The social structure of armed groups. Reproduction and change during and after conflict

ABSTRACT

Current research on civil wars and conflict increasingly turns to the inner structure and functioning of state and non-state armed groups and their impact on aspects such as violent practice, internal cohesion and the dissolution of these groups during the conversion to peace. The first aim of this introduction is to set out the theme of this Special Issue on the social structure of armed groups and previous research within the field. The second aim is to introduce the contributions within the Special Issue, alongside possible trajectories of future research on the ‘meso-foundations’ of civil war and conflict.

KEYWORDS:

• Social structure
• recruitment
• cohesion
• dissolution
• networks
• armed groups

1. Introduction

The last two decades of peace and conflict research have seen the emergence of two major and inter-related approaches to wartime violence. The first was a turn to the ‘micro-foundations’ or ‘micro-politics’ of armed conflict – challenging previous macro-analyses, which, due to their lack of nuance, paid little attention to the evolution of war, the changing wartime motives of belligerents and the exigencies created by war itself. In a similar vein and in order to remain in the wording of analytical levels, the second approach introduced a stronger focus on the intermediate or the ‘meso-foundations’ of armed conflict. Within this analytical perspective, the inner structure and functioning of armed groups are increasingly under the scrutiny of researchers of civil war and conflict. Throughout the last decade, many conflict scientists started to analyse meso-level structures with regard to their impact on aspects such as violent practice, internal cohesion and the dissolution of these groups during the conversion to peace. Here, armed groups are not viewed as unitary strategic blocs or collections of individuals but as entities consisting of different organisational modes and structures of cohesion, as well as social groups and network characteristics, all of which have an impact on the group’s external and internal behaviour. This Special Issue adds to this fast-emerging literature by gathering contributions which draw mainly on sociology, political science and history to understand the social and organisational structure of insurgent and incumbent militaries, and by discussing their patterns of change and impact on practices within and outside the respective groups.

The social and organisational structure of armed groups during civil war, we argue, should be seen as a dynamic system, as a structure that is a constant process of both reproduction and change, even during times of massive societal upheaval, such as violent conflict. Therefore, the impact of the social and organisational structure and its internal differences on a group’s practices will be addressed within three timeframes: 1) during the inception, 2) during different stages of the conflict and 3) during the transition to peace and after the end of the war. Before providing an overview of the issue’s contribution to this field of research – which views armed groups as complex, ever-changing and non-unitary structures consisting of interrelated networks and social groups – we will start by setting out the themes and academic context of this Issue. To do this, the introduction paper is divided into four sections. The turn to the micro-foundations of conflict will be discussed in the first section, while the second section will lay out the emergence of corresponding meso-level approaches to violent conflict with regard to organisational modes, armed network characteristics and internal cohesion. Following this, the third section turns to the subtopic of the reproduction and change of the social structure of armed groups and their internal social composition and structuration, before presenting the contents of each of the Issue’s contributions and an outlook on possible trajectories of future research within the fourth and concluding section.

2. The micro-foundations of conflict: space, motivation and wartime change

Throughout the early 2000s, discussions in conflict research have largely revolved around the question of whether internal conflicts are caused either by justice-seeking ethnic, political or religious motives (the so-called grievance approach)¹ or by greed, in which personal economic opportunities render participation in collective violence thoroughly rational and beneficial for the individual.² Representatives of greed theory in particular usually either took macro data to speculate over individual motives or tried to calculate the economic and societal conditions (‘opportunity structures’) under which it becomes rational for an individual to personally engage in collective violence; this is as opposed to staying away from the violence and simply hoping for a group to succeed in its struggle for justice or to deliver public goods (the so-called ‘free rider problem’).³ However, using large-N macro data on, for example, social inequalities, poverty, education, natural resources or primary commodity exports of a country turned out to be too speculative to explain why wars broke out in the first place. On top of that, researchers could not explain the complexities of civil war, the ever-changing employment of (violent) practices by armed groups as well as their internal differences.

Therefore, researchers increasingly turned to disaggregated, so-called micro-level statistics in order to improve causal inference as compared to large-N studies⁴ and also did something that greed theorist Paul Collier viewed as a rather expandable exercise: collecting fine-grained data through surveys and qualitative interviews with belligerents in order to better understand their motives and the logic behind different patterns of violence.⁵ Within the micro-level approach, studies have not speculated upon the motives of rebels as indicators for causes of war through analyses of macro-sociological and socio-economic data but have rather turned the relationship upside down by using subnational and disaggregated data to explain the rebels’ practices and changes across time and space.⁶ Thereby, instead of uncovering an all-encompassing cause, studies pointed towards a multiplicity of motives for and drivers of collective violence, which may change over the course of conflict and which differ from region to region.

