LIBERIAN WARTIME COMMAND STRUCTURES

The first central assumption in past literature is that formal wartime command structures continue as informal postwar ex-combatant networks. For instance, Bjarnesen (Reference Bjarnesen2020: 9) writes that ‘a postwar rebel network is … an overarching constellation of ex-combatants who have preserved or established links to one another in a patron-client manner, based on wartime structures’. Themnér (Reference Themnér2018: 46, 60) in turn equates command structures with ex-combatant networks and argues that ‘despite large-scale efforts to break up the command-and-control structures of armed groups by investing in disarmament, demobilisation, and reintegration programmes (DDR), ex-commanders often continue to have influence over their ex-fighters’. References to demobilisation programmes and common mentions of lingering chains of command imply at least the partial continuation of wartime organisations in more informal guises in the post-conflict. Following the idea that command denotes authority to make decisions (King Reference King2019), these enduring chains of command allow former commanders a varying degree of control over their wartime subordinates long after the end of hostilities (Themnér Reference Themnér2018: 46; for a conflicting view, see Utas Reference Utas2012: 8; for the original argument immediately after war, see Hoffman Reference Hoffman2007a).

While much past literature agrees on the continued importance of wartime structures, it is remarkable that these wartime structures are rarely discussed in detail. One of the notable exceptions comes from Hoffman, who acted as an expert witness for the defence in the Special Court for Sierra Leone. According to him, the Sierra Leonean Kamajor militias – some of whom later crossed to Liberia to fight in the rebel group Liberians United for Reconciliation and Democracy (LURD) against the former rebel leader Charles Taylor's Government of Liberia (GoL) – should not be ‘considered a military organization with a system of military command and control’ but rather ‘as the militarization of a web of social relations’ (Hoffman Reference Hoffman2007b: 639–40). At stake was responsibility for war crimes. If there was no clear chain of command, it would be unreasonable to hold elites accountable for what happened on the ground. The prosecution instead relied on a British officer, who observed the Kamajors through what Hoffman calls ‘conventional wisdom’ about violence and war. In effect, the officer saw the militia as comparable to Western armed forces (Hoffman Reference Hoffman2007b: 641; for a similar view on the LURD, see Bjarnesen Reference Bjarnesen2020: 73).

Hoffman thus offers two opposing ideal types of wartime command structures: the first ideal type presents a temporarily militarised patronage network, the second a Western-type rigid military organisation. The literature on Liberian former combatant networks that assumes the endurance of wartime command relationships can by and large be divided between these two ideal types. The remainder of this section focuses on the small groups consisting of frontline commanders and their former subordinates, while the competing understanding of a militarised patronage network is discussed in the section that follows.

An early view of Liberian ‘rebel armies’ in rural contexts described them as ‘flat structures’ where everyone began ‘from zero in military rank and was subsequently promoted’. The end result was closer to ‘egalitarian’ football teams than ‘the hierarchical image of military organisation’ (Utas Reference Utas2003: 161–3). This implies that the warring factions were new constellations, which did not exist before the war. The literature on networks of former commanders and their subordinates in turn has equated these constellations with the tight and often coercive patrimonial relationships between a commander and their clients (Murphy Reference Murphy2003; Hoffman Reference Hoffman2007b: 652; Podder Reference Podder, Özerdem and Podder2011). These kinds of small tactical-level units with personal trust-based relationships are in military sociology called primary groups, where bonding between soldiers explains why they continue to fight even when the odds have turned against them (Shils & Janowitz Reference Shils and Janowitz1948). In a similar manner, even the members of former combatant networks are perceived to have developed ‘military comradeship that often survives demobilization’ (Themnér Reference Themnér and Utas2012: 210), especially after having ‘fought side by side during war’ (Themnér & Karlén Reference Themnér and Karlén2020: 5).

While the Liberian factions no doubt constituted primary groups, it is nevertheless important to recognise that the kinds of military primary groups investigated in the West have existed as a part of a well-developed and specific institutional context. As the pioneering study of Shils and Janowitz already made clear, these groups functioned when ‘the individual's immediate group, and its supporting formations, met his basic organic needs, offered him affection and esteem from both officers and comrades, supplied him with a sense of power and adequately regulated his relations with authority’ (Shils & Janowitz Reference Shils and Janowitz1948: 281). The institutional context in Liberia was and remains very different from those typically studied in military sociology. This makes it problematic to assume the existence of similar kinds of cohesive groups in Liberia during war, let alone after (Käihkö Reference Käihkö2018a).

