un site de ressources pour la paix

Iréné est un site de ressources documentaires destiné à favoriser l’échange de connaissances et de savoir faire au service de la construction d’un art de la paix.
Ce site est porté par l’association
Modus Operandi

En librairie

Transformation de conflit, de Karine Gatelier, Claske Dijkema et Herrick Mouafo

Aux Éditions Charles Léopold Mayer (ECLM)


Brussels, novembre 2007

Conditions for successful work in the field - Accompaniment

Focusing on the activity of Accompaniment, what are the main aspects that will make it an effective activity

Mahony and Eguren provide a valuable examination of deterrence in general and of accompaniment as deterrence (1) specifically in Unarmed Bodyguards. Here, I include their ideas of the necessary conditions for successful accompaniment.(2)

”Accompaniment cannot directly threaten very much. Its presence is more of a hint - a suggestion that consequences may occur.” A series of conditions must be met:

  • 1) The accompaniment and the activist have to communicate clearly to the aggressor what types of actions are unacceptable. If the message is complex or refers to documents, the accompaniment must know that the aggressor understands the content of the documents. Subtleties must be articulated. ”Deterrence cannot work if the aggressor does not know which actions will provoke a response.”

  • 2) Deterrence commitment must be articulated: the aggressor must know in advance that an activist is accompanied and that there will be consequences to an attack. The problem here is that those giving the order may know but not inform the death squad carrying it out.

  • 3) The aggressor must believe that an organisation is capable of carrying out its resolution. The chain of communication from the accompaniment to the international community to governmental pressure must be clear and effective. In practice, each link is uncertain and results cannot be guaranteed.

  • 4) The aggressor must seriously consider an attack and then decide not to carry it out because its perceived costs are higher than its benefits. Usually it is impossible to find evidence of this.

One additional condition is that the accompaniment must know who the aggressor is. Death threats are often anonymous or the identity of an attacker must be deduced from little evidence. International reaction, in this case, may be mistargeted. Or an accused government may claim it has no control over a specific aggressor, which is difficult to disprove. Deterrence is demonstrably effective only if the potential attacker knows who the accompaniment group is, what it will do and what the consequences of an attack will be. Deterrence strategy requires access to information - clear analysis of who the attacker is and what political pressures will influence him or her.(3)

Some aggressors may not care about international pressure. There may even be a faction within the state apparatus, which politically opposes the ruling party and would attack human rights activists or international observers to discredit the seated government. [For example, PBI experienced that CERJ (4)members out in isolated villages were facing local thugs who seemed impervious to pressure, and that Civil Patrols which patrols, who attacked unarmed GAM members in front of the press and blatantly threatened even police and government representatives, were unaffected by foreign presence.(5) ]

Deterrence fails when the aggressor decides that the attack is worth it, because other benefits outweigh the political costs. All that is left is to apply the threatened consequences as firmly as possible after the attack, in the hope of changing the calculation next time around.”(6)

Additionally, the activist must not need to be in hiding for any reason. ”Semiclandestinity and accompaniment are both valid security strategies when used separately, but the combination is somewhat problematic. The mere presence of the foreigner makes hiding more difficult, and the protective function of the accompaniment is lost if the potential attacker is unaware of it.”(7)

Accompaniment cannot be used without a strong Emergency Response Network and/or other means of informing and swaying the international community. The stronger the international interest in a particular region, country, organisation, or individual, the more likely it is that accompaniment can deter an attack.