Stathis Kalyvas pioneered micro-level research by using spatially aggregated data from the Greek civil war in order to understand the employment of different types of violence by warring factions.⁷ Instead of taking indiscriminate or selective violence as an indicator for hidden greed or grievance scripts, he pointed to an endogenous and war-related factor to explain the spatial variation: territorial control. Patterns of violence do not simply relate to pre-war motives or ideologies but are, to a large extent, also the result of emergent dynamics of warfare itself.⁸ Kalyvas’ study sparked a turn to micro-level data and is closely related to similar research methodologies such as the ‘disaggregation of civil war’, which also disaggregates below the country level but retains the power of large-N comparative data sets across countries to ensure external validity.⁹ Studies based upon data sets like these have improved the understanding of micro-foundations and local dynamics in civil war as well as inequality and conflict – especially when it comes to the analysis of (horizontal) inequalities as a source of conflict.¹⁰

Large-N research using too coarse country-level data saw no differences between armed groups and different locations and was not sensitive enough for changes during wartime in terms of mass-level preferences. Research on the micro-foundations of conflict, by contrast, pointed to the complex social and political processes of conflict and to the fact that much of the action in times of war is framed by war itself. Taking a look at actual motives and life courses of combatants and military commanders makes clear that recruits usually enter armed groups long after the conflict has started, driven by war-time-related exigencies, constraints, benefits and needs. In addition, it showed that warring organisations and belligerent motivations are to a certain degree in a constant state of flux, or, as Stathis Kalyvas put it: ‘Neither organizations nor preferences are given ex ante and fixed throughout the war. Change is synonymous with war’.¹¹ Social processes in civil war are heavily affected by warfare itself. Elisabeth Wood, for example, pointed to processes such as polarisation, the fragmentation of market structures and local economies and the reshaping of gender roles as well as the formation of new relations in the militarisation of civilian governance, amongst others.¹² Like Kalyvas, Wood makes a strong argument for the study of social interactions that are the key for understanding change during and the dynamics of civil war.¹³

According to this view, recruits entering armed groups in civil war are not just ‘loose molecules’ – as Robert Kaplan called them – or free agents on an open market¹⁴ deciding to start new lives as combatants after profound deliberation over individual benefits, only then to enter the warzone after a certain period of travel.¹⁵ Members of armed groups can be driven by various motives, such as an attempt to gain political recognition, honour, future prospects or a ‘pleasure of agency’¹⁶; by hopes of achieving social securities through membership in armed groups or the desire to protect their own property¹⁷; or by a perceived threat against their own community and identity group.¹⁸ Micro-foundation studies made clear, inter alia, that war needs to be taken into account as a factor in recruitment. Few members enter armed groups when the war is not already in full swing and has not already deeply impacted their immediate milieu and community.¹⁹

Therefore, for many members of armed groups, the experience of violence within their local community and their closest milieu is of major importance, as is the wish to retaliate.²⁰ Since warfare threatens the lives of civilians, others also enter armed groups for the sake of physical protection, since in warzones these might quite simply be the safest places to be.²¹ Instead of reflecting a deeply felt commitment to the official cause or the ‘master cleavage’, conflicts consist of multiple small cleavages rather loosely aligned towards a master narrative²², and the wish to enter an armed group is often driven by a desire to survive the exigencies created by the war itself or to receive essential social, economic or even medical goods within warzones.²³ This, at least, seems to be the case for low-ranking combatants.²⁴

Exigencies of war constantly (re)create the conditions of their own creation; in so doing, they potentially lead to their own escalation. Analysing the micro-level exigencies of civil war also brings a too-narrow focus on individual motives into question as this is just one side of the coin of recruitment. Amongst others, one simple argument for moving away from a sole focus on individual motives is the long-ignored fact that, in many conflicts, forced recruitment and even abduction are quite common.²⁵ Often, forced recruitment – especially of minors – is even part of a close-knit socialisation into violence, ultimately altering the pre-recruitment motives of belligerents.²⁶ Studies, therefore, should not take only individual pre- or post-recruitment motives into account, but also organisational aspects, in-group socialisation and its disciplinary apparatus as well as ‘modes of access’ or ‘pathways’ of different social groups into armed factions.²⁷ To do so, studies on the meso-foundations of civil war increasingly make empirical use of subnational, disaggregated and so-called ‘micro-level’²⁸ data as well as surveys and qualitative data but focus analytically on the intermediate level, the organisational and social structures and internal differences as well as network compositions of state and non-state armed groups, in order to explain both their internal and external behaviour during various stages of civil war and conflict.

3. The meso-foundations of conflict: organisations, networks and cohesion

One of the earliest and certainly most influential studies addressing organisational structures was Jeremy Weinstein’s book Inside Rebellion on the impact of recruitment strategies on insurgent violence against civilians.²⁹ Weinstein argues that groups that have a lot of resources at their disposal during their inception are likely to use economic benefits to recruit. Therefore, they are also likely to attract more opportunistic recruits, who are in turn also likely to exhibit higher levels of abusive violence against the civilian population. By contrast, groups without many economic resources during their inception are more likely to attract ideologically committed recruits who are not only willing to make long-term commitments to the organisation’s goals but also less likely to mishandle civilians. By arguing in this way, Weinstein turned attention towards the organisational aspects of wartime violence against civilians. According to Weinstein, the tactics, strategies and behaviour of insurgent groups are directly shaped by whom and how groups recruit and which rules they set to discipline their behaviour.³⁰

