System Justification in the Social Explanation of the Violence against Minoritized Groups

Violence against minoritized groups continues to be pervasive despite social norms prescribing nonviolence because it is legitimized in democratic societies. It is likely that, when individuals are faced with the occurrence of violence within their social environment, they justify it because the individuals tend to perceive and defend all social events as just, legitimate, and necessary. We addressed this possibility in two experimental studies in which we manipulated the minoritized group victim of violence (black vs. women vs. gays vs. control). In Study 1 (N= 104), participants blamed more women for their own victimization; blamed the perpetrator less when the victim was black people and depicted homophobic violence as a social issue similar to general violence in society. Study 2 (N = 217) went further by showing that these effects occur especially when participants were asked to respond as thought by society. We discuss explanations for violence as examples of individuals’ tendency to justify the social system and provide new insights for the secondary victimization research.


Introduction
Violence does not victimize the population randomly.Black individuals, women and gay people are the most often victimized social groups across the world (e.g., Carroll, 2016;Global Violent Deaths;2017;Mapping Police Violence, 2019).The widespread of physical, sexual, emotional, and psychological violence against minoritized groups is a pervasive social drama with strong negative consequences for society's quality of life since it reflects and perpetuates the hierarchization of social groups (see Kunst et al., 2017).Despite the pressure of social norms prescribing nonviolence and anti-prejudiced values as an organizing principle of social life in democratic societies (e.g., WHO, 2009;Banhs & Crandall, 2013;Banhs & Branscombe, 2011), the high frequency of aggression towards social minorities seems to be descriptively normative, suggesting a degree of social legitimacy (d 'Oliveira, 2019;Paiva & Pereira, 2021).
A consolidated line of research in the social psychology of legitimacy (e.g., Jost & Major, 2001) has shown that people use seemingly unbiased justifications to perceive social inequalities as legitimate (for reviews, see Costa-Lopes et al., 2013;Abad-Merino et al., 2018).Research within system justification theory (SJT), for example, has shown that people in general, whether they come from advantaged or disadvantaged groups, are motivated to justify the social system in ways that are perceived as just, legitimate, and necessary (Jost & Banaji, 1994;Jost, 2019).Although the legitimation of violence has not been the direct focus of SJT, previous studies have highlighted the need to consider the central role of legitimating myths in justifying unjust actions, practices, and policies, including individual and state violence (e.g., Shaw et al., 2017).
Accordingly, analyzing the legitimacy of violence in the context of SJT can shed light on the social psychological processes underlying individuals' motivation to justify the social system that promotes, accepts, and perpetuates violence, particularly against minority groups.Based on SJT, we propose that "people exhibit system-justifying tendencies to defend and rationalize existing social, economic, and political arrangements" (Jost, 2019, p. 1), and are also likely motivated to explain violence against minority groups in ways that justify the reasons for it.In this article, we present two studies that address this possibility.

