The eradication of Cham Muslim women's ethnic identity in Cambodia, 1975–79

Abstract

Between 1975 and 1979 the genocidal regime of Democratic Kampuchea (DK) in Cambodia targeted minorities including the Cham Muslim population. To hold the regime to account for its crimes against the Cambodian people, the Cambodian government in 2001 formed the Extraordinary Chambers of the Courts of Cambodia (ECCC). Using transcripts of testimonies and judgements from the ECCC, this article examines Gender-Based Violence (GBV) among the Cham Muslim population. The study shows that Alexander Hinton's arguments to explain GBV using cultural frameworks are insufficient in this case. Indeed, Nicole Rafter has proven it is important to take into account the broader genocidal context to the violence. This article argues that ECCC documentation proves that GBV cannot be explained by cultural contexts alone and instead needs to be understood as a means to destroy the Cham and Cham culture.

In 1970, the US air force under President Richard Nixon dramatically stepped up its bombing of Cambodia as part of ‘Operation Freedom Deal’. The coup against King Norodom Sihanouk by his defence minister, Lon Nol, provided the catalyst for this increase to replace ‘Operation Menu’ (1969–70).¹ Lon Nol's beleaguered government suffered significant setbacks and was characterised by high unemployment and widespread government corruption.² Believing the Americans would provide economic and military aid, Lon Nol broadly supported an expanded US military presence on Cambodian soil, and the ongoing US air force bombing campaigns which resulted in the deaths of over 150,000 Cambodian civilians.³ The bombings also advanced the invasion of the country by the North Vietnamese army: the People's Army of Vietnam (PAVN). PAVN operatives, along with members of the mostly Vietnamese National Liberation Front (NLF), propagated an anti-American, pro-revolutionary sentiment. Sihanouk gambled by asking royalist supporters to back the Khmer Rouge (KR), led by Pol Pot, hoping that at the right moment he would be able to take back power. James Tyner notes that it was this, combined with the aerial bombardment, that created a ‘groundswell of support’ for the KR.⁴ This paved the KR's rise to power, particularly in the Eastern Zone, where bombings were most harsh.⁵ Despite Congress ordering a stop to the bombings in 1973, the KR relentlessly ‘liberated’ Cambodia until, by 1974, just Phnom Penh and several provincial capitals remained outside of its control.⁶ Because Lon Nol's government was so dependent on American aid, when the US reduced its air drops, it was unable militarily (running out of ammunition), economically (lack of food), and politically (lack of support) to survive.⁷ On 17 April 1975 the KR captured the capital, Phnom Penh, taking over a legacy of destruction that resulted from the actions of the Americans and Vietnamese.⁸

Urban residents were forced to march along highways into the countryside, where they were to become agricultural labourers, at least superficially on a level with Cambodia's already mainly rural population. Lawyer Silke Studzinsky has argued that the DK regime sought to create a ‘gender-neutral state’ where both men and women would contribute equally in the development of the new ‘Khmer nation’.⁹ Those who did not at least outwardly comply would be ‘smashed’, the DK euphemism for being killed. A mysterious sociopolitical arm of the government, known as Angka, ‘the Organisation’, regulated private life in its entirety.¹⁰ By the fall of the regime on 7 January 1979, around 25 per cent of the population, up to 2 million people, had either been killed or had died of starvation or illness.¹¹

The DK regime made a distinction between ‘New People’ and ‘Base/Old People’. The former included anyone ‘evacuated’ from Phnom Penh after 17 April 1975. Base People, almost always rural, poor, including the Stieng, Mnong and Stre upland minority groups (Khmer Loeu) who were most affected by the US bombings, received a form of special treatment under the regime: retaining, until roughly 1977, private property and animals, receiving larger food rations, and working fewer hours than those coming from the cities. In contrast, New People were forced to build their own communities and were generally segregated from ‘Base’ territory.¹² Certain ethnic minorities, including the Vietnamese, Chinese, Malay, Cham, and occasionally Thai populations, were considered New People, even if their communities had lived in Cambodia for centuries.¹³

History of the Cham in Cambodia

The Cham in Cambodia are predominantly descendants of individuals from the Hindu-Buddhist Champa civilisation, a maritime power once centred along Vietnam's central and southern coast.¹⁴ They are an Austronesian ethno-linguistic group, have matrilocal family structures, and today have communities throughout Cambodia. Along with the Malay, the Cham form one of the two major Muslim groups of Cambodia; Cham-Malay became a creolised identity formed specifically in Cambodia.¹⁵

Nicolas Weber and Alberto Pérez-Pereiro note that the Cham arrived in Cambodia in waves.¹⁶ Đại Việt emperor Lê Thánh Tông invaded Vijaya as a retaliation against the Champa king's attack on Hoa-chau in 1469. It was this capture and destruction of the Champa capital in 1471 that led to the first mass wave of Cham migration to Cambodia, although Cham were present in Cambodia from at least the eleventh century.¹⁷ More came in the seventeenth century after a preliminary Nguyễn incursion into the Kingdom of Pāṇḍuraṅga, the last independent polity of Champa. Later, in the eighteenth century, they came to escape the Trịnh-Nguyễn and Nguyễn-Tây Sơn conflict, and yet more arrived in the nineteenth century to escape the Nguyễn further conquest of Pāṇḍuraṅga.¹⁸ ‘The evidence suggests’, according to William Noseworthy, that there was ‘at least some form of Islamic practice in what is now Vietnam and Cambodia [as a cohesive unit] by the tenth century’.¹⁹ The ‘Champa Sea’, a trade path between Arabia and China, inspired Islamic influence in Southeast Asia from the seventh through the fourteenth centuries. Vietnamese victories forced the Cham away from coastal areas where, as Philipp Bruckmayr notes, ‘Islam was making rapid progress’, especially in the seventeenth century.²⁰ Islam also spread via familial networks: Frances Bradley argues that ‘human-to-human relationships’ among families, between ‘people and learning centres’, increased awareness of and raised Islamic religiosity in Cham communities.²¹ In 1813, the first extant mosque was built in Phnom Penh.²² Despite the complete annexation of Champa in 1832, this ‘new religion had already firmly implanted itself among the Cham’.²³

