Version classiqueVersion mobile

Politique et religions en Asie du Sud

Christophe Jaffrelot
Aminah Mohammad-Arif

Minorités religieuses

Muslims and ethnic identity in Sri Lanka

Les musulmans et l’identité ethnique à Sri Lanka

Sharika Thiranagama


Les musulmans sri-lankais sont la seconde minorité tamoulophone la plus importante à Sri Lanka, représentant environ 8 % de la population. Ils sont aussi le second groupe ethnique d’habitants permanents des régions de langue tamoule du Nord et de l’Est, autrefois objet de litige, zone de combat durant 25 années de guerre civile achevée abruptement et dans le sang en 2009. Des vies musulmanes sri-lankaises ont été inextricablement entraînées dans la guerre civile, tandis que leurs perspectives et leur situation au sein des plus grandes questions entourant le conflit ethnique, la guerre, et la réconciliation, ont été politiquement et académiquement négligées.

Cette étude examine premièrement la spécificité historique de l’identité des musulmans sri-lankais pour montrer comment et pourquoi ces derniers se considèrent comme une minorité ethnique et pour quelle raison leurs revendications sont importantes. Deuxièmement, l’article interroge la négligence contemporaine à l’égard des musulmans sri-lankais dans les études sur le conflit ethnique et la guerre civile, à cause de la représentation rigide qui en est donnée, les rendant victimes et sujets d’une guerre séparatiste menée autour de patries ethniques. Troisièmement, cette recherche se concentre sur les nouveaux contours de l’expérience sociale contemporaine et les identifications en tant que musulmans de plus en plus convaincantes pour comprendre ce qui est en jeu pour ces communautés.

Texte intégral

1Sri Lankan Muslims are the second largest Tamil-speaking ethnic minority in Sri Lanka, comprising approximately 8 % of the population. They are also the second largest resident ethnic group in the Tamil speaking and formerly disputed regions of the North and East, the warzone for the 25-year-old civil war which ended abruptly and bloodily in 2009. Sri Lankan Muslim lives have been inextricably caught up in the civil war, even as their perspectives and location within the larger issues surrounding ethnic conflict, war, and reconciliation, have been politically and academically neglected. This work of this essay is to elaborate and make clear these three statements and their compelling nature for contemporary Sri Lankan Muslim identities.

  • 1 E.g.

2Within South Asia, Sri Lankan Muslims are a curiosity. They are Tamil speaking but do not call themselves Tamil Muslims as in Tamilnadu, seeing this as a denial of their rights as minority, seemingly running counter to progressive opinion within India who de-emphasize difference and protest the racialization and marginalization of Muslims within India. For Sri Lankan Muslims, emphasizing difference is precisely seen as a progressive argument that allows the rights of all minorities in Sri Lanka to be acknowledged. Because of this anomaly, Sri Lankan Muslims are also largely ignored in the literature on Muslim communities in South Asia. This neglect has meant that if thought about, they are mostly imagined to be largely contiguous with other South Indian Muslim communities. Most Indians instinctively repeat the phrase Tamil Muslims with regard to Sri Lankan Muslims, a verbal tic that has been the subject of numerous annoyed newspaper articles and internet comments by Sri Lankan Muslims.1 Unfailingly, in many presentations of my own research, South Asianists rarely question the details of my ethnographic work with Sri Lankan Muslims and their suffering in the civil war; rather I have been questioned time and again about describing Sri Lankan Muslims as an ethnicity. Surely, I was asked recently, this is a colonial categorization which I was repeating and re-entrenching. On another occasion I was questioned whether Muslims would call themselves an ethnicity or whether this was an externally ascribed identity.

This UN map is reproduced with the permission of the United Nations Publications Board: SriLanka, Map N°4172 Rev.3, March 2008 (The UN Cartographic Section,​Depts/​Cartographic/​english/​htmain.htm).

3This essay also stands to argue that Sri Lankan Muslims are no different from Sri Lankan Tamils or Sinhalese in asserting themselves as an ethnicity, all these three categories emerge simultaneously in British colonial rule (Nuhuman [1999] 2003). However, few commentators would suggest in the context of Sri Lanka’s brutal ethnic conflict that the ethncisation of Tamil and Sinhalese collective identities has little reality, in the ways that Sri Lankan Muslim ethnicity is questioned. Here I point to how these “externally ascribed identities” have become internally compelling ones for contemporary communities in Sri Lanka, of which Sri Lankan Muslims are an example not an exception. The first section of this essay elaborates the historical specificity of Sri Lankan Muslim identity in order to both demonstrate the disjunctures between the experiences of Sri Lankan and Indian Muslims (even though religious practices bring them together) and, to show how and why Sri Lankan Muslims consider themselves an ethnic minority and why that is not a case of false consciousness but claims that have to be taken very seriously.

4Secondly, this essay takes the contemporary neglect of Sri Lankan Muslims within accounts of the ethnic conflict and civil war. Despite the fact that the everyday lives of northern and eastern Muslims have been inextricably caught up in the ongoing conflict and from a perspective of being neither Sinhalese nor Tamil, they are rare subjects within writings on the conflict. Their stories have been rendered invisible, not because of their silence — anyone working with Sri Lankan Muslims can testify to the enormous desires of Muslims to speak about and narrativise their lives under war — but because of a rigidity of representation of the conflict and who are considered to be victims and subjects of the conflict within a separatist war waged around ethnic homelands. Muslims, while considered an ethnicity, have no one place that is considered Muslim, even though there are many places where Muslims are concentrated. Wide ranging regional dispersal of Muslims across Sri Lanka means that any proposal for an area which is solely Muslim, a Muslim self-governing region (MSR) that was proposed by some eastern Muslim politicians, has found little favor among island-wide Muslims, as it would involve the forced movement of one or another Muslim community (McGilvray & Raheem 2007). There can be no “Muslim Homeland”. In a war fought over ethnic territory and homelands, where do Muslims belong? I argue that the ethnic conflict, the civil war and the relationship of Muslim to their Tamil majorities and the separatist guerillas the LTTE (also known as the Tamil Tigers) has also led to particular contours of social experience and identifications as Muslim which are increasingly more compelling for understanding what is at stake for Muslim communities. This has become ever more salient in Sri Lanka’s now post-war environment which continues to constitute minority identity as a problem, now in relation to a Sinhala majoritarian state.

Sri Lankan Muslims as national minorities

5Sri Lankan Muslim identifications have two major features. The first is the appellation of Muslim as a category encompassing both religion and ethnicity. Muslims are a Tamil-speaking “ethnic” minority, the other major ones being Sri Lankan Tamils and Malaiyaha (Hill Country) Tamils. Ceylonese/Sri Lankan Muslims see themselves as both “indigenous” to Sri Lanka (as opposed to Muslim communities who migrated in more “recent” history such as Malays) and of a different ethnicity to Sinhalese and Tamils. In Sri Lanka, the ethnic category Tamil is both a linguistic and ethnic category (not a religious one) and the ethnic category Muslim is both a religious and ethnic category (not a linguistic one) and the only pan-island category to be so (Nuhuman [1999] 2003). It is Muslimness as a separate ethnic identity here that is of special significance, rather than as a religious identity per se, given that ethnicity came to comprise a set of legal entitlements and cultural rights which have considerable legal and political significance. This must be understood in a context where firstly, all ethnic identities in Sri Lanka are of recent not primordial provenance (Spencer 1990; Gunawardena 1990; Tambiah 1986; Wickramasinghe 2007), and secondly, ethnicisation has been the dominant paradigm for collective identities in Sri Lanka and thus has an enormously compelling force in social imaginaries which cannot be ignored.

