Culture, Medicine, and Psychiatry

, Volume 33, Issue 2, pp 290–312

Conflict Nightmares and Trauma in Aceh


    • Department of AnthropologyHarvard University Graduate School of Arts and Sciences
  • Mary-Jo Delvecchio Good
    • Department of Social MedicineHarvard University Medical School
  • Byron J. Good
    • Department of Social MedicineHarvard University Medical School
Special Section: Trauma and Dreams

DOI: 10.1007/s11013-009-9132-8

Cite this article as:
Grayman, J.H., Good, M.D. & Good, B.J. Cult Med Psychiatry (2009) 33: 290. doi:10.1007/s11013-009-9132-8


In both the Acehnese and Indonesian languages, there is no single lexical term for “nightmare.” And yet findings from a large field research project in Aceh that examined post traumatic experience during Aceh’s nearly 30-year rebellion against the Indonesian state and current mental distress revealed a rich variety of dream narratives that connect directly and indirectly to respondents’ past traumatic experiences. The results reported below suggest that even in a society that has a very different cultural ideology about dreams, where “nightmares” as such are not considered dreams but rather the work of mischievous spirits called jin, they are still a significant part of the trauma process. We argue that it is productive to distinguish between terrifying and repetitive dreams that recreate the traumatic moment and the more ordinary varieties of dreams that Acehnese reported to their interviewers. Nightmares that refer back to conflict events do not appear as an elaborated feature of trauma as the condition is understood by people in Aceh, but when asked further about their dreams, respondents who reported symptoms suggestive of PTSD were more likely to report PTSD-like dreams, memory intrusions that repeat the political violence of the past.


IndonesiaAcehTraumaDreamsNightmaresPolitical violence


Hendra was a 25-year-old psychiatric nurse in the Indonesian province of Nanggroe Aceh Darussalam (referred to here simply as Aceh) at the northwestern tip of Sumatra island. He had been working for the International Organization for Migration’s psychosocial support program in postconflict areas of the province since the beginning of 2006. During a staff training seminar when I was introducing ethnographic methods for studying postconflict areas in Aceh, I discussed our research on trauma and dreams in Aceh—the subject of this paper—as an example.1 Over coffee at the end of the discussion, Hendra approached me and said, “I think I can help you with your research.” He then briefly told me the following story.

In 1992, a few years into the Indonesian government’s brutal crackdown against Acehnese separatists when the province was declared a Military Operations Zone for 9 years, Hendra was 10 years old and growing up in a rural area of Aceh Utara (North Aceh) district, not far from Exxon Mobil’s natural gas operations. One day he and his friends were playing child war games about 300 m behind his house, where they found the bloodied corpse of an unknown man with his neck sliced open and his head almost completely severed from the body. Hendra saw the body and its deadly open wound in full detail, and stood transfixed for a while before running to tell the adults. That very night, Hendra told me, he was beset with “strange and terrifying” dreams. It was too far in the past for Hendra to remember the details of these nightmares but he does remember waking up every night in fear, unable to fall asleep until morning, sitting in bed in a daze, usually insisting that his parents remain to help him feel safe. His parents would read koranic verse to help him fall asleep again, and they brought Hendra several times to a religious instructor who gave him water to drink that had been blessed by having koranic mantras chanted over it. After three months of nightmares and prayer sessions every evening, along with periodic visits to the instructor healer, little by little the nightmares went away. Hendra’s own experience has helped him summon the great reserves of empathy required to work with the conflict victims he assists today.

This paper examines the relationship between traumatic experience and dreams or nightmares in Aceh. It begins, however, with acknowledging a paradox. There is no single lexical term in either the Indonesian or the Acehnese language for “nightmare.” In our field research we used the Indonesian term mimpi buruk, which simply means “bad dream.” When I asked two young Acehnese men how they would translate mimpi buruk into their language, they paused, gave me a curious look and then together decided on the circuitous loempoe hana got (literally, not a good dream). The absence of a discrete sign for “nightmare” in Acehnese warns us from the very beginning of this discussion about nightmares and traumatic experience in Aceh to be aware of the potential of a “category fallacy”: the assumption that a symptom or category of symptoms (such as repetitive nightmares) typically linked to an illness (or, in this case, traumatic experience) in our own society will be found in an essentially similar form linked to the same disorder across social or cultural groups, time and place (Hinton et al. 2007; Kleinman 1988). At the same time, Hendra’s story of what we would consider recurrent nightmares following a traumatic experience as a child is a reminder that experiences may be similar across cultures even if unmarked by cultural categories or interpreted differently.

The acknowledgment of the complexity of the relationships among “dreams,” “nightmares,” and “trauma” suggests several questions that animate this paper. First, how do people in Aceh organize and interpret their dream life generally, and how are nightmare-like dreams understood in this context? Second, are dreams a significant feature of an Acehnese posttraumatic response to violence, and how are nightmares related specifically to symptoms of posttraumatic stress disorder (PTSD)? And finally, if survivors of violence have dreams that recall their traumatic experience, are they incorporated into Acehnese understandings of their dreams, and, if so, how?

To address these questions we look at the data from a large field research project in Aceh that examined past traumatic experience during the conflict and current mental distress. The questionnaire included several closed and open-ended questions about dreams related to past conflict experience. In this paper we briefly provide a background first to the war in Aceh and then to our research conducted following the peace agreement signed in August 2005. We summarize the findings from this research briefly, then examine in detail what our quantitative and qualitative data tell us about conflict-related dreams in Aceh and their relationship to Acehnese dream life generally.