In 1955 King Sihanouk abdicated the throne and formed the Sangkum Reastr Niyum (People's Socialist Community), a political party aiming to form a Khmer nation.²⁴ Based on an understanding he put forward in the text ‘Our Buddhist Socialism’, the Sangkum would comprise ‘Khmer Islam’ (state-Islam), ‘Khmer Loeu’ (Highland Khmer), and ‘Khmer Krom’ (Lowland Khmer).²⁵ Rather than achieving the ‘pluralist Cambodian nation’ that William Collins once advocated, in which minorities are a recognised part of the ‘national consciousness’, the very notion of ‘Khmer Islam’ creates a space for a religious minority, removing the ethnic dimension of identity, but not yet religious identity.²⁶ In addition, while ‘Khmer Islam’ (Kaum Tua-oriented) became the ‘state-Islam’, this sparked a ‘Cham Islam’ (Kaum Muda-oriented) movement. This ethno-nationalist split meant that followers of Cham Islam could, as Noseworthy argues, ‘maintain piety while also keeping Cham ethno-linguistic identity’.²⁷ The latter quickly spread through Phnom Penh, Battambang, Kampong Cham, and into southern Vietnam. With no direct link to Khmer royalty, many Cham modernists especially came to be viewed as a threat to the ‘Cambodian-ness’ (Khmer-ness) of society.²⁸ One major focus of DK hostility toward the Cham was the career of General Les Kosém. A Cham Muslim, Les had the support of Sihanouk and access to arms and funds for his battalion. The DK regime exaggerated their accusations against him, and by extension the Cham community, of not only supporting the Lon Nol government, but of working with the CIA, in addition to reports that Les’ militias were massacring KR villages in what would become the Eastern Zone.²⁹

The Cham under the DK regime

During the DK regime, significant portions of the Cham population were killed either directly or indirectly through disease, malnutrition and starvation.³⁰ Ysa Osman, an expert witness at the Extraordinary Chambers in the Courts of Cambodia (ECCC, Khmer Rouge Tribunal) and a Cham himself who has worked with the Documentation Centre of Cambodia (DC-CAM), calculates the 1975 Cham population to be 700,000, one-tenth of the Cambodian population.³¹ By 1979, this had reduced to 200,000, a 71 per cent reduction, in comparison to 23 per cent for ethnic Khmers.³² The DK regime based their discrimination against the Cham on a series of stereotypes, believing them to hold ‘improper’ jobs, more common in the Cham community, such as independent fishers, money-lenders, educated religious intellectuals, and landlords, in addition to being the ‘weak link in the State’.³³ They spoke Cham, Malay, Arabic, lived independently in large villages, and had a ‘distinct culture’.³⁴ It was this ‘threat’, as Catherine Barnes argues, that underlay subsequent DK policies towards the Cham. The goal, while never achieved, was to ‘destroy the Cham as a community’.³⁵ Sos Kamry, Grand Mufti of Cambodia, claimed to have seen a document, ‘The Advanced Cooperative Plan’, which stated that there was a plan ‘to totally smash the Cham’ by 1980.³⁶ Khmer women and Cham women were forced to cut their hair short and wear only black clothing, in contrast to colourful Cham sarongs, with a red, or blue and white krama (scarf).³⁷ According to long-standing local understandings of ‘being pure’ (adat putih), and also explained by the Muk Sruh Palei (Lady who Protects the Village) and Kaboun Ong Chin (the Cham code for men), women were to keep their hair long.³⁸ Numerous transcripts describe women's experiences of being forced to commit haram (forbidden) acts, such as being forced to eat pork, to test whether or not the DK regime had successfully broken their religious convictions.³⁹ The Cham language was forbidden, sang mâgik (mosques) were closed or destroyed, hakem (religious supervisors/senior imam), imam and tuon (religious teachers) were killed.⁴⁰ This destruction of Islam included burning copies of the Qur'an and forced marriages between Muslims and non-Muslims.⁴¹ Those who refused would either be killed immediately, sent for ‘re-education’ in pagodas-turned-prisons, or be listed for marriage.⁴² Women, young and old, were subjected to rape in at least three contexts: prior to execution, as an ‘instrument of torture’ at prison sites, and through forced marriage.⁴³

Extraordinary Chambers in the Courts of Cambodia (ECCC): Case 002

The ECCC was formed in 2003 to try those who ‘were most responsible for genocide’, as well as the regime's ‘most senior leaders’, under Cambodian and international law, in addition to violations of human rights, including Gender-Based Violence (GBV), by the regime.⁴⁴ Case 002 at the ECCC concerns the mass killings of Cham Muslims, in addition to forced marriage and rape (see appendix A). Case 002 has yet to be the focus of any published detail-driven historical analyses, owing to historians to date being primarily concerned with the framework for the Case. In the Case's Closing Order, the judges concluded that: ‘It is clearly established that under the DK regime, crimes against humanity of rape were committed in diverse circumstances.’ Nevertheless, they also noted that ‘it cannot be considered that rape was one of the methods used by the regime's leaders to implement the common purpose [of genocide]’.⁴⁵ The conclusion is vital to note because in many other genocide tribunals, including in Rwanda and East Timor, rape rulings were not only discussed, but scholars subsequently concluded that rape was used intentionally as punishment and part of the purpose of the genocide.⁴⁶ This latter conclusion supports the view of Nicole Rafter that rape can be a ‘tool of genocide itself’ in genocidal contexts insofar as it is ‘designed to destroy the victim and her group’.⁴⁷

Victims were invited to come forward to the ECCC to document their experiences, but only a small selection of these were heard in the courtroom. Transcripts reveal that victim cross-examinations took a speedy ‘linear format’: often due to lack of time in court victims emphasise the harm they have suffered, give a ‘general description of routine suffering’ and finally relate it to their specific injustices.⁴⁸ The individuals being prosecuted were frequently not present at the Court, often due to health issues such as ‘back pain’ or ‘headache’, creating delays.⁴⁹ Translations from Khmer into English and French were produced simultaneously, and a transcript recorded. Because of my lack of knowledge of the Khmer language, I will refer to the English translations.⁵⁰ Even these translated sources do not regularly appear in secondary literature at present. I will prioritise the victim over the perpetrator, primarily because victims’ voices are often silenced through judicial proceedings. It is important to note that the ECCC's investigations took place in Khmer and were conducted on behalf of the Cambodian State, so there was understandable distrust within the Cham community and a reluctance to testify. No testimonies were recorded in Cham or Malay, minority languages, of which the former continues to be overlooked and even erased to this day, contributing to a ‘second wave’ of genocidal violence.⁵¹

It is important to use with caution the testimonies of some of the participants at the ECCC, whose credibility as a Court has been challenged. Expert witness Ysa Osman, himself a Cham, was accused of bias by the defence. Osman's research was allegedly ‘driven by a desire to confirm his preconceived belief that the Cham were victims of persecution and genocide’, his evidence ‘relying exclusively on testimonies that cannot be verified’.⁵² Nevertheless, many of the interviewees in his books did appear at the ECCC and did generally repeat previous statements consistently. The general work of historians on periods before the genocide is, in part, what paved the way for making people more comfortable telling their experiences.

