Ethnic and Intercommunity Conflict Series
General Editors: Seamus Dunn, Professor of Conflict Studies and Director, Centre for the Study of Conflict, and Valerie Morgan, Professor of History and Research Associate, Centre for the Study of Conflict, University of Ulster, Northern Ireland
With the end of the Cold War, the hitherto concealed existence of a great many other conflicts, relatively small in scale, long-lived, ethnic in character and Intra-rather than inter-state has been revealed. The dramatic changes in the distribution of world power, along with the removal of some previously resolute forms of centralised restraint, have resulted in the re-emergence of older, historical ethnic quarrels, many of which either became violent and warlike or teetered, and continue to teeter, on the brink of violence. For these reasons, ethnic conflicts and consequent violence are likely to have the greatest impact on world affairs during the next period of history.
This new series examines a range of issues related to ethnic and inter-community conflict. Each book concentrates on a well-defined aspect of ethnic and intercommunity conflict and approaches it from a comparative and international standpoint.
Rather than focus on the macrolevel, that is, on the grand and substantive matters of states and empires, this series argues that the fundamental causes of ethnic conflict are often to be found in the hidden roots and tangled social infrastructures of the opposing separated groups. It is the understanding <of these foundations and the working out of their implications for policy and practical activity that may lead to ameliorative processes and the construction of transforming social mechanisms and programmes calculated to produce long-term peace.
Coming out of Violence Project
General Editors: John Darby, INCORE (Initiative on Conflict Resolution and Ethnicity) University of Ubter, and Kroc Institute, University of Notre Dame, and Roger Mac Ginty, University of Lancaster
Titles include:
John Darby and Roger Mac Ginty (editors) THE MANAGEMENT OF PEACE PROCESSES
The Management of Peace Processes
Edited by                                         NOTICE: TH!S MATERIAL MAY BE
John Darby                                  PROTECTED BY COPYRIGHT LAW
INCORE, University of Ulster and               (TITLE 17 U.S. CODE)
Kroc Institute, University of Notre Dame
and
Roger Mac Ginty
University of Lancaster
p<*»Uni™fityUbr»it«.
HIHI DQ2557157W
p3iQľ3V6
Introduction: Comparing Peace Processes
John Darby and Roger Mac Ginty
Example is the school of mankind, and they will learn at no other.
(Edmund Burke, Letters on a Regicide Peace, 1796)
Westphalia was a peace, not only the end of a war. In itself, it was an extraordinary, drawn-out business, much more of a 'peace process' than a rapid treaty. These days, we are suddenly very interested in how peace is made. The delegates in Westphalia had been negotiating for four long years before the documents were ready to sign, during which many more thousands died. The process was handled by something like our own 'proximity talks', the technique used at Dayton for Bosnia, in Northern Ireland before Good Friday, in some of the Israel-Palestine talks. The combatants sorted themselves out into two sides. The 'Imperialists', broadly Catholic and pro-Habsburg, settled into the cathedral city of Muenster, while the northerners and Protestant rulers housed their delegations at Osnabrueck, a few miles away.
Courtiers galloped back and forth between the cities with fresh drafts or new clauses, and sometimes an august diplomat made the journey. The main delegations never actually met at all. Finally, on 24 October, 1648, the grand signings took place - separately. The terms of the Peace were proclaimed and read to the crowds. Cannons fired from the walls of Muenster, and the cathedral bells boomed over the flat, sodden countryside. The war was over.
(Neal Ascherson, Independent, 8 September 1998)
1
2   Management of Peace Processes
Introduction
Mention the term 'ethnic' and you are likely to hear one of two responses: references to either folk songs, beads or rustic costumes, or to violence. The second response has probably gained dominance in recent years. International developments since the end of the Cold War -in the former Soviet Union, Sri Lanka, Bosnia, Rwanda and Burundi -may mark out the 1990s as the decade when ethnic violence crept out from under the carpet and came to haunt us again.
