Violence: post-accord problems during peace processes John Darby A peace process is often described as a mountain, but a mountain range is a more accurate metaphor. The mountaineering team comprises people who have been at war with each other and now find themselves reluctantly roped together. For most of them, the ending of violence may be a sufficient objective. If this is attained, the travelers are able to peer over the summit. What greets them is far from a tranquil panorama of gentle hills and plains. Instead the view reveals new mountains, some apparently more formidable than the one just climbed. Nor are the new mountains ranged in an obvious sequence from violence to final settlement. Each has its own distinctive terrain and sets its own priorities. Unanticipated peaks may emerge from the mists and demand the climbers' immediate attention. If they are to accomplish their common task they must overcome mutual suspicions and fears, and pick up the necessary skills as they go along. Those who negotiated the ceasefire may not be equipped to negotiate a political accord or to achieve economic regeneration and redistribution. The metaphor has its merits. It provides a comprehensible insight into the complexities of peace processes everywhere, highlighting both their common problems and those peculiar to each setting, but it has one major flaw. It presents peacemaking as an essentially linear progression between war to peace, one that approaches the task as a series of separate disputes, each of which has to be resolved separately. There is nothing per se wrong with this approach. Indeed it reflects the reality facing those trying to negotiate an agreement -- the need to overcome immediate stalemates, the management of compromise, the dominance of short-term problems over fundamental disputes. If there is to be progression toward a peace accord, it is often necessary to set aside underlying distrust in order to make progress on specific issues -- demobilization, disarmament and reconstruction (DDR), constitutional arrangements, policing reform, territorial disagreements. That is why peace accords are dominated by these concrete themes rather than by such abstract concerns as the need to build trust or improve post-war relationships. Yet it is at least theoretically possible for negotiators to reach agreement on every issue, to travel systematically through each mountain in the range, and then discover that the negotiations have confirmed their mutual suspicions rather than created good working relations between ex-enemies. The end of the journey may be greeted, not by cheers from a grateful public, but by heightened suspicions and refusal to work together. Successful peacemaking is about process as well as content, a central theme of all the contributors to this volume. What's known? What's not? There is a substantial and rapidly growing literature on the relationship between violence, ethnicity and conflict resolution. Gurr (1993) and Geller and Singer (1998) have analysed patterns of international violence during the twentieth century and since. Others, notably Sollenberg and Wallensteen (2002) and PIOOM (2000) have charted the changing patterns in contemporary violence throughout the world, and have updated the patterns on a regular basis. As a result of these and other studies we have a much more detailed understanding of how these patterns have changed, most notably away from international war and towards internal violence. We also know more about the optimum conditions for intervention in situations of violence. Doyle, Johnstone, and Orr (1997) have described the changes in multi-dimensional peacekeeping in Cambodia and El Salvador, primarily from a UN perspective. Hampson's 1996 study of five peace settlements, again all cases involving UN involvement, singled out the role of third parties and emphasized the importance of post-accord implementation. The comparative studies by Darby and Mac Ginty (2000; 2003) highlight the complexity of different approaches necessary as peace processes move from violence into negotiations, and from peace agreements to reconstruction. All of these have confirmed the central part played by violence in determining the timing and chances of success for peace initiatives is fully acknowledged. It has been argued that, in order to counter it, military support from external states may be needed to supply `the requisite political will and `muscle' required to bring parties to the table and to end violence' (Crocker, Hampson and Aall, 1999, 26 and 669). In the main, most studies of violence and ethnic conflict have concentrated on how entry points may be found in the conflict cycle for external intervention. Zartman (1989) and Haass (1990) emphasized the importance of taking advantage of the `ripe moment' when the parties in conflict have reached a `hurting stalemate' when the costs and prospects of war outweigh the costs and prospects of settlement for the major combatants. It is still remarkable that the dominant message from research on these peace processes is that modern wars are more likely to terminate at the negotiating table than on the battlefield (Wallensteen and Sollenberg 2004). But what if violence continues after a peace agreement has been reached? Although substantial research attention has been paid to the origins and dynamics of ethnic violence, to the first moves towards negotiations, and to spoiler violence, the threats to post-accord reconstruction is under-researched. Quite a lot is known about policy options on such matters as policing reform, decommissioning and demilitarization, but less about policy options to deal with violence. There is a healthy debate about the part played by spoilers -- defined by Stedman (1997, 95) as `leaders and parties who believe that