The ‘problem’ with language in a Tamil Sri Lankan context is both epistemological and political, this being a society that is suspicious of how outsiders define indigenous knowledge and the construction of meaning. There are dangers in imposing ill-considered views when these views can pose a threat to the foundations of identity within spheres of conflict. The scope of knowledge that is attainable will always be in question with ethnographies outside home territory, and epistemological enquiries must acknowledge these limitations. In many respects this study is empirical in nature because it constitutes a view of uneasy post-war dynamics that relies more on sensory experience and information than on a priori knowledge or the ability of language to portray suffering in adequate depth. Hence, primacy is given to expressions of suffering that, as already stated, employ other means – that is to say, rituals that transcribe past experience of pain in the body and the current oppressions on civilian life in the north. There have been significant contributions by medical anthropologists to the study of the language of pain and suffering within a context of violence (Argenti-Pillen 2003; Daniel 1996; Das 1997; Scarry 1985). The inexpressibility of pain inscribed on the body through torture has been eloquently analysed by Scarry (1985) in her seminal work on torture and war. She contends that pain resists objectification through language, that it transcends the language of the sufferer, making an evocation of the experience impossible: the pain of torture deconstructs, unmakes and destroys language (p. 20). Imprisonment and torture were endemic in the north and east of Sri Lanka, especially after the mid-1990s when the army began occupation of these areas. Thousands disappeared following their arrest, and their whereabouts have never been established. Consequently, many families have had to live with an uncertainty that has affected their ability to mourn. One of the central doctrines of the practice of torture is the ownership of power and control through fear (Scarry 1985; Daniel 1996; Feldman 1991; Taussig 1987; Robben 2000). What makes torture so effective is this ability to instil fear and thereby destroy trust, confidence and a sense of self, in a demonstration of power over both individual and society – not by the individual torturer but by his/her command authority. Language is subverted through torture’s insistence on meaningfulness; however, this is fictitious because the demand for

‘knowledge’ (information) is in fact meaningless or even non-existent. There is no satisfactory outcome for either the torturer or, of course, the sufferer. Daniel (1996) in his interviews with torturers in Sri Lanka states that the torturer in seeking truth ultimately fails in his endeavour due to the absence of any possible truth in this context. The ‘truth’ is, in fact, his assertion of power over the individual, not the obtaining of useful information. S. Turner (1995) describes how torture, and the employment of other means of silencing communities through the instillation of fear, means that people become reluctant to hold information that can be used against them in the future. The mere possibility of torture is sufficient to silence whole communities. In modern warfare, the abusive techniques of the torturer have also meant that visible scarring is minimised, making it easier to deny the pain and suffering that has been inflicted. When language has the potential to betray or endanger the self, it restrains its use and lends it suspicion, making it a perilous liability. This is not only the case with torture, but occurs in any society that has lost trust. In northern Sri Lanka, articulating an opinion – especially a critique – posed huge risk; even asking questions could easily incur suspicion. In this situation, conversation tended to remain superficial (and therefore safe), unless taking place among trusted friends. In a society cognisant of the constraints of colonialism and the hierarchies of the caste system with its strict codes of conduct, maintaining caution and avoiding controversy has some familiarity. But this both shifted and grew in the face of overt violence and covert surveillance, existing through the mechanism of instilled fear. Infiltration by both army and militant groups was wholescale: the army camps were not situated apart from the general population but, instead, Tamil civilians lived among those who were effectively their captors. Similarly, the various paramilitary offices were embedded in villages and towns, but difficult to spot: I visited an LTTE ‘camp’ that was located in an ordinary street in the house of a Tamil woman living abroad and unrecognisable as such from the outside. This naturally increased fear of talk, especially when lines were so crudely drawn between loyalty and disloyalty, and dissonance was consequently rare. As stated previously, there were seldom opportunities to meet people individually and discussion of more controversial subjects did not often occur. There were repercussions too for arrangements to see interviewees on their own, as this was considered unwise. This rendered interviews very limited in scope and had enormous implications for the research. What proved more productive was attention to the activities of daily life, including the observation of formal rituals. Performance and the use of dramatic interpretation are familiar to most Tamils. I came across many examples where performances – in a variety of formal and informal settings – provided clear perspectives on how people were dealing with the war and its consequences. I attended numerous local performances not only of traditional folk dance and music but also specially created shows depicting events from the war. One of these took place at a local Ayurvedic training college and explored the effects of the war on the civilian population though a drama performance. It was at once vivid, shocking, frightening and moving in its