The people of Aceh have suffered more than a century of intermittent war and political violence (Reid 2006). Following a protracted war against Dutch colonial intervention from 1873 until around 1910, violence resumed in Aceh with the brutal Japanese occupation of the Dutch East Indies during World War II, followed by the revolutionary struggle for Indonesian independence from 1945 until 1950. After playing an important role in the Indonesian fight for independence from the Dutch, the Acehnese responded to what they saw as broken promises for relative autonomy and, in 1953, turned against the increasingly centralized and insistently secular Indonesian government in Jakarta, headed by President Sukarno. The ‘Darul Islam rebellion’ was resolved only when Sukarno established Aceh as a province with increased autonomy in 1963.

Indonesia’s second autocratic president, Suharto, rose to power upon the nationwide slaughter of hundreds of thousands of suspected Communists, roughly 3,000 of whom were killed in Aceh throughout 1965 and 1966 (Siegel 1979). During the Suharto regime, more centralization and the exploitation of natural gas reserves in Aceh further accentuated longstanding grievances against Jakarta. In 1976, an explicitly secular rebellion, demanding full independence from Indonesia, was launched by the Free Aceh Movement, GAM (Gerakan Aceh Merdeka). Journalists and other observers of the Aceh conflict typically cite 15,000 civilian deaths during the most intensive years of the Indonesian military’s counter-insurgency efforts from 1989 until 2005, but in 2006 the Indonesian government’s own Aceh Reintegration Agency documented more than 28,000 conflict-related deaths (Frödin 2006).

On 26 December 2004, an unprecedented earthquake and tsunami struck the coast of Aceh, leaving more than 130,000 persons dead and a half million displaced, in a province that had a population of roughly 4.4 million people. The shock and incredible losses sustained by these natural disasters at last consolidated the political will needed on both sides to negotiate a peace agreement in Helsinki on 15 August 2005.2 Our data on nightmares and conflict trauma refer specifically to the most violent years of the conflict in Aceh, from 1989 until the peace agreement in 2005, but the overall context necessarily includes the generations of conflict that preceded it.

Methods of Our Field Study

During our first visit to Aceh in June 2005, foreign humanitarian aid workers were still working in a high-security environment, had access to tsunami areas only and were reigned in within city limits by nightly curfews. Following the Helsinki peace agreement, security in Aceh improved dramatically, particularly with the withdrawal of 15,000 so-called “inorganic” Indonesian military and police troops from the province and the formal demobilization of GAM’s military forces and their surrender of 3,000 weapons as specified in the MOU.3 Curfews were lifted and travel restrictions were relaxed, allowing humanitarian organizations involved in tsunami relief to develop programming for equally devastated conflict areas in Aceh away from the coast. Throughout the year 2006, teams of researchers from the International Organization for Migration (IOM) and the Department of Social Medicine at Harvard Medical School, led by the authors of this paper, carried out a Psychosocial Needs Assessment (PNA) in high-conflict villages in 14 districts throughout the province in order to evaluate and eventually provide support for the psychosocial and mental health needs in communities that were deeply affected by years of conflict.4

The PNA research was designed to provide empirical data that could serve as a basis for developing mental health and psychosocial services to support these communities’ efforts at recovery. Specifically, we sought to determine the level of conflict-related traumatic experiences suffered by members of these communities, to assess levels of psychosocial and mental health problems and to identify high-risk subgroups in the population. The study design included two components: key informant interviews and a formal survey of randomly selected adults aged 17 years and older. The qualitative key informant interviews were designed to explore the historical context of the conflict, how it affected communities over time and whether certain segments of the population were more vulnerable than other segments. The formal survey interview was designed to measure past traumatic events associated with the conflict experienced by respondents, to assess current stressor events and to determine levels of current psychological distress associated with these experiences. The survey combined open-ended questions, developed for Acehnese who had experienced decades of conflict and a tsunami, with widely used scales, allowing levels of traumatic events and psychological symptoms of the Acehnese respondents to be compared with findings from other studies of conflict and postconflict populations. The quantitative measures included a 43-item Harvard Trauma Events scale, adapted specifically to represent typical forms of trauma experienced in the communities being surveyed. Levels of emotional and psychological distress were assessed with a 25-item version of the Hopkins Symptom Checklist for Depression and Anxiety (HCL-25), a scale used widely in community assessments of emotional distress associated with disasters or violent conflicts. The 42-item Harvard Trauma Questionnaire (HTQ) is a measure of trauma-related symptoms that includes a 16-item core used to assess PTSD. For both the HCL-25 and the HTQ, respondents were asked to describe whether they have experienced the symptoms during the past week ‘not at all,’ ‘a little,’ ‘sometimes,’ or ‘often.’ Care was taken to incorporate common ways of expressing psychological distress in Indonesia, specifically in Aceh, into these questions. Items on the HCL-25 and HTQ were translated using common Indonesian terms—such as bingung (feeling confused), melamun (daydreaming or ‘spacing out’) and pusing (a combination of feeling dizzy and having a headache). The HTQ item that asked about recurrent nightmares was glossed with the Indonesian term mimpi buruk (literally, bad dream). A significant part of the survey was devoted to open questions allowing for the specific interpretation of Acehnese experiences, including items designed to capture popular discourses about experiences of nightmares, ghosts, spirits and hearing voices of people who had died.

Fieldwork was conducted throughout 2006 in three phases, covering 14 mainland districts of Aceh.5 The field researchers were all Acehnese, consisting of lecturers from local universities in Banda Aceh, psychiatric nurses and recent public health and counseling college graduates. All research staff participated in a three-day training prior to fieldwork. During the fieldwork, a team would visit one randomly selected village per day; six surveyors conducted three interviews each with randomly chosen respondents aged 17 or older, and team leaders conducted key informant interviews with community leaders. Data entry was carried out in Aceh, and quantitative data analyses for the overall PNA1 and PNA2 studies were carried out at Harvard using SAS. Qualitative dream data from the PNA questionnaire were coded for themes and content by a team in Aceh, and some additional ethnographic fieldwork in early 2007 was conducted by the first author specifically for this article in order to find out how Acehnese understand their dreams, particularly their bad dreams. Data analysis for this article was conducted using SPSS version 13.0 for Macintosh. The data reported here are from the 13 districts sampled during the first two phases only and are representative of populations in the high-conflict areas of those districts.