The Explanations for Violence and System Justification Beliefs
Revista Interamericana de Psicología/Interamerican Journal of Psychology 2023, Vol., 57, No. 3, e1743 ARTICLE | 3 Most Western societies assume core values of justice and equality.At the same time, the unfair distribution of wealth has increased in almost every region of the world, including the most democratized ones (Word Inequality, 2018), in which hierarchical relationships between social groups are still perceived as legitimate.To understand the seeming paradox, the social psychology of legitimacy has proposed some analytical models and theories highlighting the central role of legitimation of social inequalities.For example, based on a review of the theory of aversive racism, theory of social dominance, and the system justification, Cost-Lopes et al. ( 2013) refer to legitimation as the psychological and social processes by which social attitudes, behaviors and arrangements are justified according to normative standards.Based on the system justification theory (Jost & Banaji, 1994), they highlighted legitimation as involving at least three levels of analysis: individual (by defense of self-image), group (by maintaining the hierarchy) and system (by justification of the system as a whole).Tyler (2006) pointed out that a set of beliefs can explain and make sense of a social system to justify the differences of power, authority, wealth or status present in society.Accordingly, people can be encouraged voluntarily to accept and defend the social rules, decisions or arrangements that are considered legitimate.
In the context of violence and its relation to perceptions of justice, this phenomenon can also be presented as a process by which the occurrence of the violence and aggressive behavior are justified (Paiva & Pereira, 2021).For example, acceptance of violence can reflect the individuals' more general tendence to perceive the way society functions as natural, fair, and necessary.In fact, Chapleau and Oswald (2014) have shown that the stronger this tendency to justify the system, the less moral outrage people feel in the face of rape.In this sense, the way violence is perceived and socially explained may be related to the category of the victim's affiliation, especially if the victim is a member of a minority group.Because people behave toward social minorities in ways that legitimize their social disadvantage (see Pereira et al., 2010;Jost & Banaji, 1994), it is likely that violence against members of minority groups is equally legitimized because the acceptance of violence, especially in a social context where it is systemic and trivialized, is consistent with people's tendency to justify the social order, i.e., a place where privileged groups remain privileged and protected, while the disadvantaged continue to occupy an underprivileged position and remain defenseless against the offensives of dominant groups.
Accordingly, while on the one hand violence is socially condemned, on the other hand it can be legitimized under certain conditions (e.g., Paiva & Pereira, 2021).Like legitimation of social inequality and exclusion (Costa-Lopes et al., 2013), violence can also be legitimized through system justification motivations for maintenance of the status quo (Chapleau & Oswald, 2014), namely in situations where this violence is systemic, as in the current Brazilian context (e.g., Lobo, 2020).This tendency to defend the system and consequently maintain the status quo has been adequately explained during more than 25 years of research into the system justification theory (Jost & Banaji, 1994;Jost, 2019).
This theory states that individuals are motivated to defend, support and justify prevailing social, economic and political arrangements, thereby legitimizing the social system on which they depend (Jost & van der Toorn, 2012;van der Toorn & Jost, 2014).The theory predicts that even members of socially disadvantaged groups in certain contexts are motivated to justify the system, perceiving it as good, fair and legitimate (Jost & Banaji, 1994;Just & Toorn, 2012).Examples of actions that justify the system are stereotyping, the use of some ideologies to explain social facts, and blaming victims for their misfortune (Costa-Lopes et al., 2013).In this sense, justifications may be beliefs, attitudes, or actions that implicitly or explicitly legitimize events that occur in all aspects of social life and contribute to the preservation of social hierarchies and inequalities (Jost & Banaji, 1994;Jost, Burgess & Mosso, 2001).The question we are asking here is: what are the justifying factors that individuals use to explain the reasons for violence against minoritized groups?
Answering this question is crucial to know the extent to which representations of violence can reflect the more general tendency of individuals to legitimize the social order.Equally crucial is that violence in a society is not perpetrated indiscriminately against individuals (Martin et al., 2009).On the contrary, violence is systematically directed against specific targets, with women, blacks, and sexual minorities being the most vulnerable (Cerqueira et al., 2021).Analyzing the content of attributions of responsibility for violence against these targets can help us better understand whether these attributions follow a pattern that denotes a broader process of perceiving how society functions as legitimate.
Regarding the content of the reasons people give for violence, we predict that there should be at least three categories: blaming the victim for one's situation; blaming the perpetrators; personal disclaimer by perceiving violence as systemic and thus inevitable.We assume that the use of these types of explanations for violence is consistent with the motivation of the individual to defend and justify the system.In fact, such justification occurs when individuals blame the victims for their own suffering (see Campbell & Raja, 1999).This phenomenon has been studied especially based on just world theory (BJW, Lerner, 1980), which predicts that people behave as if they believe they live in a just world, where they deserve what they have and have what they deserve.
According to this theory, when people are confronted with situations of injustice that threaten their BJW, they are motivated to defend their belief in a just world (e.g., Dalbert, 2009;Jolley et al., 2018).For example, when confronted with victims of violence, people try to defend their belief in a just world by blaming the victims for their misfortune (Tavares et al., 2022).Thus, it is likely that one of the factors that legitimizes violence is the attribution of blame to the victims for what occurred to them.This phenomenon occurs when individuals are confronted with victimization of minoritized groups (De keersmaecker & Roets, 2017).This happens, for example, when individuals perceive that black people, themselves are racists (Ryan, 1971(Ryan, /1976)), or woman are guilty for being raped (Howard, 1984), or even when they blame gay individuals for the violence they suffer (see Ford et al., 1998;Wakelin & Long, 2003).Accordingly, victim-blaming can be readily invoked by people to explain the reasons for violence in society.
Another way to justify violence is when people do not blame the aggressors their behavior, as has been shown in research on police violence against black people.Indeed, recent studies have shown the denial of the offenders' responsibility when victims of police aggression are black, police officers are given lower penalties and victims lower compensation (Johnson & Lecci, 2020;Lima et al., 2018;Correll et al., 2014).Similarly, in cases of femicide, there is a tendency for the media to excuse the aggressors from responsibility (Correia et al., 2017), relegating the cases of victimization to the scope of the couple's private life (Paiva & Pereira, 2021).This suggests that there is a tendency to blame the aggression victims indirectly.This blaming seems to translate into an acquittal for the aggressors.Regarding homophobic violence, the lack of blaming of perpetrators is reflected in the discrepancy between the high number of murders of gays and the lack of conviction of perpetrators for hate crimes against gays (Oliveira et al., 2020), as it were possible to exist homophobia without homophobes (Teal & Conover-Williams, 2016).
Finally, a third way to justify violence occurs when individuals perceive social violence as systemic within a society and deny their responsibility for the occurrence of the facts (see Camino et al., 2001;Abad-Merino et al., 2018).A similar phenomenon appeared in the literature on societal prejudice.For example, in a study on cultural stereotypes, Devine (1989) showed that participants denied having negative stereotypes about black people, while perceived negative stereotypes about Black people as systemic in society and thus inevitable.Similarly, Camino et al. (2001) found that white participants recognized that racism in Brazil occurs because racism is systemic and unavoidable, but they perceived themselves as not racist.This idea that prejudice against minoritized groups is systemic was experimentally shown by Lima et al. (2019), when participants supported discrimination against black people after being asked to respond according to how society works, separating themselves from the discriminatory meaning of their actions.
In sum, people are motivated to justify the way society is organized and the events that take place in it (Jost, 2019), and thus are likely to appeal to justifying factors that help them explain the reasons for violence (Chapleau & Oswald, 2014).We propose that individuals appeal to at least three blame-based reasons to justify violence: the victims, the perpetrators, and the violence as systemic.We also suggest that the frequency of each of these types of justification may depend on the victim's membership in the group.We developed two studies in which we tested these hypotheses.In Study 1, we asked participants to express their personal opinions about violence in one of three experimentally manipulated target group situations (Black people; women; gay individuals) and then assessed the extent to which participants' responses provided reasons to justify violence.In Study 2, we went a step further and asked half of the participants to express their opinions about the reasons for the occurrence of violence, as in Study 1, while we asked the other half to describe society's opinion about violence to examine whether concerns about self-concerns about the anti-prejudice norm, which dictates that "good people are not prejudiced" (Lima et al., 2019), is an inhibiting factor in the situation of expressing one's personal opinion compared to the situation in which participants express society's opinion.

