Rebel Governance and Legitimacy: Understanding the Impact of Rebel Legitimation on Civilian Compliance with the LTTE Rule

Abstract

Based on extensive fieldwork in Sri Lanka, we analyze how the Liberation Tigers of Tamil Eelam (LTTE) acquired legitimacy and how legitimation impacts civilian perceptions of the rebels. Despite the LTTE’s reliance on coercion to induce compliance, civilians also supported the LTTE and their imagined state of Tamil Eelam voluntarily. Different LTTE strategies and acts helped creating legitimacy. Effective forms of legitimation were rooted in Tamil nationalism, tradition, charismatic leadership, sacrifices made by LTTE cadres and the people’s need for protection. However, the strong reformative socio-political agenda of the LTTE largely failed to engender legitimacy among the population.

Introduction

This article focuses on the ways in which rebels acquire legitimacy among their constituencies and beyond. Once rebels become involved in governance and start to behave like a rebel government, they face similar legitimacy issues as incumbent governments, as they raise the expectations of civilians living under their control. Even if a rebel government relies upon coercion to implement its rule, it may still want to acquire legitimacy to increase civilian compliance, either by employing conscious and deliberate legitimation strategies or simply because legitimacy is accorded to them by the population for a variety of other reasons. The costs for a government or rebel group of relying only on coercion are usually high and its effects last only as long as the coercion is effectively applied, while an element of legitimacy may provide sustainability. The different forms of legitimation during periods of rebel governance and their mutually complementary or opposing effects on the compliance of civilian populations in rebel-held territories have remained under-researched in the current rebel governance literature.

Civilian compliance during civil war is usually based on a mixture of coercion and persuasion; on fear and sympathy (Kalyvas 2006, pp. 101–104). To explore the balance between coercion and legitimation, we investigate the case of the Liberation Tigers of Tamil Eelam (LTTE) in Sri Lanka. The LTTE is known for its use of coercion in ruling the Muslim and Tamil civilian populations in the territories under its military control. ¹ In this article, we will demonstrate how, in addition to coercive techniques, the LTTE also induced civilian compliance in its quasi-state due to the legitimacy it possessed or acquired. Apart from forced recruitment, many sons and daughters of Tamil families joined the LTTE voluntarily as cadres or general members of the movement. Many families provided food and shelter to the LTTE cadres or supported the movement through financial means and local intelligence. Among those living in the Northeast, there was also respect for the LTTE leadership and admiration for the sacrifices made on behalf of the Tamil community. The movement used coercive techniques extensively, but also made significant efforts to create a broader sense of legitimacy among civilians that lived inside rebel-held territories.

Insights into LTTE governance have come from authors in various disciplines who investigated the LTTE’s political governance project in Jaffna from 1990 until 1995, and in the Vanni from 1995 until the end of the war in 2009 (Stokke 2006, Sarvananthan 2007, Fuglerud 2009, Mampilly 2011, Klem and Maunaguru 2017, Terpstra and Frerks forthcoming). Mampilly showed how the LTTE was involved in the provision of various public services such as the police, judiciary, health care and education (Mampilly 2011). With regard to the police and judiciary, Mampilly asserts that his ‘civilian informants confirmed that the rebel police had a high degree of legitimacy and viewed the force as an uncorrupt and important stabilizing factor in the region’ (Mampilly 2011, p. 116). At the other end of the spectrum, Thiranagama (2011, p. 39) emphasises that one has to be careful with evaluations of legitimacy given the strong political repression that characterised the society in the LTTE-controlled territories. In more general terms, Wedeen stresses that one has to make a clear distinction between what social scientists in the tradition of Weber conceive as ‘a charismatic, loyalty-producing regime’, on the one hand, and an ‘anxiety-inducing simulacrum’ on the other (1998, p. 506).

We analyse coercive techniques as ‘actions or practices of persuading someone to do something by using force or threats’ (Stevenson 2010). Threats, intimidation and a show of force by rebels may induce civilian compliance, including by extortion, population expulsion and forced conscription. But in addition to coercive techniques, many rebel groups also deploy legitimation strategies, described as ‘a series of actions – speech, writing, ritual, display – whereby people justify to themselves or others the actions they are taking and identities they are expressing or claiming’ (Barker 2001, 2003, pp. 163–164). In addition, constituents may themselves comply with rebel rule because ‘doing so is believed to be right, fair and appropriate’ (Whalan 2013, p. 56). We argue that even though the LTTE’s rebellion was also based on coercion and an enforced agenda of social and political reform in the Tamil community, broader civilian compliance existed, shaped by legitimation based on Tamil nationalism, tradition, charismatic leadership, enemy images and sacrifices of LTTE cadres. Some of these sources were actively disseminated by the LTTE, others proved to operate on their own among the populace.

While we recognise that rebel governance is frequently ‘multi-layered’, involving influences of several parties, our approach is a ‘single-layered’ analysis. ² This approach is well-suited, because a study on legitimacy requires an investigation into the effect of legitimation strategies on a specific target audience. We therefore focus on the perceptions and behaviour of those living inside territory under full or partial LTTE control from 1983 until 2009, although we recognise that the LTTE invested in its portrayal among an international audience too (Bhatia 2005, Nadarajah and Sriskandarajah 2005, Terpstra and Frerks forthcoming). Within a target audience legitimation strategies may resonate differently, for example, if we consider differences between elderly and youngsters, or different castes.

Segments of the Tamil population may have had their own reasons and inducements to do, or do not, accord legitimacy to the LTTE. In this connection, we consider legitimation a two-way process in which both top-down and bottom-up aspects play a significant role. It would be a mistake to believe that all legitimation effects observed emerged as a consequence of a calculated ‘rational actor’ behaviour on behalf of the LTTE. We indeed assert that the LTTE used a number of deliberate legitimation strategies, but recognise that other forms of legitimacy originated bottom-up or were a consequence of behaviour or acts that were not intended as a legitimation strategy, but nevertheless contributed to legitimacy. We also should not forget that the operations and behaviour of the Sri Lankan Armed Forces and of the Indian Peace Keeping Forces (IPKF) contributed to the legitimacy of the LTTE and the Eelam project. ³

We are obviously aware that our contribution is entangled in a complicated context of war and a militarised society (Mel 2007). This makes it difficult to make clear-cut distinctions or cause-effect statements. Likewise, it is also not always possible to distinguish clearly the respective impact of coercion and legitimacy in engendering compliance. It is evident that the exercise of power by the LTTE was based on a combination of coercion and legitimacy, and one needs to keep in mind that they are interrelated both from the perspective of the LTTE and in the minds of the population, even though we distinguish them analytically. A final note we want to make in advance is that the war and the LTTE rule encompassed a long period of time with obvious variations internally and externally, and shifting military fortunes. Though we are well aware of these variations, we cannot treat them in-depth within the scope of this article, apart from highlighting a few major events in our chronological section on the LTTE below. Moreover, the war has been documented extensively in the broader literature on Sri Lanka and we do not need these details for the sake of our main argument. ⁴

Our contribution to the fields of Sri Lanka studies and rebel governance is threefold. First, our analysis of the specific elements of legitimation and how these can produce varying and opposing effects on civilian compliance has remained a relatively underexplored aspect in the current literature on rebel governance. We have added detailed data and insights to this literature making it more robust in terms of empirical substance. For Sri Lanka this data was still largely missing. Second, though Schlichte and Schneckener (2016) have dissected different sources of legitimacy, we have identified how these operate in a real case setting and also can complement or oppose each other. We go beyond the tangible top-down service provision dimension highlighted in much of the rebel governance literature and address the legitimising effect of the symbolic repertoire in a detailed manner. Third, we believe we collected fairly unique primary data that can especially elucidate the bottom-up aspects of rebel legitimacy by representing the views and opinions of the population in rebel-held areas in a more in-depth manner than done so far. Such a type of research was not possible during the LTTE’s or Rajapakse’s rule due to safety considerations for both researchers and interviewees.