Sample Demographics

A total of 1792 interviews were conducted in 13 districts of Aceh. Table 1 provides an overview of our sample; the distribution by sex and age suggests the efficacy of the random sampling method used by the research teams. The majority of respondents live in their own homes and have had at least an elementary school education. The majority of respondents are also married; the significant difference between widows (13.8% of women) and widowers (1.7% of men) is most likely an artifact of men being disproportionately killed during the conflict.
Table 1

Demographics of study participants (total N = 1,792; n males = 940; n females = 851)

Gender (% total sample)





Age (% total sample)









Marriage status (%)

    Never married







    Currently married







    Divorced or separated














Schooling (%)

    No schooling















    Middle school







    High school







    Associate’s degree







    Bachelor’s degree














Housing (% total sample)

    Live in own home


    Live with friend or relative


    Live in abandoned/destroyed home


    Rent housing


    Live in barracks or tent


Traumatic Experience During the Conflict

Stumbling inadvertently on dead bodies as Hendra did is just one of dozens of varieties of traumatic experience that characterized the violence inflicted on Acehnese civilians during the conflict. The most difficult part of the questionnaire for both interviewers and respondents was the Traumatic Events Checklist. The Indonesian military and police forces had an explicit strategy to “destroy GAM down to its very roots,” which in practice meant targeting the rural villages of Aceh where most GAM forces came from and on which they depended for food, shelter and logistical support (Kompas 2003).

A few examples from our findings illustrate the profound effects the conflict had on civilian populations. Seventy-five percent of the total sample report having lived through combat experiences and 53% experienced having to flee from danger in their community. Thirty-nine percent of the sample had a family member or friend killed. Five percent of all women respondents reported that their husband was killed in the conflict, and four percent of all respondents had lost children. Forty-one percent reported having their property confiscated or destroyed, and 25% experienced extortion or robbery. People were forced to fight (18%) or else be punished for not fighting (11%), and were forced to search for members of their community in the forest (28%). Thirteen percent of respondents were publicly humiliated and five percent were forced to humiliate another person. Forty-four percent of the sample witnessed physical punishment; the qualitative data suggest that this was usually by force and that most physical punishment was accompanied by public humiliation. These experiences of humiliation were more commonly experienced by men than women, but it should be acknowledged that while rape and other forms of sexual violence were included on the traumatic events checklist, we found few respondents willing to acknowledge having personally experienced violent forms of sexual humiliation that were widely known to have been systematically committed during the conflict in Aceh. Men also experienced higher levels of violence inflicted on the body. Forty-four percent of men in our sample, compared with 13% of women, reported beatings to their body; 25% of men, compared with 11% of women, reported being attacked by a knife or a gun.

These statistics provide clear evidence of the magnitude of suffering and terror experienced in these communities, but the narratives found in both the open response questions on the survey and the key informant interviews have emotional and testimonial qualities that numbers cannot measure. The qualitative data of the study are, first and foremost, filled with stories about men and women being brutally interrogated, intimidated and threatened for information they could not provide and then severely beaten (or worse) for not having answers. Some vivid additional examples include suffocation with plastic bags, public displays of sexual humiliation, drownings in septic tanks and sewage canals and individuals being forced to injure or humiliate friends and loved ones. Women described being forced to watch with their children as their husbands and sons were mutilated and killed. An entire village was forced to watch while a dozen homes were set afire with their residents trapped inside. Stories of being forced to provide labor or being forced to serve as human shields are common. Many of these scenes are reflected in the nightmare narratives discussed below. In addition, many communities reported having schools and public buildings burned or destroyed and having been extorted for money by both the guerilla and the government security forces, leaving them bereft of community resources. All community officials were invariably required to provide information about and take responsibility for the actions of their village populations to both sides during the conflict, creating a hopeless sense of entrapment and a reluctance to lead.

Estimating Levels of Depression and PTSD

The overall study (PNA1 and PNA2) found extremely high levels of psychological symptoms in this population. In the larger study, we used the methods for calculating the percentage of persons suffering depression and PTSD recommended by Mollica et al. (2004). (Following these procedures allows us to compare findings for the Aceh sample with similar samples from high-conflict areas such as Bosnia and Cambodia.) Using recommended cutoff levels (mean scores of 1.75 for depression and 2.5 for PTSD), 44% of the total population suffered high levels of depression symptoms and 17% suffered high levels of PTSD symptoms. Using algorithms to estimate diagnoses, as recommended by Mollica et al. (2004), 33% of the total population met the criteria for major depressive disorder, and 19% for PTSD. Levels of symptoms of PNA1 respondents were much higher than those of PNA2 respondents, suggesting that symptoms declined with time and with a reduction in levels of current stressors (in particular, with the inorganic troops, most responsible for the violence, having left the province). Using odds analyses, the single factor that most highly predicted level of psychological symptoms was level of traumatic events experienced by an individual.6

Table 2 estimates the prevalence of PTSD in the sample reported in this paper, using cutoff scores from the HTQ data. We follow the standard procedure recommended by Mollica et al. to use a mean score of 2.50 as a cutoff on the HTQ to identify a person as suffering a PTSD (Mollica et al. 2004). The standard estimate of PTSD prevalence in this sample is 19% using the 2.5 cutoff score, with significantly more women (21%) than men (17%) scoring above 2.5.
Table 2

