Generating distrust through intelligence work: Psychological terror and the Dutch security services in Indonesia,1945–1949

POWs in an undeclared war

From the perspective of the Dutch authorities the decolonization war in Indonesia in many ways resembled a run-of-the-mill colonial rebellion, the kind of which they had seen and quashed before. ⁶ But as for their imperial counterparts, Britain and France, in their own decolonization conflicts, the interlude of the Second World War – marked by the Dutch imperial military defeat and the Japanese occupation of Indonesia – had also brought profound changes. Whereas imperial dissolution and Indonesian independence had once seemed an option only for a far-away future, it now suddenly became thinkable. ⁷ The official purpose of the Dutch war in Indonesia, therefore, quickly became not the restoration of the pre-war status quo, but rather the restoration of order in anticipation of a regulated Dutch-led transition to an autonomous Indonesian state within a Dutch-Indonesian union still dominated by the Netherlands. The ultimately misguided, but surprisingly long-lived belief among higher Dutch authorities, both in the metropole and in the colony, was that the rebellion was only supported by a small minority of ‘extremists’ with little popular support, while the great majority of Indonesians were fundamentally well-disposed to ongoing Dutch stewardship. ⁸

In a similar way that the British fought their decolonization wars under the label of ‘Emergencies’ and the French referred to the war in Algeria as an ‘operation for the maintenance of order’, ⁹ the Dutch authorities never officially acknowledged the four-year conflict over the decolonization of Indonesia as a war. From their point of view, the conflict remained an internal affair, as is most conspicuously clear from the designation of the two large military campaigns (in July 1947 and December 1948) as ‘Police Actions’. The stated purpose of the military mission was not re-occupation, but the restoration of ‘peace and order’. But what a focus on the two ‘Police Actions’ obfuscates is that in the time between and after these campaigns, a more insidious and often vicious conflict of insurgency and counterinsurgency developed, claiming large numbers of casualties on both sides. ¹⁰ Even though Dutch military authorities professed only to be fighting a small group of insurgents, the impact of their campaigns was much wider. As part of their effort to get to grips with an elusive insurgency, Dutch forces arrested large numbers of Indonesians – during their daily patrols, their larger-scale sweeps, and especially during the two Police Actions – in the hope of isolating the minority of ‘extremists’ from the silent and supposedly loyal majority. ¹¹

The practice of arrests and internments would never assume the scale nor the systematic nature of attempts at population control prevalent in other colonial counterinsurgency campaigns, such as the French campaign in Algeria or the British ‘new villages’ policy in Malaya. An important difference is that the Dutch authorities in Indonesia never instigated large-scale internment or forced displacement of the civilian population. At least in theory all prisoners were enemy combatants or a more or less legitimate ‘threat to peace and order’. ¹² Nevertheless, the number of arrestees was considerable, with the so-called ‘de facto prisoners of war’ reaching a maximum of approximately 40,000 during the Second Police Action and the number of ‘political internees’ mostly fluctuating somewhere between 10,000 and 20,000. ¹³ Moreover, precisely because Indonesia did not witness the establishment of internment camps or walled-off ‘new villages’ on the scale that we know from Algeria or Malaysia, turn-over of prisoners was constant. Several thousand remained imprisoned for a long time, but many were released or deported to Indonesian-controlled territory after a while, only to be replaced by new prisoners. The number of people who passed through one or more of the prisons and camps was therefore much larger, although a total figure is impossible to establish. Many were only arrested for questioning, not for long-term imprisonment. Records of these arrestees never found their way into the aggregates that the central authorities drafted at irregular intervals. ¹⁴

Prisoners who were kept for longer and made their way into the central system of camps and prisons were classified into three categories: ‘de facto prisoners of war’, ‘political internees’, and ordinary ‘criminals’. ¹⁵ Among the ordinary ‘criminals’, the authorities classified anyone who could be charged with a clear criminal act – murder, plunder, sabotage – for which sufficient evidence was available to try and convict them. Even if these acts had been committed as part of the Indonesian guerrilla war effort, they were tried as if the pre-war circumstances still applied. ¹⁶ In other words, the other two categories, of ‘de facto prisoners of war’ and ‘political internees’ were generally reserved for those who the authorities wanted to intern but could not convict for lack of evidence. ¹⁷