However, if boiled down, organisational aspects in his model do not really count as these are directly caused by the extent of (economic) resources available during a group’s inception. Organisational variables run the risk of ultimately shrinking to mere placeholders for initially available resources: ‘Whether a group is filled with activists or opportunists then constraints the choices leaders make as they organize military operations, govern civilian areas, and struggle to retain their members in the course of conflict’.³¹ As shown by Francisco Gutiérrez, this rather static model fails to account for variances between groups that have similar conditions at inception.³² Moreover, as highlighted by Yvan Guichaoua, not only does it not account for differences during inception, it also cannot explain the often erratic trajectories of armed groups and what such organisations are sociologically made of.³³ In the end, Weinstein treats armed groups as far too static and as unitary blocs within which few or no internal differentiations can be found that are relevant for its external behaviour. It also leads to the problematic conclusion that leaders are powerless before their recruits. They can only be held responsible for the fact of whether they started a rebellion under favourable or unfavourable conditions. Current research moves to more dynamic approaches in studying the complex competitive logic behind civilian abuse.³⁴ It also shows the importance of understanding the role of ideology in shaping organizational modes.³⁵

Furthermore, recruitment is not just a matter of acquiring combatants via money or ideological commitment. There are various modes and pathways of access into different levels within the inner hierarchy of armed groups. Recruitment into armed groups does not always consist of a centrally coordinated policy from above that is met by a rational decision made by socially loose molecules, ‘activists’ or ‘opportunists’, from below. As already pointed out, for example, forced recruitment is very common and the socialisation into violence, through the disciplinary apparatus of armed groups, may ultimately alter belligerents’ initial recruitment motives. Another point is the fact that it is often difficult to draw the line between a full-time member, a part-time supporter, a combatant, a mercenary or someone who joins for only a brief period.³⁶ Insurgencies in particular consist of loose alliances, of cores and peripheries, of different ranks and their respective modes of access.

On top of that, armed groups often not only make use of ethnic, religious, communal, political and social networks but ultimately consist of militarised networks, at the top as well as within the rank-and-file of a movement.³⁷ In terms of the institutionalisation and possible peer pressure at their disposal, these networks can differ sharply, as pathways via networks may include religious circles, political and ethnic associations or commitment to friends and family members. Many belligerents refer to contacts within their immediate milieu, such as a sports club, within associations or to a warrior heritage within their family of origin as the main reason for joining an insurgent movement.³⁸ Beyond different modes of recruitment, close solidarity bonds like those within a recruit’s social milieu often regulate access to armed groups, not just among the rank-and-file of a movement, but also within the higher echelons.³⁹ In many cases, individuals are not entering an insurgency, but instead they are part of an already existing network that militarises people.

Hence, social networks pre-condition certain trajectories of access and also determine some initial patterns of discipline, securing loyalty to ‘the cause’. Membership of a network is a strong predictor, explaining why individuals enter certain armed groups over others. Ana Arjona and Stathis Kalyvas, for instance, conducted a study in which they compared the motives of combatants for joining Colombian insurgent groups as opposed to those joining the incumbent state military.⁴⁰ Surprisingly, the motives to join one over the other do not differ significantly. Recruit motivations within both militaries did not differ in terms of classical explanations for joining an armed rebellion, such as economic incentive or communal grievances. Instead, the spatial and social proximity to a group turned out to be a reliable predictor for whether someone would join one group or the other. In accordance with earlier research, Arjona and Kalyvas pointed towards the importance of pathways (or ‘possibilities’) and modes of access, whether the network and institutions of a military faction reach into a recruit’s social milieu, or, quite simply, are within his or her spatial reach. That is why more than half of the respondents within the survey stated that they had known somebody within the respective group beforehand. In many conflicts, the line between insurgent and incumbent militaries is difficult to draw anyways and, over time, belligerents often become members of both. The binary distinction ‘state’ and ‘non-state’ is problematic in many regards – ignoring amongst others the ‘quasi-state’ character of armed groups – and sometimes tells us more about the observer and the politics of legitimacy within international politics than about the actual practices and structures of respective groups ‘on the ground’.⁴¹

However, by reproducing social networks, rebellions also reproduce their inner hierarchies and pre-existing structures of cohesion. Paul Staniland’s study of networks of rebellion showed how pre-war structures of cohesion are being both reproduced and changed by wartime dynamics.⁴² Staniland differentiated between armed organisations in terms of organisational control within their own ranks and with regard to the local populace in order to deliver a social-institutional explanation for their structure during inception as well as changes over time: ‘Insurgent groups are built by mobilizing prewar politicized social networks. These preexisting social bases provide information, trust, and shared political meanings that organizers can use to create new armed groups. The initial organization of an insurgent group reflects the networks and institutions in which its leaders were embedded prior to violent mobilization. […] These social bases have a variety of possible structures, however, which lead to different initial insurgent structures’.⁴³ The organisational structure of an armed group is to a large extent a continuation of loyalty structures, interdependencies, hierarchies and popular support within its social base. Civil war is not just synonymous with change (as Kalyvas had it), but with reproduction and change. According to Staniland, these initial structures determine certain pathways over time, but still organisational evolution and changes in network structures do occur, which explains why some insurgent groups are better equipped for a lasting rebellion than others.⁴⁴

In a similar vein, Klaus Schlichte developed a theory according to which insurgent armed groups can be understood as quasi-state structures, which have to legitimate their violent order, thereby following certain trajectories in order to institutionalise their rule and to prevent fragmentation.⁴⁵ According to Schlichte, armed groups have to defend their legitimacy in three political arenas: within their own followership, within the populace and within the international community. Depending on the structure of their rule, armed groups follow certain organisational trajectories and pursue different strategies in legitimising and institutionalising their violent order, which is constantly threatened by the ‘shadow of violence’ calling for the containment of its effects. Violence always runs the risk of spinning off and delegitimising the political goals of a group. To build their rule and to prevent decay, armed groups employ different ‘centripetal techniques’ such as subjectivisation, the shuffling of offices and the closure of social space and base their rule upon an increased patrimonialisation or formalisation.