Our findings are based on extensive fieldwork in the Northeast of Sri Lanka in 2014, 2015 and 2016. In 2014–15, we focused on LTTE governance and conducted a total of 76 semi-structured interviews of about two-and-a-half hours with community members in nine different locations in the Trincomalee district (33 interviews), 10 locations in the Batticaloa district (25 interviews) and 4 locations in the Northern province (18 interviews). A total of 20 key informants such as civil society leaders, community leaders, NGO workers, religious leaders, doctors, ex-LTTE cadres and supporters and local government officials were also interviewed. In 2016, we conducted 62 interviews about popular support for the LTTE and symbolic legitimation in the Trincomalee district and throughout the Northern Province. The locations included zones then under full control of the LTTE, government-controlled (‘cleared’) areas and grey zones. The interviews with key informants were done by ourselves, while we used a team of eight local field researchers for the semi-structured interviews, in view of language demands. This team was working for a local research NGO and wishes to remain anonymous. They were trained by us in several workshops, while we also held debriefing sessions. Notes were made of all interviews and the resulting Tamil reports were translated in English by a professional translator, after which they were coded and analysed by us with the help of NVIVO software. The translations and working with different researchers may have led to some loss of content and perhaps a less systematic probing then we ourselves would have done. In order to redress, this we have triangulated our different data-sets where possible and also extensively used secondary material.

Theoretical Considerations

Rebels and Civilians

In theories on rebellion and insurgency in civil war, civilian collaboration is usually seen as a central concern for the strategic success of a rebel group (Kalyvas 2006, pp. 87–110). In the words of Mao Tse-tung: ‘The richest source of power to wage war lies in the masses of the people’ (1965, p. 186). On the basis of evidence collected from three continents throughout three decades, Fall shows that civilian support is the essential element of successful guerilla operations (1967, p. 345). Since civilians can provide food, information and be a source of new recruits, the leadership of rebel groups will attempt to consolidate support among its constituents. A rebel leader without followers or civilian support will probably not get far in achieving political and military goals.

However, the assertion that every insurgency needs popular support to achieve its objectives is contested. As O’Neill points out, there is a tendency in the literature on insurgency and counter-insurgency to equate all types of insurgencies with the ‘protracted popular war’ strategy (2005, pp. 45–67). In the ‘conspiratorial strategy’ for example, in which an insurgency seeks to remove the ruling authorities through a limited but swift use of force, broad-based popular support is not necessary for political or military success. While acknowledging these variations, we focus in this paper on an insurgency that followed a ‘protracted popular war strategy’ and invested in the compliance of its followers, whether through coercion or ‘softer’ forms of persuasion.

Consequently, we focus on the interactions between rebel elites and the civilian population during civil wars. As Kasfir points out, however, the process of engaging civilians is filled with obstacles, probably most prominently because the rebels are at war with the incumbent government (Kasfir 2005, p. 273). Military strategic considerations may, for example, force insurgents to retreat from territory where they have just built a positive relationship with the population. If anything, the relationship between the rebel leadership and the civilian population is contingent on the dynamics of war. The empirical study of patterns of interaction between the rebel group and the civilians will facilitate the understanding of the factors and dynamics influencing the use and effect of legitimation strategies in rebel governance.

Rebel Governance

In this paper, we adopt the notion of governance to gain a better understanding of the interactions between rebels and civilians during civil war. As Kalyvas puts it: ‘Insurgency can best be understood as a process of competitive state building rather than simply an instance of collective action or social contention’ (2006, p. 218). Insurgents attempt to develop ‘counter-states’ through ‘political consolidation’ (Kalyvas 2006, p. 218). Similarly, Kasfir explains that the ‘relations between guerillas and civilians can be understood as problems of politics under special circumstances’ (2002, p. 4). He adds ‘guerrilla governance refers to the range of possibilities for organization, authority and responsiveness created from the daily interactions between guerrillas and civilians’ (2002, p. 4). These interactions may vary from ad hoc arrangements to elaborate and enduring regulatory structures. These interactions may also vary from coercive measures to ‘softer’ measures inducing civilian compliance. Civilians may respond to these measures by agreeing to the insurgent’s requests or refusing them. Passively or actively, civilians may demand adaptations of the governance structures (Kasfir 2005, p. 274). In other words, within this interaction both the rebels and the civilians are agents who shape governance structures, and at the same time the governance structure constitutes the opportunities for their own conduct. As we use it here, governance contains no prescriptive implications. Instead, governance comprises the conduct of rebel–civilian relations, regardless of morally good or bad behaviour or treatment by either side. ⁵

According to Kasfir, a number of scope conditions must apply before we are able to observe the phenomenon of ‘rebel governance’ (2015, p. 25). First, rebels must control territory – even if that control is contested. Second, civilians have to reside in that area. Third, the rebel groups must have acted violently and continue their hostility or at least threaten to do so in the territory they govern. Kasfir (2015, p. 25) elaborates that if these conditions are met, at least some elementary form of governance usually emerges. Governance in itself does not equal ‘popular support’, although they may go hand-in-hand. The rebels must show the civilian population that their rule has some degree of predictability and that they have rules they enforce responsibly over time.

Rebel Legitimacy

We follow Mampilly (2015) and Schlichte and Schneckener (2016) that rebel governance depends not only on instruments of coercion and service provision, but also on legitimacy in the minds of the ruled. ⁶ We will, therefore, investigate how the LTTE attempted to legitimise their existence, their rule and their use of coercion through various symbols of, and claims to, legitimacy. According to Suchman, legitimacy is a ‘generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions’ (1995, p. 574). Legitimacy involves the ‘complex moral right of the state to impose rules on its subjects, while the latter agree to comply with such rules’ (Podder 2013, p. 19). Political legitimacy resides in a ‘tacit social contract’ between state and society, or the ruler and the ruled (Podder 2013, p. 19). For rebel groups, unlike state actors, legality is usually not a viable option as a formula for legitimacy (Schlichte and Schneckener 2016, p. 417). Despite the absence of juridical legitimacy or international recognition, rebel groups may hold varying levels of legitimacy among the members of a particular constituency (Mampilly 2011, pp. 48–92, Péclard and Mechoulan 2015).

Rebel groups can draw from different sources to legitimise their rule (Schlichte and Schneckener 2016, pp. 416–419). We highlight five legitimation strategies that are relevant to our argument. First, claims can represent the socio-economic and political aspirations of a community, for example, a social class, ethnic group or a clan. Rebel leaders claim their positions as representatives of these communities and their grievances. They give a (radical) voice to it. A second strategy projects outside threats and particular enemy images. By presenting the opponent as particularly brutal and inhumane, they may create solidarity and show that their (violent) actions are a necessity. A third element is the charisma of the successful warrior and leader. Charismatic authority exists if there is belief in the extraordinary quality of a specific person (Weber 1978, p. 241). Fourth, the readiness of leaders and fighters to sacrifice their lives for a common cause, may work to legitimise the rebel group towards its constituency. This is usually expressed through heroism and martyrdom. Fifth, the claims of rebel groups are often rooted in ‘communal myth-symbol complexes’ and popular belief systems, traditions and cultures. In Weber’s explanation, command and obedience are legitimised on the basis of the sanctity of immemorial traditions, which govern the authority relationship (Weber 1947, p. 341).

For all these legitimation strategies it is important to underline that without the recognition of the followers the legitimacy of a movement or its ideas does not exist (Weber 1947, p. 359). Hence, the perceptions and practices of the civilians living inside the territory of the rebel group determine whether these legitimation strategies are working. To evaluate their effectiveness, we need to learn whether these legitimation strategies ‘resonated’ amongst the target audience (Benford and Snow 2000, pp. 619–622).