  • 2 McGilvray identifies the roots of these Muslim communities in both the pre-Islamic seaborne trade b (...)

6The second important peculiarity is the wide regional dispersal of Muslims all across the island with nearly a third of the total population, concentrated in the Eastern coast.2 It is this regional differentiation and demography that gives Sri Lankan Muslim identity formation and history some of its distinctiveness, both within Sri Lanka (Kearney 1988; McGilvray 1998) and in relations to Muslim groups across the southern subcontinent (McGilvray 1998). Sri Lankan Muslims propound in both urban and rural areas. Sri Lankan Muslims can also be subdivided regionally into “Eastern”, “Northern”, and “Southwest” blocs, though Muslim dispersal across the island, especially in highland Kandy, makes the southwest bloc more permeable. East coast Muslims form around a third of the total population of Sri Lankan Muslims, and the East coast as a whole represents the most considerable concentration of Muslim population. While Sri Lankan Muslims speak Tamil at home, and are mostly Sunni Muslims (Sufi practices were more widespread but are increasingly declining Hasbullah n.d.) there are real differences between these communities. As McGilvray (1998: 446) remarks, “the numerous Moorish farmers on the Northern and Eastern coast are not only distanced geographically but separated socio-economically and culturally as well, from the more affluent and cosmopolitan centres of Muslim trade and political influence in the central and western parts of the island”. These dispersed populations vary greatly in occupation, kinship structures, and wealth and are differently situated vis-à-vis other ethnic groups — Tamils are the majority population for Eastern and Northern Muslims and Sinhalese the majority for southwestern and central Muslims (McGilvray 1998; Ismail 1995). Moreover, despite characterizations of Sri Lankan Muslims as a trading community (based on the self-perception of South Western elites and larger South Asian stereotypes of Muslim communities), the vast majority of Muslims are engaged in fishing, farming and other occupations such as tailoring in northern Jaffna (see Ismail 1995: 79 for a good critique). Historically, the elite has been drawn from South Western Muslims. However, increasingly throughout the twentieth century, Eastern Muslims have become the moral center of Muslim politics. This dispersal and heterogeneity has meant that from the start Muslim identity formation has contended with the effects of a pan-island ethnic identity and deep regional schism and contestation over that identity, which is both situated in a specific colonial history and has had long lasting consequences in postcolonial Sri Lanka. Let us take both these features as they have produced a distinct collective identity.

The formation of ethnic identities

7Religious difference was the key marker by which Muslims were identified long before ethnicity as a concept became key to all social groups in Ceylon, now Sri Lanka. The Portuguese recognized Muslim groups as “Mauros/Moors” with whom they had a long history of competition across world trade routes and it was then that the label “Moors” was first used, though it (as Nuhuman [1999] 2003 points out) is not a category that Muslims designate themselves by. It was in fact under the British (1796-1948) in the context of racialization of collective identities that the category and identification of Sri Lankan (Ceylonese) Muslim emerged. Officially there are five different Muslim groups who were differentiated on the basis of “racial origin” in the colonial period: Ceylon/Sri Lankan Moors, Indian/Coast Moors, Malays, Borahs and Memons. While there has been some intermarriage between these different groups, Muslim ethnic consciousness has primarily coalesced around the group referred to as Ceylonese/Sri Lankan Muslims.

8As Nissan and Stirrat have cogently argued, ethnic/racial categories were not the basis of inclusion or exclusion prior to the eighteenth and nineteenth century, instead social groups were asymmetrically clustered around pre-colonial monarchies in “dissimilar ways” (Nissan & Stirrat 1990: 26). After the British conquest of the Kandyan Kingdom in 1815, the whole island was placed under a common administration in the 1833 Colebrooke-Cameron reforms and its heterogeneous population was differentiated in a system of communal representation (see Spencer 1990; Ismail & Jeganathan 1995). Social groups became differently entitled racial groups in relation to each other as well as the larger social body (ibid). This was embedded within systems of limited representation and the ordering of Ceylon as a racialized geography.

9The 1833 reforms instituted Ceylon as first 5, then late 9 distinct provinces with their mini capitals and answering to Colombo. Different areas were administrated as racio-linguistic regions, in either Sinhalese (southern and central provinces) or Tamil (northern and eastern provinces) according to the preponderance of speakers. Particular regions became understood, and more importantly governed, as the territory of a particular group with a common understanding of inheritance, property, caste, and kinship e.g. the Thesawalamai of the Jaffna Tamils. As Wickramasinghe (2007) remarks, the Colebrooke-Cameron reforms, intended to break down social and cultural division and introduce a more homogenized European nation-state space, in fact introduced constant differentiation at a local level, reinforcing assumptions about Tamilness and Sinhalaness while at the same time creating new configurations that gave rise to new supra-local ethnic identifications (e.g. the eastern province historically was not connected to the pre-colonial northern Jaffna kingdom). Another prominent example is the collapse in the first half of the twentieth century of a historically highly significant distinction between Low-Country and Kandyan Sinhalese into an island-wide identification as Sinhalese. This “racialised geography” of rule (Stepputat 2008) with its simultaneous differentiation and pan-island homogenization of ethnicity is the precursor of today’s post-colonial ethnic conflict, in its demarcation of Tamil and Sinhalese as bounded, politically significant, ethnic collectivities with their own naturalized homelands.

10The major exceptions to this regionally dimpled application of law, custom and governeance were Muslims. Sri Lankan Muslims were instead subject to an island-wide Mohammedan code, though practically speaking, this Mohammedan code was adapted to the various different regional practices of different Muslim communities. The regional affiliation of Muslims or Moors was considered incidental to their ethnic classification, even though in fact at the village level in the north and east, Muslims shared common kinship and inheritance practices with neighboring Tamils. Muslims thus could claim regional homes, but were not accorded to be living in places that were “their own”. They became thus an ethnic collectivity that emerged without a place of its own. This has had enormous importance in the postcolonial era, where nationalist imaginaries which as Liisa Malkki ( [1992] 1999) among others have pointed out has consistently linked people and place as the compelling components of recognized “national” identities (ibid.).

  • 3 In 1923 when territorial representation was more widely introduced out of 37 unofficial members, 23 (...)