As shown by Sarah Zukerman Daly in her comprehensive study on Colombia, organisational aspects and geographies of initial recruitment – whether insurgencies recruit combatants from the same locality in which they were deployed later on – strongly affect the levels of post-war organised violence and the likelihood of a relapse into civil war.⁴⁶ In doing so, Zukerman Daly points to the necessity of taking into account the continuities and discontinuities of geographies, organisational structures and trajectories of armed groups, even after official – in quotation marks – ‘reintegration’.⁴⁷ Post-conflict developments are intimately connected with structures and trajectories that originated before recruitment and have continued throughout the course of the conflict. Organisational structures and their social effects do not just fade away during militarisation or after the war’s end. To conclude this section, meso-foundation studies show that analysing armed groups means taking a comprehensive and holistic look at the intermediate level, at different social groups, their modes of access and cohesion, their network characteristics, their corresponding patterns of recruitment and in-group disciplining and socialisation as well as their fragmentation and internal hierarchies. All these aspects are in a constant process of reproduction and change during inception, during different stages of conflict and during a group’s breakdown or conversion to peace.

4. Social structures: reproduction and change

Armed groups are social fields that to certain degrees reproduce hierarchies and differentiations prior to their own inception.⁴⁸ Following Pierre Bourdieu’s concept of social fields, many positions within the internal structure of insurgent groups are homologous with pre-existing societal hierarchies.⁴⁹ By and large, insurgencies consist of a military and political elite at the top and their patrimonial network within the mid-range, while the rank-and-file, almost without exception, are recruited from the lowest strata of society: a war-displaced peasantry and working class, for instance. Biographies of commanders within the leadership and of the so-called ‘first movers’ who started rebellions mostly refer to deviants within a country’s elite.⁵⁰ Usually, the mid-range in command receives the least attention in academic studies, even though they are decisive in implementing command, in running rebellions and are the ones who run the risk of spoiling peace solutions if they are for example not properly included in disarmament, demobilisation and reintegration (DDR) programmes.⁵¹

A society at war does not turn social relations and structure completely upside down, but perpetuates them, while also transforming them in many respects.⁵² As argued elsewhere, power and disciplinary practices in armed groups are shaped by the habitus of commanders and soldiers; they are socially differentiated.⁵³ ‘Habitus’ refers to the totality of behavioural dispositions that have the tendency to reproduce behavioural patterns learned during the agent’s socialisation process in certain social milieu, not restricted to but especially linked to that of their family of origin. That disciplinary practices – that is, how commanders and soldiers try to make others or themselves loyal to command – are socially differentiated means that they reflect and are shaped by classificatory discourses of different social groups within the organisation’s fold, which in turn are structured by a soldier’s and commander’s habitus.

Non-state armed groups, in particular, are not uniform entities in terms of social composition but consist of multiple hierarchical layers and come from many different kinds of social groups, such as former state military commanders, commanders-turned-intellectuals and politicians, bureaucrats, former state military rank-and-file soldiers, soldiers-turned-white- and-blue-collar workers and peasants, amongst others. Moreover, as in any other social field, there is competition among these groups with regard to the overarching goals, organisational structure, internal rules, military strategies, ideological setup and frames and focuses of identity of the movement. Furthermore, the position of an individual within this armed network and its internal hierarchy becomes mediated by an agent’s possession of different forms of ‘capital’. Pierre Bourdieu’s four main forms of capital can help with the understanding of the social status and its changes within armed groups.

• Economic capital relates to any kind of possession with a monetary or exchange value within the field.

• Social capital refers to relevant connections to people holding influential and field-relevant positions.

• Cultural capital consists of all sorts of skills, knowledge and education, as well as the possession of culturally relevant objects. In the field of insurgency, for instance, military socialisation (a form of knowledge) is highly valued, as is the level of formal education in high school and university.