It is a condition for all teams that the accompanied person is an activist and is unarmed. PBI has set conditions on themselves too: ”PBI would not do political organising or form groups, would not initiate activities that Guatemalans themselves could initiate, would not attempt to cover the entire national territory, and would at all costs avoid any indiscretion or disclosure of information that might put others in jeopardy.”(8)

Accompaniment must continue uninterrupted as long as the threat exists to the person or group, provided that it is wanted. According to Labour organiser Sergio Guzman, Guatemala, ”It’s not that the threats necessarily stop when you have accompaniment. Accompaniment questions the threat… You call off the accompaniment when you feel you’ve reached a politically different situation. It doesn’t mean the systematic violence is over. It’s more subjective when the accompaniment has fulfilled its task of calling the violence into question.”(9)

PBI closed its accompaniment project in El Salvador in 1992 after five ”precarious years.” ”The war was over, and although violence and inequality continued in many forms, protective accompaniment was no longer the service Salvadorans wanted from foreign NGOs.”(10)

Each organisation must set conditions regarding safety and risks, knowing that accompaniment involves risks as described in the following examples. In El Salvador between 1987 and 1989, PBI members ”were inside movement offices while the army surrounded them… death squads set off bombs at night while accompaniment was inside. On 14 different occasions, PBI volunteers were detained, interrogated and invited to leave the country. In hundreds of other instances they were stopped by soldiers on the street, interrogated and intimidated. Yet the more the government harassed foreign volunteers, the more the Salvadoran civilian movement valued the accompaniment.” (11)Grenades were thrown into the Peace Brigades house in Guatemala, and three team members were knifed (12) by an unknown assailant.(13)

A last condition involves having other activities in place. ”Accompaniment is much more than an immediate tactic… It required substrategies for communicating with the army, building political clout, making diplomatic contacts, recruiting and training adequate volunteers, finding funding, and developing an emergency response network. These substrategies are conditioned by basic principles but are also designed to alleviate resource limitations and actively change the political context.”(14)


High visibility is not necessarily a condition for effective presence. The parties whom one deems potentially violent must absolutely be aware that peace team internationals are among the citizen population but they need not know where those individuals are at any particular time. The Michigan Peace Team in Chiapas used this factor quite uniquely.(15) Team members were asked to enter the country unobtrusively on a tourist visa. Further, they were required to travel at night as they entered the villages that invited them and thereafter to remain indoors during the daytime so as not to be seen. The Chiapan host villages believe that only if the internationals are not seen but known to be in the area can they protect more than a few.(16)

In most cases, however, teams have made their presence and position known to as many parties as possible, for example Cry for Justice participants strolling through the streets of Haiti.

Emergency response network

A list of 10 occurrences, which BPT felt necessitated the use of their Alert Network, is given in fiche 16. Questions developed by them regarding conditions for use are below:

  • 1. Are other organisations/agencies working on case?

  • 2. Is it possible to co-ordinate actions with them?

  • 3. Is event in question a single case or repeat? Is it an exemplary case? Are people involved in the events known to you? Exemplary cases or cases indicating a worsening of the situation should have priority.

  • 4. How serious is the case? Is there a danger for life or health? Threats to life or health have priority over other human rights violations.

  • 5. Did you double-check the information? Did you witness the event yourself? Are there at least 2 independent sources? How reliable are sources?

  • 6. Who wants the team to activate the network? Do persons or groups concerned want the team to take action? Do they want the case to be made public? Never act against the will of the people concerned. Co-ordinate with them which facts might be made known.

  • 7. Would taking action on the case be an additional danger to the people concerned? To third parties? To the team? Never endanger people, even if it is only a slight possibility, without having asked them. If the safety of the team is concerned, consult with the co-ordinator.

  • 8. How often has the alarm been triggered? Alarms cannot be triggered too often. Their effect and the willingness of people to take action wear off easily.(17)


BPT member Erik Torch itemised conditions for peacebuilding in Kosovo/a: ”In working on peacebuilding there are several points that we need to bear in mind. First and foremost, building peace needs to be focused on the relationship. To do this will require personal time spent with people as well as planning and implementing projects…It also means that when designing such projects a lot of listening must be done with interested community leaders, activists and NGO’s to make it something that they see as worthwhile and not simply imported and forced upon them. Secondly the work has to be looked at through the lens of sustainability…Thirdly it must be done within three contexts or spheres: locally (Kosovo/a), sub-regionally (the South Balkans) and regionally (Europe) since the war involved all three.”(18)


In their empirical analysis, Müller and Büttner make the observation that nonviolent interventions ”do not automatically combine the peace strategies.” [Peacekeeping, peacemaking, peacebuilding.] Most are ”action-based” and use one of the strategies. The study shows interventions to have greater de-escalating effect if they are ”process-oriented,” aiming to affect the conflict dynamic while developing methods during the project itself, usually applying a combination of the three peace strategies.(19) All three strategies are necessary in severely escalated conflicts and must be parts of an integrated process of conflict resolution.