Two ways of estimating PTSD, by gender* and total sample

PTSD measure

% Male (n = 940)

% Female (n = 851)

% Total sample (N = 1,792)

Mean PTSD score ≥2.5 (“symptomatic”)




* Chi-square test, men versus women, p < 0.05

Local Idioms: Trauma and Stres in Aceh

During questionnaire interviews respondents were asked the following question before the symptom checklists: “The conflict has brought unique pressures upon the people of Aceh during the past number of years. Have these pressures had an effect on your feelings, energy, or your health in your daily life? Can you explain what this effect has been?” The responses to this question are interesting because respondents describe for interviewers how they think of mental illness in their own words before the standardized checklists suggest the symptoms to them. Their answers yield a list of local idioms of psychological distress (see Table 3).
Table 3

Respondent descriptions of psychological distress

Loss of spirit or energy (kehilangan semangat)

Frequent sadness

Exhausted for no reason

Crying often

Unable to work



Suspiciousness/mistrust of others

Uneasy, restless

Hard to socialize

Unable to sleep at night


Shaking uncontrollably




Body hurts

Remembering what happened

Frequent headaches

Hard or slow to think

Heart problems (many varieties)

Thinking too much



    Racing heartbeat


    “It feels as if my heart has fallen”


    Weak heart


    Heart attack (upon hearing bad news)


The first thing one learns when talking to people in Aceh about mental health is that the English words “stress” and “trauma” have been thoroughly absorbed into their understandings of mental illness. One key informant told his interviewer: “Before the conflict, no one around here knew the word trauma.” As in English, these two words have gained such a broad currency in the local language that it is hard to know exactly what someone means when they say stres or trauma. Nevertheless, deeper investigation yields some broad generalizations—and indicates substantial differences between Acehnese- and English-language use of the terms. The meanings of stres and trauma overlap, both denoting deep psychological distress brought on by external events such as war, a death in the family, or a natural disaster, but judging from the use of these words during key informant interviews and everyday conversation, trauma is a temporary condition from which one can recover. In contrast, the word stres—unlike in English, where stress might suggest something as light as the effects of a bad day at the office—denotes a more serious, long-term condition that may require psychiatric care at a hospital. One might be suffering from trauma but still be present and at least appearing functional in the community, whereas when someone has stres, he or she is noticeably debilitated from performing everyday social roles. Hence the word trauma tends to appear more often in interview transcripts (e.g., “Everyone in this village is still trauma from the conflict”).

Beyond these two broadly defined categories of mental distress, the answers to the question quoted above yield a distinct list of common symptoms that describe local understandings of trauma. Many respondents refer directly to somatization of mental distress, which is to say not just that psychological distress brought on by the conflict frequently manifests as physical illness, but also that people in Aceh understand that some of their physical ailments, particularly in the heart, are caused by psychological distress. Most people mentioned “fear” (takut) as a common symptom of trauma, but rarely was fear mentioned as a general condition. Usually, respondents tied their fear to something specific like “I am afraid of crowds” and “I get scared whenever I see military uniforms/hear a motor vehicle/hear a noise that sounds like gunfire.” Table 3 summarizes the symptoms, physical and psychological, that were most frequently mentioned to describe Acehnese understandings of psychological distress.

It is remarkable that, although many respondents mentioned insomnia, not even one respondent mentioned nightmares or dreams of any kind as a sign of psychological distress resulting from conflict experience. If Table 3 offers us a schematic overview of how trauma looks and feels to the respondents in our study, then a preoccupying question for the rest of this paper is to what extent, if at all, nightmares are a part of this “trauma ontology” in Aceh.

Respondents’ Descriptions of Their Dreams

The absence of a discrete sign in the Acehnese language for a concept that resembles “nightmare” further suggests that a person’s dream experience might not be an important marker of traumatic disorder for many people in Aceh. And yet, when respondents were asked as part of the HTQ if they had frequent repetitive mimpi buruk during the past week, 21% of the sample responded with either “sometimes” or “often.” When asked whether respondents were visited by spirits of the dead in their sleep, 24% of the total sample said yes. Following that, an open-ended question gave respondents a chance to tell interviewers in more detail about any nightmares they have had since the conflict, and 458 respondents (26% of the sample) took the time to tell us something about their dreams. Women were slightly more likely to describe a dream (28%) than men (23%) (χ2p-value < 0.01). Responses were coded for their affect and the presence of various themes. These themes are summarized in Table 4 and described below.
Table 4

Can you tell me about any nightmares that you have had frequently since the conflict?


% of respondents reporting a dream (total N = 458) whose report included a theme

Seeing spirits of the deceased

54 (n = 247)

Of those seeing sprits of the deceased, what percentage were pleasant visits?

52 (n = 247)

Neutral or positive content, not a nightmare


Violent conflict events


Frequent and/or repetitive dreams


Violence afflicts the respondent in dreams


Manifestation of guilty feelings


Dream contains a message


Disturbed or tempted by evil spirits or ghosts


Insomnia and other sleep disturbances


Nightmares occur shortly after violent events only


Tsunami themes


Religious themes


“Not a Nightmare…”: Pleasant Visits with the Deceased

Although the question specifically asks about mimpi buruk, if we loosely define a nightmare as a dream that inspires negative affect (sadness, fear, pain), many of the responses to this question (36%) instead discuss ordinary dreams or dreams that made respondents feel good. In particular, among the respondents who describe the visits in their dreams from people who died in the conflict or tsunami (N = 247), 52% described them as pleasant encounters. Some respondents went out of their way to correct the interviewer and emphasized that it was not a nightmare but, rather, quite the opposite: “I dream about my husband, and it is pleasing and beautiful, not a nightmare.” Their comments suggest pleasure, comfort, wonder and yearning for more visits with the spouses, children and friends they miss. When respondents told interviewers that their deceased brother or sister visited them to say thank you for a proper burial and to ask them to take care of surviving orphaned children, one might think of these visits as therapeutic rather than pathological, as a process that helps bring closure to traumatic loss (Mollica 2006). The following are examples of such reports.

