Introduction
In 2013, residents of a neighbourhood controlled by one of the largest armed groups in Barranquilla (Colombia) found a human head. A few streets away, in a neighbourhood controlled by another gang, the trunk of the body was found. The arms and legs, with identifying tattoos, were found in front of the victim's house.Footnote 1 Similar scenes of dismembered bodies scattered in different neighbourhoods of the city have been repeated on at least 20 occasions since 2013 in Barranquilla.
Dismemberments in Colombian cities are usually committed by large groups of people, involving between three and eight individuals and other resources. This is because dismemberments involve the kidnapping of the victim, vehicles to transport them to the place where they are dismembered, a place to commit the crime, people who then carry out the act, and the distribution of the body parts around the city. Kidnapping alone is a high-effort, multi-person activity.Footnote 2
Selective homicides, on the other hand, usually involve one or two people who are relatively available to hire in Colombian cities. Why, then, do organised criminal groups (OCGs)Footnote 3 choose to dismember people when this is a more complex and costly practice than targeted killings? Researchers typically explain dismemberments in three ways: as acts of pure cruelty, as consequences of criminal group rivalries, or as attempts to conceal evidence and hinder investigations. However, these explanations cannot account for cases where OCGs publicly display dismembered bodies in territories under their undisputed control.
In this article, we address this puzzle by connecting the literatures on criminal governance and extra-lethal violence, understood as acts of violence that violate shared norms regarding the treatment of individuals and their bodies.Footnote 4 We argue that dismemberment, one form of extra-lethal violence, is used by OCGs as a way of establishing or sustaining criminal governance regimes. We argue that dismemberment serves as a tool of violent communication targeting three distinct audiences (rivals, group members and civilians), with the ultimate goal of generating compliance through fear.
First, dismemberment demonstrates an armed group's capacity and willingness to use extreme violence against rivals, deterring competitors and asserting dominance. Second, when used as exemplary punishment against civilians who violate the group's rules, dismemberment becomes an instrument for regulating behaviour and maintaining social control over the governed population. Third, it reinforces internal discipline by punishing group members who break internal codes or defect to rivals. In sum, we argue that the fear produced by dismemberment helps delineate acceptable behaviour and reinforce the criminal group's system of norms and punishments.
We show the logic of this argument by drawing upon an original qualitative database on dismemberments in Barranquilla and over five years of fieldwork. We triangulated this information with a review of more than 800 local press stories and interviews with members and former members of criminal organisations, civilians and local authorities.
This article makes two key contributions to the study of organised crime in Latin American cities. First, it challenges the prevailing assumption that brutal violence is merely a byproduct of inter-group rivalries. While some dismemberments do occur during conflicts between armed groups, our findings show that at least half of these acts were committed either during periods of criminal hegemony or against civilians unaffiliated with rival gangs. This suggests that brutal violence serves purposes beyond inter-criminal competition.
Second, the article also contributes to show that gruesome violence can be strategically deployed to cement criminal governance. Rather than destabilising armed groups or attracting institutional attention, dismemberments were effective in instilling fear among civilians and demarcating the boundaries of acceptable behaviour. In this way, brutality was not only not the result of instability, but was functional in generating stability.
In the remainder of this article, we first advance some hypotheses elaborated in the literature on civil wars and organised crime to explain dismemberments. Next, we connect the literatures on criminal governance and extra-lethal violence to build our theoretical argument. We then show how we created our dataset and used interviews and press reviews to supplement it. In the empirical section, we first introduce the context of criminal violence in Barranquilla and we then point out three empirical implications of our argument that are supported by the database and the interviews. Finally, we explain how alternative explanations are less effective in explaining our case and conclude by suggesting future avenues of research.
Why do Dismemberments Happen?
Dismemberment is a form of overkill against bodies. It involves the mutilation of the victim's body parts (usually the head or limbs). Sometimes dismemberment is, in itself, the way of executing a person, while in other cases this procedure is done after the homicide.Footnote 5 Although dismemberments can be committed by single individuals, they have also been used in the context of civil wars and by OCGs.Footnote 6
In civil wars, this phenomenon is puzzling. Jeremy Weinstein, for instance, affirms that some acts of ‘brutality … are above and beyond what is required to send a signal of the costs of defection’.Footnote 7 Nevertheless, there is evidence to suggest that some deeds are done more viciously during civil wars than during peacetime.Footnote 8
To account for this puzzle, some scholars have argued that armed groups use terrorist tactics to generate a general sense of insecurity.Footnote 9 Others argue that the events of chaotic violence in civil wars desensitise combatants, making them more prone to extreme violence.Footnote 10 These explanations are divided between those that attribute brutality to instrumental reasons and those that attribute it to factors other than rationality.Footnote 11
Nevertheless, the applicability of these explanations to criminal violence is not straightforward due to differences in the logics of violence between civil wars and criminal violence.Footnote 12 OCGs seek to hide their violence to evade law enforcement;Footnote 13 dismemberments, however, are likely to elicit strong institutional reactions and media attention.
Paradoxically, in some cities OCGs have publicised their dismemberments and openly taken responsibility for them.Footnote 14 Furthermore, dismemberment is a resource-intensive practice compared to targeted killing.Footnote 15 Given these factors, why would OCGs opt to use dismemberment in certain situations?
Three main arguments have been developed to answer this question. First, some authors argue that, given that OCGs would want to remain below the authorities’ radar, dismemberments are a strategy to conceal the victim's identity and/or delay or derail the authorities’ investigations.Footnote 16 Second, some authors argue that extreme violence is used by OCGs as a way of exerting cruelty on the victim's body. This cruelty can be functional to building their organisational identity and creating a sense of unity.Footnote 17 Cruelty can also be the result of spontaneous situations outside the control of criminal organisations.Footnote 18 Finally, extreme violence, including dismemberment, is functional for OCGs to claim territorial control against rivals.Footnote 19 By employing extreme violence against rival groups, OCGs send a message to rivals, the state and the community about their dominance and authority over certain markets or territories.Footnote 20
In the final part of the paper, we grapple with all these rival explanations empirically and theoretically. We contend that, while these are useful arguments to account for some cases of dismemberments, they are all partial explanations that leave cases underexplained. Our argument, by contrast, combines these perspectives and offers a rationale that can be used beyond idiosyncratic cases.
