Pii: s0030-4387(00)00016-8

Mediating Conflicts of Need, Greed, and Creed Whilecivilwarsareoftenseenastheproductofunfulfilledbasic needs, internal ethnic conflicts are commonly driven by privategain and collective beliefs as well. Such combinations of motives— mixing need, greed, and creed—pose especially complex challenges formediators and underscore the importance of prevention over cure.1 For oncethe three combine to spark and nourish a conflict, mediation becomes atough job of uncertain entry and long duration.
The Nature of Ethnic Conflict
A perceived collective need that is denied is the basic condition for conflict. Denied need refers to a broad range of grievances, from relief frompolitical repression to redress for economic deprivation. The claims of sometheorists notwithstanding, it is not possible to establish a hierarchy of needs.2Perceived needs are flexible and satisfied at different levels under differentcircumstances, and needs satisfied at one time do not always remain so.
Above all, satisfaction of needs—like all other satisfactions—is a function ofexpectations, which are themselves malleable. Nevertheless, conceptualizingconflict in terms of needs is useful, for it points to the basic dimension ofgrievances, hence of solutions.
1 These categories draw on the creative work of Paul Collier at the World Bank; see Paul Collier and Anke Hoeffler, “Justice Seeking and Loot Seeking in Civil War” (Washington, D.C.: World Bank, 1999); and Paul Collier,“Economic Consequences of Civil War,” Oxford Economic Papers, vol. 51 (1999), pp. 168 – 83. See also I. WilliamZartman, “Managing Ethnic Conflict,” Foreign Policy Research Institute Wire, vol. 6, no. 5 (1998).
2 Abraham Maslow, “A Theory of Human Motivation,” Psychological Review, vol. 50, no. 3 (1943), pp. 370 –96; Edward Azar, “Protracted International Conflict: Ten Propositions,” in The Understanding and Management ofGlobal Violence, ed. Harvey Starr (New York: St. Martin’s, 1999).
I. William Zartman is the Jacob Blaustein Professor of International Organization and Conflict Resolution at the
Nitze School of Advanced International Studies of The Johns Hopkins University. His most recent book is
Peacemaking in International Conflict: Methods and Techniques, co-edited with Lewis Rasmussen (U.S. Institute
of Peace, 1998). He has been mediating the civil war in Congo-Brazzaville for the Carter Center.
Nondiscrimination in meeting needs is a public good, and therefore solving a given conflict may require unhindered access to common or equaljustice. But justice per se is not really one of the issues in a conflict so muchas a concept through which issues and grievances can be analyzed andresolved.3 Resolution of need-based conflict requires problem-solving skills(including detachment) not always available to parties caught up in it.
To the extent that people feel themselves to be targets of repression and deprivation, discrimination can become a cause for rebellion and asource of solidarity among the rebels.4 People may feel targeted because oftheir political beliefs, social position, or ascriptive membership, but whateverthe cause of the discrimination, it provides the coin of identity for theconflicting party. While a conflict can be resolved in its initial stages byremoving the grievances, such a response in the middle stages may not beheard by those in revolt (particularly by the leadership) as they focus theirattention on the more important challenge of building unity and solidaritybehind their revolt.
One of the sources of a sense of discrimination is creed, referring to generalized beliefs and identity feelings. Ethnic conflicts (and, by definition,religious ones) are creed-based conflicts. Creed itself is a “need,” as allindividuals need to feel some level of identity, through ascriptive member-ship and/or belief systems. Such needs vary according to the individual andcontext, the latter being a social phenomenon of greater interest to thepresent discussion than the former. People have a greater need to know whothey are in some circumstances than in others. Three such circumstanceshave a particularly important impact on the need for identity: rapid orprofound change, breakdown of other identities, and discrimination.