Because the authorities did not refer to the Dutch-Indonesian conflict as a war, it refused to recognize enemy combatants as ‘prisoners of war’. But in order to placate international opinion, they nevertheless treated any prisoners captured in battle, ‘as much as possible’, in accordance with the 1929 Geneva Convention. Among these so-called ‘de facto prisoners of war’, the Dutch authorities subsumed ‘all armed Indonesians captured at the front and during military actions, also if the evidence cannot be delivered that they are members of one of the struggle organizations’. ¹⁸ The ‘political internees’ comprised all other Indonesians who were not clearly a part of one of the military organizations, but who were arrested either at the front or in the presumably ‘pacified’ backcountry for ‘subversive activities’ or other actions that were deemed a risk to ‘public order and peace’. The rights and protections of these ‘political internees’ were much less closely upheld than those of the POWs, and forced labour, insufficient or inferior food and severely overcrowded prisons were common. They were normally interned either under article 20 of the State of War and Occupation – ‘temporary’ internment as a threat to peace and order – or under article 169 of the old colonial Criminal Law Code, which criminalized membership of illegal clubs and associations. ¹⁹

At several points in time, the authorities attempted to reduce the number of prisoners by releasing those no longer deemed a threat to ‘peace and order’ into the hands of their family, or by expelling them to Indonesian-controlled territory. In all cases, it was the civilian authorities (especially the public prosecutors) who pleaded with their military counterparts to begin or accelerate the separation of the genuinely ‘dangerous’ internees from those who, with the right guidance and possibly through some form of social rehabilitation, could be released back into society. But the military authorities were much less eager to free detainees. In case of doubt, they generally seem to have erred on the side of caution and opted for continued imprisonment. ²⁰ Nevertheless, many prisoners were eventually released after thorough questioning, although this might take well over a year as the officials charged with the selection process were severely overworked and plagued by a chronic lack of evidence – many prisoners had been sent from the field with little more than their name and, if they were lucky, a note about the circumstances of their arrest. ²¹ The problem was that at every layer of the imprisonment system, when confronted with only a bare minimum of documentation about particular prisoners, the tendency was to send such individuals on to the next stage rather than take the responsibility for releasing them. Central to the ultimate fate of a prisoner, therefore, was often not their processing by the system, but rather the initial decision to arrest them.

‘Reason for arrest: suspicious attitude’

The confusion and disagreement at the higher levels of the military and civilian government about what to do with the various groups of prisoners, and where to place them, is indicative of a wider tenet of the Dutch counterinsurgency effort in the Indonesian war of decolonization: the extent to which it evolved ad hoc and was determined by local or regional developments over which the central authorities in The Hague or in Batavia (Jakarta) had only tenuous control, chasing to keep up with or retro-actively sanction new developments. ²² Although the regulation defining the different classes of prisoners and internees also represents an attempt (post hoc) to set some guidelines about who should be arrested – enemy combatants, ‘subversive’ elements, common criminals – the fact of the matter is that these guidelines were so abstract that the decision really was up to the assessment of lower-level actors in the field. The result was a chaotic and highly diverse practice, depending on regional factors and the developing intensity of the conflict over the years, the culture and tradition of different army and police units, or the idiosyncratic decisions of individuals.

One way of bringing some order to this chaos and getting a notion of the reasons why certain people were arrested, and why others were not, is to look at the interrogation reports from the various intelligence and security services. Arrests were frequently made by or on behalf of the intelligence services, or if the arrests were made by others, the local intelligence units were normally the first to receive and interrogate the prisoners before either releasing them or relaying them to higher authorities. Their records and conclusions were normally crucial to the further processing of prisoners, determining whether or not to keep or to release the prisoner.

