Previous PageTable Of ContentsNext Page


On January 19, 1999, as Muslims around the world were celebrating the end of the fasting month, a fight broke out on the island of Ambon, in Maluku (Molucca) province, Indonesia, between a Christian public transport driver and a Muslim youth. Such fights were commonplace, but this one escalated into a virtual war between Christians and Muslims that is continuing as this report goes to press. Much of the central part of the city of Ambon, the capital of Maluku province, and many neighborhoods (kampung) in other parts of Ambon island and the neighboring islands of Ceram, Saparua, Manipa, Haruku, and Sanana have been burned to the ground. Some 30,000 people have been displaced by the conflict, although the figure is constantly shifting.

The death toll by early March was well over 160 and rising rapidly as army reinforcements, brought in to restore order, began firing on rioters armed with sharp weapons and homemade bombs. The head of the Christian documentation center in Ambon told Human Rights Watch by telephone on March 10 that eighty-three Christians had been killed between January 19 and March 9, 1999, twenty-three of them at the hands of the military. Nur Wenno, head of Muslim relief efforts at Ambon’s largest mosque, said there were no precise figures on the Muslim death toll, but it was over one hundred.

Questions as to who was accountable for the violence in Ambon and surrounding islands focused on three issues: Who started it? Why did it escalate so fast? What, if anything, could the government have done to halt it? And what should the government be doing now?

The Indonesian press, senior Indonesian officials and opposition leaders, and many Jakarta-based diplomats believe the violence was provoked as part of a nationwide strategy of rogue military officers linked to the Soeharto family to disrupt the forthcoming parliamentary elections in June and create the conditions for a return to military rule. The June elections, which promise to be the freest Indonesia has had since 1955, would, if fairly conducted, almost certainly lead to a further diminution of the military’s power, which has been on the wane since President Soeharto resigned in May 1998. Local leaders in Ambon tended to see the violence as locally instigated for narrow communal goals. In either case, the government of Soeharto’s successor, Habibie, seems to have been half-hearted about investigating allegations of provocation at either the national or local level.

Why did the violence spread so quickly? Ambon was portrayed in the Indonesian media as a land where relations between Christians and Muslims had always been harmonious, the tranquility of interfaith relations protected by an alliance system called pela, where for centuries, a village of one faith had been twinned with a village of the other, where Christians helped build mosques, and Muslims helped build churches. The reality was very different. Tension between the two communities, Ambonese Christians on the one hand, and Ambonese Muslims and Muslims from various migrant groups on the other, was so high that it would have taken very little provocation to ignite an explosion. Once the violence began, it quickly fed on itself, dragging out historical grievances, creating new injuries, and generating new, deeply felt communal suspicions.

What might the government have done differently? A key question revolves around the use of lethal force. The conflict in Ambon separates into two distinct phases, demarcated by a decision to fire on demonstrators. From January 19 to about February 14, most of the deaths on both sides were caused by traditional or homemade weapons — machetes, long knives, spears, arrows shot from slingshots, molotov cocktails, and fishing bombs (illegal devices exploded under water to capture large quantities of fish). Many people also burned to death when houses or vehicles were set on fire. From February 14 onwards, most of the deaths took place when security forces, whose numbers by March had risen to 5,000 on an island with a population of about 350,000, began implementing shoot-on-sight orders. There is no question that an extremely grave security threat existed, and the security forces were initially accused by both sides of standing by and doing nothing as the different sides were attacking each other. When they finally did intervene, they shot lead bullets rather than attempting to use any methods of non-lethal crowd control.

A second question relates to the composition of the security forces used. Both sides have made allegations of bias, with the Muslims tending to accuse the police of favoring the Christians, and the Christians tending to accuse the army of siding with the Muslims. The accusations of bias were based in part on non-military attributes of the soldiers and police involved (geographic origin, religion, ethnicity) but also on their behavior in the field. Muslims accused Christian police in one case of opening fire near a mosque; Christians accused Muslim soldiers in another of helping Muslims attack a Christian village. Those accusations need to be thoroughly examined by an impartial body. Moreover, the government should deploy security forces with a view toward minimizing perceptions of bias, a point we elaborate on below.

The first two questions are directly related to the protection of human rights in a situation of civil strife. There is a third question of the government’s response, however, which has arisen in other outbreaks of communal violence, such as a serious ethnic conflict that erupted in West Kalimantan in late 1996 and early 1997. This is the Indonesian government’s belief in top-down conflict resolution: that if the local government brings religious or customary leaders together and has them sign a peace pact or participate in a traditional ceremony, the conflict can be solved. This approach can have unfortunate consequences, because when the pact inevitably breaks down, the participants often believe that the bad faith of one of the parties must have been responsible — and mutual distrust and suspicion grow deeper.

The conflict in Ambon has also displaced tens of thousands of people from their homes. Human Rights Watch is concerned that the provincial government, which has done the best it could to provide emergency shelter for the displaced, may be moving too fast to decide on transmigration — that is, resettlement of the displaced on other islands — as the optimal long-term solution to the problem. We are also concerned about the provincial government’s unwillingness as of early March to allow international humanitarian agencies in to help with the distribution of assistance to the displaced and other victims of the unrest.