Higher-level goals such as prevention do not seem achievable through the sole approach of having teams of volunteers in crisis regions. If NGOs working for peace do not want to lose sight of these higher-level goals, team activities need to be integrated into a broad-based approach involving activities by a number of different actors.(20)

Clear goals

An internal condition for measurable or successful peace intervention is the presence of concrete and clear goals, which match the activities that are planned and possible. This was a painful issue in 1995 in Croatia and in 1997 in Serbia, when BPT experienced at first hand the run-up to surges in escalation. In Croatia, the volunteers thought it a defeat not to have secured a foothold from which they could exert some influence. By contrast, the organisers had never even expected this of the project. In the deliberations about how to react to a possible escalation in Kosovo/a, the question of the possibilities for achieving prevention was not even discussed—it was so far beyond perceived capacities.(21)

BPT’s peacebuilding work also suffered from the lack of a comprehensive plan that would have focused its work more systematically. Guiding notions such as human rights, non-violent conflict resolution, the channelling of information, and the provision of skills grew out of ideas about possible helpful roles and the subsequent practical realisation of these based on team experience.(22)

Other examples of poorly formulated goals would be the Gulf Peace Team and Mir Sada. GPT lacked refined goals, had not formulated their strategic objectives on a reasonable projection of the numbers they could mobilise, and had estimated the impact they could have on sheer optimism. Likewise, imprecision about the goals was one reason for the failure of the Mir Sada intervention. Christine Schweitzer suggests that the vagueness of goal formulation is obvious in the original appeals of Beati and Equilibre as well as the common Mir Sada appeal:

  • To stop the war, starting with a ”cease fire” during the Mir Sada period.

  • To be in solidarity with each person suffering from this war, regardless of his/her ideology, sex, religion or ethnic origin.

  • To represent civil interposition against violence.

  • To support and encourage a multi-ethnic population to live together in Bosnia.

  • To implement negotiations that will go beyond armed conquest and will impose both respect for, and the safeguard of, human rights under international law.

Nowhere was the aim of stopping the war elaborated.(23)

Without goal specificity, activities will be vague in focus and morale will suffer from uncertainty about even small successes and an undermining sense of failure.

Clarity of concepts and principles

Serious flaws in the Mir Sada plan are also found in a lack of agreement about neutrality, no agreement about the appropriateness of talking to Serbian and Croatian leaders, a vague understanding or agreement about the term nonviolent interposition, and even what to do when they arrived in Sarajevo. Lack of clarity about neutrality and how to achieve it usually means there will be none. In the case of Mir Sada, Schweitzer describes the results thus: ”During our stay in Prozor there was the lasting rumour that the Bosnian troops did not attack Prozor because of our camping there. But we did not actually do anything for the Bosnian side, which was shelled every day with grenades from a place about two miles from our camping site. A half interposition is not a successful example of interposition, but taking sides in a war!”(24)

Non-partisanship rests on good communication with both or all sides in a conflict and on a carefully selected physical position of the intervention. Mir Sada succeeded at neither.


Essential to conscionable intervention is the condition that locals welcome the team and have autonomy in creating their own solutions to problems. Galtung warns that intervention must not be left entirely to the outside: the broader the role defined for a third party the more it does to turn the local population into clients, taking away what might have served them in building a conflict resolution capacity, leaving them with solutions rather than challenges.(25)

”Colombians need to decide Colombian issues by themselves,”(26) says Bennett of WfP. ”We don’t advise Mexicans on what to do,”(27) says Poen of SIPAZ. ”We support them in working out their own problems.”