The surviving records of such interrogations are, it appears, highly incomplete. A complicating factor is the fragmented nature of the Dutch intelligence services. ²³ Apart from the traditional (political) police force and its investigation service, ²⁴ for most of the conflict there were two competing military intelligence services. On the one hand there was the Netherlands Forces Intelligence Service (NEFIS), which was responsible for general intelligence gathering of both military and political nature. NEFIS was not intended to have executive authority, which meant that in theory – though not always in practice ²⁵ – it had to request local police or military forces to execute the arrests that it desired. On the other hand, the so-called Intelligence and Security Groups (Inlichtingen en Veiligheidsgroepen, IVG) were integrated into the regular army structure, directing the work of designated but only minimally trained officers embedded within army units in the field, and were responsible for collecting regional and tactical military intelligence. These groups were much more autonomous, with the express authority to conduct arrests and to direct patrols. It was only in September 1948 that the NEFIS and the IVG were merged into one organization, the Central Military Intelligence Service (CMI). This restructuring does not seem to have solved the institutional distance and the difficulty of communication between the central intelligence agency and the field operation of local intelligence groups. Finally, both the navy and the marine corps retained their own intelligence services (MARID and VDMB, respectively). ²⁶

The various security services typically relied on a top layer of ‘white’ Dutch officers who drafted reports and were assisted on the everyday level by personnel recruited in the Indies among the indigenous or Eurasian population, with better linguistic and cultural fluency. Van Doorn and Hendrix in their analysis of Dutch military violence in Indonesia characterized the interrogation room in segregated terms: a white officer sat behind a desk making notes, while a native ‘aid’ was responsible for meting out the ‘harsh’ techniques used to encourage the prisoner to talk. ²⁷ These lower-tier intelligence workers were mostly delegated to the intelligence services from the KNIL (the colonial army). They consisted both of Eurasians and indigenous soldiers recruited from ethnic groups considered particularly loyal to the colonial regime, especially the Ambonese and the Menadonese. ²⁸ The security services further relied on a range of more or less official ‘translators’ as well as ‘informants’ to provide information and to perform some of the dirtier work of interrogation. These informants were sometimes recruited among loyal, or socially excluded and endangered groups, but were often also enlisted with the promise of compensation (money or food). One consequence of this use of intermediaries to carry out questioning and interrogations was that their Dutch superiors usually had to rely on second-hand interpretation. Another consequence, for the Indonesian informants, was that they exposed themselves (and their families) to retaliation from Indonesian forces. ²⁹

The prevalence of Eurasian and indigenous recruits among the ranks responsible for the ‘dirty’ side of the interrogation work reflects a broader tendency to blame the KNIL (the colonial army) for most of the more brutal violence during the decolonization war, rather than the KL (the Royal Dutch Army consisting of volunteer and conscript recruits from the Netherlands). ³⁰ Not only were the KNIL soldiers (supposedly) more hardened and desensitized than their KL counterparts after experiencing the Japanese occupation and the early period of revolutionary violence by Indonesians against anyone suspected of sympathizing with the colonizer. They were also trained in the more ruthless traditions of colonial military warfare and had much more to lose than their Dutch counterparts, as their very homeland was at stake. Not surprisingly, therefore, there is indeed reliable evidence that KNIL personnel were more likely to commit atrocities. ³¹ Nevertheless, we should also be aware that part of the reason for this overrepresentation is in fact the tendency to delegate the ‘dirty’ work to KNIL units – hence also their overrepresentation in the lower ranks of the intelligence units.

The archival documentation of interrogation reports is as fragmented as the intelligence services themselves were. Much can be found in the archives of NEFIS/CMI, where the reports drafted by the various branches throughout Indonesia were collected. However, for reasons that are unclear, the reports of arrest and interrogation, though extensive, do seem to be incompletely handed down. ³² Also for unclear reasons, these collections rarely continue for the period after the reorganization in which the NEFIS was subsumed into the CMI. More importantly, in the case of the IVG, the majority of interrogation reports have not survived. What we can find in the archives of the Dutch forces in the Netherlands Indies are daily and weekly reports, ³³ but these generally contain only the bare information that the officers had collected, which often included the numbers of casualties as well as prisoners. They reveal only little about the reasons and grounds for arrest, and even less about the subject of individual interrogations. The more extensive notes that we saw from Sergeant Sytsema, the intelligence officer from the introduction to this paper, are an exception: they only made their way into the archive because they were carried home by a soldier from his battalion and were handed over to the military archives in 2001. ³⁴ Hence, they offer a rare insight into field practices of the IVG, of which documentary evidence has otherwise largely eluded us.