Shadowing this spectre was another phenomenon, less frequently observed. In the last few years a number of durable and traditional ethnic conflicts seemed to lurch along the continuum from violence towards settlement, or at least compromise. The 1990s witnessed a unique moment in history, when a number of persistent ethnic conflicts was transformed into a less violent stage.
•   Perhaps the most dramatic has been the rapid move in South Africa towards free elections and majority rule in April 1994. The constitutional settlement has significantly shifted the emphasis to the need for fundamental social restructuring and more equitable distribution of resources, and the process towards a more lasting peace continues.
•   In the Middle East, talks between Israel and the Palestine Liberation Organisation (PLO) have provided sufficient momentum for the transfer of territory and a level of autonomy. The Oslo Accords have subsequently been undermined by political suspicion and by violence, but it is premature to write off the peace process there.
•   The long-standing conflict in the Basque Country is showing cautious signs of moving towards resolution. After years of violence, concessions by the Spanish government have given the Basques greater, if not complete, control over their own affairs. Euskadi 'ta Askatasuna (ETA) ended its violent campaign in September 1998, by which time it had lost the support of most Basques.
•   The declaration of the Irish Republican Army (IRA) ceasefire in August 1994 ushered in the start of a cautious peace process in Northern Ireland. Despite the ending of the IRA ceasefire, the process continued and culminated in the Good Friday agreement in 1998. The post-agreement problems continue, but so does the momentum of the process.
These are the front-page cases. Behind the headlines the process of making peace has been transformed during the 1990s. Over the course
Introduction   3
of the 11 years between 1988 and 1998, at least 38 formal peace accords were signed which aimed at resolving internal conflict in 33 states. Some of these agreements were fresh attempts to resolve conflicts which had existed before and during the Cold War. Other more enduring conflicts, such as Northern Ireland, New Caledonia and Western Sahara, had scarcely been touched by the struggle between the superpowers.
These agreements were not evenly spread geographically. Nineteen took place on the African continent and 12 of these involved Sub-Saharan African states. It is also worth noting that the African agreements included several of the more successful attempts at conflict settlement, including those in South Africa, Namibia and Mozambique. Elsewhere, there were two each in South Asia and Central Asia/former Soviet Union, four apiece in Europe and East Asia/Padfic, three in the Middle East, and five in Latin America/Caribbean. To look at it from another aspect, most post-war conflicts were in the Global South, and all but three agreements - Northern Ireland, South Africa and Israel-Palestine - came from Third World or formerly Second World states.
All of the accords dealt with conflicts which were mainly concentrated within national boundaries, as distinct from wars between neighbouring states. This does not mean that external powers were not involved in their settlement; indeed, none of the conflicts was free of external influence. Nor does it mean that all the conflicts concerned the entire geographical territory of a state; some were confined within a region of a state. By this perspective 29 of the accords were national pacts, while 9 addressed sub-national disputes, usually between the state and an ethnic minority.1
What's in an accord?
In all of these places the ending of violence has brought new problems as well as new opportunities, not least of which is the threat that it may return.; When violence ends in durable violent conflicts, the policy agenda shifts radically from military containment towards a new set of problems which have not previously been faced: how to include ex-paramilitaries in political negotiations; how to handle such sensitive issues as amnesties, political prisoners, the decommissioning of weapons and the policing of divided societies; and the problem of economic reconstruction:
Every conflict has its own distinct culture, history, attitude and social development, but the issues confronted in peace accords are often quite similar. The accords share universal themes - ceasefires, the inclusion
4  Management of Peace Processes
of armed participants in negotiations, disarmament, amnesty for political prisoners and reform of the security branches of government. The issues of autonomy and human rights are constants, and elections are usually necessary for the ratification of the agreement. Socioeconomic development, if it features at all, is usually cast in a secondary role. Internationally, the removal of foreign forces and the return of refugees may draw formerly hostile third-party states into the negotiations.
