Towards a Narrative Understanding of Victimhood: The Perception of Intergroup Conflicts in Light of Past Ingroup Victimization

: The present study explores the role of linguistic compositional characteristics in transmitting collective victimhood beliefs. Experimentally manipulated excerpts of history textbooks were used to examine the perception of the victim position of national outgroups and its intermediary social psychological processes with Hungarian ( N = 415) and Finnish ( N = 116) participants. The results reveal that the narrative composition of the victimhood narrative had a significant effect on the perception of the target groups’ victimhood position. The evaluation of the groups changed according to which variant of the story was introduced. The results demonstrate that the perception of a perpetrator group can be changed purely by means of narrative construction and that their actions can acquire a “victim tone”. This effect is present in both the Hungarian and Finnish samples, suggesting that narrating an event of victimhood has certain universal characteristics, although their effect is partially dependent on the national-historical-cultural context.


Introduction
Collective victimhood has been an important area of interest in social psychology over the past decade, with an array of scholarly papers from different cultural backgrounds and different theoretical approaches (see Bar-Tal et al., 2009;Noor et al., 2012;Vollhardt, 2020). However, little is known about how collective victimhood consciousness is represented in historical narratives; and how the structural and compositional linguistic properties of the story may affect the interpretation of historical events. The present study attempts to develop a research setting where the narrative structural composition of an intergroup conflict is experimentally manipulated to modify the perception of victim roles.
Narratives are a central means of social communication that convey conceptions of the national past that are passed down as normative schemas and which form a dynamic historical continuity (Liu & László, 2007;Sani et al., 2007;Wertsch, 2008). This paper contributes to the field of collective victimhood research by emphasizing the narrative psychological aspects of collective memory and its relation to national history and history education in two different national historical settings.
The study observes from an intercultural point of view how characteristics of historical representations contribute to both the phenomena and the research context of collective victimhood. The paper focuses on Hungary and Finland, two EU-member countries, which are both located on the Eastern border of Europe, and whose histories share many common themes. The 20th-century history of both Hungary and Finland share the theme of oppression and a fight against alien forces (e.g., revolt against the Austro-Hungarian Monarchy in Hungary and historical battles against Russia in Finland) and losses (e.g., territory losses in WWI). However, the two countries have constructed a different understanding of their national history. Hungarian historical writing emphasizes historical losses and defeat (László, 2013), while Finnish historical writing accentuates the perseveration of independence through historical challenges (Hakoköngäs & Sakki, 2016).
The present study aims to demonstrate the impact of the narrative structural composition ) on the perception of the actors' victim position. We will examine the effect in two seemingly similar historical contexts (Hungarian and Finnish) and their contribution to understanding the phenomenon.