Liberian former combatants with past military experience readily recognise that the armed groups they belonged to were ad hoc constructs in comparison to the Armed Forces of Liberia (Käihkö Reference Käihkö2015, Reference Käihkö, Haldén and Jackson2016b), itself not very organised in comparison to modern Western militaries. Overall, it needs to be recognised that the concept ‘military’ denotes a specific relationship between a modern state and a bureaucracy dedicated to upholding the monopoly of (typically, but not exclusively, external) violence. This regulated relationship offers a source of legitimacy and esteem, while the organisation and its relationship with overarching state structures helps to meet basic needs of its members. Yet bureaucratisation is evident in the ideal military unit which consists of functionaries – cogs in a larger machine – that can be replaced after attrition. This is not how most networks or groups work in real life. In fact, the ideal type of a military unit is the exact opposite of a fluid and personal network, and hence the least likely case to apply a theory of network-based governance on.

It is equally important to emphasise how tight military patrimonialism was in Liberia predominantly reserved for often-underage bodyguards who moved together with their commanders. While coercion contributed to isolating Liberian combatants from the surrounding society, for most combatants their armed groups or units became nowhere close to the micro-societies and surrogate families Peters (Reference Peters, Özerdem and Podder2011) witnessed in neighbouring Sierra Leone, let alone the total institutions of Western armed forces that isolate their members and subject them to almost total control (Shils & Janowitz Reference Shils and Janowitz1948; Goffman Reference Goffman1987). In comparison, Liberian combatant narratives feature a remarkable freedom of movement during the war between different commanders and units, as well as between frontlines and safer areas behind them (International Crisis Group 2002: 9; Bøås & Hatløy Reference Bøås and Hatløy2008: 44). For many, fighting was far from their main occupation during the four years of war (Utas Reference Utas2003; Cheng Reference Cheng2018: 150). The evidence suggests that most combatants in Liberia did not belong to tight-knit wartime primary groups comparable to Western military units, or necessarily only one such group.

While tight military patrimonialism did exist in Liberia, accounts that focus on child soldiers are extreme and hence unrepresentative of overall organisation of the Liberian armed groups (Hardgrove Reference Hardgrove2017: 99–100; Käihkö Reference Käihkö2017). Another problem with the primary group-like depictions of wartime command structures is their emphasis on trust but neglect of coercion. With weak ideological means, Liberian armed groups had little choice but to rely on force against their members. Because of widespread insecurity, it is often difficult to distinguish forced participation in war from voluntary involvement (Utas Reference Utas2003: 16; Podder Reference Podder, Özerdem and Podder2011: 59–60). Even if not directly forced to join the armed groups, many did so in order to shield themselves or their families from abuse (Utas Reference Utas2003; Bøås & Hatløy Reference Bøås and Hatløy2008: 44–5; Hardgrove Reference Hardgrove2017: 95). Previous research suggests that coercion has a negative effect on group identification (Bourke Reference Bourke1999: 56), and hence cohesion. How coercion affected loyalty and continued association between combatants and their abusive commanders in post-conflict Liberia remains unexplored. Age, place of origins and the time spent in the armed groups too have received limited attention. Finally, it is far from clear that the people – who appear to have been male – who belonged to these commander-combatant networks ever engaged in combat together. It even seems that not everyone in their post-conflict networks even know each other, which alone questions the assumption of a unit-like cohesive primary group that literally fought together (Themnér & Karlén Reference Themnér and Karlén2020: 18).

Group formation is never a natural process, but one which requires intensive social action and organisation (Malešević Reference Malešević2017). Wanting resources contributed to the ad hoc organisation of almost all Liberian armed groups, as did the kind of warfare these groups fought. As common elsewhere (for only one example, see Waite Reference Waite1969), even in Liberia non-state groups were organised around charismatic individuals. The successful frontline commanders took care of their forces (Hoffman Reference Hoffman2007b; Hardgrove Reference Hardgrove2017: 99). This required access to resources, which were typically controlled by political elites. In Liberia, commanders had to associate with elites who could provide resources, including arms, ammunition and food. Even these wartime organisations thus had several layers of hierarchy.

Finally, the war influenced command structure. Wars of differing intensity fought in varying contexts require different kinds of organisation (Sinno Reference Sinno2008), but also lead to diverse social processes. Liberian former combatants describe the war they fought as a rebel war. Its relatively low intensity is captured in the way the three battles in the capital city of Monrovia at the end of the war were called ‘world wars’. Battles were characteristically sporadic and took place when one side tried to take over a village or a town from the other or ambushed them on a road. With aimed shots rare, casualties remained low. Battles tended to end when one side ran out of ammunition (Brabazon Reference Brabazon2003: 9). In general, access to arms and ammunition correlated with the ebb and flow of fighting (Hazen Reference Hazen2013). Like in wars more generally (Bourke Reference Bourke1999: 6, 360), most who partook probably belonged to the supporting tail rather than the tooth, and never engaged in combat.