11Muslim ethnic identity instead took root not through regional identity but through the new centralized Executive and Legislative structures also instituted by the 1833 reforms. A limited number of unofficial members were to be included in the Legislative Council, but selected on the basis of their representation of different “races”. Communal representation based on race was consistent throughout the hundred years of the Legislative Council, even when overlaid with different ways of constituting the unofficial members (Nissan & Stirrat 1990).3 This system existed for nearly a hundred years before 1931 when communal representation was totally abolished in favor of territorial representations and universal suffrage introduced. The state thus provided a stage where struggles over representation could only be legitimate insofar as they were racialised. Older collectivities became transformed as their significance took on practical and legal consequences. As Wickramasinghe (2007: 44) remarks, while “colonial knowledge did not imagine identities or construct them… it opened up a new realm for political identities to blossom”. Enumerating these new social identities and fixing them in space solidified and fixed social identities and made them possible to be categorized, systematized, identified with, and also maneuvered around as objective features (Cohn 1998; Wickramasinghe: ibid.).

12The beginnings of a debate on separate Muslim racial and political representation date from the late 1860s. By 1866 Moor representatives had begun to serve on a few municipal councils but in 1833 the Island’s Legislative Council had contained only three unofficial representatives, a Tamil, a low country Sinhalese and a Burgher (Asad 1993; Nissan & Stirrat 1990; McGilvray 1998). The Tamil representative Ramanathan was presumed to also represent the Tamil-speaking Moors minority. In 1889 after considerable agitation two new members were added: a Moor member and a Kandyan Sinhalese. It was precisely this development that prompted the famous Ramanathan-Azeez exchange that was seminal in the formation of Muslim politics. In 1885, before the selection of a “Moor” member, Ramanathan made a speech to the Legislative Council, later followed by an 1888 publication of this speech as a paper to the Royal Asiatic Society, Ceylon Branch. This article entitled “the ethnology of the Moors of Ceylon” argued that the “Moors” were not a race (read thus not eligible for a seat on the legislative council) as they were Tamils, who looked like other Tamils and were indeed Tamil speaking, “the proof cumulatively leads to no other conclusion than that the Moors of Ceylon are ethnologically Tamils” (quoted in Ismail 1995: 67). Ramanathan’s intent was understood from the beginning as political rather than ethnological, I.L.M Azeez, a prominent Moor leader and lawyer immediately rejoindered “[Ramanathan’s] object in calling Moors Tamils in race was to dissuade the government from appointing a Moorish member in the Council… to make them understand that there was no such necessity for taking such a step, as the Moors did not form a distinct race” (Azeez in 1889, quoted in Samaraweera 1979: 374).

13Mimicking Ramanathan’s translation of political interests into ethnological ones, in 1907, Azeez published a reply to Ramanathan which attempted the first systematic formulation of a clearly racialised history of the Sri Lankan Muslim community. Azeez argued that the Moors were indeed a distinct race from Tamils with Arab (not South Indian, origins): they were instead the descendants of Hashemites from Arabia who had acquired Tamil through marrying Tamil women and for the purposes of trade. Azeez straddled two dilemmas, eager to claim distinctness from Tamil political hegemony, he also had to domesticate Muslims to defend them against a Sinhala nationalist movement that had denounced Moors and particularly Indian Muslim migrants, as outsiders and foreigners (Ismail 1995). As Ismail (ibid.) concludes, Muslim identity formation in this period was beset already by the contradictions forced upon it and embraced by it by “Moorish” relations to “Tamil” and “Sinhalese” elites and the construction of racial identity as political representation by the British. In order to be recognized as a community with distinct entitlement, Moor elites were forced to simplify the complex heritage of Sri Lanka’s “Moor” community. This erased Persian influences, downplayed the importance of language and the similarities between Muslim and Tamil village level practices outside of the Southern based urban Muslim elite, and promoted instead “a hypostatised Arab ‘racial’ pedigree… to separate Moorish from Tamil and Sinhala ‘races’” (McGilvray 1998: 450).

14This new emergence of elite Muslims into public politics was accompanied by Muslim religious revivalist movements, which had taken their inspiration from earlier Hindu and Buddhist revivalist movements and the larger Islamic revivalist movements across South Asia. Hindu and Buddhist religious revivalist movements had focused on state protection and establishment of religious boards and temples in the face of Christian proselytisation. These revivalist movements also set up Buddhist and Hindu English language schools in an attempt to break the dominance of missionary control of education. Muslim revivalism was similarly pre-occupied by ideas of “upliftment”, modernity, state acknowledgement, and education.

  • 4 In November 1884 Siddi Lebbe oversaw the opening of the first Muslim English educational school “Al (...)

15Muslim elites had not participated as heavily as Tamil and Sinhalese elites in the politico-administrative and educational structures of colonial Ceylon as a result of their reluctance to be educated in Christian Mission schools, the primary vehicle for indigenous elites for English literacy (McGilvray 1998: 448). In the late nineteenth century, this became the focus of new small mercantile and middle class Muslim elites based in Colombo and around Kandy who linked religious revivalism and educational societies (Nuhuman 2003; Zackariya & Shanmugaratnam 1997).4 Papers such as the Ceylon Mohammedan on the 3rd of January 1901, stressed that “in order to take the proper place among our fellow countrymen we should educate our children” (Samaraweera 1979: 372) and were pressing for educational reform, English language education, Islamic reform, and increased political authority. However the religious and educational revivalist movements clearly favored Southern Muslim elites, with not a single school established for East coast Muslims (Samaraweera 1978; McGilvray 1998). The large numbers of Muslims living outside of the southwest were largely excluded from these developments.

  • 5 These fissures were also more general, both Kandyan Sinhalese and the Jaffna Depressed ( “lower”) c (...)

16As Ceylon instituted universal suffrage, mass politics and headed towards independence, elite politicians found themselves having to confront the enormous political gulf between themselves, members of an Anglophone elite (that bridged ethnic differences) (Obeysekere) and their voting constituencies. As others have demonstrated elsewhere (Spencer 1990) Southern Sinhalese politicians increasingly turned to the language of ethnicity, purified Buddhism, and the feeling of being exploited as “sons of the soil” by colonial rulers and minorities who were seen as being privileged under colonial rule. Similarly, Southern Muslim urban elites attempt to transform themselves from a regional elite to a national elite, speaking on behalf of all Muslims throughout the nineteenth and twentieth century (Samaraweera 1979), also found themselves enmeshed in constant struggles over representation. A widened franchise in the 1930s and Independence in 1948 made divisions between Colombo elite leadership and provincial Muslims even more pronounced; “differences between the leadership in Colombo and the Muslims in the regions (living outside the centre), became sharper due to conflicts between the older and younger generations on the issues of power and control over community affairs” (Zackariya & Shanmugaratnam 1997: 9). In discussions surrounding the 1931 Donoughmore constitution instituting territorial representation, Eastern Muslims supported territorial rather than communal representation in an attempt to wrest control away from Southern Muslim elites (Asad 1993).5 Thus, Muslims both mobilized themselves around the same kinds of communal logic that Tamil and Sinhala national elites pursued in trying to constitute themselves around seemingly homogenous ethnic identities, and also along a similar internal/vertical struggle for communal representation (Spencer 1990). This struggle had two major consequences, 1) the increasing prominence of Eastern Muslims in dictating the shape of Muslim politics, and 2) the transmutation of conflict between Tamil and Muslim politicians for limited political representation as minorities (always couched implicitly against each other) into open conflict as Tamil Muslim relations became subsumed by the violence of the civil war. In the northern and eastern war zone areas, the question of who Tamils and Muslims are to each other, politically and socially, has become one of the most pressing questions that the current ethnic conflict and war has forced.