We examine all of these issues in the report, based on a fact-finding trip to Ambon in February 1999, documentary material collected and interviews conducted during that visit, and subsequent communications with Christian and Muslim leaders in Ambon.1 Based on that material, we make the following recommendations to the Indonesian government:

1. Ensure that its security forces respect the Basic Principles on the Use of Force and Firearms by Law Enforcement Officers and that troops assigned to Ambon are fully equipped with non-lethal methods of crowd control. Of particular importance to Ambon is the principle that “Law enforcement officials, in carrying out their duty, shall, as far as possible, apply non-violent means before resorting to the use of force and firearms. They may use force and firearms only if other means remain ineffective or without any promise of achieving the intended result.”

2. Investigate accusations of bias in the behavior of security forces. In many outbreaks of violence thus far, notably the shooting on March 1 of four people outside a mosque and the conflict on the island of Haruku on February 14, Muslims have accused Christian Ambonese police personnel of taking part in attacks against them. Likewise, the Christians have accused troop reinforcements sent from the Wirabuana command of the Indonesian army, based in Ujung Pandang, Sulawesi, of siding with Muslim villagers in several clashes. The fact that the Wirabuana command is led by a Muslim Ambonese, and that some of the Muslims involved in the conflict in Ambon are ethnic Bugis and native to the area around Ujung Pandang, does little to allay suspicions of bias. An impartial independentinvestigation, not necessarily of every outbreak of violence that has taken place, but at least of three or four major clashes where bias has been alleged, would be useful.

At the same time, independent investigators should work with local village heads (raja) and community leaders to examine reports that have been produced by both sides to identify points of convergence and divergence and use these findings to understand how perceptions have fed the conflict.

3. Avoid at all costs the imposition of a state of "civil emergency" in Ambon and surrounding islands. This option is currently being weighed by Cabinet ministers in Jakarta and has been recommended by some local leaders in Ambon. With the very clear exacerbation of the situation caused by the presence of security forces with shoot-on-site orders, additional measures that allow the military to bypass normal civil rights safeguards are likely to make things even worse.

4. Make absolutely clear in all public pronouncements and interviews that both Christians and Muslims have suffered terrible losses. There has been a distressing tendency in both the Indonesian and international media to quote sources from only one side of the conflict. That reporting feeds back into the communal tensions in Ambon, helping fuel one side's anger against the other.

5. Find and prosecute any provocateurs. If General Wiranto and other senior government leaders have enough information to acknowledge, as they have, that provocateurs played a role in the initial outbreak of violence, they have an obligation to make public the nature of their evidence and make every effort to ensure that those individuals are found and prosecuted to the fullest extent of the law.

6. Undertake a thorough study on the underlying political, economic, and demographic causes of tension and prepare recommendations on how to address them that can be discussed and debated in Ambon.

7. Ensure that international, non-religious-based humanitarian organizations are allowed full access to Ambon and surrounding islands to assist the wounded and displaced. The need is not so much for supplies of food and medicine but to find a way to distribute existing supplies safely and impartially.

8. Ensure that the rights of internally displaced people in Ambon are fully protected in accordance with “Guiding Principles on Internal Displacement” prepared by the Office for Coordination of Humanitarian Affairs of the United Nations.


Ambon is the name of a city and an island, and the term “Ambonese” describes a cultural area that embraces many of the islands in the district of Central Maluku, Maluku province, Indonesia. Those islands include Ambon, Saparua, Haruku, Buru, Manipa, Nusalaut, and Ceram. The population of indigenous Ambonese since the sixteenth century has been relatively evenly divided between Christians and Muslims.2 For the most part, the two live in separate negerior villages, and even within mixed villages, they tend to live in separate kampungs or neighborhoods. In addition, many ethnic Butonese migrants from southeast Sulawesi, a large island to the west of Ambon, and ethnic Bugis and Makassarese, from south Sulawesi, have settled in their own kampungs. These migrants are overwhelmingly Muslim, and they dominate small-scale retail trading and transportation networks.

Tension between Muslims and Christians in Maluku province had been growing for decades, the result of the declining influence of traditional authority mechanisms; the influx of migrants; and the “greening” or perceived Islamicization of the central government in Jakarta. The outbreaks of communal violence elsewhere in Indonesia in the aftermath of President Soeharto’s resignation in May 1998 served to heighten distrust between the two communities.

Both the pela alliance system and the authority of traditional local leaders, called raja, had been undermined long before the current conflict erupted. The pela system had received a fatal blow at the time of Indonesian independence in 1949, when a largely Christian political elite, many with military or administrative ties to the Dutch colonial administration, opted to establish the Republic of the South Moluccas (RMS), rather than join the new Indonesian state. A brief war ensued which the RMS lost in December 1950. In the course of the conflict, many Muslim villages were razed by RMS forces, and the destruction was not forgotten. In addition to the war, a steady influx of migrants from other parts of Indonesia resulted in the establishment of new settlements that were completely outside the pela system, which applied only to Ambonese Christians and Muslims.