That said, the material that we do have is extensive enough to reveal something about the practices on the ground. For one thing, crucially, the intelligence officers often worked with only the thinnest of information, filling in the gaps with guesswork and preconceived notions. ³⁵ In their daily and weekly reports, like any intelligence organization, they presented an image of a meticulously thorough, all-knowing, and impartial or disinterested force. But if we delve into the material on which they based these reports, another story emerges.

The example of Sergeant Sytsema, for instance, shows that even a field intelligence officer involved in the daily routine of running patrols frequently had only a bare minimum of information about the prisoners he interrogated. On one day in June 1948, he reported to his superior that a patrol had brought in 20 prisoners:

All were delivered to me without any further explanation, making a fruitful investigation very difficult. […] I prefer that no prisoners are brought in of whom nothing is known. These people have been picked up at random, without any reason. ³⁶

After thorough interrogations, Sytsema freed most of these prisoners within two days, but regarding one of the prisoners he was less sure. He sent this man on to his superior, noting: ‘I cannot get much more out of him. You will probably know better what to do with him.’ ³⁷ We can only guess where this prisoner ended up or if he was ultimately released. Nevertheless, in the majority of cases Sytsema did in fact release those prisoners who were brought to him without any evidence or information, though not until he had submitted them to a painstaking interrogation. ³⁸ Many of his colleagues in the field, by contrast, were less scrupulous even in these decisions, as we can gather from the recurring complaints by the officials higher up in the prison system that they were overwhelmed by large numbers of prisoners without accompanying dossiers, who they did not know what to do with. ³⁹

The scarcity of evidence with which local intelligence officers were working is also reflected in their interrogation reports. NEFIS personnel and other intelligence officers in the local branches around the archipelago were supposed to record their interrogations according to a more or less standardized format. ⁴⁰ The reports opened with some basic information about the subject of the investigation, which, it seems, was normally drawn from the interrogation itself rather than from previous research. This information consisted of the subject’s name, his ⁴¹ ethnicity, date and place of birth, the location of his arrest, his occupation, who had arrested him, and an assessment of the reliability of his testimony. Sometimes it also included an evaluation of the subject’s ‘intelligence’, and it always closed with the ‘reason for arrest’. What followed was a short ‘chronology’ of the subject’s life and actions – his upbringing and education, his activities during the Japanese occupation (1942–1945), his undertakings since August 1945, and finally the circumstances of his arrest. More often than not, this ‘chronology’ was very concise, adding only marginally to the list of essentials. Finally, the rest of the report was made up of information that the subject provided about the economic, military, and political situation in the areas where he had resided.

What interests us here is the information provided to explain the ‘reason for arrest’ of the suspect. Sometimes, the intelligence officers merely entered ‘suspect’ or ‘suspicious attitude’. One Salikoen, for example, who was arrested by a patrol near Surabaya in April 1946 and delivered to the security service of the marine corps, was registered as ‘suspicious’, but the officer interrogating him did not expand on this or offer further material reasons. It was merely noted that Salikoen in his interrogation ‘made a good impression, is sturdy and willing to work with us’. The officer recommended to ‘keep [him] for further questioning’. ⁴² In another case, the same officer interrogated a man called Hidin for his ‘suspicious attitude’. Hidin had worked on a plantation on East Java since even before the Japanese occupation. The plantation had now been taken over by Indonesians, and Hidin had continued to work there. Other than that brief account, his ‘suspicious attitude’ was not elaborated on in the report. ⁴³

At other times, the interrogators were a little more precise when providing the ‘reason for arrest’. Recurring motivations were: ‘suspect of murder’, ‘rampok’ (plunder/robbery), ‘theft’, ‘membership of an extremist organization’, ‘member of a struggle organization’, or ‘espionage’. ⁴⁴ Whereas suspicions of murder or theft were often related to a specific crime, for which the interrogators had outside witness reports, ⁴⁵ the more common suspicions of membership of a military group or an extremist organization were much more abstract and expansive. Any anomaly or even coincidence could lead to a ‘suspicion’ arising in the eyes of the patrols or intelligence officers. Possession of weapons was of course a major incriminating factor, which, as we saw, was in itself considered sufficient proof of membership of a ‘struggle organization’. But the definition of ‘weapons’ was pliable. One man, for example, was arrested because a patrol found a rusty old sword from the Royal Netherlands Indies Army (KNIL) in his house. ⁴⁶ Another common reason for suspicion was simply to be found in the wrong place at the wrong time. One 12-year-old peasant boy was arrested during a raid on Kejawan (Surabaya, East Java) on suspicion of ‘conspiring with the PRI [Pemuda Republik Indonesia = Youth of the Indonesian Republic]’. He was unarmed, and was released several days later, but only after he had agreed (presumably under pressure) to work as a guide for the local military patrols. ⁴⁷ Another man, called Ali, was arrested near the village of Buncitan (near Sidoarjo, East Java) because a company of the marine corps had been fired at from that particular village. Ali lived and worked on a fish farm just outside Buncitan. He claimed that he had in fact heard the shots but had thought ‘that the Dutch were at it’, and believing that it would not concern him he had continued working his fishing nets. He had become suspect to the Dutch patrol simply because he had not decided to flee – either when the shots were fired or when the Dutch patrol approached. ⁴⁸