Muslim politics in the twentieth century

17In 1989 a new party the SLMC (The Sri Lanka Muslim Congress formed in 1981) contested the Provincial Council elections, under the leadership of the charismatic MHM Ashraff (Jeyaraj 2000). The SLMC contested the election despite an ordered boycott by the Tamil militant organization, the LTTE (Liberation Tigers for Tamil Eelam) and were rather successful which fuelled Tamil militant anger even further. Ashraff himself had formerly been a member of the (elite) Tamil parliamentary party, the TULF (Tamil United Liberation Front). They were not the only ones; another Muslim only party had also recently announced its inauguration: the East Sri Lanka Muslim Front (ESLMF), which later became the Muslim United Liberation Front (MULF) which attempted to more explicitly negotiate with the separatist Tamil guerilla group LTTE. The SLMC and MULF were the first Muslim-only parties, prior to 1989; Muslim politicians had instead positioned themselves within the two major national parties, the UNP (United National Party) and the SLFP (Sri Lanka Freedom Party) (Ismail 1995; McGilvray 2008). The initial impetus for the formation of the SLMC had come as a result of protest against the UNP government’s alliance with Israel and its proposal of merging the Muslim constituencies of Amparai within the Sinhalese dominated Uva province, effectively disenfranchising a significant proportion of Eastern Muslims. But the emergence of the SLMC and the MULF also represented the increasingly politically articulate and enumerated voice of Eastern Muslims, a voice which grew through a violent “dialogue” with Tamil nationalist politics and which dismissed the Muslim elite in the south as insufficiently protecting their community within increasingly militarized domains. Ashraff’s provenance in Tamil nationalism and his disillusionment reflected a larger trend of Muslim disenchantment and the profound and repeated inability of Tamil nationalist parties and militant groups to accommodate Muslim identity and rights. The contestation of the 1989 elections despite the boycott would itself be the point where the LTTE reversed its lukewarm response to Muslim rights into out and out aggression as it began to perceive that Muslims could play a significant and distinct political role which could not be subsumed by itself. How had it got to this point?

18The Muslim community’s vulnerable position as a minority sandwiched between Sinhala and Tamil nationalisms continued to dictate contests over Muslim political identity throughout the twentieth century. In 1915, Sri Lanka saw its first anti-minority riot. A Sinhalese pogrom against Muslims in the South saw Muslim elites seeking protection from the British state, with Tamil and Sinhala nationalist politicians clearly excusing the Sinhalese rioters (Roberts 1994; Ameer Ali 1981). Fear of such riots cast a long shadow. The first elections for Independent Ceylon in 1948, clearly favored Sinhala majoritarian parties and seeing this and the defeat of many Muslim candidates, the southern Muslim elite began to follow a policy of accommodation with the Sinhala dominated state (Ismail 1995). Muslim politicians developed considerable ingenuity at coalition politics, switching between the UNP and the SLFP governments in order to win concessions and protection for Muslim minorities. This accommodative politics saw the nominal participation of Muslim politicians within State political structures but not much in actual rewards for ordinary Muslim communities (McGilvray 1998: 456).

19Moreover, for northern and eastern Muslims their concern was not how to negotiate life in Sinhalese areas but how to manage a highly intimate neighborly life with Tamils (not Sinhalese) their ethnic majority. Muslims comprise a third of the eastern province — a sizeable minority. In the North Muslims are only 5 % of the population and have always been a more vulnerable community.

Village level relations between Tamils and Muslims

20Village level relationships between Tamils and Muslims are culturally intelligible while segregated. Tamils and Muslims not only share a language in the north and east, they also — unlike Muslims and Sinhalese, and Tamils and Sinhalese — share kinship and neighborly structures. McGilvray (1998, 2008) has extensively documented how East Coast Tamils and Moors share a common kinship structure organized into matrilineal clans, similar understandings of sexuality and substance ideologies and so on. East coast Muslims and Tamils both operate under the kutI system, where kuti’s are exogamous matri-clans and men acquire the kuti of their wives on marriage. Muslim Mosque and Hindu temple membership are passed on through these matri-clans (McGilvray 1998: 467). While residential separation, reduction of intermarriages, and a bifurcated school system have led to less direct interaction between eastern Tamil and Muslim communities (McGilvray 2008), there is real regionally attenuated mutual intelligibility between the communities. As one Northern Muslim told me, in the regular regional tensions and debates between eastern and northern Tamils in Jaffna University in the 1970s and 1980s, eastern Muslims aligned themselves with eastern Tamils in challenging Jaffna dominance over education. This fact is also apparent if one considers the contrast between northern and eastern Muslims. Northern Muslims and Tamils do not operate under a matri-clan system but under a more bilateral caste and mosque structure with similar sorts of dowry patterns. Regional differences between north and east thus transcend ethnic differences. Northern Muslims and Tamils, and, eastern Muslims and Tamils, share more with each other intra-regionally than inter-regionally.

21Moreover, there are differences between Tamil and Muslim relations across region too. The size and scale of the Eastern Muslim community gives it a different confidence in its dealings with Tamil neighbours. In the East, the general pattern of residence is alternating Tamil and Muslim villages. One commonplace saying compares Tamil Muslim relations to “puttu”, a staple favoured by both Tamils and Muslims, long steamed rolls of rice flour cakes glued together with alternating stripes of grated coconut. Tamils and Muslims are said to live together like rice and coconut, integral but separate. In comparison, there were very few Muslim-only villages in Northern areas, with Mannar district with the greatest concentration of Northern Muslims the only exception. Northern Muslims and Tamils shared villages which were internally segregated. Muslims lived within Tamil majority villages, so their representation of villages is of Tamil and Muslim co-existence in even more intimate form. The violence that has entered Tamil and Muslim relations in the north and east has entered it at the level of everyday life. It is not about elite political forms, or particular forums inaccessible to most as in the colonial period. Ethnicity and ethnic identity have come to matter in Sri Lanka because it has come to encompass everyday life in ever more intimate form. The ethnic discrimination that Tamil communities faced under the Sri Lankan state, is inhabited by Muslim communities not only in relation to a state which is largely indifferent to its minorities except in pursuit of its own short sighted military goals, and, by the violence of Tamil nationalist politics. The battlefields have not been far off, they have been the villages and towns and the ordinary public and private spaces of civilian life.