In 1974, with the passage of a new law on local government, local leadership was gradually transformed from a clan-based system, represented by the Ambonese raja, to a territorially based system of village heads, the lowest rung on the Indonesian administrative ladder. In one sense, the new system was more egalitarian, because it opened up the possibility that migrant communities of ethnic Bugis, Butonese, and Makassarese could be represented, and some candidates for village head appealed to these communities for votes. On the other hand, it meant that many of the village heads lacked the authority the old raja had enjoyed, and when conflict broke out, there were fewer people at a local level with the ability to stop it.

The migrant influx also tipped the demographic balance in favor of Muslims. Migrants from Sulawesi had been coming to trade in Ambon since the sixteenth century, but migration picked up sharply in the 1970s, and with it, increasing tension with the Ambonese population.3 Ethnic Bugis, who had traditionally settled along the coast in self-contained communities, began to settle in the city of Ambon, displacing other traders, taking over the transportation sector, and in the view of some Ambonese, creating slum areas and contributing to urban crime.4 Bugis also began to make themselves felt politically in the 1980s and 90s, with tightly organized Bugis associations that local politicians ignored at their peril. Their political rise coincided with what Ambonese Christians saw as an affirmative action policy undertaken by the national government in the early 1990s to redress the marginalization of Muslim entrepreneurs in comparison to their ethnic Chinese competitors. Whatever the rationale for this policy in Muslim majority areas, in Ambon it created anger and frustration among Christians, as they saw not just economic opportunities but also civil service jobs going more and more to Muslims, many of them migrants. As Christians were eased out of the positions they had traditionally held in the local government, teaching profession, and police, theyturned to the private sector, only to find that migrant groups from Sulawesi, among others, had sewn up the market. Christians began to feel that their political, economic, and cultural existence in Ambon was threatened.5

Communal relations, then, were not good, even before the violence erupted, and everyone we talked to in Ambon spoke of regularly recurring fights between Muslim and Christian kampungs. The neighborhoods seemed to live in a state of barely repressed hostility, but the frequent fights were quickly settled.

The atmosphere, however, changed perceptibly for the worse after a series of possibly provoked communal incidents broke out elsewhere in Indonesia in late 1998. On November 22, 1998, a dispute between local gangs over a gambling establishment, at which Christian Ambonese acted as security guards, turned into a communal riot as rumors spread that the Ambonese had destroyed a local mosque, and Muslims youths trucked into the area then burned some two dozen churches. On November 30, a Christian youth congress in Kupang, West Timor, held a congress, followed by a march, to protest the church-burnings. In the middle of the march, a truckload of youths appeared whom no one seemed to know, and in no time an ethnic Bugis neighborhood, including the mosque, was burned to the ground.

Both incidents were widely believed to have been provoked by the military, because the army in particular was perceived to be the beneficiary of civil unrest: a traumatized population might see the army, rather than a democratically elected government as Indonesia might have next June, as the only guarantor of security. The local government warned religious leaders around the country, as Christmas and the Muslim fasting month approached in December 1998, to be on alert for provocation and to resist being influenced by rumors.

One meeting of religious leaders was held in Ambon in mid-December. The atmosphere was so tense, according to one participant, that the Muslims left convinced that the Christians had decided that the only way to address the problem was to rid the province of Muslim migrants. Leaders of both communities set up “posko,” an acronym defined either as “communication post” or “command post” depending on the militancy of the definer. These posts, with networks of mosques and churches connected by cell phone or regular telephone, were intended to alert the respective communities to any danger of provocation. In fact, once a fight broke out, they served as much to spread rumors and mobilize communities.

In such an atmosphere, it did not take much to cause a conflagration, although precisely how it started is still a matter of speculation.

1 We would like to acknowledge the invaluable assistance provided by Binny Buchori of the Indonesian nongovernmental coalition, INFID, based in Jakarta, who facilitated our visit and conducted some of the research. Any errors or misinterpretations in this report, however, are solely the responsibility of Human Rights Watch. 2 The commonly held misperception that Ambon is predominantly Christian, and predominantly Protestant at that, may be rooted in the fact that Christians traditionally dominated the civil service, including the teaching profession, and the police,and the Moluccans who fled to the Netherlands following the failure of the RMS movement were more than 90 percent Christian. 3 On the history of migration into Ambon, see Gareth J. Knoop, “A City of Migrants: Kola Ambon at the End of the Seventeenth Century,”Indonesia (Ithaca, NY), No.51, April 1991, pp. 105-128. 4 M.J. Papilaja, "Apa, Mengapa, & Bagaimana Kerusuhan Ambon: Sebuah Kajian Empirik," February 1999 (e-mail communication from Ambon, received March 1999). 5 Ibid.

Previous PageTable Of ContentsNext Page