Abstract
Non-state war actors have trajectories. While recent research contributions stress the role of material interest as the driving force for the formation of non-state war actors, this article attempts to sketch an alternative explanation for their formation. Based on a data-set of 80 cases and comparative case discussion it focuses on the relationships that leaders and staff members of armed groups entertain before the actual formation of such groups. Three mechanisms of formation are consecutively distinguished, depending on the degree and kind of social relationships that precede their formation. The article discusses the causal settings of each mechanism and reveals in how far state policies in medium term and long run horizons produce what they want to curb: the formation of violent challenges.
How do armed groups come into being? The perilous acts of organizing a violent subversive group, the refusal of safer alternatives, the danger of being killed – all these risks tend to make the occurrence of armed groups rather unlikely. And while research on motives of recruits has produced various insights, the logic of the formation of armed groups remains somewhat enigmatic.Footnote1
In this article I will investigate the formation of armed groups by using the concept of figuration that was introduced into political sociology by Norbert Elias.Footnote2 The term shall designate all social settings, groups and less structured collectives that consist of interdependent individuals. In figurations single actors are linked by asymmetrical power balances, as they exchange favours or commodities, as they maintain emotional ties, and even as they fight. This way of conceiving armed groups has the advantage that it goes beyond methodological individualism, which has become the major obstacle to bridge the gap between large-N studies on political violence and field research-based studies of anthropologists or the work of historians. Also, it is open enough not to predetermine the kind of relations between the members of armed groups.
As this contribution argues, there are at least three mechanisms by which armed groups come into being. Based on a statistical overview, on single-case stories, and, most importantly, on systematic comparison, these three mechanisms also allow explanation of many features of the further trajectories of armed groups.Footnote3 First is the mechanism of repression. Violent repression exerted by government forces causes political opposition to evolve into armed action. Leaders of these groups are usually not militarily experienced but instead are politicians who have acquired their positions through descent, formal education, and long political activity. Groups that emerge from this mechanism become armed.
Second is the ad hoc mechanism. It is activated when neo-patrimonial settings experience crises. Individuals who feel excluded from clientelist networks of a political class begin to organize violent actions against state agencies. Groups formed through this mechanism are new creations that can include older modes of organization but have difficulty becoming stable due to the vagaries of war.
The third mechanism is often linked to situations of open political violence. This spin-off mechanism is tied to state policies, but its main characteristic is that the group's activities become free from state control. Originally, the formation of these groups is a state project. In times of war, governments or single-state agencies often employ informal, non-regular armed forces they can deploy for objectives that regular forces are unwilling or unable to achieve. In many cases, these informal troops are initially under government control but later develop a life of their own.
These mechanisms have been generated by systematic comparison between 80 cases described in the database. Although they do not cover all cases with the same precision, it is obvious that they delineate pathways that repeatedly appear in the empirical record. In the form they are presented here, they are similar to Weberian ideal types.Footnote4 In that sense they do not form a classical typology that would sort an empirical manifoldness into a complete list of mutually exclusive types. The claim connected with these three types is to discern three major mechanisms that lead to the creation of armed groups in the period after 1945. This means that there might be more such mechanisms, which could not be detected by the systematic comparison of the accounts of the 80 groups looked at here.
Furthermore, the three mechanisms do not exclude each other in the course of a war. Two or all three can occur within a social conflict at various points in time. Therefore, purpose of the mechanisms presented here is not to explain the entire variety of all cases but to determine those constellations of conditions under which the formation of armed groups is particularly likely to take place.
I will argue that these sets of conditions differ in many regards. But they overlap as well. It is not surprising to observe that the emergence of armed groups in the period under consideration here are closely related to the political dynamics of postcolonial states, especially when shrinking distributional capacities lead to exclusion. Secondly, the formation of an armed group is always a very internationalized process, as the role of exiled personnel and of foreign governments will show. Thirdly and most importantly, all mechanisms show that a decisive element in the process of formation is produced by states themselves. The production of violence expertise, of the capacity to use arms and to organize armed forces, is almost always learned within state institutions.
This last commonality links this article to several ongoing academic discussions. It might relativize the ‘newness’ of so-called new generations of warfare in which states allegedly play a less important role.Footnote5 And secondly, while these findings confirm the central role of state crises in the outbreak of civil war,Footnote6 it relativizes the ‘otherness’ of armed groups. To a large extent, political violence revolves around political rule, but it does not necessarily entail the end of stateness as a political form.
In the following presentations emblematic stories will be used to illuminate the causal relations at work in each of them, followed by a short discussion of similar cases in order to outline the consequences of these formative processes for the figuration's life. Depending on the mechanism of formation, organizational features of armed groups begin to differ as they build divergent hierarchies. There is also a relation between methods of funding their activities, and thirdly, practices of violence by armed groups seem to be related to the way they come into being.Footnote7
Where Leaders and Staff Originate – A Statistical Overview
Nobody rules alone, and only very simple forms of political domination can do without staff. Gerontocracy and primary patriarchal forms are among the only figurations in which specialized personnel is not required, as the rulers themselves suffice to enact and enforce their decisions. But any political setting in which social relations surpass immediate and face-to-face interaction needs some sort of support staff, that is people who reliably acquiesce to general procedures and obey explicit orders of political leaders. The staff is usually augmented by followers and lower ranks. The distinction between leaders, staff, and followers allows the development of rough initial ideas about the foundations and development of these internal relations within armed groups as figurations.
What can be seen from a rough statistical analysis using this distinction is, first, that states are deeply involved in the emergence and logic of armed groups. This becomes apparent when examining the most frequently shared biographical characteristics of leaders and staff. Secondly, the ties between leaders and staff members, with few exceptions, predate the onset of armed violence.
Personal motivations of course play an important role in the formation of armed groups, as they do in any political organization. The chance of financial gain or motives of affection certainly rank high for the explanation of why individual members join such a group. However, if such reasons alone form the motive for political personnel to follow commands, the figuration is rather instable.Footnote8 The core element of Weber's theory of domination calls for stabilization of relations between leaders and staff, and that constitutes the idea of legitimacy.Footnote9 Both within the figuration and in its external relations, insurgencies need to overcome the costs of violence which comprise first and foremost its delegitimizing effects. Armed groups, as I argue, need to develop forms of legitimacy in order to stabilize their organization and in order to survive the vicissitudes of war.
