A Theoretical Model of Victimization,Perpetration,and Denial in Mass Atrocities: Case Studies From Indonesia,Cambodia,East Timor,and Myanmar

Putting Atrocities in Societal Contexts

TOPASC (Figure 1, left box “Macro level: Global, Regional, and Local Contexts”) prompts scholars to consider political, economic, historical, and sociological risk and protective factors that are relevant and in particular identifies nine macro-forces for consideration (loosely grouped under disciplinary headings, although of course they are interrelated). Each scholar seeking to apply the model would need to engage what is known about these forces from the interdisciplinary literature for their context. This is important to understand the process of how and why the mass atrocities occur.

The present model locates the atrocities that we consider within four post-colonial countries in the Southeast Asian region. While the primary objective of this paper is to elucidate that the emergence of mass atrocities is significantly influenced by societal context, the four cases are also chosen both to remedy past scholarly neglect and to draw out the implications of past colonization, which have been neglected in social psychological research (Hendricks & Moghaddam, 2020). Cases of genocide and related mass atrocities from Asia have been underrepresented within the broader comparative analyses (Mayersen & Pohlman, 2013), with the East Timorese, Indonesian, and Myanmar cases particularly neglected in the field (Bloxham & Moses, 2010). In addition to their geographical proximity, however, these four cases from Southeast Asia share key features that warrant comparative analysis. In each case, as we explain further, the mass atrocities were at least in part shaped by the violent decolonization processes of the mid-20^th century, and there were similar psychological processes at play in how victims were conceptualized and targeted, how people became involved in violence, and how memories of these genocides have been subsequently denied. We elaborate further on their socio-historical contexts in the “historical overview” section and emphasize which aspects are significant at the macro level and what needs to be considered at the meso level in the study of mass atrocities.

Micro-Level Outcomes of Mass Atrocities

In this article, we can only begin to address the impact and antecedents of atrocities at the individual level (Figure 1, right box “micro—individual or subgroup processes”), but these are comprehensively addressed by a range of genocide scholars (e.g., Baum, 2008; Minow, 1999; Waller, 2007). For victims, of course, outcomes of atrocities can include death, physical injuries, and mental illness, as well as changed political attitudes, such as mistrust, processes such as radicalization, and (in the event of reconciliation) compassion or forgiveness (Bilali & Vollhardt, 2019; Mukaman & Collins, 2006). For perpetrators, atrocities can give rise to psychological dissonance and moral injury, alienation from relationships and mainstream identities, and lasting psychological ill-being (e.g., Schwab, 2010; Waller, 2007). In other cases, no lasting effects are observed, however, and understanding the factors at play is an important potential contribution of the present theoretical model as well as an important direction of future empirical research (Louis et al., 2015, 2018). As elaborated subsequently (see also, Baum, 2008), we distinguish between the active perpetrators of the atrocities and their constituents: passive actors’ roles differ greatly based on whether they are kept carefully unaware of the atrocity (in which case they may profit from it more readily) or whether they are aware, in which case they may experience ill effects such as distress, polarization, and political instability, seeking to legitimize or criticize what is occurring.

Theorizing the Intersectionality of Mass Atrocities

The present article seeks to make a contribution to the literature on the psychology of mass atrocities by theorizing the construction of victims and perpetrators beyond the dyadic level. ³ It is common to speak of mass atrocities in dyadic terms: in the Holocaust, Nazis murdered Jews; in Rwanda, Hutus murdered Tutsis; and in the United States, White Americans (Europeans) murdered Indigenous peoples or First Nations. Such descriptions are quite true on one level: they have an elegant simplicity and focus. However, they can be criticized on a number of grounds, such as neglecting the agency and, in some cases, the reciprocal violence of targeted groups (Baum, 2008; Waller, 2016).

For the purposes of this article, the primary critique we advance is that such approaches neglect important intragroup differences that are of great interest in terms of understanding and preventing perpetration (see also, Himawan et al., 2022; Perlman et al., 2015). Only a minority of actors, whom we will call perpetrators in this article, mobilize to commit violence personally themselves. Within a nation that is responsible for atrocities, most active perpetrators are men, for example, and particular military or police units may become specialists in carrying out the atrocities of their group (Baum, 2008; Waller, 2007, 2016). Other members of the group, whom we will call constituents, may passively ignore and profit from the atrocity, actively facilitate it, or contest it. Turning to victims, particular ethnic groups are often selectively targeted within a broader disadvantaged community; these may or may not include women and children; where they do, the type of atrocity may change: for example, sexual and gender-based violence such as rape may become more common. As we shall see in the section on victimization, the choice of victims to include or exclude non-combatants and children is an important dimension along which atrocities vary, demanding theorizing and research. Furthermore, the denial or acknowledgment of atrocities is often selective: Within a particular society, the violence may be an open secret in some circles (e.g., among some military and political elites) but kept out of the media through censorship and intimidation (Cheterian, 2018). Our contention is that some social circles are carefully constructed to be ignorant, or to have plausible deniability, to allow a culture of impunity to persist for atrocities that violate the wider social norms of the group.

In seeking to understand the formation of these subgroups and subcultures and their role in atrocity perpetuation, we borrow from the scholarship of intersectionality (Case et al., 2012; Cole, 2009; Collins & Bilge, 2020; Crenshaw, 1989; Rosenthal, 2016). The concept of intersectionality was originally put forward by Black feminist scholars and advocates to make the point that experiences of privilege and disadvantage for group memberships such as gender and race are not additive. Instead, social categories such as manhood and womanhood, or Whiteness and Blackness, depend upon each other for meaning. In examining outcomes from discrimination, interaction effects mean that intersecting identities of disadvantage may be disproportionately disadvantaged or advantage disproportionately advantaged. Furthermore, qualitative differences in the character of privilege or disadvantage may also be present. For example, Black lesbian women may have a different experience of all three of their identities by virtue of their intersectional identities. Their womanhood may be contested by others because of their lesbianism, for example, while they may be marginalized within feminist and queer spaces because of racism.

The original approach of many scholars of intersectionality was to explore the unique experiences of intersectional groups with case studies and qualitative research, often published in feminist and women’s studies journals and books. However, intersectionality has the potential to change quantitative and mainstream psychology as well, and we will argue that it is relevant here to any understanding of mass atrocities. The three tenets of intersectional research (Buchanan & Wiklund, 2021) are: First, that structures of oppression and inequality (such as sexism, racism, ableism, and heterosexism) co-create each other; Second, that experiences of privilege and disadvantage vary depending upon these intersecting dimensions so that experiences of individual dimensions should not be understood in isolation; and third, that scholarship has a social justice element: theory and praxis working in tandem so that a scholarship of mass atrocities is part of either a liberation struggle or a system-justifying one. In the present analysis, we use an intersectional lens to understand intragroup differences in the definitions and lived experiences among victims, perpetrators, and constituents. In the next section, we elaborate on our proposed model using examples from mass atrocities in Southeast Asia. We begin by giving a brief overview of the historical contexts.