Only capturing and defending territory demanded concentration of forces in this kind of warfare. This was however hampered by poor communication technology and training. As a result, command relationships became compressed and unit sizes limited. While internal stratification in the form of ranks existed in Liberia, the absence of anything resembling non-commissioned officers means that narratives typically only mention ‘fighters’ and ‘generals’. Core units probably consisted of a commander and around a dozen bodyguards (best evident with government militias; see Käihkö Reference Käihkö2017: 57–8). While overall units could be larger, enforcing participation to missions remained challenging. In practice all larger operations required various commanders to bring their forces together. Compression of command relationships meant that even top commanders had to lead from the front and share dangers with fighters. This kind of charismatic authority does not lend itself well to institutionalisation. Several commanders from all three warring factions were killed in the process. Losses can contribute to cohesion, for instance through evoking solidarity (Barkawi Reference Barkawi2017: 162) and survival guilt (Merridale Reference Merridale2005: 238–9). The deaths of even lower-level commanders nevertheless risked and repeatedly led to demoralisation and disarray in frontlines. The lack of trust within armed groups is evident in the manner such deaths were often explained as results of internal power struggles, not enemy action.

Wartime relationships are generally characterised as instrumental for survival in violent situations (Shils & Janowitz Reference Shils and Janowitz1948: 283). As a result, they are described as ‘utilitarian and narrow but no less passionate because of their accidental and general character’ (Gray Reference Gray1998: 27). Clearly all such relations are products of a particular context, as is their continued existence. Contexts also play a role in whether these networks endure in the absence of coercion and threat of violence. As noted, group formation is not natural, nor is a group's continued existence when its primary function becomes unnecessary. Even maintaining groups requires social action and resources. The utilitarian nature of wartime relationships suggests that they are often accidental in the sense that they are formed because of war and the risks it poses.

End of violence makes many wartime relationships redundant as the threat that has united fundamentally different people diminishes (Gray Reference Gray1998: 15). This, as well as other factors such as physical distance and difficulty of communication help to explain, for instance, why many soldiers tend not to continue to keep in touch with each other after war (Moskos Reference Moskos1970: 145). Overall, it was perhaps unsurprising that two years after the end of the war in Sierra Leone, Hoffman (Reference Hoffman2004: 350) observed that ‘[d]espite the recent end to the hostilities, none of the factions continue to exist with even the rather minimal levels of coherence of the war years … it is not clear that any of the various factions would necessarily re-appear in the form they took during the war’. Why would Liberian wartime command structures have survived longer?

Without a more thorough understanding of wartime structures, it becomes difficult to explain how exactly former commanders maintain control over their former subordinates, and the sources of cohesion that keep these wartime constellations together. If nothing else, the ad hoc nature of the Liberian armed groups, the presence of coercion, the uncertainty that the members of identified networks ever fought together, and the utilitarian nature of wartime relationships raise the question how representative depictions of unit-like cohesive primary groups are in contemporary times. While some Liberian former combatants continue to associate with their former commanders, it feels perilous to assume that former commanders who still maintain a degree of control over their subordinates ‘are somewhat representative of the broader population of ex-commanders in Liberia’ (Themnér & Karlén Reference Themnér and Karlén2020: 13). The next section turns to investigate the social embeddedness of former combatants, and the second ideal type of wartime command structure.

FORMER COMBATANTS AND NETWORKS AS SURVIVAL GROUPS

The second assumption in some past portrayals of Liberian former combatants is that they form a homogeneous community united by their wartime identity, at least to an extent isolated from the surrounding society. Munive (Reference Munive2014; see also United Nations Panel of Experts on Liberia 2013: 12–13) criticises these views, summarising them as follows:

The notion that former combatants have failed to reintegrate suggests that they are not embedded in their surrounding societies in a way that resembles the clear-cut distinction between military and civilian in contemporary Western societies. In fact, the equation of the terms military and former combatant in this literature implies that both are distinguishable from civilian non-combatants, and that the former combatant identity is strongly connected to combat (Themnér Reference Themnér and Utas2012; Bjarnesen Reference Bjarnesen2020: 2; Themnér & Karlén Reference Themnér and Karlén2020: 5). Circumventing that the boundary between civilians and combatants in Liberia was blurred by the thousands who never took part in the war but nevertheless participated in the subsequent DDR programmes (Munive Reference Munive2014: 338–9), it is perceived that former combatants remain ‘mobilized’ in ‘postwar rebel structures’ or ‘former rebel networks’ (Bjarnesen Reference Bjarnesen2020: 3).

The second ideal type of Hoffman's wartime structure is a militarised network. As he observes, ‘the patronage networks which dominate everyday existence have not been replaced in wartime, they have simply become militarised. Ex-combatants remain dependent on their commanders even after disarmament’ (Hoffman Reference Hoffman2007b: 660). Unlike the primary groups discussed in the previous section whose interpersonal cohesion is perceived to have been formed by warfare, these militarised networks are not new creations, but based on pre-war networks that acquired other functions during war. As a result, when Hoffman describes how one of his informants began to mobilise labour after the war, the mobilisation was done through an ‘overlapping network of social and military ties’ where ‘the connections between kinship and command were embodied in the same persons’ (Hoffman Reference Hoffman2011a: 187). If indeed these combatant networks pre-date war, it should hardly be surprising that they survive into its aftermath. And while all this again suggests the importance of social embeddedness, Hoffman simultaneously witnesses more impersonal relationships: the social dynamics of labour ‘fully subsumed by erasing the necessity of distinct institutional identities’ as ‘policing such classifications is no longer necessary’ (Hoffman Reference Hoffman2011a: 188).