The ending of violence is often regarded as the principal objective in a peace process, or at least as an important first step towards settlement. In many processes, including South Africa, Israel-Palestine and Northern Ireland, ceasefires were preceded by secret negotiations between the representatives of guerrillas and either government or other intermediaries. Even when secret negotiations led to ceasefires, the resulting talks were almost inevitably carried out to a background of disruptive and destabilising violence by more extreme spoiler groups and dissidents. Ceasefires never eliminate violence completely, and the continuation of violence may feed the distrust which is a feature of most internal peace negotiations. In particular, it takes time to.remove the suspicion that those representing ex-paramilitaries are prepared to call their soldiers into action to influence the negotiations or to remind other negotiators of their power. This highlights differences about whether security-related issues such as amnesties or the release of prisoners should be introduced at an early stage as confidence-building gestures for paramilitaries, or if arms decommissioning should be a precondition in order to reassure constitutional politicians.
Serious negotiations often do not begin until each side has attempted to maximise its bargaining position through military offensives. Paramilitary groups may attempt to demonstrate their strength and staying power in order to persuade the government that military victory is impossible. In Angola, both the opposition National Union for the Total Liberation of Angola (UNITA) and the government Popular Movement for the Liberation of Angola (MPLA) launched military offensives to win back territory and deter the other from securing a military victory in preparation for negotiation. In the case of more established regimes like those in Mexico, the Philippines, Russia, Israel, Guatemala and Cambodia, government policies shifted between limited and total warfare. The campaigns of militant rebels such as the Moro National Liberation Front (MNLF) in the Philippines, the PLO in Israel, the IRA in Northern Ireland, the Coalition Government of Democratic Kampuchea (CGDK) in Cambodia and the Guatemalan
Introduction   5
National Revolutionary Unity (URNG) ultimately led to more forthright negotiations. However, new offensives can also interrupt the peace process, and did in Angola, Liberia and Rwanda.
A key component of many settlements is the early release of prisoners. Some go further and include the integration of former combatants into the national armed forces and/or police, a transition often linked to the demobilisation of troops and decommissioning of weapons. A new unified armed force was established in Angola under the Bicesse Accord. A proportional balance between the military was agreed in Mozambique, while the People's Liberation Army of Namibia and the Southwest Africa territorial force were merged into a National Defense Force in Namibia. There are many other examples: Tuaregs in Niger, the new South African security arrangements, the United Tajik Opposition in Tajikistan, the CGDK including some Khmer Rouge in Cambodia, and the amalgamation of the MNLF into the Armed Forces of the Philippines. These attempts at integration did not always succeed. Notable failures include those attempted by the Mondou Accord in Chad and the Conakry Accord in Sierra Leone. In Sudan defecting troops fired on the government troops with whom they had recently amalgamated. Amalgamation was unable to prevent genocide in Rwanda.
In many cases the accords outlawed discrimination or recognised the rights of a group to exist, while some countries have taken the further step of forming truth and reconciliation commissions. The 1996 Philippines agreement and the agreements in Latin America placed a special emphasis on the protection of human rights - indeed Guatemala has probably the most extensive consideration of human and indigenous rights of any agreement. Reconciliation commissions were incorporated into the agreements in Mozambique and South Africa, and discussed in Chad. Discriminatory legislation was repealed in South Africa and Namibia, and a Victims Commissioner was appointed in Northern Ireland. There - as in a number of other cases including El Salvador, Guatemala, Nicaragua and Haiti - the peace agreement went far beyond political and constitutional agreement. The agreements included undertakings to review security arrangements, emergency legislation and the criminal justice system, as well as professionalising the police and armed forces.