The Indonesian Context

In Indonesia, large-scale massacres of hundreds of thousands of accused communists were carried out by the Indonesian military with co-opted civilian militias after an alleged failed coup attempt on 30 September 1965, which was blamed on the Indonesian Communist Party (Partai Komunis Indonesia/PKI ⁴ ). The coup itself was a short-lived and poorly-planned attempt by a group of Leftist middle-ranking soldiers to get rid of the Army’s right-leaning leadership in Jakarta by kidnapping six generals. Five top generals were assassinated in this coup attempt but the September 30^th Movement, as they named themselves, was suppressed by Army special forces within a few days (Roosa, 2006).

In the years prior to this coup, Indonesia experienced a range of risk factors for genocide, including a divided society, political upheaval, and worsening economic conditions (Ingelaere et al., 2013; Nyseth Brehm, 2017; Ogunyemi, 2020). After declaring independence from the Dutch at the end of the Second World War in 1945, Indonesia had become sharply polarized along socio-political lines, with Communist, Islamist, and “developmentalist” groups competing for dominance. The developmentalist group, who were essentially market-oriented capitalists, became strongly associated with the military and right-wing political groups. These groups had emerged some decades earlier amid growing nationalist resistance to Dutch colonialism and were, at heart, competing visions for an independent Indonesian polity (Cribb, 2001; Melvin & Pohlman, 2018). These socio-political cleavages sharpened by the mid-1960s, particularly between the Army and the Communists, as Indonesia’s economy declined rapidly, with hyperinflation and spiraling food prices threatening widespread famine (Cribb, 2001; McGregor et al., 2018).

Under the leadership of Major General Suharto, after the September 30 coup attempt, a genocidal campaign was launched by the Indonesian Army against PKI members and supporters (Cribb, 2001; Melvin, 2018; Roosa, 2006). The Army carried out a mass propaganda campaign that demonized Communists and incited civilian participation from anti-Communist religious and other groups, leading to the killings of an estimated 500,000 people between October 1965 and March 1966. Furthermore, hundreds of thousands were detained without trial in the years after 1965, where many died from torture, starvation, and forced labor (Cribb, 2001; McGregor et al., 2018). Sexual and gender-based violence was also pervasive during the massacres and in the detention camps (Pohlman, 2015). The Indonesian military received financial, logistical, and other forms of support to eradicate the “Communist threat” from the United States, Australia, the United Kingdom, and other Western governments (Simpson, 2008).

This period of mass atrocities, led by the Army, also initiated a three-decade period of military rule called the “New Order.” Despite the end of the military regime in 1998 and democratization, the developmentalist (military-led oligarchy) and Islamist groups in Indonesia remain deeply committed to anti-Communist narratives and still teach that Communism is a threat to Indonesian values (Adam, 2018; Putra, Holtz, Pitaloka, et al., 2018, 2019).

The Cambodian Context

The genocide in Cambodia happened from 1975 to 1979, during the Khmer Rouge “Democratic Kampuchea” (DK) rule under the leadership of Saloth Sar, known as Pol Pot (Kiernan, 1996). Prior to the Khmer Rouge seizing power, Cambodia also underwent a period of significant political upheaval tied to the violent decolonization of the former French colonial territories of Indochina. Cambodia declared its independence initially at the end of WWII and the start of the first Indochina war (1946-1954). Under King Norodom Sihanouk, the Kingdom of Cambodia saw bitter divisions emerge between right-wing/anti-Communist groups and left-wing/Communist groups that supported resistance to the French. Although Sihanouk attempted to maintain Cambodia’s neutrality, the second Indochina war (the American War, 1955-1975) spilled over the border from Vietnam, eventually leading to Sihanouk’s overthrow in a coup in 1970. From 1970 to 1975, the American-backed Republic changed leaders multiple times as Cambodia became more deeply embroiled in the conflict between the north Vietnamese Communist forces, who had crossed into Cambodian territory, and joint United States–South Vietnamese forces who invaded to drive the Viet Cong farther west into Cambodia (Chandler, 2008). The former king built connections with the Communist Party of Kampuchea (CPK), popularly known as the Khmer Rouge, to fight the new Republican government. The Republican government had the backing of the United States (Shawcross, 1986) and was blamed for the U.S. bombings along the border of Cambodia and Vietnam between 1969 and 1973 which killed more than 70,000 civilians and drove waves of Cambodians out of agricultural areas into the cities (Edwards, 2004; Shawcross, 1986). With strong popular support, the Khmer Rouge seized control of Phnom Penh in April 1975 and, over the following 6 months, forced the population out of the cities and towns into rural areas (Chandler, 2008; Tyner et al., 2014).

Inspired by communist ideology and the Chinese revolution, the mission of the Khmer Rouge regime was to turn Cambodia into an idealized peasant society (Edwards, 2004; Kiernan, 1996; Maguire, 2005). In the quest to return to “Year Zero,” the hypothetical time when Cambodia started without a class division, the regime embarked on a radical and genocidal social engineering project in which the entire population was forced into agricultural labor, and the country’s social, economic, and political infrastructure destroyed to remake the nation (Kiernan, 1996). Sexual and gender-based violence was also widespread, as families and communities were separated and forced into new agricultural communal working groups, and forced marriages were imposed by the regime (LeVine, 2010). Perceived enemies, ranging from those who had lived in the cities, to Buddhist monks, to those deemed ethnically non-Khmer, particularly the Cham Muslim and Vietnamese populations, were considered impure and frequently targeted for torture and murder (Hinton, 2005; Kiernan, 1996). The regime set up a country-wide system of interrogation centers and prisons, where these enemies, and anyone thought to have opposed the regime, were tortured and often killed (Chandler, 1999). In 1977–1978, the DK regime instigated attacks on neighboring Laos, Thailand, and Vietnam, which led to its demise when Vietnamese forces invaded in December 1978. An estimated 1.7 million people died through massacres, forced labor, and starvation: approximately one-fifth of Cambodia’s pre-1975 population (Kiernan, 1996, pp. 456–460).

The East Timorese Context

Mass atrocities in East Timor began at the end of 1975 when Indonesian forces invaded to annex the territory. East Timor had been a Portuguese colony from the mid-1700s but the Carnation Revolution in Portugal in 1974 led to the rapid decolonization of that country’s overseas colonies. In East Timor, the withdrawal of the Portuguese precipitated a brief civil war in 1974-1975, in which the leftist Frente Revolucionária de Timor-Leste Independente (Revolutionary Front for an Independent East Timor/Fretilin) defeated pro-Indonesian oppositional forces and declared independence. A month later, supported by the United States, Australia, and a number of other Western and regional governments, Indonesia invaded, the rationale being that Fretilin was communist and the annexation was necessary to prevent the “fall” of another Southeast Asian nation to communism so soon after the American defeat in Vietnam (Fernandes, 2015; Simpson, 2005).