Whether these legitimation strategies should be conceived of as fully intentional, unintentional or somewhere in between is subject to debate and has already been alluded to in the introduction. At its very basic level, our reading is an existing intention on behalf of the LTTE leadership to strategically portray its struggle to various audiences, with the objective to achieve political and military results. One such audience was the international community. Part of the LTTE leadership travelled around the world to show what their perceived Tamil Eelam would look like and tried to convince their target audience that support for their cause was indeed necessary. For this purpose they also established an extensive propaganda machinery abroad, including internet. Another target audience, was the Tamil population living under full or partial control of the LTTE in Sri Lanka itself. The LTTE leadership was dependent upon the compliance of Tamil civilians to continue its struggle, particularly on the long run. In this connection Balasingham writes:

While the Sri Lankan state was intensifying its military domination and respression in the Tamil homeland, the LTTE embarked on a plan of action to expand and consolidate the organisation (…). It was during this time a programme of political action was undertaken to mobilise, politicise and organise the broad civilian population towards the national cause (2004, p. 33).

In order to achieve civilian compliance the LTTE had to expand its coercive practices, intentionally, towards other ‘softer’ measures of persuasion: legitimation strategies. With a combination of coercion and legitimation strategies, compliance becomes what Podder refers to as ‘quasi-voluntary’, which is ‘compliance that is motivated by a willingness to comply but backed up by coercion, in order to ensure that others will obey the ruler’ (Podder 2017, p. 688).

The LTTE – War, Discipline and Governance

The conflict between the government of Sri Lanka and the LTTE surfaced internationally as an overt, violent conflict in July 1983, when 13 Sinhalese soldiers were ambushed in Tirunelveli in North Sri Lanka by the LTTE. Riots broke out in Colombo killing thousands of Tamils (estimates go up to 3,000 casualties) and damaging the livelihoods and homes of probably 30,000. An estimated 100,000 Tamils were displaced in Colombo and 175,000 fled abroad.

The July riots in 1983 were a watershed in many ways. Wickramasinghe describes the repercussions of ‘Black July’:

The riots of 1983 left a lasting imprint on the collective consciousness of the Tamil people. For many it led to exile and refuge in foreign lands, for others to a heightened sense of alienation from the state that spawned radicalism; for others it led to an erasure of identity, a refusal to be incorporated in a given identity. (…) The events of 1983 made ‘terrorists’. For the insurrectionist groups they were a bonanza in that their ranks suddenly multiplied. (2006, p. 287)

Black July and its aftermath started a full-blown war that was to last over 25 years. The fight by the rebel groups between 1983 and 1987 became known as Eelam War I. Three other ‘Eelam Wars’ were to follow. Over the years the LTTE forcefully gained dominance over the other Tamil militant groups to the point where it claimed to be ‘the sole representative of the Tamil-speaking population’. In order to reach that position, it decimated competing groups. Hundreds of cadres from these rival groups were killed from 1986 onwards (Wickramasinghe 2006, p. 289).

Between the 1980s and the end of the war, the military wing of the LTTE transformed from a guerrilla organisation to a type of regular army with a conventional fighting force (Richards 2013, pp. 16–17). By 2002, the military chain of command consisted of a deputy commander under Prabhakaran and a number of special commanders in charge of specific subdivisions including the LTTE army (the ground forces), the navy (the Sea Tigers), the air force (the Air Tigers) and the LTTE intelligence wing, as well as the Black Tigers, the LTTE’s suicide commandos (Richards 2013, p. 16). The LTTE had a disciplined structure with Prabhakaran as its supreme leader. The chain of command was authoritarian and hierarchical. All cadres were obliged to swear an oath of allegiance to the struggle for Tamil Eelam, and just as importantly, to Prabhakaran himself (Hussain 2010). Nothing significant in the LTTE happened without Prabhakaran’s knowledge (Swamy 2002, p. 319).

Discipline in the LTTE military was extremely rigid. The LTTE cadres were obliged to follow a strict moral code of conduct, which included prohibition of cigarettes and alcohol, while relationships with the opposite sex were also regulated (Samaranāyaka 2008). Even a simple smile or wave to the opposite sex could lead to punishments from superiors. ⁷ In addition, LTTE cadres were not allowed to get ‘too friendly’ with civilians. ⁸ The performance of cadres during battle was closely monitored and any unsatisfactory results and breaches of the codes of conduct would be punished. ⁹ To avoid any leaks of information, cadres that passed LTTE military training would be provided with a cyanide capsule that they were instructed to swallow before being captured by the Sri Lankan Army (Roberts 2007).

In areas under its control (particularly in ‘the Vanni’), ¹⁰ the LTTE set up its own administrative structures, such as the police, the judiciary and tax collection, while allowing the continuation of the Sri Lankan state services in the general administration and the provision of social services. These state services, however, functioned in close collaboration with, if not under the complete control of, the LTTE. As Kasfir points out ‘the best example of the opportunity to elicit popular support on a patterned basis is the establishment of a zone safe from enemy attack in which civilians are not caught between two competing authorities’ (Kasfir 2002, p. 7). In similar terms, Kalyvas points out that the possession of territorial control creates opportunities for rebel groups to generate loyalty that are not possible without a level of territorial control (2006, p. 128). The Vanni was exactly such a zone of full territorial control. The LTTE was able to establish and control a ‘base’ area up until the last years of the war in 2008/2009. ¹¹ This was also facilitated by a cease-fire agreement (CFA) brokered by the Norwegian Government that was officially in force between 2002 and 2007. The agreement consolidated and ‘froze’ the territorial control of the LTTE. Though the CFA would have to be followed up by negotiations on a political solution, several rounds of talks soon ended in failure.

After the defection of Eastern Commander Karuna and his followers from the LTTE in 2004, the balance of power in the Eastern part of the island shifted to the Sri Lankan Government, while the CFA slowly unravelled due to countless violations by both sides. In 2007, President Rajapakse was able to regain military control of the East and evict the LTTE. After that, the tables turned against the movement. The Sri Lankan Government formally abrogated the CFA to pursue a military solution to the conflict. A bloody military campaign with high military and civilian losses, initially against fierce LTTE resistance (Eelam War IV), finally led to the defeat of the LTTE in May 2009 and the complete extinction of its military and political leadership, including supreme leader Prabhakaran.

Coercive Techniques

As indicated above, the LTTE partly relied upon coercive techniques to induce compliance from the civilian population. As Kasfir explains, coercion can form the basis of a rebels’ political order if civilians consistently obey rebel rules because they fear the consequences, with the rebel leaders commanding their compliance on a regular basis (2015, p. 38). Although many Tamils felt a deep sense of loyalty to the LTTE, the people would often talk about the Tigers in a nexus of respect and fear. LTTE violence and intimate surveillance were palpable in the everyday imagination of Tamil communities in the North (Bremner 2013, p. 48). Coercion and the use of force helped bring new recruits, including children, into the LTTE (Becker and Thapa 2004, p. 5), particularly in the Eastern Province. If a family resisted, threats and harassment could follow (Becker and Thapa 2004, p. 5). Towards the end of the war, forced recruitment and abduction became commonplace. Harendra de Silva, chair of the National Child Protection Authority, estimated: ‘in 1994, I found that one in nineteen child recruits was abducted (…) in 2004, the reverse is true and only one in nineteen is a volunteer’ (Becker and Thapa 2004, p. 16). This shows that the LTTE relied upon coercive techniques and legitimation strategies from the beginning until the end, but the balance between the two shifted over time. Particularly in the last years increased military pressure led to additional forced recruitment of fighters.