Historical narratives and group identity
Social identity theory implies the process of social categorization, in which meaningful categories are created to establish a clear distinction between the ingroup and the outgroup (Tajfel, 1978;Tajfel & Turner, 1979), but also enables the group to distinguish between friends and enemies. Collective remembering of historical events is closely related to social identity. As a result, historical events are continuously negotiated and reinterpreted to form a "usable past" (Pennebaker & Gonzales, 2009;Wertsch, 2008), which fulfills both an epistemic and an identity function of historical remembering (Licata & Mercy, 2015). The former serves the group's need for origin, while the latter helps to provide a positive self-image. Group history tells us who we are and where we come from (Liu & Hilton, 2005) and, at the same time, places events in a particular perspective that displays the group in a favorable light (László, 2013). Group members play an active role in this process by maintaining and protecting the favored representations of history (Klar & Baram, 2016;Wertsch, 2002).
Historical representations are organized in a narrative form (Liu & László, 2007). Wertsch (2008) claims that collective memory itself is a thematically organized and schematized representation of events that form a group narrative. These narrative templates are specific to each cultural tradition and provide a frame of reference for interpreting the past and present challenges faced by group members. Narrative templates exert a considerable effect on collective memory by adjusting new experiences in the narrative frame, ignoring those inconsistent with it (Sahdra & Ross, 2007;Wertsch, 2002). For example, studying Russian national grand narratives, Wertsch (2008) argues that the so-called 'expulsion of foreign enemies' narrative template can be identified as a common plot in Russian narrative tradition: a peaceful setting is disrupted by foreign forces, leading to suffering and devastation, which is overturned by the heroism of the Russian people. From the 13th century (e.g., the Mongol invasion) to the 20th century (e.g., Hitler's invasion), Russian national historical tales and cultural memory bear the marks of this specific narrative template, which plays a central role in the meaning formation of events in the last decade as well (Wertsch, 2008(Wertsch, , 2012. Negative events and historical traumas pose a threat to group identity and rupture the continuity of group history by challenging the core self-beliefs of the members (Alexander, 2012). Collective traumas disrupt the group's psychological integrity, and group members might struggle with reintegrating traumatic experiences into their historical templates (Hirschberger, 2018;Volkan, 2001). Making sense and integrating the negative events into the group history is crucial for both the perpetrator and victim groups, but for different reasons (Hirschberger, 2018). The primary need for perpetrators is to restore the group's moral image, whereas victims need to find a way to reinforce their sense of agency and security (Bilali & Vollhardt, 2019;SimanTov-Nachlieli & Shnabel, 2014).
Encountering recurrent defeats and traumas may lead to the prevailing feeling of being victimized, making the group unable to mourn their losses (Volkan, 2001). This phenomenon is referred to as self-perceived collective victimhood or collective victimhood consciousness Vollhardt, 2020). At the core of self-perceived collective victimhood are widely shared group beliefs that consist of the perception that the group was unjustly, deliberately, and undeservingly harmed in the past by the fault of outgroups David & Bar-Tal, 2009;Schori-Eyal et al., 2014). Collective victimhood affects the interpretation of historical events and attribution of responsibility . It may impair the group members' ability to take the perspective of and be empathically concerned with other groups (Demirdağ & Hasta, 2019;Noor et al., 2008). Furthermore, collective victimhood prevents group members from seeing the ingroup as a responsible agent for historical conflicts, even when the atrocities were committed by the ingroup (Hirschberger, 2018;László, 2013;Mészáros et al., 2017;Noor et al., 2012Noor et al., , 2017. The victimized group is often characterized by compulsive alertness that is accompanied by the perception of the world as a dangerous place, a defensive and distrusting attitude, and a sense of vulnerability and helplessness Eidelson & Eidelson, 2003;Noor et al., 2017;Schori-Eyal et al., 2014). These so-called dangerous beliefs (Eidelson & Eidelson, 2003) or worldviews might impact how the group relates to other groups (e.g., Bilewicz & Stefaniak, 2013), immigration (e.g., Mészáros & Szabó, 2018), or even international politics (e.g., Lerner, 2020). Vollhardt (2009) distinguishes between exclusive and inclusive forms of victimhood consciousness. Inclusive victimhood is an adaptive way of perceiving collective victimhood that allows individuals to consider others' suffering as well. Individuals with inclusive victimhood show higher empathy toward others' hardships and are more concessive concerning the "victim" category membership (Adelman et al., 2016;Gordijn et al., 2001;Vollhardt & Bilali, 2015). However, individuals who think in terms of exclusive victim consciousness are more prone to sense that their group's hardships are unique and cannot be compared to other groups' suffering, which leads to an inflexible, monopolized category of victimhood. Comparative victim beliefs may target former or current intergroup conflicts, but they can also convey a general feeling of distinctiveness and uniqueness of the suffering that might result in devaluation of hardships of even unrelated groups. The limited available data show that competition for the victim status can also occur between groups that are not responsible for the group's past or present suffering, especially in cases when victimization is not given due recognition. Numerous studies pointed out that individuals with higher global exclusive victimhood are less prone to accept other groups' suffering (see Bilewicz & Stefaniak, 2013;De Guissmé & Licata, 2017).
Despite its negative consequences on intergroup relations (i.e., Hirschberger, 2018), placing the ingroup in a victim role can be an adaptive strategy in terms of group identity. Noor et al. (2012) argue that the accentuation of the ingroup's victim position in an intergroup conflict can be considered advantageous, as being a victim provides a morally superior position, which can invite empathy from outside parties and legitimize the ingroup's efforts or atrocities Vollhardt, 2009). Gaining recognition of the victimized position is especially important if the group's position is not obvious, and in some cases, the victims may behave as perpetrators and perpetrators as victims. These diffuse roles often motivate groups to accentuate their suffering and lessen their responsibility for the crimes committed Hirschberger et al., 2016), for which narration serves as a powerful tool.