‘Espionage’, another word for mobility

The final cause for arrest that was commonly cited, suspicion of espionage (often in conjunction with suspected membership of an illegal organization), is particularly revealing about the Dutch fears and prejudices. Spies and infiltrators are ubiquitous in the interrogation reports of the intelligence officers throughout Indonesia. Generally, the main reason for suspicion of espionage to arise was mobility. The mere fact of moving from one place to another made one a suspect, especially if it meant moving across the demarcation line between Dutch- and Indonesian-controlled territories. Regional or even interregional mobility made one an anomaly in the eyes of the Dutch intelligence services. It did not fit with their notion of presumably normal Indonesians’ lifestyles and relations. But if we look more closely at some of the cases in which the Dutch suspected espionage, we can see that these (classically colonial) preconceptions were problematic, and that they skewed the way the intelligence services tried to distinguish enemy combatants from the mass of the civilian population.

The problem for the intelligence services was, of course, that the ‘real’ spies to Dutch eyes looked deceptively like common traders, or people on their way to visit family members. In fact, Indonesian authorities and irregular forces made active use of this confusion, disguising their spies and informers in Dutch-controlled territory as common peasants, ⁴⁹ or simply recruiting established traders in exchange for some additional income. ⁵⁰ But as a consequence, the security forces jumped on anyone who travelled outside of what the Dutch thought was their respective hometown. One Hasnah, for example, was arrested in May 1948 as a spy for the Indonesian Republic. She had recently returned to her native Padang (West Sumatra) after an absence of just over a year. In fact, as she tried to explain to her interrogator, she had only been away because she had been abducted by Indonesian irregular forces, who from their side had suspected her of being an informer for the Dutch – because she had previously been temporarily interned on suspicion of possession of weapons, and was thus seen to have communicated with the Dutch. After having cleared her name with the Indonesians, in early 1948, she had seized the opportunity presented by a temporary ceasefire to return to Padang, only to fall under Dutch suspicion once again. ⁵¹

The suspicion towards travellers seems to originate from a belief that most common Indonesians in the countryside were sedentary, connected to their homeland and not inclined to leave their villages unless there was a good reason. For one thing, this perception ignored the fact that especially in the densely populated farming areas of Java, a combination of farming and trading, or travelling for seasonal labour in one of the larger industrializing towns, had already become quite common in the late colonial period. ⁵² And what is more, the perils of a war that oscillated back and forth to many common Indonesians provided precisely such a good reason to move about. The story of a man named Rasman, a native of Kencong (East Java), is striking. He fled his home in September 1947 after Dutch forces had started shelling his town and he settled with several other refugees on an old rubber plantation called Tempursewu. In May 1948 the village head told the evacuees that they could now safely return, which Rasman did with 48 others originating from the same area. But as soon as he returned to Kencong he was arrested as a potential spy. ⁵³ Rasman’s fate seems not to have been an anomaly: refugees were more frequently targeted by the intelligence services. We can already read about this phenomenon in an internal report from the Indonesian army in Kediri (East Java) dated 4 June 1947. The local army unit warned Indonesian civilian authorities that many refugees had recently come under the false impression that it was safe to return home, only to be arrested, interrogated, and tortured upon arrival by Dutch soldiers looking for information about the situation behind enemy lines. ⁵⁴