What emerges is a contradiction: a relationship that is simultaneously socially embedded into networks that precede war, yet where individual characteristics appear unimportant. The resulting assumption that former combatants constitute a community risks essentialising former combatants and isolating them from their surrounding societies; as Malešević (Reference Malešević2003: 284) claims, ‘“identity” is a concept that simultaneously imprisons and detaches persons from their social and symbolic universes’. McMullin (Reference McMullin2013b) has observed the same with discourses about Liberian former combatants.

To some extent the assumption of a uniting identity among former combatants may perhaps be traced to disciplinary boundaries and narrow aims of past studies, where focus on armed conflict and combat has reinforced the view that this group is socially isolated. The social isolation is implicit in the way factors such as education, ethnicity, military training, religion, social background, wartime faction and rank and whether the former combatants ever participated in combat have often been omitted in past studies (for a notable exception, see Bøås & Hatløy Reference Bøås and Hatløy2008; several are also included in Themnér & Karlén Reference Themnér and Karlén2020: 30–1; ethnicity and faction in Cheng Reference Cheng2018: 153). These kinds of factors may have real consequences, and need to be included in analysis. For instance, both ethnicity and religion became central factors in wartime Liberia, whereas social background likely influenced the fate of combatants after demobilisation, including their continued association with each other. Treating former combatants as a uniform group isolated from their surroundings risks leading to oversimplified policy recommendations.

As noted, Hoffman sees that social dynamics of labour erase the importance of other factors. This raises the question whether ex-combatant identity too becomes irrelevant. It is noteworthy that several descriptions of former combatant networks explicitly note that these networks do not exclusively consist of former combatants. In addition, the former combatants in these networks come from different wartime factions (Hoffman Reference Hoffman2011a: 184–7; Persson Reference Persson and Utas2012: 107–14; Cheng Reference Cheng2018: 153; Bjarnesen Reference Bjarnesen2020: 9–10). The members of these networks could thus have belonged to competing wartime chains of command, or none at all.

According to Bjarnesen (Reference Bjarnesen2020: 9), former combatant ‘networks are not static entities that would be mapped easily … they are not formalized static groups but fluid and flexible constellations that continuously change in size and composition’. This observation can be taken as criticism against the notions of tight commander-combatant networks investigated in the previous section and methods such as social network analysis used to study them. Yet if these more fluid networks include former combatants from different wartime factions as well as non-combatants, one wonders their precise link to wartime command structures, and why these networks are explained through the war. This kind of emphasis on past conflict risks oversimplifying things and reinforcing problematic discourses about former combatants (see McMullin Reference McMullin2013b: 396–7).

Long after the end of the war a more parsimonious explanation for former combatants to associate with each other is not past chains of command but vulnerability in a context characterised by scarcity (Utas Reference Utas2003; De Vries & Wiegink Reference De Vries and Wiegink2011; Bjarnesen Reference Bjarnesen2020: 148; Mitton Reference Mitton, Mac Ginty, Vogel and Brett2020; Themnér & Karlén Reference Themnér and Karlén2020). Much has been made in past literature about the ideas of ‘wealth in people’, or the importance of investing in social relations in West Africa. Ultimately, relationships constitute a crucial survival strategy (Bledsoe Reference Bledsoe1980) to the extent that groups whose primary function is safeguarding the existence of their members can be described as survival groups. Of course, these kinds of strategies are a recurrent necessity in economies of scarcity around the world. Minority groups in Europe and the USA alike rely on extensive kin networks for support (Stack Reference Stack1974; Friberg Reference Friberg2020), while Soviet citizens depended on networks of favours among friends called biat (Edele Reference Edele2019). A more contemporary example of the importance of networks – strategic social interactions for an inherently instrumental purpose – comes from their importance in job-hunting (Gershon Reference Gershon2017).

Liberians are not exceptional in having to rely on each other to make do. Even there most assistance during times of need takes place first within a family, and then between friends, colleagues, neighbours, and church and mosque goers, among others. Faced with the loss of their wartime status in a poor post-conflict environment, it is understandable that many former combatants were reluctant to let go of the past. Some would have wanted to, but had lost their families, were too old to depend on them or were rejected by their families and communities immediately after the war (Utas Reference Utas2003: 63–64; see also Hardgrove Reference Hardgrove2017: 104–5).