The narrative construction of collective victimhood
An attempt to elaborate on a traumatic event and to make sense of the unacceptable loss begins with narration. Most of the studies examine the forms, the antecedents, or the consequences of collective victimhood, and less attention is paid to its narrative organization (see Vollhardt, 2020), however, narratives are primarily responsible for the elaboration and transmission of a traumatic event (Fülöp et al., 2014;Pennebaker & Susman, 1988). History education as a stage of socialization and history textbooks as a form of shared discourse is central to transmitting historical knowledge and assisting in acquiring a normative explanation of history (Aldrich, 2006;Bilewicz et al., 2017;Liu & László, 2007). Besides its role in constructing national identity and the ability to integrate complex perspectives of historical actors, historical textbooks often provide dichotomized and ethnocentric viewpoints of historical events that might lead to a distorted and selective view of history (Alridge, 2006;Carretero & van Alphen, 2014;Psaltis et al., 2017c;Wertsch, 2002). They can be seen as tools for identity politics; a legitimation for the prevalent collective memory (Baranovicé, 2001;Sakki, 2014). This draws attention to the importance of the narrative construction of historical events (Bilewicz & Liu, 2020;Liu & László, 2007), particularly the way how the events are narrated, in addition to what they tell. Historical narratives plot the sequence of facts in order to make a story (White, 1990) that satisfies both the identity and the epistemic needs of the group at the same time (Licata & Mercy, 2015;Paez & Liu, 2011). For a positive distinction of the ingroup, it also needs to define apparent intergroup boundaries, especially for events where the ingroup's victim or perpetrator position is diffuse (Hirschberger, 2018), and averting blame becomes ponderous (Wohl et al., 2006).
One way to construe the victim and perpetrator roles is to capture them through the actions and outcomes associated with roles ( Bar-Tal et al., 2009). According to the United Nation's definition, the victim is usually identified as a person who individually or collectively has suffered different forms of harm and substantial losses, while the perpetrator is someone who individually or collectively committed a crime or a violent action that led to suffering or deaths (Melander et al., 2004). However, the victim and perpetrator roles are also outcomes of psychological construction, albeit being rooted in historical facts. Consequentially, they are to some degree changeable in accordance with situational factors (Nadler & Saguy, 2004). In this sense, it is also feasible to assign victim and perpetrator roles through the narrative structural elements of the story, such as semantic roles  and the various predicates associated with them. It is possible to modify the perception of group positions (victim vs. perpetrator) in a conflict without changing the historical facts (in an interpersonal setting, see Bohner, 2001). Harmful actions may be reframed as just defense or retribution regardless of the actual content of the actions in situations where roles are made ambiguous (Hirschberger et al., 2016).
In the following, we present the narrative compositional elements (narrative categories) that may be relevant in the victim-perpetrator relation and their way of contributing to the narrative organization of collective victimhood consciousness.
Agency refers to the perception of one's capability to act in an effective and active way (Bandura, 2001). It can be captured through the feeling of being able to affect one's surroundings and reach one's goals (Hamilton, 2007;Ferenczhalmy et al., 2011). Agency and its linguistic correlates (i.e., activity and intention) are one of the major categories in narrative psychology that can be expressed by different voices. Active and passive voice refers to whether the subject or object in the sentence performs the action of the verb (Formanowicz et al., 2017). The active voice indicates activity and capability for influencing the actual situation, while the passive voice obscures agency by placing the actor in the background and the object in the foreground (Penelope, 1990). A high level of agency in the narratives indicates that the group takes responsibility, plays an active role in reaching the desired goals, and assumes a capability to influence the outcome, while lack of agency implies a lack of control, incapability, and general powerlessness (Ataria, 2015;Yamaguchi, 2003). In an intergroup context, a relatively low ingroup agency compared to the outgroup indicates defensiveness and leads to perceiving the ingroup as not being responsible for the events. Traumatized groups often show a low level of agency and position themselves as helpless agents who are unable to control the situation (Erős, 2007;Fülöp et al., 2014;Herman, 2015;Volkan, 2001).
Evaluation is essential in narratives (Labov & Joshua, 1967), which can be clearly stated or implied and can be realized in various ways. It is a linguistic tool that might divide actors into morally defined categories by assigning positive or negative attributions to them or their actions. Intergroup distribution of positive and negative evaluations serves a distinctive view of the ingroup and outgroup . It may contribute to enhancing the positive moral image of the ingroup and devaluation of the outgroup. Previous research has found that collective victimhood is associated with asymmetry in intergroup evaluations (e.g., Hirschberger et al., 2016;Szabó, 2020) that leads to portraying the ingroup as a positive and the outgroup as a negative actor of the events (Korostelina, 2010;Schori-Eyal et al., 2017).
In narratives, the psychological perspective represents the agent's mental states (thoughts and feelings) that function as a narrative tool for perspective-taking and triggering empathy (Keen, 2006). Traumatic events necessarily generate an intense expression of thoughts and negative feelings (Conejero & Etxebarria, 2007;Pennebaker & Harber, 1993), which may invite empathy (Hogan, 2001) and bolster the identification with the actor (e.g., de Graaf et al., 2012). The expression of mental states in a discourse setting may contribute to the structuring of the event in a way that promotes the acceptance of the actor's perspectives. Fülöp et al. (2014) pointed out that negative emotions such as fear, disappointment, and sadness paired with inner thoughts with positive propositional content (the subject matter of the thought) are more often attached to the Hungarian group in history textbooks and lay historian accounts regardless of the event's valence. At the same time, opposing outgroups are usually assigned negative, hostile emotions, thoughts, and intentions, which depicts them as deliberate actors and underlines their responsibility in the events Vincze & Rein, 2011). The results provide further evidence that Hungarian national identity (Mészáros et al., 2017;Szabó, 2020;Ting-Toomey & Dorjee, 2019) and history writing are essentially victimhood-centered, with an emphasis on losses and defeats (Kovács et al., 2012;László, 2013;Vincze et al., 2021).
Concerning the structural compositional properties of the historical narratives, László et al. (2002 found that different intergroup distributions of narrative markers resulted in different group representations: the low level of ingroup agency associated with negative ingroup emotions such as fear, sadness, and disappointment is indicative of the victim's position. On the other hand, a high level of agency with hostile intentions and negative cognitions displayed by the negatively evaluated outgroups are distinctive narrative features associated with the perpetrator role. This intergroup pattern of narrative structural properties indicates the collective victim representation of a group (László, 2013). Recently, Van Dijk (2020) and Lewis et al. (2021) came to a similar conclusion, pointing out that a "legitimate" victim is presented as a passive, vulnerable, helpless, violated actor in the stories who show empathy-induced emotions such as scare or fear.