The evolution of Muslim politics

22It was not surprising then that it was in eastern Sri Lanka that the first Muslim-only parties emerged. The East coast as a whole represents the most considerable concentration of Muslim population. Tamils form 42 % of the East, Sinhalese 26 % and Muslims 32 %. Some districts are also Muslim majority areas: Amparai district is about 40 % to 42 % Muslim. This means the possibility of concentrated and powerful voting constituencies and influence in municipal structures. Furthermore, Muslims emerged holding the delicate balance of power between Sinhalese and Tamil constituencies, when state sponsored dry-zone colonization from the 1950s onwards settled a frontier of poor low-country Sinhalese peasants in eastern areas and altered the population balance of the Eastern province cutting into Tamil and Muslim constituencies. In the current conflict, any possible devolution of the northern eastern region thus has to take account of the Muslim minority in the Eastern province, and any merger of the North and East as a single Tamil region rests on Muslim consent.

  • 6 In the 1950s high profile Eastern Muslim MPs such as M.M. Mustapha, M.S. Kariapper and M.C. Ahmad b (...)

23Muslim politicians and youth also participated in the rise of Tamil nationalism in the North and East (Ismail 1995: 86) hence Ashraff’s initial beginnings in the Tamil TULF.6 This became ever more uneasy as Tamil nationalist politicians failed to negotiate beyond exclusivist ethnic ideologies. Tamil nationalist parties could not sufficiently address the concerns of Muslims and sought to mainly subsume Muslims under a broader umbrella of Tamil-Speaking people that tacitly sought to suppress Muslim identity as Tamil. When M.M. Mustapha and other Muslim politicians then moved to switch between the two different national parties, UNP and SLFP, they were immediately accused of being thoppi pirutu, or turncoats, (literally, switching hats) by Tamil politicians, despite the fact that some prominent Tamil MPs were also members of the UNP. This set in place a script trotted out whenever convenient by Tamil politicians of Muslim “collaboration” with the “Sinhalese”.

  • 7 All (with the possible exception of TELO) had Muslim cadres with some reaching very high positions (...)

24As disenchantment with the Tamil parliamentary parties spread through the North and East, Tamil militant groups came to occupy a more prominent place within political landscapes in the 1970s and 1980s. Whilst, the small Muslim minority in the North did not pose any significant threat or problem for Tamil militant politics, all Tamil militant movements realized in the eastern provinces that the Muslim minority in one way or another represented a key constituency. The first phase of militancy saw the attempt to incorporate Muslims into the militant struggle as combatants through a larger ideological notion of Tamil speaking peoples which would then both hold the North and East as their traditional homelands: “subsumed from time to time under various politically-convenient concepts such as ‘Tamil speaking people’, ‘Eela-Muslims’, and ‘Islami Tamils’” (Taraki [1990] 1991: 28). Indeed, Muslims had sheltered Tamils in the 1983 anti-Tamil riots, and the intensification of the government military campaigns against the Northeast did lead to alliances between the two communities, with some Muslims even joining the major militant movements.7 However, from the mid-1980s, militant movements began to increasingly impose taxes on Muslim businesses and advance more exclusivist ethnic ideologies. In April 1985 the first riots between the Tamil and Muslims broke out in the east, with violence spreading outwards from Kalmunai to other areas. Muslims began to fear for their own safety their rights as minority communities with well-articulated rights of their own, not subsumed under a majoritarian Sri Lankan Tamil identity.

25These issues were not addressed under the Indian occupation by the IPKF (Indian Peace Keeping Force – 1987-1990) and the Indo-Lanka treaty of 1987. Instead Muslims saw further diminution of their rights. The (Tamil militant) EPRLF led Eastern-province government persecuted some Muslim communities. Muslims also complained that the Indian Army refused to acknowledge the historic importance and independence of the Sri Lankan Muslim in brokering devolved powers. The Indians treated them as similar to the Indian Muslim community, and, its rank and file soldiers also in some instances persecuted Muslims (Jeyaraj 2002). Muslims thus in this period did often support the increasingly dominant militant group fighting the Indians, the LTTE (Liberation Tigers of Tamil Eelam): there were many Muslim cadres in the LTTE, and as Taraki points out, the LTTE survived the IPKF offensive in the east chiefly because eastern Muslims maintained supply lines through Muslim areas to feed guerrillas in the jungle (Taraki [1990] 1991: 71). The IPKF itself also identified some Muslim villages as Tiger strongholds (ibid.). Despite this assistance, when the LTTE gained control of the north and east after the Indian Army’s withdrawal in 1990, it adopted a distinctly racist attitude to the Muslim community fuelled by the successful contestation of the 1989 elections by the SLMC. As war began with the Sri Lankan Government in June 1990, the LTTE embarked upon a series of massacres of Muslims in Eravur and Kattankudy. In 1990 it evicted Muslims from the North. LTTE massacres and harassment of Muslims was accompanied by the machinations of the Sri Lankan state Special Task Force in arming young Muslim men as “Muslim home guards” to counter-attack Tamil civilians to destabilize Tamil communities. The cycle of killings and counter-killings in the East make grim reading. Tamil and Muslim relationships have broken down at a village level leaving great scars within the East that were re-stoked continually by the LTTE.

  • 8 See article on this in: OpenDocument (...)

26Throughout the war years, the LTTE deliberately sought to break Muslim economic power, land occupation and political strength. This was because the Muslim community in the East always posed a deep unsettling crisis at the heart of the nationalist project for a unified, pure Tamil North-East homeland. In the 2002 peace talks for example, the LTTE successfully ensured that Muslim political parties were only allowed limited participation, clearly excluding them as stake holders (see McGilvray & Raheem 2007 for an extended discussion). Eastern Muslim communities were trapped between Sri Lankan Army and LTTE fighting and bombardment resulting in mass displacement (such as in Mutur in 20068) with neither caring to guarantee the safety of Muslims. In 2007, the Sri Lankan state mounted a campaign to “clear” the East of the LTTE, prior to its big final campaign against the LTTE that resulted in its elimination in 2009. This resulted in (the splinter LTTE group transformed into) the parliamentary party TMVP in charge of the East, with a paramilitary wing working in alliance with the Sri Lankan state. Far from violence against Muslims abating, Muslim communities reported constant harassment and discrimination by the TMVP cadres. At the same time, the circulation of small firearms in the east led to call kinds of conflicts, including within the Muslim community around Sufi and Sunni debates, becoming militarized (Hasbullah n.d.).

New forms of community

27If, as I have suggested above, Muslim ethnicity has emerged in the postcolonial state in relation to being a third minority, sandwiched between Sinhalese and Tamil majorities, and through an ongoing relationship with Tamil identity formation — particularly Tamil nationalist — then here I wish to turn in more detail to my suggestion that the civil war and the relationship of Muslim to the separatist guerillas the LTTE has also led to particular contours of social experience and identifications as Muslim. My own research has been with Northern Muslim communities, and here I wish to just briefly describe how Muslim communities from the north became transformed into “Northern Muslims” in order to illustrate the larger transformations of Muslim ethnic identification drawn out in this essay.

28Firstly, Muslims from the north possess neither the numerical strength of the East coast Muslims nor the historical dominance of urban southern elites. The only northern district to have numerical strength is Mannar province, 26 % Muslim, The question of a “place of their own” is also thus more heightened amongst northern Muslims. However, while Muslims only formed 5 % of the total population of the north, they were significant minorities nonetheless; the Mannar district is given its distinctive character within the Northern Province by its large Roman Catholic population and its Muslim community, and the sacred Catholic and Muslim sites (Talaimannar) within it. Jaffna Town’s Moor Street complex was famously part of the town with its distinct architecture and history.