peace emerging from negotiations threaten their power, worldview, and interests, and use violence to undermine attempts to achieve it' -- in disrupting peace processes, and the subject continues to attract considerable research attention. A substantial literature is emerging about dealing with past violence through truth commissions, and there is also growing interest in the rise in violent crime during post-accord periods, but still a shortage of empirical evidence and case studies. On the other hand there has been very little work so far into post-settlement peace-building and the difficulties facing post-accord states in generating security, economic growth and a healthy civil society. Finally, the relationship between international terrorism and internal ethnic violence is under-rated and under-researched. Violence and peace processes All peace processes must battle against the threat of violence, but its effects are not universal or constant. They vary between different settings, and between different stages in each setting. Although a peace process usually follows open warfare between a highly militarized state and guerrilla-type opposition, the level of violence may intensify during the pre-negotiation phase as combatants try to optimize their negotiating positions. The negotiations themselves are often accompanied by the emergence of more extreme dissident groups. Negotiators also have to confront a new range of priority issues, including demands for the early release of prisoners, disarmament and demilitarization and policing reform, as well as the re-integration of militants into society and consideration of their victims, in accordance with the "issue accumulation" principle highlighted in the introduction to this volume. Although these patterns generally apply, they are also heavily influenced by local distinctions. A peace process is not a predictable sequence from violence to settlement. Even when a peace agreement has been signed, violence continues to produce problems in four principal ways (Darby 2001): 1. Violence by the state The achievement of a peace agreement removes the glue that had held together disparate elements within government during negotiations. Implementation is an exasperating and lengthy process. Even politicians involved in the negotiations may be expecting, and some may be working toward, its downfall. Political and security interests within government have different priorities, and often different views about the durability of an agreement. Höglund and Zartman (In press, 2005) argue that three main elements associated with the state may be tempted to explore a conflict rather than a peace track: `the military, militias and decision-makers discontented with the turn to a settlement track'. In reality the boundaries between them is often blurred. State actors opposed to an agreement, for example, may work in association with militias, or benefit covertly from militia activities, as they did in Congo-Brazzaville and Haiti. In Guatemala too the sharp rise in crime which followed the 1996 peace accords were `often the work of members of the existing police force or the army', as well as unemployed ex-combatants (Economist, June 28 1997). Who is to police the behavior of state actors during peace processes, especially after agreements have been reached? International organizations and other states usually work on the assumption that the state and its agencies, however corrupt and partial, have the prime responsibility for the maintenance of law, order and governance during peace processes. But the state is also a primary actor in the conflict. It controls the greatest number of armed personnel and most of the weaponry, as well as the official agencies responsible for information and propaganda. Further, illegal use of force by the state is often covert, always denied and usually difficult to prove. So how are abuses by the state and by state personnel and agencies to be discovered and countered? If a government abuses its power by continuing to use illegal violence after a peace agreement, as happened in Serbia-Montenegro, the principal onus for monitoring state behavior falls on agencies operating wholly or partially outside the state. Human rights NGOs and international election monitoring played an important role in Latin America, South Africa and other peace processes. One of the strongest guarantees against the abuse of state power is a vigilant press. State control of the media in Burma and Zimbabwe eventually became part of the conflict itself. 2. Violence from Militants The declaration of a cease-fires by militants is never unanimous, and its continuation is contingent on achieving concrete rewards during the subsequent negotiations, such as early prisoner releases and the dismantling of the state's security apparatus. These rewards are rarely immediate. Consequently the pendulum may swing back towards the militants who opposed negotiations in the first place. Four principal militant and ex-militant groups are contained within the range of the pendulum, all of them potential spoilers. Dealers (those who are prepared to negotiate); Zealots (those whose goal is to spoil the process by bringing it down by violence); Opportunists (violent groups which may be persuadable under some circumstances to end violence); and Mavericks (those whose violence is primarily motivated by personal rather than political objectives) (Darby 2001). If peace processes are to survive the different forms of violence undermining them, different policy approaches are required towards these different interests. The key actors in any peace process are the Dealers, whose decision to negotiate initiated the process and without whose involvement the process ends. The key policy objective, it follows, is to