• The fourth capital is symbolic, represented by codified forms of superiority such as ‘honour’. Symbolic capital can be understood as a meta-resource, giving weight to other resources, defining the value of the other forms of capital when claiming a symbolically codified superiority within a certain social field.⁵⁴

According to Bourdieu, the position of agents within a social space consisting of multiple social fields both determines and is the result of their access to these four types of capital. However, of course, not all positions within an armed group are pre-determined by the pre-war social status as mediated by an individual’s possession of capital. War at the same time heavily alters the valuation of various forms of capital, adding value to some and subtracting value from others.⁵⁵

The social field of an armed group during civil war, in particular, sets its own rules and valuations of capitals. For instance, many individuals may rise in rank from a relatively low social status due to an increased value of combat experience and a collectively perceived ‘warrior ethos’. They are able to claim symbolic superiority through their display of strength, brutishness and fearlessness, which enabled them to rise in rank up to top-commanding positions. However, these commanders have to constantly not only display and replicate their image as fearless and rough soldiers but often also face the high volatility of their symbolic capital.⁵⁶ And commanders who are unable to maintain their high status after war are usually also the ones who prefer the continuation of or a relapse into war.⁵⁷ Being a strong and battle-hardened soldier is especially short-lived not only in the literal sense but it is also a comparatively highly field-bound resource, which – depending on the mode of transfer and the political conditions within the post-conflict society as a whole – runs risk to drastically fade after the end of a war.⁵⁸

Hence, the social structure of armed groups in civil war consists of multiple layers and internal differentiations. It is not simply a unitary bloc but consists of various social groups which have entered the field on different pathways and through different modes of recruitment, and it may comprise multiple cores and peripheries, more or less loosely structured and ever-changing alliances and loyalties, as well as recruits and groups that ‘hop on and off’.⁵⁹ There is a high degree of factional and internal fluidity with a frequent shifting of alliances and sides by competing individuals, units and groups.⁶⁰ Therefore, it holds true that war is in many regards synonymous with change.

However, as said, while insurgencies are characterised by constant change, there still remains a lot of continuity in social relations and hierarchies. Former military commanders or members of the political elite seldom take up arms and fight at the frontline. This activity is usually reserved for those coming from the lower social milieu. And while there is a lot of pressure and often even direct force to make them fight, the importance of ‘symbolic violence’ has not received much academic attention. As argued elsewhere, for many soldiers, fulfilling the role of the one who has to carry out the fighting is normal due to their self-perceived position in society.⁶¹

In many regards, combat soldiers during war perceive it as a ‘normal thing’ to become exactly that: combat soldiers. Or as one respondent in an interview put it, naturalising arbitrary social orders: ‘We [as ordinary soldiers] simply followed, what we were meant to be’.⁶² Soldiers and commanders within non-state armed groups are not just part of an insurgency, but also of a society’s symbolic universe, in which they themselves tend to accept their inferior positions while their commanders equally naturalise social hierarchies. As a KPNLF regiment commander put it by distinguishing between himself and ‘ordinary refugees’ becoming combat soldiers (word-by-word transcript with upper-case letters indicating intonations and brackets indicating a speech pause and its duration):

Respondent:

The file and rank was recruited (.) really from the (..) ORDINARY refugee [hm] (2) it was (..) it was not very difficult AT ALL.

Interviewer:

To recruit?

Respondent:

To recruit (2) BECAUSE (.) you know (1) those years (..) more than 10 years (2) it was STRANGE (..) special years (2) because I think that (..) PEOPLE seem to lose a LOGIC (.) of a NORMAL of human being (3) the psychology was really in war time (..) you (.) you see people do things that they NORMALLY don’t do [hm] (2) rape, stealing (..) you know KILLING (2) and it (..) when the PEOPLE (..) I don’t say that the people LOVE to do war (1) but for THEM (..) BECAUSE it is wartime (..) it is NORMAL that they are recruited to be a soldier of the resistance movement [hm] (3) so there is no problem AT ALL (.) you know.

Interviewer:

It is a (.) logical choice

Respondent:

Logical choice (..) YEAH (2) and life in the refugee camp (..) being a soldier (.) the life is much better than the normal life of a refugee [hm].⁶³

In times of war, it was a ‘logical choice’ for these ‘ordinary refugees’ to be recruited as combat soldiers. While pointing to the ‘normality’ of becoming a combat soldier, the commander also puts himself at the top of the hierarchy. Maybe, as Tarak Barkawi argued, citing Field Marshall Lord Wavell, ‘many battles and campaigns have been won by men who had little idea of why they were fighting, and, perhaps cared less’.⁶⁴ Amongst many others, symbolic, incorporated and naturalised hierarchies are an important source of ‘glue’ for internal cohesion.

The social structure of armed groups, and its reproduction and change during times of conflict, also matters during ‘reintegration’.⁶⁵ So far, there are very few systematic studies on the intermediate level and the importance of social differentiations and network characteristics during the conversion of armed groups. Usually, the focus lies on ‘insurgents’ and here on individuals and the question of proper incentives for entering a life as civilians.⁶⁶ Macartan Humphreys and Jeremy Weinstein, however, showed how, amongst others, ex-combatants’ previous membership of abusive units and their socio-economic characteristics impact the likelihood of either a successful or failed ‘reintegration’ while stating that there is little evidence that internationally funded programmes facilitate better demobilisation.⁶⁷ Unfortunately, instead of focusing on the whole social field of insurgent or incumbent militaries or its organisational and relational character, the focus in most studies lies on groups as unitary actors or individual characteristics of ex-combatants from the lower ranks, who on top of that have been part of demobilisation programmes. This focus is too narrow as war is a complex form of social conflict. There are also only few sociological studies on civil wars and conflict. Until today, in an academic field dominated by political sciences, economics and – to a certain degree – area studies, war and violence have mostly been avoided by sociological analysis that still treats both as a pre-modern phenomenon and not as something that is an integral part of the formation of modernity and its social conflicts.⁶⁸