People thus fled their home towns ahead of advancing Dutch forces, or they were forced out as a consequence of the destruction wrought by Dutch shelling. ⁵⁵ In some cases, this even happened several times as shifting fortunes saw Dutch and Indonesian forces alternating as occupiers. ⁵⁶ When such refugees tried to return, their movements were nevertheless suspect in the eyes of Dutch intelligence officers. In the words of the head of the IVG for the Semarang area (Central Java) in December 1946: ‘Bizarrely, some inhabitants of the Semarang kampungs [native urban neighbourhood], who fled the city during the riots of November ’45, are currently returning.’ He suspected these people of having joined the Indonesian forces and of now returning to conduct espionage, disguised as ordinary city dwellers. ⁵⁷ As this officer and others seem to have seen it, the Dutch forces were merely in Indonesia to fight the ‘extremists’ terrorizing the general population, meaning that anyone innocent should have no reason to flee. Hence, people like Rasman or the refugees from Semarang were flagged as ‘infiltrators’ rather than refugees.

In a sense, the tendency to target anyone crossing the demarcation line for arrest, interrogation and imprisonment, might be interpreted as an improvised, hapless and ultimately failed attempt at population control. As is evident from the case of the British war in Malaya, which is widely regarded as the model of a relatively successful (colonial) counterinsurgency, for such a campaign to be effective it needs to rely centrally on population control on a massive scale, through mass internment or the forced relocation of civilians. Although the ‘Malayan Emergency’ is famous for being the conflict in which the phrase ‘winning hearts and minds’ was introduced, the internments and forced relocations seem to have had more impact on the eventual success. ⁵⁸ Considering the size of its colony and its own limited resources, Dutch authorities in Indonesia were unable (and unwilling) to mount an operation of a comparable scale, but the actions and considerations of the intelligence groups in the field do betray a similar logic of population control. They were faced with the challenge of separating the proverbial ‘fish’ that was the guerrilla fighter from the ‘water’ that was the wider population. One way of doing so was to target anyone who behaved in an unexpected, supposedly anomalous way. Another way was attempting – however impossible it turned out to be in practice – to treat the demarcation line as an impermeable border also for the civilian population. This is one explanation for why the authorities, whenever the prisons and camps started to overflow, were somewhat more forthcoming in releasing prisoners as long as they could be expelled across the demarcation line. If the enemy combatants could not be separated from the population, they could at least be removed to the Indonesian-controlled territory, while the Dutch could get on with their task of establishing ‘peace and order’ in their own territory.

Psychological terror and the ‘common’ Indonesian

Equally important in trying to separate the enemy combatants from the larger population was what I described in the introduction to this paper, in analogy to Neil MacMaster’s analysis of the war of decolonization in Algeria, as the psychological terror that was exerted by the perceived ubiquity of the intelligence apparatus and its informers. Although the total number of prisoners and internees taken by the Dutch may have been comparatively low, many more people passed through the hands of the intelligence services only to be released after a relatively short time. But this short time could still compromise the arrestees in the eyes of their countrymen. As we saw in the case of the woman named Hasnah, and as can also be seen in a number of other cases in the interrogation files of the intelligence services, a short stint in a Dutch prison could lead to an arrest by Indonesian forces or guerrillas on suspicion of cooperation with the Dutch, subsequently leading to another arrest by Dutch authorities, and so on. ⁵⁹ Meanwhile, both Dutch and Indonesian information services warned the population of the ubiquity and perfidy of enemy spies and informants, thus stoking further fear among the population. This warning was supported by the actual Dutch practice of recruiting secret informants, but in the propaganda of the Republic grew much larger than the facts bore out.

The fear among Indonesian officials and military that former prisoners of the Dutch might have become informers was further exacerbated by the common practice among intelligence officers to actively solicit the cooperation of the people they were interrogating. They offered their prisoners the opportunity to become spies or informers, sometimes in exchange for their speedy release and/or monetary reward. ⁶⁰ Conversely, the Indonesian army and Indonesian irregular forces also tried to pressure potential defectors into choosing their side, by offering advantages for signing up, spreading propaganda about the alleged cruelty of the Dutch forces, as well as threatening anyone suspected of cooperation with the Dutch. ⁶¹ In short, both sides were trying to pressure an often reticent population into showing its colours. If one thing troubled them, it was a population that refused to pledge its allegiance – Indonesians because they needed the support of the wider population for their guerrilla effort, the Dutch because they desperately sought to distinguish enemy from civilian. The wider population often became caught in the middle of these two suspicious forces, while always expecting ‘informants’ of either side to be among their midst.