Alexander Hinton, another expert witness, was accused of being ‘biased, unreliable, and unverifiable’.⁵³ His sources are based largely on second-hand evidence, in particular the work of Ben Kiernan. The Grand Mufti of Cambodia, Sos Kamry, was initially reluctant to testify at the ECCC, citing health reasons. He later testified, though not under religious oath.⁵⁴ Sos Kamry is the only witness of ‘The Advanced Cooperative Plan’ and was unable to give any details on the nature of the document, its author, the year he saw it, or the context in which the plan ‘to totally smash’ the Cham was written.⁵⁵ Nevertheless, Sos Kamry's perspective is particularly useful in documenting the treatment of the Cham during the DK regime.

In an ECCC statement, International Co-Prosecutor Nicholas Koumjian wrote that ‘around the world, it is a recurring phenomenon that sexual violence is grossly under-reported’.⁵⁶ Women have to break societal taboos to testify, especially because of the nature of their community moral codes, fearing that they or their families would be dishonoured.⁵⁷ Theresa de Langis, Nicole Rafter, Farina So and Judith Strasser all note that while survivors of GBV generally do not speak about their experiences ‘due to feelings of shame for themselves and their families’, it is particularly hard for women to speak out about such traumatic events and their social consequences.⁵⁸ Rafter points out that it is women who ‘may have to endure the wrenching experience of bearing children conceived during rape’.⁵⁹ Strasser gives a list of the consequences of forced pregnancy, including abortion and gynaecological issues.⁶⁰ De Langis correctly warns that in dealing with survivors’ testimonies, there is always the risk of ‘re-traumatising and/or retaliation’ for these women, who consider such accounts to be ‘shameful stories’.⁶¹ The vast number of female victims who came forward to document their experiences is therefore significant, especially given the cultural taboo of speaking out and the potential for subsequent societal imposed isolation and shame. Over 350,0000 people observed or participated at the ECCC, of which at least 50 per cent were women.

The role of GBV in genocide

Although the prevalence of gender-based violence in the DK is now widely acknowledged, the question of why it occurred is still not answered satisfactorily. There is a scholarly debate over the extent to which DK ideology was based on indigenous sources. For David Chandler, Karl Jackson, and Elizabeth Becker, the intellectual genealogy of DK regime ideology was derived from multiple sources, including Maoism and Khmer nationalism.⁶² For Tyner, the regime was closer to capitalism than communism, in that workers did not own the means of production.⁶³ Ben Kiernan has argued that DK ‘ideology’ had no parallels and should be regarded as a sui generis phenomenon. In other words, it was utterly unique, leading to the assertion that the ideological policy of the state was to implement as its primary purpose: the building of a Khmer state.⁶⁴ The Case 002 Closing Order notes that ‘groups espousing communist ideology had been in existence in Cambodia for many years beforehand’, but that the DK regime was unique for ‘re-writing its own history’.⁶⁵

Anthropological studies place a greater focus on cultural explanations for sexual violence, whilst allowing a fusion with external influences, including Maoism. For instance, Alexander Hinton and Ian Harris seek to explain GBV through both local understandings, including the Khmer moral code for women (Chbab Srei), the theme of disproportionate revenge (kum), and Maoism.⁶⁶ Cham religious communities have similar codes and practices to the Khmer, including the Muk Sruh Palei, similar to the Khmer code for women; there are potential repercussions by spirits and saints for infringing upon customary practices (adat), and I would suggest that Cham women at the very least had an idea of the Khmer cultural practices, even if not observing them themselves.⁶⁷ The similarities between some Cham and Khmer cultural practices, for example the tendency to emphasise filial piety, made them mutually entrenching, facilitating the DK regime's abuse of them. While Cham and Khmer cultural factors provide part of the context for GBV, the genocide provides the other. Genocidal actions still have to be enacted along culturally specific lines and factors. Rafter argues that genocidal rape, or ‘state rape’, differs from non-genocidal rape in that it is ‘a means used by perpetrators to achieve their aim of reconstituting a society’.⁶⁸ Put simply, GBV is a tool of genocide itself. Although Rafter's study notes that there is not yet enough evidence to suggest there was ‘genocidal rape’ during the DK regime, her arguments, with the benefit of ECCC transcripts, can be seen in fact to apply here in the case of the Cham.

GBV falls within the context of genocide under the DK because it specifically targeted the fabric of Cham Muslim communities, tearing at the fibres of what made those communities both ethnically (Cham) and religiously (Muslim) cohesive. This fits the UN definition of genocide, which places emphasis on acts committed to destroying ‘group’ identity.⁶⁹ The ECCC, in the Closing Order to Case 002, acknowledged that ‘senior leaders of the regime shared this intent to destroy, in whole or in part, the Cham group’.⁷⁰ This is important because the systemic nature of GBV was targeted firstly at the destruction of religious identity—which was in this case an aspect of ethnic identity—and, then explicitly at the ethnic identity more broadly, and finally relied upon deliberately inflicting on the group conditions of life calculated to bring about its physical destruction in whole or in part.⁷¹ Fundamentally, GBV was used as a weapon to destroy the entire community.⁷² The rape of one woman was symbolic of the defilement of the ummah-at-large. Jess Melvin describes how ‘reckless’ killings in Indonesia were designed to ‘eradicate the harmony in the community’, in a similar vein to the DK regime's desire to break up society.⁷³