As noted, the former combatant identity can be used strategically. Ultimately the longevity of any identity is connected to the hardships and privileges involved. In Liberia as well as in many other contexts, the relevance of former combatant status varies. In some cases, the status of being a combatant is positive, and kept meaningful through positive incentives, such as pensions, societal recognition and state-sponsored veteran's associations that create and maintain memory, as well as wartime organisational structures. In other cases, the status becomes a stigma that many seek to disassociate themselves from. In Liberia the celebration mostly concerns the regular soldiers of the armed forces, not the former combatants who belonged to non-state armed groups (Kaufmann Reference Kaufmann2013; Käihkö Reference Käihkö, Haldén and Jackson2016b).

As noted, the idea that some former combatants continue to rely on their wartime identities and former combatant networks for survival is not new. Viewing these networks as survival groups among others nevertheless allows de-emphasising former combatants’ strict wartime role and permits – if not necessitates – their members to be analysed as more than mere former combatants. This is threatened by essentialisation of the identity, where everything former combatants do is explained by their wartime past. As many veterans even in our own societies can attest (for instance, see Bourke Reference Bourke1999: 345–69), this is again a more general phenomenon than one that concerns Liberia alone.

Understandably there are many reasons for continued association between former combatants, and it would be misleading to claim that the war has ceased to matter to all Liberians who participated in it. The transition from war to peace was difficult for many combatants (Utas Reference Utas2003; Hardgrove Reference Hardgrove2017; for a more general view, see Bourke Reference Bourke1999). Like many others whose lives were interrupted by warfare, even many Liberian former combatants feel disadvantaged because of the war. Aside from whatever physical and psychological traumas they might suffer from, many of my informants have poor relationships with their families and lack professional experience and education. Unable to go to school during the war, several remain illiterate.

In some ways the situation was different for commanders, who risked losing the often-unprecedented status they had enjoyed during war. Only a few were able to gain formal positions during the interim government that ruled from the end of war until the Johnson Sirleaf government assumed power in 2006. The majority were subsequently replaced by political loyalists of the new incumbent, and joined the precarious existence of not only most former combatants, but the majority of Liberians (Moran Reference Moran2006). As one LURD commander explained the dilemma, he felt trapped between his wartime status and the meagre opportunities offered in the post-conflict: he risked losing the status gained during war if he was seen sweeping his own floor, let alone accepting any menial paid work.

Yet when I repeated these words to several other former combatants they scoffed and emphasised how the war had been over for a long time in 2013, when these discussions took place. While former combatants could indeed continue to rely on their wartime comrades for various kinds of support (Bjarnesen Reference Bjarnesen2020: 149; Themnér & Karlén Reference Themnér and Karlén2020), those who continued to cling to the wartime identity risked increasingly being seen as losers. While many commanders sought to maintain their status, they simultaneously struggled to provide opportunities for their wartime subordinates. In 2012 it was clear that many former commanders maintained wartime stratification. They associated more with each other than with former combatants, who they preferred to keep at arm's length. This did not escape the notice of former combatants, many of whom complained bitterly about the way they had been treated after the war by their wartime superiors.

To summarise, the notion that former combatants remain mobilised in ‘rebel’ structures or networks long after the war suggests that they share an identity that unites them in a community separated from their surrounding society. This notion of weak social embeddedness is however questioned by Hoffman's idea of militarised networks, as well as observations that former combatants too belong to various networks (Bjarnesen Reference Bjarnesen2020: 8). Both suggest weaker links to wartime structures on the one hand, and former combatant identity on the other than those offered by more essentialised views of former combatants, present especially in policy discourse (scrutinised by Munive Reference Munive2014).

THE DISCIPLINING EFFECT OF WARTIME EXPERIENCE

The third assumption in some of the literature about Liberian former combatants concerns the disciplining effect of wartime experiences. In what serves as a reversal of past discourses that have reduced former combatants to a threat (investigated for instance in McMullin Reference McMullin2013b; Munive Reference Munive2014: 340–1), Liberian former commanders have been described as entrepreneurial and ‘reliable managers’, and their former fighters as ‘good workers’ (Utas et al. Reference Utas, Themnér and Lindberg2014: 2), ‘hardworking, well organized, and disciplined’ (Bjarnesen Reference Bjarnesen2020: 124), and ‘[h]ardened by their experiences from the previous war, accustomed to working collectively, and used to taking orders’ (Themnér Reference Themnér and Utas2012: 205).

While the DDR programme itself likely had a disciplining effect (Munive Reference Munive2014), these views assume that it is first and foremost wartime demands that have forged former combatant networks useful even to the post-conflict labour market. Following Reno's (Reference Reno2010: 141) train of thought, these networks should not be demobilised, as much of the peacebuilding literature proposes, but rather restructured and maintained in order to alleviate the economic grievances of former combatants (Themnér Reference Themnér and Utas2012; Utas et al. Reference Utas, Themnér and Lindberg2014). There are obvious ethical dilemmas connected to supporting former combatants who may have committed human rights violations during war. This is especially the case if this entails reconstituting, if not creating, wartime structures long after war. More importantly, there are strong theoretical and empirical reasons that question the assumption of any major disciplining effect of a ‘rebel’ war.