I’ve never had nightmares, but I have had dreams about my relative who passed away and I invited him out to sea to go fishing just like we used to do before he died. (Male, 57)

I dreamed that my husband came to visit all of us, he spoke and joked around with the children as if he was still alive. Especially at the end of the fasting month, that’s when I dream most often that my husband comes to visit. (Female, 45)

I never had a nightmare, but I once dreamed that I met my husband and he told me that he would return someday. In my opinion that’s a good dream. (Female, 56)

I dreamed that I was speaking with my son. He was sitting under a tree near the mountain, and he told me that I shouldn’t worry, that he was healthy and has everything he needs. (Female, 40)

I met him in my dream and felt inspired because he was doing well. (Female, 44)

As the gender assignments to the quotes above suggest, women were more likely to describe pleasant dreams than men. Of all 127 respondents who described pleasant dreams, 61% were women (χ2p-value < 0.05).

Dreams of Violent Conflict Events and Violence Afflicting the Self

For every ordinary or pleasant dream described in answer to the open question about nightmares, there are as many descriptions of nightmares that repeat a common set of violent conflict events that respondents experienced or witnessed themselves. Troops—usually TNI (Indonesian military forces) or BRIMOB (Indonesian mobile police brigade), sometimes “unknown assailants” (“OTK”; orang tak dikenal), much less frequently GAM forces—enter a respondent’s village and then beat, shoot, capture or kidnap the respondent or the respondent’s close relative or friend and bring him or her to the nearest military post, to be tortured until bloody or dead. Dreams end at the moment when the victim is the most bloody, is in the most pain and screaming or dies, and the respondent almost invariably wakes up in a “cold sweat” and cannot fall back to sleep until sunrise. Other common narratives about the conflict include getting caught in the line of fire (kontak senjata) during armed battles between GAM and TNI, and being endlessly chased, stalked and hunted by the TNI or BRIMOB. Without any special probing questions from the interviewing staff, many respondents volunteered enough information suggestive of the textbook posttraumatic nightmare, a repetition of past events with little or no variation in content, occurring frequently and terrifying enough to prevent respondents from returning to sleep:

I dream that my son who joined GAM died after being shot and tortured sadistically by the TNI … he was screaming to me for help and I wake up suddenly and can not fall back asleep. (Male, 42)

I see the dead thrown into the river. (Male, 32)

I dream about the battles in my village, see my father-in-law taken by the TNI and buried alive. (Male, 55)

I saw it all over again in my dream, my son being tortured by them. I felt so sad, and this happened every single night, for weeks. (Female, 47)

One day I go off to sell my produce, then I am captured by the security forces, then I am beaten until I am dead. I have this dream often, especially when I’m sick or not feeling so healthy. (Male, 54)

Back in 2005, I would always dream that I was being electrocuted in exactly the same way I saw when my friend was tortured at the military post one year ago. (Male, 20)

I often dream about when I was tortured while held captive at the TNI post and forced into the irrigation canal in front of the post for one day and one night. (Male, 35)

I always dream about the time when BRIMOB caught me and my husband. They beat my husband in front of our house and then they wrote “GAM” on the front wall. (Female, 30)

In the dream it’s as if I am repeating over again the time when I was stripped naked and dragged all around by them on the ground. (Male, 33)

My dream is so terrifying. I see bodies gory and smeared with blood, others decapitated and shot up. What’s so strange about the dream is that these were all things I saw in my village with my own eyes that one time … all this violence carried out by the TNI. (Female, 21)

Every night I dream about the time when I was tortured or when I saw their necks were slit and others were shot on sight by the TNI. After waking up I can’t sleep until the morning. (Male, 47)

Although slightly more men than women reported dreams about violent conflict events, the difference was not statistically significant. Fortunately, some respondents (6%) mark their nightmares with the past tense, noting that these vivid and troubling experiences, though once frequent, occur rarely now or have not presented since the violence stopped. (e.g., “Yes, I had nightmares about the conflict, but not since the peace agreement,” “I always had nightmares right after a conflict event in my village,” and “I have nightmares whenever the security situation is bad.”)

Dreams Containing Messages

Many of the dreams cited above feature messages for the dreamer, and our fieldwork in Aceh shows that these messages are an important component of the dream experience. In the structured questionnaire of our research, only 26% of the entire sample shared with us something about their dreams. We found that it can be quite difficult to elicit dream narratives from people in Aceh during conversations with friends. People rarely talk about their dreams.7 This is because messages received in dreams may bear truths (kebenaran) to the dreamer that are cherished and held in secret.

An exemplary dream from Aceh Selatan recounted in Mike Griffiths’ book Indonesian Eden illustrates the principle:

A lady had a dream in which two orphaned kittens approached her and begged for food. She consented and the kittens expressed their gratitude. The next day while working in her ladang [fields], she saw two tigers at the forest’s edge. Recognizing the significance of her dream, she prepared food and left it at the place where she saw the tigers, whistling as she left. After that she continued to leave food out, and periodically the tigers came to eat perhaps learning to associate her call and whistle with the opportunity for easy food. (McCarthy 2001)

It is believed that the message to feed and look after the tigers ultimately protected the community’s fields and the villagers themselves from tiger attacks, a rare but real threat to villages near the mountain forests of Sumatra.