International organisations and peace processes
Over the last decade the number of UN-brokered agreements has declined in relation to other approaches to peace accords. The United
6   Management of Peace Processes
Nations was directly involved in 16 of the 38 accords agreed between 1988 and 1998; the other 22 had no substantive UN involvement. Within this broad pattern, however, there has been a significant shift away from UN involvement. Six of the ten peace agreements (60 per cent) signed during the first five years resulted from UN initiatives. Only 10 of the 28 agreements (36 per cent) signed during the following six years involved the UN. The change in pattern has been accelerating; of the 15 agreements reached since the start of 1996, all but 2 were agreed without UN assistance.
Economics and strategy both had a say in this abrupt shift. The United Nation's peacekeeping budget dropped from $3 billion in 1995 to $1 billion in 1998, and the international political will for extensive peacekeeping operations has greatly diminished. At the same time the UN's peacekeeping strategy, notably in Cambodia and El Salvador, was gradually shifting towards a multi-dimensional approach. This was based on a stronger attempt to secure consensus rather than to impose force. It required co-operation between the state and the international community, and between UN peacekeepers and a host of other actors, including civilian police and non-governmental organisations. The UN was involved in peacemaking, peacekeeping and peace-building, sometimes simultaneously. The UN mission was 'not merely to create conditions for negotiations between the parties, but to develop strategies and support structures that would bring about a lasting peace'.2
It seems likely that the UN's peacekeeping role will continue to diminish. In the immediate future the most common form of peacemaking will fall more heavily on the Non-Governmental Organisation (NGO) community, individual states and protagonists. More significantly, the shift from UN peacekeeping towards internally negotiated peace processes is the most significant development in international conflict reduction since the mid-1990s.
Peace processes - what are they?
There is no such thing as a peace process', thundered Ian Paisley at the Democratic Unionist Party (DUP) conference in 1998, not for the first time. Nor was he the only politician in Northern Ireland to suggest this. Every peace process is punctuated by claims that it has ended or never started, for reasons of either frustration or tactics. It is not an easy charge to sustain or refute, because there is no universally
Introduction   7
agreed definition of a peace process. When does a process start? What constitutes an irretrievable breakdown? When can one claim it has been successfully completed?
It is important to distinguish between the common characteristics of peace processes and their essential characteristics. Ceasefires and secret negotiations may stimulate negotiations, but often fail to do so. The core dispute is often, but not always, about demands for and opposition to greater devolution. Some peace processes are heavily influenced by local external actors, other hardly at all. There are no rules about how many internal parties are sufficient to deliver a durable agreement, although most recent negotiations have tended to include as many parties as are willing to abjure violence.
Looking at these common features, it is possible to identify five essential criteria which define a peace process.
1.  The protagonists are willing to negotiate in good faith
At some stage the protagonists have made a conscious choice (often at leadership level) to engage seriously in negotiations, as distinct from approaching them as the continuation of war by other means. This is not always an easy decision, due to residual mutual suspicions, and the ultimate test is the Protagonists' willingness to sustain a process despite inevitable setbacks.
Willingness to negotiate does not imply an inevitable outcome. What is inevitable is the tension between the negotiators' needs both to co-operate and to compete with each other. The success or failure of a peace process is determined primarily by the management of this tension.
2.  The key actors are included in the process
The definition of 'key actor' is determined by the local peculiarities, but three empirical observations are constant. First, meaningful negotiations include the existing government of the area in question. Second, they include those paramilitary groupings which command significant support or an electoral mandate; third, they include all elements which have the power to bring about the downfall of an agreement. The third consideration - the inclusion of veto-holders -is the most controversial and subjective. It makes a distinction between mainstream paramilitary groups and spoiler groups, and recognises that the distinction may be altered by time. Significant spoiler groups may opt out or be excluded, to be admitted later under certain conditions - a morally ambiguous concession which, it might be argued, encourages spoiler violence. The uneasy principle
8   Management of Peace Processes
of 'sufficient consensus', devised during the South African process, provides a useful test for admission.