5. Paper presentation

The papers within this Special Issue approach the study of the social and organisational structure of armed groups in civil war from different angles, in different conflict regions across the globe and throughout various stages of conflict. Francisco Gutiérrez Sanín highlights how crucial it is to analyse organisational structures if one wants to understand the trajectories of armed groups. His paper points to an organisational continuum between more loosely structured armed networks and state military-like armies. His central thesis is that the successes as well as the ultimate demise of the Revolutionary Armed Forces of Colombia (FARC) can only be understood by reference to the adoption of a ‘militaristic blueprint’ during its 7th conference in 1982, which led to strict militaristic discipline and a high level of centralisation in decision making. This change in mode of organisational structure allowed the FARC’s leaders to tame centrifugal tendencies. Without reference to this organisational principle and change, it is impossible to understand the expansion of the FARC, its relations to narco-business and other rent-seeking economies, or its patterns of violence. However, interestingly, the decision to adopt a more state military-like blueprint explains not only the successes of the group during times of isolation and heavy competition but also its ultimate demise due to massive rates of desertion and a worsening of its relations with the local populace.

Ernesto Cardenas, Kristian Skrede Gleditsch and Luis Carlos Guevara turn to the impact of network structures on Disarmament, Demobilisation and Reintegration (DDR) programmes. They argue that more cohesive and tighter networks employ higher levels of supervision and control within their units, which increases the probability of successful ‘reintegration’. Using an innovative network analysis to map and characterise the structure of the United Self-Defense Forces of Colombia (AUC) and the Revolutionary Armed Forces of Colombia (FARC), Cardenas et al. show how the probability of remilitarisation rises in smaller and more isolated components and that military units with a high degree of centrality within the network play an important role in the risk of conflict recurrence.

Yvan Guichaoua, Andrew Lebovich and Nicolas Desgrais deliver an analysis of two groups active in Northern Mali, the MAA-Tabankort and the MSA, both of which are presently aligned with the government. Yet while the MAA-Tabankort originally belonged to a jihadi outfit, the MUJAO, the MSA was integral to the separatist movement that launched the rebellion in 2012. Building upon the works of Paul Staniland’s Networks of Rebellion and Fotini Christia’s Alliance Formation in Civil Wars, the paper shows how both groups followed a dual logic: the MAA was designed mainly to preserve business interests, which is why it had major incentives to distance itself from the MUJAO (which was also chosen as its protector for business reasons in 2012) after the French intervention.⁶⁹ It also had to distance itself politically from members who stayed active in the jihadi outfit after the French intervention. French consistent counter-terrorist efforts exerted huge pressure on the MAA. Similarly, the creation and evolution of the MSA followed a dual logic: it was based on the preservation of the interests of a Tuareg tribe that was originally part of the separatist coalition, but economically active in an area distant from the stronghold of the separatists and exposed to pressures from France and Niger. Here again, top-down pressures and local interests combined to produce a split among the separatists and the formation of a new, ad hoc, local coalition. The paper shows how alliance formation and de-formation consider anticipated dividends on the national political scene and in the local political economy, which include sticky, identity-based expectations from leaders’ constituencies as well as strategies pursued by local rivals. Despite an apparently messy wartime dynamic, armed groups can navigate three options, produced by a history of rebellions and state responses, specifically: separatism, the jihadist agenda and alignment with government.

Drawing on Bourdieu’s four types of capital (economic, social, cultural and symbolic), Claire Metelits turns to the study of the Polisario Front, which is part of one of the longest-running disputes in African history, involving the fight for the independence of the Western Sahara from Morocco. Based on field research in the United States, Algeria and the Western Sahara, Metelits, in her fine-grained analysis, delivers a field-theoretical approach to the resilience of the Polasario Front. Making use of Bourdieu’s theoretical tools, Metelits shows how the usage of different types of capital may explain the armed group’s continued existence despite shifting regional and global international politics. However, interestingly, economic capital turns out to be the weakest resource within the portfolio of this long-running insurgent movement, thereby carrying less explanatory value for its resilience than symbolic or cultural resources. The paper argues for a balanced view on a complex set of capitals, instead of focusing – as greed theorists certainly would – solely on one type of capital. Bourdieu’s categories may help to balance the discussion and point to a new direction of research, which includes complex portfolios of ‘capitals’ and their interplay.

Likewise drawing on Bourdieu, Philipp Münch argues that the social structure of insurgent movements yields strong explanatory power in understanding the stances taken by an armed group. His contribution builds upon a case study of the two largest insurgent groups in Afghanistan – the Taliban and the militant wing of the Hezb-e Islami-ye Afghanistan (HIA). He locates these groups among the inhabitants of the most powerful positions (i.e. the ‘forces of conservation’) and their challengers (the ‘forces of heresy’). Combining this relational perspective with an analysis of the social background of the two groups, Münch explains how – within the current political setting – the Taliban are more likely to be a force of ‘heresy’, while the HIA’s social structure in terms of similarities in habitus and strong personal connections to government actors makes them more likely a force of ‘conservation’. Thereby, Münch convincingly manages to relate the practices of armed groups to broader social structures and struggles on the political field.