So much has been written about these three elements that they need only be noted briefly here. Times of deep change strike at the very notion ofone’s identity and accentuate a need to know who one is as uncertainty swirlsabout. Identity at such times is not just a “taking” affair, but also a matter of“making” and doing, imposing demands on others to respect or honor therequirements of one’s newly (re)affirmed identity. Creed then becomes aspecific aspect of identity, as belief systems and actions give content to simpleidentification. Similarly, when other creeds fail, new ones arise to fill the voidand gather energy from their offensive momentum. Thus, Islamic fundamen-talism capitalized on the failure of Arab socialism, and ethnic identities havegained force from the failures or even just the challenges of nation-building.
But selective, targeted deprivation is the most frequent cause of identity-based conflict. Collective needs for identity turn deprivation intodiscrimination. When deprivation hits identifiable parts of the population, or 3 I. William Zartman et al., “Negotiation as a Search for Justice,” International Negotiation, vol. 1, no. 1 (1996), 4 Ted Gurr, Minorities at Risk (Washington, D.C.: U.S. Institute of Peace, 1993).
when those parts perceive themselves to be selective targets, they takeoffense, using discrimination as a source of solidarity. When discriminationcontinues, the goals of the rebels turn from the redress of substantive griev-ances to procedural demands for control of the system, because redress at thehands of others is no longer trusted. Procedural demands are harder toresolve, since they can be satisfied only by a reallocation of positions as wellas benefits, providing both an additional grievance and a cause for greatersolidarity. And because stronger grievances and solidarity are mutually rein-forcing, problem-solving becomes more difficult.
Creed adds fear for security to need as a source of conflict. Not only do creed-based groups perceive discrimination in distribution (too few ben-efits or too much repression), but they also fear for their very existence,creating a reciprocal fear in the dominant group—a vicious cycle known asthe security dilemma—that lies at the basis of much ethnic conflict.5 A group(or government) feeling threatened at a low level takes measures to assure itssecurity, thereby decreasing the security of the threatening group, which inturn takes measures to bolster its security, forcing the originally targetedgroup to respond, and so forth. This security dilemma best explains thevicious violence between groups that formerly lived harmoniously in Bosnia,Kosovo, Burundi, Rwanda, Congo-Brazzaville, and elsewhere. By producingthe material that holds groups together and gives them solidarity, creedcreates the conditions for a security dilemma to erupt, conditions not pro-vided by need alone.
In broad conceptual terms, creed-based conflicts call for access to separate or preferential justice, either to redress the discrimination or toachieve preferential treatment as demanded by the creed. Demands of thelatter type are extremely difficult to satisfy because they call for compensatoryjustice or “affirmative action,” that is, positive discrimination toward theaggrieved group. Although a return to equality is less difficult to achieve, itusually is not satisfying to the rebel group, which seeks compensation forpast grievances as well as removal of future ones, and in procedural as wellas substantive terms.
Deprivation-based grievances produce political entrepreneurs who articulate demands and organize demand-bearing groups to carry out theconflict.6 Under their leadership the process of conflict continues its circularroute: conflict aims at redress of grievances, but also establishes solidarity,which in turn is necessary for the effective pursuit of conflict. Sometimesleaders are merely the selfless agents of group demands, but in other cases 5 Barry Posner, “The Security Dilemma and Ethnic Conflict,” in Ethnic Conflict and International Security, ed.
Michael Brown (Princeton: Princeton University Press, 1993); and Stephen John Stedman, “Negotiation andMediation in Internal Conflict,” in The International Dimensions of Internal Conflict, ed. Michael Brown(Cambridge, Mass.: MIT Press, 1996).
6 I. William Zartman, ed., Elusive Peace: Negotiating to End Civil Wars (Washington, D.C.: Brookings their personal ambitions become a source of new demands separate fromneed and creed. Such greed is the basic impetus for political entrepreneurswho turn collective need into an instrument of action and solidarity. Themore greed can mask itself in general (need) or specific (creed) grievances,the more it can attract a following and hide its personal nature. Greed is oftennot oriented toward solutions or problem-solving, but toward private gainand continuation of conflict, which is the source of its legitimation. Greed-based leaders of ethnic conflict include Slobodan Milosevic, Charles Taylor,and Rauf Denktas¸.