3.  The negotiations address the central issues in dispute
The central issues dividing the participants, usually constitutional, are sometimes deferred until procedural or confidence-building measures have been agreed. But agreement to include the fundamental issues is an essential criterion of a peace process.
4.  The negotiators do not use force to achieve their objectives
Violence precedes peace processes and continues as an unavoidable background during them, and the willingness to continue negotiations in spite of disruptive violence is a crucial test. A peace process cannot progress if any of the participant parties continues to be systematically involved in political violence. Consequently it must include explicit agreement by participants not to engage in violence in order to gain an advantage in the negotiations.
5.  The negotiators are committed to a sustained process
A peace process is distinguished from a truce or a treaty by its duration. Above all else, it is a process. It must be able to withstand pressures, sometimes violence, from without. Peace processes are often suspended, sometimes for a substantial period. The decision as to whether it continues or has been terminated depends on the extent to which the key actors regard it as a feasible approach to resolving their differences.
Outside these general principles peace processes follow greatly varied directions. Pre-negotiation contacts may be used to test the ground, and may involve external or internal mediators. The official process usually begins with a public announcement and often with a ceasefire. Once started, the rules and sequence of negotiation are determined by negotiators who, by definition, have little experience of negotiation. It is not essential to start with a denned constitutional or political outcome for the process, but a peace process cannot be regarded as completed unless a political and constitutional framework has been agreed. Even if it is, the detailed implementation presents other opportunities for failure. Throughout, the process is likely to run into periods, sometimes extensive, of stalemate. The ultimate test of its durability is its ability to retain all of its key characteristics and to leave open the possibility of restoring momentum.
The study of peace processes
Until the 1970s the study óf ethnic conflict rested uncomfortably in an academic space somewhere between sociology and international
Introduction   9
relations, until Glazer and Moynihan's Ethnicity? published in 1976, identified a space for it. Since then it has elbowed itself into an expanded empirical and theoretical space, tracking the growth of ethnic violence since the 1980s.
The initial emphasis was on country-specific studies, and especially on the causes and dynamics of ethnic violence.4 The five conflicts which are the focus of this book, for example, have attracted a level of empirical research which shows no sign of abating. These empirical building blocks have illustrated a dynamic ethnic cycle from conflict prevention through violence to conflict management and resolution and, ultimately, to post-settlement peacebuilding. The substantial and rapidly growing literature on conflict resolution has been reviewed by Miall, Woodhouse and Ramsbotham (1999).
Paralleling these empirical studies is a growing number of important analyses of the changing patterns of international violence, notably by Gun (1993) and Geller and Singer (1998). Two recent longitudinal studies in particular have moved on from analysis of recent and current worldwide violence towards the analysis of its ending. The Uppsala project identified 66 armed conflicts which were terminated between 1989 and 1996. The terminations were classified into three groups:
a.  Peace agreement           19
b.  Victory                        23
c.  Other outcome             24
The project did not attempt to distinguish ethnic from other conflicts, and, although the vast majority was internal conflicts, the total also includes international wars. 'Peace agreement' was defined as 'an arrangement entered into by warring parties to explicitly regulate or resolve the basic incompatibility'.5
The 1997 PIOOM (Interdisciplinary Research Program on Causes of Human Rights Violations) World Conflict map identified 22 'peace accords and ceasefires' since the mid-1990s. Most of these remain volatile, and repeated violations are reported. To quote PIOOM, 'A Peace Accord does not automatically mean an end to violence. A number of High Intensity Conflicts still figure on the current list of Low Intensity Conflicts, while others have fallen back to the level of Violent Political Conflicts.'6
These quantitative studies have been supplemented by a welcome rise of interest into how it might be ended. Zartman (1995) focused on negotiation within the process, while Doyle, Johnstone and On (1997) explored multidimensional peacekeeping in Cambodia and El Salvador,
10   Management of Peace Processes
primarily from a UN perspective. Hampson's 1996 study of five peace settlements, again all involving the UN, singled out the role of third parties as especially important and laid proper emphasis on the importance of implementing as well as signing agreements. There is still a dearth of comparative work on the successful management of peace processes. As Zartman put it, 'negotiation theory must be further developed, tested against situations of internal conflict, and refined accordingly, so that deductive guides to the potentialities for negotiation can be established'.7
The coincidence of these developments across the world, and the emergence of a literature on internal peace processes, provides an opportunity for comparative research. It also provided the foundations for the 'Coming Out of Violence' research project on which this book is based.