Turning to the analysis of organisational factors in the explanation of combat cohesion, Siniša Malešević shows how micro-level solidarities depend largely on macro-level organisational processes. Malešević’s contribution is based upon the case study of two armed forces involved in the 1991–95 Wars of Yugoslav Succession (between the Croatian Army and the Bosnian Serb Army). Drawing on in-depth interviews with former combatants, Malešević analyses how important social cohesion was for the Croatian Army in winning the war and how the micro-level solidarity was intimately linked to larger organisational developments and structures. Similarly to Gutiérrez Sanín, Malešević points to the importance of organisational aspects in the analysis of micro-level practices and solidarities of armed groups, thereby making a strong point about the analysis of different kinds of meso-level structures and their impact on the micro- and macro-level dynamics and outcomes of civil wars.

Offering an in-depth description of the Movement for Democracy in Liberia (MODEL) based on historical ethnography, Ilmari Käihkö points to the need to analyse sources of legitimacy during different stages of the existence of armed groups. His comprehensive contribution argues that much of the armed group’s behaviour depends on a harmonisation of legitimacy on the micro, meso and macro levels. The case of MODEL also illustrates the difficulties of doing this in practice. Not only did this ad hoc group have few resources it could dedicate to this task, but the different kinds of legitimacy all evolved over the course of the conflict. These different kinds of legitimacy, at all three levels, conflicted to some extent with each other.

And finally, Regine Schwab turns to the emerging literature on ‘rebel governance’ with an analysis of three regions in Syria to compare rebel-juridical structures. In the existing literature, insurgent courts have usually been analysed in the context of one hegemonic faction, such as the Afghan Taliban or the Tamil Tigers in Sri Lanka. What has been missing so far is an analysis of different pathways of formation, stabilisation and ‘success’ of rebel courts. As exemplified by the three case studies of judicial institutions in Eastern Ghouta, Idlib and Aleppo, these are shaped by the distribution of power between ‘same-side’ groups, yielding unipolar, bipolar or multipolar constellations. Another finding is that, whereas all courts emerged as part of the organisational structure of armed groups, they developed differently in terms of the degree of autonomy they were able to achieve.

Acknowledgements

Daniel Bultmann acknowledges support from the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG); grant reference: BU 3104/1-1.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) BU 3104/1-1.

Notes

1. Gurr, Why Men Rebel.

2. Collier, “Doing Well out of Civil War”; and Collier and Hoeffler, “Greed and Grievance in Civil War.”

3. For theoretical critique, compare: Kalyvas and Kocher, “How ‘Free’ is Free Riding in Civil Wars.” On ‘opportunity structures,’ compare: Fearon and Laitin, “Ethnicity, Insurgency and Civil War.”

4. Weidmann, “Micro-level studies.”

5. In his famous paper on ‘doing well out of civil war,’ Collier argues that qualitative data taken from interviews is completely useless as it cannot be trusted; Collier, “Doing Well out of Civil War” As argued by Sidney Tarrow, the turn to qualitative methods is not really a turn but a re-turn, cf. Tarrow, “Inside Insurgencies,” 590.

6. However, micro-level data may include variables on meso-structures such as organisations, thereby, at times, producing confusion when talking about data and analytical levels. ‘Micro’ with regard to data usually only refers to being subnational and disaggregated, which in a three-leveled analysis may by definition include meso-structures.

7. Kalyvas, The Logic of Violence in Civil War; and “Wanton and Senseless.”

8. Many studies followed his lead; for a most recent comparison, see for example, Balcells, Rivalry and Revenge.

9. Cederman and Gleditsch, “Introduction.”

10. Cederman et al., Inequality, Grievances, and Civil War.

11. Kalyvas, “Ethnic Defection in Civil War,” 1063.

12. Wood, “Social Processes of Civil War.”

13. Cf. Tarrow, “Inside Insurgencies,” 596.

14. See Ilmari Käihkö’s contribution in this special issue.

15. Kaplan, Coming Anarchy.

16. Wood, Insurgent Collective Action.

17. Kalyvas, “Urban Bias in Research on Civil Wars,” 173.

18. Gutiérrez, “Dilemmas of Recruitment.”; Cederman et al., Inequality, Grievances, and Civil War; Fearon and Laitin, “Violence and the Social Construction of Ethnic Identity.”; and Smith, Nationalism.

19. However, tracing warfare unambiguously back to its ‘first movers’ might in most, if not all cases, be an impossible exercise.

20. Fuji, Killing Neighbors; and Balcells, “Rivalry and Revenge.”

21. Nordstrom, Paths to Domination, 271; and Kalyvas and Kocher, “How ‘Free’ is Free Riding in Civil Wars.”

22. Kalyvas, “The ontology of political violence.’

23. Guichaoua, “Introduction.”

24. Most of the literature does not study the motives of higher ranking members of armed groups such as mid- or top-range commanders. There is a strong focus on understanding foot soldiers, in whose discourse and action the logic of violence is believed to be found.

