War & Peace in Sri Lanka: the Government's Reform Proposals

தமிழ்த் தேசியம்

"To us all towns are one, all men our kin.
Life's good comes not from others' gift, nor ill
Man's pains and pains' relief are from within.
Thus have we seen in visions of the wise !."
- Tamil Poem in Purananuru, circa 500 B.C

Chandrika's 'Devolution Proposals'

War and Peace in Sri Lanka
The Government's Reform Proposals and Beyond
Sumantra Bose

The constitutional reform proposals presented by the Chandrika Kumaratunga government have been broadly welcomed in Sri Lanka, India and internationally. This generally positive response is in itself not unexpected or inexplicable. When one of the parties to a seemingly intractable and interminable civil war, which has exacted a horrifying human toll, finally puts forward a detailed blueprint for a political resolution of the stalemated crisis, a certain level of optimism is neither avoidable nor unwarranted.

Yet it might be prudent to temper that optimism with a good measure of circumspection and even a sizeable dose of skepticism. First of all, there is the perhaps unfortunate but nonetheless inescapable reality that any "peace process" in Sri Lanka that does not involve the participation, in a central role, of the Liberation Tigers of Tamil Eelam (LTTE), is likely to remain an exercise in futility. The relapse into massive violence that has followed the breakdown of the ceasefire between the government forces and the LTTE on April 19,1995, is a maximally unfavorable context for peacemaking efforts of any kind. Prospects of a change for the better in the foreseeable future appear dim, since both belligerents seem committed for the time being to a trial of military strength.

But uncompromisingly hardline positions and a grim context of full-scale warfare are not the only impediments to a lasting peace in Sri Lanka. The flurry of accolades that have greeted the government's set of proposals have detracted from probing critiques of whether these proposals do, in fact, constitute an optimally adequate, appropriate and effective approach to the core problem of Sinhalese-Tamil relations in modern Sri Lanka.

With the Colombo government having more or less successfully appropriated the mantle of "democracy" and "progress," as well as the rhetoric of equality and justice (largely thanks to the LTTE, whose sense of public relations leaves something to be desired), there is a risk that any doubts voiced about whether the proposed reforms constitute the antidote to the ethnic conflict will get dismissed as Sinhala extremist-reaction or Tamil-intransigient-rejectionism. Yet there do appear to be a number of serious questions about the viability and promise of these proposals as the basis for a lasting peace in Sri Lanka.

One year ago, I wrote an article welcoming the apparent commitment of the then newly-elected Kumaratunga government and the Tiger leadership to a process of dialogue, and outlining in both general and concrete terms the fundamental changes that would have to be executed in order to alleviate and eventually resolve the Sinhala-Tamil conflict. In that article, I argued that only "a comprehensive and radical restructuring of the Sri Lankan state," and a "systemic transformation" of the ideological and institutional foundations of the island's polity, could possibly mitigate the deep divide between Sri Lanka's two major communities, and potentially pave the way to a non-military solution.

The appeal of the government's proposals lies precisely in that they appear to herald such a "comprehensive and radical restructuring," and such a "systemic transformation." Yet changes in the power-structure, however ambitious and "far-reaching," cannot in themselves resolve a complicated and polarized civil war, unless they can address in an optimally adequate, appropriate and effective manner what I called "the core issue in this conflict.. the political empowerment of the Tamil people who live in the North and East of Sri Lanka." Whether the overall thrust and substantive content of the government's scheme have the capacity to accomplish this outcome is open to question.

The basic model for the governments plan appears to be the aborted Bandaranaike-Chelvanayakam (B-C) Pact of 1957. Much like the present plan, the B-C Pact envisaged wide-ranging decentralization of administration and devolution of powers to one or more regional authorities to be established in the Tamil majority North and East of the island. The regional body (or bodies) were to be entrusted with a range of subjects, including agriculture, cooperatives, land settlement and development, education, health, industries, fisheries, housing and social services, electricity, irrigation schemes and roads. The regional council(s) were to have the right to tax and borrow to finance projects in their fields of responsibility, and in addition, the central government was charged with providing a certain level of funds from the state's general revenue. To reduce linguistic discrimination, the Pact further stipulated that Tamil should be the official language of all administrative work in the northern and eastern provinces.

Many aspects of the latest proposals bear a striking resemblance to this framework of devolving responsibility in most spheres of everyday governance to regional authorities, while retaining certain critical powers at the "center" and ensuring linguistic impartiality. However, the current proposals are wider in scope than the B-C agreement, in that they seek to divide up the entire island, not merely the North-East, into a cluster of eight regionally organised administrative entities, and reconstitute Sri Lanka as an "Union of Regions."