One should be careful about assuming that the institutional and social context in Liberia is the same as in Europe, where Weber, among others, saw all discipline originating from the military (Waters & Waters Reference Waters and Waters2015: 62–71). Conscription emerged as an important way to educate citizens and subjects (Lachmann Reference Lachmann, Hall and Malešević2013: 47–8), with militaries acting as schools for the nation, and in the Soviet Union even for socialism. Drilling discipline into the conscripted male population throughout Europe was perceived as valuable schooling for productive work in the factory (Sheehan Reference Sheehan2008: 17). Likewise, labour reserves of semi-skilled but disciplined industrial workers could also be turned into military reserves when needed (Lederer Reference Lederer2006: 256–7; Edele Reference Edele2019: 86). With the experiences of the First World War in mind – a war of material where industrial capacity played an unprecedented role – even Weber connected military discipline to centralisation of war equipment (Waters & Waters Reference Waters and Waters2015: 69–70).

As already discussed, this was not the institutional context in Liberia, nor did warfare waged there depend on comparable discipline. The ad hoc nature of the warring organisations and the demands of a ‘rebel’ war meant that there was little of the drill and routine associated with everyday life in the military. The comparatively small scale of warfare enabled, if not demanded, individual heroism in ways the industrial scale and quality of the World War all but denied. All belligerents in Liberia struggled to instil discipline to control their fighters (Käihkö Reference Käihkö2015). This was partly because arms and ammunition allowed them to live from the land, and to thus acquire independent sources of sustenance. Centralisation could also be perceived as a threat. Not unlike other authoritarian leaders (Ullrich Reference Ullrich2017), even Taylor resorted to the practice of divide and rule, and for instance subcontracted his war effort to competing militia commanders with equal rank to maintain constant competition among them (Käihkö Reference Käihkö2017). No institutions dedicated years to transforming citizens to soldiers, as even the armed forces’ participation in the wars after the murder of President Doe in 1990 was limited. The role of the state was thus fundamentally different. Despite attempts by elites, nationalism and similar unifying countrywide ideologies too had limited significance.

Empirically, rather than showing discipline, warfare in Liberia more often illustrates the lack of it. This does not mean that the war equalled anarchy. Nevertheless, those who fought in Liberia usually describe their wars with ‘guerrilla’ and ‘rebel’ labels. This kind of warfare was perceived to require traits not associated with formally trained government soldiers, but militias and rebels (Käihkö Reference Käihkö2017). The terms ‘militia’ and ‘rebel’ are used synonymously in Liberia, and ‘a rebel in the popular imaginary of the Mano River region lives in the bush and inflicts violence on the populace’ (Hoffman Reference Hoffman2011b: 38). As already noted, the positive traits such as legitimacy associated with the military do not by and large extend to former combatants (Hardgrove Reference Hardgrove2017: 105–6). In everyday Liberian parlance ‘rebel behaviour’ remains anything but positive. Often mentioned in connection with motorcycle riders who are commonly but incorrectly considered to mostly be former combatants, rebel behaviour denotes reckless, rude, and potentially violent conduct (Bjarnesen Reference Bjarnesen2020: 121). As McMullin describes, former combatants are scapegoated for more general antisocial behaviour, which becomes ‘compartmentalized from the social, political, and economic relations … and is instead characterized as behaviour modelled on (and therefore authored by) ex-combatants’ (McMullin Reference McMullin2013b: 396–7, 411). This scapegoating rarely helps former combatants in the post-conflict era. This is especially the case with women (Coulter Reference Coulter2009; Vastapuu Reference Vastapuu2017; Bjarnesen Reference Bjarnesen2020: 125–7), most of whom however avoided the former combatant label (Bøås & Hatløy Reference Bøås and Hatløy2008: 38).

Much of the stigma associated with former combatants derives from the widespread notion that the archetypal former combatant is not disciplined at all. Those described as former combatants are typically perceived to be impatient and undisciplined. During my fieldwork, many Liberians believed that former combatants could best be found in drug ghettoes. Used to fast money, they were assumed to be ready to employ dubious means to get it. None of these traits are positive for any tasks without immediate benefits. Of course, not all former combatants were identified as such, and some have successfully framed their wartime experience as security expertise. While obscured by lumping together all former combatants, transforming wartime experience into something useful in the post-conflict was easier for those who had served in the armed forces and other formal security outfits, than for those in rebel and militia forces. It is likely that legitimacy from the association with the state again plays a role here, which is one reason why the broader discourses on veterans and former combatants tend to differ significantly. Of note is how we tend to ignore the disagreements of who counts as a veteran even in our own societies. The recognition is ultimately political, not least if the status is connected to material or symbolic benefits.