The argument here is not only that all rulers strive to rouse and groom belief in the legitimacy of their rule. For stable rule it is necessary that at least staff members believe in the legitimacy of the respective order. It is by the type of legitimacy that political forms can best be distinguished. This, as I will argue, also holds true for armed groups. Their inner functioning, their internal dynamics, and also their external behaviour can best be understood and explained by the role and quality of legitimization within the figuration. It has often been said that it is extremely difficult to empirically ascertain legitimacy. Empirically, hypocrisy can often be observed in hierarchical relations and might be wrongly taken for legitimacy. Also, personal interest, weakness, or helplessness prompt people to follow a given order. Indeed, all these motives are empirically observable in figurations such as those of armed groups. But no political figuration can endure for long without a certain degree of legitimacy, at least in the eyes of its staff. Furthermore, it is not probable that such weak motivations as indifference or helplessness could explain participation in an armed group in its early, most dangerous phases.
The empirical record shows that many armed groups who were able to fight in civil wars for extended periods had strong forms of inner legitimacy. It is likewise implausible to assume that a group could keep its organization intact through long periods of fighting connected with suffering, hardship, and huge costs in so many regards without such inner bonds.
The formation of an armed group is the process by which the kernel of a figuration is formed. As will be seen in the three mechanisms, this formation always takes place within a context that hands down a minimum of legitimacy to armed groups. There are always rules, relations, and meanings on which the formation's initiators can draw. This does not necessarily imply that the formation will be successful. An uncounted number of armed groups never surpassed their early stages. The pre-existence of these ties, however, means that there is a foundation on which to build first attempts toward legitimization. As will be seen in many examples, this inner legitimacy provides essential cohesiveness for the internal hierarchies of armed groups. Loss of legitimacy leads to a group's ultimate failure.
The form of legitimacy matters not only for the analytical purpose of distinguishing political organizations. It also reveals a great deal about the entire internal logic of any political organization. The main distinction here is between rule-based authority on the one hand and personal authority on the other. Whereas the former can be identified, for example, in bureaucratic forms of rule, the latter either involves charismatic legitimacy – the belief in a specifically exceptional power or qualities of the incumbent, or the belief in the sanctity of old forms and ties, that is on traditional legitimacy. All these forms of legitimacy, rule-based and personal ones, can be found in armed groups, although never in pure forms, but always blended to varying degrees. Nevertheless, the distinction not only allows tracking changes in the inner organization, it is also fundamental to discern varieties of their original formation.
One way to start the investigation of relations within a figuration is to look at numerical distributions of biographical experiences of leaders, staff, and followers. They indicate – of course very tentatively – some features of the processes of armed groups' formation. A first look at leaders of armed groups, for example, shows a surprising number of shared features in their biographies ().Footnote10
TABLE 1
BIOGRAPHICAL CHARACTERISTICS OF ARMED GROUP LEADERS
No specific profile can be deduced from this list. However, one can see that states, in many ways, are involved in the ‘production’ of the leaders of armed groups. It is supposedly within state institutions that core skills needed for armed rebellion are transmitted. The high percentage of academically educated, for example, suggests that skills such as abstract reasoning and knowledge of bureaucratic and organizational techniques are important preconditions for becoming the leader of an armed group. Also, some degree of military expertise, usually acquired in state institutions, is a characteristic of the profile of many armed group leaders. And finally, the experience of violence in state prisons and prior political conflicts also suggests a causal relation between encounters with state violence and the resort to arms as a political strategy. The data also leads to the assumption that conflicts in which armed groups are involved are indeed political. This is suggested by the high percentage of those leaders who have been active as political opponents before they became leaders of violent political groups. Furthermore, it is interesting to note that in opposition to what it is often stated in academic as well as the media discourses, few armed group leaders have criminal backgrounds (6.8 per cent) or religious education (9.6 per cent). The ‘crook’ and the ‘mullah’ do not rank prominently among those who instigate and organize armed opposition against regimes in power.
Further findings hint at relations between leaders and staff that are not purely instrumental. As indicated above, recent literature suggests that only interest in personal enrichment induces people to participate in armed rebellion. But greed is an insufficient explanation for what drives the formation of an armed group. Much more conclusive evidence about the mechanisms of how armed groups are formed comes from information on shared experiences of leaders and staff.
The empirical record shows that a very high number of staff members share experiences of political battles with their leaders. Many have the same ethnic background, and in an equal number of cases staff members and leaders know each other from attending the same educational institutions. In 50 out of 56 cases in which there is sufficient information on the matter, there is at least one example of ties between staff members and leaders such as a shared political past, a common ethnic background, or socialization at the same school. We know of 28 groups, out of a sample of 80, in which leaders and staff members had been active in the same political organizations before the armed group was formed. In 17 groups of the same sample, leaders and a number of staff members went to the same schools or universities, and there are also many cases in which both share the same ethnic background or even have family ties.Footnote11
The data discussed here is useful for preliminary speculation on how armed groups come into being. As such, however, they cannot reveal the processes by which formation of armed groups take place. Further evidence is needed for reconstruction of these evolutions. Based on the intense case studies and the comparative discussion of these findings with accounts on additional cases, it is possible to construct the three mechanisms of armed groups' formation discussed in the following.
These mechanisms do not cover all cases of formation. The claim here is that most cases can be explained by one of these mechanisms. The three mechanisms of armed group formation will be elucidated in the following by giving typical stories that will render their causal connections clearer. Evidence from further cases will then bolster the claim that formulation of these mechanisms is not merely an unjustified generalization of single instances.
From Party Politics to War: The Mechanism of Repression
During the 1970s, the years of explosive economic growth in the Philippines, the Muslim Independence Movement (MIM) emerged.Footnote12 The commercialization of agriculture, migration towards cities, and increasing levels of education all fuelled the political dynamics of which the formation of the MIM was a part. At about the same time, foreign actors began to engage with Philippine Muslims. Two hundred young men from the island of Mindanao were given grants by Gamel Abdul Nasser in Egypt, allowing them to study at the famous Al-Azar University in Cairo. However, students who acquired academic degrees, whether in Egypt or the Philippines, could not use them because the old elite was still controlling the access channels and excluded the ‘Moros’, the Spanish colonial term for Muslim Filipinos. The way up the social ladder was closed. As a consequence a political party was founded whose main goal was autonomy for the Muslim regions of the country. Due to a land shortage in the early 1970s, Christian gangs attacked Muslim farmers, as a consequence of which Muslim gangs formed and took revenge on Christians. The government reacted by declaring a state of martial law. This turned the political party MIM into a guerrilla movement called Moro National Liberation Front (MNLF). Its existing contacts in the Middle East alleviated its situation by quickly beginning to serve as channels for foreign support. Within months, hundreds of thousands fled from rural areas into cities. The war between government forces and the MNLF ended in 1976 when, in the Libyan capital Tripoli, a peace treaty was signed that contained provision for legal autonomy in the Muslim regions.
