On 23 May 1965, the immensely popular President Sukarno addressed 120,000 enthusiastic communists cramming the national stadium.

The 45th anniver­sary celebrations for the Communist Party of Indonesia (PKI) looked like a national holiday. Five months later, all those people were being hunted down in a nationwide campaign to exterminate the PKI.

Within Indonesia, the ‘crushing' of the PKI has since then always been a matter of public celebration. But that this involved murdering so many unarmed civilians remains a taboo. A militarised New Order regime made the crushing a central part of its legitimacy. Contrary to the expectations of human rights activists, however, democracy since 1998 has not brought about a change of narrative. The tenacity of denial suggests that statist explanations do not do justice to reality. As one experienced student of the episode has put it, the central dilemma facing scholars remains what it has been for many years: determining ‘the relative importance of army initiative and local tension'.^[763]

On the one hand, growing evidence indicates that nearly all the killings were coordinated by the military. Earlier accounts that attributed a large proportion of deaths to uncontrolled violence by outraged religious militias and other independent civil groups are increasingly discredited.

Within the technical constraints of a Third World state, the soldiers did an effective job of (a) disposing oflarge numbers of communist sympathisers in a short time; (b) ensuring they (the soldiers) would not be blamed for it afterwards; and (c) fixing an impression of unspecified terror permanently in the public mind. Radio broadcasts immediately created a warlike atmosphere that justified extreme violence; local officers forced political party represen­tatives to sign off on death lists. Afterwards, military teams toured rural areas to survey political affiliation and run indoctrination sessions. In the longer term, military censors checked that news media and school textbooks rein­forced the effect of terror created by the half-secret killing campaign. Even today, military-backed militias remain available to intimidate human rights groups intent on honouring or exhuming victims. The garrison system that made this possible - heavy on intelligence capabilities - remains in place, though the system of seconding (retired) military officers to key senior government positions is mostly gone.

On the other hand, there is no doubt that social involvement in the killings was substantial. Civilians enthusiastically took part in killing people for local reasons - prestige, envy, payback, even fun. This adds complexity and even ‘multi-causality' to the story.^[765] Civilians provided the death lists, and the labour to do much of the dirty work with them in hand.

This chapter is divided into four parts. The first reviews three theoretical approaches that have inspired explanations for the 1965 killings, evaluating them in terms of their success at dealing with the state and society question. The next two sections apply the best of them, a contentious politics approach, to the problem at hand. One of these, the second, sketches the context before a marked turning point in 1963; the third reconstructs an arc of escalating tension, climax and de-escalation over the subsequent three to four crucial years. A fourth section ends the chapter with observations on the nature of state-society relations that produced mass violence in Indonesia in this period.