29On the 30th of October 1990, the LTTE expelled all the 75-80,000 Muslims from the five districts in the North under its control, Vavuniya, Mannar, Mullaithivu, Jaffna and Kilinochchi. They were given twenty-four to forty-eight hours to leave. In Jaffna, the area on which I worked, Muslims were given only two hours to leave, roads were blocked and a specific route allocated through LTTE checkpoints. At the checkpoints, Muslims’ possessions were further examined and appropriated. In some areas, local LTTE cadres were moved to other areas to carry out the eviction so that they would not protest at having to evict their neighbours. The order came from the highest ranks of the LTTE and no clear explanation was offered. One popular supposition is that the LTTE, faced with the numerically and politically stronger eastern Muslim minority, decided as a warning to evict the much smaller and more politically vulnerable Muslim minority in the north. It was a military operation without any popular backing or participation. By November, there were no more Muslims in the North. This ethnic cleansing is known as “the Eviction” and the community of Muslims created by this act of mass expulsion are formally “IDPs” (Internally Displaced People) and refer to themselves as “Northern Muslims” and “ahathi” (refugees). Sixteen years on, the majority of Northern Muslims, 65,000 or so, live in refugee camps and settlements in the Puttalam District in the North-central province, curving under the disputed territories of the North that they still call home. Through two peace processes and ceasefires, the collective right of Northern Muslims to return, and an LTTE guarantee that they will not be evicted again was never brokered.

30The Eviction for Northern Muslims is truly, in Veena Das’ categorisation, a “critical event”, events which have the capacity to interpenetrate and transform multiple institutions and narratives and after “new modes of action came into being which redefined traditional categories such as codes of purity and honour, the meaning of martyrdom, and the construction of heroic life” (1995: 6). Northern Muslims were “born” in the aftermath of an unthinkably traumatic event that broke one set of communities in the North and created another. Previous to the Eviction, there were no Northern Muslims; there were only Jaffna Muslims, Mannar Muslims, etc. The use of the term “Northern Muslim” as meaningful collective self-description only appeared after the Eviction and it denoted a community traumatically born through Eviction, it gave them an origin in a place, a region after they had lost it. They are a community created around 1) common origins in the Northern districts of Sri Lanka, 2) a shared collective experience of “The Eviction” and 3) a common experience of internal displacement within refugee camps or as the UNDP called them “welfare centers”. The strength of this collective identification and the density of stories of the Eviction in Puttalam cannot be underestimated. Diverse families, individuals and villages found that their lives became pulled together in such a way that, even though their pasts were dissimilar and multiple, in 1990 the LTTE ensured that their futures would be intertwined. Thus, ethnic identification continues to transform from the more standard story of colonial differentiation and electoral politics that I traced above, as ever new stakes and histories which draw together new collectivises. The Eviction is one such major event, it redrew the ethnic map of the north, making it Tamil in a way it had never been before, it tore apart Tamils and Muslims giving them now a history of fissure rather than sometimes uneasy fusion and co-existence, and it created a new community of people with a common genesis. Muslims from the north had become Northern Muslims.

  • 9
  • 10 Muslim % 20Population % 20 (...)

31The significance of this identification as Northern Muslim, was that it denoted a very specific community of experience around a traumatic origin born directly out of the war. This can be seen in the complex ways in which Northern Muslims related to the sizeable Puttalam Muslim community in the area they settled in as IDPs. When the 65,000 Muslim refugees arrived they transformed the demography of the Puttalam district considerably. In 2001, the Puttalam district was 73.8 % Sinhalese, 6.9 % Sri Lankan Tamils, 0.3 % Indian Tamils, 0.3 % others, and 18.7 % Sri Lankan Muslims (Consortium of Humanitarian refugees 2003). In 1981, Sri Lankan Muslims were only 8.8 %.9 Whereas Sinhalese are obviously the large majority of the district, settlement patterns provided a different ambiance to the area. The majority of the Sinhalese lived in outlying rural areas; the Puttalam urban area was 67.4 % Sri Lankan Muslims for example, and the Kalpitiya division where another large number of refugees settled was 57.9 % Muslim.10 These figures mean that in the areas where the IDPs live, the spoken language on the street was Tamil and the physical presence was overwhelmingly Muslim.

  • 11 For the refugees, 90 % still on dry rations provided by the World Food Programme and the Sri Lankan (...)

32While they were initially welcomed to Puttalam, by the time I first visited the camps in 2003, relationships between locals and IDPs had soured significantly. The influx of IDPs dramatically changed the political and social economy of a historically underprivileged and economically stagnant area, albeit one that experienced a brief period of agricultural intensification after the IDP arrival (Shanmugaratnam 2000). Locals began to resent the IDPs and their seeming “well-off status”, receiving rations in a desperately poor region, leading to constant conflicts some violent.11 IDPs also complained of continual discrimination in the administrative and bureaucratic structures of Puttalam. Despite this new Tamil-speaking population, the bureaucracy had remained rigidly Sinhala-speaking and Puttalam locals of all ethnicities were Sinhala-competent unlike Northern Muslims, who tended to be only conversant in Tamil and bitterly resented the language discrimination they faced. Puttalam Muslims like most South-West Muslims are minorities for Sinhalese not for Tamils, and this has considerably shaped the ways in which they understand themselves. Northern Muslims instead complained bitterly about the birth certificates for their children being written in Sinhalese.

  • 12 Northern Muslims were officially designated IDPs (Internally Displaced Persons) by the government a (...)

33I have written about life in Puttalam extensively elsewhere (Thiranagama 2007, 2011). Here I just wish to highlight the complexity of the relationship between Northern Muslim IDPs and the local community, including Puttalam Muslims. Distinctions between the “locals”/host community and IDPs, however, went both ways. IDPs themselves argued for social and moral distinctions between “locals” and ahathis (refugees) even across ethnic lines.12 Puttalam’s pre-existing Muslim minority were seen as different from refugees — despite their shared ethnicity and language — on the basis of villages of regional origin. For refugees, not all Muslims were of the same kind. Being with “your own people” meant primarily being with other Northern Muslims, preferably those from the same district as yourself. Intermarrying between Puttalam Muslims and Northern Muslims has become common over the last 20 years, but it is not openly advocated. IDPs represented themselves as more “sophisticated” and “refined”; they talked of being more “educated” than local Muslims. In 2008, as my friend Nazima took me through a settlement with many new Jaffna Muslim houses, she talked about the style of the new houses as “Jaffna style” with large verandas, pillars, and open doors, very different, she suggested disparagingly, from Puttalam Muslims, who built “dark and small houses” with indirect entrances like the other Puttalam locals. Northern Muslims drew clear status distinctions between themselves and locals. At the same time, Puttalam Muslims complained about the lack of piety among Northern Muslims and the seeming laxness of their control over women. When IDPs arrived, the young girls who were accustomed in their northern homes to walk to evening tuition classes and ride bicycles found their movements restricted after local Muslim complaints and their families’ embarrassment at being called bad Muslims. These sorts of tensions, attitudes, and differences meant that Muslimness as an ethnic category, if not as religious practice was, for most I interviewed, hyphenated by place and a local/refugee distinction even as intermarriage threatened it. For example, Shanmugaratnam (2000) relates how Alankuda village, prior to the Eviction, had been a Puttalam Muslim village with one mosque, and was seen by locals as one village. After IDPs moved into Alankuda and established settlements, the village changed radically. It became seen by the refugees who lived there as composed of twelve mosques and twelve clustered settlements. Each settlement and mosque was centered around a village of origin in Mannar or Mullaithivu. Thus though joined together by religious practice and by marriageability, nonetheless, the civil war and differing landscapes of minority relations, separate Northern and Puttalam Muslims.