Importance of historical representation in constructing victimhood consciousness
Many scholars argue that an adaptive way of constructing historical representations would be to form complementary intergroup perspectives and multilayer narratives in history teaching, addressing the complexity of historical settings (McCully, 2012;Psaltis et al., 2017c). However, the victimhood-based narration of events that László (2013) found in Hungarian history teaching materials usually develops a dichotomized, prefabricated point of view, not leaving space for in-between interpretations of the events (Adelman et al., 2016).
Nevertheless, not all nations with experience of historical traumas present their history from the victim's point of view. Finnish history writing, for instance, tends to emphasize a positive, agentic outlook on national history, highlighting the perseverance of the Finnish people despite their historical challenges (Hakoköngäs & Sakki, 2016;Hakoköngäs et al., 2021). Analysis of Finnish history textbooks found the representation of the Finnish nation as being strong and unified against external threats, and wars being portrayed as either positive or neutral (Hakoköngäs & Sakki, 2016), even though the Finnish national history includes several periods of oppression, and the fight for independence is a central theme in Finnish history (e.g., the Winter War between Finland and the Soviet Union in 1939-1940, in which Finland lost a significant proportion of its territory). Some scholars argue that the reason for constructing a more adaptive way of historical event representation is that the Finnish history curricula strongly emphasize developing critical historical thinking, even though the former patriotic ways of teaching remain intact (Rantala, 2012;Sakki & Pirttilä-Backman, 2019). These efforts have a clear contrast with Hungarian history education, which has been characterized by a mythical and conservative, schematic narration of events with little regard for outgroup perspectives (Jakab, 2008), and a nationalistic approach in general (Agnes & László, 2007;Kaposi, 2010).

Purpose of the present study
Numerous studies have examined collective victimhood consciousness in general (Bilewicz & Stefaniak, 2013;Schori-Eyal et al., 2017) and in a Hungarian context (see Fülöp & Kővágó, 2018;Szabó et al., 2020;Vincze et al., 2021). Nevertheless, little is still known about collective victimhood and its relation to narrative composition, especially in a cross-cultural context.
The study's central hypotheses (H) are as follows: Based on previous studies (Jenei et al., 2020;László, 2013), we hypothesized (H1) that a victimhood-oriented narrative composition (a particular intergroup distribution of specific narrative markers) may alter the perception of a target group's victim position, even when its role (whether it is a victim or a perpetrator) is ambiguous. Accordingly, we expected that the group presented in the story with a relatively low level of activity and intentions associated with empathy-induced negative emotions (e.g., fear) is more likely to be identified as a victim, even if the group is the initiator and beneficiary of the conflict.
We also assumed that the effect of the narrative composition is mediated through specific worldviews, such as vulnerability, distrust toward others, and feelings of helplessness, which are associated with a victimhood consciousness. We hypothesized that the more the participants see the target group in the story as vulnerable, distrustful, and helpless, the more likely they perceive the target group as a victim (H2).
As several studies indicate that identifying the victim in a story can be driven by specific narrative properties, we expected that the perception of victimhood is principally the result of the narrative composition regardless of the participant's nationality (H3).
On the other hand, we supposed that the effect of the narrative composition might depend on the collective victimhood consciousness of the participants. Numerous studies point out that individuals with higher exclusive victimhood are less prone to accept other groups' suffering. Although both nations share a similar history of oppression and a series of defeats, victimhood is much less present in the Finnish national identity, unlike in Hungary, where a victimhood-centered national identity is prevalent. Thus, we expected that the collective victimhood of participants would significantly reduce the effect of narrative composition on the perception of the target group's victim position in the Hungarian sample but not in the Finnish sample (H4).