Obstacles to Mediation
In most cases (two out of every three in the twentieth century), internal conflicts end in a one-sided victory rather than a negotiated com-promise, but such outcomes are notoriously unstable. Victories over ethnicrebels tend merely to push them underground, where they lick wounds, buildmyths, and bide time until circumstances permit their resurgence. Further-more, half of the negotiated solutions were achieved by mediation involvingthird parties. The challenges of negotiations go far to explain the duration ofethnic conflicts and suggest that their abiding nature stems from tactical andsituational concerns rather than the two sides’ inherent, need-based irrecon-cilability.7 The obstacles to mediation can be overcome, but only throughskillful attention to their causes.8 1. Elements of compromise are characteristically missing in ethnic conflict. In their demands for separate or preferential justice, ethnic rebelsseek terms that are ipso facto repulsive to the other side. A formula foragreement based on a shared sense of justice is difficult to find when separatejustice is demanded, and terms of trade are hard to identify since the ethnicrebels have nothing to offer to the government except an end to the rebellion.
Eritrea and Kosovo present instances where the preferential demands werenot palatable to the Ethiopian and Serbian governments, respectively, andwhere the rebels had nothing with which to buy the government’s satisfactionexcept the possibility of ceasing its agitation. Preferential justice is simply notsubject to compromise, as the debate over the legal aspects of identity inSudan illustrates.9 In ethnic conflict, from Algeria to Palestine to East Timor,recognition is the top and bottom line— once achieved, the ethnic rebellionhas won and the government lost. All these aspects make the stuff of anegotiated or mediated agreement difficult to achieve.
7 Edward Azar, The Management of Protracted Social Conflicts: Theory and Cases (Aldershot, U.K.: Dartmouth 8 I. William Zartman, “The Unfinished Agenda: Negotiating Internal Conflicts,” in Stopping the Killing, ed. Roy Licklider (New York: New York University Press, 1993).
9 Francis Deng, War of Visions (Washington, D.C.: Brookings Institution Press, 1995).
2. Elements of context are also missing. Conflicts cannot be negoti- ated just any time; rather, the context must lend itself to a search for a bilateralsolution. Parties—whether the government or rebellion—that are winning orhave an expectation of achieving eventual victory are not likely to be inter-ested in coming to terms with the enemy. Normally, they need to findthemselves in a mutually hurting stalemate, in which each side’s hopes ofvictory are stymied and the continuing blockage hurts.10 But in internalconflicts, such a stalemate is a harbinger of victory for the ethnic rebellion,since its separateness and equality are implicitly recognized. The more typicalsituation is a soft stalemate, which yields no solution, but rather a stable,bearable, de facto compromise, thereby preventing victory by either side andkeeping the conflict alive. Such is the situation in the Western Sahara and inPalestine, conflicts of long duration.11 In the absence of a mutually hurtingstalemate to push the parties to negotiate, a mutually enticing opportunitycan theoretically serve to pull them in the same direction. But, again, ininternal and particularly ethnic conflicts, such an opportunity is almost alwaysunobtainable, since procedural grievances and preferential justice leave littleroom for mutual enticements.
3. The elements of agency are also frequently missing.
A mutually
Negotiation and mediation require a valid spokesman for both sides, yet the position of spokesman is characteristicallya source of conflict within the ethnic group. This is true for all stalemate
three phases of ethnic revolt. In the beginning, ethnic groups represents
tend to be pluralistic and divided, producing several leaders victory for the
seeking to deal with the government in various ways and usually splitting over the question of tactics. Later, during the rebellion.
consolidation period, a leader seeks to unite his group behindhim, but is in competition with others for the position ofsupreme spokesman. Even when the struggle seems to be won and victory ornegotiations are near, break-away leaders are tempted to make a separatedeal for a part of the aggrieved group on terms more favorable to thegovernment than those the mainstream would offer. Indeed, negotiation itselfcan delegitimize a leader as his rivals hew to a harder line. The divisionbetween Ibrahim Rugova and the Kosovo Liberation Army, among the vari-ous Palestinian liberation groups, among the national liberation parties inRhodesia prior to its emergence as Zimbabwe, and between the successiveEritrean liberation fronts are all examples of intra-movement struggles for 10 I. William Zartman, Ripe for Resolution (New York: Oxford University Press, 1989); I. William Zartman, “Ripeness Revisited,” in Conflict Resolution, ed. Alexander George and Paul Stern (Washington, D.C.: NationalAcademy of Science, forthcoming 2000).