34IDP life has transformed Muslim life indelibly (Thiranagama 2011). Not least, Northern Muslim identity has emerged integrally through the civil war. The birth of a community through a traumatic act, but also through the sedimentation of this loss into everyday forms of life, exemplifies firstly, how ethnicisation in Sri Lanka has been continually transforming throughout the last thirty years of war and how people continually exceed the categories by which they are (often violently) defined. Since the end of the war in 2009, despite twenty years of absence, Northern Muslims are returning in their thousands to their former homes in the North, opening a new chapter on Tamil Muslim relations that still remains to be told.


The Tamil National Alliance (TNA) has said the Sri Lanka Muslim Congress (SLMC) could hold discussions with the TNA leadership instead of participating in the talks between the government and the alliance. TNA parliamentarian Suresh Premachandran has told the media that the TNA had a good relationship with the SLMC and kept the major Muslim party informed of the developments in the talks between the government and the TNA. Premachandran made this comment in the backdrop of the SLMC insisting that the party be given an opportunity to represent the Muslim community at the talks between the Government and the TNA on finding a political solution to the ethnic issue [italics added, May 30 2011, The Sunday Leader].

35By way of conclusion, let me return to some of the central propositions set out initially. This essay has attempted to demonstrate that ethnicity, and the appellation of Muslims as an ethnic identity, has been central to Muslim identity formation. While being a Muslim is also constituted by religious practice, it is translation of religious practice into an ethnic identity that in Sri Lanka guarantees particular distinct political rights as well as religious ones which all communities are entitled to. This makes Sri Lankan Muslims like not unlike Tamils and Sinhalese, given that all three of these communities are constituted around a recent colonial history of substantializing “race” into representation. Throughout the twentieth century mass politics has transformed ethnicisation and representation into encompassing a wider franchise of political possibility. Ethnicity has also become less a matter of elite representation and more a question of everyday political practice and life as ethnic conflict and civil war have interpenetrated ordinary lives in cataclysmic form. It is not Muslim religious practice per se that has come under attack, it is Muslimness as a distinct community, with community spaces, internal and separate community organizations and as a distinct political community with a powerful and mature voice within national and local politics that is of significance. The comparison is with Christians who are also 8 % of the population but are subsumed within Tamil and Sinhalese communities and have distinct religious practices and spaces but are not a political community.

36It is this distinct identity which has led to the increasing violence enacted against Muslim communities in the war zone, as well as their neglect within academic and political writing which has seen only Tamil and Sinhalese subjects and victims. The configuration of the civil war as a war about homelands, where the North and South are considered natural homelands for Tamils and Sinhalese, has little conceptual space for Muslims who do not have a distinct homeland but nonetheless live in great number all across the island. Not least, the civil war has transformed Muslim identity in the North and East, the emergence of a distinct Northern Muslim community, one such compelling example. In Sri Lanka now in 2011, as we contemplate the end of the civil war and the new lineaments of a resurgent Sinhalese majoritarianism, we cannot afford to ignore Sri Lankan Muslims as central figures in any postwar reconciliation. The major fissure of the last thirty years has also been the transformation of Tamil and Muslim relations, and, any possible settlement for minorities has to also include Muslims, the minority of the minority.



Ameer Ali, A.C.L. (1981), “The Genesis of the Muslim Community in Ceylon (Sri Lanka): A Historical Summary”, Asian Studies, 19, pp. 65-82.

Asad, K.M.N.M. (1993), Muslims of Sri Lanka Under the British rule, New Delhi, Navrang.

Brun, C. (2003), “Local Citizens or Internally Displaced Persons? Dilemmas of Long Term Displacement in Sri Lanka”, Journal of Refugee Studies, 16 (4), pp. 376-397.

Consortium of Humanitarian Agencies (2003), “National Peace Audit, Puttalam District”, November,

Dewaraja, L.S. (1994), The Muslims of Sri Lanka: One Thousand Years of Etnic Harmony, 900-1915 AD., Colombo, Lanka Islamic Foundation.

Gunawardena, R.A.L.H. (1990), “The People of the Lion: The Sinhala Identity and Ideology in History and Historiography”, in J. Spencer, ed., Sri Lanka: History and the Roots of Conflict, New York, Routledge, pp. 45-86.

Hasbullah, S.H. (n.d.), “Sufi Traditions in Eastern Sri Lanka: Past and Present” [Available from author].

Hasbullah, S.H. (2001), Muslim Refugees: The Forgotten People in Sri Lanka’s Ethnic Conflict, Vol. 1, Nuraicholai, Research and Action Forum for Social Development.

Hirschon, R. (1989), Heirs of the Greek Catastrophe: The Social Life of Asia Minor Refugees in Piraeus, New York, Oxford University Press.

Ismail, Q. (1995), “Unmooring Identity: The Antinomies of Elite Muslim Self-Representation in Modern Sri Lanka”, in P. Jeganathan & Q. Ismail, eds., Unmaking thenation: The politics of identity and history in modern Sri Lanka, Colombo, Social Scientists Association.

Jayawardena, K. (1985), Ethnicity and Class Conflicts in Sri Lanka, Colombo, Sanjiva Books.

Jeganathan, P. & Ismail, Q., eds. (1995), Unmaking the nation: The Politics of Identity and History in Modern Sri Lanka, Colombo, Social Scientists Association.

Jeyaraj, D.B.S. (2000), “Obituary: A Pioneering Leader, M.H.M Ashraff 1948-2000”, Frontline, 17 (20). September 30th-October 13th, (last accessed February 2011).

Jeyaraj, D.B.S. (2002), “Need for SLMC – LTTE MOU”,

Malkki, L. (1995), “Refugees and Exile: From ‘Refugee Studies’ to the National Order of Things”, Annual Review of Anthropology, 24, pp. 495-523.

Malkki, L. ( [1992] 1999), “National Geographic: The Rooting of Peoples and the Territorialization of National Identity Among Scholars and Refugees”, in A. Gupta & J. Ferguson, eds., Culture, Power, Place: Explorations in Critical Anthropology, Durham, N.C., Duke University Press.

McGilvray, D.B. (1982), “Sexual Power and Fertility in Sri Lanka: Batticaloa Tamils and Moors”, in C.P. MacCormack, ed., Ethnography of Fertility and Birth, London, Academic Press, pp. 25-73.