Respondents affected by the conflict take away a variety of messages from their dreams. One woman who described having frequent dreams of being shot dead just like her husband told us that, after the peace agreement, her husband has visited her often in her dreams and tells her not to follow him down that path, that he is already too far away. Deceased husbands tell their wives to have patience in facing the difficult future ahead and grant them permission to remarry. Deceased siblings tell their brothers and sisters to take care of their surviving children, not to let anyone else take them away and to always look out for their parents’ welfare. Deceased parents remind their children to pray five times a day and to always take care of one another. Other messages are more specific:

A very old man instructed me to open up a small business to support my family. (Male, 50)

My son who died came to me in a dream and told me that I could find his corpse near the elementary school. (Female, 32)

My brother asked me to bury him near our mother’s grave. (Male, 46)

My younger brother came to me in a dream with a decomposed face, and he asked me to go to the forest to find his corpse and bury it in a better place. I’ve had no other dreams, but this dream about my younger brother doesn’t bother or scare me because he is my own brother. (Male, 31)

After I took out the corpse that was thrown in a ditch, its spirit came to me in a dream and thanked me for helping him … it felt so real. (Male, 43)

There was no statistically significant difference between men and women in the reporting of dreams that contain messages.

Dreams of Being Tempted, Disturbed or Terrified by Evil Spirits

The generic term jin refers to spirits with bad intentions, which range from deceitful tricksters that try to steer people away from piety to indescribably dark and terrifying figures that try to kill the dreamer. Jin are the source of ordinary frightening dreams, i.e., not the posttraumatic nightmares described above. A disturbance or temptation brought by jin makes the whole enterprise of interpreting one’s dreams a treacherous exercise, because it can be difficult to sort the truths (kebenaran) that some dreams reveal from the temptations and lies of a jin. There are a few rules that can help dreamers sort truths from temptations, such as noting the time of the dream. If one has a dream during predawn hours, after about 4 a.m., this is automatically assumed to be the product of jin and should be ignored. A dream reveals truth only if the dream repeats itself a minimum of three times. But one can never be sure, as the following story, which also incidentally features tigers from the mountain in Aceh Selatan and which we heard from a friend’s uncle, shows.

A man dreamed that he was visited by an old man wearing all white clothes, who instructed him to go to a nearby mountain where he would find a cave with lucrative swallow’s nests. The old man told him that in order to reach the cave he would have to walk to the mountain alone, on a Friday morning, and he would meet a tiger that would show him the way. The tiger would bring him to a large stone outcropping and behind the huge stone he would find the entrance to the cave. Although the dream was clear, the man did not immediately follow the message because he only had the dream once, and he was terrified of tigers. He did not look for the cave, nor did he tell anyone, because he felt it might be a message but he still was not sure. After having the same exact dream a second time, the man sought a trusted dream expert who would help him understand. The expert told the man to be patient, if there is a fortune to be found in such a cave, then the dream would surely repeat itself a third time. Indeed, the dream occurred a third time, so he decided to take the risk and follow the dream’s instructions. Although instructed to go alone, he brought the expert with him because of his fear of encountering the tiger. In fact, they did come across a tiger along the way, and the man said to the tiger, “If indeed you are the guide, please show us the way to our destination but stay far ahead of us so that we can’t see you … we will follow your footprints only.” The tiger’s footprints lead them to a huge stone outcropping, around which they could see swallows flying. The expert encouraged the man to go around and look for the cave entrance. It was already 11:30 a.m., however, and to go any further would certainly cause them to miss Friday prayers. After discussing the matter they decided to go home for Friday prayers and look for the cave another time.

The man and the dream expert decided to avoid what may have been a trick by a jin trying to prevent the man from fulfilling his religious obligations by tempting him with massive profit. To date he has not returned to the cave. Though not a nightmare per se, dreams like the man in Aceh Selatan had are unpleasant because they tease, perplex and distract.

Jin bring other unpleasant dreams that approach our understanding of the word nightmare as it is used in English. By far the most common nightmare described is a less elaborate version of the kind of sleep paralysis described in many other parts of Southeast Asia. A large, tall, black and obscure figure sits on the chest of a sleeping person and pushes down so hard and/or takes the person into such a tight and choking embrace that the person cannot move, cannot breathe and cannot shout for help. Usually dreamers of this terrifying figure are half-awake and aware of their paralysis. Acehnese informants variously describe this phenomenon as digeunton (pressed on) or dicekek (choked or strangled), and everyone has experienced this at least once in his or her lifetime, including many respondents in our sample:

I often see a large tall black figure, always present in my dreams. (Female, 22)

I often wake up in the middle of the night and feel as if a person is strangling my neck. (Male, 72)

Many nightmares feature respondents being chased or haunted by unidentifiable ghosts or spirits:

I’m always chased by someone, but his shape is not clear. (Male, 40)

Strange people show up in my dreams, sometimes they have no face, sometimes no body, almost every night like this. (Female, 17)

Suddenly I was brought to a graveyard by a horse, afraid the horse will bite me, so I run away. In my dreams I am always taken to a spooky place … it scares me. (Male, 58)

I dreamed I was being chased relentlessly by a large tall dark black man with large red eyes, he chases me on and on.… Who wouldn’t be scared cold? Although it was just a dream it felt so real. (Female, 25)

I don’t remember exactly, but I often see people peeking and spying upon me. (Female, 45)

…choked by a ghost, and I scream and run far away but I never get there. (Female, 17)

I’m driving a moped near the sea and then suddenly chased by a huge snake. (Female, 40)

I always dream about the ghosts and monsters I see on the TV mystery programs. (Female, 25)

There was no statistically significant difference between men and women in the reporting of dreams that would be ascribed to the work of jin.