The LTTE communicated its demands through letters, house-to-house visits, radio announcements and community meetings. Refusals led to more coercion, including threats of violence, the burning of houses and abduction (Becker and Thapa 2004, p. 16). As one of our respondents from Kilinochchi puts it:

the ‘one person per family’ rule was prescribed in order to increase manpower for the war and they forcibly recruited the youth into the movement. They kidnapped people and they imposed compulsory taxes. They collected funds forcefully, for example by abducting people and demanding ransom. They abducted or murdered people who were not in favor of the movement. ¹²

As Kasfir further observes, rebels usually force civilians to ‘choose between them and the existing government’ (2002, p. 6). As a consequence, choices made by civilians during a guerilla war are never completely free from coercion, and hence, some level of coercion will always underlie rebel governance (Kasfir 2005, p. 272). The LTTE’s battle for a Tamil nation was a homogenising, constraining discursive process that made it difficult for ‘regular’ Tamil people to dissent (Brun 2008, p. 420). There was particularly little room for criticism from the civilian population. ¹³ A shopkeeper from Mullaitivu (in the Vanni) told us: ‘We were compulsorily bound to our flag, our symbol, our rule, our administration and so on’. ¹⁴ Another respondent stated:

There was a gap between the citizens and the [rebel] state, so the citizens were not able to influence much of the rebel rule, as it was no democracy. It was imposed from above by the sole leader and the military and intelligence commanders. ¹⁵

The LTTE did not allow others much say into its affairs. By making itself the sole representative of the Tamil people, the LTTE silenced competing Tamil voices (Brun 2008). With its assumed sovereign right to express the will of ‘the Tamil people’, there was little room for collective action from the people itself, unless it matched LTTE strategic interests (Orjuela 2004, p. 177) or was orchestrated by them.

The use of coercive techniques also became apparent from the LTTE’s choices of identity exclusion. The Sinhalese, and at a later stage the Muslims, were excluded from the LTTE’s governance project and the nascent LTTE state. Muslims (as another Tamil speaking community) and ethnic Tamils (Hindu or Christian) have strong historical and linguistic ties as communities living in adjacent areas (Sitrampalam 2005a). Initially, the LTTE comprised both Tamils and Muslims, but later they started to exclude Muslims and even forcefully expelled them from the Northern Province in 1990. A respondent of this study from Oddamavadi, a city in Batticaloa district explains:

In the early years there also were Muslim LTTE cadres. But what they did after a while is killing their own Muslim cadres. Also those Muslim leaders that had a good reputation and gained support among the Muslim community here. ¹⁶

Other respondents also explained how the LTTE pushed the different communities further apart and exacerbated minor tensions that existed between them. This was partly due to the suspicion of the LTTE towards the Muslims. As one of the Tamil respondents explains:

Muslims were mainly living in army-controlled areas. So Muslims had contacts with the army. If we as Tamils had contacts with the Muslims, the LTTE would not like that, they would also suspect you. They would be afraid that you would pass on certain information to a Muslim who would pass that information on to the army. So the LTTE clearly created another enmity between Tamils and Muslims. That is what pushed us further apart. ¹⁷

Apart from this friction between Tamils and Muslims, the unity of Tamil nationalism did also get fractured in the Eastern Province throughout the 1990s and 2000s (Fuglerud 2009, pp. 204–206). Within the Tamil population the social division between the North and East is an old one, based on variations in caste and cultural traditions. In the East, there is an image of the Jaffna Tamil being arrogant and high-handed (Fuglerud 2009, p. 206). This fraction became prominently apparent with the Karuna split when eastern Tamils shifted away from the LTTE.

Legitimation Strategies

Legitimation strategies are employed by rebels to guide and inspire civilians living under their auspices. As such they provide an interpretation of events taking place, and they organise individual and collective experience in rebel-held territories. Rebels use legitimation strategies deliberately to justify their actions and to express their identities. As one of our respondents explained, the LTTE had slogans about the struggle like: ‘“we want our land back”, or “we have to fight”’. ¹⁸ He further elaborates: ‘they wanted to boost their image, and they wanted to get that feeling in the community. They did it very deliberately’. ¹⁹ Below we analyse the LTTE’s legitimation based on Tamil nationalism, tradition, charismatic leadership, enemy images and sacrifices of LTTE cadres. For each of these forms of legitimation, we explore the effects on civilian compliance and the perceptions of civilians inside LTTE-territory.

Tamil Nationalism

Contestation in civil war often centres around identity and the politics of ‘inclusion and exclusion’ (Demmers 2016). This regards the question of ‘defining who belongs and who does not belong to the nation (state), who is indigenous and who is foreign’ (Hagmann and Péclard 2010, p. 554). The most common currency of violent imaginaries in late modernity are nationalism and/or ethnicity (Schröder and Schmidt 2001, p. 11). The LTTE used both as common identity markers for their envisaged Tamil Eelam – the independent Tamil state in the North and East of Sri Lanka that it insisted had to be established in order to protect the rights and security of the Tamil-speaking community. ²⁰ In its political programme, the LTTE styles itself as ‘the national liberation movement of the Eelam Tamils waging a relentless military and political struggle for the total independence of the Tamil homeland’ (LTTE 2005, p. 292). It elaborates:

The LTTE holds the view that Eelam Tamils possess all the basic elements that define a concrete characterization of a unique nation. We have a homeland, a historically constituted habitation with a well-defined territory embracing the Northern and Eastern Provinces, a distinct language, a rich culture and tradition, a unique economic life and a lengthy history extending to over three thousand years. As a nation we have the inalienable right to self-determination. (LTTE 2005, p. 299)

This emphasis on protecting and liberating the Tamil nation is a salient, recurrent theme in the LTTE’s discourse and propaganda. In its struggle for a hegemonic position in the North and East the LTTE assumed the authority of defining what ‘Tamilness’ was, and presented itself as the ‘sole representative’ of the Tamil nation (Brun 2008, pp. 417–419). Symbols of the movement and symbols of the struggle for the independent state with a distinct national identity were conscious and well-planned components of the LTTE’s symbolic repertoire that evolved over time.

How did those living under LTTE rule perceive these symbols? At meetings, they sang the Eelam national anthem (‘Look the Flag is Rising’). A national bird, a national tree and a national flower also symbolised the separate nation. These symbols were known and respected by the public. According to a respondent from Kinniya: ‘There was absolute royal respect for the national Tiger flag and the national flower, the November flower. We also saluted and worshipped these. We willingly did so by ourselves’. ²¹ Being Tamil always involves some negative image of what a Tamil is not: Sinhalese (Brun 2008, pp. 411–412). Several of our respondents that lived in LTTE-controlled areas asserted that despite its authoritarian character they actually felt better represented under LTTE rule than under the Sri Lankan Government. In contrast to the government, the LTTE evoked feelings of attachment to the Tamil communities, based on family bonds, and ethnic identification. The LTTE cadres would be sons, daughters, cousins, brothers, sisters or former schoolmates to them, even though many were forcefully recruited. Tamil communities opted for the LTTE even if they would not (fully) endorse their modus operandi. ²² As one of our respondents in Jaffna points out: ‘There were several negative things as well, but we respected that they were fighting for us, we respected that they were sacrificing their lives for us. Not everything was right, but we accepted that’. ²³

It has to be noted, however, that identity categories themselves are not static or fixed, but change over time, particularly during war (Butler 1997, p. 99, Thiranagama 2011). Due to the LTTE’s rule the meaning of being Tamil changed in various ways and forced each Tamil to choose their loyalties. As Thiranagama (2010) explains, the LTTE was deeply concerned about ‘traitors’ within the Tamil community. Prabhakaran noted in a Heroes Day speech: ‘The traitor is more dishonorable than the enemy’. The boundaries of the categories that defined treasonous acts in areas under LTTE control grew wider creating an ever-looming possibility of destructive punishment, which Thiranagama refers to as the ‘intimate power’ that the LTTE possessed towards both the cadres and the Tamil civilians (2010, pp. 139–140). In other words, the LTTE offered its subjects unity and meaning through the Tamil nation, but at the same time, it possessed an unbounded force of coercion to punish those Tamils who did not show the behaviour that the movement demanded (see also Klem and Maunaguru 2017, p. 648).

Traditional Sources of Legitimacy

As Weber observed the ‘validity of a social order by virtue of the sacredness of tradition is the oldest and most universal type of legitimacy’ (1978, p. 37). The command of rebel leadership and the obedience of civilians living under rebel rule are then legitimised on the basis of the sanctity of immemorial traditions (Weber 1978, p. 215). References to common descent and the continuity of a collective memory were used by the LTTE leadership to legitimise its struggle for Tamil Eelam.