Extra-lethal Violence and Fear in Criminal Governance
Benjamin Lessing defines criminal governance as the ‘imposition of rules or restriction on behavior by a criminal organization. This includes governance over members, non-member criminal actors, and non-criminal civilians.’Footnote 21 Even though armed groups are able to impose these rules, they may want to reduce some forms of visible violence in their governed zones so as not to trigger the intervention of the state.Footnote 22
Such governance is exercised by large drug cartels, street gangs and prison gangs alike.Footnote 23 Given this variety of OCGs, the way they govern is also multifaceted.Footnote 24 A less explored aspect of criminal governance has to do with the mechanisms used to establish and sustain it over time.Footnote 25 Such governance can sometimes be accepted and recognised as legitimate by civilians.Footnote 26 To achieve this, armed groups can use various strategies, such as engaging community actors or reducing their violence so that they are recognised as ‘less evil’ than others.Footnote 27
However, these approaches do little to help understand the role of brutal violence in criminal governance. Indeed, as Svetlana Stephenson argues,Footnote 28 the violence of criminal groups can produce a state of fear and uncertainty that perpetuates their more predatory character. How is it, then, that brutal violence can contribute to sustaining criminal rule rather than destabilising it? Lee Ann Fujii's concept of extra-lethal violence helps to connect the strategic motivations of criminal violence with the fear instilled by dismemberments.Footnote 29
Organised criminal violence is typically instrumentally motivated.Footnote 30 In cases of dismemberments committed by OCGs, this logic also seems to hold true. Several OCGs have publicly displayed their victims or inscribed messages on their bodies.Footnote 31 In other words, not only was there an aim to punish the victims, but there was also a communicative logic in the act.Footnote 32
Fujii defines extra-lethal violence as the ‘face-to-face acts of violence that transgress shared norms about the proper treatment of persons and bodies’.Footnote 33 Extra-lethal violence is not randomised violence; instead, it is directed against specific people and displayed to a particular audience.Footnote 34 This type of violence can be committed for instrumental or expressive purposes, as well as being premeditated or spontaneous. A single episode of extra-lethal violence may feature some or all of these dimensions.
The central point of extra-lethal violence is to put violence on display, thereby constituting a particular social order: those who commit these acts are asserting their identity in front of an audience through the fear it produces.Footnote 35 When extra-lethal violence is exercised, those who exercise it are asserting their authority in the most graphic terms possible. These acts, then, are not only a form of punishment against the person killed, but seek to ‘instill fear beyond the victims of the attack’.Footnote 36 Simply put: the purpose of extra-lethal violence is not only homicide, but the instillation of fear in an audience to ensure compliance within the social order that perpetrators seek to establish. Although they are naturally ambiguous spaces, armed groups manage to send brutal signals that demarcate what is allowed in their zone of governance.Footnote 37
Our argument is that the dismemberment of bodies is functional for the OCGs to establish and sustain the social order they desire in their criminal governance regime. This form of extra-lethal violence serves three purposes. First, it contributes to the establishment and protection of criminal governance by eliminating rivals. As the literature cited above has already shown, brutal violence functions as an exercise in deterrence. The organisation in charge of dismembering its rivals is demonstrating that it has the capacity to use force, and that other members of the rival organisation may suffer a similar fate. Thus, by exerting extra-lethal violence on rivals during turf wars, armed groups are incorporating a criminal ethos, consisting in communicating to their rivals that they are prepared to use overkill tactics, maximising the fear that they arouse in them.Footnote 38 Some OCGs in Mexico, for instance, commonly record videos of the dismemberment or disperse the body parts of dismembered rivals.Footnote 39 El Tren de Aragua, a Venezuelan criminal group, has been known to use similar tactics in Colombia, Chile, Perú, Venezuela and Brazil.Footnote 40
Second, dismemberments are functional to punish non-compliance with the codes of behaviour delineated by armed groups. In other words, when criminal governance is not challenged, dismemberments are not primarily directed at rivals, but at the governed community (civilians). This shows that criminal governance regimes are not only protected through violence against competitors, but also through disciplining the community itself.Footnote 41 There are multiple examples of dismembered bodies which have appeared in public with signs accusing the victims of snitching (i.e. of being ‘sapos’) in Honduras or Ecuador,Footnote 42 just to mention a few. In Mexico, two dismembered bodies appeared with a message: ‘This is what happens when you kidnap undocumented migrants’,Footnote 43 demarcating there what is not allowed.
Finally, dismemberments are also useful as an internal governance mechanism. Because they exist illegally, armed groups face multiple organisational challenges. Lack of internal discipline can lead to problems ranging from fractionalisation to individual desertion of members to join rivals.Footnote 44 Dismemberments then function as a means of disciplining members who have betrayed the armed group or broken rules of behaviour and deter them from exiting the organisation. In Chihuahua, Mexico, for instance, one armed group dismembered two people and left a message saying: ‘traitors’.Footnote 45
The public display of extreme violence, such as distributing human remains across neighbourhoods, aims to instil fear in the community. However, this fear generates compliance rather than panic or paralysis primarily due to the selective nature of such violence. Unlike terrorist actions where any member of the community might be targeted,Footnote 46 this form of violence specifically punishes those who transgress established norms.