The Cham are a matrilocal community. To kill one woman means destroying a string of matrilocal ties. At the same time the killing of educated Islamic teachers destroys not only the religion's framework but also its ability to function, preventing women from entering into their traditional role at the centre of a Cham household. The murder of males deprives females of their protection, while the attack and rape of females destroys the male's own sense of community and identity: gendered violence is aimed specifically at ‘destroy[ing] the psychology of both sexes’ within a wider community.⁷⁴

Gender-based crimes were systemic, but like all crimes under the DK regime, were regionally specific. Particularly in the Eastern Zone (where US bombings were the harshest) and Battambang province, where large Cham, Malay, and Cham-Malay populations resided, violence was considerably worse.⁷⁵ Caricaturing the traditional context of an arranged marriage, forced marriages were designed to disempower the parents, who during the regime were excluded from the wedding and its arrangement, and to submit the new couple to Angka, the centre of the DK regime. Both rape and forced marriage are considered in the following sections as part of GBV more generally, and I shall be arguing that their cause is not only cultural, but also contextual. All these factors need to be considered together in order to better assess the genocidal policies of the DK regime.

Motivations for GBV including rape

This section will examine the motivations for using GBV, particularly rape, considering first religion in relation to community identity, then cultural codes and the use of fear to create an ethnic Khmer state. All the evidence seems to suggest that it was ignoring both the Chbab Srei, the Khmer traditional moral code of conduct for women, and the Revolutionary Codes of Conduct (RC), regime-issued directives by which cadre were to abide, that facilitated such violence. The former helped create an attitude in which women were supposed to be inferior to men, causing particular violence in the Cham case where women are superior in certain spaces, such as within the home or matrilocal contexts. The RC's selective ignorance and general militarised chauvinism additionally contributed to sexual violence. In the cases of DK cadres who were Cham who committed GBV, ignoring the Muk Sruh Palei, in which forbidden acts (amal) are tied to customary practices (adat), law (hukum), and shari'a, forcing other Cham to commit haram may have played a role in their committing rape, in the sense of perversion of moral precepts.

Angka vs Islam: The destruction of community identity

Transcripts reveal Cambodians had very little idea what or who Angka was, but believed it had the power to control the whole of society.⁷⁶ Alongside other factors, the regime used religion to gain submission to its precepts. For the Cham population, mosques were transformed into pig pens, kitchens or prisons, and 132 masjids were destroyed to ‘distort their meaning’ and distance the community from their social reality: to destroy ethnic identity.⁷⁷ Muslims were forced to commit haram, acts absolutely forbidden in Islam; these included cooking or eating pork, tending to pigs, forced marriage tantamount to rape, sexual touching before marriage (a form of sexual assault), and marriage between Muslims and non-Muslims.⁷⁸ Entire Cham villages were razed to the ground in Koh Phal (Island of Harvest) and Svay Khleang in October 1975, the former renamed Koh Phes (Island of Ashes). Names, particularly those denoting sifats, the exclusive qualities of Allah, were shortened with the aim of causing malfeasance, forbidden in the Qur'an.⁷⁹ More generally, the Cham were persecuted based on the accusation that they were disloyal to the regime.⁸⁰

According to Article 20 of the DK Constitution, kept secret until 1977, ‘every citizen of Democratic Kampuchea has the right to worship according to any religion’.⁸¹ Having ‘freedom of religion’ was dependent on not violating the previous 19 clauses, including the integrity of the State (Article 1) and engaging in ‘dangerous activities against the State’ (Article 7). Nevertheless, Article 20 did not mean an actual choice of religions, since de facto the only approved religion was the ideological ‘religion’ of Angka, as ‘Reactionary religions which are detrimental to DK and Kampuchean people are absolutely forbidden’ (Article 15). Since Islam and Buddhism were interpreted as threats to the primary principles of Articles 1, 7 and 15, they were in effect banned, leaving a void to fill in the religious sphere.⁸² In the vision of the DK regime, this void was filled by Angka, which sought to unify Muslims, as Khmer, around itself, the secular ‘organisation’ (regime world centre), in especial opposition to Cham Islam.⁸³ The ideological construction of Angka would have had some resonance with religious audiences. Similar concepts to Angka's philosophy are found in Shafi'i Sunni Islam, which by the 1970s was the dominant school in Cambodia. While Sunni Islam has a different concept of a ‘world’ centre, the positionality of a centre exists with al-Mec (Mecca), the Kaaba and Masjid al-Haram. One of Allah's 99 names is ‘al-Basir’, the ‘all-seeing’. It is possible that some younger practitioners who left religious traditions to join the DK cadres valued these ideological similarities.

To ‘smash’ the Cham population meant more than to kill; it meant ‘to destroy knowledge itself’.⁸⁴ Religious beliefs became a weapon against the Cham through which to harm them, by destroying artefacts, copies of the Qur'an and Kitab, Cham manuscripts, books, buildings, killing Islamic teachers and intellectuals.⁸⁵ Removing Islamic teachers—the imam, katip (both charged with performing marriage ceremonies), and bilal—would be the equivalent of destroying the religion itself and has the significance of being another form of GBV by attacking the community's spiritual foundations upheld by educated males, since becoming a cleric is the fulfilment of the highest level of proper masculine behaviour in Islam. Without the imam, the religious precepts associated with shari'a cannot be interpreted; without the katip, there is no-one to sing the adhan (call to prayer), ensuring there cannot be a proper prayer on Friday in a masjid. Such removal also prevents women from fulfilling their symbolic communal roles at the centre of Cham households.⁸⁶ Sometimes children were separated from their parents so that they would have no knowledge about the way things were before the regime.⁸⁷ Religious teaching was violently obliterated in order to make way for Angka's behavioural prescriptions.