The occupation by the Indonesian military, in collaboration with pro-Indonesian Timorese militias, led to violent clashes, famine, and other widespread human rights violations, including sexual and gender-based violence (Loney, 2018). The initial invasion period lasted until 1979, after which Indonesia declared the territory “pacified,” but East Timorese resistance led by Fretilin continued in the face of ongoing terror campaigns against the civilian population by the Indonesian military (Comissão de Acolhimento, Verdade e Reconciliação [CAVR], 2005). Throughout the occupation, the Indonesian military government received arms and other support from a range of Western governments (Job, 2018; Tanter et al., 2001). Between 100,000 and 200,000 excess mortality deaths are estimated from 1975 to 1999 or around a quarter of the population (Kiernan, 2003; Silva & Ball, 2006; Staveteig, 2007). Following the end of the New Order regime in Indonesia, the people of East Timor voted for independence in 1999 amid widespread violence by Indonesia-backed militias, finally gaining their independence in 2002 (Fernandes, 2021; Robinson, 2009).

The Myanmar Context

In Myanmar, the accusation of genocide refers to the most recent series of violent outbreaks in 2016-2017 and afterward, associated with mass killings and the violent displacement of Rohingya Muslims (see e.g., Crisis Group, 2017a; Human Rights Watch, 2020). Myanmar was colonized by the British in the 19^th century after three wars and gained independence in 1948. Military rule from 1962 to 2010 and again since 2021 has both been justified by and given rise to long-running insurgencies against the central government. As we examine in the following section, the Rohingya Muslims have been targeted as a specific ethnic and religious community over many decades in post-colonial Burma, ⁵ but our focus here is on the most recent series of incidents from 2016 against Rohingyas, in which thousands of civilians died and approximately 1 million people fled their homes (UNHCR, 2022).

Violence against Rohingyas had been on the rise during the 2010s but escalated during military operations in 2016 badged as bringing stability to the region (Crisis Group, 2019; Hussam, 2019). Many Rohingyas were driven to flee to the neighboring country Bangladesh, where more than a million Rohingya Muslims now live as refugees. Military forces, in collaboration with paramilitaries and ultra-Buddhist-nationalists, carried out systematic attacks against Rohingya villages (Human Rights Council, 2019). An estimated 24,000 Rohingya Muslims were killed between 2017 and 2018 (Kartal, 2018), and 10,000 women were raped and sexually abused (Anwary, 2020), with many killed after they were raped. Hundreds of thousands of Rohingya Muslims were also tortured, and over a hundred thousand Rohingya houses were burned (Kartal, 2018). Consistent with the mass atrocities described earlier, children and women were among the victims (Lone et al., 2017).

Societal Context Analysis

In each of the cases we have identified, there are macro-level societal factors at play (see Figure 1, left box). Politically, we can ask about the construction of the nation-state in the context of Southeast Asia, and its alliances and enmities. Colonial actors such as England, France, and the Netherlands established states without respecting the historical boundaries of the local peoples’ political entities (Englebert, 2000; Young & Brown, 1995). As a result, post-colonial nation-states forced ethnic and cultural groups that are disparate, and even inimical to each other, to contest sovereignty as one actor rather than living as independent polities as they had done historically.

Post-colonial states were also drawn into relationships with historical rulers’ great power struggles, such as the Cold War: the rivalry between Western Europe and the United States with Communist Russia and China (e.g., Henry, 2016; Woods, 2018). These struggles became justifications for atrocities, as we shall see below, and the atrocities were facilitated by international arms trades and military subsidies to murderous regimes, which were also legitimized by great power struggles (Craft & Smaldone, 2002).

In domestic political economies, unequal privileging of certain regions, ethnic groups, or religious groups by colonial administrators, for example, due to racism or religious beliefs, sets the stage for lasting cleavages, grievances, and insecurities which also created volatility and conflict (Bettache, 2020; Englebert, 2000; Young & Brown, 1995). The impacts on international and national political systems are mirrored by other impacts on economies, histories, and social structures. In particular, we can consider the presence of weapons as well as concrete material and symbolic destabilizing forces.

In the countries considered here, traditional economies and trade networks were distorted to force colonies into export-import systems with host empire countries (e.g., Amin-Khan, 2012; Anderton & Carter, 2019). These export-oriented economies were more volatile and often neglectful of food security, increasing the prevalence of price shocks and famines that fuel social unrest. The socialization of leaders and elites into European values and the transformation of education systems, away from historical cultures, faiths, and kinship systems, destabilizes internal peace-making practices and alliances, introduces new cleavages, and exacerbates the risk of internal exploitation and authoritarianism, further giving rise to volatility and conflict (e.g., Bettache, 2020; David & Okazaki, 2006; Englebert, 2000).

At a collective level, violent conflict within living memory from both colonial policies (e.g., repression and massacres) and decolonization (e.g., civil wars) gives rise to destabilizing forces (Manz, 2008), including narratives of historical threat and grievance (e.g., Bentley, 2015; Ogunyemi, 2020). On an individual level, legacies of violence may give rise to trauma and increased risk of perpetrating future violence (e.g., Schwab, 2010; see also micro-outcomes). Colonial practices may introduce or reinforce both internalized racism or ethnic hostilities (Bettache, 2020; David & Okazaki, 2006), and gendered constructions of violence and domination that construe murderous violence as a legitimate defense of own group security and mass rape as a legitimate part of intergroup conflict (e.g., Mukaman & Collins, 2006; Sanford, 2008). We touch on these themes in the discussion of the group processes (i.e., meso level) below, and otherwise, we refer readers interested in deeper consideration of societal forces to other accounts of the impact of colonization and decolonization drawn from genocide studies and history, politics, and economics (e.g., S. E. Brown & Smith, 2022; Moses & Stone, 2007; Sanford, 2008).

How Victims Are Chosen

As a first psychological process, we examine in the four mass killing contexts how and why the victims were targeted and demonized. We first review psychological theories of victimization, and then analyze each context in detail, before revisiting the theoretical implications of our case study analyses.

In one detailed model, Reicher and colleagues identify five steps in how inhumane acts against other groups are construed as morally right (Reicher et al., 2008; see also, Amiot et al., 2020): (Step 1) Identification of the ingroup and limitations of membership; (Step 2) Exclusion of the outgroup, those who are different; (Step 3) Positioning the outgroup as threatening and endangering ingroup identity; (Step 4) Representing the ingroup as morally righteous; and finally (Step 5) Celebration, in which the destruction of the outgroup is positioned as essential for the defense of virtue. The genocide of Jewish people during World War 2 can be taken as an example. Initially, Jews were considered different and then excluded from part of the German identity. Following that process, they were then argued to be a threat, particularly to German values and virtues, in response to which genocidal extermination policies were then justified (Reicher et al., 2008). It must be noted, however, that the order of the steps is understood to be flexible (Pettersson, 2019).