Communication with all parties

PBI has demonstrated again and again how absolutely essential it is that peacekeeping activities include communication with authorities involved in the conflict. ”An effective deterrence strategy can be hindered by an inability to communicate with the state. Salvadoran officials dismissed accompaniment groups as subversives; thus PBI and others were hesitant to identify themselves publicly. Likewise, it took years in Sri Lanka and Guatemala for PBI to build up a relationship with the government.”(28)

Director Robert Poen says of SIPAZ, ”We have what would be called collegial relations with local organisations. We try to connect with all groups: human rights, civic, Zapatistas, paradistas… We have contacts with the paramilitary… We try to reach out to all points of view without discrimination. It’s risky and complicated to talk to one group and then go to the next group and find that they won’t talk to you. The tendency is for people to assume that we’re pro-Zapatista. We’re trying to overcome that.”(29)

”The effectiveness of nonviolent peacekeeping is probably to a decisive extent dependent on how constructive the relations to the individual parties are, which further forms of pressure can be activated, and how far and how effectively pressure from civil society is exerted on the conflict parties.” These activities require ”reliable, long-term work and cannot be achieved though short… actions” (30) One difficulty in this is that beyond a certain stage of escalation, conflict parties view outsiders only as ”friend or foe.” Social relations are often what enable teams of nonviolent intervenors to monitor or go between mutually threatening groups to prevent renewed escalation.(31)

Attitude of aggressor

There are groups who impede work toward peace and may take direct action to prevent it. Individuals or groups perpetuate war even without obvious gain. ”Rejectionists” become so strongly identified with a cause and make such sacrifices for it that its end is a threat to their identity. ”Irreconcilables” are willing to suffer in order to inflict pain on others in return for pain experienced. An end to war represents their undoing; perhaps they’ll be tried for their actions. War is the only means by which they can survive.

”Too often… peacemakers appear surprised by rejectionists’ irreconcilable violence and allow it to interrupt steps that have been carefully constructed to bring the majority to agreement. When they do so, the actions succeed. When peacemakers signal rejectionists that their disruptions will undermine momentum toward peace, they reinforce rejectionists’ resolve to carry out such acts.”(32)

An aggressor might not fear international condemnation or repercussions of actions. Further, when a military group is threatened, it may become even less responsive to the need for a good relationship with the international community.(33)

”The processes of nonviolent conflict management, resolution, and transformation work best where state systems are democratic and/or have high levels of political, economic, and social legitimacy. Where regimes are controlled by military and paramilitary groups, they tend to believe that it is more efficient to rule by terror rather than persuasion. In these circumstances the opportunities for normal adversarial politics, played according to widely accepted rules of the game, are minimal. State-sponsored terror and political repression force individuals, interest groups, and political parties to either withdraw from the political system or to engage in violent or nonviolent resistance.

…The problem facing those seeking alternatives to the politics of terror is how to generate safe political action spaces while minimising the risk of arbitrary arrest, torture, disappearance, or death. The construction of such action spaces is a prerequisite to nonviolent problem solving. A number of problems are associated with generating creative resistance to terror:

  • How to turn victims into protagonists

  • How to overcome individual and collective fear

  • How to develop deterrents to political and military threats

  • How to promote a political system that enhances the positive consequences of political activity while minimising the negative.”(34)

“Relevant to all human rights pressure is the principle that it is about power as well as justice… in postwar or postterror transitions, there is a strong tendency for governments to respond to such pressure not by doing the right thing but by doing something. That something often involves throwing the least costly scapegoats to the wolves.”(35)

Any role chosen by NP will fail if the conflict is not carefully evaluated by careful assessment of the attitude of the belligerents

Emergency response network

As has been said, the strength of other tactics depends upon the breadth, speed and reliability of an emergency response network, which ”mobilises in the shortest possible time relevant international publicity which cannot be ignored… Here use is made of repressive power, which third parties can exert on a particular conflict party. However, this power must first be activated and be prepared to let itself be mobilised to act for particular values, such as human rights against weapons export. The use of this power is hence not always available …but rather somewhat precarious.” (36) The network and its reliability must be functional before entrance into a conflict area.