No wonder, then, that many ‘common’ Indonesians preferred to stay on the fence, weighing their options and biding their time. Of course, when examining the fear engendered by the intelligence services, we should also remember that field intelligence units are known to have been among the main perpetrators of brutal violence during the Dutch operations in Indonesia. As Van Doorn and Hendrix described in their seminal work Ontsporing van Geweld (Derailment of Violence) in 1970, and as has since been further confirmed by other historians, these units enjoyed considerable freedom to do as they saw fit. In many cases their methods included torture and intimidation, arson, and extra-judicial killings. ⁶² Indonesian propaganda eagerly included reports of such abuses in order to sway international public opinion, and likely also to dissuade the population from communicating with the Dutch intelligence services. ⁶³ Still, it is unknown exactly how widespread these practices were. There is a possibility that the focus by Van Doorn and Hendrix and other historians on intelligence units as central perpetrators of atrocities has created the impression that such abuses were more systematically and widely applied than was actually the case. But in any event, as I argue here, even the everyday practice of arrests and interrogations already constituted a form of psychological terror – even when carried out ‘by the book’.

Sergeant Sytsema describes another telling example of the reticence of the local population that he had to police. In early November 1948 he sent out a patrol to arrest a number of notorious ‘gang members’ who were rumoured to reside nearby (but who ultimately eluded him). When the patrol entered the village of Kawungadeng, the soldiers – who in this case must have been indigenous Indonesian recruits – conducted a little experiment: they greeted the population with the ‘merdeka [freedom] salute’, in an attempt to trick the villagers into thinking that they were dealing with an Indonesian ‘gang’. Sytsema reports:

The reaction was funny. People expressed their delight in the arrival of this ‘gang’ and were immediately willing to supply food, if desired. But as soon as they found out that they were dealing with a bunch of disguised soldiers, they changed their tune, but with the same willingness. Quite a sign that the population swims with all tides. ⁶⁴

Such impenetrable behaviour increasingly frustrated the Dutch, ⁶⁵ as we can see from the clear signs of exasperation coming from even the initially restrained Sytsema, leading to more forceful (although hardly more effective) behaviour. On 5 December 1948, for example, about ten months into his assignment as NCO of intelligence in Puraseda, Sytsema sent out a patrol that randomly brought in approximately 40 civilians from the surrounding area after a group of churchgoers had been fired at near his encampment. Two days later, on 7 December, he noticed that a call from his local commander for volunteers to cut down the shrubbery around the town had attracted very few of the villagers. But ‘after a short “tour” in the village’, as he reported to his superior, ‘the turnout was suddenly surprisingly large’. And finally, after another two days, Sytsema organized a propaganda meeting for the population and its dignitaries, which again saw a disappointingly low turnout of no more than 50 attendees, most of whom had been ‘accidentally picked up by the [passing] patrols’. Nevertheless, Sytsema’s commander, Major C.P. van Lede, went on to sternly address the crowd, giving them ‘an urgent warning, in case the population continues to refuse to collaborate’. ⁶⁶ The pages of Sytsema’s daily notes in this period ooze with almost continual frustration.

Reliable and readily available information was of paramount importance, and the intelligence officers felt responsible for providing the military with this precious resource. But the people who were the potential informers were not forthcoming. Hence, if and when people did not step forward voluntarily and if the carrot of remuneration did not work, recourse to the stick of psychological and physical pressure became a common solution. The population had essentially become trapped between a rock and hard place, as the head of the IVG in the Semarang area poignantly illustrated, in December 1946, when warning his superiors about the danger of spies and infiltrators entering the city of Semarang as part of travelling groups of salesmen:

Seeing as the presence of spies in a group of salesmen guarantees the safety of the rest of the group against raids from PT or TRI patrols [Indonesian Military Police or Army], because these spies outside of the city are always in possession of a ‘soerat keterangen’ [identification], the salesmen will not readily betray them, also out of fear of retribution. Sudden razzias among groups of salesmen only yield results after protracted sharp interrogations, when the bona fide salesmen have started to fear being taken for spies themselves. ⁶⁷