In collective violence, we propose that the process of deploying collective blame is central to the construction of threat. In collective blame, all members of an outgroup are deemed to be responsible for the actions of individual group members (Lickel et al., 2006). For example, terrorist attacks carried out by Muslim extremists in the United States and other Western Countries were followed by hate crimes targeting Muslims, because some blamed Muslims collectively as responsible for the terrorist attacks (Bruneau et al., 2017). Often this narrative of collective blame and threat rests on social identity processes (Reicher et al., 2008; Tajfel, 1982). Individuals’ social identities define them not just as individuals (“I”) but also as members of in-groups (“we”), which are generally favored over other groups to which the individuals do not belong (“outgroups”). In some cases, out-group members are perceived as homogeneous and interchangeable (Medin & Ortony, 1989). Such a mechanism is called psychological essentialism (Medin & Ortony, 1989). Often, a group characteristic that is essentialized is endowed with the same immutable attributes and entity (Wagner et al., 2009), so that all group members are considered to be the same. Unsurprisingly, essentialism and collective blame of the out-group both reinforce and reflect intergroup prejudice and outgroup dehumanization (Bandura, 2002, 1999; Bruneau et al., 2017). A process of mutual radicalization and retaliation occurs, as groups increasingly harm those who were not directly involved in the initial tensions (Louis et al., 2020; Moghaddam, 2018).

Of course, such processes are not inevitably unleashed (Reicher et al., 2008; Tajfel, 1982); they depend upon the intergroup context. Specific reasons why groups are excluded or believed to be a threat can also vary in different cases of mass atrocities. In any one case, the victims can be demonized and deemed a source of evil (Reicher et al., 2008; Vollhardt & Campbell-Obaid, 2016) or dehumanized, a process by which humans are associated with animals and considered as less than human (Bandura, 1999, 2002; Halperin & Pliskin, 2015; Owens et al., 2013; Waller, 2016), or perceived as both evil and less-than-human. In the Bosnian genocide, for example, historical narratives of previous conflicts in which Bosnian Muslims killed Serbians were used to rationalize anticipating future murderous attacks. This “kill or be killed” narrative facilitated the mobilization of perpetrators to take part in violence against Bosnian Muslims (Winton & Unlu, 2008). The Rwanda genocide can be seen as a clear example of dehumanization, where Tutsis were portrayed as cockroaches by Hutu Power extremists (Baum, 2008; Ndahiro, 2014).

In the present approach, out-group derogation and violence toward out-group members are made more or less likely by ingroup leadership, which in turn both responds to and constructs group members’ beliefs about the intergroup relationship as threatening (illegitimate, unstable, and eroding the positive distinctiveness of the group) or reassuring (legitimate, stable, and securely positively distinct). This fluid responsiveness to context is aligned with the TOPASC model. While some scholars of social identity have at times glibly used a decontextualized language that implies that social identities inevitably produce favoritism and discrimination, the theory is quite clear that when an intergroup hierarchy is legitimated and stable, out-group favoritism by disadvantaged groups may develop along with other system-justifying tendencies (Hornsey, 2008; Jost, 2019; Jost et al., 2015). System-justifying beliefs function to preserve the status quo, for example, if the out-group is understood to serve as a positive model for the disadvantaged group to emulate. In other words, social identity processes, including the confirmation of who is part of the group, who the allies are, and who the enemies are, rely on the cultural context (Hopkins & Reicher, 2011).

The social identity processes identified as arising out of conflictual, unstable contexts, and the dynamics of collective blame and mutual radicalization, are captured in what we will present below. We analyze the patterns of victimization in the four case studies, before briefly considering novel theoretical implications that emerge from the data. As part of our attempt to bring an intersectional lens to the understanding of how and why victims are chosen, we will seek to explore the specific contingencies of disadvantage that define some victim groups as legitimate targets of mass atrocities, while others are protected (see also, Heberle et al., 2020). We will also seek to link the group processes for the atrocities committed to the political, economic, and historical contexts given.

Theoretical Analysis of Victimization in Context

As introduced above, we consider that collective blame (Lickel et al., 2006) is central to the construction of threat narratives that justify victimization, as provocative or historical incidents are taken as evidence of malignity for the entire outgroup. The out-groups are perceived as homogeneous so that they can be treated as interchangeable (Medin & Ortony, 1989; Wagner et al., 2009). In addition, when individual or small groups of out-group members resist or engage in attacks, mutual radicalization is observed, justifying reciprocal acts of violence (Bruneau et al., 2017; Louis et al., 2020; Moghaddam, 2018). An initial violent act by a small group, such as the September 30^th plotters in Indonesia, is the pretext for indiscriminate state violence; failure to support such mass atrocities may even become a further pretext for widening the definition of victims.

Our model, TOPASC, is contextually sensitive, in which through our cases, we learn that the application of the five-step social identity model (Reicher et al., 2008) depends on the societal context and the steps’ order can be flexible rather than sequential. In many cases, the basic problems began around identity construction; those who are part of the ingroup and therefore are Indonesians, Cambodians, or Burmese, and those who do not fit with in-group values. In this, compared to other countries, the case of East Timor is rather unique as East Timor was forcedly integrated as part of Indonesia. During this integration, however, East Timorese who were believed to belong to a group or engaged with an ideology related to communism were deemed unfit. In addition to that, those who opposed the Indonesian military/government were considered a supporter of separation, thus a threat to the nation (Kiernan, 2003; Robinson, 2001).

In all cases, those who are unfit with ingroup values are then demonized and/or dehumanized and deemed as a threat or dangerous for ingroup identity and values, whereas ingroup values are believed as sacred, noble, and morally sound. In our cases, in Indonesia and East Timor, communists were seen to be savage and evil, as well as traitors (Indonesia case) or separatists (in the East Timor case). In Cambodia, the Vietnamese were believed to be internal enemies; Muslims and other religious communities were portrayed as a hindrance to productivity or parasites; intellectuals, potential leaders or aristocrats, and professionals were categorized as inherently counter-revolutionary: Chinese in Cambodia as ethnically impure (Kiernan, 2003; Williams & Neilsen, 2016). In Myanmar, Rohingyas are perceived as not part of the “national races” and as a threat to the Buddhist nation (Anwary, 2020; Wade, 2017). Of these labels, the moral superiority and virtue of the ingroup (Reicher et al., 2008) are positioned as justifying the massacre, imprisonment, or ethnic cleansing of the out-groups and as essential for the in-group to survive and thrive (Step 5).

In terms of the flexibility of the five steps, this can be seen in the case of communism in Indonesia. First, the PKI / communism is perceived as savage and traitors (Step 3). This construction of an evil enemy (Putra et al., 2019) led to a re-evaluation considering communism to be incompatible with Indonesian values (Step 2). Communism, which had been part of Indonesian ideologies from the early 20^th century until 1965, was then banned.

The Cambodian case presents a unique in-group—out-group relational context. In Cambodia, the outgroup was not merely understood as a single outgroup, but as multiple out-groups with different labeling. It was Cambodians against the Vietnamese, religious communities, intellectuals or educated people, and the Cambodian Chinese (Kiernan, 2003). Hence this case explains that the victims of genocide or mass atrocities can be members of groups simply because they are considered a threat (and do not fit) within the imagined nation.

Nonetheless, we also note that members of a group are targeted not simply because the group is a threat to the nation or ingroup. As two separate countries with no history of conflict, East Timor, a poor country (up until now one of the poorest countries in Southeast Asia), was not a threat to Indonesia. The perception of threat might have been triggered by global anti-communist propaganda, during the cold war, against the dangers of communism (Hansson et al., 2020). It was then that a Fretilin-controlled East Timor could be constructed as a potential threat by the Indonesian government. However, allegations that Fretelin was communist have often been viewed as a pretext for Indonesia to occupy East Timor (Kiernan, 2003). It seems likely that threat is an antecedent, a symptom, and an outcome of dehumanization and collective blame, rather than playing a purely causal role.