At the top of the Muslim opposition were men who started out as politicians. Nur Misuari, the first head of the MNLF, had been active in a communist youth group and studied political science. A typical politician turned guerrillero, his political skills were useful in the late 1970s for negotiation of a peace agreement. But as the new governor of Mindanao he was unable to prevent dissatisfaction among younger MNLF members and staff persons when implementation of the peace agreements failed.
Salamat Hashkim, the leader of the splinter group Moro Islamic Liberation Front (MILF), had once been Misuari's deputy and belonged to the group of Al-Azhar graduates. After his return he worked as a librarian but was also politically active organizing Muslim cultural circles in his home area. Through these activities in the late 1960s he came into contact with disgruntled Muslim politicians who had lost their status in the clientelist networks of the Philippine oligarchy.
Both leaders proved to be politically apt and able to negotiate complex treaties encompassing the regulations of autonomy for some southern regions of the Philippines. These treaties also encompassed the integration of thousands of fighters into the military and police forces. Despite multiple rounds of negotiations, the conflict was not brought to an end until 2006, partly because defecting factions such as the Abu Sayyaf Group (ASG) disrupted peace agreements and partly because internationalization of the war as a part of the post-9/11 ‘war on terror’ aggravated the conflict and led to further military escalation.
But it is not the possible ways of settling a war that are of interest here. It is rather the MILF's process of formation that reveals typical elements of the mechanism of repression. Governments in societies that have become unstable are overburdened with tasks that result from social change. If they represent an old oligarchy, these regimes lose their legitimacy in a rapidly changing social and political landscape.
In order to preserve their privileged positions, these regimes often resort to violent repression of their opponents. This measure usually has more unforeseen consequences than planned ones. It supports hardliners in oppositional parties and does not encourage attempts to find peaceful solutions. Political activists feel threatened and abstain from political activity or go underground. This renders non-violent politics even less likely, as could be seen in the case of the MIM, which became an armed liberation movement after state repression had begun.
The radicalization of existing parties is alleviated by the fact that state repression is often indiscriminate. As Stathis KalyvasFootnote13 has convincingly shown, this is always the case when information on opponents is scarce. This typically occurs when government troops are deployed in areas they are not familiar with and thus cannot tightly control. When rebellious forces offer at least partial protection, civilians shift their support towards them. This has been the case in Mindanao too. Indiscriminate government repression made the option of joining the rebellion a more rational alternative than flight and non-action. The young fighters, mostly between the ages of 15 and 25, were largely reporting that they joined the rebellion to defend themselves and their families against the Philippine government.
In contrast to armed groups that emerge as ad hoc formations, the MNLF could draw on pre-existing structures. The earlier political activity of the MIM had secured not only a broad network of politicians, it was also backed locally within the Muslim communities of Mindanao and other islands in the southern Philippines. The political experience of its leaders enabled it to hold a political line throughout the years of war, even if during negotiations and implementation of the agreements these structures proved too unstable to endure internal disputes resulting from disappointments and envy among its leaders and followers.
Armed groups formed by the mechanism of repression do not always fragment like the MNLF. Many succeed in keeping their organizational boundaries and build strong bonds of inner legitimacy. The Liberation Tigers of Tamil Eelam (LTTE) in Sri Lanka is a case in point. Having been formed in much the same manner as the MNLF, the LTTE did not falter but became one of the biggest and best organized armed groups with complex internal organization and strong transcontinental branches.Footnote14 Furthermore, most leftist guerrilla groups in Latin America have been shaped by this mechanism. More often than not they were under oligarchic rule with strong militarist traditions. Similarly, colonial rulers often reacted with repression against what were called at the time ‘liberation movements’. The Frente da Libertação de Moçambique (FRELIMO), the beginning of the anti-French Front de Liberation Nationale (FLN) in Algeria, and the foundation of the Viet Minh during Japanese occupation of French Indochina are identical in this regard.
The common features of armed groups formed through the mechanism of repression thus do not extend throughout their entire lifespan. Other mechanisms can alter their fate in many directions. The usual sequence of the repression mechanism of the formation of armed groups can be summarized as follows ().Footnote15
FIGURE 1 SEQUENCE OF THE REPRESSION MECHANISM OF THE FORMATION OF ARMED GROUPS
The fate of groups that emerge via this mechanism is not preordained. They seem, however, to be more successful in gaining political power than armed groups on average. This might have its basis in their social ties and organizational forms that exist prior to widespread armed conflict. Legitimacy is established before violence sets in. Furthermore, the practices of violence they exert differ from those employed by groups which come about through the two other mechanisms presented below. Also, these groups experience different forms of violence. More often than in the groups of the other two kinds, indiscriminate violent repression such as massacres and collective punishments continuously characterizes the conflicts in which these groups engage. Hence, the last stages of the mechanism are seen repeatedly during wars, constantly producing violent resistance.
The Ad Hoc Mechanism: Initiatives of the Disappointed
The National Patriotic Front of Liberia (NPFL) that pushed this West African country into a civil war, lasting with brief interruptions almost 15 years, was led by a bona fide opportunist.Footnote16 Charles Taylor, born in 1948 in marginal Nimba County, travelled to the United States in 1972 and earned a degree in economics. After returning to Liberia he became a senior official in the regime of Samuel Kanyon Doe. Taylor did not stay long in a government position. Just three years later, he was charged with embezzlement of US$900,000 and escaped to the United States.
Between 1986 and 1989 Taylor travelled throughout West Africa before finding in Blaise Compaoré and Houphouët Boigny, the heads of state of Burkina Faso and Côte d'Ivoire, respectively, two powerful supporters for his plans. They introduced him to Colonel Ghadaffi of Libya who provided training grounds and military expertise to aid formation of the first troops of the NPFL. Taylor also made contacts with businessmen from France and other foreign countries that desired access to Liberia's rich mineral deposits and other valuable natural resources.Footnote17
The main Liberian support Taylor had for his plans came from exiled politicians, driven out by Doe's tactics designed to protect his own power. This group, however, was not homogenous. Indeed a number of Taylor's early supporters, it became clear, had their own agendas. After a few months this led to the first divisions within the NPFL.
Among these exiled politicians was Moses Duopu, a former minister in Doe's government and main recruiter of the first round of members. Duopu had a joint past with Charles Taylor as student activist. Tom Woewiyu, a long-standing opponent of the Doe regime, was yet another old friend from student days and became the Defense Minister in Taylor's shadow cabinet. Both later challenged Taylor's position. Duopo was killed in June 1990, while Woewiyu survived.
Within six months, the NPFL was split into subgroups and it was only through general mistrust and ruthless internal violence that chains of command could be maintained. Although Taylor's plan had worked as intended in so far as indiscriminate violence by government troops drove thousands of fighters into his ranks, the ties within his ad hoc formation were not strong enough to turn this support into a coherent organization, or to contain the competing interests of its most powerful members.