As Schalk rightfully notes there is a difference between ‘historisation’ for the historian and the activist. For the historian ‘historisation’ is often the conscious attempt to ‘range a phenomenon into a continuity consisting of periods, or more generally the past and present’ (Schalk 1997, p. 35). For the activist who wishes to support his or her political and/or religious movement, however, ‘historisation’ may become a conscious attempt to use selected fragments of history to fulfil certain strategic interests. In similar terms Schröder and Schmidt point out how ‘elements of history are decontextualized and reinterpreted as part of a communal legend’ that serves to legitimise violent confrontations, thereby creating an ‘imaginary of internal solidarity and outside hostility’ (2001, p. 11). The LTTE claimed that their ideology was rooted in Tamil tradition and culture. However, they gave their own specific reading of that tradition and culture, embracing or emphasising particular historical elements and de-selecting or rejecting others (Schalk 1997, p. 35).

Schalk (1997, p. 37) states that the ideology of the LTTE was presented as a traditional one in the teleological sense: Prabhakaran saw the LTTE ideology and struggle as continuing in a long historical martial tradition. The name and logo of the insurgency, the ‘Tigers’, is an example. LTTE members would generally explain that the ‘Tiger’ refers to the old royal emblem of the Chola Kings (Hellmann-Rajanayagam 1994, p. 56). In the collective memory, those were great imperial Kings under whom Tamil culture and Tamil power expanded and flourished (Hellmann-Rajanayagam 1994, p. 56). Songs about the LTTE connected the greatness of Prabhakaran with these Kings. The LTTE created an image of Prabhakaran as able to step into their footsteps and conquer the north-east of Sri Lanka. The geographical borders for the imagined Eelam, the ‘historical’ homeland of the Tamil population, coincided with the areas where the Tamil-speaking community historically lived (Hellmann-Rajanayagam 1994, p. 57, Sitrampalam 2005b). In that sense, the LTTE presented the ‘greatness’ and ‘sanctity’ of these ancient Tamil Kingdoms in order to legitimise its claim to liberate and govern the Tamil population.

However, the ideology of the LTTE was not completely ‘traditional’, as it included revivalist and several contemporary elements (Schalk 1997, p. 40). Anti-colonial ideologies of Subhasism and Dravidism became main components, flavoured with elements of socialism. The LTTE’s ideology was also an ‘incitement to dedicate life to a cause’ and ‘a moral doctrine that teaches discipline’ (Schalk 1997, p. 54). As Schalk explains:

It is not an ‘out-pouring’ of a Tamil martial tradition. It projects roots, but it is not dependent on pre-colonial roots. (…) [Prabhakaran] sees the fighters as (free) cutantira cirpikal, ‘artisans of independence’. I can see that their making use of the past is pragmatic and instrumental. The question they put is: What do we need of the past for shaping the present and the future? (Schalk 1997, p. 40)

However, even if the historicity of some of the claims by the LTTE might seem doubtful, they resonated within Tamil society. As Schröder and Schmidt highlight, the ‘symbolic meaning of prior wars is re-enacted and reinterpreted in the present, and present violence generates symbolic value to be employed in future confrontations’ (2001, p. 9). This legitimation strategy was particularly visible in notions such as ‘sacrifice’, ‘martyrdom’, ‘our soil’ and ‘the motherland’. As Hellmann-Rajanayagam points out:

They [the LTTE] are seen as the legitimate, traditional protectors of the people, because their rhetoric runs along lines which are (…) deeply entrenched and thus strike a related chord among the population: we protect you, your nation, your honour, your women. (Hellmann-Rajanayagam 1994, pp. 136–137)

Those living under the LTTE rule attended festivals and commemorations of the LTTE. One of our respondents noted that they ‘considered them as their tradition and as a part of their duties towards their motherland’. ²⁴ The LTTE used references to selected traditional elements in the Tamil society strategically, and those seem to have resonated among the civilian population. At the same time, however, the LTTE proposed several ways of living that broke with Tamil traditions, and entailed a different direction for social life under LTTE rule. These constituted a discontinuity with the past and were met with more resistance as discussed in the last section on the LTTE’s sociopolitical reform.

Legitimation Through Charismatic Leadership

The performance of charismatic leadership is another element legitimising a rebel group’s rule (Schlichte and Schneckener 2016, p. 418). Leadership was one of the factors to shape and secure the emergence and expansion of the LTTE as a rebel movement (Staniland 2014, p. 149). As Richardson notes, ‘Prabhakaran himself deserves much credit for providing the leadership skills and building the organization that achieved this transformation’ (2005, p. 479). But, as Weber reminds us, without the recognition of followers the leader’s special ‘gift’ of charisma does not exist. There is considerable evidence that many Tamil civilians considered Prabhakaran charismatic.

During our interviews, several respondents explained that they worshipped, and still worship, Prabhakaran:

We accept Prabhakaran as our leader. We are proud to use symbols where our son’s face is engraved. We […] stand up and salute when we see a picture of our leader. Our leader protects our motherland and the Tamil race. ²⁵

Another pointed out:

Although we did not have the privilege of seeing our leader personally; an order from him is sufficient for us. We would give our life in that place. We will bow our heads to the Tiger flag. We salute our heroes. We worship the feet of our leader. ²⁶

More emphatically, ‘We respected our leader’s picture as our God’. ²⁷ With regard to heroes’ day, a respondent said: ‘We would not have redemption until our leader comes and talks to us’. ²⁸ The leadership of Prabhakaran provided a solid point of reference for the Tamil community. From the early days until the end of the war, the Tamil leader possessed a mythical status and guided the Tamil community in the northeast. In 2016, several respondents even stated that Prabhakaran was still alive and would return one day.

Legitimation Through Sacrifice and Martyrdom

Reports of sacrifices of cadres were extensively propagated to induce compliance with the LTTE’s cause. The death of a martyr may reaffirm the cohesion of a particular group, may legitimise the group’s convictions and strengthen their self-respect (Hellmann-Rajanayagam 2005, p. 115). The sacrifices made by LTTE cadres were an important element in the LTTE’s symbolic portrayal of its struggle. ‘Martyrdom was a critical factor in drawing popular support among the Sri Lankan Tamil people. [The…] devotional commitment indexed by the suicidal act was evaluated highly […] by many a Tamil person’ (Roberts 2009, p. 222).

On the individual level, new forms of Tamil identity came into existence. In the statemaking endeavour, new routes to status and power through self-images of glorious heroism emerged around those who were willing to sacrifice their lives for the Tamil nation (Thiranagama 2011, pp. 59–63). At the same time, this put a high pressure on those Tamils that did not join or support the LTTE, shifting them into identities of ‘cowardice’, ‘selfishness’ or even ‘traitors’, who chose themselves over the greater good of the Tamil nation and collective sacrifice (Thiranagama 2010, 2011, p. 38). Countless commemorations of fallen cadres took place in which these heroes or martyrs were accorded a type of ‘sainthood’ (Bavinck 2014, p. 11). Bavinck noted in his diary in Jaffna in 1994:

For the Tigers, the practice of declaring special days of mourning or celebration forms a means to involve the general population more closely in their cause. Because of their complete control they succeed in this. Today I saw the plaited pal leaves and the black flags, which are the usual sign of such days, at several places in town. (2014, p. 15)

Another excerpt in Bavinck’s diary concerns Prabhakaran’s appreciation of the Black Tigers:

Today is the day of the Black Tigers, the suicide commandos. Everywhere one sees flags, posters and little commemoration chapels with photographs of the fallen heroes. Loudspeakers are blaring martial music all day long. The great leader wrote an article in the papers in which he very idealistically calls these Black Tigers spiritual heroes, who give their lives for others. (2014, p. 20)

How did the civilian population respond to this legitimation strategy? One of our respondents explained: ‘The fallen cadres were commemorated every year, and we would also go there every year to commemorate them’. ²⁹ In reference to these commemorations another respondent mentioned, ‘[The Tamils] considered it as their tradition and as a part of their duties towards their motherland’. ³⁰ The respect that the cadres and their families earned for their readiness to sacrifice their lives for a common cause created legitimacy among the Tamil constituency.