Cultural codes and models

For both Khmer and Cham, the next in importance guides to conduct, subordinate however to those of religion, are the cultural codes of the Chbab Srei/Proh and Muk Sruh Palei/Kaboun Ong Chin and shari'a, respectively. It has been argued that the existence of the codes, and their interpretation, helped facilitate the violence. Hinton is the primary exponent of the use of a cultural model to explain DK ideology. As an anthropologist he builds on Daniel Goldhagen's argument of ‘eliminationist anti-Semitism’, of a single German model being responsible for the Holocaust, inspired by pre-existing values and beliefs.⁸⁸ Hinton replaces the cultural contexts: ‘Khmer’ for ‘German’, and ‘Cambodian Genocide’ for ‘Holocaust’. Historical and ideological context was believed to morally justify the abuse of an ‘enemy’ group. Hinton claims that specific cultural traits made ‘the Khmers more conducive to genocide’ than other groups, and that it was these ‘intrinsic cultures’ (he refers to Buddhism, hardly an ‘intrinsic culture’) that were deliberately manipulated by DK leaders.⁸⁹ As such, ‘cultural phenomena’ need to be examined in themselves, rather than attempting to understand them through a psychological or social lens.⁹⁰ For Hinton,

The [regime] attempted to motivate its minions to kill by invoking ideological discourses that played upon Cambodian cultural models related to revenge [kum], power, patronage [non-Khmer subordinate to Khmer], status, face, and honour.⁹¹

This is accurate. However, to understand perpetrators’ actions, as Rafter does, we must also consider the psychological term ‘splitting’.⁹² Splitting refers to an ‘inability to see a group as partly good and partly bad’; rather the victim group is seen as having nothing in common with the perpetrator group, a common feature of genocides, where victims are reduced to merely the ‘single, simplistic identity of an enemy’, in effect dehumanised.⁹³ Rafter's perspective better reflects the relational dynamic of genocide: the empowered and the powerless. Similarly for Eve Zucker, order is achieved through creating outcasts, or as the regime described them, ‘bad elements’.⁹⁴ This is where the employment of GBV can be seen as an instrument of DK policy: the use of fear to create an ethnic Khmer state. But the violence needs to be understood within the particular context of the existing cultural moral codes governing sexual and familial relationships.

Theresa de Langis has argued that both the Chbab Srei and the RC were responsible for facilitating gender-based crimes, but not in relation to the Cham.⁹⁵ The male counterpart code, the Chbab Proh, states that ‘madness with women should be avoided’.⁹⁶ In other words, it demonstrates the male obligation to protect women. If the Chbab Srei was important in shaping actions for DK cadres against the Cham, then the Chbab Proh would have also been important, therefore eliminating the possibility that either code alone can explain GBV. At least some of the Cham knew the Chbab Srei, and therefore also its male counterpart, given its importance in Khmer society. It is therefore included here as one factor which may have influenced the subjugation of women to DK cadres.

The Cham had their own codes with which they would have been more familiar. The Muk Sruh Palei is a manuscript from the Eastern Cham community of Pāṇḍuraṅga that became popular in the Western Cham community of Cambodia in the mid to late nineteenth century. The manuscript is commonly written in Akhar Thrah (Akhar Srak in Cambodia), although there are Jawi versions as well.⁹⁷ The text provides moral and practical guidance for girls and young women with resemblances to Islamic sensibilities but few direct references to Islamic texts.⁹⁸ Much like the Chbab Srei, the rules of the Muk Sruh Palei are to be followed in order for the individual and community to live in peace and gain prosperity. In the context of the Cham matrilineal culture, a child belongs not only to a singular set of parents, but to what Po Dharma terms the ‘extended family’, or community.⁹⁹ The moral treatise occupied a ‘prestigious role’ in Cham society and, in a similar way to the Khmer Chbab Srei, demanding the submission of young women to the ummah, demonstrating Cham society as matrilocal yet patriarchal. By contrast, Khmer society tends to be patriarchal and patrilocal. According to the moral code, unmarried Cham women are expected to tend the home, engage in domestic occupations such as weaving or seamstressing, and care for children, while men ‘earn a living for the family’.¹⁰⁰ The term Adat Inâ is interpreted by Po Dharma as meaning ‘custom of mother's side’, literally explaining Cham family organisation, where descent is through maternal ancestry (the gep batian). It is mothers, not fathers, who have strong influence over their children, embedding the Muk Sruh Palei and Kaboun Ong Chin in their minds.¹⁰¹

Although the Khmer traditional codes assume male superiority, these codes alone, with their intention to maintain harmony and prosperity, clearly cannot be deemed responsible for the eruption of sexual violence that took place under the DK. Still, the notion of a synthesis of cultural codes with the RC in order to ensure Angka's total control over sexuality is worth considering. The RC states the obligation of cadres to ‘never gamble in any way’ (Code 8), ‘not to touch the people's money’ (9), in addition to ‘fight[ing] bravely against enemies’ (12).¹⁰² The lower-level, in addition to upper-level, cadres were at least casually aware of these values, if they did not know the exact wording. The Chbab Proh states, ‘at night you shouldn't be so quiet. You have to be confident with what you are doing and not have to hide it from other people’. However, the code goes on: ‘don't be careless with passion, it will drive you into unhappiness’. Depending on whether ‘passion’ and ‘being so quiet’ are interpreted as euphemisms, the Chbab Proh may have encouraged especially lower-level cadre to show off to their comrades in a chauvinistic fashion despite socialism generally rejecting such displays.¹⁰³ Indeed, men were encouraged to be chauvinistic with ‘the right to marry many wives’.¹⁰⁴ The codes both demand the submissiveness of women to men, most fundamentally in a woman's responsibility to be maternal and not to speak out. Nevertheless, according to Nancy Smith-Hefner, Cambodian males generally claim they either know little of the male code, or that they ‘don't really follow it’, whereas girls ‘cannot forget that kind of thing’, suggesting a general level of subservience in the case of the Chbab Srei. ¹⁰⁵

All the evidence seems to suggest that it was ignoring or misinterpreting the precepts of both the Chbab Srei and the RC that facilitated GBV against women. If the RC were intended to be strictly applied, rather than being merely a type of propaganda, then they would have prevented GBV against women. Code 6 of the RC explicitly states, ‘thou shalt do nothing improper respecting women’ or ‘be smashed’.¹⁰⁶ But according to the executive director of DC-CAM Youk Chhang, there were reports of at least 156 cases of rape where ‘the rapists were not punished’ by DK cadres working in co-operatives and prisons, evidence that neither Code 6 or the RC as a whole were interpreted or punished with any great strictness or severity.¹⁰⁷