The formation of the NPFL thus shows in an extreme manner the crucial problem of armed groups formed by the ad hoc mechanism. Typically, they consist of members of the political class who have been driven from power during severe crises in post-colonial regimes. Ad hoc groups are created exclusively for the purpose of taking power by force without relying on existing social or political organizations. They might try to borrow legitimacy from older traditions and to impress external observers by using satellite telephones and press officers, but such tactics cannot disguise their chronic instability.
In Weberian terminology, ad hoc groups are ‘voluntary associations based on self-interest’ (Zweckvereine)Footnote18 that is their inner logic is the shared assumption of its members that military success is necessary to seize the spoils of power. This shared interest is the weak bond holding the initial group together. Only the charisma of the leader would be a viable source of legitimacy that could overcome this fatal weakness. But in Taylor's case it was the intervention of West African states that prevented him from immediately conquering Liberian state power. The ultimate source of warrior charisma, sudden miraculous success, was thus beyond his reach.
Structures of these groups are chronically endangered as any change in the situation might give incentives for members to exit the shaky coalition. This is the primary reason ad hoc groups fragment as frequently as they do. All members who act with strategic views closely observe any change of options. They leave as soon as they have enough supporters to project a chance to get a bigger share through acting independently. Another problem of ad hoc groups is lack of internal control, leading to uncontrolled violence with its de-legitimizing effects on the group as a whole.
This fragility accompanies ad hoc formations from the very beginning. They have sharply limited time frames to organize a wide range of skills in order to take best advantage of their first military strikes. In the case of the NPFL, for example, it seemed vital that they took action without delay as the regime of President Doe was still cut off from official external support. According to unconfirmed reports, the International Monetary Fund and the World Bank were about to resume services to the Liberian regime. Such a step would have increased Doe's options considerably. The most important competencies required to organize this early phase are the ability to raise funds, to organize political support, and to build sufficient military strength. When these competencies cannot be combined in one person, as was the case with the NPFL, centrifugal tendencies are strong from the beginning. Taylor dispensed with the political connections to heads of states of neighbouring countries and had access to the informal business networks that ruled Liberia's economy. But for military expertise he had to rely on others.Footnote19
The ad hoc mechanism can be summarized as follows: when neo-patrimonial regimes – or any political systems in which clientelist networks structure political life – come under serious strain, members of the political class will be selectively barred from the spoils of power (). An increase of export profits or the loss of international monetary credits can lead to this effect.
FIGURE 2 THE AD HOC MECHANISM OF ARMED REBELLION
Those who are excluded tend to organize armed violence when they find propitious conditions, such as sanctuaries offered by neighbouring regimes willing to offer support. If the central figures are able to enlarge their group's competence by the addition of military expertise, the likelihood that exclusion from clientelist rule will eventually lead to armed opposition increases.
Armed groups produced by this mechanism often become structured along patron-client lines themselves. Two reasons account for this. First, their economy does not differ from the political economy of the system in which they operate. William Reno's account of Taylor's predecessor makes this point very clear: the economy of extraversion that underlay Doe's rule, this strategic mix of fees for business concessions and rents from plantations plus mineral exports, development aid, and political loans was and would continue to be the basis for the rule of Taylor or any other president of Liberia. Apart from the political loans, Taylor already had control of such an economy in his para-state, ‘Taylorland’.Footnote20 So even when ad hoc groups succeed, they are doomed to replicate the structures they fought against.
This in turn means that their future rule will be threatened by the same dangers they created for their predecessors. Only to the degree the leaders are able to form much deeper and more stable lines of allegiance in their newly conquered territory can they diminish the risk of being overrun by future challengers to their power. The need to create legitimacy first and foremost among the followers does not vanish with the success of these groups' leaders. It is merely postponed.
Furthermore, it is not accidental that ad hoc groups typically develop in settings where systemic change like decolonization has triggered a long-term process of appropriation of inherited institutions, which defines the post-colonial situation. The states of sub-Saharan Africa are one main theatre of these processes. Charles Taylor's NPFL in Liberia and Foday Sankoh's Revolutionary United Front (RUF) in Sierra Leone are not the only salient African cases. Laurent Kabila's Alliance des Forces Démocratiques pour la Libération du Congo (AFDL)Footnote21 follows much the same pattern. Supported by governments of neighbouring states, single particularly ambitious politicians organized exiled members of the political class and enlarged that kernel by including further personnel with military expertise.Footnote22
Intellectuals often play a prominent role in ad hoc groups, but also in armed groups that come about by the mechanism of repression. This has much to do with the close relationship universities maintain with the political field throughout the world. Academic institutions not only reproduce a state-class of bureaucrats and technocrats. Especially in times of political crises and growing social contradictions, they are also hotbeds of political opposition. As Derluguian pointed out,Footnote23 what matters are not only the academic skills, such as the ability to formulate political programs, to address a Western public in a foreign language, or to set up an organization. Of equal importance are the contacts, the social capital, accumulated during the long years in the cultural milieus that are part of the individual's resources. Skills, cultural capital, contacts, and social capital are necessary for successful leaders of armed groups.
In the faltering years of the Soviet Union, such intellectual networks often merged with violent entrepreneurs chasing opportunities in weakened states, and sub-proletarian militancy rendered violent escalation in these situations more likely. That the ad hoc mechanism is not restricted to the post-Cold War period can be seen by a number of further cases in which the mechanism occurred in almost its ideal-typical form. A perfect example is Uganda's National Resistance Movement (NRM), formed by the current president of Uganda, Yoweri Museveni, when he was defeated in contested elections in 1980. The NRM also started with a few dozen fighters in February 1981, who were soon joined by those who were, or feared becoming, victims of indiscriminate repression by government forces. At the end of the year 1981, the NRA already had 1,000 members and began to reject volunteers since they did not possess enough guns. Initially formed as a group of left-wing oppositionists in neighbouring Tanzania, some of the NRA's leading members had learned guerrilla tactics in camps of the FRELIMO, which at the time was fighting Portuguese colonial troops in Mozambique. But apart from a rather rhetorical leftist jargon and some signs of Maoist guerrilla strategy, the NRA was not determined by the usual Cold War tactics of the time.Footnote24
Armed groups that come into being by the ad hoc mechanism are thus not a product of narrow periods of time. There are structural histories behind their emergence, and these structures extend to the inner workings of these groups. They rely on inter-state rivalries, and they usually appear in patrimonial regimes. For their emergence this institutional setting is causally much more important than the fluctuations of global conflict. The same applies, it seems, to the third mechanism.