These performances helped construct and maintain a political community, which could be seen as a type of nationbuilding (Hellmann-Rajanayagam 2005, p. 117). One of our respondents elaborated this point:

During public events, Tiger flags were raised and symbols were worshipped. People were proud of this. All of us should salute the national flag of our soil. It is our duty and responsibility. Isn’t it one of our special duties to perform poojas to the pictures and memorials of our heroes who sacrificed their lives for our race? People got involved in this activity with a lot of interest and enthusiasm. ³¹

Several other respondents also emphasised this sense of community with regard to the LTTE heroes:

All the heroes are our brothers. So, the people and the Liberation Tigers together celebrated it impressively, displaying pictures and lighting lamps in the houses. The skills, abilities, and achievements of the heroes were revealed on the heroes’ day. It is a day to be proud of. ³²

However, this legitimation strategy did not work effectively under all conditions. Though sacrifices often served to legitimise the LTTE’s struggle against the Sri Lankan Government, Tamils questioned them too. In 2004, for example, resistance against forced recruitment led to struggles between the LTTE and the Karuna faction that had split off (Sanchez Meertens 2013, pp. 104–105). Some Tamil parents actively opposed recruitment of their children for this internal battle (Sanchez Meertens 2013, pp. 104–105). In other words, while the willingness of the LTTE to fight the Sri Lankan Government provided them with legitimacy, people were reluctant to sacrifice their children for a fight against the Karuna faction.

The exodus from Jaffna in 1995 provides another example. Civilians soon realised the LTTE did not continue to defend the people’s homes on the Jaffna Peninsula. Instead, the LTTE withdrew and forced the civilian population to come with them (Thiranagama 2011, p. 67). This led to disappointment and resistance against the movement. As the University Teachers for Human Rights (UTHR) shows:

In one of those long bread queues in Chavakachcheri one man blew up: ‘We are being treated as slaves. If this is their behavior now, how would it be when we get Eelam?’ Unlike in other times (…) others joined in. (…) the [LTTE] police arrived and ordered the first speaker to get into their vehicle. The man was vocal in his refusal. Finally he was dragged inside. (UTHR in Thiranagama 2011, p. 68)

Civilians perceived the move to withdraw completely from the Jaffna peninsula as selfishness of the movement, and not as a sacrifice. People openly criticised the exodus and the capacity of the Tigers to establish and run its Tamil Eelam – behaviour hardly ever shown in public earlier.

We must therefore conclude that the sacrifices of the LTTE cadres indeed served to legitimise the LTTE rule, but only to the extent that they were made in the struggle against the Sri Lankan Government. When such sacrifices served the ‘motherland’ and the ‘Tamil nation’, they were rooted within in a historical collective memory of the Tamil population. When they were not, they weakened the legitimacy of the Tigers.

Legitimation by Invoking Outside Threats and Enemy Images

Many Tamils witnessed or suffered abuses by Sri Lankan security forces. Staying within LTTE territory also meant that Tamil civilians would experience the shelling by the Sri Lankan forces. Furthermore, government abuses prior to the cease-fire included unlawful detention, interrogation, torture, execution, enforced disappearances and rape (Becker and Thapa 2004, p. 5). Propaganda films produced by the LTTE portrayed the soldiers of the Sri Lankan Army as alcohol abusers who dance while the Tamil people are suffering (Brun 2008, p. 407). Most films also included a history of the movement and the portrayal of the battle sites. They showed the LTTE’s achievements in the violent struggle and the atrocities against Tamils carried out by the Sri Lankan army (Brun 2008, p. 411). In many films, the Black July pogrom of 1983 was cited as the reason for the LTTE’s struggle against the Sri Lankan state (Brun 2008, p. 411).

The LTTE decided what people were allowed to watch and what was forbidden. Only limited media was available at the time in the Vanni, and it was controlled by the movement. ³³ As one respondent observed, ‘There was hardly any access to other media, we had to watch it. But like I said, most people also liked watching it’. ³⁴ Through its control over the propaganda channels the LTTE possessed what Demmers refers to as the ‘power to define’ the ‘legitimate’ course of action (2016, p. 137). The use of symbols and propaganda produced and reproduced a particular narrative to which no alternative was allowed. To disseminate these images the LTTE managed websites, a radio and propaganda machinery and organised parades, theatre and music shows. One of our respondents explained:

The LTTE had their radio channel in Jaffna, Mullaitivu and Kilinochchi. (…) They were doing documentaries, films, and short films. They would release good films that would describe what was being achieved. That was the propaganda. These films were also taken on the battlefield. (…) They would show it very fiercely, you know, they take all the Hindu songs and music, and they would get very vibrant. ³⁵

How did the civilian population respond to this legitimation strategy? The history of abuses helped the LTTE to build and sustain credible enemy images of the Sri Lankan security forces that resonated strongly among Tamils. By presenting the Sri Lankan forces as brutal and inhumane, the LTTE tried to show that its own violence was needed and legitimate. A 1993 study of adolescents in Vaddukoddai in the North, for example, found that at least one quarter of the youths had witnessed violence personally. In response, many persons joined the LTTE, seeking to protect their families or to avenge real or perceived abuses (Becker and Thapa 2004, p. 5). As one of our respondents insisted:

The LTTE was fighting to fulfil the aspirations of the Tamil people and seeking their freedom. People accepted them. The injustices and atrocities of the Sinhala army were increasing day by day. Under such a situation, the aim and activities of the Tigers were reasonable. ³⁶

Due to the ongoing hostilities the credibility of this threat was high. One of Thiranagama’s respondents states:

The Sri Lankan air force was bombing us. People were dying and the houses were being smashed. The Navaly Kovil (temple) was hit. In this situation when you see the bodies of those who have been killed brought past you, when you see them, then you feel like I must join the LTTE and fight against the Sinhala army and the Sinhala government. I thought about it, I came close to the decision to join. ³⁷

Most of the residents of the North and East would have witnessed the consequences of the Sri Lankan army’s attacks on people close to them (Thiranagama 2011, p. 56). Consequently, the LTTE’s invocation of outside threats and enemy images seems to have worked effectively as a legitimation strategy. They related to the experiences of those living in LTTE territory. Through the media, LTTE images stressing outside threats persuaded civilians that protection against the Sri Lankan forces should remain in the hands of the movement.

Sociopolitical Reform: Gender Roles and Caste – Challenges to Legitimation Strategies

The LTTE also imposed sociopolitical reforms on the Tamil society. These reforms were not legitimation strategies intended to persuade the civilian population, but reflected on the one hand ideological convictions of the leadership and were on the other hand strategic choices to increase its opportunities to include a larger group of the Tamil community into the movement, in both civil and military functions. From the 1990s onwards the LTTE imposed new rules regarding gender and caste. Through this policy they included both women and civilians from all castes as potential members of the movement. In addition, the LTTE envisaged Tamil Eelam as a ‘people’s state’ that was secular, democratic and based on socialist principles. This state was to be created by the will of the people and administered by the people (LTTE 2005, p. 308). The objectives of national liberation and social revolution were to ‘integrate both national struggle and class struggle, interlink both national emancipation and socialist transition of our social formation into a revolutionary project aimed at liberating our people from both national oppression and social tyranny’ (LTTE 2005, p. 309).

The LTTE claimed that it had ‘inspired and awakened the Eelam national consciousness and organized and united all freedom loving, patriotic sons and daughters of Tamil Eelam to fight for the cause of national liberation and social emancipation’ (LTTE 2005, p. 308). Hence, the LTTE’s ideology combined both (LTTE 2005, p. 308). They referred to the ‘principle of self-determination’ with regard to the first, and to ‘socialist transformation’ with regard to the latter. The LTTE presented their objective as a fight against ‘oppression’, ‘exploitation’ and ‘social injustice’ that characterised the traditional Tamil society. While following certain traditional lines, the LTTE also deliberately attempted to reform Tamil society. Though members of the lower castes, youths and women could acquire new positions in this envisaged society, for others (males, higher castes, older generations) these reforms threatened their positions in society and the status quo.