25. Weinstein and Humphreys, “Who fights.”

26. For an analysis of how ‘membership matters’ in socialising allegiance even under conditions of forced recruitment compare Gates, “Membership matters.”

27. Bultmann, Inside Cambodian Insurgency, 15–22. For an approach on how institutions and behavior of combatants intervene compare Green, “Armed group institutions and combatant socialization.”

28. See supra note 6.

29. Weinstein, Inside Rebellion.

30. Weinstein and Humphreys, “Handling and manhandling civilians.”

31. Weinstein, Inside Rebellion, 126.

32. Gutiérrez, “Telling the difference.”

33. Guichaoua, “Who joins ethnic militia,” 22.

34. For example, the role of competition between warring factions, compare Metelits, Inside Insurgency.

35. Gutiérrez and Wood, “Ideology in civil war.” Laia Balcells and Stathis Kalyvas pointed to the ‘Marxist paradox’ according to which ideologically highly commited groupshave low rates of success. Balcells and Kalyvas, “Revolutionary Rebels.”

36. Guichaoua, “Circumstantial Alliances and Loose Loyalties.”

37. Hoffman, “The meaning of militia.”; and Giustozzi, Koran, Kalashnikov and Laptop, 43–69.

38. Compare Argo, “Why fight.”; Schlichte, “When the ‘Facts’ Become a Text.”

39. Bultmann, Inside Cambodian Insurgency, 165–73.

40. Arjona and Kalyvas, “Recruitment.”

41. Schneckener, “Militias and the Politics of Legitimacy.” See also Ilmari Käihkö’s contribution in this special issue noting that armed groups are political actors governing certain territories in which they are functioning as quasi-states. On the notion of armed groups as quasi-states, see Jackson, Quasi-States. In line with this, armed groups often, amongst others, comprise of former members of the military elite of the ‘incumbent’ military alongside their patrimonial entourage – both of whom often also return into at least similar positions after the war’s end. Cf. Bultmann, Insurgent Pathways to Peace.

42. Staniland, “Organizing Insurgency.”; and Networks of Rebellion.

43. Staniland, Networks of Rebellion, 9.

44. Staniland distinguishes between integrated groups with strong central discipline and robust local control, parochial groups with weak central discipline and robust local control, vanguard groups with strong central discipline and weak local control, and fragmented groups with both weak central and local control, which ‘find it very difficult to change and are likely to be quickly marginalized’ (ibid., 11.).

45. Schlichte, In the Shadow of Violence.

46. Zukerman Daly, Organizing Violence after Civil War.

47. The term ‘reintegration’ should be used in quotation marks not only because of the previously mentioned often blurred distinction between incumbent and insurgent militaries and the quasi-state nature of most armed groups governing territories (supra note 41) but also because it implicates a return from the outside of society. As Torjesen already pointed out, in most cases, this would imply that armed groups are cut off from social relations during war, which they clearly are not, even if they are operating from geographically remote areas. Cf. Torjesen, “Towards a theory of ex-combatant reintegration.” And it also implies that only ‘insurgencies’ are in need of post-conflict conversion programmes.

48. Bultmann, “Analyzing the Cambodian insurgency as a social field.”

49. Cf. Bourdieu, “Field of Cultural Production.”

50. Tilly, Politics, Schlichte, In the Shadow of Violence, 30–56.

51. Alden, “Making Our Soldiers Fade Away.”

52. For a comprehensive study of the patterns of perpetuation and change within the wartime social condition compare Lubkemann, Culture in Chaos.

53. Bultmann, Inside Cambodian Insurgency.

54. Bourdieu, “Forms of Capital.”; Logic of Practice; and Rehbein, “Differentiation of Sociocultures.”

55. Economic resources for instance have the tendency to lose in value, while social resources gain in value, cf. Schlichte, “Profiteure und Verlierer von Bürgerkriegen.”

56. Bultmann, Insurgent Pathways to Peace.

57. Bultmann, Insurgent Groups during Post-Conflict Transformation.

58. While this was certainly the case in for example Cambodia, this might change from case to case depending on the post-conflict valuation of a warrior ethos and the mode of ‘reintegration’. For a slightly different transition under different political conditions, see for instance Hensell and Gerdes, “Exit from war.”

59. Giustozzi, Koran, Kalashnikov and Laptop; and Guichaoua, “Circumstantial Alliances and Loose Loyalties.”

60. Wood, “Social Processes of Civil War,” 542.

61. Bultmann, “Normality of Going to War.”

62. Interview with a rank-and-file soldier of the KPNLAF. 9 August 2015. Battambang, Cambodia. Original in Khmer. Ibid., 104.

63. Interview with a regiment commander of the KPNLAF. 2 June 2011. Phnom Penh, Cambodia. Original in English, word-by-word transcript with brackets indicating speech pauses. Partly cited from Bultmann, Inside Cambodian Insurgency, 135.

64. Barkawi, “Subaltern Soldiers,” 24.

65. See above 56.

66. For example, Willibald, “Does money work.”

67. Humphreys and Weinstein, “Demobilization and Reintegration.”

68. Malesevic, Sociology of War and Violence.

69. See supra note 42; and Christia, Alliance Formation.