ensure that they remain in the process. Their involvement as fully engaged partners is essential for effective reconstruction. Opportunists and Zealots are not always easy to distinguish. Both are usually smaller and more extreme groups that have either been excluded from, or excluded themselves from, the process. The essential difference is that the Opportunists may ultimately be open to participating in the process, while the Zealots are dedicated to its failure. The dilemma then is how to keep the door open for Opportunists while isolating the Zealots. If this is accomplished, the process must find a way of dealing with them, as well as with any Mavericks who also continue to use violence. One measure of success is the degree to which violence can be criminalized by dealing with it through the courts. This becomes more feasible if the Dealers feel strong enough to condemn it. Dealers never find this easy, as it allows Zealots to charge them with betrayal, and to present themselves as the true patriots. Timing is crucial. In Northern Ireland the universal fury at the Omagh bomb in August 1998, and the marginalization of the bombers, allowed Sinn Féin to condemn a republican bomb for the first time. However, the strength of Hamas and the Islamic Zealots in Israel-Palestine prevented Arafat from such an unqualified condemnation. 3. Violence on the Ground After the declaration of a ceasefire, confrontations with ethnic rivals or the security forces often replace the discipline of war. In Northern Ireland, for example, the peace process saw a return to more direct violence between Catholics and Protestants, especially during the annual marching season. The level of conventional crime is also likely to rise. There is an inverse relationship between political violence and criminal activities in times of intense violence. The reasons are varied. Policing is dominated by security concerns. Many hostile areas are not patrolled, leading to an under-recording of crime statistics. Paramilitary organizations claim, and often exercise, the right to police the areas under their control. If an agreement is reached, these conditions may be reversed. The criminal aspects of paramilitary campaigns often transmute into crime syndicates, particularly involving drug smuggling and dealing. The security services, more geared to political than conventional crime during the war years, are often ill equipped to deal with them. Underlying all of this, people have become accustomed to routine violence. Expectations of immediate economic and other improvements, raised during the period of negotiations, are often disappointed after the accord is signed. A significant rise in post-accord crime is clearly a matter for concern, but does it necessarily pose a threat to the agreement itself? Roger Mac Ginty's judgment is that crime has not been identified as the major factor contributing to the outright failure of a peace accord, but it can erode confidence in the agreement among the general population. As the sapping of Unionist support for the Good Friday Agreement in Northern Ireland demonstrates, the accord may survive, but be replaced only by an unstable and uncertain limbo. There are two more specific dangers: first, an authoritarian reaction to crime from government may undermine `the liberalizing intentions and provisions of a peace accord'; and second, the encouragement of vigilantism, a post-war problem deserving greater attention (Mac Ginty, (In press, 2005). In Guatemala vigilantes were responsible for the killing of dozens of suspected criminals in the six months after the accords; more than twenty lynchings followed in 1997 (Economist, June 28 1997). 4. New Issues: Demobilisation, Disarmament and Reconstruction (DDR) and Policing All peace processes give birth to a number of issues that were unthinkable while the fighting continued. Three of the most common are demobilisation, disarmament and reconstruction (DDR), and reform of the police and the security forces. Even if a peace agreement is signed, it can only signify agreement to sketch out the general parameters of how some of these issues should be approached. But the fine details and implementation - what du Toit (2003) has called post-settlement settlements - form the new agenda for dispute. They are among the main tasks of the post-accord years. Virginia Gamba passed somber judgment on recent international approaches to DDR: Key objectives of a peace process are to secure peace, to ensure demobilization, to ensure disarmament, and to assist post-conflict reconstruction and development. If these objectives are not realized, peace cannot be consolidated. Since 1989, almost all cases of multinational peace making and peace support operations have not fully realized the above mentioned objectives (Gamba, (In press, 2005). Many of the worst failures highlighted by Gamba were in Africa, so it is ironic that Africa may be providing a model for other continents and regions through the New Partnership for African Development (NEPAD). In Gamba's words, `the key of success on the ownership formula that lead to the signature in Lusaka was that NEPAD was not presented as a hard and fast plan but as a hard and fast idea'. Most of all, it provided a mechanism for coordination not only between different functions (demobilization, disarmament and reconstruction), but also between the neighboring countries in the region. Police reform can be an equally difficult and contentious post-war issue. The two main options facing societies are the integration of ex-militants into the new force, and reform of policing structures. The integration of ex-militants into the security system removes potentially disruptive soldiers and weapons from the street, legitimizes irregular policing activities, and provides