This effectiveness stems from what Stathis Kalyvas termed the ‘perception of credible selection’.Footnote 47 Even when mistakes occur in targeting, armed groups carefully maintain an image of selective punishment. A police officer's statement illustrates this dynamic: ‘Good people can rest easy because everyone they kill is because they got into trouble with them. Those who owe nothing fear nothing.’Footnote 48 Similarly, a civilian who saw multiple dismembered bodies noted: ‘Of course it is scary because they are very awful images, but at least it is not that they are going to dismember you for no reason; at least we know what we have to do to avoid trouble.’Footnote 49 In sum, dismemberment, being both sporadic and perceived as selective, produces compliance rather than widespread panic or paralysis in the population. We will now demonstrate the relevance of this argument using evidence from one Colombian city.
Research Design
Case Selection and Empirical Evidence
We use a database provided by the Policía Nacional de Colombia (hereafter ‘National Police’) for an identifiable universe of dismemberments in Colombia between 2006 and 2023. Of these, we focus only on municipalities with an exclusive presence of OCGs, i.e. without rebel groups. We have considered only municipalities with more than 15 cases of dismemberments, since otherwise it would be hard to discern trends. Thus, we reduced the universe of municipalities to five cities: Barranquilla, Medellín, Bogotá, Cali and Buenaventura.
Among these municipalities, we chose a convenience sample. Barranquilla is the only city in which dismemberments virtually disappeared until 2013. Although there were several turf wars between 2006 and 2013, there was no dismemberment in Barranquilla during these years. This temporary disappearance of dismemberments after the end of paramilitarism allows us to identify a clear point of origin in the analysis. Given that our purpose is to understand (i) when dismemberments appear, and (ii) how dismemberments happen after criminal governance is established, Barranquilla offers an ideal scenario to understand the phenomenon over the last ten years. In the Supplementary Appendix we expand on other cases we considered.
Our data is essentially composed of two sources: an original qualitative database and interviews conducted during our fieldwork.
Qualitative dataset: To systematise the information gathered from diverse sources, we established a qualitative database where each observation represents an individual case of dismemberment. The total count of dismemberment cases in the city varies depending on the source we consult. We employed a triangulation approach, combining data from various government entities (the National Police, the Fiscalía General de la Nación [Attorney General's Office], the Defensoría del Pueblo [Ombudsman's Office], the Instituto Nacional de Medicina Legal y Ciencias Forenses [National Forensic Medicine Institute]) and a private database generously provided by the late William Colina Páez, a journalist from the city's largest local newspaper who covered criminal violence for over 15 years.
Our analysis covers the period from 2006 to 2023. The total number of dismemberment cases fluctuates significantly across these databases over this period, ranging from a minimum of 17 to a maximum of 27 cases. This variation is attributed to differences in how institutions define dismemberment.Footnote 50
With these insights, we constructed an original dataset, entitled ‘Dismemberments_Barranquilla’ and provided as part of the supplementary materials for this article. Each observation in this database corresponds to a dismemberment case reasonably linked to an OCG. In order to consider it as such, we triangulated the versions offered by the press, the National Police, the Ombudsman's Office and interviewees. We excluded cases where evidence was insufficient to make a judgement. Additionally, we compiled a collection of news stories from local and national press, totalling around 800 stories, ranging from eight to 40 stories per dismemberment depending on coverage levels. This extensive collection allowed us to confidently identify cases where organised crime was unlikely (at least two dismemberments) and select 25 cases where, through factors like family statements, newspaper research, or details emerging years later – such as the imprisonment of the perpetrators – we could reasonably determine OCG involvement. These 25 cases constitute the observations in our database.
The database comprises ten columns: year of dismemberment, neighbourhood where body parts were distributed, victim profile, specific reason for dismemberment, function of the dismemberment, whether the dismemberment was made public or not, specific details about body part distribution, specific details about the body parts, presumed armed group responsible and the age of the victim. In the Supplementary Appendix we explain the coding decisions for each of those variables.
Interviews: Our second main source of information was 34 interviews conducted over more than five years of fieldwork in Barranquilla and nearby municipalities between 2018 and 2023. The interviewees were active members of the National Police (seven), active or former members of OCGs in Barranquilla (14), civilians living in areas where dismembered bodies were identified or relatives of victims (seven), and officials in charge of security in Barranquilla, the department of Atlántico or other public institutions (six). Details of ethical dilemmas, of how coding decisions for the interviews were arrived at, of the questions we posed to our interviewees and context for the extracts included in the article are available in the Supplementary Appendix.
We posit three empirical implications of our argument (viz. that the dismemberment of bodies functions for OCGs to establish and sustain social order in their criminal governance regime): three observations that, if our argument were true, we should be able to identify.
Observable Implication 1: Victims of dismemberment should have various profiles: following Lessing,Footnote 51 we attempted to identify four different profiles that OCGs try to govern: members of their own OCG, members of other criminal organisations, non-criminal civilians, and criminal civilians. For reasons given in the Supplementary Appendix, we ended up re-classifying the last two profiles into the single category of ‘civilians’. If our argument were true, we would expect that the victims of dismemberment are not just rivals of the armed group, but also members of the whole governed community that Lessing identifies in criminal governance regimes.
Observable Implication 2: Dismemberment should be used as part of a communicative logic: in addition to committing the dismemberments, we expect that the OCGs are going to take public credit for them. Given that our argument states that the purpose of dismemberments is not just to punish the victim but to instil fear in an audience, it would make little sense to assume that OCGs will not openly disclose to that audience what they have done and why.
Observable Implication 3: Profiles of dismemberment victims change according to the context of violence: given that dismemberments can be used either for establishing criminal governance or for sustaining it, we expect the profiles of the victims to vary according to the context. If our argument were true, we expect that in situations of competition among OCGs victims are likely to be members of rival organisations, while, in established criminal governance regimes, the victims are more likely to be members of the governed community who have breached the codes of behaviour established by the OCG.
We chose observable implications not only as a way of measuring the explanatory power of our hypothesis, but also to help us adjudicate between alternative explanations, as recommended by Tasha Fairfield and Andrew Charman.Footnote 52 Table 1 explains how the data we collected were useful in supporting our explanation and weakening (or not) rival arguments.