From our perspective, two other novel insights that emerge from our theoretical model, TOPASC (Figure 1) are to consider the psychological processes at play in the development of narratives of intergenerational or historical threat, in all four contexts, and the differences between the wide-ranging and diverse victim groups constructed as enemies in Cambodia and the more circumscribed victim groups in Indonesia, East Timor, and Myanmar.

First, based on the descriptions provided earlier about the targeted victims, the construction of historical intergenerational threats appears to significantly contribute to the promotion of collective blame against victim groups. In turn, where the groups are perceived as homogeneous, this perception facilitates justifying the massacre or imprisonment of children or stripping them of their citizenship rights. The case of the Communist victims in Indonesia offers an extreme illustration: the notion that Communism is passed down through generations has resulted in ongoing legal restrictions and the continued stigmatization of those with Communist family affiliations (Conroe, 2018). In this sense, the psychology of temporal distance seems of understudied importance to the construction of intergroup violence (Liberman et al., 2007). Narratives of group continuity, which can be protective against prejudice and bigotry (Jetten & Wohl, 2012), are replaced in the present example with what could be called narratives of threat continuity, which justify atrocities and genocide (see also, Ogunyemi, 2020). But, what are the conditions under which such narratives emerge?

As highlighted in TOPASC (Figure 1), it is the post-colonial context that directly contributes via previous experiences of conflict and historical narratives of existential threat in the four cases. An important element in many or most contexts in which threat narratives emerge is the legacy of racism and prejudice, infusing systems of political, social, and economic functioning in post-colonial societies as elsewhere (e.g., Salter et al., 2018). Power-holders in post-colonial contexts may internalize a colonial mentality of supremacism that legitimates violence and domination (Bettache, 2020; David & Okazaki, 2006). Historical continuity of some forms of prejudice (e.g., anti-Semitism) is associated with localized incidences of violence across centuries (e.g., Voigtländer & Voth, 2012). However, we would comment that while this relationship between centuries-old prejudices and outbursts of violence is clearly descriptively valid, it does not clearly specify predictively which stigmatized groups within a specific society will be targeted for mass atrocities and when. In addition, as we have limited data sources regarding the narratives, we recommend that future studies explore the construction of narratives regarding how children were targeted specifically. We return to this point below.

As a second reflection, we highlight the wide construction of enemies in Cambodia (versus narrower in our other three case studies) which increased the number of victims and thus is of applied importance as well as theoretically interesting. Previous research has identified that “tight” identities and norms serve to increase uniformity and consensus at the expense of punitiveness toward dissenters and comparative difficulty integrating newcomers (e.g., Gelfand et al., 2011). The Cambodian case suggests that in some cases, a narrower construction of the ingroup may even facilitate more widespread targeting of genocides, by permitting a diffuse construction of the “enemy” as anything outside this tight, narrowly-defined identity (see also, Louis et al., 2018). However, in some cases, tight definitions of in-groups defined, for example, by strict ritual purity codes may persist for decades or centuries without promoting violence (Wibisono et al., 2019). The conditions in which tight versus wide identity constructions develop into wider and more lethal out-group targeting remain to be addressed in future research.

Greater attention is also needed to identify the specific communities of disadvantage that are targeted for mass atrocities. Working in the American context, Heberle and colleagues (2020) have proposed that state-perpetrated violence by the police and military consolidates interlocking systems of privilege. While this is undoubtedly the case, to us there is more work that needs to be conducted in this context and others, to examine the constraints that sometimes limit state-perpetrated violence to single-victim encounters (e.g., when one person of color is killed by police in a town), compared with wider outbreaks. For example, such research into the intersectionalities of mass atrocities might ask, why, when men are comparatively privileged, are men from disadvantaged groups disproportionately targeted by the state for murderous violence? Or granted that there are inter-generational threat narratives that target particular groups: Under what conditions do family-wide and village-wide massacres re-emerge after periods of no violence or single-target violence?

Using our theoretical model (Figure 1), hypotheses can be mapped that conflict narratives rooted in historical conflicts, and sociological constructions of gender, contribute to targeting male victims for murderous violence and female victims for rape, for example (see also, Mukaman & Collins, 2007; Sanford, 2008). Such narratives also are likely to play a role in shaping wider processes of victimization to identify families and villages as threats. However, our model locates group processes and identity construction—active processes by leaders and group members—as the engines of atrocities and mass atrocities. Longitudinal research that explores changing practices in mass atrocities over regions and time in relation to leadership and followership discourses (e.g., using natural language processing) has the potential to illuminate factors that widen or narrow victimization processes (see also, Bilali & Vollhardt, 2019; Putra et al., 2021; Williams & Neilsen, 2016). By locating such research within the TOPASC model, scholars may benefit from identifying macro-, meso- and micro-level variables to track jointly over time, allowing the moderating factors that specify one trajectory or another to be identified.

Who Were the Perpetrators and How Did They Mobilize to Commit Atrocities?

As noted earlier, the social identity perspective of intergroup conflict and atrocities emphasizes the construction of violence as an essential or legitimate response to outgroup threat—one that protects the ingroup and enacts its values (Reicher et al., 2008; Tajfel, 1982). In this sense, as we elaborate in the discussion below, the in-group identity and the mobilization to atrocities are co-constructed with the out-group identity and the justification of victimization. In analyzing the perpetrators, we also reflect on the coordinated and intentional nature of mass killing for certain segments of the perpetrator group, and how perpetrators are selected (or self-selected) and deployed, in contrast to other sectors of the population who are not understood to be expected or legitimate killers.

From this perspective, each group’s norms, its social rules or standards for behavior, align to reflect and create perceptions of threat, and to support intergroup conflict and violence (Amiot et al., 2020; Louis et al., 2020). Narratives are put forward that justify and call for violence by group leaders and through media and social media. Elements of these narratives that have mobilizing power include grievances (variously considered by different theorists as relative deprivation or anger, threat, and injustice); moral conviction; and the effectiveness of the action to protect the in-group and advance its goals (van Zomeren et al., 2011, 2012; Walker & Pettigrew, 1984). Each of these aspects is clearly identified in the case studies below.

Yet social identity theorists have also sometimes been called on to consider how people choose among the diversity of possible pro-group tactics available in response to threats from genocide to negotiation (e.g., Louis, 2009). Similarly, scholars have been pushed to explain the minority of ingroup members who actually take collective action, often under 5% of the population (e.g., Himawan et al., 2022; Louis et al., 2020). In the case studies below, we explore the role of intersectionality in suggesting answers to these research questions. Without an understanding of active perpetrators as a subgroup of their constituent community, genocide prevention is inevitably hampered (Austin & Koppelman, 2004; Baum, 2008; Waller, 2007, 2016) Understanding the interaction among dispositional, situational, and social factors is important in understanding the development of perpetrator individuals and groups, as we shall see below.