The consequence of this pressure from both sides was a psychological terror enacted upon the population which resulted in high levels of mutual distrust. As one Zainal Ngabidin put it very simply during his interrogation by the security service of the marine corps of Surabaya: ‘[T]he various population groups are distrustful of each other because of the presence of spies.’ ⁶⁸

Such mistrust could be pernicious and may have contributed to longer-lasting regional and communal tensions in the multi-ethnic society that became the independent Republic of Indonesia. ⁶⁹ It not only set neighbour against neighbour, but also, and more importantly, ethnicity against ethnicity. In the interrogation reports from the NEFIS in Padang (West-Sumatra), for example, we can frequently read evidence that especially Ambonese and Menadonese – two ethnic groups considered particularly loyal to the colonizer – but also Javanese were at particular risk of arousing the suspicion of the Indonesian forces. Again we can conjecture from the reports that the Dutch authorities were happy to spur on, or that at least they anticipated, such tensions. Two of the most central questions that were asked of all prisoners were, firstly, how they assessed the ‘attitude of the population’ towards the Dutch authorities, and, secondly, what they knew about the ‘relations between the population groups’. The answer to the former question, as it was recorded in the interrogation reports, was surprisingly often that ‘the population’ was fundamentally well-disposed toward Dutch reoccupation, but that it was afraid of being terrorized by Indonesian armed groups. The interrogator’s wish (or rather his assumption) seems often to have been father to the thought – and it was an assumption that Indonesian arrestees were only too happy to confirm. ⁷⁰ The second question seems to have had the objective of cataloguing weaknesses in the Indonesian position that might be exploited, by teasing out tensions between various groups. Even if such tensions were rarely instrumentalized in a straightforward way for military recruitment or alliance, it did aid in determining who to target as potential informers. ⁷¹ The mutual distrust was thus reinforced as a consequence of the perceived ubiquity of intelligence informers.

Conclusion

In this paper I have studied the actions and perceptions of the Dutch intelligence and security services in Indonesia during the war of decolonization, focusing especially on their practices of arrest and interrogation. As in any counterinsurgency effort, intelligence services played a central role, because they were (at least in theory) tasked with distinguishing civilian from enemy combatant. The illustrious tradition of the colonial political police, ⁷² combined with the conviction that they had learned valuable lessons from the state-of-the-art intelligence practices of the British and the Americans in the Second World War, ⁷³ had created an initial confidence among the leadership of the intelligence apparatus that they would be equal to their assigned task. But this confidence gradually eroded as it became clear that they were not achieving the results that they had hoped for. Hampered by a lack of reliable informers, running up against their own prejudices and misconceptions, and most of all, confronted with a largely silent and reticent population, the intelligence apparatus increasingly failed to deliver the information that the army and police forces needed. The consequence was exasperation among those responsible for carrying out arrests and interrogations, and at least in some cases an increase in arbitrary and more brutally violent behaviour.

Even though the conflict in Indonesia did not involve attempts at population control of the magnitude or pervasiveness that we know from other colonial counterinsurgencies such as in Malaysia, Algeria, or Kenya – something that considering the size and population of Indonesia would have been nigh on impossible – some of the practices of the intelligence services can nevertheless be explained by a similar logic. Exerting control over the civilian population through large-scale imprisonment or forced relocation was not a centrally directed strategy. But the actions of local officers and regional commanders, especially the tendency to police mobility and to restrict border-crossing, could be interpreted as ad hoc, hapless, and ultimately failed attempts at population control: in a sense, it was population control ‘on the fly’.

Secondly, the practices of arrest and interrogation, combined with the recruitment of local employees and informants and with (threats of) violence in case of refusal to cooperate, and in conjunction with an Indonesian strategy of often violently discouraging any cooperation with the Dutch, had the effect of creating a psychological terror among the population, which in turn generated widespread mistrust. Partly this effect seems to have been deliberate, partly it was an inadvertent outcome. It can be assumed that the mistrust and mutual animosity among Indonesians that developed in the years of the revolution had a lasting effect for the stability of the independent Indonesian state and society in its early decades. Exactly how this legacy of (at least in part Dutch-engendered) mistrust transformed into later regional and ethnic conflicts remains to be studied.