Some of these reform policies met resistance from Tamils. The LTTE was involved in a battle on two fronts during the early 1990s – the fight within Jaffna society and the fight against the Sri Lankan army (Hellmann-Rajanayagam 1994, p. 142). The LTTE confronted this challenge head-on. Schalk similarly observes:

[The LTTE] grew up with this innate conservatism which certainly became an obstruction and obstacle to the changes that the LTTE wanted to introduce in Yalppanam [Jaffna]. The LTTE went against this mainstream conservatism. I observe especially its radical program for the benefit of women that indeed would change fundamental structures of Yalppanam society, like the caste system, legal system and the negative influence of religion. (Schalk 1997, p. 39)

Despite the LTTE’s references to the traditional ‘roots’ of its struggle and its ideology, its ideology included in fact a number of very ‘untraditional’ reformative sociopolitical goals. For example, age is respected in traditional Tamil society and older people in society had difficulties with the disrespect that the young LTTE cadres expressed. As one of our respondents pointed out:

I had a friend, a teacher that was living in the Vanni for a long time. He was a respected person. But since the LTTE rule youngsters would come by and show no respect to him. You know they had their guns around their necks and were thinking that they could do anything. For example they talked to him very impolitely: in singular person instead of plural. ³⁸

He adds that ‘the younger people were making the rules and were not always respecting the authority of the elderly people in the communities’. ³⁹

Contrary to more traditional roles of women in the Tamil society, the LTTE actively promoted the emancipation of women. With regard to women’s emancipation the LTTE political programme states:

Tamil women are also victims of social oppression, oppression generating from their own social structure. The notorious practice of the dowry system, the male chauvinistic domination and other forms of socio-cultural practices that degrade women and deprive them of human dignity and equality are typical features of our repressive social system. Our liberation movement is committed to the emancipation of women from national and social oppression. The LTTE assures that the dowry system will be legally proscribed and the equality of status and opportunity for women will be constitutionally guaranteed. (…) Education will be made compulsory for all girls and the practice of sexual division of labor will be abolished. (LTTE 2005, pp. 312–313)

There are indications that the movement worked with relative success on a number of these issues. LTTE female cadres eventually went through the same training as their male colleagues and were fully deployed in offensive missions and as suicide commandos. Their attire and behaviour were far removed from those of a traditional Tamil girl from Jaffna. It is debatable whether this can be called emancipation or is in fact a form of subjugation to a male, nationalist and oppressive order (Frerks 2016, pp. 93–113, see also the references mentioned in note 7). Samarasinghe contends that the LTTE started to recruit women because of a shortage of male recruits, while the liberation struggle was simultaneously presented by the LTTE as a strategy for emancipation (1996, p. 214). In particular, some of the more conservative families were displeased with the new roles their daughters would have to take up in society.

Furthermore, caste relations came under strong LTTE surveillance. While the LTTE was breaking the relations between Muslims and Tamils, it deliberately attempted to overcome the existing boundaries between the different Tamil castes. Without the Muslims, the envisioned LTTE state was to become united as one Tamil community regardless of the salient pre-war caste differences among various groups of Tamils. As Bremner puts it, a ‘new history of a homogeneous nation was being superimposed upon existing relations’ (2013, p. 48). For example, to enforce the homogeneity of the Tamil nation, the LTTE encouraged Tamil women to marry Tamil, and not Sinhalese or Muslim men. ⁴⁰ Aside from that, in LTTE law discrimination on the basis of caste became a crime. Relatives that obstructed inter-caste marriages would be punished (Bremner 2013, p. 48).

Initially these caste policies were met with resistance from within Tamil society, particularly from the Vellala caste. As one of our respondents puts it,

in those times […] we had a strong caste system in our society. Prabhakaran is Karayar, that means he belonged to the fishermen. But some other parties were from the upper class, the Vellala. So half the people, the upper class, did not like Prabhakaran. ⁴¹

He adds:

It [caste] was not erased. (…) the people they did not change, they have the mentality within them. This mentality of how can we accept a Karayar as our leader? That is their idea, but not all people would say that. ⁴²

Although the LTTE did not fully end caste, it was able to affect a much higher degree of mobility between castes than was possible before LTTE rule (Malathy 2013, Manoharan 2015). This policy resulted in significant representation of lower castes in the administrative, political and military wings of the LTTE (Manoharan 2015, p. 177). Manoharan speculates that the LTTE could have created a casteless society if it had continued to survive as a rebel ruler (2015, p. 178). The LTTE framed their mission so that it would not create enmity between the castes, and that the movement would be able to overcome caste differences. Though this did not appeal to all castes equally, it seemed to have served to avoid internal caste struggles within the LTTE. The LTTE consciously accommodated higher castes to avoid Vellala resistance.

Overall, we must conclude that the reform agenda of the LTTE hardly helped legitimise the LTTE or bring its envisaged Tamil Eelam closer. Rather, reform was part of the LTTE package that had to be accepted and followed. Though the reforms may have appealed to a small section of the Tamil population, generally it was not the reforms that helped the LTTE to legitimise their rule. Instead, legitimation was enhanced by claims surrounding common Tamil descent and a common Tamil nation, sacrifices, martyrdom, charismatic leadership and the LTTE as protector from Sinhalese threats. The LTTE created a degree of compliance despite their proposed reforms, not by virtue of them. Yet, despite the resistance from within Tamil society, the reformative agenda provided the LTTE with some strategic advantages. It gained a larger fighting force by including Tamil women and Tamils from all castes. As they tended to be less conspicuous, women in particular provided strategic advantage to carry out activities as couriers, spies and reconnaissance soldiers (Frerks 2016, p. 105).

Conclusion

We investigated the case of the LTTE to see how the movement legitimated its rule and how its struggle for Tamil Eelam enhances our understanding of the legitimation aspect of rebel governance. Our study confirms Kalyvas’ argument that control creates possibilities for rebel groups to generate loyalty that are not possible without a level territorial control (2006, p. 128). Territorial control provides a rebel group with the opportunity to create a ‘tacit social contract’ with the civilian population. In the case of the LTTE the population may not have endorsed their modus operandi completely, but Tamils in LTTE-controlled territory accepted its rule and struggle as unavoidable vis-à-vis a Sri Lankan state that was invariably seen as inhumane and oppressive. LTTE propaganda actively tried to portray an enemy image of the Sri Lankan state that resonated with the people’s own experiences. Beyond the rebel governance literature that has focused on the merely instrumental exchange of public services in exchange for types of civilian compliance (Weinstein 2006), this article has shown that symbolic repertoires of nationalism, tradition, charismatic leadership and the portrayal of sacrifices and martyrdom may significantly shape the character of rebel–civilian relations and engender legitimacy. The devotion of Tamil civilians to their leader is quite remarkable compared to other cases of rebel governance. Prabhakaran seemed to possess a type of God-like status. The martyr cult – through a mixture of coercion and persuasion, provided the LTTE a devoted segment of followers that were not necessarily only after future material gains.

Second, our study testifies to Kalyvas’ argument that civilian compliance during civil war is usually based on a mixture of coercion and persuasion; on fear and sympathy (Kalyvas 2006, pp. 101–104). Our empirical evidence showed that broad-based compliance of Tamil civilians under the LTTE rule was exactly based upon such a mixture. Moreover, we showed that persuasion took place through a variety of material and symbolic forms of legitimation. Successful LTTE legitimation was based on Tamil nationalism, the needs of Tamil civilians for protection against the Sri Lankan state, the performances of charismatic leadership and the portrayal of sacrifices by LTTE cadres. Symbolism, tradition and the mimicry of state performances were essential elements of legitimation. A salient finding of our field studies is that LTTE symbols, traditions and remembrances remain alive to this very day. They do not seem to have lost much of their appeal among the Tamil civilians, despite fundamental sociopolitical changes on the ground caused by the defeat of the LTTE. This observation shows that the LTTE’s repertoire of legitimation strategies was effective. It did not rely only on force and coercion, as these are absent now, while many of our respondents still talked about the LTTE with admiration.