A Theoretical Analysis of Perpetrator Selection and Mobilization

As seen earlier, the data show strong support for a social identity perspective account of intergroup conflict and atrocities. In each case, we see that violence is constructed by perpetrators as an essential or legitimate response to outgroup threat—one that protects the in-group and enacts its values (Reicher et al., 2008; Tajfel, 1982). The group’s norms strongly support group members to mobilize for intergroup violence (Amiot et al., 2020; Louis et al., 2020). Narratives by leaders openly call for violence and justify it, expressing anger and hate at enemy threats and injustice, advocating with a moral conviction for atrocity, and sparking the effectiveness of violence to protect the ingroup and advance its goals. In this sense, unsurprisingly, we find that a social identity analysis provides a very useful lens with which to examine intergroup harm-doing in Southeast Asia.

Yet as noted earlier, social identity scholars have also sometimes been pushed to consider how people choose among the possible tactics from genocide to negotiation or conciliation (e.g., Louis, 2009), and how the minority of group members who actually take collective action, often under five percent of the population, are selected (e.g., Louis et al., 2020; see also, Himawan, E. M., Louis, W., & Pohlman, A. (2022)). As we can see through the cases in Indonesia, Cambodia, East Timor, and Myanmar, there are a number of intersectionalities that are made salient to illuminate these questions.

One thing to be emphasized is that large-scale atrocities occurred not because of the absence of the state, but rather because of active violence by the state and in particular by military forces (see also, Dutton et al., 2005; Melvin, 2018; Waller, 2016). The hierarchical chain of command in military forces legitimizes and demands the obedience of lower-ranked personnel upon pain of court-martialing (Dutton et al., 2005). In most nations, military personnel have a “license” and a skill set to kill to defend against national threats (Waller, 2007). For these reasons, when a target is dehumanized, demonized, and positioned as a threat to the nation, military personnel may be vulnerable to use as tools to participate in atrocities to eliminate the threatening targets.

Moreover, one aspect of the social psychology of mass atrocities that is highlighted in the data is the importance of preexisting roles and subgroups (see also, Hornsey & Jetten, 2004), particularly those for state actors. The norms and threat perceptions within these state-backed groups (often formed by propaganda outlets targeting these subgroups) are critical in mobilizing for atrocity or alternatively putting a brake on would-be perpetrators. Within these subgroups, our case studies also show the vital importance of leadership and conformity—topics we return to below. Nevertheless, the range of military groups involved in the mass atrocities varies significantly, from the more widely-mobilizing, immersive, and inclusive genocidal behavior of the Indonesian context of the 1960s, which included a range of civilian actors alongside soldiers, to the narrower mobilizations that supported violence in East Timor. For perpetrator groups as for victim groups, the construction of wider versus narrower perpetrator mobilization is an important dimension to theorize and study.

In each case, military murderousness existed in dynamic with two distinct and important civilian forces, which legitimized and directed military violence, and which were threatened and subjected to it. First, civilian perpetrators volunteered or were recruited from subgroups, communities, and militias that were positioned in existing social cleavages and tensions as “enemies” of victim groups (e.g., Anwary, 2020; Robinson, 2001, 2003). Ethnic, religious, and conflict historical divisions structured the networks by which military invitations to participate in violence were channeled. This is particularly highlighted in Roosa’s (2020) analysis of the different composition of the militia groups supporting the 1965 violence in different states of Indonesia, from thugs to student groups to religious organizations. Based on existing cleavages, then, enemies of the targeted group were liable to be approached and recruited for violence, under threat if need be.

We also have earlier highlighted the gender intersectionality, and the historical and political forces at the macro-level that feed into the construction of perpetrators: the mobilization of women perpetrators is nearly absent in some contexts (i.e., Indonesia, East Timor) but more present in others (e.g., Cambodia; see Jacobsen, 2018). It seems to us that the moderating factors that determine how widely the mobilization of perpetrators occurs lie at the level of macro-variables, such as historical, economic, and political forces arising out of post-colonial relationships. These macro-level processes have been documented by scholars from within particular fields (e.g., Anderton & Carter, 2019; Cribb, 2001), but to our knowledge, mass data have not been deployed to test a full model linking individual, group, and societal levels over time (for one comparative analysis, see Rodriguez & Kinne, 2019). This is an exciting direction for future research, as we revisit below.

A final aspect of intersectionality that is of interest in the mobilization of perpetrators is in relation to international audiences, and the construction of their tacit or explicit support for mass atrocities (Kiernan, 2002; Robinson, 2003). Just as within a country ethnic, religious, and political rivals of a targeted group were liable to be approached by the state for recruiting to support the violence, the same dynamic could be observed in international partnerships. Cold War anti-Communism apparently spurred Western countries such as Australia and the United States to actively support mass atrocities in Indonesia and East Timor (e.g., Simpson, 2008). In Cambodia, a conflict between Vietnam and the United States profoundly shaped the nature and rise of the Khmer Rouge (Hughes, 2009). In this regard, the present data support the call to move beyond binary in-group–out-group distinctions to consider ideologies and interconnections among multiple groups in simultaneous interaction (Louis et al., 2020). This is an important direction of future research and theorizing in intergroup harm-doing.

In integrating future research at the individual level, TOPASC (Figure 1) also directs our attention to consider past perpetration as a predictor of future perpetration (e.g., Schwab, 2010), but also victimization as a possible antecedent for radicalization to violence, including for state actors (e.g., Dutton et al., 2005; Winton & Unlu, 2008). Another predictor, we propose, would be profits accruing to the passive constituents of the perpetrators (e.g., ordinary citizens) from ignorance, such as state taxes, royalties, and other economic gains which may reinforce the likelihood that cultures of impunity will persist, as civil society actors turn a blind eye to state violence (Bass, 2016; Gutman, 2018). By examining the macro-, meso-, and micro-levels of analysis over time, in empirical research, future scholars can build on the present case studies with comparative research.

Denial, Justification, and Commemoration

In this final section, we consider the collective denial, rationalization, meaning-making, and memories of intergroup violence in the four contexts. As scholars have noted, collective denial of atrocity has many purposes (Bilali & Vollhardt, 2019; Putra et al., 2019). We reflect on processes of denial and the construction of innocence/ignorance for sectors of the population. While perpetrators and victims are keenly aware of atrocities, in some cases perpetrators’ neighbors, families, religious and political groups, and civilian agencies are expected to know and to celebrate the atrocities, while in others they are to be kept in the dark through censorship and taboos, to create a culture of impunity.

For perpetrators, denial can be strategic to protect a group from retaliation or sanctions (Bandura, 1999, 2002); it may shield individuals and leaders whose past actions now are judged as evil (S. Cohen, 2001; Putra et al., 2021); or it may reflect initial or dissonance-fueled moral conviction that the threat was genuine and the atrocity in some sense necessary or legitimate (Bilali & Vollhardt, 2019). For victims, and for society as a whole, denial may serve a system-justifying function, instilling a sense of meaning and control that may be threatened if the scale of an atrocity is acknowledged (Jost, 2019; Jost et al., 2015).

