiterated in interviews with two other local leaders from Porce Canyon (12, 13), Amalfi -November 2018 and a fisherman in who had lived in Porce Canyon (14), Tarazá -December 2018.17.The situation is eerily similar to that of Jesus Maria Valle, the famous human rights defender who brought national attention to the massacres of Ituango in 1997 and accused the Colombian government of complicity.Valle was offered a similar helicopter ride along with a group of officials to Ituango by Uribe (who later accused him publicly as an 'enemy of the Armed Forces') to show where the graves were and prove his claims.Last minute Valle was informed that the helicopter ride was full and that he would have to drive alone.Valle decided against it.He was killed a year later in his office in Medellin (Sánchez-Moreno 2018, 23).18. Interview with national analyst (15), Bogota -September 2018.19.Interview with local activist (8) Ituango -November 2018.