Hinton states the RC were ‘an important part of the genocidal bricolage of many DK perpetrators, and effectively motivated them’.¹⁰⁸ For him, GBV was made possible by regime policy from the upper echelons. Yet to make such an intentionalist argument is to potentially miss the role of middle and lower-echelon cadres working within the regime's boundaries, or ubiquitous variations within regime policy and application. Cadres were granted significant leeway in interpreting and applying the RC. Indeed, Kosal Path and Angeliki Kanavou's study of DK notebooks revealed that while local leaders drew upon the ‘ideological indoctrination’ of the Angka Loeu (the Communist Party's inner circle), they transformed and translated this into its application at the local level.¹⁰⁹ One enlightening testimony by Him Huy, a guard at S-21, reveals this distinction: despite being ‘taught that principle since the beginning’, there was still ‘one man who raped a female prisoner’.¹¹⁰ The ‘principle’ Him Huy refers to is ‘morality with females’, which he claims was taught to all staff at the detention centre, and anyone who violated it ‘would be arrested’, though not necessarily punished.¹¹¹ While the cadre should have been punished according to the RC, he was later released. One cadre raped at least two women before killing them, forcing a subordinate to bury the bodies.¹¹² These are not isolated examples, and it is therefore difficult to maintain that ignorance of the RC and cultural codes was not a significant factor in promoting a culture of GBV. The Cham and Khmer cultural codes alongside the RC promoted a culture of implied and explicit chauvinism. Still, nominally, chauvinism is criticised by the Chbab Srei, the Chbab Proh, the Kaboun Ong Chin and Muk Sruh Palei, and the RC, and women are allegedly protected. Localised practices, and even systemic GBV, therefore, relied upon a selective ignorance of parts of existing precepts, while simultaneously accepting the general militarised chauvinism that the codes seem to have promoted. Rather than the Codes themselves being responsible for GBV, it is the way they were used in a context of general prevailing violence that is important.

The ‘Other’ and the ‘Khmer Nation’

Osman notes that the regime did not kill Cham individuals who successfully assimilated as ‘Khmer’.¹¹³ Transcripts show that those who lied, however, were rarely targeted using GBV, although they still had to break from Cham cultural tradition. This suggests that for the regime, GBV was reserved for those clinging to their non-Khmer identity. A unit chief from Kampong Cham testified that ‘Cham would report on other Cham. They wanted to be considered loyal people by the regime.’¹¹⁴ He continued, ‘if we refused to eat pork,’ a practice against halal dietary prescriptions, ‘we would risk our lives. There was plenty of fish but we were forced to eat pork since they said there was only one Khmer nation.’¹¹⁵ Significantly, he claims that ‘Cham people were referred to as enemy number one’. ‘Later on, I noticed that Cham people were taken away and killed.’¹¹⁶

One Cham woman detained at Trea Village Security Centre described how membership of the Cham group was verified: ‘A cadre asked, “What race are you?” I lied.’¹¹⁷ She continues to describe the consequences for the women who described themselves as Cham after being taken away: ‘I heard shouting from a woman begging them not to rape her’, then ‘I saw them falling into the [burial] pit.’¹¹⁸ This is an indication that the regime used sexual violence as a form of symbolic punishment, fitting with Rafter's argument of sexual violence used for group humiliation.¹¹⁹ It represents the regime's ownership of women and reinforces the power group as invincible. Another woman from Kampong Cham reported on her interrogation by DK soldiers at the same location:

They tied me up and questioned me and asked whether I was Cham or Khmer, and I said I was Khmer. After a few rounds of back and forth, they believed I was a Khmer girl.¹²⁰

For those who refused to claim they had assimilated: ‘A DK cadre walked a woman with a headscarf heading to the river front. Then a cadre held the headscarf of the woman and slashed her throat, and pushed her into the river.’¹²¹ Osman provides another testimony of this very clear ethnic and gender-based violence against a group of 200 girls in Kroch Chhmar district:

District Secretary Hor asked each one of us, ‘What is your race?’. I lied and said I was Khmer. Hor pulled me off to one wall … Those who answered that they were Cham were taken down below. They took the girls outside five at a time to a pit … stripped the girls naked, put them on a plank across the pit, and cut their throats …. until they were all gone.¹²²

The principles of kum and measured retaliation help explain how the DK regime managed to assimilate or remove the Cham quite so effectively.

Kum

Hinton's argument focuses partly on the principle of kum, literally translated as ‘disproportionate revenge’. In essence, it refers to not losing face in the short-term, most accurately described by Haing Ngor: ‘if I hit you with my fist and you wait five years and then shoot me in the back one dark night, that is kum. Cambodians know all about kum.’¹²³ We can assume that some, if not most, Cham would have been familiar with kum as it was a popular Khmer concept, while Islam has a similar principle in ‘measured retaliation’, except that it is better to forgive and amend than to retaliate, in order to gain reward from Allah.¹²⁴ Additionally, Cham and Khmer communities had lived side by side for hundreds of years, so most of the Cham population would have been well aware of the significant cultural practices of the majority of Cambodian society. The revenge principle could be used to explain GBV in two ways. Firstly, it aided the regime's use of GBV to subordinate the position of the Cham: subservience to the regime could be rationalised by the victim through the belief that whatever injustices the regime may inflict upon them, they would eventually get their revenge, as in Islam there is an expectation that individuals will receive punishments after death, such as being sent to Jahannam (hell) or the realm of al-Nar (the Fire).

Secondly, kum justifies Angka's crusade against ‘non-Base’ minorities. For Hinton, ‘class rage’ is key: for peasants, ‘those who have’, or landlords, were the ‘familiar oppressors’ who served as a sort of experiential point of reference for DK announcements about ‘class oppressors’ who charge high interest rates. To be sure, regime ideology could use kum to encourage the population to turn against the Cham population, for they, alongside urbanites, foreigners, and other minorities, posed a seemingly existential threat to the nation:¹²⁵ they allegedly were not only ‘capitalist exploiters’, but they were also not ‘real Khmer’.¹²⁶ Ultimately, the DK regime's ideology relied upon a combination of Marxist-Leninist basic socialist revolutionary (democratic centralist) ideology, Maoist interpretations of agrarianism and, local Khmer belief in the all-powerful nature of the state, and notions of social fabric, as well as interrupting some of those notions and resurrecting them under Angka. Forced marriages, a subject more widely discussed in GBV literature, provide further evidence of this.