Its reach and scope put it in a league quite apart from the "Indo-Sri Lanka accord" concluded between Indian prime minister Rajiv Gandhi and Sri Lankan president Junius Jayewardene in July 1987. This latter agreement, for all its grandiose pretensions, merely granted the Tamils limited and conditional concessions within the framework of an explicitly unitary state.

Indeed, given the intimate connection between the two issues, such a strategy makes a lot of sense. But there is a problem in the government's apparent inability or unwillingness to explicitly acknowledge the absolute centrality of the "national question" to Sri Lanka's future, and in its apparent subordination of the proposed resolution to that question to a scheme of administrative reorganization (however "far-reaching").

What are the Tamils (who, according to Sri Lanka's president, "have genuine grievances to which solutions must be found") getting from this deal that the (overwhelmingly Sinhalese) inhabitants of the Central Province, the Western Province, the Southern Province, the North-Central Province, or Uva or Sabaragamuwa provinces, also not getting? The president's explicit denial that her government's proposals are "federal" in nature is both rather ridiculous and somewhat disturbing. Above all, it smacks of a tendency to evade or sidestep the political essence of the problem, and instead reduce it to administrative-legal terms (this, incidentally, was the crippling flaw in the Austro-Marxist Karl Renner's otherwise excellent work on the issue of nationalities in the Habsburg Empire).

If the proposal is couched in such overarching, sweeping terms partly or largely as an elaborate charade, i.e., in order to preempt and deflect hardline Sinhala nationalist opposition, the subterfuge is so transparent as to be positively useless. Hardline Sinhala nationalists may be narrow-minded and obsessive, but they are not idiots. Indeed, any suspicion that the government is "giving in" to Tamil demands under cover of some grandiose devolution project is likely to increase, not reduce, the habitual paranoia and alarmism of this constituency.

It is certainly difficult to find fault with any effort to strengthen institutions of lower-tier administration and grassroots governance in Sri Lanka. Indeed, there is yet considerable space for expanding the role and arena of responsibility of the Provincial councils in the overwhelmingly Sinhala regions of the country, where, unlike in the war-torn North-East, they are functional organs of government. But it is hard to discern what tangible benefits can be derived by parceling out a small country in this manner into a gamut of regional units. Indeed, there is a possibility that this may spawn bloated and unnecessary bureaucratic and political establishments at the "regional" level.

This manifest overkill on "regionalization" is all the more odd because there is hardly any popular demand for such changes in the Sinhala-majority areas. The last time a major proposal was mooted to constitute Sri Lanka as something of an "Union of Regions" (as proposed by the Kumaratunga government) was in 1927, when the Kandyan Sinhalese chiefs petitioned the British to set up three autonomous regions within the framework of a quasi-federal polity-- the units being the Tamil-majority North and East, the Kandyan Sinhalese highlands and adjacent areas, and the low-country Sinhalese lands of the western coast and deep south.

At the time, this was a reasonable proposal worthy of serious consideration (it was turned down). But things have changed very dramatically since then. While a pan-Sinhalese identity has been consolidated in the postcolonial period (as one scholar has written, "low-country and Kandyan Sinhalese are inextricably mixed up today"), it is the antagonism between Sinhala and Tamil collective identities that has emerged as by far the most salient cleavage in modern Sri Lanka.

In attempting to reconstruct Sri Lanka as an "Union of Regions," rather than, as social and political realities demand, a "Union of Peoples," the government's proposals may run the risk of falling between the cracks. In other words, they may fail to mollify hardline Sinhalese nationalist opinion, while falling far short of Tamil aspirations that they be recognized as a "nationality" with a "historical homeland" consisting of the island's North and East.

There is, of course, a very important territorial/regional dimension to the Tamil claim to "self-determination''-- indeed, this is the crux of the conflict in Sri Lanka. But the issue is far too thorny and complex to be dealt with as an administrative reshuffle, and that too in which the Tamil autonomous entity comprises only one of eight such units. There is also no hint in the proposals of the possibility of "asymmetrical federalism" (as in Canada, where Quebec enjoys the status of a "distinct society" within the Canadian federation), which may go some way towards meeting Tamil claims to a distinct identity and to self-rule as a people.