11 See Khadija Mohsen Finan, Le Sahara occidental (Paris: Findational nationale des sciences politiques, ethnic spokesmanship involving debate over tactics as well as elements ofgreed and creed.12 Nor are these struggles purely intra party. Each side in the conflict also contests the legitimacy of another side and of a particular person to speak forthem, preferring their own candidate. Thus, each side enters into the debate overthe tactical question within the other side. The struggle for leadership in Chech-nya is a striking example and has served to renew and prolong that conflict.13 4. Furthermore, elements of third-party entry beyond a mutually hurting stalemate are also missing.14 Ethnic conflicts are internal affairs in which medi-ation is by definition perceived as meddling. Mediation automatically strengthensthe rebels because it suggests that the government is unable to handle its owninternal problems. It is therefore resisted by the government for procedural aswell as substantive reasons. Third-party intervention to help a weaker side—usually the rebellion— only exacerbates the solidarity of the government side.
Kosovo was a case in which humanitarian efforts to help the minority made thegovernment feel, or claim to be, more justified in its repression.
As a result, nongovernmental organizations (NGOs) pursuing second- track diplomacy are often more likely candidates for mediation than are othergovernments. However, second-track diplomacy is unable to provide any ofthe constraints and inducements for a solution that official agencies can offer,either to block alternative paths or to reward cooperating parties. It istherefore obliged to rely on the only remaining weapon, simple persuasion,which is no more effective than the presence or absence of more attractivealternatives allows it to be.15 Thus, the Carter Center, mediating in theCongo-Brazzaville dispute in 1999, suffered the same weakness as the specialrepresentative of the secretaries-general of the United Nations and the Orga-nization of African Unity two years before: the inability to prevent troops fromAngola from offering a better alternative to one side than reconciliation withthe leaders of the other, largely ethnic parties.16 5. Finally, identity and solidarity tend to be dependent on conflict.
Creed requires protection (separation) or assertion (superiority), which isachieved by conflict, and conflict is also the way to achieve the solidaritynecessary for effective action. As a result, normal cost-benefit calculations onwhich negotiation behavior is based may no longer work.
12 Barry Rubin, Revolution until Victory: Politics and History of the PLO (Cambridge, Mass.: Harvard University Press, 1994); Stephen John Stedman, Peacemaking in Civil Wars (Boulder, Colo.: Lynne Rienner, 1991); RuthIyob, The Eritrean Struggle for Independence (New York: Cambridge University Press, 1997).
13 Gail Lapidus, “The War in Chechnya,” in Opportunities Missed, Opportunities Seized, ed. Bruce Jentleson (Lanham, Md.: Rowman and Littlefield, 1999).
14 Mohammed Maundi et al., Getting in the Door: Entry into Mediation (Washington, D.C.: U.S. Institute of 15 I. William Zartman and Saadia Touval, “Mediation in the Post Cold War Era,” in Managing Global Chaos, ed.
Chester Crocker, Ren Osler Hampson, and Pamela Aall (Washington, D.C.: U.S. Institute of Peace, forthcoming 2000).
16 I. William Zartman and Katerina Vogeli, “Prevention Gained and Prevention Collapsed: Competition and Coup in the Congo,” in Opportunities Missed, Opportunities Seized.
The Need for Preventive Rather Than Remedial Action
This is a formidable list of obstacles to the mediation of ethnic conflict, and some may find it so daunting that they would simply write offethnic conflicts as protracted or primordial by nature and beyond any reme-dial attention. But there are many other ethnic situations that exist peaceably,without conflict, and, indeed, long periods of stability and coexistence haveat previous moments interrupted most ethnic conflicts. The protracted socialconflict school cannot explain these apparent exceptions, or the differencebetween conflict and nonconflict situations, in time or place. Even in times ofprofound change, competing identities, and targeted deprivation, some situ-ations produce conflict while others do not.