As can be assumed, the thousands of Liberian former combatants portray great variation, as do their wartime experiences. Former combatants are not the same, and even their discursive representation varies (McMullin Reference McMullin2013b: 388). Ultimately, limited concrete evidence exists to support the idea that whether positive or negative, variation in former combatants’ discipline results from more than individual differences.

REMOBILISATION OF LIBERIAN FORMER COMBATANTS

The fourth assumption concerning Liberian former combatants is that they are good mobilisers and easy to mobilise. As it is the assumed capacity to violence that makes former combatants stand out, I first focus on election mobilisation and then remobilisation to war. These are also the two activities most literature has focused on, even if studies of other kinds of labour might have been helpful for investigating their discipline, for instance. I discuss former combatants’ struggles and social embeddedness and contend that electoral and more violent mobilisation cannot be understood without considering vulnerability, inclusion to state structures and social embeddedness. More recent cases of the Liberian 2017 general elections and the 2010–2013 mobilisation to fight in Côte d'Ivoire suggest that like combatant networks, even their members’ power to mobilise diminishes with time.

Former combatants and their commanders have repeatedly been connected to voter mobilisation (Hoffman Reference Hoffman2007a: 421; Christensen & Utas Reference Christensen and Utas2008; Themnér Reference Themnér and Utas2012: 214–15; Themnér & Karlén Reference Themnér and Karlén2020). In my discussions with Liberian politicians in the run-up to the 2017 elections, former combatants were however repeatedly understood to be more untrustworthy than other youth when it came to political support. Informants associated with the elections emphasised the importance of familiar and popular ‘pioneers in a community’ able to sway voters one way or the other. In comparison, former commanders were perceived to not only lack economic status, but more importantly any moral standing. Viewing wartime command chains ineffective, one long-standing member of Taylor's political party described former commanders as ‘non-entities’. While the supporters of one former head of an armed group wished that he had drawn more on his wartime reputation in the first elections after the war, by 2012 they already felt that this opportunity had passed.

While it is common for former commanders to boast that they can mobilise great numbers of former combatants if needed, several of them ultimately admit that people with steady employment or other kind of security would have no reason to answer their call. Simply put, they could mobilise those with little to lose but something to gain. Even the politicians I talked to offered several concrete examples where former combatants had been mobilised by commanders for political work, but where the end result was a catastrophe as soon as money ran out.

If successful mobilisation depends not on wartime status but rather vulnerability and poverty, this is something that any so-called big man – anyone with resources and hence power – could achieve (see even Cheng Reference Cheng2018: 85). In this sense it is crucial to emphasise how the term ‘commander’ included even those who mobilised but never actually commanded forces in battle (Hoffman Reference Hoffman2007b: 652–3), and that the above informants discussed former combatants in an essentialising manner best avoided.

Early on in my research in 2012 it became clear that many ‘generals’ introduced by local research assistants were these kinds of mobilisers, not commanders in any military sense. This implies the conflation of ‘generals’ and ‘big men’ in the minds of many Liberians, and suggests that the structural causes behind successful mobilisation are connected to vulnerability and the necessity to find the means to survive. The primacy of vulnerability in these relationships can also be seen in numerous examples where more successful individuals leave groups when their reciprocal nature makes them lose rather than win. As noted, this is also how many commanders appear to feel about their wartime subordinates.

Explaining mobilisation through vulnerability is however far from enough, as this suggests that former combatants constitute a destabilising force comparable to lumpen youth. While the ideas of commanders as remobilisers and vulnerability as a more personal cause seek to give an explanation to violence, they end up with the same problem: while they explain why some belonging to these categories mobilised for violence, they never address why the vast majority do not. One survey conducted in 2008 found that only two of the 466 former combatants asked would join an armed group if fighting started again in Liberia (Bøås & Hatløy Reference Bøås and Hatløy2008: 51).

In the end no single factor can fully explain remobilisation to war (Podder Reference Podder2017). Yet what occurred in the south-eastern Grand Gedeh County in 2010–2013 may offer some possible explanations.

In late 2010 the contested results of the Ivoirian presidential elections led to a civil war where Liberian combatants were mobilised by both sides. The mobilisation patterns roughly followed the factions of the Second Liberian Civil War: whereas former Taylor supporters were approached by those supporting Alassane Ouattara, the Laurent Gbagbo side relied on their former acquaintances who had fought in an Ivoirian-supported militia, parts of which later became the Movement for Democracy in Liberia (MODEL) rebel group in 2003. Most of the latter hailed from the Liberian side of the border in south-eastern Grand Gedeh County (Käihkö Reference Käihkö2018c), and most of the former probably from north-eastern Nimba County. As in the mobilisation to fight against Taylor in the 1999–2003 war, even a decade later social embeddedness played a role. To assume otherwise normalises violence to a degree unseen in Liberia (Käihkö Reference Käihkö2018b).