There may thus be a basic flaw in the structure and emphasis of the government's "package." The reform proposal is an interesting and innovative one, but it would be better-suited to a society composed of a multiplicity of territorially based ethnic and linguistic groups (as in, say, India). Indeed, those concerned with a rejuvenation of Indian federalism and more equitable center-state relations in India should find this document to be of considerable interest. But it is not optimally geared to the specific challenges of the Sri Lankan situation, where a protracted armed conflict between two highly articulate and organized groups (one identifying with the state, the other violently opposed) has fostered a highly regrettable situation (which, hopefully, will not endure forever) of political polarization along ethnic lines.

Reservations can also be raised about specific recommendations in the peace proposal. Paramount among these is the question of physical security, for all communities in Sri Lanka but especially for the Tamil population of the North and East, who have borne the brunt of the descent into violence since 1983. According to the draft document, "defence" and "national security" remain under the exclusive jurisdiction of the central government.

This is simply too glib to pass muster. A durable peace settlement is not possible in Sri Lanka in a scenario in which the central state authority wields a monopoly, or near-monopoly, of police and military resources.

Indeed, a major reason for the LTTE's growth and popularity among the Tamils is its effectiveness as a deterrent and retaliatory force to the violence of the (almost entirely Sinhala-Buddhist) armed forces. Most ordinary Tamils harbor a deep-seated fear of the "Sinhala army''-- the mass exodus of Tamil civilians from those areas of the Jaffna peninsula into which the army temporarily advanced during its military offensive (code-named "Operation Leap Forward") in July 1995 was not exactly a thundering statement of confidence in the state's forces.

They are thus unlikely to be willing to entrust their "defence" and "security" exclusively to an army, navy and air force at whose hands they have suffered immensely, and which they believe is systematically biased against them as a people. A "regional police force" (as suggested in the proposals) whose head will be appointed by the President "in consultation" with each region's chief minister, and which will remain subordinate in important matters to a "national police service" controlled by the central government, is just no substitute for effective and adequate security guarantees.

One long-term solution may be to drastically reduce the size of the armed forces, and change their composition to reflect the multiethnic make-up of Sri Lanka's society. But this will have to await serious progress towards peace. in the meantime, there are rumors that Kumaratunga has informally agreed to the formation of a Tamil "regiment" or "brigade" in the army. Apart from the fact that one Tamil military unit will scarcely make any substantive difference, there is the question as to who will be recruited into such an unit. If any attempt is made to formalize the existing cooperation between the army and remnants of several anti-LTTE Tamil groups by creating a "Tamil brigade" out of these latter elements, this is likely to be a farce, and a rather dangerous one at that.

There are also indications that there may be some not immediately obvious riders discreetly attached to the proposals. It has been reported, for instance, that though the regional councils are generally to be left to their own devices, "the President is expected to have powers to impose federal rule in case a regional government is found to be promoting 'an armed rebellion or insurrection'." Apart from opening a window to manipulation of regional governments from the center, this seems to betray a deeply ingrained distrust among officialdom in Colombo of Tamils in general. It is unlikely that a modus vivendi can be reached in a climate in which Tamils continue to be regarded as a bunch of unpatriotic, untrustworthy secessionists.

Then there is the rather nebulous and indeterminate position, in the proposals, of the "Tamil homeland." This is a defensive and somewhat problematic political construct, but nonetheless deeply rooted by now in the Tamils' collective psyche. The integrity and indivisibility of this "homeland" (i.e., the North and East) is the sacred cow of Tamil politics, and a "non-negotiable issue" to almost every Tamil organization, including the LTTE. All that the reform proposal undertakes in this regard is that "one of the [eight] Regions will be constituted by redemarcating the existing boundaries of the present North-East Province with a view to reconciling Sinhala, Tamil and Muslim interests.',

If "redemarcation" and "reconciliation of interests" implies redrawing of borders, this issue is likely to be a major stumbling block. Firstly, a partial dismantling of at least the most recent Sinhala settlements in the North and East would be necessary to assuage Tamil concerns about state-sponsored "colonization" of their "homeland." Of particular significance in this regard is the future of Weli Oya, a settlement complex established in the North, near the junction of Mullaithivu and Trincomalee districts, with the express purpose of disrupting the territorial contiguity of the north-eastern region.