Explanations for this discrepancy come from two different directions.
On the one hand, the contextual conditions for conflict lead to an outburst ofviolence only when political entrepreneurs consciously throw a match intothe tinder.17 On the other hand, conflicts have been prevented by specificmeasures applied before the conflict broke out. That is, third-party efforts canprevent creed-based conflicts from developing, in part through measures toblock would-be political entrepreneurs. These efforts may not usually bethought of as mediation, but that merely indicates that the usual definitionneeds to be expanded if the challenge of containing conflict is to be met.
Identifying measures that can really forestall violence is an ongoing chal-lenge, but some that follow from the previous discussion are evident.
1. Standard-setting efforts. Mediation by human rights groups and global forums to establish criteria for creed-blind opportunity, access, andallocation, as well as for avoidance of creed-based dominance of functions, isan important and growing method to prevent perceptions of discriminationand ethnic grievances. Standards alone will not ensure nondiscrimination, ofcourse, but by setting up visible, normative guidelines they can constitutetargets and yardsticks by which actors can judge themselves and be judged.
One of the most important efforts has been the Helsinki Declaration with its basket of human rights, which gave rise to a European Commissionerfor Human Rights and the Organization for Security and Cooperation inEurope. The Helsinki experience was adapted in the Kampala Document bythe Conference on Security, Stability, Development, and Cooperation inAfrica, which is even more explicit on standards for equitable ethnic treat-ment.18 Critics noted at the time that the World Bank, in its programs inRwanda in the late 1980s, could have attached standards for equitable ben-efits to its agricultural development programs and thus helped avoid the 17 Zartman, “Managing Ethnic Conflict”; Jane Holl et al., Preventing Deadly Conflict (New York: Carnegie 18 Olesegun Obasanjo, The Kampala Document (New York: African Leadership Forum, 1991); Francis Deng and Terrence Lyons, eds., African Reckoning (Washington, D.C.: Brookings Institution Press, 1998); Francis Dengand I. William Zartman, Norms for Africa (Washington, D.C.: Brookings Institution Press, forthcoming 2000).
genocide of the 1990s.19 Such norms and standards are not self-implementingor self-enforcing, and so it is unrealistic to expect too much from them. Butlike other norms, they trumpet an ideal that is hard to ignore even ifdisobeyed. And as the community of obeyers grows, pressure mounts on thedisobeyers.20 Enforcement can then follow, by peers and then institutions.
This is a long and bumpy process, to be sure, but once norms are establishedthey cannot be dismissed lightly.
2. Preempting Need from Creed. Needs can also be addressed before they become creed-based conflicts. To eliminate deprivations entirely may bea counsel of perfection, but the more inequalities in the distribution ofdeprivations are reduced, the more manageable the challenge of preventingviolence becomes. Studies have shown that populations accept austerity andstructural adjustment if the changes have been fully explained beforehandand if the government can persuasively demonstrate competence.21 Absentthese conditions, “IMF riots” take place, which can attract targeted popula-tions if the deprivation has been accompanied by discrimination, as in theRevolutionary United Front rebellion in Sierra Leone in the 1990s, or ifspecific populations are subject to neglect at a time of rising expectations, asin Chiapas in the 1990s, the Moroccan Rif in the 1950s, and Algerian Kabyliain 1980.22 All of these creed-based rebellions could have been avoided by proper government measures of equitable distribution. However, the medi-ator’s role is particularly difficult, since it involves convincing sovereigngovernments to do what they should be doing on their own. In none of thecited cases was a third-party role as clearly indicated as the role of thegovernment itself. In Mozambique, the National Resistance Movement (Re-sistencia Nacional Moc¸ambicana—ReNaMo) that fed on rural disruption andneglect was brought into dialogue with the Mozambique Liberation Front(Frente de Libertac¸a˜o de Moc¸ambique—FreLiMo) government by externalmediation before the resistance was able to consolidate a particular ethnicbase. It must be said, however, that the mediator benefited from a mutuallyhurting stalemate, induced in part by a local drought, that brought home theurgency of settlement to both parties.23 3. Preempting Greed from Creed and Need. The role played by the pyromaniac in setting off ethnic conflagrations makes it important to separate 19 John Eriksson, “An Institutional Framework for Learning from Failed States,” in Evaluation and Develop- ment, ed. Robert Picciotto and Eduardo Wiesner (New Brunswick, N.J.: Transaction Publishers, 1998); ReneLemarchand, The World Bank in Rwanda (Bloomington, Ind.: University of Indiana African Studies Program,1982).