This kind of mobilisation through former commanders to Côte d'Ivoire nevertheless turned out to be suboptimal: several commanders were known to have ‘eaten’ money given to them to mobilise fighters. A number of my informants also narrated the story of ‘Saddam’, a former LURD commander who hailed from a town close to the Ivorian border whose inhabitants were described as ‘like in Ivory Coast’. Saddam was apparently killed by his own men – ‘people he didn't know’ – when returning to Liberia after he failed to pay them. These independent but generally consistent narratives suggest how mobilisation by a former commander got him killed, but also how those he mobilised were people he had not associated with during past wars. This corresponds with other narratives which describe how the mobilisation to Côte d'Ivoire did not exclusively concern former combatants. In 2010–2011 a wave of youth – a category that in Liberia denotes marginalisation rather than age – made their way across the border from Liberia. Yet only a year later after Gbagbo had been disposed, no similar mass mobilisation took place. This begs the question why.

One partial explanation originates from the Liberian civil wars, where fighters from all factions were united by the goal of inclusion into existing state structures (for a more general discussion of inclusion into patronage networks, see Themnér Reference Themnér2015). Here the inherent uncertainty of war can be beneficial for those who seek to change the status quo, and especially those who have little to lose. If one succeeds in picking the winning side, there is at least a possibility of inclusion into the state after victory. Considering the insecurities inherent in everyday life in Liberia even during times of peace, their most potent antidote is the state (Käihkö Reference Käihkö2017). Even further, the importance of association with the state is strengthened by the view that it is ultimately seen as the only legitimate source of wealth, and hence security in an economy of scarcity characterised by insecurity (Marchal Reference Marchal, Carayannis and Lombard2015). In contemporary Liberia, government jobs are widely perceived to offer a chance to relax (Hardgrove Reference Hardgrove2017: 51–3). Many officials are simply assumed to come in to sign their attendance, leaving the actual work to ‘undermen’. In addition, and especially in the security sector, government connection is also seen to give status and hence protection against harassment, as well as contacts useful in private business activities. Strengthening the importance of inclusion as a factor for mobilisation, a government position is also the stated desire of those who help to mobilise voters to get politicians elected as well as those assisting with security during rallies and political events (Christensen & Utas Reference Christensen and Utas2008; Bjarnesen Reference Bjarnesen2020: 104). This said, the idea of inclusion may need to be nuanced somewhat when it comes to remobilisation of former combatants long after war.

One way to do this is to draw from Mitton, who reverses the question of remobilisation: despite the unchanged economic, political and social condition, why did Sierra Leoneans not return to war when one could have expected them to do so? Ultimately, he finds that especially those former combatants coerced to fight were as reluctant to use violence as they had been before they were forcefully mobilised. Even those who had fought for ideological reasons admitted that the war had made them worse, rather than better off (Mitton Reference Mitton2013; see also Bjarnesen Reference Bjarnesen2020). Reluctance to use violence and a similar kind of understanding that war concerns ‘destruction, not construction’ made many Liberians uninterested in renewed fighting. The idea of inclusion also explains why the subsequent mobilisation against Ouattara in 2012–2013 attracted precious few in Grand Gedeh (and possibly no-one elsewhere). Not only had those mobilised across the border in Grand Gedeh fought and lost with Gbagbo in 2010–2011, but with Ouattara consolidating international support the chances of ousting him appeared slim. This did not encourage mobilisation to fight against him.

Here a sharp differentiation between contracts and employment in security forces drawn by Liberian former combatants comes into play: it is only the latter that can offer long-term security. This preference is also reflected in the answers to the 2008 survey mentioned above, where 10 of the respondents answered that they would seek to join the armed forces or police forces if fighting restarted in Liberia (Bøås & Hatløy Reference Bøås and Hatløy2008: 51).While still overall an insignificant portion of the overall respondents, this is still five times more than those who would have preferred to join a non-state armed group.

While inclusion offers a partial explanation to the absence of mobilisation to Côte d'Ivoire in 2012–13, it does not explain why some still participated in the conflict. The Ivoirians and Liberians who I know participated in these attacks professed close connections to the Gbagbo regime. Several enjoyed personal links to Ivorian government officials. In fact, most of the Liberians who fought in 2012–13 had opted to stay in Côte d'Ivoire after the partition of the country in 2003, rather than to return to Liberia as rebels to fight against Charles Taylor. Significant parts of their lives remained in Côte d'Ivoire when the post-election violence began in 2010. With (Liberian) Krahn associated with Gbagbo, many of these former combatants’ futures were tied with his regime in ways that make it difficult to see their participation in the 2012–13 conflict as apolitical. While social embeddedness in Côte d'Ivoire may have been enough to bond together a small number of people, these cross-border attacks were never condemned by Grand Gedehians during my fieldwork there in 2012–2013 and 2017. This suggests the possibility that they remained legitimate and within the bounds of local moral order. This warrants the consideration of macro-level explanations that concern many more than former combatants alone (see also Blok Reference Blok2001: 14–28).