Trauma and Nightmare Descriptions

When asked to share their nightmares since the conflict, less than 64% of the 458 people who answered the question gave responses that approximated a nightmare. These dreams were coded for their content, noting whether their nightmares refer explicitly to conflict violence or something more commonly ascribed to the work of jin during one’s sleep. Table 5 shows what percentage of the respondents who described different kinds of dreams meet or exceed the cutoff score of 2.5 we used for estimating PTSD in this sample. The results show that those who described nightmares in general and those who described nightmares about conflict violence are significantly more likely to meet the criteria for PTSD than those who did not. For example, 44% of the respondents who described a general nightmare met the standard 2.5 HTQ cutoff score criterion for being PTSD symptomatic, whereas only 22% of the respondents who did not describe nightmares met this criterion. Forty-eight percent of the respondents who described nightmares based on conflict events met this same criterion, whereas among those who did not, only 30% met the 2.5 cutoff score. This result does not change significantly when the analysis controls for nightmares typically ascribed to the work of jin. Both of these cross-tabs (general nightmare and conflict nightmares) had highly significant chi-square values (p < 0.001). On the other hand, respondents who described dreams that featured pleasant visits with deceased friends and family were less likely to be symptomatic for PTSD than those whose dreams did not feature such a visit, suggesting that this kind of dream helps respondents work through their traumatic experience. Finally, nightmares typically ascribed to the work of jin show no association at all with the criteria for PTSD (p = 0.47). This result does not change when the analysis controls for nightmares that refer to conflict violence.
Table 5

Estimating PTSD among respondents who described their dreams


Was the dream a pleasant dream?*

Was the dream a nightmare?**

Did the dream feature conflict violence?**

Did the dream feature evil spirits or ghosts?a

% Total sample (N = 458)


% Yes (n = 127)

% No (n = 330)

% Yes (n = 291)

% No (n = 167)

% Yes (n = 160)

% No (n = 298)

% Yes (n = 30)

% No (n = 428)

Mean PTSD score

    ≥2.5, “symptomatic”




















Note: Yes versus no (independent-samples t-test): * p < 0.05; ** p < 0.005; a no significant difference

Table 6 presents the same results by comparing the mean PTSD scores. Respondents who described a general nightmare had an average PTSD score of 2.39, which was significantly higher than the average PTSD score of respondents who described a dream that was not a nightmare (2.02), with a p-value of <0.005 for an independent-samples t-test. The same significant difference pertains to respondents whose dreams referred to conflict events (average PTSD score = 2.44) compared with respondents whose dreams did not (2.16), while the scores of those who described dreams typically ascribed to the work of jin are nearly the same as the scores of those respondents who did not describe jin-inspired dreams.
Table 6

Average PTSD scores among respondents who described their dreams


Was the dream a pleasant dream?*

Was the dream a nightmare?**

Did the dream feature conflict violence?**

Did the dream feature evil spirits or ghosts?a

% Total sample (N = 458)

% Yes (n = 127)

% No (n = 330)

% Yes (n = 291)

% No (n = 167)

% Yes (n = 160)

% No (n = 298)

% Yes (n = 30)

% No (n = 428)

Mean PTSD score










Note: Yes versus no (independent-samples t-test): * p < 0.05; ** p < 0.005; no significant difference

Another way of looking at the data is to ask whether people who meet or exceed the PTSD cutoff score of 2.5 are more likely to have reported bad dreams, in particular, conflict dreams (see Table 7 Of the 165 respondents who told us something about their dreams and had PTSD scores ≥2.5, 78% (128) described bad dreams and 47% (77) described dreams that featured conflict violence, whereas only 56% of those who scored <2.5 described bad dreams and 28% described dreams with conflict violence. Both of these cross-tabs had high statistical significance (χ2p < 0.005). To put it more simply, of all the respondents who told us something about their dreams, those who were symptomatic for PTSD were roughly 1.4 times more likely to have described a general nightmare and 1.7 times more likely to have described a nightmare featuring conflict violence than respondents who were not symptomatic. Meanwhile, PTSD-symptomatic respondents were 11% less likely to describe a pleasant dream than respondents who were not symptomatic, and were neither more nor less likely to describe dreams featuring evil spirits.
Table 7

Type of dream described by mean PTSD score above and below 2.5

Dream type

Mean PTSD score

% Total sample (N = 458)

% <2.5 (n = 293)

% ≥2.5 (n = 165)

Pleasant dream*




General nightmare**




Nightmare with conflict violence**




Nightmare with evil spirits or ghostsa




Note: Score <2.5 versus ≥2.5 (chi-square test): * p < 0.05; ** p < 0.005; no significant difference

Discussion: No Nightmares in Aceh?

In a 1978 article about curing rites, dreams and domestic politics in Aceh, anthropologist Jim Siegel (2000a) suggested that there is no word for “nightmare” in the Acehnese language because all dreams, apart from “true” dreams with a message, are unpleasant and, by definition, brought by jin. Dreams with a message, called loempoe in Acehnese, tell the dreamer something about the future. They are deemed true when the dream repeats itself, and the source is known to be the spirits of loved ones or God himself. Many of the dreams recounted above are loempoe, such as the dreams about a mountain cave with swallows nests, instructions on where to find a loved one’s corpse and how to bury it and reassurance from deceased husbands that, with patience, the future will be alright. Some respondents shared with us these special messages when we asked them about their “nightmares” related to the conflict, and many went out of their way to tell us they were not scary, not mimpi buruk, because they featured visits and signs from people who died during the conflict and whom they miss dearly.