additional police when they are most needed to deal with increased transitional violence. It certainly eased the transition in South Africa, but led to increased confrontations in the Palestinian territories. Reform of policing structures also raises tensions. The police are likely to be `reluctant reformers', so change is unlikely to be achieved rapidly. As Murray (In press, 2005) remarked, The emerging police force therefore will now be tackling criminals, who in the past may have acted as protectors of, and were protected by, the communities of which they form part. In addition, the steps necessary to restore some form of normality (searches, curfews, checkpoints etc.) are exactly those most likely to engender antagonism. Whether integration or reform is adopted, the changes must be accompanied by significant re-training and re-structuring. Policing of a peace process requires that the force must be acceptable to the community at large, which often implies significant changes in its composition. As Murray indicated, behind `almost all of the concerns expressed and observed therein, was the overarching concept of trust'. Violence as catalyst Occasionally certain atrocities provoke universal condemnation and galvanize popular reaction against the perpetrators. Responses to the Omagh bomb in 1998 provided a springboard for post-accord progress in Northern Ireland. Security cooperation between British and Irish governments improved; for the first time Sinn Féin's Gerry Adams condemned a republican bomb `without any equivocation whatsoever'; the Unionist leader David Trimble uncharacteristically complimented the Irish Republic. Within four days of the bomb those responsible for planting it, the `Real' IRA, had declared a ceasefire. Other groups soon followed. Violence, or rather spoiler violence, had become temporarily unfashionable. The responses to Omagh were dramatic, but not unique, in peace processes. The Basque Country and South Africa have experienced similar episodes. Instead of destabilizing negotiations or an agreement, they became a stimulus for negotiation (du Toit 2000). Three factors are critical in converting atrocities into catalysts for peace. They are quality of leadership, the cohesion of the groups in negotiation, and timing. If these factors are in place, the atrocities enable public outrage to be harnessed, rather than simply vented. The likeliness of this happening depends on a range of factors including a relative symmetry of power between, and cohesion within, the main parties. A more precise understanding of how violence influences post-accord reconstruction, as it affects every other stage in the process, provides guidelines for providing the right incentives to `tip the balance toward cooperation in deeply-rooted ethnic conflicts' (Sisk, in press, 2005). South Africa and Northern Ireland All peace processes are subject to these four forms of violence, but the balance between them varies according to local circumstances. In Israel-Palestine, for example, violence from the state and from militants continues to be the major threats to peace (cf. Oberschall chapter in this volume), while Guatemala and El Salvador still struggle with post-war crime. Any peace process might be presented as a four-pillar bar graph, each pillar representing one form of threat In South Africa and Northern Ireland the bar graphs might be presented in this manner: South Africa and Northern Ireland are often regarded as two of the more successful peace processes, a view expressed with greater conviction outside the two places than within. Both processes started in the early 1990s, so have advanced a considerable distance along the conflict cycle. Each produced a peace accord, but is experiencing post-war problems. Although abuses by the state were a feature of the wars in both places, allegations of illegal state violence have declined markedly in post-war years. The reason is to be found in the unusually high external scrutiny on both places. In Northern Ireland, civil rights agencies had alerted the world to collaboration between loyalist paramilitaries and security forces during the early stages of the process. The close involvement of the governments of the United Kingdom and the Irish Republic, the European Union, and the United States ensured that the process was conducted to acceptable democratic standards. International NGOs uncovered the South African government's covert support of Inkatha Freedom Party and others even after negotiations had started. Both peace processes were played out in the full view of the world media, helping to ensure that any violence by individual members of the security forces or police was the subject of proper scrutiny. Attempts by militants to overturn the settlements have also been generally unsuccessful, especially in South Africa. In Northern Ireland republican organizations opposed to the Good Friday Agreement began a bombing campaign in 1998, culminating in the Omagh bomb. Their failure to sustain the campaign was partly a result of popular revulsion against the bomb and support for the peace process among republicans. In both Northern Ireland and South Africa the key to the failure of dissidents, however, was the dominance of the pro-agreement leadership within their respective communities. Sinn Féin and the ANC respectively were able to sustain their support, and to withstand rival zealot movements. It should be emphasized, however, that this support was conditional on successful implementation of peace settlements. During the war years, crime rates were low in both South Africa and Northern Ireland, but rose after the war ended. There is a tendency to underestimate conventional crime during wars, and a greater post-war willingness to report it. Post-war homicide rates in South