Forced marriage

The frequency of forced marriages is the most widely acknowledged form of GBV in the existing scholarly literature on the genocide that took place under the DK regime.¹²⁷ Between 1975 and 1979 at least 250,000 women aged between 15 and 35 were forced to marry, though generally not before the age of 20 (25 for men).¹²⁸ Marriages were dictated and arranged by Angka. Part of the reason the DK regime was able to enact its forced marriage policy so widely was that arranged marriages were already standard practice in Cambodia, including in Cham communities. The regime's modus operandi was to take a pre-existing cultural-cum-religious setting, the key one of arranged marriage, and then intentionally disrupt its established order with the power of the state, under Angka, by taking aspects of the existing order to a radical extension of their logical ends, in the form of forced marriage.

Traditionally, the practice of Cham marriage is a means of continuing family lineages in accordance with Islamic practices and the customary law of the Cham community. Older women, as part of the role of the Muk Sruh Palei, play a prominent role in arranging the ceremony in addition to ‘confirming the successful consummation’.¹²⁹ The wedding typically lasts several days, involves the community/ummah-at-large with time for the bride and groom to decorate their house, prepare for family life together, and have the wedding itself in the bride's home. In traditional Cham marriage structures, the parents of both bride and groom are central. The parents of the girl make marriage overtures to a boy, before asking him if he accepts their daughter in marriage, with the expectation that he will say yes. During the ceremony an imam, or possibly katip, reads from the Qur'an and prays for the happiness and health of the couple. Marriage was the prelude for a young woman's entry into motherhood and adulthood. Cham marriages are mainly matrilocal, and consummation is monitored by the eldest female (muk). After the ceremony, there is a feast and the groom is taken to his wife's family's house, where he will now also live, having brought pots, pans and dishes, generally for communal use.¹³⁰ When couples move out from the parental home, they live neolocally, in the wife's parents’ house.

DK marriages by contrast were sombre affairs, and whilst retaining the traditional weight placed on consummation, this was monitored very differently. A line of women and a line of men would stand in pairs in front of the flag of Democratic Kampuchea. A local cadre would say a few lines about the couple's responsibilities to Angka and the nation in terms of reproduction. For the ‘fortunate’, small gifts of tobacco or fruit would be exchanged, before any invited guests—generally the family were not invited—would be given a bowl of rice.¹³¹ Anywhere between two and one hundred couples could be married at one ceremony, usually in public locations or pagodas, reverted from prisons for the occasion. Often an individual was given notice that they were to be married less than one day before the ceremony, with the person to whom they were to be ‘coupled’ only revealed on the actual day; the impact of the suddenness and surprise created tension and most importantly, uncertainty. Where traditionally a couple would have time to make preparations for the ceremony, both mentally and in practical and social terms, now instead a blind marriage had been thrust upon them, the manner of its announcement no different from that of summons for a crime. If Angka ordered a marriage, ‘you must marry’.¹³²

Traditionally, the role of the Cham family was subordinate to the community, the ummah and the palei, and a family's obligation was to the village cause rather than the family alone.¹³³ Nonetheless, in the past, women may well have been able to voice an opinion regarding potential suitors, within the protected context of the family, for marriages which would have been arranged within the private sphere and conducted in accordance with both shari'a and adat Cam. In the DK, the radical interjection of state authority to project marriage into the public sphere removed this ability as well as removing the ceremony from the religious sphere. In previous generations, marriages between Khmer and Cham took place, but anyone marrying a Cham woman or man would have converted to Islam.¹³⁴ Under the DK, the young couple would no longer live with the bride's parents and the traditional matrilocal Cham family organisation was disrupted. As Nagakawa Kasumi describes, there was a strong pro-natalist element to DK policy, but above all these state marriages were about the destruction of religious identity and ‘the achievement of revolution’.¹³⁵ Forcing marriages without the customary conversion was a key feature of a programme to destroy the religion. One Cham man from Kampong Cham reported of his marriage in 1977 that he was forced to marry a Khmer woman in a pagoda, a haram religious space, then eat pork, a haram food, after the ceremony: ‘We were advised to become a Khmer nation. There was another Cham person who was paired up with a Khmer girl.’¹³⁶ As forced marriage was a ‘central policy’ of the regime, GBV became a ‘core element of the DK modus operandi’.¹³⁷ A Cham phrase states that one (a man) cannot have a ‘life’ (hadiep) without a ‘wife’ (hadiep), so because all life stems from the relationship between women and men, when the DK regime asserted control over marriages, they de facto asserted pseudo-religious control over all life for the Cham.¹³⁸ Forced marriage, therefore, represents the ultimate appropriation of parental and kinship rites by the state.

Two key factors underlying this type of GBV were the element of violent surprise in the summons to marriage, and the humiliating way in which consummation was enforced and monitored. The fear instilled by the way in which individuals were summoned to be married is evident in testimonies. Women did not know who, when, or where they would marry, only that they would have to marry and consummate or feign consummation, in conditions of violence, whether physical or emotional. One man from Kampong Cham province reported that in 1977:

I was called to go with district cadres, and they told me that at this hour, this day, that I had to go with them in the cooperative. I was thinking ‘What did I do wrong?’. I learnt that it was an arranged marriage ceremony. Nobody knew who would be his or her future husband or wife.¹³⁹

It is first important to note the use of the term ‘arranged marriage’, which hints that the DK regime was intentionally disrupting and manipulating a pre-existing cultural practice. What is also notable is not only the surprise of this man at the prospect of the marriage ceremony itself, but his fear after being called to attend. Clearly he did not expect to be married, but to be punished for an unknown crime. In the event the punishment was marriage, rather than death; but clearly severe consequences for some unknown crime were not unexpected. We do not know what his fear specifically related to, although we know of the DK's hostility towards extra-marital relations and there is evidence that if individuals committed ‘moral offences’ they would often be married, so this was evidently considered a form of punishment.¹⁴⁰ The other main means of attacking an individual's ethnic identity via marriage was the way in which consummation was monitored after the ceremony.