Sample
The total sample (N = 551) included 415 Hungarian adults (139 males, 276 females; M age = 31.54; SD age = 13.06; range between 18 and 76 years), and 116 Finnish adults (26 men, 84 women, and 6 identified otherwise; M age = 29.97; SD age = 9.32; range between 18 and 69 years). In the Hungarian sample, 47% of the participants finished some form of higher education, while for the Finnish sample, this was 67%.
The participants were recruited via online social media advertisements and personal requests at the University of Pécs, and the University of Eastern Finland in the autumn of 2019.
The participants were informed of the goals and qualities of the data collection, the method and extent of the data storage, and the protection of their anonymity. The option for aborting the study was provided for every participant, and formal participation consent was requested. The participants were not rewarded for their participation in any manner.
Since both Finnish and Swedish are recognized as official languages in Finland, participants were asked to indicate their native language and permanent place of residence. Participants whose native language was Swedish (n = 8) were excluded from the Finnish part of the data because linguistic characteristics play a central role in the experimental design. Any other participants with missing data from the main experimental evaluation (n = 12) were also excluded from the overall dataset.

Research design and procedure
In the experimental design, participants were randomly assigned to read one of two fictional stories of an intergroup conflict presented as history textbook excerpts (for the English translation, see the Appendix, the original Hungarian and Finnish narratives are provided as supplements to this paper) using the PsyToolkit online platform (Stoet, 2010(Stoet, , 2017. The target groups in the stories have been chosen to be unrelated to Finnish or Hungarian history so that we could examine the effect of narrative composition without the effect of the historical link. The story's plot displayed the Kyrgyz and the Uzbek nations and their long-lasting territorial conflict, from which the story flashed a violent episode. We produced two event versions where Kyrgyz and Uzbek were assigned to different group roles (i.e., victim or perpetrator). Victim and perpetrator roles can be accomplished in at least two ways: through the frame of the story (i.e., who initiated the conflict, who is the beneficiary and the injured party of the conflict) or by the narrative structural composition of the story (i.e., intergroup distribution of activity and passivity, negative and positive emotions, or evaluations).
Two versions of events have been developed accordingly. The story's framing was the same for both versions: the Kyrgyz group initiated the conflict and was the beneficiary (the Kyrgyz military troops marched into a predominantly Uzbek town, and following bloody events, the territorial dispute was finally settled in Kyrgyz interests) (see Figure  1). However, the story versions differed according to the assigned victim role induced by the narrative composition. In the congruent narrative, the Kyrgyz group was represented as more active, negatively evaluated, and having more hostile intentions than the Uzbek group, who had more positive and empathy-induced emotions and showed a low activity level. Based on the previous research (László, 2013), we assumed that the indicated intergroup distribution of narrative markers would lead the Kyrgyz group to be perceived as a perpetrator and the Uzbek group as a victim. In the incongruent narrative, the relative allocation of victimhood narrative markers was reversed: the Kyrgyz group was placed in a "victimized" role by attributing lower activity, higher frequency of empathy-inducing emotions, and positive evaluations compared to the Uzbek group, who was represented as a negatively evaluated and highly active agent with hostile intentions and a lack of empathy-inducing emotions leading to perceive as rather a perpetrator of the event. The intergroup distribution of employed narrative markers is shown in Table 1. The congruency/incongruency refers to the correspondence/contradiction between the victim role induced by the story frame (Kyrgyz group as the initiator and the beneficiary of the conflict) and the narrative composition (intergroup distribution of the victimhood narrative markers). We expected that keeping invariant the story frame, the narrative composition of the story can affect the perceived victimhood of the target groups in the story.  The narratives were designed initially in Hungarian and translated to Finnish afterward, using the committee method (Brislin, 1980), which involves a group of bilingual researchers who translate from a source to a target language (Nasser, 2005). Special attention was paid to the specific linguistic markers in the text in order to convey the same psychological meaning and the same intergroup distribution.
After reading one of the story versions, participants were asked to answer questions concerning the two target groups in the story and complete short scales.