Africa were dramatically high, but surprisingly constant in the years following the 1994 settlement. So was the number of reported rapes. Burglaries, on the other hand, having fallen in frequency in the immediate post-war years, rose by 95.4% between 1996 and 2003. The pattern was similar in Northern Ireland, albeit at a much lower per capita rate. The number of rapes was constant between 1995 and 2003, but the annual number of murders, attempted murders and culpable homicides almost tripled during the period. Burglaries decreased as they had in South Africa, by 19.8% between 1995 and 2000, and subsequently rose by 80.4%. Disguised among these statistics was a disturbing rise in punishment shootings and beatings by paramilitaries, and attacks aimed at Northern Ireland's tiny (less than 1%) racial minorities. There were 212 reported racist incidents in the last eight months of 2003, 25% higher than the previous year. In general, violent crime remained high in South Africa after the 1994 agreement, but shows some cautious signs of stabilizing. Violent crime appears to be rising in Northern Ireland five years after the Good Friday Agreement, suggesting that post-accord countries should prepare for an imminent rise in violent crime, but that this may level off if political stability continues. Punishment beatings and shootings by paramilitary groups in Northern Ireland have been persistent drag anchors to post-accord reconciliation and reconstruction. Vigilantism was also responsible for some of the worst violence in South Africa, especially by PAGAD (People Against Gangsterism and Drugs), who engaged in tit-for-tat killings with criminal gangs as well as in no-warning restaurant bombs. PAGAD was responsible for twenty bombings around Cape Town, including eight in 2000, leading the Minister of Justice to step up protection for judges and prosecutors involved in cases against PAGAD, and to consider a ban. PAGAD's response was to threaten `one prosecutor, one bullet', a play on the radical mantra `one settler, one bullet' used during the anti-apartheid war (New York Times, September 9 2000). If South Africa and Northern Ireland were spared the worst consequences of post-accord violence from governments and militants, their experiences parted when it came to policing reform and DDR. South Africa accelerated the process of creating a more representative police force by integrating various private armies into the South African army and making the racial composition of both the army and the police more racially representative, while Northern Ireland's police force was reformed more gradually. Between 1994 and 1999 more than 22,700 former liberation fighters and black homeland soldiers had been integrated into South Africa's military, transforming a force which had been 38% black to one with more than 55% black members. In Northern Ireland the Patten Report ushered in a period of gradual reform, including a controversial decision to reserve 50% of new places for Catholic applicants until the religious imbalance -- only 8% of the Royal Ulster Constabulary was Catholic - had been remedied. Northern Ireland's experience of demobilization and disarmament also proceeded at a much slower pace than in South Africa. The decommissioning of weapons became one of the most formidable obstacles in Northern Ireland, yet it was bypassed at a brisk trot in South Africa. It led to the fall of a power-sharing executive, and a return of direct rule. It remained the most potent obstacle to post-war reconstruction. Why did Northern Ireland experience such difficulties in implementing DDR and policing reform, in contrast to South Africa? The Unionist community appears to have moved away from its initial support for the Agreement to disillusion and opposition, whereas support for the South African settlement has remained solid. The reason appears to lie in structural factors. South Africa experienced a transfer to rule by a substantial black majority, encouraging whites to accept what they saw as an irreversible change. The Unionist majority in Northern Ireland was forced to make what some regarded as over-generous concessions to a substantial minority of more than 40 per cent. The former transition, once the agreement has been made, is more likely to stick, but it is difficult not to put part of the explanation down to the quality of the respective qualities of leadership. 9/11 and the War Against terrorism Developments following the events of September 11th 2001has elevated terrorism from the local and provincial to the international stage. President Bush's declaration of a War against Terrorism had a number of important consequences. It was declared globally, but fought locally. It became more difficult to distinguish between the war against al-Qaida and long-standing guerrilla struggles in Indonesia, Palestine, Sudan and other places. This has affected the behavior of states facing internal violence, and it has affected the behavior of militants confronting states, in two quite different forms. The problematic nature of external intervention in domestic conflicts is a persistent theme of several contributions to this volume, and highlighted in the introduction. b. The effect of 9/11 on governments Since the War against Terrorism was announced, a number of governments have opted to adopt tougher approaches against their dissidents rather than to negotiate with them. Human Rights Watch (2003) claimed that `many countries around the globe cynically attempted to take advantage of this struggle to intensify their own crackdowns on political opponents, separatists and religious groups, or to suggest they should be immune from criticism of their human rights practices'. It detailed seventeen such countries. The intensification of anti-terrorist measures