Table 1. How the Observable Implications Relate to our Hypothesis and to Rival Explanations
Why Focus on Dismemberments?
Extra-lethal violence by organised crime takes various forms beyond dismemberment. In Mexico, for example, cartels have adopted practices like hanging bodies publicly (known as ‘colgados’) or leaving bodies with signs on public roads.Footnote 53 In Barranquilla, it is becoming more common to throw bodies into the river, where they later appear floating on the surface.Footnote 54 According to Fujii's definition, these acts are also examples of extra-lethal violence. So why focus on dismemberment?
First, unlike some forms of extra-lethal violence that appear idiosyncratic, dismemberments are widespread across Colombia and other Latin American countries. Public displays of dismembered bodies have occurred in cities like Bogotá (Colombia), Asunción (Paraguay), Santiago (Chile), and Chilpancingo (Mexico), to name a few. This suggests that studying dismemberment can provide insights not only for scholars interested in Barranquilla or Colombia but also for those of Latin American violence more broadly. Second, dismemberments share key characteristics with other forms of extra-lethal violence. For example, the public display of hanged bodies also involves more people and resources than a targeted killing, emphasising the performative aspect of the violence. Thus, insights gained from studying dismemberment can be extrapolated to understand other forms of extra-lethal violence. Finally, dismemberments may be relatively easier to observe compared to other forms of extra-lethal violence. While dismemberments that occur as a form of enforced disappearance pose challenges (which we address in the following section), public dismemberments provide a more accessible and detailed universe of data for analysis. This makes them a valuable subject for studying the broader phenomenon of extra-lethal violence.
Scope Conditions and Limitations
As a first scope condition, we believe our argument can be useful for understanding other forms of extra-lethal violence that are intentionally publicised. Central to the puzzle we identify is that armed groups not only commit these forms of extra-lethal violence, but that they make them explicit to an audience. We believe that, even if they are not dismemberments, any form of extra-lethal violence by organised crime that is publicised in this way can be anchored in similar explanations. Secondly, we believe that our argument is useful for explaining extra-lethal violence in territories where criminal governance regimes exist (or are being established). Consequently, it is to be expected that if an armed group has no interest in establishing a criminal governance regime, it will not use extra-lethal violence in the way we are describing here.
The data we collected for this study comes with two core limitations. First, dismemberment is a rare event. In ten years, there have been only 25 dismemberments in a city of over one million inhabitants, indicating that dismemberment is not a common form of murder. Despite this limitation, three reasons suggest that this study remains valuable. (i) While we cannot directly assess the effectiveness of dismemberment as a strategy on target populations, the fact that armed groups have been consistently committing dismemberments for more than a decade indicates its perceived functionality. If dismemberment were entirely ineffective, it is likely that armed groups would have abandoned the practice in favour of other strategies. However, its sustained use suggests it serves a specific purpose within these groups’ operations. (ii) Interviews consistently revealed that both civilians and members of armed groups associate dismemberments with significant fear and awareness of the reasons behind them. This suggests that dismemberment has an impact on how audiences internalise norms or, at the very least, instils a deep fear of punishment. The psychological and social effects of dismemberment may, therefore, be more profound than those of other forms of violence. (iii) It may be precisely because dismemberments have such a significant and visible impact that they are not more widespread. One dismemberment can generate more horror, discipline and lasting effect than several targeted killings. Additionally, the strategic rarity of dismemberments could result in preventing the authorities from responding to a wave of such incidents, making the practice more sustainable for criminal groups. Further research is necessary to explore this hypothesis.
The second core limitation relates to potential survivor bias. Almost all observed dismemberments were public, raising the possibility that other dismemberments occurred out of public view. This lack of data makes it difficult to define the total universe of dismemberment cases and draw comprehensive conclusions. However, two reasons suggest this limitation does not invalidate our findings. (i) Cases without records of dismemberment could be considered forced disappearances, with impacts and motivations differing from those we aim to explain. Our study, therefore, focuses solely on public dismemberments, while acknowledging that another logic may underlie dismemberments that armed groups attempt to conceal. (ii) Known cases of forced disappearance in Barranquilla over the last ten years have been relatively low – 21 cases between 2013 and 2023 in the entire department of Atlántico.Footnote 55 Even assuming all these cases were dismemberments, then at least 50 per cent of the total victims of dismemberment would still be public. However, uncertainty remains regarding unreported cases, so this limitation must be acknowledged.
Dismemberments in Barranquilla: Brief Context
Barranquilla is the most populous city in northern Colombia and the fourth nationally. With its high level of economic development, it has been described as ‘the capital of the Colombian Caribbean’. In 2006, the Frente José Pablo Díaz of the Autodefensas Unidas de Colombia (United Self-Defence Forces of Colombia, AUC) demobilised. This ended a cycle of violence in the city resulting from Colombia's armed conflict. However, this did not mean the end of criminal violence. During the second half of the 2000s, the Ombudsman's Office condemned the presence of several OCGs that took advantage of the vacuum left by the paramilitary group to implement new systems of extortion and drug distribution in the outskirts of the city.Footnote 56 Until 2013, the city had been in the midst of several disputes between criminal groups.
However, towards the end of that year, following a dispute between three armed groups (‘Los Rastrojos-Caleños’, ‘Los Costeños’ and the ‘Autodefensas Gaitanistas de Colombia’ [AGC]), ‘Los Costeños’ gained criminal hegemony in the city. This hegemony remained virtually uninterrupted, with the exception of 2016–17, when the AGC returned to the city financing a group called ‘Los Papalópez’, and 2022, when a different local armed group also challenged their hegemony.Footnote 57
Observable Implication 1: Profiles of Dismemberments
The first observable implication of our argument is that victims of dismemberment have different profiles. Specifically, we believe that if our argument is correct, there should be at least three victim profiles of dismemberment:
1. Rivals: The literature on organised criminal violence has stated that the use of gruesome homicides is associated with the punishment of members of rival groups or a sign of establishment in a territory.Footnote 58 Given that there have been some moments of competition for territory in Barranquilla, we assume that among the victim profiles there will be members of rival organisations. Their dismemberment would function to deter other members of the organisation from continuing to challenge the hegemony of one of the armed groups or to demonstrate the territorial control that the group responsible for the dismemberment exercises.