Measures
Victimhood of the target groups. Perceived victimhood of the target groups was measured with one item ("In your view, how likely are the Uzbeks/Kyrgyzes to see themselves as the victims of the events?"), and participants indicated their response on a 7-point scale ranging from 1 (very unlikely) to 7 (very likely).
Dangerous Beliefs. A modified Individual Group Belief Inventory (Eidelson, 2009) was used to assess the particular set of beliefs associated with perceived collective victimhood (see Bar-Tal et al., 2009). Participants were asked to evaluate both target groups in each experimental condition concerning their perceived vulnerability (e.g., "The Uzbeks/Kyrgyzes feel unsafe."); perceived distrust towards other groups (e.g., "The Uzbeks/Kyrgyzes think they should be cautious of other groups' intentions."); and perceived helplessness (e.g., "The Uzbeks/Kyrgyzes think they cannot influence their own future."). Participants indicated their responses on a 7-point scale ranging from 1 (completely disagree) to 7 (completely agree). The scale showed high reliability for both the Kyrgyz-specific and the Uzbek-specific versions for both the Hungarian (HU) and Finnish (FI) samples (HU: α = .90-.93; FI: α = .89-.90).
Empathy. Empathy was measured by two subscales of the Interpersonal Reactivity Index (Davis, 1980) to control the trait empathetic attributes of the participants. Emphatic concern (e.g., "I often have tender, concerned feelings for people less fortunate than me") and the ability for perspective-taking (e.g., "I believe that there are two sides to every question and try to look at them both") were measured on a 5-point scale from 1 (completely disagree) to 5 (completely agree) and were later aggregated on a unified scale. A total of 14 items formed a reliable scale (HU: α = .80; FI: α = .78).
Collective victimhood consciousness. We also measured the participants' beliefs concerning past historical victimization of their own nation. A national group-specific version of the global exclusive victimhood beliefs subscale of the Global Collective Victimhood Scale  was used. Exclusive victimhood measures the perceived uniqueness and distinctiveness of the ingroup's victimization (e.g., "While all experiences of victimization are somewhat different, the experience of the Hungarians/Finns is truly unique.") (HU: α = .79; FI: α = .55). Although Cronbach's alpha values, especially for the Finnish sample, are somewhat low, they are still acceptable (see e.g., Hinton, 2004).
Manipulation check. After reading one of the story versions, participants indicated the possible source of the story, whether it was derived from an Uzbek or a Kyrgyz history textbook ("Which nation's textbook do you think the excerpt is from?"). Numerous studies show that groups prefer to display themselves as victims (see e.g., Barnard, 2003;Basic, 2015;Korostelina, 2010;Stojanov & Todorov, 2020) in an intergroup conflict due to several benefits of victim position (see Bar-Tal et al., 2009;Crawford & Kuperman, 2006;Noor et al., 2017).

Data analysis
A manipulation check was performed using the chi-square test to determine the effectiveness of the manipulation. The hypothesized conditional process model (see Figure 2) was tested with PROCESS macro model 5, v3.5 (Hayes, 2018) in SPSS version 22. We examined the significant effect of the narrative structural composition on the judgment of the target group's victimization in a single model, but separately for the two sub-samples using a bootstrapping approach with bias-corrected 95% confidence intervals (n = 20,000). Due to the differences in variable scaling, all scale-type variables were entered in a mean-centered way (for an overview of the method, see Hayes, 2018). Since the primary question of the study was how the perceived victim position of a perpetrator might change as a result of the narrative structural composition, in the model, we examined only the victim position of the Kyrgyz group as an outcome. The story versions were coded as 0 = congruent and 1 = incongruent, treating the congruent condition as a reference level in the model. The narrative structural composition was the predictor (IV), with perceived dangerous beliefs of Kyrgyz group as mediator (M). Dangerous beliefs were measured as a composite score of vulnerability, distrust, and helplessness. The outcome variable (DV) was the perceived victimhood attributed to the Kyrgyz group, and the exclusive victimhood of the participants constituted the proposed moderator (W). Trait empathy and the dangerous beliefs attributed to the Uzbek group were entered into the model as a covariate. Direct and indirect effects were deemed to be significant if the CIs excluded zero (Hayes, 2018) Note: C n (covariates) = trait empathy; dangerous beliefs of the Uzbek group

Manipulation check
The chi-square test was used to determine the successfulness of the linguistic manipulation. The frequency distribution was in line with our expectations in both the Hungarian (χ 2 = 70.1; p ≤ .001 Φ = .44) and the Finnish (χ 2 = 79.4; p ≤ .001 Φ = .83) samples. Participants categorized the congruent story version more often as being derived from an Uzbek textbook in both samples (N HU_UZBEK = 130, N HU_KYRGYZ = 70 and N FI_UZBEK = 51, N FI_KYRGYZ = 5), while the incongruent story version as being derived from a Kyrgyz textbook (N HU_UZBEK = 52, N HU_KYRGYZ = 136 and N FI_UZBEK = 5, N FI_KYRGYZ = 55). The result shows that the participants were able to differentiate the target groups' points of view from which the story was presented. It seems that the semantic roles of the target groups (i.e., victim or perpetrator) created by the narrative composition oriented the evaluation of the two groups according to the initial assumptions, which is assumed to be a result of the experimental manipulation of the narrative structural properties.