was most evident in Muslim countries or countries with Muslim minorities. The Philippines and Israel-Palestine provide two examples. The Philippines was among the first countries to respond to 9/11 by offering air space to the United States. The rewards were substantial. `It's 4.6 billion dollars and still counting', acknowledged President Arroyo (Inter Press Service, March 25 2002). Part of the American aid came in the form of advisors, arms and military aid. 660 US troops were sent to THE ISLAND OF Basilan to help seize control from the Abu Sayyaf group. At the same time the Philippine government reversed its policy of trying to end its war with Muslim rebels. In Israel the government became more willing to defend its tougher security approaches against Palestinian violence after 9/11. These approaches included a willingness to assassinate alleged leaders of Hamas and other groups sending suicide bombers into Israel, the use of torture to gain information on such groups, the expansion of Jewish settlements, and the building of the Israeli separation wall, also called security fence. These policies were not new. The use of torture, for example, had been defended by the Israeli government before the War against Terrorism (New York Times, September 14 1999; Los Angeles Times, December 4 1999). The novelty was the Israeli government's new confidence that the policies were less likely to be criticized by the international community. Stronger actions by governments against terrorists have reached far beyond groups associated with al-Qaida (Independent, June 13 2003). The Spanish government cited international terror in its decision to ban the Basque independence party Batasuna, in 2002, and leaped prematurely to the conclusion that ETA had planted the Madrid bomb in March 2004. New legislation in Thailand authorized the government to use military force in emergency cases such as terrorist attacks, but also `to tackle unrest in the South' (Citation June 23 2003). In other countries moves to combat violent opposition were stimulated by the prospect of American aid. The US provided $20 million military aid to crush the Marxist rebellion in Nepal after the government offered air space to the United States following the 9/11 attacks. `The United States and other western countries showed very little interest in Nepal's internal affair before the 9/11 terrorist attacks', according to Dhruba Kumar, a strategist analyst working with Center for Nepal and Asian Studies (Poudel, 2002). The global threat cited as justification for these and similar actions is often unconvincing. As Richard Falk put it, `the anti-state armed movements that are being targeted are by and large preoccupied with their own territorial struggles of a political character against a particular government in power, and their al-Qaida links are not significantly related to the sort of mega-terrorist vision of Osama bin Laden, but are designed to facilitate their local struggles.' (Falk 2003, 12). Falk goes on to argue that one of the results of the US's approach to what he terms `megaterrorism' is that violence by the state, even when deliberately directed at civilians, is not regarded as terrorism: This purported exemption allows governments around the world to rely on large-scale violence against their civilian populations, and avoid the stigma of terrorism, while at the same time tending to taint all reactive violence from oppressed peoples, even in resistance to foreign occupation, as terrorism (Falk 2003, 9/10). c. The effect of 9/11 on militants It seems certain that the War against Terrorism has provided a boost for resistance struggles in Afgnanistan, Iraq, the Philippines, Palestine and elsewhere. Suicide bombings have spread. Recruitment has risen and the level of ferocity increased. It has had the opposite effect in other places, where sympathy for militants engaged in ethnic violence has diminished. This has particularly affected those resistance groups depending on financial and other forms of support from diaspora populations living in the United States, Australia and Europe. The result, at least in the short term, has been to restrict the activities of such groups, as demonstrated in Northern Ireland and Sri Lanka. Since the Good Friday Agreement in 1998 dissident republican groups in Northern Ireland have attempted to resuscitate the traditional republican sympathies among the Irish diaspora, particularly to solicit arms and funds. Their efforts were considerably affected by 9/11. American supporters of armed Irish republicanism were often conservative in domestic politics, and hostile to terrorism internationally. Increased security curbed their ambitions to launch a bombing campaign in Britain (The Observer, February 3 2003. Tougher legislation within the Irish Republic applied further pressure. In Sri Lanka the Liberation Tigers of Tamil Eelam (LTTE), faced similar problems after 9/11. Before September 11 2001 they operated freely, but since then the pro-LTTE support has gone underground and fundraising has been much more difficult (Ganguly, 2002). Shastri Ramachandaran (2002) rated the War against Terror high in the list of factors that pushed the Tigers to the negotiating table. First and foremost, after 9/11, knowing that their flow of weapons and funds for weapons would get choked, they came out with a strong plea for peace..... Secondly, there was a sharp drop in the flow of funds and arms following the global War against Terrorism'. The result has been a more conciliatory posture from the LTTE. Its leader, Velupillai Prabhakaran, went to great lengths to condemn the September 11 attacks in New York and Washington. Within days, he halted all attacks on Sri Lankan forces, and three months later declared a unilateral ceasefire. The ceasefire