While some other settlement schemes may be subject to negotiation at some future date, if and when better sense prevails on all sides, the continuation of Weli Oya in this form will in all probability be inadmissible to almost the entire spectrum of Tamil political opinion. The recent failed Tiger attack on Weli Oya, while a military setback for the rebels, may well have earned them not inconsiderable political and propaganda points with their constituency. The fact that a couple of hundred Tamil boys and girls gave their lives trying to take out the single most hated symbol of "colonization" cannot have failed to arouse a chord of sympathy even among Tamils who have deep misgivings about the LTTE.

The other crucial question in this regard is how to assure the security and rights of the Tamil-speaking Muslims, the largest minority group in the North-East (17 per cent of the North-East's population as a whole, a full third in the eastern province). The government apparently plans to tackle this problem (and placate its coalition partner, the Sri Lanka Muslim Congress) by detaching Amparai district, which has a Muslim plurality, from the rest of the east coast and carving out an autonomous Muslim council of some kind in this district. There are several problems with this approach. Firstly, it will enrage Tamil nationalists to no end, and possibly precipitate further Tamil-Muslim bloodshed (the Tigers are a force to reckon with even in Amparai, and neighbouring Batticaloa district has a Tamil majority of 70 per cent).

But more importantly, this policy cannot be a viable solution to the status of the Muslim community in the North-East. The pattern of Muslim settlement in the eastern districts of Batticaloa and Trincomalee, and in the northern district of Mannar, resembles a series of quite non-contiguous blotches on the map of the North and East. In most of these places, moreover, Tamil and Muslim communities live cheek by jowl in highly mixed areas -- for example, the volatile coastal town of Kattankudi in Batticaloa, which has a Muslim majority, is surrounded on all sides by Tamil-dominated villages.

It is thus not possible to reconcile Muslim and Tamil "interests" through "redemarcation," at least not without making a complete mess of the geographical coherence of the eastern province and the mixed zones of Mannar. Some other formula for Tamil-Muslim coexistence has to be found in the longer term. One option might be to explore the creation of a network of autonomous municipal and cantonal assemblies and administrative organs in the Muslim dominated pockets, with specified powers and responsibilities within the framework of a "Tamil Autonomous Region" governed by a "Tamil National Authority."

In any serious negotiation process, the status of Trincomalee port, the prize of the east coast, would be yet another major point of contention. At the moment, the government's "package" solves (or rather avoids) this issue at the stroke of a pen, by decreeing that "international ports and harbours" will remain under central control. This is not a satisfactory resolution. Some kind of middle ground will eventually have to be found between Tamil fears that they will be left with a truncated "homeland" minus its economic hub, and Sinhalese anxieties that the Tamils will seek to monopolize the benefits of Trincomalee harbour. As a long-term arrangement, there does appear to be a case for some mechanism of joint control over Trincomalee.

Finally, the government's proposals focus almost exclusively on the notion of decentralization to administrative regions, to the neglect of reconfiguring institutions of executive and particularly legislative power at the central level so as to better reflect Sri Lanka's multinational reality. This is not a trivial or inconsequential matter. The major reason that Tamils are so obsessively attached to the idea of territorially defined autonomy, or to outright independence, is that they have been systematically frozen out of power-sharing at the central level by the Sinhala-Buddhist elite ever since the 1950s.

The rules of Westminster-style parliamentary politics have been twisted and manipulated in such a manner as to result in the permanent marginalization of Tamils from decision-making. This can only be rectified to a limited extent by proportional representation in parliament -- the minorities (Sri Lankan Tamils, the Tamil plantation workers and the Muslims) together account for barely a quarter of the population, and they are deeply divided. Given this demographic configuration, PR in itself cannot offset the possibility of anti-minority rule in the name of the Sinhala-Buddhist majority, nor can it ensure an effective voice to each of the communities of Sri Lanka.

That purpose would be far better served by expanding Parliament into a bicameral institution, in which a "House of the Peoples" would be created with equal representation from the Sinhala, Tamil and Muslim communities. Most multinational states worldwide have some such institution-- Sri Lanka is a stark exception. This chamber could be entrusted with mediating and arbitrating any inter-community disputes, and vested with the authority to accept or reject proposals to change the confederal constitution of the state.

As of now, the government, fully in keeping with its administrative-legal approach to political questions, and its penchant for setting up "commissions" at the drop of a hat on every conceivable issue, proposes to entrust the task of mediating disputes between the center and the regions, and among regions, to an appointed "Permanent Commission on Devolution."