In any new marriage, even before the DK regime, consummation was expected to be ‘proven’, but the DK's policing of consummation laid marriage rites open to creating a state-sanctioned culture of GBV. Under the DK regime consummation was overseen by child spies (Khmer: chhlop), thus disrupting the religious order in Cham Muslim communities, as well as the communal order of authority. A couple would be provided with temporary accommodation for two to three days, before being separated. During this time they would be spied on overnight by chhlop, who would lie beneath the floorboards and ‘monitor whether they consummated the marriage’.¹⁴¹ In almost all cases, this led to enforced sex or rape, motivated neither by sexual attraction nor by a sense of community but by mutual fear of reprisal.

Those who refused to marry would be classed as ‘moral enemies’.¹⁴² In the Court a local chief was asked, ‘If anybody did not want to marry, then the person wouldn't have to marry?’, and testified with a simplistic ‘Yes’.¹⁴³ However, there is no evidence in ECCC transcripts of any Cham women or men who refused to marry surviving, although two transcripts detail Khmer individuals who survived after refusing a forced marriage.¹⁴⁴ Even if people did not actually have to marry, their fear is testament to the control the DK regime could exert over its citizens by this means. There is evidence that at least Khmer people falsely claimed they were agreeing to marriage:

The unit chief came to ask me if I got married voluntarily. I replied [that] I got married voluntarily. Actually, I was just trying to provide the answer upon their request. I did not get married voluntarily, but I had to agree to their request.¹⁴⁵

Therefore, forced marriages were to serve a greater ideological purpose for the regime: the continuation of a process that had begun in the 1950s to forge ‘the Khmer nation’, requiring either assimilation of the individual into Angka's notion of the true Khmer, or their removal. As the Case 002 judgement, focusing on the plight of the Cham, concludes, although some individuals may have been consulted on their marriage, allowed to select or refuse potential partners, ‘the general practice was that individuals had no choice as to whether they would marry’.¹⁴⁶ State control of marriage was thus being used to destroy community and personal identity. In the case of the Cham, this meant they were no longer to be governed by the qadi (Islamic judge). The Angka's intention was to attempt to break the foundations of shari'a and adat Cam in Cham communities, by destroying the performance of communal rituals and those who had the authority to conduct them. Since Cambodian independence in 1953 and Sihanouk's ministerial reign from 1955, many Cham had been trained to say that they were ‘Khmer Islam’ when speaking to state authorities. The genocide of the Cham, therefore, was rooted not only in the DK regime, but in post-colonial Cambodian society as a whole, supported by state policies in the 1950s and 1960s.¹⁴⁷

Conclusion: Beyond cultural factors

Cultural factors and contexts must be combined with an understanding of GBV to better assess the genocidal policies of the DK regime. According to Nicole Rafter's generic study, sexual violence was used as a tool by genocidal regimes to destroy the victim group; it is ‘a means used by perpetrators to achieve their aim of reconstituting a society’.¹⁴⁸ GBV as a feature of genocide demonstrates the power of the regime to control and humiliate. Those who refused to assimilate would, along with their culture, be removed. This, evidence has shown, was the case with the Cham under the DK. The Khmer Rouge regime sought to destroy Cham ethnic and religious identity and to create a new non-religious identity around Angka.¹⁴⁹ In many cases, women were symbolically raped before their death, or were raped in prisons before being returned to their worksites. In such situations, rape is not merely about punishing the individual, but about instilling terror in the victim's community.¹⁵⁰ The combination of sexual torture and killing destroyed what sense of ummah remained. Consequently, under pain of death, Cham women were willing to deny their identity—to declare themselves ‘Khmer’ and eat pork.¹⁵¹ The rape of one woman would become symbolic of the defilement of the community-at-large, destroying a string of matrilocal ties. Killing educated Islamic teachers would effectively destroy the religion, preventing the fulfilment of women's symbolic communal roles at the centre of Cham households. The system of forced marriage negated the close role played by parents of the bride and groom in Cham marriage and, along with separating children, ensured the destruction of the family.

GBV was thus used as a tool to instil fear, to force physical, even if not psychological, unity among a ‘Khmer’ population. It is of course important to remember that those testifying that the regime aimed to destroy the Cham, such as Sos Kamry, are themselves members of that community. Likewise with DK cadre testifying the reverse: as one deputy secretary put it, ‘If there was actually a purge planned from the Centre to kill the Cham people, then they would all have been killed.’¹⁵² I assert that the primary intention of the regime was, at least after 1977, to remove at least the Cham that had not fully assimilated as Khmer. As stated in the conclusion to Case 002: ‘orders to purge the Cham [in at least the Northern Zone] came from the upper echelon’.¹⁵³ When people are viewed not as human beings but merely as mechanistic tools of the state, a converted worker may be allowed to live and serve the state's purposes. There is no need for an explicit policy of total extermination on the lines of the Nazi ‘final solution’. Purgation of a community, whether complete or incomplete, explicit or implicit, is a genocidal purpose.

As has been shown throughout this article, scholars have frequently attributed the DK regime's use of GBV to local cultural contexts, including Cham and Khmer cultural codes, religion, and kum, the principle of vengeful justice. There is some localised support for these interpretations. No doubt Khmer awareness of the Chbab Srei encouraged the assumption among Khmer men that women ought to be subservient and this contributed to their silence in the face of abuse. Yet to consider this a complete explanation for the GBV, without taking into account the genocidal context, risks making the incorrect generalisation, as Hinton does, that Cambodians are inherently violent. Moreover one cultural code cannot be responsible for systemic rape, even if connected to other cultural factors, and this interpretation overlooks for example, the male obligation to protect women mentioned in the Chbab Proh. Rather, the DK regime was faced with a task of post-war reconstruction in which gendered violence was ‘part of the project of forming a new, ethnically homogenous nation’, as Rafter describes.¹⁵⁴ Angka intended to regulate all sexual relations.¹⁵⁵ While Rafter did not believe there was evidence that the DK regime ‘systematically’ used rape as a tool of the regime, there is evidence that this very violence was used with the intent to destroy not only singular victims, but collective Cham identity, both actually and symbolically, and to replace it with a unified, ethnically homogenous, ‘Khmer’ population subjugated to Angka. In the case of the treatment of the Cham, therefore, rape cannot solely be explained by cultural factors, but can only be accurately described in conjunction within its genocidal context: the intention to subjugate ‘the other’ to the regime, Angka.