Delegated Violence: The Spin-Off Mechanism
One of the most renowned militias of the former Yugoslavia was clearly a state creation. It was the Srpska Dobrovoljačka Garda (SDG, Serbian Volunteer Guard), led by Željko Raznatović, also known as Arkan. Raznatović was born in 1952 as the son of an air force officer. He was a problematic child, and his father had reportedly approached the secret services to request they assume care of his son, who then became a member of secret units of various federal agencies. In the 1970s, Arkan operated mostly abroad and had arrest warrants in several Western European countries. He escaped from prison several times, for example in Belgium and the Netherlands, allegedly with the help of Yugoslavian secret services. Throughout the 1970s and 1980s he was almost certainly under the protection of these secret agencies.
In 1990 Arkan was again arrested, this time by local police forces in Croatia for having smuggled arms into the Krajina region. Once free, he founded his Serbian Volunteers Guard. Much earlier, Arkan had become a member of the fan club Delije (Heroes) of Belgrade's famous soccer club, Red Star. This job was also reportedly undertaken on behalf of the Ministry of the Interior, as the members of this club had become hooligans and were seen as a problem for public order. Arkan, who at the time was officially merely the owner of a bakery, invited fans on tours in other countries whenever Red Star played matches in the European leagues, and thus he soon became a hero among the Heroes. Most members of this fan club seem to have come from the deprived housing areas of Belgrade where the social contradictions of socialist modernization were particularly hard felt. High unemployment rates among youth, declining household incomes, and a constantly deteriorating supply of staple goods marked life in the areas of Novi Beograd. But when the SDG was founded in 1991, recruitment for Arkan's guard was not restricted to socially deprived youths, but included people of all ranges of age and education.Footnote25
The Serbian Ministry of the Interior provided Arkan's militia a training camp in Erdut in East Slavonia. Certainly with its consent, if not on its order, the SDG participated in the siege and capture of various cities in Slavonia and Bosnia and soon gained a reputation for particularly brutal treatment of non-Serbian civiliansFootnote26 as one of the main instruments of the policy of ethnic cleansing. Among Serbian paramilitaries the SDG was well-equipped with light artillery, trucks, and arms from barracks, including tanks according to one source. Although there is divergent information concerning estimates of its size, ranging between three and ten thousand, it seems likely that altogether around 10,0000 men underwent training in Erdut,Footnote27 and that in most cases war participation was not longer than one or two years, while many members stayed only a couple of months.Footnote28
During the years of war and particularly after the Dayton Agreement in 1995, Arkan had tried to diversify his position. Erdut, the location of the SDG's training camp, became the brand name of a wine company owned by Arkan, and he became the owner of a shipping company, a radio station, and the casino in the Hotel Yugoslavia. His main economic activity, however, developed under the embargo during the war years. Apparently in growing competition with Milosevic's son, Marko, Arkan tried to monopolize the illegal import of petrol and derivatives from Romania and Bulgaria. Other sources also report his activity in further criminal markets, partly in cooperation and partly in competition with Kosovo-Albanian networks.
Arkan increasingly became a political entrepreneur, making efforts to improve his parliamentary career. His Party for Serbian Unity (Stranka srpska jedintsva, SJS), founded in 1992, however, did not fare well, and he failed to gain a seat in the Serbian Parliament after 1993. Three years later Arkan became the owner of Serbia's first league soccer club, FK Obilić. Politically, Arkan had failed, and his erstwhile power base was eroded after the end of war in Bosnia-Herzegovina.
With a pending international arrest warrant and growing competition from rival mafia networks with closer contacts to Serbian authorities, the lives of Arkan and his wife were almost restricted to his fortified house and the lounges of Belgrade's international hotels. In January 2000 he was killed by two men in the foyer of Belgrade's Hotel Intercontinental.
The trajectory of the SDG is typical of armed groups that start out as state militias and become independent of their sponsors. Seemingly, one condition for the fact that many delegated violent organizations develop a life of their own is that the political situation becomes murky, as was the case in Serbia during the 1990s. Another is that leaders of these groups have political ambitions and use the power they assemble as a stepping-stone to an independent political career.Footnote29
The desire for personal enrichment, whether by legal or illegal means, was definitely a strong character trait. But even Arkan had shown another face. Many of his actions reveal that his ambition was not just to enrich himself and to prove his masculinity in combat. His eagerness to appear in the media, as well as his attempt to build up political support, suggest further motivations. Arkan's political career, however, was not successful, and his appearances as a statesman might have been instrumental for opportunities that were accessible only through public positions. But his desire for public recognition cannot be overlooked. He succeeded in this regard insofar as his image among militaristic nationalist circles comes close to that of a hajduk, those outlaws who served, during the long periods of foreign rule in the Balkans, as a positive role model for social bandits.Footnote30
Whereas spin-off groups such as that of Arkan are not without imprints of local historical traditions and political cultures, the main mechanism that leads to their creation is orchestrated by states. In situations in which governments feel they cannot exclusively rely on their army, they tend to either tolerate or deliberately create other informal armed forces. One reason for this can be that the chain of control is interrupted due to deep political crises and stalemate situations, as was the case in Lebanon in the mid 1970s. Parts of the armed forces then become focal points for the creation of militias.
Another reason, probably more often the case, is the deliberate decision of governments or single state agencies to create a second layer of organized violence because the regular forces are seen as either not sufficiently effective or not trustworthy to undertake missions which are in violation of international ethical standards. Perhaps for this reason, militias and paramilitary troops created in this manner are particularly prone to commit severe human rights violations in civil wars: they feel legitimized to cross moral boundaries since they know they have state backing, and these transgressions are seldom if ever sanctioned.
Very often, regular forces are not trusted to be ruthless enough to commit these deeds, given their ethos as professional soldiers. The creation of paramilitary forces is thus always the result of individuals' ambitions and strategic decisions within state organs. Exactly how these causal moments mix in each case is a question that can only be answered by empirical investigation. Nevertheless, judging from the similarities between cases, the main elements of the spin-off mechanism can be summarized as follows ().Footnote31
FIGURE 3 SPIN-OFF MECHANISM OF GROUP FORMATION
Spin-off groups form in wars or war-like situations. Politically motivated violence is already openly visible, and for individual reasons or in order to support the official state forces, the commander of the state military delegates the right to exert violence to newly created groups. When these groups are successful in the sense that they form, organize, and succeed in combat, their leaders potentially accumulate enough warrior charisma and resources to sever the chain of command to official superiors. The newly formed group then gains its own momentum.Footnote32
Spin-off groups usually encounter several problems not shared with groups emerging through the mechanisms outlined above. Spin-off groups typically do not stand in an openly hostile relation to state forces, although the relation is never an easy one. Many officers of the Yugoslavian army (JNA) detested the militias created by the Serbian Ministry of the Interior.Footnote33 The ‘Cossacks’ who participated as informal troops in the short war that led to the separation of the Transdniestrian Republic from Moldova in 1992 were viewed with contempt by Alexander Lebed, the commanding general of the 14th Army, then stationed in Tiraspol.Footnote34 The uneasy relationship between spin-off groups and formal state forces is somewhat ameliorated by personal relations and the fact that they share elements of the same military habitus.