2. Members of the organisation: The literature on criminal governance has shown that OCGs, in addition to establishing norms of behaviour for communities, have established codes of behaviour for their own members.Footnote 59 Failure to comply with these codes often results in exemplary punishments that can be brutal in order to deter other members from following these behaviours.Footnote 60 Our argument assumes that, in an environment where dismemberment is used by criminal groups, their own members who violate the organisation's codes of behaviour may also be victims of this form of violence.
3. Civilians: In this category we include all persons who, without being members of armed groups, are punished for breaking rules established by the armed groups in territories where they exercise criminal governance.Footnote 61 When armed groups exercise territorial control, they not only establish codes of behaviour for their members, but many have also established a system of rules for people outside their structure.Footnote 62 Some armed groups have established punishments for those who engage in theft within the community, sexual violence, drug use, or so-called ‘social cleansing’ campaigns in the name of stopping waves of violence.Footnote 63 Consequently, if we argue that armed groups in Barranquilla have used dismemberments to sustain their criminal governance, then we should have evidence that some of these killings were committed against civilians for failing to comply with the norms established by the OCGs. This would coincide with recent literature identifying that armed groups exercise violence against civilians even when they control the territory.Footnote 64
For the purposes of this article, we have grouped these categories according to the function that the dismemberment fulfilled for the armed group. Thus, there were dismemberments that fulfilled an external function (deterring rivals), an internal function (disciplining members), and a function of enforcing the established social order (through violence against civilians). Table 2 shows the dismemberments classified by the function they served.
Table 2. Reasons for Dismemberments in Barranquilla (2013–23)
Observable Implication 2: Communicative Logic of Dismemberments
External Function of Dismemberments
We classified six of the 25 dismemberments identified as fulfilling an ‘external’ function. The victims’ body parts were spread across different neighbourhoods, including La Luz, La Chinita, Rebolo, Siape and Las Flores. These were all neighbourhoods being disputed for territorial control by gangs and close to the river, which allowed access for the entry and exit of drugs into and out of the city.Footnote 65 This information was confirmed by the National Police, who stated that these cases of dismemberment were ‘related to neighbourhood-level distribution and trafficking of drugs’.Footnote 66 In one case of dismemberment, the victim was filmed and the video was later circulated through WhatsApp. In the video, the person committing the crime was shouting, ‘This is happening to you for being a traitor. Here in the neighbourhood, we rule, but you decided to ally yourself with the competition.’Footnote 67
All dismemberments that we classified as fulfilling an external function were made ‘public’. That is, in all cases the bodies of the victims were displayed in the city. There were three main ways of doing this. The first consisted of leaving the body parts in full view of the entire community on public roads or in areas heavily frequented by the inhabitants. Four cases were made public this way. The second consisted in piling up the body parts in bags, small coolers or sacks and leaving them on public roads. One case was displayed in this way. Finally, in another case, the dismemberment was broadcast live on social media.
The way in which the body was dispersed also suggests that the purpose was to convey a message. In four of the six cases of dismemberment, the limbs were left in one neighbourhood, the torso in another, and the head (which usually appeared days after the first remains) in the neighbourhood where the victim lived. The neighbourhoods chosen for the display of the dismembered parts were the same areas where the armed group had a continuous presence. In some cases, messages were written explaining the reasons for the killing, in addition to leaving the dismembered body in public view.
The use of dismemberments as a communicative act against rival groups was confirmed by several members of the organisation. One of them stated that ‘It is not just about punishing the person, but you have to show why. You have to be careful to think about where and how the bodies are left so that … when [the body] is discovered the people will know who this happened to and why.’Footnote 68 Another interviewee stated: ‘The head could appear close to where he could be recognised … but the other parts close to where he committed what he was accused of.’Footnote 69 A police officer also explained that symbolism could be employed, such as taping over the mouth when the dismembered person was accused of leaking information; or messages could be left close to the body.Footnote 70
Internal Function of Dismemberments
We classified seven of the dismemberment cases in Barranquilla as fulfilling the function of ‘internal discipline’. Of the seven people in this group, the reason given by the armed group for dismemberment was the theft of money from the organisation (two cases) and betrayal (five cases). Money theft usually involved the victim being in charge of illegal activities, such as extortion or drug trafficking, and not reporting all the money collected to their superiors.Footnote 71 Betrayal usually consisted of allying with rival armed groups, either to pass on information about the armed group or to change sides in the midst of a competition for territory.Footnote 72 In some cases, victims were also accused of using violence without authorisation (e.g. murder without permission).
In times of competition with other armed groups, both ‘Los Costeños’ and ‘Los Papalópez’ dismembered members of their own organisation, accusing them of being affiliated with rival OCGs. One of the members of the organisation said that ‘We were all told when the other lot started the war that whoever changed sides was going to be punished. Then it was really tough because the others could offer you more money or even threaten you, but you knew this lot could dismember you.’Footnote 73 Another reason for dismembering a member of the organisation was because they leaked information to the authorities. A member of the organisation who allegedly handed over evidence to incriminate one of the leaders of ‘Los Costeños’ was dismembered, and the video of the event was distributed among members of the armed group.Footnote 74
In contrast to the external function, five of the seven dismemberments that served an internal function were not made public. In these cases, the remains were buried or hidden rather than distributed throughout the city. Nevertheless, these acts still conveyed a message: victims’ families were promptly informed of the dismemberment, and the community learned the reasons for the killings.Footnote 75 As one threatened member of the armed group recounted:
A friend of mine said to me: ‘We were ordered to picarte (dismember you), but since we like you, we are advising you to leave. You have 24 hours.’ And right then they told me why. [It was because] they realised that I was playing for two teams.