The Coming Out of violence Project
Background and aims
'Coming Out of Violence' set out to explore the transformatiqn from violence to agreement in ethnic conflict. It monitored the peace processes as they evolved in five selected areas, in order to identify those factors which accelerated or frustrated them, and to distinguish between the regional and universal. One of the paradoxes of ethnic conflict is that each conflict is essentially parochial, with its own history and social context, but shares with other conflicts similar characteristics and problems. 'Coming Out of Violence' aimed: to provide an empirical base for comparative study; to develop a better theoretical understanding of peace processes; and to assist those - policy-makers, politicians, international organisations, NGOs and community groups -who are currently dealing with these problems, and who will be in the future.
In each area under study a distinguished academic or team monitored the peace process as it evolved or collapsed over a two-year period. All were working to an agreed methodology, which made allowances for local differences. The partners participating in the study were: John Darby and Roger Mac Ginty, 1NCORE, University of Ulster; Pierre van der Post du Toit, University of Stellenbosch; Tamar Hermann, Tami Steinmetz Center, Tel Aviv University, and David Newman, Ben Gurion University; Ludger Mees, Universidad del Pais Vasco-Euskal Herriko Unibertsitatea, Bilbao; and Paikiasothy Saravanamuttu, Centre for Policy Alternatives, Colombo.
Introduction   11
The principal aim of the research, and the factor which determined the selection of the five case studies, was to reach a fuller understanding of how peace processes are initiated, how they evolve and what factors support or frustrate their success.
All research involves selection bias. The five areas participating in 'Coming out of Violence' were not selected at random. They were chosen to achieve certain objectives. It is, of course, possible to argue that a different selection would produce different conclusions, well illustrated by the apt title of Barbara Geddes's article 'How the Cases You Choose Affect the Answers You Get*.8 Our approach is to explain as honestly as possible our criteria for selection, and to let the reader judge. The deficiency of many comparative studies', as Horowitz put it, 'may not be selection bias as much as it is the failure to be expliciť.9 Three main criteria determined the selection of the five cases.
First, they were durable ethnic conflicts. Previous studies of peace processes have often focused on those which involved UN interventions. The focus of this research was on internal peace processes, those where the main motor came from within the parties in dispute rather than international organisations.
Second, the five areas were selected because they had embarked on a peace process or were likely to do so. No two peace processes start and finish in the same time span - indeed it could be argued that the year 1994 marked both the elections which ratified the South African process, and only the start of the Northern Ireland process with the first ceasefires. This raises the risk that external factors, which affect all peace processes, had altered the context to the point where comparisons are invalidated. In effect, however, all five processes were creatures of the 1990s, and were played out within the same post-Cold War international context.
Finally, the 'Coming out of Violence' project was distinctive in adopting an evolutionary as opposed to a retrospective approach. This creates a number of problems in the selection of cases. It is difficult to forecast if a process is likely to continue - peace processes are notoriously vulnerable to violence and political extremism. The advantages, however, are considerable. The participating researchers were able to record the unvarnished views and reactions of participants at various points as processes stumbled along the road to agreement or breakdown. Such concepts as 'ripeness', for example, are more sensitively assessed contemporaneously than retrospectively.