Effects of the narrative structural composition on the perceived victimhood of the target group
Mediated and moderated effects of narrative composition on the perceived victimhood of the target group were tested using PROCESS macro model 5, v3.5 (Hayes, 2018) with the goal of examining how the effect of narrative composition comes to be and under what circumstances its effect exists or not (Hayes & Rockwood, 2020). Since the primary question of the study was how the perceived victim position of a perpetrator (the indicator and the beneficiary of the conflict) might change as a result of the narrative structural composition, in the model, we examined only the victim position of the Kyrgyz group as an outcome.
The result (see Figure 3 and Table 2)

Journal of Social Psychology Research
Note: Path values represent unstandardized mean-centered regression coefficients. The first coefficient represents Hungarian, while the second represents Finnish model coefficients. * p < .05, ** p < .01, *** p < .001

Discussion
The primary aim of the study was to examine the effect of the narrative structural composition on the perception of a target group's victim position. The results show that the victimhood-oriented composition of the story (incongruent) can change the perceived victim position of the perpetrator group (Kyrgyz), which is mediated by dangerous beliefs. The result indicates that for identifying the victim, participants took into account the degree of vulnerability, distrust, and helplessness of the target group as important characteristics of the victim role. Furthermore, this effect was present in both samples, proving that the narrative composition principally operates on the linguistic level.
However, the analysis revealed that the direct effect of the narrative composition was dependent on the participants' exclusive victimhood consciousness in the Hungarian sample but not in the Finnish sample. The result implies that the higher level of exclusive victimhood consciousness reduces the likelihood of accepting the victimized depiction of the perpetrator group. There are some possible explanations that may account for this result. One is that the type of event (defeat and a significant loss of territory) might resonate with the Hungarian historical narrative templates (victimhoodorientation), which eliminated the effect of the narrative structural composition (Wohl & Branscombe, 2008). The other possible explanation is that individuals with a high level of exclusive victimhood are more prone to making a sharper distinction between victim and offender (Shnabel et al., 2013) and less likely to tolerate role ambiguity produced by the narrative composition. In intergroup conflicts where the group was both a victim and a perpetrator in the course of history, a perceptible separation of victim and perpetrator roles could be essentially important (see Bar-Tal et al., 2009), leading to different interpretations of ambiguous information (Staub, 2006).
Results show a slightly different picture for the Finnish sample. The manipulation also showed an effect on estimating the victimhood of the target group; nevertheless, the direct effect did not depend on the participants' own collective victim beliefs. On the other hand, victimhood beliefs assigned to the Uzbek group significantly affected the perceived victimhood of the Kyrgyz group. It suggests that the greater an individual perceives the Uzbek group as vulnerable, distrustful, and helpless, the less likely to see the Kyrgyz group (perpetrator) as a victim. The result indicates that for Finnish subjects, the perception of the Kyrgyz group's victim position is relative to the other group's (Uzbek) mental states (i.e., dangerous beliefs). It seems that the Finnish participants interpreted the position of the Kyrgyz group from a holistic point of view, focusing on the meaning derived from the level of the story content. A possible explanation for this result can be found in Finnish history education, which places a strong emphasis on critical thinking and the presentation of events from multiple perspectives (Hakoköngäs & Sakki, 2016;Rantala, 2012;Sakki & Pirttilä-Backman, 2019), which might result in a more balanced evaluation of intergroup roles in a historical setting.