Interviewer: So, you were working also for [a different armed group]?
Yes. The mistakes one makes! They arrived in the municipio and offered me more money, so I started selling their stuff too. But in this world [of armed gangs/OGCs] you know that you can only do this till they catch you. I knew that eventually I would have to flee.Footnote 76
The decision to dismember without public display seems puzzling, since the OCG could have carried out a targeted killing instead. However, we argue that dismemberment serves as an established rule that must be enforced to maintain the credibility of punishment for defection among rank-and-file members. A former member of a criminal group explains:
Sometimes in this business people do things they do not want to do. I believe that those people [those dismembered whose remains were hidden] were usually liked by the boss or had been in the group for years. So, it is a question of having to kill them but not humiliate them, which is something else.Footnote 77
This same person said: ‘Even though the body did not appear, we were notified of what happened. Sometimes we were even sent photos or videos.’ Another person said: ‘There was a certain respect [for people in the group]. They can punish you for what you did, but then only the people in the group need to know. They weren't going to go round the neighbourhood telling [everybody] about it because nobody was interested.’Footnote 78
Through dismemberment, the group enforces its own regulations, fostering the perception of a rational-bureaucratic order where rules are applied regardless of rank.Footnote 79 While the earlier case of the escapee shows this system is not infallible and remains subject to personal networks,Footnote 80 his flight itself implicitly acknowledges that even friendship would not exempt him from this norm.
In the two cases of internal discipline in which the bodies were displayed in public, the individuals were accused of stealing money from the organisation. One member of the criminal group stated: ‘Usually nobody keeps an eye you, and you have a profit margin that you keep. If you get greedy and take more than you should, you may be putting your life at risk.’Footnote 81 When asked why in these cases the bodies were put on display, several interviewees mentioned that these people were usually new members or not well liked by the command: ‘Whoever dared to do something like this was usually a rookie. What I think is that as they were rookies the bosses didn't have to respect them. What's more, this way they were sending the message to future members that they shouldn't think they could behave in this way.’Footnote 82
Dismemberments to Enforce Social Order
Surprisingly, we found that most of the dismemberments in Barranquilla (12) fall into the category of ‘dismemberments to enforce social order’. The reasons identified for these dismemberments were fourfold: punishing someone for revealing information to the authorities (what the armed groups call ‘ser sapo’, contextually translated as ‘law of silence’); killing a member of the armed group – it could be in a fight or as the result of a personal grudge; engaging in criminal activities without the organisation's permission – such as stealing or becoming a hired killer; and getting romantically involved with the partner of a member of the armed group.
Within this group of people, we could distinguish between those who were killed for committing crimes outside the armed group's supervision (murder, robbery or drug dealing, six people) and those who were killed for different reasons (six people). In all cases, they broke rules that the armed group had established for civilians in the neighbourhoods where they exercised criminal governance.
We classified as ‘public’ 11 of the 12 dismemberments in this category.Footnote 83 The most common way in which these dismemberments were made public was with the display of limbs, torso and head in different neighbourhoods, similar to what was observed with homicides that served an external function. Ten cases appeared in this way. One police officer, referring to these cases, mentioned that: ‘by doing that, the homicide itself is sending a message through to many neighbourhoods. In all those neighbourhoods, people then knew that if they collaborated with us, something similar could happen to them.’Footnote 84
Six of the victims of enforcement of social order dismemberment between 2013 and 2022 were not accused of committing any crime. In these cases, the reasons identified for their homicide were twofold: romantic involvement with the partner of a member of the armed group and breaking the ‘law of silence’.Footnote 85 The latter was the most common reason for civilian dismemberment. In some neighbourhoods, ‘Los Costeños’ had established ‘the law of nobody saw and nobody knew anything’ (‘la ley del nadie vio y nadie supo nada’).Footnote 86 Basically, this meant that in neighbourhoods under the control of this armed group, people were not allowed to talk about ‘drug seizures, captures of criminals, criminal attacks, or any other matter of public order’.Footnote 87 An inhabitant of the Las Flores neighbourhood said that ‘Here we already know the issues that cannot be touched. To talk about it is to put a gun to one's head.’Footnote 88 The dismemberments of these people were in response to the fact that they revealed information to the National Police about some of the events happening in these neighbourhoods. The distribution of body parts throughout different neighbourhoods, as one former member of an armed group put it, served the function of ensuring that everyone under their control knew that such activities were forbidden: ‘A killing like that puts everyone in a panic. When things are like that not only do people begin to behave themselves in the [victim's] neighbourhood, but people further away hear about it too and avoid making trouble.’Footnote 89
Among those six persons killed because they were accused of committing crimes, there was one case in which the person was dismembered for murdering a member of ‘Los Costeños’. In the other cases, all the killings were because the people were carrying out criminal activities in the OCG's territory, but without its permission. In one case, the person killed was working as a hitman and, days before he was dismembered, he received a threatening phone call saying ‘In La Luz [neighbourhood] I am in charge, so you cannot work for yourself. We are in charge here.’Footnote 90 After his dismemberment, some of his limbs appeared in La Luz. In another case, the victim's body appeared in La Luz with a sign that read: ‘This will happen to the other criminals of this neighbourhood. This will happen to all those who continue to do harm here. Robbery is not allowed.’Footnote 91 In two other cases, people were killed after the OCG discovered that they were reselling the drugs they bought as consumers, without being authorised to deal.