has been carefully policed, and Prabhakaran criticized Tiger squads who were involved in minor incidents in the east. It is easy to exaggerate the influence of international developments in Northern Ireland, Sri Lanka and elsewhere on restraining local militant groups. Local factors also contributed. Nevertheless the War against Terrorism appears to have had three parallel, and sometimes contradictory, effects: A. Some governments, encouraged by the growing concern about terrorism, have distanced themselves from the possibility of negotiations with militant minorities in their own countries. Others have justified tougher security approaches as part of the War against Terrorism. B. The tougher security approaches have angered many Muslim groups, and violent resistance has intensified, especially in parts of Asia and the Middle East. C. Paradoxically, some armed militants, especially those depending on diaspora support, have been forced to limit their campaigns. D. Lybia and Pakistan decided to abandon their support of terrorists and to assist in the international crackdown on terrorist organizations. One might argue about the balance between benefits and costs. What is indisputable is that the War against Terrorism has significantly altered the local climates within which ethnic conflicts are conducted. The question is whether this is temporary, or represents a climate change. Policy Implications Violence is unavoidable at every phase of peacemaking, from ceasefire to reconstruction. Far from providing an excuse to avoid negotiations, it is a factor that can actually advance them. For each set of actors, the use of violence is primarily a matter of costs-benefits calculation rather than spontaneity. Sharp rises in the intensity of violence, for example, `force those at the negotiation table to make critical choices - essentially whether to recoil and return to the fight, or hunker down, continue talking, and weather the political storm' (Sisk, (In press, 2005). Whether the choice is positive or negative depends on how militants or the state respond to it. Höglund and Zartman (In press, 2005) have demonstrated how, under certain circumstances, mutual fear may `accelerate negotiations and the implementation of a peace agreement'. Both sides fear the costs of war and the consequences of peace. Strategic rethinking about negotiations is possible during the relatively short moments when both these considerations point in the same direction. A Policy package Policy approaches to violence during peace processes need to combine a comprehensive approach to all forms of violence with an awareness of the subtle differences between them. This applies both to the dynamics of violence during peace processes and to the issues associated with it. a. Violence as Process It is important to analyze the distinctions between different forms of post-accord violence, because each form requires a different policy approach. The best safeguards against state violence, for example, are external monitors -- international and regional organizations, non-governmental groups and a strong independent media reporting illegal actions to the outside world -- a reality recognized by the attempts of some governments to exclude or control their access to scenes of violence. An awareness of the distinction between different forms of violence is even more necessary when dealing with militants, because different types of spoilers require different treatments. Policy should reflect these distinctions. It should aim to support those ex-militants who have entered negotiations, to leave the door open for other militants to enter the process under strict conditions, and to apply the rule of law to isolate and sanction the zealots and mavericks who continue to use violence. A further reason for distinguishing carefully between different forms of violence is that the appropriate policy for some of these problems may actually exacerbate others. For example, the traditional approach to post-war crime -- an increased police presence on the streets - may not only provide targets for zealots, but also lead to confrontations between police and activists, as has happened in the Basque Country and Northern Ireland. b. Violence as Content While it is clear that some policy initiatives may exacerbate other policy approaches, such linkages also offer real opportunities for policy development. Mutual concessions between government and militants on issues of deep mutual suspicion need to be carefully choreographed. Peace settlements are packages which are often assembled towards the end of negotiations, after the components in the package have been negotiated separately. Announcements about partial agreements run the risk of undermining a comprehensive settlement, by subjecting it to premature scrutiny while other talks continue. However, there are strong arguments for deliberately and visibly linking together carefully selected paired components during negotiations. Paired changes, carefully phased, demonstrate that sacrifices have been made by all parties. Reciprocity helps to build up confidence in tandem between related concerns, as pointed out in the introduction about normative principles in peace settlements. Among these paired components are: early prisoner release and the effect of it on victims; demobilization and disarmament; and differing views about policing reforms. The early release of prisoners is characteristically one of the earliest demands from militants. Victims' grievances, including truth commissions, are usually addressed much later, during the implementation of peace accords. Prisoners' release can help to build confidence during the early suspicions of a peace process, but it also reminds the relatives and friends of their victims -- police, soldiers, security