These reasons have, we hope, produced what Collier and Mahoney called an appropriate 'frame of comparison' for the research.10 The
12   Management of Peace Processes
reader should be aware of what falls outside the frame, those subjects which we excluded from our concern, as well as what is included. Those interested in the UN's role in peacemaking, for example, should be aware that this is not the subject of this book, and should read the findings in conjunction with studies which address that concern. Nor should research into emerging peace processes be regarded as a laboratory study. Its aims are more modest. It is to provide insights into an increasingly important category of peace processes.
The approach
The monitoring procesy involved the collection and classification of newspaper cuttings, official documents, relevant legislation and policy papers, and academic commentaries. An e-mail discussion allowed useful and speedy debate on critical methodological and definitional problems. Each participant established a reference group of politicians, activists and others who have special insights into developments. The aim was to record immediate responses to developments rather than retrospective wisdom, and to provide an inside commentary on the problems involved in peacebuilding. The end result was a lorfgitudinal survey of emerging or collapsing peace processes.
The themes
The success or failure of any emerging peace process depends on the interaction between a wide range of variable influences. The balance and the interaction is different in different settings and may alter significantly as each evolves, but elements of all are likely to have influence on the outcome. For the purposes of this research the main influences were divided into six main themes:
•   violence and security issues
•    economic factors
•    the role of external actors
•    popular responses
•    symbols and ritual
•    progress towards political/constitutional agreement
Violence and security issues
The most obvious threat to any peace process is that violence may start up again. Indeed it seems likely that a combination of factors would make its return inevitable: an entrenched culture of violence; the continuing presence of arms; failure to move towards successful negotiations and compromise; and unwillingness to remove the security apparatus erected
Introduction   13
during the period of violence. The key question then is the resilience of the peace process itself, and its ability to continue, despite a resumption of violence, and perhaps to engender further ceasefires. The ending of a ceasefire does not necessarily mean the ending of a peace process.
Economic factors
One of the ironies of any peace process is that the ending of violence often carries a peace deficit rather than a peace dividend. The promise of new investment may not compensate for jobs and incomes lost as a result of the termination of hostilities - police, soldiers and extra security personnel - thus leaving discontented and volatile armed elements to threaten the stability of the peace process.
The role of external actors
External actors, including the UN and regional organisations, may also have a negative or positive effect on the outcome of a peace process. If a neighbouring country feels that its interests may be threatened by either instability or change induced by ethnic violence, it may be tempted to intervene or influence events. Alternatively, a common desire to encourage a stable accommodation may act as a spur towards agreement. Similarly, diaspora populations may finance warring parties or act as mediators between them.
Popular responses
A common feature in many societies suffering from ethnic violence is a popular desire for peace. This is often frustrated by its inability to influence either those using violence or those engaged in political discussions. The success or failure to mobilise popular opinion, and the strength of civil society, are often important factors in determining the outcome of a peace process.
Symbols and ritual
Ethnic conflicts rapidly take on symbolic expression - flags, songs, cultural affiliations, religious and linguistic revivals. The symbols and rituals add an additional layer of complexity that has to be unravelled. Symbols have an equal importance in binding wounds during and after a peace process.
Progress towards political/constitutional agreement
Political discussions aimed at reaching agreed constitutional and political structures sometimes start while violence is still prevalent, or follow
14   Management of Peace Processes
the declaration of a ceasefire. A common sticking point is an insistence on preconditions - insistence on abjuration of violence or decommissioning, or the inclusion of certain parties in the negotiation process. Once the parties reach the table, they are confronted by disagreements about the management of the process and ultimately about their contradictory constitutional and political objectives. Underlying the process is basic mutual mistrust.