In intergroup situations, perpetrators often deny any wrongdoing at first (Leach et al., 2013). Nevertheless, an apology is sometimes considered an important aspect of group reconciliation (Iqbal & Bilali, 2018; Putra et al., 2019). This is because an apology demonstrates the perpetrator group’s willingness to acknowledge their past transgressions to the victim group. Yet as we will see below, there are often interesting silences, taboos, and intersectionalities, in which some subgroups are constructed to know the secrets of mass atrocities, versus constructed (via censorship, silencing, and taboos) as innocent or sheltered or ignorant, in the broader society of supporters and allies. Such silencing is an important facilitator of a culture of impunity, whereas acknowledgment of atrocities is a pre-condition for contesting the legitimacy of the events. At the same time, the acknowledgment and commemoration of past atrocities may have distinct psychological impacts on perpetrators and victims. Such acknowledgements validate and affirm the victim identity, and open the door to healing, recognition and compensation (e.g., Minow, 1999). For perpetrators and their constituents, however, the impacts may be mixed (Bonanno & Jost, 2006). For one subgroup, previously ignorant, acknowledgment may threaten a positive identity in which the atrocity never took place. For active perpetrators, atrocity acknowledgment may be a step toward accountability, with legal and social repercussions, but also the possibility of confession, healing, and access to treatment. For reformists within the perpetrator group, acknowledgment is essential to affirm a new identity in which norms of violence, impunity and censorship are broken and new norms of responsibility to prevent future atrocities are consolidated.

A Theoretical Analysis of Denial, Justification, and Commemoration

Taken together, the four case studies show strong denial and moral disengagement among perpetrators, in which inhuman acts are presented as morally justified and absolved from sanctions, consistent with the work of past scholars (Bandura, 1999, 2002; Staub, 2003). Compiling the arguments for denial in the present study, we found that victims were positioned as evil, impure, a threat, or insurgents so that mass atrocities were justified in defense of the state. Violence was seen as in line with the values of the community and “liberating” other civilians; the accused perpetrators, in particular the executioners, stated that they were acting righteously, they did not have other choices or they were just following orders; and systematic killing was denied by promulgating a narrative that both sides of the conflicting groups had become victims including the perpetrators (see also, Putra et al., 2021).

As noted earlier, collective denial of atrocity functions for perpetrators both on a group and individual level, shielding both from retaliation or sanctions. It may also reflect genuine moral conviction that the threat was genuine and the atrocity in some sense necessary or legitimate (Bandura, 1999; S. Cohen, 2001; Waller, 2007). Yet as we have seen earlier, knowledge and ignorance do not arise in neutral contexts: Advantaged group members are motivated to turn a blind eye to history (Bonam et al., 2019) and use their systemic power to create ignorance and undermine challengers (Phillips et al., 2015). For example, there may be active commemoration of the individuals who have fallen on one side of the conflict, and conspicuous silence concerning the victims on the other. Some atrocities may be named and some perpetrators challenged in international courts of law (e.g., lower-ranking state actors or civilians), while other atrocities are omitted and other perpetrators live out their lives in peace, even with honor in their communities (van den Herik, 2012).

Within perpetrator groups, clear intersectionalities emerge such that those perpetrators with more power—and in particular, those protected by state roles in the military—are much less likely to face censure, prosecution, and judgment. In our cases, for example, Pol Pot and Suharto were never officially found guilty for the mass atrocities over which they presided. This culture of impunity rests in part on the active silencing of critical voices through censorship and repression (Gutman, 2018). Since press freedoms (alongside the rule of law) are associated with lower human rights violations by state agents (Rodriguez & Kinne, 2019), limits to press freedom should be understood as part of a suite of measures enabling impunity and atrocities to persist (see Haldemann & Unger, 2018). However, advantaged group members also may be complicit in the silencing, avoiding awkward questions and conversations, and avoiding opportunities to learn about or confront the violence that cements their privilege (Bonam et al., 2019). Furthermore, disadvantaged group members may be complicit in the silence, avoiding the threats to control and meaning that acknowledgment of atrocities would evoke (Jost, 2019; Jost et al., 2015).

When individual perpetrator group members do admit to mass atrocities, many say that they were just following orders. Such diffusion of responsibility (Bandura, 1999) is commonly found in most incidents of atrocities (Bilali & Vollhardt, 2019; Reicher et al., 2008). However, other group members have maintained their moral conviction that violence was justified even decades after the event (Zulkilfi et al., 2012). This long-term zealotry is possible when an outgroup is demonized (Putra et al., 2018) and their values are thought to undermine ingroup values, highlighting that legitimizing narratives are still being propagated in Myanmar and Indonesia today.

Our theoretical model, TOPASC (Figure 1) helps us to theorize and analyze the comparative context for the four atrocities. Political and historical factors—in particular, domestic military regimes, constructed in relation to external threats and propped up by great power partners—determine the degree of repression in a society, which in turn empowers perpetrators’ collective denial strategies (e.g., Bass, 2016). Political and historical factors also enable the contestation of denial, and the gradual emergence of acknowledgment and commemoration, for example, after regime changes, or in relation to the actions of international actors such as the International Court of Justice or humanitarian groups (van den Herik, 2012). In turn, these political actors are embedded within socio-economic international systems that continue to financially reinforce (through aid or military subsidies) or seek to undermine (through sanctions) the ongoing human rights violations. Again, longitudinal comparative research which empirically tests the hypotheses that these factors are linked over time to changing degrees of denial and acknowledgment would be a valuable direction of future scholarship.

TOPASC also directs our attention to the individual level, connecting in with the scholarship of genocide and trauma (see also, Baum, 2008; Waller, 2007, 2016). For example, the literature has shown, with clinical case studies and survivor research, that acknowledgment is central to the victim and perpetrator healing and reconciliation (e.g., Minow, 1999). Cracking the social code of silence may be required for perpetrators to find help from post-traumatic stress, domestic violence, and suicidality (see also, Bonnano & Jost, 2006). Similarly, a series of case studies around the Truth and Reconciliation Commissions show that working toward commemoration may be healing, but also may be revictimizing, for victims, families, and communities (Bevernage, 2018; Boehme, 2020). Comparative studies which track the macro-, meso-, and micro-levels of analysis over time, to test the hypothesized moderators empirically, would be a fruitful direction of future research.

Contributions of the Present Article

In terms of how the emergence of mass atrocities is related to the societal context, the deep-rooted impact of colonialism on current-day politics and social psychology in Southeast Asia deserves more attention. It is not uncommon for horrific atrocities, such as the Rohingya persecution or the massacres in East Timor, to be portrayed as independent disasters in “backwards” parts of the world. But these events emerge from histories of colonial exploitation inflicted by the European empires on societies that are still deeply entwined politically, economically, and socially with Great Power struggles.

To connect macro-level and meso-level processes promoting mass atrocities, the present paper draws upon the scholarship of intersectionality (Cole, 2009; Collins & Bilge, 2020; Crenshaw, 1989; Overstreet et al., 2020; Rosenthal, 2016) to better understand the psychology of mass atrocities, and the relationship between macro-level factors and the group processes of victimization, mobilization, and denial. Genocides and mass atrocities have the potential to reinforce the co-created structures of oppression, as scholars of intersectionality have highlighted (Buchanan & Wiklund, 2021). To our knowledge, the lens of intersectionality has too rarely been brought to the study of violence (cf., Heberle et al., 2020; Kimmel, 2017).