staff and, most of all, civilians -- of the hurt they have suffered. At its most emotive level this includes the problem of the disappeared, when family members are still ignorant of where the bodies of victims have been disposed. So early releases of prisoners must be accompanied by recognition of and compensation for the harm done to the bereaved and the wounded. The reciprocation between disarmament of militants and demobilization of security forces is more obvious. It is natural for combatants to want to test their opponents' sincerity before disarming themselves, which can bring a peace process to stalemate. Consequently there is need to orchestrate reciprocal concessions on demilitarization and decommissioning during peace negotiations, and to do this very early in the proceedings. Similarly with the reform of policing. Whether ex-militants are integrated into the police force or the force is reformed internally, early implementation is likely to enhance the chances of successful policing and security reform. The End of the Peace Process? The War against Terrorism illustrates the ease with which external events can influence internal peace settlements. Even before the 9/11 attacks, a view was growing that the optimism for peace processes during the 1990s has not been justified by subsequent developments. Part of the gloom arose from the erosion of peace after so many agreements have been signed. Edward Said entitled his 2000 collection of essays on the Middle East `The End of the Peace Process'. Since then, the model of peace processes developed during the 1990s has been further undermined by the War against Terrorism and the responses to it. This model was dominated by an inclusive approach to peacemaking, one that sought opportunities to negotiate rather than confront; its characteristics were compromise and optimism. The War against Terrorism has encouraged an alternative model, one that sees the possibility of victory over dissent; its characteristics are strength and the presentation of stark choices. The two models co-exist. The question is: Are we moving toward a form of peacemaking that is predominantly driven by security interests, rather than by opportunities for a negotiated settlement? I am grateful for the support provided from a number of sources during the conduct of this research, notably the United States Institute of Peace and the Joan B. Kroc Institute for International Peace Studies. The chapter also benefited from responses from Thomas Pettigrew, Roger Mac Ginty, Dominic Murray, Mirak Raheem and Irene Zirimwabagabo. Bibliography Chester A. Crocker, Fen Osler Hampson and Pamela Aall (eds), Herding Cats: Multiparty Mediation in a Complex World, (Washington DC: United States Institute of Peace press, 1999). John Darby, The Effects of Violence on Peace Processes, (Washington DC: United States Institute of Peace Press, 2001). John Darby and Roger Mac Ginty (eds), the Management of Peace Processes, (London and New York: Macmillan, 2000). John Darby and Roger Mac Ginty (eds), Contemporary Peace Making (London and New York: Palgrave-Macmillan, 2003). John Darby (ed), Violence and Reconstruction, (Notre Dame IN., Notre Dame Press, (In press, 2005) Michael Doyle, Ian Johnstone and Robert Orr, Keeping the Peace: Multidimensional UN Operations in Cambodia and El Salvador (Cambridge, UK: Cambridge University Press, 1997). Pierre Du Toit, `Why Post-Settlement Settlements?', Journal of Democracy , 14 (3), (2003). Richard Falk, The Great Terror War (Olive Branch Press: New York, 2003). Meenakshi Ganguly, `Breaking off the battle', Asia, February 18 2002. Virginia Gamba, `Post-agreement demobilization, disarmament and reconstruction: Towards a new approach', in Darby, (In press, 2005) R. Geary and J. Morrison, `The perception of crime', in P. Stringer and G. Robinson (eds), Social Attitudes in Northern Ireland, (Belfast: Blackstaff Press, 1992). Daniel Geller, Daniel and David Singer, Nations at War: A Scientific Study of International Conflict, (New York, NY: Cambridge University Press, 1998). Ted Robert Gurr, Minorities at Risk: A Global View of Ethnopolitical Conflicts, (Washington DC: United States Institute of Peace Press, 1993). Richard N. Haass, Conflicts Unending: The United States and Regional Disputes (New Haven: Yale University Press, 1990), 232. Fen Osler Hampson, Nurturing Peace: Why Peace Settlements Succeed or Fail (Washington, DC: United States Institute of Peace Press, 1996). Kristine Höglund and I. William Zartman, `Violence by the State: Official Spoilers and their Allies', in Darby, (In press, 2005). Human Rights Watch, Annual Reports, http://www.hrw.org/campaigns/september11/opportunismwatch.htm,. Roger Mac Ginty, `Post-accord crime', in Darby (In press, 2005). Dominic Murray, `Post-Accord Police reform' in Darby (In press, 2005)) Keshab Poudel, `What a difference a Year makes', Nepalnews.com, 22, 12. Sep 13-19, 2002 Shastri Ramachanandaran, `Tamil Tigers out of the Woods', Times of India, November 11 2002. Edward Said, The End of the Peace Process, (New York: Pantheon Books, 2000) Timothy Sisk, `Political Violence After Peace Talks: Searching for the Silver Lining' in Darby (In press, 2005). S.J. Stedman, `Spoiler problems in peace processes', International Security, 22, 2, Fall 1997. S.J. Stedman, D. Rothchild and E.M. Cousens (eds), Ending Civil Wars: The Implementation of Peace Agreements, (Boulder Colorado: Reinner, 2002). Peter Wallensteen and Margareta Sollenberg, `Armed Conflict 1989-2000', Journal of Peace Research 38 (5), 2003. Marie Joelle Zahar, `Spoilers in Peace Processes: Voice, exit and loyalty in the post-accord period', in Darby (In press, 2005) I. William Zartman, Ripe for Resolution: Conflict and Intervention in Africa (New York: Oxford University Press, 1989).