In almost all of the cases identified, the armed groups not only dismembered their victims, but then implemented a complicated scheme of distributing the remains throughout different neighbourhoods of the city. According to the interviewees and the location of these neighbourhoods, this distribution was not randomly executed, but aimed to deliver particular messages: the location of the head was usually close to where the person could be identified, while the rest of the parts appeared in different neighbourhoods that were also governed by the OCG. At the same time, there was always an effort to make the motive for the killing known, either through messages explaining the motive, or by tasking some members of the armed group with making the information known. As a former member of an OCG mentioned:
You had nothing to do with these deaths … but even so, sometimes you had to go to a neighbourhood and spread the word that this person was [killed for being] a ‘sapa’ (‘snitch’). That way I think everyone was in panic and nobody wanted to talk to the police.Footnote 92
Observable Implication 3: Profiles of Dismemberment Victims Change According to the Context of Violence
Several authors report that levels of organised criminal violence increase in times of competition for territory.Footnote 93 When armed groups do not have other actors challenging their dominance, then they will potentially seek to reduce violence to avoid attracting media or institutional attention, as it may affect their business.Footnote 94 However, if another armed group is challenging its criminal governance regime, the OCG has much greater incentives to increase levels of violence against its rivals.Footnote 95
In this scenario, if our argument is valid, dismemberments would then need to be used differently by armed groups according to the context. We would thus expect an increase in dismemberments in Barranquilla during the years when the OCGs were establishing criminal governance in competition with other armed groups (2013) or when this criminal governance was challenged by other actors (2016–17 and 2021–2). During the last ten years, there have been five years of established criminal governance and five of contested criminal governance. Figure 1 helps to situate the patterns of dismemberments by year.
Figure 1. Patterns of Dismemberments in Barranquilla (2013–22)
In total, 76 per cent of the dismemberments happened during the years when there were competitions for territory. All dismemberments of rivals occurred in 2013, 2021 and 2022. In addition, all dismemberments of gang members accused of betrayal occurred between 2016 and 2017, except for one, which happened in 2021. In other words, all the events that could be related to the competition for territory (either the dismemberment of rivals, or of members of the organisation suspected of having allied themselves with rivals) happened during the years of dispute.
In fact, the first dismemberments in Barranquilla occurred in 2013, the year of the first clashes between ‘Los Rastrojos-Caleños’, ‘Los Costeños’ and the AGC. This likely suggests that ‘Los Costeños’ used dismemberments to establish its criminal governance regime by deterring its main rivals in 2013. Subsequently, during the later years of competition, the dismemberments were a way to defend the criminal governance they had already built. At the same time, the fact that all dismemberments tend to increase during years when there is armed competition also suggests that armed competition may generate incentives to use them, even if not against rivals. Dismemberments can serve as a way of signalling a willingness to use force against all non-compliant civilians and thus also deter people from changing their allegiances.
It is also noticeable that, even though most of the dismemberments against civilians happened during years with inter-criminal competition (seven out of 12), during years with an established criminal governance the proportion of dismemberments directed against civilians was even higher (five out of six). This would imply that, during years without territorial disputes, civilians continue to be those at major risk of suffering this punishment.
Alternative Explanations
There are three alternative hypotheses that could be useful in explaining the cases of dismemberment by criminal groups. Although some of their premises have been contradicted by the data presented so far, we will dedicate the rest of the article to demonstrating briefly why we consider them inadequate for explaining the case of Barranquilla and, probably, any other cases where OCGs frequently use dismemberment.
Hiding the Identity of the Victim
As mentioned at the beginning of this article, some authors suggest that dismemberments serve the function of hiding the identity of the victim and, with it, information that could be useful for the authorities to find the perpetrators.Footnote 96 Although this hypothesis seems to be very useful for thinking about individually committed dismemberment cases, it has also been used to explain the actions of some armed groups during civil wars.Footnote 97 Dismemberment may not be discovered for a long time and, when it does, body parts may be difficult to identify. It makes sense that OCGs, which have as one of their central objectives to conceal their violence so as not to attract state and media attention, would use this tactic.
This hypothesis is unlikely to apply to Barranquilla. Of the dismembered bodies in Barranquilla, 75 per cent were found on the day of or, at most, a few days after the dismemberment. Furthermore, we observed that the perpetrators often left the victim's head close to their place of residence. Therefore, it is unlikely that the perpetrators were attempting to delay the identification of the victim or conceal their identity. On the contrary, they appeared to want the victim to be identified in order to convey their message. This was also confirmed by some members of the OCGs interviewed.
Crime to Disfigure the Victim
A second hypothesis that could explain these cases is that dismemberment happens because the perpetrator wants to unload all his violence on the victim's body. According to this idea, explanations relying on the assumption of an instrumental or strategic logic are insufficient because the purpose is to disfigure the body rather than simply to kill.Footnote 98 Dismemberments happen because people have learned to be cruel during war, not because they want to use the event for ulterior purposes.Footnote 99 We believe that this explanation cannot be immediately dismissed if we consider that some of the crimes could be heavily influenced by strong emotions (such as those that were the result of betrayal or infidelity). Moreover, there is no doubt that dismemberment is an act of extreme violence to the victim's body.Footnote 100
However, we believe that, although it may be part of the story, this explanation leaves out cases in which, as a sign of respect for the victims (who were almost all members of the organisation), the bodies were not put on public display, by order of the OCG's leaders.Footnote 101 If they were not going to display the victim's body, they could have used selective killing. Nevertheless, they carried out dismemberment.
We argue that this constitutes compliance with a rule and not an act of hatred against the person. Therefore, it is not just a dehumanising crime (although it is), but a crime that serves an ulterior purpose: to enforce the norms that underpin criminal rule. Dismemberment makes it possible to maintain the rational-bureaucratic order of a given regime of criminal governance; otherwise there would be no point in carrying out these concealed dismemberments, or the punishment of others that could have been better dealt with by targeted killings (such as of those who violated the law of silence or who were minor criminals in the neighbourhood). Dehumanising is an inherent part of dismemberment, but by itself it does not solve our puzzle.