McGilvray, D.B. (1998), “Arabs, Moors and Muslims: Sri Lankan Muslim Ethnicity in Regional Perspective”, Contributions to Indian Sociology, 32 (2), pp. 433-483.

McGilvray, D.B. (2008), Crucible of Conflict: Tamil and Muslim Society on the East Coast of Sri Lanka, Durham, N.C., Duke University Press.

McGilvray, D.B. & Raheem, M. (2007), Muslim Perspectives on the Sri Lankan Conflict, Washington, D.C., East-West Center.

Nissan, E. & Stirrat, R.L. (1990), “The Generation of Communal Identities”, in J. Spencer, ed., Sri Lanka: History and the Roots of Conflict, New York, Routledge.

Nuhuman, M.A. ( [1999] 2003), “Ethnic Identity, Religious Fundamentalism and Muslim Women in Sri Lanka”, Lines Magazine, and (accessed September 2010).

Roberts, M. (1994), Exploring Confrontation. Sri Lanka: Politics, Culture and History, Chur, Harwood Academic Publishers.

Rogers, J.D. (1990), “Historical Images in the British Period”, in J. Spencer, ed., Sri Lanka: History and the Roots of Conflict, New York, Routledge.

Samaraweera, V. (1979), “The Muslim Revivalist Movement, 1880-1915”, in M. Roberts, ed., Collective Identities, Nationalism, and Protest in Modern Sri Lanka, Colombo, Marga Institute.

shanmugaratnam, N. (2000), “Forced Migration and Changing Local Political Economies: A Study from North-Western Sri Lanka”, Noragric Working Papers, Noragric/Agricultural University of Norway,

Spencer, J. (1990), “Introduction: The Power of the Past”, in J. Spencer, ed., Sri Lanka: History and the Roots of Conflict, New York, Routledge.

Stepputat, F. (2008), “Forced Migration, Land and Sovereignty”, Government and Opposition, 43 (2), pp. 337-357.

Tambiah, S.J. (1986), Sri Lanka: Ethnic Fratricide and the Dismantling of Democracy, London, IB Tauris & Co Ltd.

Taraki ( [1990] 1991), The Eluding Peace (An Insider’s Political Analysis of the Ethnic Conflict in Sri Lanka), Paris, ASSEAY.

Thiranagama, S. (2011), In My Mother’s House: Civil War in Sri Lanka, Philadelphia, University of Pennsylvania Press.

Thiranagama, S. (2007), “Moving on? Memory, Generation and Home for Displaced Northern Muslims in Sri Lanka”, in J. Carsten, ed., Ghosts of Memory: Essays on Remembrance and Relatedness, Oxford, Blackwell Publishing, pp. 126-149.

Wickramasinghe, N. (2007), Sri Lanka in the Modern Age: A History of Contested Identities, London, Hurst.

Zackariya, F. & Shanmugaratnam, N. (1997), “Communalisation of Muslims in Sri Lanka: An Historical Perspective”, in Alternative Perspectives: A Collection of Essays on Contemporary Muslim Society, Colombo, Muslim Women’s Research and Action Forum.


1 E.g.

2 McGilvray identifies the roots of these Muslim communities in both the pre-Islamic seaborne trade between South, South East Asia and the Middle East both Arabic and Persian, and Arab Muslim mercantile trade in the southern subcontinent subsequent to the dawn of Islam in the 17th century (McGilvray 1998: 436). As early as the 15th century and by the 17th century there were large populations of Muslim farmers established on the East coast (McGilvray 1998). There is clear early archaeological evidence for 7th and 8th century “Muslim” settlements in the North (Hasbullah 2001). Further, there was large scale Muslim migration to the central inland of the island, after the Kandyan kingdom offered protection from Portuguese persecution (Dewaraja 1994), as both the Portuguese (1505-1658), and the Dutch (1658-1796) regime who followed them, sought to impose specific persecution on Muslim communities designed to break their trading practices and alliances across the Indian Ocean.

3 In 1923 when territorial representation was more widely introduced out of 37 unofficial members, 23 represented territorial electorates and 11 special communal electorates (Nissan & Stirrat 1990: 29).

4 In November 1884 Siddi Lebbe oversaw the opening of the first Muslim English educational school “Al Madurasathul Khairiyyatul Islamiah” in Colombo. Siddi Lebbe then went on to found the Muslim Educational Society in 1891 to emphasize the need for modern Muslim English education.

5 These fissures were also more general, both Kandyan Sinhalese and the Jaffna Depressed ( “lower”) caste league also petitioned the Dounoughmore commissioners asking for separate representation from low-country Sinhalese and Sri Lankan Tamils respectively.

6 In the 1950s high profile Eastern Muslim MPs such as M.M. Mustapha, M.S. Kariapper and M.C. Ahmad became members of the Tamil dominated Federal party and stressed their relationship to the broad linguistic nationalism of Tamil speaking areas and a larger Eastern-Muslim consciousness (Ismail 1995: 86)

7 All (with the possible exception of TELO) had Muslim cadres with some reaching very high positions such as Jaan Master of PLOTE, Cader of EROS and Farook of the LTTE In addition cadres, both Tamil and Muslim, took movement nicknames that crossed ethnicity (i.e. Tamil cadres with “card names” such as Raheem, and Muslim cadres with non-Arabic Tamil names).

8 See article on this in: OpenDocument&RSS20=18-P


10 Muslim % 20Population % 20in % 20Puttalam

11 For the refugees, 90 % still on dry rations provided by the World Food Programme and the Sri Lankan state in 2000 (HASBULLAH 2001) and the majority living in appalling living conditions, this resentment was highly unwelcome and restrictions on employment sent them further into impoverishment.

12 Northern Muslims were officially designated IDPs (Internally Displaced Persons) by the government and UNHCR, but when talking about self-perceptions I follow their usage where they refer to themselves as ahathi (refugee).

Table des illustrations

Légende This UN map is reproduced with the permission of the United Nations Publications Board: SriLanka, Map N°4172 Rev.3, March 2008 (The UN Cartographic Section,​Depts/​Cartographic/​english/​htmain.htm).
Fichier image/png, 440k


She is an Assistant Professor of Anthropology at the New School for Social Research and currently Visiting Assistant Professor of Anthropology at Stanford University. Her research has focused on various aspects of the Sri Lankan civil war. Primarily, she has conducted research with two different ethnic groups, Sri Lankan Tamils and Sri Lankan Muslims. Her research explores changing forms of ethnicisation, the effects of protracted civil war on ideas of home in the midst of profound displacement and the transformations in and relationships between the political and the familial in the midst of political repression and militarization.

– 2011 In My Mothers House : Civil War in Sri Lanka, Philadelphia, University of Penn­sylvania Press.
– 2009 S. T. & T. Kelly, eds., Traitors : Intimacy, Suspicion and the Ethics of State-Building, Philadelphia, University of Pennsylvania Press.

© Éditions de l’École des hautes études en sciences sociales, 2012

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search