Secondly, it is typical for the organizational development of spin-off groups that they desperately lack popular support. From a sample of nine such groups only one reportedly enjoyed considerable popular support, whereas for the overall sample of armed groups popular support was reported to be more than 50 per cent.
Thirdly, spin-off groups are much more inclined to use ruthless violence with resultant high costs to their legitimacy. This might be related to the fact that their staff usually includes high numbers of military personnel. Among the nine cases investigated more closely, six leaders have a military education, whereas the overall rate for the entire sample is 43 per cent.Footnote35 Also, in six of the nine cases at least some staff members came directly from positions in state armed forces. In the entire sample of 80 armed groups, this applies to only one-third of them. Although spin-off groups become increasingly independent from state control, they might still feel legitimated by the power originally invested in them by state agencies.
Like the other two mechanisms by which armed groups form, the spin-off mechanism is not inevitable. It can stop at any point if conditions for its unfolding are not sufficiently developed. Most of these groups do not survive the regime that has produced them as they are unable to attract sufficient support with their generally vigilante and nationalist rhetoric. Extreme violent practices characterize them in comparison to the two other types. Their programs and rhetoric seems to be overshadowed by their violent practices.
Conclusions
The three mechanisms of formation outlined produce slightly different pathways for the future development of the armed figurations they bring about. And although these mechanisms do not determine the odds of these figurations to institutionalize and turn their violent power into domination, the divergence of structural conditions nevertheless accounts for different probabilities.Footnote36
Groups that suffer excessive violence from repressive regimes seemingly do not suffer noticeably from legitimacy deficits. Repressive states de-legitimize themselves, and in turn, targeted groups can garner popular support, especially if they provide effective protection. These groups also generally benefit from social ties and legitimate forms of organization that precede the outbreak of violence. Given their relatively strong legitimacy base right from the start, groups that come about by the mechanism of repression have much better chances to survive and to be ultimately successful than the two other types.
Ad hoc groups, in contrast, usually have weaker ties at the beginning. They consist of connections that are products of circumstances rather than relations cultivated over time. Consequently, their internal functioning is precarious. Shared interest alone does not suffice to create stable organization, and ad hoc groups are therefore more prone to fragmentation and decay. In propitious settings, such as strong support by other states, they can institutionalize and defeat government armies.
Spin-off groups typically consist of organized kernels, relying on state resources during their beginnings, accompanied by respective organizational capacities. While internal hierarchies initially remain uncontested, spin-off groups have enormous problems overcoming the de-legitimizing effects of the massive violence they often inflict. Their retrogressive discourse is seldom able to raise huge popular support. Their fate depends on the abilities of post-war states to reintegrate them into armed forces. Often they linger on in ambivalent relation to state armed forces.
Despite their differing outcomes, these three mechanisms as processes have many features in common. All three processes of formation of armed groups have a reverse side. They are always to some degree internationalized. Other states are often involved. The experts in violence have partly acquired their skills in remote institutions. Political ideas around which a group's program is centred have a long-standing international history, and even the very act of founding an armed group sometimes takes place on another continent or in another country.
However, the formation of armed groups always takes place in a local arena, as large and far-reaching as the internationalization of these processes might be. While the political project might be connected to events in other states, it is first and foremost aiming at political change in single states. It is, as was shown, not satisfying to point to particular features in the political economy or to material interest to explain the formation of armed groups. The more challenging question is how and under which conditions these groups form. The three mechanisms distinguished share the common feature that these formations often occur in critical situations of post-colonial states, in either a crisis of distribution due to a shortage of resources or in a crisis in which exclusion and political violence already play a role. These situations can be seen as culminating phases of deeper structural change, and through the three mechanisms it is thus possible to say more about the situational and structural background of the formation processes.
It also has become clear that the formation of armed groups is genuine social action in the sense that it has identifiable actors who act intentionally. There is always somebody who makes decisions regarding the creation of armed organizations, be it in reaction to state repression, to being driven from the ruling class, or in a situation perceived as desperate enough to permit the creation of unlawful armed forces. The resort to arms, as legitimate as it may seem to the individuals involved, is always based on a decision. It does not happen without it.
A third commonality concerns expertise in exerting and organizing violence. It is overwhelmingly within state institutions that future insurgents learn how to fight by military means. Rough aggregated data alluded to the state itself as a main source of this capacity. The evidence given in the exemplary stories delivered further evidence for this interpretation. Particularly those armed groups that originate through processes resembling the second or third mechanism sketched above involve personnel who have acquired their knowledge of how to use arms and how to organize armed forces in state institutions.
Fourthly, the formation of armed groups seems to be bound to pre-existing milieus, locales, and micro-arenas. When studying the accounts of various armed groups it is astonishing to see that inevitably there is a structure from which groups develop. Oppositional milieus, universities and armies, prisons and schools appear to be the institutional settings in which focal points crystallize that start the formation of armed groups. The reason for this is certainly the simple organizational requirement that shared interest does not suffice for an organization to emerge. There must be other social ties that allow for the aggregation of interest. This is even more evident when high politics, such as incumbency of power positions, is at stake, and when the outcome is settled by violent means.
Milieus and institutions are necessary as it is within them that focal points occur that serve as the basis of armed groups. At the centre of these figurations are ties that stem from shared experiences such as many years in armies, schools, prisons, or in exile, of having seen the same things and undergone the same or very similar experiences.
These findings might be taken as starting points for further questions. One is to look back at the usual distinction between state armed forces and non-state armed groups. A deeper investigation than the one delivered here might reveal that violent challenges to incumbent regimes in numerous if not most cases are closely related to the failure of regimes to provide enough space for political change. Furthermore, capacities created by states do in numerous regards enhance the chances of armed groups. This applies to soft skills as well as to military expertise and hardware produced by states and used by armed groups. States seem to be confronted with the challenge to control their violent apparatus in the long run: it seems that the very means to counter violent challenges lay at the basis of future formations of armed groups. The dilemma of how to control means of violence without producing even more remains unresolved. But this insight may raise scepticism about current attempts to render the world more secure by creating greater expertise in the exertion of violence.