Yet as our analysis shows (see also, Baum, 2008; Waller, 2007, 2016), the victimization and mobilization of perpetrators do not affect all members of advantaged and disadvantaged groups equally. Subgroups of victims, perpetrators, and constituents are differently exposed to, engaged in, and ignorant or aware of the atrocities taking place. Thus, in addition to shifting the needle on broad risk factors for a society to experience mass atrocities (e.g., by increasing the risk of famines), historical, economic, and political forces arising out of colonization and decolonization define subgroups at more or less risk of perpetration and victimization. This is an important direction of future research, in our view.

Engaging colonial and post-colonial macro-level factors has the potential to illuminate social psychological theorizing more broadly, not just in the countries identified in the present research, but more broadly, including in the United States, England, France, and other colonial powers. In this sense, our theoretical model, TOPASC (Figure 1), draws from previous scholarship in genocide studies (Austin & Koppelman, 2004; Baum, 2008; Waller, 2007, 2016) but provides a novel integrative model and multi-level, group-centered analysis that offers fruitful directions for future research in intergroup behavior broadly.

Directions for Future Research

While the present model engages complexity and interdisciplinary integration to build on contemporary theories of intergroup relations which remain at a dyadic level, the true interdisciplinary scholarship that comprehensively connects a range of variables to one another across the levels of analysis, and considers feedback loops as well as linear causality, has typically not been done. In case studies and historical analyses, as well as in the thriving literature on genocide studies (e.g., Baum, 2008; Waller, 2016, 2007), scholars have examined particular variables or clusters of variables that link political, economic, historical, and sociological forces to groups’ identities and norms, and to individuals’ trauma, well-being, and social and political attitudes and actions. To our knowledge, there is no comparative research that examines the three central elements of our model—victimization, mobilization, and denial or acknowledgment—in relation to one another over time. However, the present world of big data offers the opportunity, through analysis of online and social media discourses as well as data from political and economic large-scale projects, to attempt to join the dots in longitudinal quantitative analyses that could also aspire to model and predict the trajectories of mass atrocities (see also, Verdeja, 2016). This is an exciting direction for future scholarship.

It is important to note that, due to the non-empirical nature of this work, we are unable to provide a detailed and comprehensive elaboration of all aspects of TOPASC depicted in Figure 1. Although we have addressed both macro- and micro levels, our primary focus in this paper is on the meso level, specifically group processes. Therefore, we recommend that future research will provide a balanced emphasis on macro, meso, and micro levels. In addition, perspectives from the victims are still underrepresented in our work. For example, how does the victimized group acknowledge this past violence? How does the acknowledgment by the perpetrating group affect the members of the victimized group, either positively or negatively? Based on TOPASC, addressing these questions should involve dialogue with individual-level factors (e.g., victim mortality rates, levels of mistrust, radicalization, and attitudes toward forgiveness), as well as meso and macro-level contexts.

Furthermore, although we have provided explanations of how perpetrators approach past transgressions, whether through denial or acknowledgment, we recommend further comprehensive analysis for future studies. This analysis might encompass factors related to public visibility and international awareness of the violence. Future investigations also could explore differences between denial or acceptance of past events within a country and on the international stage.

From a social identity perspective, we have made the point that whether genocidal processes are unleashed is proposed to depend upon the intergroup context (Hopkins & Reicher, 2011; Reicher et al., 2008; Tajfel, 1982), and specifically upon the process of leadership and beliefs about the legitimacy, stability, and security of the group’s positive distinctiveness. These are the potential mediators or moderators that might change to link changes in macro-level factors in the TOPASC model to changing victimization, mobilization, and denial. When political, economic, historical, and social forces affirm that an intergroup hierarchy is legitimated and stable, these system-justifying beliefs should mean that outbursts of violence would not be expected (Hornsey, 2008; Jost, 2019; Jost et al., 2015). Examining changing leaders’ discourses in relation to group members’ beliefs and state-perpetrated violence is an interesting direction for future research. Experimental research that links the variables to establish the direction of causality would also be of interest.

TOPASC also models the victimization, mobilization, and denial processes as linked to one another in feedback loops, which reflects our approach and the social identity perspective that these are outcomes of socio-structural forces and beliefs. However, we have also in previous research highlighted the particularities that arise from leaders, factions, and internal discussions and debates that profoundly impact upon groups’ choices of tactics (e.g., Bilali & Vollhardt, 2019; Louis et al., 2018; Moss, 2014; Waller, 2007). Comparative research that examines when and how the three processes unraveled their connections and become disconnected, as part of deradicalization, restraint from violence, and cessation of atrocities, will make an important contribution to deradicalization research.

In a similar vein, we are not aware of any comprehensive discourse and qualitative analysis that has tracked changes across all of the “isms” (i.e., sexism, racism, classism, ableism) intersecting within a particular context of mass atrocities or genocide. Comparative research would be very valuable to examine patterns contrasting contexts in which genocidal violence attaches simultaneously to multiple targets (e.g., perhaps in the case of the Nazi Holocaust: Jews, people with disabilities, Communists, etc., and in the Cambodian case) versus when the genocide plays out in more narrowly targeted violence (e.g., in Indonesia in the 1960s). In particular, discourse analysis contrasting the construction of gender for contexts in which women perpetrators are empowered (e.g., Cambodia) versus generally excluded from performing violence (e.g., Indonesia) would also be of great interest. What TOPASC offers is a framework for rich interdisciplinary hypothesis generation as well as empirical analysis that centers on social identity principles and the group level of analysis, but extends to consider individual agency and well-being. Such a model we hope has great potential to draw scholars to the study of mass atrocities and other important intergroup behaviors, from rioting to insurrection.

Citation Statement

While in this paper the authors have tried to discuss the four cases raised from many perspectives, our academic disciplinary backgrounds, as reflected in our earlier position statement, might influence the overall biases in the cited literature. For example, we have drawn primarily on psychological and historical sources in our analyses, and we are more familiar with the scholarship from Australia, Indonesia, and traditional WEIRD sources than (for example) primary scholarship from Cambodia in the Khmer language.

Constraints on Generality

The present theoretical model “TOPASC: A Theory of the Psychology of Atrocities in Societal Contexts” can be used to understand and analyze a wider context. However, it needs to be adapted and adjusted to the social context where it will be used.

Our generality in explaining mass atrocities is also constrained by the four cases we use, in which we specifically focus on analyzing mass atrocities from Southeast Asia: Indonesia, Cambodia, East Timor, and Myanmar. While these cases may shed light on the issue of mass atrocities in other contexts, caution should be exercised in interpreting the relationship between one case and another. Many risk factors for mass atrocities in colonized countries are not present, for example, in the colonizing empires. At the same time, we consider that persistent denial, justification, and failure to commemorate mass atrocities abroad marks the contemporary societies of the former western empires, and may contribute to the ongoing legitimization of violence in neocolonial invasions and violence in the present day. This is a promising direction for future work with TOPASC.