20 Audie Klotz, Norms in International Relations (Ithaca, N.Y.: Cornell University Press, 1991).
21 Joan Nelson, “Poverty, Equity and the Politics of Adjustment,” in The Politics of Economic Adjustment, ed.
Stephen Haggard and J. Kaufman (Princeton: Princeton University Press, 1995).
22 Jeanne Favret, “Revolt by Excess Modernization,” in Arabs and Berbers, ed. Charles Micaud and Ernest Gellner (Lexington, Mass.: Heath, 1972).
23 Ibrahim Msabaha, “Negotiating and End to Mozambique’s Murderous Rebellion,” in Elusive Peace.
the political entrepreneur from his potential following. The image is onlyfigurative, since the entrepreneur is generally not known before he starts onhis adventure. But the opportunities for his appeals to take hold can bereduced by the actions of either the government or external parties. Holdingfederation-wide elections in Yugoslavia in the late 1980s rather than separateelections in each of the republics, for example, would have reduced theopportunity for ethnic campaigning.24 Reducing rent-seeking by enlistingdiamond magnates to ban diamond sales from areas of conflict would closethe opportunity to reap enormous financial gains from ethnic conflicts inAngola or Sierra Leone.25 Some suggestions can be generic, while others mustbe adapted to the specific situation. But in general it is better to deterentrepreneurs at the beginning of the adventure than be faced with the needto remove them at the end.
4. Establish Confidence- and Security-Building Measures. Confidence and security are the only solutions to the security dilemma, because partiesneed to be shown that their fears are groundless and that security can beprovided them without provoking countermeasures. Joint patrols, dialoguesessions, and enforceable laws of equal treatment are among the measuresthat can be used. Consider the Turkish experience in the late 1990s. After ahardline campaign against the Kurdish Labor Party (Partiya Karkere Kurdis-tan—PKK), the Turks extracted a call for nonviolence from the capturedleader, Abdullah O ¨ calan, while also benefiting from a public opinion survey that showed the PKK to have less Kurdish support for a separatist programthan originally feared by the Turks.26 At the same time, earthquakes in Turkeyand then in Greece increased the mutual help and cooperation between theold rivals at a time when the issue of Turkish entry into the European Unionprovided an occasion for an official rebuff and then an apparent reconsider-ation by the union. In sum, a number of disparate events pointing in differentdirections broke the logjam in two sets of internal and international ethnicconflicts. These measures tend to be dependent on actions of the internalparties themselves, but NGOs have been successful in promoting dialogueand confidence-building measures and in training nationals to manage con-flict-prone areas.27 Specific Possibilities for Mediation
Successful mediation is a matter of containing, enticing, and mending.
The mediator must be able to block the impending or escalating conflict, 24 Susan Woodward, Balkan Tragedy (Washington, D.C.: Brookings Institution Press, 1995).
25 The idea is John Collier’s.
26 Meltem Mu¨ftu¨ler Bac¸, “Addressing Kurdish Separatism in Turkey,” in Theory and Practice in Ethnic Conflict Management, ed. Marc Howard Ross and Jay Rothman (New York: St. Martin’s Press, 1999).