PBI has the following thoughts on timing for intervention:

  • 1. Intervenors need credibility in order to gain access. This can be built through long-term relationships or through the reputation of intervenors via past work or position.

  • 2. Is there hope of success given the resources of intervenors?

  • 3. Is the conflict divisible to enable intervention in only one segment? Or is there a possibility of doing test intervention in one area?

  • 4. Is peace desired by all parties? Are the parties motivated to resolve conflict? Are the parties hurting enough to welcome intervention?

  • 5. Is doing nothing worse than the prospects for intervention?

  • 6. Are domestic factors conducive to intervention?(37)

Visibility in the Field

There are multiple issues involved in deciding how much visibility is advantageous in the field. One is the practical matter of legal standing within the country. If team members have entered the country on tourist or religious visas, visibility of their peace work might give a non-welcoming government the opportunity to deport them. But if the government has agreed to the team’s presence, visibility has proved helpful. (38) Strategies of deterrence depend on high visibility of the accompanier or interpositioner.

Still other questions have to do with the effect of this spotlight on third party internationals. Does it detract from the credibility and confidence of local peacemakers or does it reinforce them?

The opinion of Michael Beer, Nonviolence International staff, is that third parties should strive for the minimum visibility necessary to get the job done. Over-exposure might bring on a political attack or a slide into dependency. Under-exposure nullifies the benefits of intervenors (39)and may decrease credibility.


  • (1) : See Mahony/Eguren 1997 p. 85, for the distinction between general deterrence and immediate deterrence.

  • (2) : Mahony/Eguren 1997: 84-87.

  • (3) : Ibid: 225.

  • (4) : Council of Ethnic Communities ”Runujél Junám.”

  • (5) : Mahony/Eguren 1997: 64.

  • (6) : end Mahony/Eguren quote

  • (7) : Mahony/Eguren1997:162.

  • (8) : Ibid: 15.

  • (9) : Moser-Puangsuwan/Weber 2000:143.

  • (10) : Mahony/Eguren 1996: 181.

  • (11) : Mahony, in Moser-Puangsuwan/Weber 2000: 148.

  • (12) : All recovered.

  • (13) : Mahony in Moser-Puangsuwan/Weber 2000:144,145

  • (14) : Mahony/Eguren 1996: 235.

  • (15) : MPT ended its team-sending project to Chiapas in 2000.

  • (16) : Heid, John, 4/01 interview with author.

  • (17) : Schirch 1995: 91.

  • (18) : E-mail from the field, 8/1999.

  • (19) : Müller/Büttner 1998: 5.

  • (20) : Müller/Büttner 1999: 5.

  • (21) : Müller/Büttner 1999: 2.

  • (22) : Müller/Büttner 1999: 7.

  • (23) : Moser-Puangsuwan/Weber 2000: 271.

  • (24) : Schweitzer, Christine in Moser-Puangsuwan/Weber 2000: 272.

  • (25) : Galtung, J., ”Participants in peacekeeping forces” cited in Schirch 1995: 47.

  • (26) : Bennett, 6/01 interview with author.

  • (27) : Poen, 5/01 interview with author.

  • (28) : Mahony/Eguren 1997: 225.

  • (29) : Poen, interview with author.

  • (30) : Müller/Büttner 1998: 23ff.

  • (31) : Ibid: 25.

  • (32) : Anderson, Mary B. 1999: 17.

  • (33) : Schirch 1995: 46.

  • (34) : Mahony/Eguren 1996: iii.

  • (35) : Mahony/Eguren 1997: 202.

  • (36) : Müller/Büttner 1998: 26.

  • (37) : Mahony/Eguren 1997: 46.

  • (38) : Schirch 1995: 89.

  • (39) : Schirch 1995: 89.