General discussion
Our results verified that (1) the narrative structural composition  of an intergroup conflict can influence the perception of the actors' victim position, even when contradictory facts are provided; (2) the effect of the narrative composition is in part independent from the national context, (3) and mediated by the victimhood beliefs (dangerous worldviews); however (4) certain factors may influence or even modify this effect. Although the structural composition of the event is able to change the person's perception in relation to the event, the historical thinking of the participants, in particular, their own sense of collective victimhood can change this effect. The results showed that a sense of exclusive victimhood may prevent individuals to perceive the perpetrator group as a victim in the incongruent narrative composition. On the other hand, the higher level of trait empathy promotes the perception of the victim position of the target group (Kyrgyz). Due to the fact, that the results above appeared only for Hungarian but not for the Finnish sample, it implies that national history and historical thinking might play an important role in event perception on the narrative level.
Many studies pointed out that the sense of collective victimhood is central to the Hungarian identity (e.g., Kővágó, 2020;László, 2013;Szabó et al., 2020;Vincze et al., 2021) and collective traumas have a profound effect on the representation of subsequent national historical events . The present study revealed that national traumas impact even the interpretation of intergroup conflicts between groups unrelated to national history.
The suggested effect of historical thinking on event perception can also be found in the results of the Finnish sample but in a slightly different way. Finnish subjects are more likely to perceive one group's position in relation to the other group, which implies readiness for taking multiple perspectives that might result in a more balanced perception of the event. In this context, history education provides a crucial social stage. Forming complementary intergroup perspectives and multilayered narratives (McCully, 2012;Psaltis et al., 2017c) is important in developing historical consciousness and empathy (Bryant & Clark, 2006;Sakki & Pirttilä-Backman, 2019). This aim has an especially crucial aspect in post-conflict societies, where social representations of history often conserve the polarized narratives of the experienced conflict (Psaltis et al., 2017b). Numerous studies are available regarding the role of history education in the process of identity formation and conflict resolution (for an overview, see Psaltis et al., 2017a). However, the way these conflicts are construed in the narratives, particularly in the history schoolbooks, is understudied. Our findings demonstrated the importance of subtle narrative structural composition as a symbolic tool in meaning construction (Semin, 2009) and the presence of its effects apart from various language cultures.
It is important to note that stressing the victim position of one group does not imply that the other group is necessarily perceived as a perpetrator, nor do the authors wish to suggest that. However, by means of the narrative structural composition presented in this study, it is possible to construe a reframed version of the event that might produce a perception of one group as a victim and the other one as an offender -or, in an opposite way, the narrative composition of the story can also be used to portray events in a less biased way. Either way, our results point in the direction that besides the factual information of a historical event, it is also worth paying attention to how it is told.

Limitations and future directions
It is important to address the methodological limitations that this kind of narrative manipulation can hold. As the present study was one of the first attempts to inspect the effect of a manipulated narrative structure in Hungarian, the thorough validation of the experimental intervention is yet to be achieved. It is important to highlight that the results could be derived from many interdependent factors (e.g., slight changes in explicit interpretations in the narratives) which we were not able to control at a satisfactory level in the present study. Despite the obtained effect of the narrative composition, the study design did not allow us to independently test the effect of each individual narrative category. The question of how these narrative structural features contribute to the perception of the target groups' victim role remains unanswered.
Another limitation relates to the imbalanced sample sizes that may limit the validity of the results. The present study aimed to examine the effect of narrative structural composition in two different national samples with similar historical backgrounds -in terms of historical challenges -albeit with a different approach to it. Consistent with our expectations, the results showed that Finnish participants' sense of victimhood did not influence the perception of the victim position of the target group. We explained this result with a low level of the participants' collective victimhood and traced it back to the non-victimizing Finnish national identity (see Hakoköngäs & Sakki, 2016). However, the low sample size and the low victimhood scale reliability for the Finnish sample might question our arguments. Nevertheless, the impact of the participant's historical consciousness (victimhood consciousness) on the perception of the target group's victim position in the story is supported by the results obtained from the Hungarian sample. As a possible future direction, it would be interesting to repeat the study with other national groups known to have a high or low collective victimhood consciousness.
As a cultural characteristic, it is important to note that Hungarian victimhood is typically historical in nature, rooted in already settled historical conflicts, and is more likely to be marked by feelings of betrayal, the sense of pride associated with victimization, the perceived threat to the territorial integrity, or the invisibility and un-acknowledgment of past suffering (Szabó, 2020). An interesting aspect of the results is that Hungarian participants generally associated a less intense victim position with the Uzbek and Kyrgyz groups. It raises the question of whether Hungarian participants are in fact more susceptible to the suffering of outgroups than they are prone to generally devaluing these instances of the victimization of others due to their pre-existing national historical templates that highlight their own unique suffering. The question is worth studying, along with the general study of the narrative framing of victimhood, which shows a clear gap in the existing literature. It is also important to keep in mind when interpreting these results that national historical curricula and differences in history teaching can be considered more of a context of interpretation and not an immediate antecedent of the observed phenomena.
The other possible explanation is that individuals with a high level of exclusive victimhood are more prone to making a sharper distinction between victim and offender (Shnabel et al., 2013) and less likely to tolerate role ambiguity produced by the narrative composition, which is an interesting direction for future research.