This list is neither comprehensive nor weighted. In some settings, for example, interventions by neighbouring states have been a critical determinant of progress, while in others tardy political progress or a continuation of violence have been more influential. It is precisely the variety of influences, and the interaction between them, which is the central justification of this project
The literature on peace processes is unbalanced in favour of United Nations intervention, and the five cases described in the following chapters may illuminate the growing proportion of peace processes which are primarily the result of voluntary internal negotiations. Among the problems inherent in selecting on-going processes is the impossibility of forecasting how they are likely to develop. In retrospect, the five areas illustrate almost the entire spectrum from success to failure. The 1994 elections in South Africa still stand out as a most remarkable voluntary transfer of power; the revolution was neither entirely radical nor entirely peaceful, but now appears to be irrevocable. There are still major obstacles to overcome before the 1998 Good Friday Agreement in Northern Ireland can be claimed as either irrevocable or completed, but it has moved the conflict on to a different level. It may be too soon to claim a great deal from the Basque process, but the 1998 ceasefire by ETA and the increased willingness of the Spanish government to contemplate negotiations bears certain resemblances to the situation in Northern Ireland two years earlier. At the time of drafting this book, many observers have declared the Camp David Accord dead and the Wye River agreement on its death bed; an air of despondency dominates, and the process has certainly decelerated, but it is sometimes overlooked that the Israel-Palestine dispute has accomplished what many other intractable conflicts have not - the transfer of both territory and authority to the Palestinians. Only in Sri Lanka is it difficult to find any evidence of change or reasons for optimism; the main antagonists have not yet reached the essential turning point which initiates all peace processes - the realisation by both sides that they cannot achieve their objectives by force, or that the price of military victory is not worth the reward.
Introduction   IS
Notes
Internal Peace Agreements: 1988-98 (date of most important accord)
UN                           Non-UN (national)           Non-UN (sub-national)
1998	Namibia	1989	Lebanon	1992	Mali
	Western Sahara I	1991	South Africa	1995	Niger
1991	Angola I	1992	Czech Rep./	1996	Mexico
	Cambodia		Slovakia		Philippines
1992	El Salvador	1993	Afghanistan		(Moro)
	Mozambique		Israel/Palestine I	1997	Bangladesh
1993	Ethiopia/Eritrea	1994	Djibouti		Russia
	Haiti	1995	Nicaragua	1998	New Caledonia
	Rwanda	1997	Chad		Northern Ireland
	Somalia I		Sierra Leone		Philippines
1994	Angola II		Somalia II		(NDF)
1995	Bosnia		Sudan		
	Guatemala	1998	Guinea-Bissau		
	Liberia		Israel/		
1997	Tajikistan		Palestine H		
	Western				
	Sahara II				
10
M. W. Doyle, I. Johnstone and R. C. On (eds), Keeping the Peace: Multidimensional UN Operations in Cambodia and El Salvador (Cambridge: Cambridge University Press, 1997).
N. Glazer and D. P. Moynihan (eds), Ethnicity (Cambridge: Cambridge University Press, 1976).
Northern Ireland, perhaps an extreme case of academic overload, illustrates the level of research activity and publications. Three registers of research into the conflict indicate no slowing-down of interest since the 1970s: in 1972 there were 175 projects, rising to 517 in 1981 and to 605 in 1993. A similar pattern applies in Israel-Palestine and South Africa. M. Sollenberg, States in Armed Conflict (University of Uppsala: Department of Peace and Conflict Research, 1997), p. 32.
PIOOM (The Interdisciplinary Research Program on Causes of Human Rights Violations) (University of Leiden, 1997).
I.  W.  Zartman  (ed.), Elusive Peace: Negotiating an End to Civil Wars (Washington DC: The Brookings Institute, 1995), p. 4. B. Geddes, How the Cases You Choose Affect the Answers You Get', in J. Stimson (ed.), Political Analysis (Ann Arbor: University of Michigan Press, 1990), pp. 131-50.
See Chapter 12 in Donald L. Horowitz's book The Deadly Ethnic Riot (California: University of California Press, forthcoming). D.  Collier and J.  Mahoney,  'Insights and  Pitfalls:  Selection Bias  in Qualitative Research', World Politics, 49,1 (October 1996), pp. 56-74 (68).