portrayal of the madness of war – more than any purely spoken or written words could convey, I suspect. With school or college groups, as well as with more therapy-orientated activities, drama was the preferred mode of articulating a wide variety of social issues. It arose from ancient forms of dance, poetry, music, theatre and religious ritual that were still extant, as well as being adapted to the current setting. The LTTE, for example, used songs, poetry and dance to propagate their message in villages and towns and opened a drama institute in Kilinochchi, the main urban centre within their territory. There was also a very popular and renowned theatre group, which toured the peninsula and beyond, putting on dance and drama shows for villagers that engaged people of all ages. Many of the local NGOs also used creative means to address suffering and were highly regarded in their work, which extended across the whole region. Despite a facility and predilection for performance, discussing problems directly is not an alien concept. In most villages in the past there was an individual who was regularly sought out for her/his wisdom, insight and expertise and who played a pivotal role in the resolution of both personal and local conflicts. There was also the parihari, mentioned previously, who had detailed knowledge of families through access to their major life events and enabled a communicative network of shared information. There was a form of narrative that communicated key social and life events across neighbourhoods and belonged largely in the public domain as: ‘a sanctioned process, a social fact . . . (that) helps maintain group unity, morality and history’ (Gluckman 1963b). Following the destruction of so many villages, these individuals, along with their fellow inhabitants, were dispersed and – without a settled, cohesive community in which to fulfil their role – gradually disappeared. Within the context of multiple dislocations, information was spread more randomly through gossip – and there were frequent complaints of gossip’s lack of restraint. Meanwhile, talking and listening had become ‘professionalised’, usually by strangers who assumed a role similar to that formerly taken by local village elders but without their degree of trust or familiarity with family or village history – psychosocial counselling was offered by most of the large INGOs to assist traumatised individuals after the war. Notions of confidentiality, privacy and individuality are linked to theories and models used in counselling. While applicable in a western setting, it was problematic in the context of dislocation described above, where information was communicated through gossip and the idea of individual privacy was not well developed. Counselling was seen as belonging in the public domain, with any discussion taking place in the counselling sessions treated as public knowledge. The discussion of private issues with a stranger, particularly one from a different culture, would not be easily countenanced, especially as, in this tense inter-war period, gossip and defamation were undermining the private lives of many. Thus, people did not attend the offered sessions for two reasons: either they did not feel confident in talking to strangers, or they were wary of further gossip. Only the local NGO workers seemed to have an awareness of these issues, as well as some key local knowledge, and their interventions subsequently acquired

a more pragmatic direction. This was possible because they were not bound by rigid methodologies of time, space and language – essential aspects of ‘traditional’ western counselling. Another silent group was made up of those who chose the radical step of joining the LTTE. Following their enrolment, none of them spoke openly, either while a member or afterwards, having taken a vow of silence about any controversial matter related to the movement. Interviews with LTTE cadres elicited little in the way of unbiased information, as they tended to follow LTTE rhetoric with externally managed and limited statements. To all intents and purposes, those who joined up disappeared from their families’ lives: their names were changed, and on the occasional visit home they were always accompanied by a fellow cadre to ensure nothing that contravened the LTTE line was said. It was virtually impossible to obtain accurate information in Sri Lanka due to the propaganda employed by all factions in relation to the political and military negotiations: silence was therefore widespread.