Not all LTTE legitimation strategies were equally successful, however. Its call for sociopolitical reform of Tamil society did not appeal to most segments of the Tamil civilian population. In fact, these policies were only successful to induce compliance from a small highly dedicated segment of LTTE-supporters. This shows that even though rebel groups may indeed use previously engrained local cultural references as Mampilly (2015, p. 91) notes, rebel groups may nevertheless choose to use ideological expressions that do not generate broad-based appeal. From a more strategic perspective this testifies to what Kalyvas highlights as the ‘five percent rule’ (2006, p. 103), according to which only five percent of a civilian population is actually necessary as active militant supporters.

Our study confirms that legitimacy is an essential feature of rebel governance. Rebels will generally be able to govern effectively if there is a ‘tacit social contract’ between them and the ruled. Examining different ways of legitimation, including deliberate strategies, and civilian responses to these improves our understanding of how rebel governance affects the relations of rebels and civilians. The article has shown that in order to secure compliance, rebel groups must carefully balance between ‘change’ by means of a reformative agenda and ‘continuity’ by means of adapting to historically contingent values, norms and beliefs. Overstretching the reformative agenda may work counterproductive in order to secure a stable form of compliance. Overall, the LTTE’s legitimation strategies helped forge civilian compliance to its rule and its struggle for the imagined state of Tamil Eelam. Next to that, we also noted forms of legitimation which were not consciously driven by the LTTE, but emerged bottom-up or as a consequence of their acts or reputation among the population. In addition, legitimation took place in a war-related context characterised by violence and force. Whether these elements of the LTTE’s legitimation, based on history, tradition, charisma, symbolism and sacrifice, will be re-enacted, adapted, reframed or rejected in future contestation over Sri Lanka’s political dispensation is an important topic for future research.

Notes on Contributors

Niels Terpstra is a PhD candidate at the Centre for Conflict Studies, Utrecht University, the Netherlands. 

Georg Frerks is a professor of Conflict Prevention and Conflict Management at the Centre for Conflict Studies, Utrecht University, the Netherlands; and professor of International Security Studies at the Netherlands Defense Academy.

Acknowledgements

We are very grateful to Nelson Kasfir for his close reading and insightful comments on earlier versions of this article. Special thanks also go to Jonathan Fisher and Paul Jackson for their comments as editors of the Civil Wars journal and two anonymous peer reviewers who helped us improve the article significantly. We are grateful to the Netherlands Organisation for Scientific Research (NWO) and the Gerda Henkel Foundation for financial contributions in different stages of this research project. We thank our fieldwork assistants and translators in Sri Lanka for their dedication and hard work. We also thank Christopher Day, Jolle Demmers, Toon Dirkx, Mario Fumerton, Caroline Kennedy-Pipe, Romain Malejacq, and Nora Stel for their useful comments on this article.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported by the Gerda Henkel Foundation; and the Netherlands Organisation for Scientific Research (NWO).

Notes

1. The LTTE directly governed segments of the Muslim population until they expelled the Muslims from the Northern Province in 1990. In the Eastern Province, the LTTE presence had mostly an indirect impact on the Muslims, but they also faced direct attacks, including the Kantankuddy massacre on 3 August 1990, where over 100 worshippers were killed. For a detailed study on the expulsion of the Muslims from the North and its consequences see: Hasbullah 2004.

2. See the introduction to this Special Issue.

3. A telling quote by Anton Balasingham from Indian journalist D. Singh reads: ‘India’s “enemy” was everywhere and at all times; they were heroes of the people and came from the people; they were nurtured, harboured and supported by the local people’ (2004, p. 122).

4. The Sri Lankan conflict has been extensively documented in countless publications, i.a.: Frerks and Klem (2005), Goodhand and Klem (2005), McGilvray (2008), Peiris (2009), Spencer et al. (2015), Wickramasinghe (2006), Weiss (2011).

5. We follow Kasfir’s explanation here, see: Kasfir 2002, p. 4.

6. See also a recent special issue in Small Wars and Insurgencies: (Duyvesteyn 2017).

7. Even though Prabhakaran initially opposed LTTE revolutionaries to get married in the earlier stages of the rebellion, this changed when he fell in love himself in the mid-1980s. He got married to Mathivathani – also known as Mathy – in 1985 and the restrictions on love and marriage for LTTE cadres were loosened compared to the early days of the insurgency (Swamy 2003). There is, however, quite some controversy on the notion of the ‘armed virgin’ and the respective roles of female subjugation and empowerment in the LTTE nationalist armed struggle (see i.a. Coomaraswamy 1996, de Alwis 1998, Balasingham 2001, Frerks 2016, Sumathy 2004).

8. Interview code: KI 05 – Trinco.

9. Interview code: KI 17 – Jaffna town.

10. ‘Vanni’, sometimes spelled as ‘Wanni’, is used as a term for the mainland districts of the Northern province, namely Kilinochchi, Mannar, Mullaitivu and Vavuniya. The other district of the Northern province, Jaffna, is a peninsula. The term is often used (in interviews and in contemporary literature) with reference to the LTTE-controlled areas in the Northern Province, excluding the Jaffna peninsula.

11. Some geographical areas were under full LTTE control, but over the course of the various Eelam Wars frontlines and control over territory did shift back and forth, particularly in the East (Gaasbeek 2010, pp. 132–133). In some periods this resulted into a patchwork of LTTE-controlled territories, grey areas, government-controlled territories (Gaasbeek 2010, pp. 132–133, Korf et al. 2010, p. 393). These frontlines were not always fixed as the LTTE would generally move in during the night time when the Sri Lankan forces withdrew to their camps. In other words, the LTTE rule and opportunities to govern were not necessarily homogenous across time and space (Korf et al. 2010, Terpstra and Frerks forthcoming).

12. Interview code: 3 – NA3.

13. As several respondents in Batticaloa remarked: ‘they were reigning with the gun, so what could we do, it was our fate’.

14. Interview code: 23 – NC7.

15. Interview code: KI 17 – Jaffna town.

16. Interview code: KI 08 – Batticaloa district.

17. Interview code: KI 13 – Batticaloa district.

18. Interview code: KI 13 – Batticaloa district.

19. Ibid.

20. See Brun’s discussion of Anton Balasingham’s book War and Peace: Armed Struggle and Peace Efforts of Liberation Tigers. Balasingham claimed this book to be ‘the definitive history of the LTTE’s political struggle’ (Brun 2008, pp. 405–408).

21. Interview code: 27 C1 – Kinniya.

22. A general attitude that we observed during our fieldwork in the Northeast was: ‘as a Tamil person I could not be against them, although I did not necessarily agree with them’.

23. Interview code: KI 17 – Jaffna town.

24. Interview code: 2.4 – Pudukuduyirippu.

25. Interview code: 29 – C3.

26. Interview code: 23 – NC7.

27. Interview code: 30 – C4.

28. Interview code: 34 – C8.

29. Interview code: KI 20 – Jaffna town.

30. Interview code: 3.2 – Kilinochchi.

31. Interview code: – 22 NC06 – Mullaitivu.

32. Interview code: 10 NB2 – Mankulam.

33. Interview code: KI 05 – Trincomalee town.

34. Ibid.

35. Ibid.

36. Interview code: 15 – NB7.

37. Respondent ‘Anthony’ in Thiranagama (2011, p. 56).

38. Interview code: 2016 – K05. In Tamil polite speech requires use of the plural when a younger person addresses an older respected person.

39. Interview code: 2016 – K05.

40. Though it has to be noted that this was not a particular policy of the LTTE. Throughout the various community groups in Sri Lanka it is generally not strongly encouraged by the family to marry outside one’s own ethnic group, caste or religion.

41. Interview code: 2016 – K01.

42. Ibid.