It would be much more useful and democratic to give this responsibility to an elected upper house of parliament in which all communities have equal representation. Not only would this be a check on the reassertion of hegemony by the Sinhala-Buddhist oligarchy, but it would potentially give the Tamils a greater stake at the "center." Other gestures, even if mostly symbolic, could help reinforce that stake. For example, it can be constitutionally mandated that if the President is Sinhala, the Prime Minister must be Tamil, and vice versa (a major improvement on mother-daughter teams, at least) , and that sessions of the bicameral parliament should rotate between Colombo and Jaffna.

Most of the omissions and shortcomings in the Kumaratunga government's reform proposals stem from one source: that they are, by and large, an unilateral declaration of intent by one party to Sri Lanka's conflict. Sri Lanka does need to be transformed in order to be saved, but the basis for such a transformation can only be laid through extensive consultation and negotiation between the main parties to the conflict.

Applause from members of the "intellectual" stratum in Colombo is not really worth very much-history suggests that some, though not all, of these elements stand ready to endorse anything that those in office may choose to do at any given time. Nor is the approval of the small Tamil groups based in Colombo worth much. Bereft of organizational strength, political credibility and popular support, these groups have little capacity to be effective spokesmen for Tamil interests.

When the veteran leader of the Tamil plantation workers, Mr. S. Thondaman, asserts that "Tamils who seek success in a political arrangement have an obligation to concede the leading role of the LTTE," he presumably means that were it not for the struggle waged by the LTTE, the Tamil nationalist movement would be insignificant and there would be no question of devolution proposals of any kind.

It is bitterly ironic, therefore, that the one Tamil movement that does have very significant clout and bargaining power has apparently decided to renounce the path of dialogue, and reverted instead to the pursuit of a costly and high-risk war option.

The Tigers may be able to hold their own militarily in what they call "Eelam War III." They are an extraordinarily resourceful and resilient force, and have demonstrated a most impressive capacity to withstand and survive challenges and setbacks.

Moreover, their strategic warmaking capabilities have been greatly enhanced by the stupendous growth since 1991 of their naval wing, the Sea Tigers (who have established effective control over most of the coastline and high seas off the North and East), and by their acquisition and deployment of surface-to-air missile systems. The Tigers are now capable of mounting ambitious strikes and operations integrating their land and marine forces, and of partly neutralizing the government's air power. Both these developments will have a major impact on the military balance on the ground.

But as I pointed out in my earlier article, there are certain constraints and dilemmas that the Tiger leadership can ignore only at its own peril. Firstly, especially because of the government's economic blockade of the de facto Tiger state in the North (a weapon of slow but potentially deadly strangulation), it is questionable whether Tamil society can indefinitely sustain such an exacting and strenuous war-effort, at least without losing enthusiasm for the political cause (however deep and widespread the popular appeal of "Eelam" may otherwise be) . The latter scenario could well spell deep trouble, in the longer run, for a political movement that relies heavily on mass support.

Secondly, the domestic expansion of the Tigers has gone hand in hand with their growing international isolation (except, of course, their ever active overseas networks consisting of expatriate and refugee Tamils) . The absence of prospects of international recognition ensures that "Tamil Eelam" is doomed to remain a juridical nonentity for the foreseeable future.

As for the government, the pitfalls of its own brand of militarism have been graphically exposed by "Operation Leap Forward" and its aftermath. In military terms, this over-hyped offensive turned out to be a damp squib. The armed forces proved unable to hold on to the territory they temporarily captured, and were compelled to withdraw within days to almost their pre-offensive positions. The offensive also largely failed in its stated objective of drawing out and killing large numbers of Tigers.

But it is the political costs that made this campaign a singularly ill-timed and counterproductive exercise. The Tigers did not offer significant resistance in the initial phase of the offensive only partly because of military considerations, i.e., their refusal to bite the army's bait and be lured into frontal combat. More importantly, the army's advance played directly into the Tigers' political strategy of convincing the Tamil people that this government, despite claims to the contrary, is no better than its predecessors in its wanton disregard for the life and property of Tamils. The LTTE high command was only too content to let the army make some fleeting territorial gains.

The Tigers gained infinitely more in political mileage from the army's indiscriminate artillery and mortar shelling of heavily populated areas in the Jaffna peninsula, which killed several hundred civilians and displaced tens of thousands. To compound this folly, the air force chipped in in inimitable style, with one of its bombers massacring 120 civilians sheltering in a church. The offensive has also led to a severe shortage of food and essential medicines throughout the Jaffna peninsula. If the Kumaratunga government's goal is to bypass the LTTE and reach out to the Tamil population in an effort to win their "hearts and minds," such offensives are likely to achieve just the opposite.