Notes
Two factors have been highlighted by this research, namely material interest and coercion. This explanation is certainly not covering all relevant forms of motivation that are empirically never pure. Interviews with war veterans always reveal a complex mix of motives and even participation not even based on real decision. The more recent literature is discussed in Scott Gates, ‘Recruitment and Allegiance. The Microfoundations of Rebellion’, Journal of Conflict Resolution, Vol. 46, No. 1 (Spring 2001), pp. 111–30. Macartan Humphreys and Jeremy Weinstein, ‘Handling and Manhandling Civilians in Civil War’, American Political Science Review, Vol. 100, No. 3, (Fall 2006), pp. 429–47, deal with the consequences of recruitment based on material incentives for practices of violence. Paul Richards, Fighting for the Rain Forest. War, Youth and Resources in Sierra Leone (London: Heinemann, 1996) has challenged the presupposition of mere material interest, implying that the logics of exclusion matter much more. Thomas McKenna, Muslim Rulers and Rebels. Everyday Politics and Armed Separatism in the Southern Philippines (Berkeley: University of California Press, 1999) stressed the role of immediate security needs.
Cf. Norbert Elias, Die höfische Gesellschaft (Frankfurt am Main: Suhrkamp, 1983).
This contribution is based on research that was carried out on a grant from Volkswagen-Foundation between 2001 and 2007 (Project title: ‘Nachwuchsgruppe Mikropolitik bewaffneter Gruppen’, see www2.rz.hu-berlin.de/mikropolitik). The author is grateful for this support and for the discussions with Astrid Nissen, Katrin Radtke, Jago Salmon, Daria Isachenko, Alex Veit, Stefan Malthaner, and Teresa Koloma Beck on whose research this paper also relies. In this research project in-depth analysis, including field research, was carried out in 14 countries. Hypotheses and theses of these theory-driven studies were tested against a dataset on 80 armed groups, further on referred to as the MAG database. The construction of the mechanisms sketched here has been based on the comparative, reiterative discussion of cases within that research group.
See Max Weber, Gesammelte Aufsätze zur Wissenschaftslehre, 5th ed. (Tübingen: Mohr, 1988) on the methodology of creating ideal types and their heuristic function, namely to allow for differentiation of empirical observation and to discuss thesis on chains of causalities.
On this discussion see Aaron Karp's contribution in this special issue.
See the contribution of Diane E. Davis in this issue, and Klaus Schlichte (ed.), The Dynamics of States. The Formation and Crises of State Domination (Aldershot: Ashgate, 2005).
On these relations to other characteristics of armed groups cf. Klaus Schlichte, In the Shadow of Violence. The Politics of Armed Groups (Chicago, IL: Chicago University Press, 2009).
‘But custom, personal advantage, purely affectual or ideal motives of solidarity, do not form a sufficiently reliable basis for a given domination. In addition there is normally a further element, the belief in legitimacy’. Max Weber, Economy and Society. An Outline of Interpretive Sociology (Berkeley: University of California Press, 1978), Vol. 1, p. 213.
Legitimacy is a contested concept in social sciences. I follow here the definition that is fundamental to Weber's political sociology, namely to conceive it as the belief in the ‘moral authoritativeness’ of a given order. Weber, Economy and Society, (note 4), p. 15. On the role of forms of legitimacy in the life of armed groups, especially of ‘charismatic ideas’, see Schlichte, In the Shadow of Violence (note 7), ch. 3.
As mentioned above the database from which this information is taken had started with 50 rough sketches of armed groups in order to identify important aspects of actors (leader, staff, follower), organizational aspects (form, agenda, life-story), and practices (funding, violence). Sources for this information have been accounts by eyewitnesses or participants gathered during field research, press reports in several languages and other secondary sources like academic case literature. The most frequently occurring qualities were then systematically investigated in a sample of 80 cases. Although this enlarged sample roughly followed the regional distribution of internal warfare after 1945, preference was given for pragmatic reasons to well-documented cases. For more detail on the database cf. Stefan Malthaner, ‘The Armed Groups Database: Aims, Sources and Methodology’, 2007, Available at http://www.ipw.ovgu.de/publikationen/inhalt/publikationen_der_mitarbeiter.html (accessed 7 July 2009).
MAG database (note 3).
The following account largely follows the impressive study of McKenna, Muslim Rulers and Rebels (note 1). See also Nikki Rivera Gomez, Coffee and Dreams on a Late Afternoon: Tales of Despair and Deliverance in Mindanao (Quezon City: University of the Philippines Press, 2005) and John T. Sidel, Capital, Coercion and Crime: Bossism in the Philippines (Palo Alto, CA: Stanford University Press, 1999) on political structures in the Philippines.
Stathis Kalyvas, The Logic of Violence in Civil War (Cambridge: Cambridge University Press, 2006), ch. 6.
Katrin Radke, ‘From gift to Taxes. The Mobilization of Tamil and Eritrean Diaspara in Infrastate Warfare’, Working Paper Micropolitics 2/2006, Humboldt University Berlin, 2006.
It is impossible to list all the groups whose emergence follows the pattern of the mechanism of repression rather than the two other patterns. Cases include the National Democratic Front of Bodoland, see Sudhir Jacob George, ‘The Bodo Movement in Assam’, Asian Survey, Vol. 34, No. 10 (October 1994), pp. 878–92; the Sudan Liberation Army in Darfur, see Gerard Prunier, Darfur. The Ambiguous Genocide (Ithaca, NY: Cornell University Press, 2005); the Maoist guerrillas in Nepal, see Philippe Ramirez, ‘Maoism in Nepal’ in Michael Hutt (ed.) Himalayan ‘People's War’. Nepal's Maoist Rebellion (London: Hurst, 2004), pp. 225–42; the Tigray Movement in Ethiopia, see John Young, Peasant Revolution in Ethiopia. The Tigray People's Liberation Front 1975–1991 (Cambridge: Cambridge University Press, 1997) and a number of groups in Latin America. The Frente Sandinista de Liberación Nacional (FSLN) in Nicaragua and the Frente Farabundo Martí de Liberación Nacional (FMLN) in El Salvador are classic examples of this, see Cynthia McClintock, Revolutionary Movements in Latin America. El Salvador's FMLN & Peru's Shining Path (Washington, DC: United States Institute of Peace Press, 1998).
An extensive account of the formation of the NPFL is given in Stephen Ellis, The Mask of Anarchy. The Destruction of Liberia and the Religious Dimension of an African Civil War ( London: Hurst, 1998). A shorter version can be found in William Reno, Warlord Politics and African States (Boulder, CO: Lynne Rienner, 1998), pp. 91–5, laying more stress on Taylor's economic motivations. On the later development of this business network see François Prkic, ‘The Phoenix State: War Economy and State Formation in Liberia’, in Schlichte, The Dynamics of States (note 6).