27 Harold H. Saunders, A Public Peace Process (New York: St. Martin’s Press, 1999).
draw the parties away from hostile perceptions and actions, and bring themtogether in a more harmonious relationship. “The mediator is in fact also aparticipant, a wielder of power who compels a recalcitrant party to make acompromise it does not want to make.”28 Mediation requires an ability tocreate incentives for need-based situations to receive evenhanded govern-ment attention, open opportunities for creed-based groups to overcome theirfears, and close possibilities for greed-based leaders to achieve their goals bydestroying other groups. Optimally, mediators need to have the power andauthority to threaten the parties with endless conflict if their solutions are notaccepted, and to ensure implementation if their solutions are accepted. Thisis a tall order, requiring the mediator to tap local resources as well as his ownand to collaborate with both the government and ethnic parties. A number ofareas for action can be suggested.
Mediation requires more than a single act, whatever its duration.
Rather, it is a process that involves the establishment of trusting relations withall parties, the opening and closing of alternatives, and the gradual with-drawal of the mediator as trust builds between the conflicting parties. All thisrequires a mix of strategies both to strengthen and soften the separateidentities of the contending sides.
The process of mediation needs to address the parties’ grievances, both substantive and procedural, in an effort to identify difficult compromisesand compensations. Once that is done, it must focus on setting up mecha-nisms for handling future grievances that may arise. The situation of ethnicrelations in Sri Lanka was a story of measures and backtracking in the 1950s,1960s, and 1970s, until the revolt of the Tamil Tigers finally broke out inunmanageable violence.29 The Dayton Accords in 1995 were likewise fol-lowed by a year of inattention when monitored implementation could havemoved the process forward.30 Constructive ambiguity is needed for creative new solutions. Sover- eignty is one of the most difficult challenges in ethnic conflict because it haslong been treated as sacrosanct. However, solutions to seemingly nonnego-tiable dilemmas have been crafted through looser and more creative conceptsof divisible sovereignty. The 1998 Good Friday Accords in Northern Ireland,for instance, contain a provision for three overlapping jurisdictions: theNorthern Irish, the British, and the Anglo-Irish. In the 1995 Dayton Accords,a Serb republic was established alongside a Muslim-Croat federation withinthe united state of Bosnia-Herzegovina. In Tatarstan, a 1994 agreementprovided all but sovereignty for Tatarstan and all but unity within Russia.31 28 “The Disturbing Niceness of George Mitchell,” The Economist, Apr. 11, 1998, p. 45.
29 Michael Kuchinsky, “Yielding Ground: Losses and Conflict Escalation in Sri Lankan Protracted Social Conflict,” in The Understanding and Management of Global Violence; Howard Wriggins, “Sri Lanka: Negotiationsin a Secessionist Conflict,” in Elusive Peace.
30 Richard Holbrooke, To End a War (New York: Random House, 1996).
31 P. Terrence Hopman, “Disintegrating States,” in Preventive Negotiation: Avoiding Conflict Escalation, ed. I.
Not all sovereign states can be divided, but in many cases stark indivisibilitycan give way to creative solutions. Jerusalem and southern Sudan still awaitsuch creativity.32 Communication and interaction are necessary ingredients in any attempt to end conflict and prevent its future occurrence. Leftto themselves, ethnic communities are bound to retreat into their own myths and histories, develop an exclusivist creed, themselves,
and be quick to take umbrage at any perceived slight.Hence, dialogue cannot be permitted to cease.33 To be sure, intenseintercommunity dialogue has not won complete reconcilia- communities
tion in Israel, Ulster, or Cyprus, but it is certain that those often retreat
situations would have been worse without it.34 The question into their own
remains open as to whether personal integration, in which creed-based groups lose their importance, or communitarianpower-sharing and local autonomy, in which they become consolidated, isthe better solution. Arguably, the second may be the temporary expedientwhile the first gradually works its effects.