For the respondents who did tell us about dreams that scared and disturbed the dreamer, the results reported in Table 7 suggest that it is productive to distinguish between conflict-related nightmares and other nightmare-like dreams typically ascribed to the work of jin. Although it is difficult to conclude with certainty from our field staff’s transcriptions alone, the examples of conflict nightmares quoted above approach the textbook definition of a posttrauma nightmare that repeats either an exact or a nearly exact version of past lived experience during the conflict. This contention is supported by the significant association between respondents’ reported conflict dreams and their PTSD evaluations, bearing in mind that the jin-inspired dreams are not associated with the same PTSD criteria.

But we are left with the curious disjuncture between the reporting of terrifying conflict nightmares and their attendant associations with PTSD, on the one hand, and the total absence of nightmares or dreams of any kind, on the other, in Acehnese descriptions of what constitutes trauma. Conflict nightmares about the past do not easily fit into Acehnese categories of loempoe, which describe potential futures, and the undesirable varieties of jin-inspired dreams. Adrianne Aron uses the term “political nightmare” to describe a similar set of nightmares among Central American refugees in the United States. The content of a political nightmare

originates in the terrifying conditions of institutionalized political repression … and bears a direct relationship to the way power is distributed and exercised in the countries where the dream was born … it is a typical dream of posttraumatic stress, of an individual whose waking hours and sleeping hours alike are intruded upon by the traumatic events of the past, when the individual was only one step ahead of death. (Aron 1996)

The assailants described by the respondents in our sample are almost invariably TNI and BRIMOB; they are the torturers carrying out a systematic campaign of psychological warfare and humiliation against Aceh’s civilian population. In a more recent essay about Aceh that addresses the conflict violence in the late 1990s, Siegel documents the bewilderment felt by people in Aceh over the extreme and inconceivable displays of sadistic violence inflicted by TNI and BRIMOB on ordinary Acehnese civilians—a level of violence that renders obsolete any capacity to recognize Indonesian forces (and, by extension, Indonesia) as anything but an incomprehensible “other” (Siegel 2000b). Conflict nightmares that replay terrifying events of the past are as foreign and unrecognizable in the Acehnese dreamscape as the Indonesian troops that perpetrated the dreamers’ original traumatic moment.

Ernest Hartmann has argued that the posttraumatic nightmare is a memory intrusion and not truly a nightmare. His research among Vietnam veterans suggests that the “memory intrusions” suffered by veterans with PTSD during their sleep are quite different from the nightmares suffered by veterans without PTSD in terms of content, repetitiveness, biology and function (Hartmann 1996). The posttrauma nightmare repeats over and over; the content never changes because it does not become “absorbed” or “connected” into accumulated memories. But ordinary nightmares can be about anything, and their meaning and importance are construed differently according to variable social and cultural contexts. In Aceh, ordinary nightmares are the work of mischievous or malicious jin, and the posttrauma nightmare is a foreign intrusion, a political nightmare that for some respondents ceased to disturb their sleep after the peace agreement, when the foreign intruders left the posts they occupied in villages all across the province. Bearing in mind that dreams of any kind were not mentioned on the list of respondents’ descriptions of psychological distress (Table 3), we might conclude that nightmares are not an elaborated feature of trauma as the condition is understood by people in Aceh. Only if one who suffers these political nightmares is asked the reason for these dreams might we expect an answer that recognizes and names the condition: “Saya masih trauma” (“I am still [suffering from] trauma”).


Singular first-person pronouns (“I” and “me”) reference the first author; plural first-person pronouns (“we” and “us”) reference the collective authorship.


GAM has since demobilized its troops and transformed into a political organization. On 10 December 2006, Aceh was the first province in Indonesia ever to hold direct elections for a provincial governor and district-level regents. Aceh will also be the first Indonesian province to form local political parties, but meanwhile it was mostly the independent candidates with a history of leadership in GAM who swept the elections in December 2006.


The term “inorganic” troops refers to military or police forces temporarily imported to Aceh from other parts of Indonesia. Inorganic forces sent to Aceh followed orders from central command authorities based in Jakarta, thus outranking the local “organic” troops, which followed orders from regional command structures based in the province. Organic troops were subject to as much scrutiny as civilian populations for their potential involvement in rebel GAM activity.


Byron Good and Mary-Jo Good were principal investigators of the project from the Harvard side. Jesse Grayman worked for IOM as the research coordinator, leading the on-the-ground investigation and data collection. Mary-Jo Good and Matthew Lakoma were responsible for the quantitative data analysis for the major reports from the study, while Jesse Grayman was responsible for the qualitative data analysis. Byron Good, Mary-Jo Good, and Jesse Grayman were responsible for publication of two reports on the larger project: Good et al. (2006), referred to as PNA1; and Good et al. (2007), referred to as PNA2.

Villages were randomly selected from high-conflict subdistricts in the following districts, organized chronologically by the three phases of fieldwork:
February 2006:

Pidie, Bireuen, and Aceh Utara

July 2006:

Aceh Timur, Aceh Tamiang, Aceh Tenggara, Gayo Lues, Aceh Tengah, Bener Meriah, Aceh Barat, Nagan Raya, Aceh Barat Daya, Aceh Selatan

November 2006:

Aceh Besar.


Detailed reports on the findings and analyses, and full discussion of the differences between PNA1 and PNA2 findings, are provided in the PNA reports (Good et al. 2006, 2007).


A notable exception to this is when people overhear an individual talking nonsense while sleeping. Called igauan in Indonesian, or wen-wen (sometimes won-won) in Acehnese, individuals never know that they were talking in their sleep unless those who overhear it tell them the next day. The dreamer who talks in his or her sleep does not remember it, but his or her friends or family who overhear will tell the dreamer about it the following day, often in a teasing manner. Wen-wen are caused by distracting spirits (jin) and are said to occur when someone goes to sleep with unfinished business from the previous day.


Copyright information

© Springer Science+Business Media, LLC 2009