Cf. Prkic, The Phoenix State (note 15). The formation of armed groups in such an international space is nothing unusual. The experience of being exiled, apparently, is strongly connected with the emergence of armed groups. Examples abound and evidence is also given by the high percentage of time spent in exile by both leaders and staff. Hypotheses as to why this is so important might be built on lack of integration in host countries, on formal education and politicization abroad, or on social ties growing stronger between fellow countrymen living in a culturally distinctive environment. The formation of first organizational kernels in exile is however not connected to ad hoc groups. The first steps in the formation of Eritrean resistance also took place in Sudan, cf. John Markakis, National and Class Conflict in the Horn of Africa (Cambridge: Cambridge University Press, 1987), p. 107.
Cf. Weber, Economy and Society (note 8), p. 41.
In anthropological terminology, these armed groups start out as ‘bands’, cf. Eugene V. Walter, Terror and Resistance. A Study of Political Violence (New York: Oxford University Press, 1969), p. 57. A counter-example to Taylor and a much more successful case is Uganda's current president, Yoweri Museveni, who controlled external political connections at the beginning of the National Resistance Army (NRA)'s rebellion and could also credibly present himself as the military leader of the rebellion. See his own biographical account: Yoweri Museveni, Sowing the Mustard Seed. The Struggle for Freedom and Democracy in Uganda (London: Macmillan, 1997).
Cf. Reno, Warlord Politics and African States (note 16).
The history of the RUF and of other armed groups that follow this pattern is given in Christopher Clapham (ed.), African Guerrillas (Oxford: Oxford University Press, 1998). On the formation of the AFDL cf. Erik Kennes, Essai biographique sur Laurent Desiré Kabila (Paris: L'Harmattan, 2003), pp. 218–21.
There are striking similarities with processes that developed in states of the former Soviet Union in the course of its dissolution. Here again, the appropriation of new states included violent confrontations between armed groups that formed almost randomly around certain persons. Many of them did not have sufficient success to receive attention from Western media, as Georgi Derlugian, Bourdieu's Secret Admirer in the Caucasus. A World-System Biography (Chicago, IL: Chicago University Press, 2005) and Valery Tishkov, Chechnya: Life an a War-Torn Society (Berkeley: University of California Press, 2004) have demonstrated in the Northern Caucasus.
Cf. Derluguian, Bourdieu's Secret Admirer (note 22), p. 61.
On the interpretation of the strategy of the NRA see Museveni, Sowing the Mustard Seed (note 19), and Frank Schubert, “‘Guerrillas Don't Die Easily’: Everyday life in Wartime and the Guerrilla Myth in the National Resistance Army in Uganda, 1981–1986”, International Review for Social History, Vol. 51, No. 1 (2006), pp. 93–111, and Jeremy Weinstein, Inside Rebellion: The Politics of Insurgent Violence (Cambridge: Cambridge University Press, 2007).
Author's interviews with war veterans in Belgrade, conducted between March 2003 and October 2005. On the documentation of the veterans of Belgrade's quarter Rakovica see Milislav Sekulic, Na krilima patriotisma. Borci Rakkovice u ratovima od 1990 do 1999 (Belgrade: self-published, 2001).
Laura Silber and Allan Little, Yugoslavia. The Death of a Nation (London: Penguin, 1996), pp. 85, 222.
In comparative regard, it is one of the particularities of the wars in Yugoslavia that the country had compulsory military service and regular military training so that almost the entire adult male population had intensive military education. The strong militarist tradition relates to the Partisan's war victory as the foundational myth of Yugoslavia and with the security situation of Yugoslavia during the Cold War as it was perceived after the Soviet invasion of Czechoslovakia in 1968. See Robin Alison Remington, ‘State Cohesion and the Military’ in Melissa K. Bokovoy (ed.), State-Society Relations in Yugoslavia, 1945–1992, (New York: St. Martin's Press, 1997), pp. 61–78. This might be seen as another indication for the fuzziness of the state/non-state distinction in the study of armed groups.
For more on the trajectories of Serbian paramilitaries see Klaus Schlichte, ‘Na krilima patriosma – On the Wings of Patriotism. Delegated and Spin-Off Violence in Serbia’, Armed Forces and Society, 2009 (forthcoming).
The trajectories of militia leaders in Serbia vary however, as they do in other cases. Cf. Jago Salmon, Militia Politics. The Formation and Organization of Irregular Armed Forces in Sudan and Lebanon, PhD thesis, Humboldt University Berlin, 2006, (http://edoc.hu-berlin.de/dissertationen/salmon-jago-2006-07-18/PDF/salmon.pdf) and Schlichte, ‘Na krilima patriotisma’ (note 28).
On the role played by the ‘Hajduck’ image in the politics of militias during the wars of Yugoslavia see John Allcock, Explaining Yugoslavia (New York: Columbia University Press, 2000) pp. 390–95. The seminal reading on social bandits is Eric Hobsbawm, Primitive Rebels. Studies in the Archaic forms of Social Movements in the 19th and 20th Century (New York: Norton, 1959).
I am drawing here again on the comparative study of Salmon, Militia Politics (note 29), p. 96.
The spin-off groups included in the sample are: the South Lebanese Army (SLA) in Lebanon, see Jürgen Endres, Wirtschaftliches Handeln im Krieg. Zur Persistenz des Milizsystems im Libanon (Wiesbaden: VS-Verlag, 2004); Interahamwe in Ruanda, see Gérard Prunier, Rwanda. History of a Genocide (London: Hurst, 1998); the Hrvastko vijece obrane (HVO) in Croatia, see Steven L. Burg and Paul S. Shoup, The War in Bosnia-Herzegovina. Ethnic Conflict and International Intervention (Armonk, NY: M.E. Sharpe, 1999); the Mchedrioni and the National Guard in Georgia, see Thornike Gordadze, ‘Les nouvelles guerres du Caucase (1991–2000) et la formation des Etats post-communistes’, in Roland Marchal and Pierre Hassner, Guerres et sociétés. Etat et violence après la guerre froide (Paris: Karthala, 2003), pp. 371–402; the Shan United Army (SUA) in Burma, see Alfred W. McCoy, ‘Requiem for a Drug Lord: State and Commodity in the Career of Khun Sa’, in Josiah McC. Heyman, 8th ed., States and Illegal Practices (Oxford: Berg, 1999); and Dostum's militia in Afghanistan, see Antonia Giustozzi, Respectable Warlords? The Politics of State-Building in post-Taleban Afghanistan (London: Working Paper Series, Crisis State Programme, LSE, 2003).
Interviews with ex-JNA officers in Belgrade March 2003, October 2005.
Anatol Lieven, Chechnya: Tombstone of Russian Power (New Haven, NJ: Yale University Press, 1999), p. 244.
MAG database (note 3).
For a comprehensive discussion on these dynamics see Klaus Schlichte, In the Shadow of Violence. The Politics of Armed Groups (Frankfurt A.M.: Campus/Chicago: Chicago University Press, 2009, forthcoming).