But the fall-out of operation Leap Forward is illustrative of a deeper dilemma that has confounded both the Indian and Sri Lankan regimes in their repeated attempts to cut the Tigers down to size. Because of the kind of movement the LTTE is, and the type of war it wages, it is just not possible to take on the Tigers without victimizing the Tamil population on a large scale. This, of course, only further undermines the state's credentials in the eyes of ordinary Tamils and consolidates support for the LTTE-- whatever the Tigers lose in military terms, the Colombo regime loses even more in political standing.

It is also highly doubtful that the Sri Lankan state has the capacity to conduct decisive military campaigns against the LTTE while its naval and aerial capabilities, essential for waging a successful land war, are increasingly depleted (indeed, almost decimated) by relentless Tiger attacks. Operation Leap Forward has not succeeded in dispelling the widespread perception that the government's forces are militarily inept. And the Tigers remain capable of absorbing significant casualties without losing steam or momentum.

Moreover, the Kumaratunga government's military strategy of going for the Tigers' jugular-- the Jaffna peninsula-- may not be the wisest military option. While the build-up of government troops in bases in and around the peninsula will keep the Tigers under sustained pressure in their most formidable redoubt, the obstacles to decisively eroding the LTTE's military and political grip on its stronghold are daunting, indeed almost insurmountable. In the meantime, the pull-out of troops from the East to bolster the beleaguered government garrisons in the North has enabled the Tigers to establish their writ over vast tracts of the eastern province.

Another massive offensive may weaken the Tigers' military strength somewhat, but it will also result in massive civilian casualties and by implication, a massive political setback to the government's efforts to project itself as an ethnically unbiased, democratic entity. If the partisans of the Colombo government believe that the Tamil people are eagerly awaiting liberation by the armed forces from the LTTE's clutches, they are deluding themselves.

Despite the coercive and oppressive aspects of LTTE rule, and the single-minded, intolerant authoritarianism that permeates the Tigers' nationalism, there is no underestimating either the depths of alienation from Colombo felt by most Tamils, nor their attraction to the ostensibly emancipatory beacon of "Tamil Eelam." As for the sons and daughters of the farmers, fisherfolk and toddy-tappers who comprise the bulk of the LTTE's legions these days, they can be counted upon to follow Velupillai Prabhakaran, their leader and idol, come what may.

There are indications that the Colombo authorities are well aware of all this. A strong sense of frustration at a botched peace process, and a deep trepidation at the prospect of another protracted war, pervades the insecure and affectedly belligerent utterances (including the use of intemperate language not quite befitting a head of state) of Chandrika Kumaratunga. However, the fact remains that Kumaratunga received a resounding mandate last year to not just make peace with the Tamils, but more specifically to talk without any reservations or encumbrances to the LTTE.

The Sinhala electorate showed commendable maturity and patience in giving the "peace candidate" a 62 per cent mandate, even after the Tigers staged a grave provocation by assassinating her opponent in the presidential election of October 1994, Gamini Dissanayake. The failure of the UNP's crass attempt to generate a "sympathy wave" from this incident (by nominating the dead man's widow as his replacement), may have been as much a reflection of the abysmal reputation of cynical, amoral political operators of the Dissanayake variety as of the public's war-weariness. But the people's deep yearning for a lasting peace was unmistakable.

At this time last year, I likened the apparent opening to peace in Sri Lanka to the appearance of a glimmer of light at the end of a very long and dark tunnel. But as the saying goes, there is often another tunnel at the end of the light. Yet it may be overly pessimistic and premature to dismiss that ray of hope as simply a mirage that appeared and disappeared. The paradox of Sri Lanka's bleeding war is that while both sides stock up on arsenals of deadly weapons, a solution crafted ultimately through political means becomes ever more inevitable with each passing year.

SUMANTRA BOSE is Social Science Research Council-MacArthur Foundation Fellow in International Peace and Security at Columbia University's Department of Political Science, where he is completing his Ph.D. He is author of States, Nations, Sovereignty: Sri Lanka, India and the Tamil Eelam Movement (Sage Publications, 1994), "State Crises and Nationalities Conflict in Sri Lanka and Yugoslavia" (in Comparative Political Studies), and "Democracy, Self-Determination and the Struggle in Kashmir" (forthcoming in Modern Asian Studies). His dissertation is titled "Democracy and National Self-Determination: A Comparative inquiry."