Where creed-based groups are homogeneous, contiguous, and undi- vided by minorities of their own, where solutions of reconciliation have beenworn out, and where the conflict has gone to a point of no return, separationmay be the only solution.35 Even negotiated under the best conditions withthe best attention to mechanisms for handling future problems, as in theCzech-Slovak and Ethiopian-Eritrean separations, problems are certain toarise and, as in the latter case, may break out in sustained violence. Thedifficulty in achieving a happy divorce points up the need to exhaust allcreative solutions for cohabitation before separation is undertaken.36 Practical Advice
Where does all this leave us with regard to Kosovo, Sudan, Congo- Kinshasa, Congo-Brazzaville, Sri Lanka, and the other ethnic conflicts sure toappear over the horizon? The lessons underscore the difficulty of mediation,it goes without saying, but they also suggest that creed-based conflicts need William Zartman (Lanham, Md.: Rowman and Littlefield, 2000).
32 Cecilia Albin, “Negotiating Indivisible Goods: The Case of Jerusalem,” The Jerusalem Journal of Interna- tional Relations, vol. 13, no. 1 (1991), pp. 45–76; Deng, War of Visions.
33 Saunders, A Public Peace Process.
34 Ross and Rothman, Theory and Practice in Ethnic Conflict Management.
35 Chaim Kaufmann, “Possible and Impossible Solutions to Ethnic Civil Wars,” International Security, Spring 36 Separation cannot be treated here in depth. See Lawrence Farley, Plebiscites and Sovereignty (Boulder, Colo.: Westview, 1986); I. William Zartman, “Putting Humpty Dumpty Together Again,” in The InternationalSpread of Ethnic Conflict, ed. David Lake and Donald Rothchild (Princeton: Princeton University Press, 1998);Chaim Kaufmann, “Intervention in Ethnic and Ideological Civil Wars,” Security Studies, January 1996, pp. 62–100.
not be permanent and irresoluble, but rather may be susceptible to carefulanalysis and intervention.
The first such lesson is: Do not repress your minorities, or they just may end up being a majority in a territory you claim. Minorities can regroupgeographically if they do not already inhabit contiguous territory, forming amajority in an area for which they then claim self-determination. Similarly,groups that break away with their territory in the name of self-determinationmay well find their homogeneity illusory and face claims from minoritieswithin their own midst. Once self-determination enters the agenda, it canbecome a runaway locomotive.
Secondly, ethnic relations require tending lest mending become nec- essary, and mending lest rending become inevitable. There is no substitutefor good governance, accountable democracy, and normal politics.
Thirdly, what history has joined together, let not momentary passions put asunder. Secession is costly, rarely as satisfying as promised, and almostnever preferable to a solution based on autonomy and federalism.
Fourthly, court-mandated counseling is preferable to divorce, and if divorce should occur, do not count on alimony. Since courts are not availableto handle internal or interstate disputes, states—alone and in community—need to take advantage of available official and unofficial mediators to helpmaintain their unity and harmony. If a state does break up, the breakawaypart should not count on any largesse from its former sovereign, which willbe preoccupied with restoring its torn soul and honor.
Fifthly, find ways to improve your own status that do not damage the status of others. Both ethnic communities and their opponents, be theygovernments or other ethnic communities, need to adopt the uncomfortablepractice of thinking of each others’ needs while pursuing their own creed andneeds. Obviously, an ethnic conflict tends to reinforce self-centeredness, butthat only makes concern for the other side more necessary.37 Lastly, if you have to call in the cops, be sure they shoot the gangsters and not the victims. Peacekeeping forces should strive to make things better,not worse. There is often an auto-da-fe´ quality to measures punishingcreed-based conflict, in which the greed-based actors are the last to feel thepunishment. Only now is awareness growing about how expensive remedialconflict management is by comparison to preventive action, in terms ofmoney, lives, and productivity.38 The high costs of punishment and rehabil-itation by the world’s policemen, whether from the United Nations, NATO,regional organizations, or the United States, should lead them todevote more attention and energy to preventing ethnic conflictsbefore they erupt into violence.
37 Jeffrey Rubin, Dean G. Pruitt, and Sung Hee Kim, Social Conflict (New York: McGraw Hill, 1994).
38 Michael Brown and Richard Rosecrance, eds., The Costs of Conflict (Lanham, Md.: Rowman and Littlefield,

Source: http://www.cerium.ca/IMG/pdf/Zartman_Mediating_Conflicts_of_Need_Greed_and_Creed-2.pdf