Gaining Territorial Control
As already mentioned, many published studies examining the extreme violence exhibited by OCGs have focused on Mexican drug cartels. The predominant conclusions among researchers are that gruesome violence is often employed during struggles for territorial control or attempts to establish dominance in new markets.Footnote 102 The purpose of such violence is twofold: to terrorise the population and rival groups,Footnote 103 and to foster intra-organisational cohesion.Footnote 104
Under the first argument, armed groups are cruel to their enemies because they have dehumanised them or because they need to make credible the threat of violence against all those who try to encroach on their territories.Footnote 105 However, we have seen that extra-lethal violence in Barranquilla also occurred during periods of consolidated territorial control: only 25 per cent of dismemberments over the past decade were directed against rivals, while 75 per cent were perpetrated against criminals outside the organisation, non-criminal civilians, and members of the armed group itself. While this partial explanation sheds light on some scenarios, we believe it may not fully account for territories with persistent high levels of dismemberment over an extended period.
The second explanation, which suggests that gruesome violence is aimed at building intra-organisational cohesion, also offers only a partial understanding. It is true that extra-lethal violence can foster a sense of community and identity within armed groups. However, our case study in Barranquilla revealed elements of dismemberment practices that deviate from the ritualistic logic described by Valentin Pereda.Footnote 106 For example, there were cases where dismemberments were communicated to organisation members without displaying the bodies, indicating a different motivation beyond ritual. One of the interviewees confirmed this:
We knew that there were people who were very respected by the bosses. [In that case], well, it was not a spectacle as it was with others, but here it was a punishment for going against the organisation. We knew about it and why [it had happened], but we didn't watch it. This is part of the respect that is shown to some members.Footnote 107
Conclusions
In this article, we have explored a phenomenon that has received relatively little attention in organised crime studies: human dismemberment. This is a rare form of violence compared to others. Nevertheless, we consider that our argument can be useful to explain other forms of extra-lethal violence that are publicly displayed by OCGs.
To evaluate this nexus between dismemberments and organised crime, we analysed the case of Barranquilla. Our methodological strategy consisted, first, in understanding how many of the dismemberments were potentially related to OCGs. To do so, we triangulated all available databases on the phenomenon in the country and in the city. Having characterised the pattern, we analysed the information based on three elements: (i) the ways in which dismemberment was presented to the community, (ii) the profiles of the victims of dismemberment and (iii) the variations over time of this form of violence. With this information, the argument we present is that dismemberments are used by armed groups for three central purposes. The first is to make the territorial establishment of an armed group credible. Dismemberments against rivals increase when one armed group attempts to challenge the territorial establishment of the other. Thus, the message sent to rivals is that there is a willingness to defend that territory with a significant level of viciousness, which is intended to deter competitors. The second purpose is to discipline members of the armed group who betray the organisation. The third purpose is to enforce the norms that the OCGs have established for civilians subject to their criminal governance.
Taken together, the dismemberments generate a sense of fear that is functional to establishing and sustaining a social order and discouraging further challenges to criminal governance. Our article thus connects the literatures on dismemberment in civil wars, on extra-lethal violence in settings of organised violence, and on criminal governance.
There are two questions related to dismemberment that our research did not resolve and that may be useful for further research. As we show here, dismemberment appears to be useful both for deterring rival organisations and for generating credibility for the system of norms produced by criminal governance. It is pertinent, then, to ask what makes some OCGs carry out dismemberment, whilst others do not. The literature on civil wars can be useful in this regard, showing how such violent acts are taught and how they become practically mandatory for members of some armed groups.Footnote 108
The second question is: how do the state and society react to dismemberments? It is worth remembering that these happened mainly in cities (in the Colombian case) and that on at least 20 occasions human remains were displayed in different neighbourhoods of the city. Even so, our fieldwork and press review shows that there has been neither a large-scale institutional response nor a public reaction against the practice. On the contrary, for ten years the authorities have tended to claim that dismemberments are not systematic events, but rather the settling of scores between criminals that should not be of concern to civilians.Footnote 109 This framing of violence has the potential to limit public solidarity as well. If it is suspected that the murdered person was involved in criminal affairs, it is likely to be assumed that the phenomenon will not be a problem for others.Footnote 110 Understanding how the media, citizens and authorities talk about and respond to dismemberments is intriguing and a fruitful avenue for further research.
Finally, an issue on which there is still much room for research is the micro-politics of dismemberments: is the decision to dismember taken by someone specific? How does the norm of dismemberment as punishment for the organisation come about and how is it adopted? How does the organisation get its members to internalise these ideas? This is one of the most difficult avenues of research to pursue given the secrecy of some armed groups and the messiness of these processes. Still, it is worth attempting to better understand these small decision-making processes in order to contribute to better policies to combat the phenomenon.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022216X25000057.
Acknowledgements
The data used in this study has been collected with the cooperation of the Centro de Pensamiento UNCaribe based at the Universidad del Norte (Barranquilla, Colombia). It is a result of the research project ‘Dinámicas de la insurgencia armada en el Caribe colombiano’ (‘Dynamics of the armed insurgency in the Colombian Caribbean region’), ref. 2022-037. We want to thank all the people who gave up their time for us to interview them and who shared their valuable experiences with us, even knowing some of the information was hard for them to share. Also, we want to thank the participants in the Political Violence Workshop at the University of Chicago for their insightful comments on earlier versions of this article, and express our gratitude in particular to Andreas Feldmann, Bengisu Şimşek, John H. Murdy, Madeleine Stevens, Ari Weil and Andrés Uribe for their detailed reading and discussions of earlier versions of this manuscript.
Data availability
An anonymised version of the qualitative dataset that we used is available in the Supplementary Appendix. An extended version of the dataset can be shared with researchers upon reasonable request, given the sensitivity and privacy of some of the data we used.
Open access