Study 1
In this study, we presented participants with a description of a hypothetical violent situation in which the victimized target group varied (i.e., black people vs. women vs. gay individuals vs. controls) in a randomized between-participants experimental design.The  2023, Vol., 57, No. 3, e1743 ARTICLE | 7 violence: blaming the victim, blaming the perpetrator, and viewing violence as systemic.
We also predicted that these reasons for violence would vary by target group of violence, which would allow us to characterize the specificity of legitimation for each target group.

Participants
A convenience sample of 104 university students at a public university in the city of João Pessoa, Brazil, participated in this study (MAge = 22.58; SD = 6.44; 54.8% male).
We used an experimental design in which participants were randomly assigned to one of four conditions that corresponded to the target violence group: black (n = 26), female (n = 26), gay (n = 26), control (n = 26).A sensitivity power analysis for main effects and interactions in ANOVA, using Webpower (Zhang & Yuan, 2018) with α = 0.05 and four groups, showed that we have 80% power of detecting a medium effect size of f = 0.36 or higher.

Experimental Manipulation and Procedures
We asked the participants to collaborate in an opinion survey regarding violence in Brazil.Because we used a between-participants experimental design, each participant answered only one of the four conditions: a control condition, in which we questioned them about violence without specifying any specific target group; and three experimental conditions, each with specific target group (black individual vs. women vs. gay people).
Specifically, in the control condition, participants read: "In Brazil, many people experience violent situations.In your opinion, what are the main reasons that contribute to these people being the target of violence?".In the black-target condition, they read: "In Brazil, we witness many cases of homicide.Statistics show that blacks are disproportionately murdered.What do you think are the main reasons why blacks are targeted by this type of violence?".In the female condition, participants read: "We see cases in Brazil where women are the target of domestic violence (e.g., they are beaten by their partners), with serious consequences for their physical and psychological integrity.
But not all women go through this situation.In your opinion, what are the main reasons why some women are affected by this situation?".In the gay target condition, they read: "We witness in Brazil situations of violence with great cruelty.When you look at the history of these cases, it turns out that in many instances the victims were gay people.
What do you think are the main reasons why some homosexuals are subjected to this type of violence?"

Dependent Variable Measurement
We asked participants to write on a blank sheet why this type of violence occurs in Brazil.We categorized participants' responses based on the three predicted categories (perception of violence as systemic; blaming the victim; blaming the perpetrator).
Because participants wrote multiple sentences with different content within the same paragraph, it was not only possible to assign each elicited sentence to a category (i.e., violence as systemic; blaming the victim; blaming the perpetrator), but also to quantify how many sentences characterizing each category elicited them, which allowed us to assess the intensity with which these categories were mentioned.Specifically, we calculated the strength of evocation for each type of category most frequently mentioned by participants by counting how many times each participant mentioned different elements of the same category, which is a strategy typically used by people to affirm their positions on socially relevant issues (Tracy, 2019).We then used the number of invocations as the reason for violence as the dependent variable.
To validate our categorization and quantification of each category, four independent intergroup relations experts indicated the number of times each participant mentioned elements in their responses that blamed the victim and the perpetrator and viewed violence as systemic.The experts did not know the research hypotheses and did not know to which condition the participants were assigned.They read each participant's responses and indicated for each of them how many responses referred to elements that blamed the victim, blamed the perpetrator, and blamed the system.This allowed us to calculate the degree and internal consistency of the raters' assessments.This consistency was high for the three types of blame: blaming the victim, α = 0.87; systemic violence, α = 0.83; and blaming the perpetrator, α = 0.71.This high consistency allowed us to calculate the intensity of the type of explanation attributed to the violence.Thus, this measure indicated the intensity with which participants attributed each type of explanation to violence.

Ethical Concerns
The two research we present in this manuscript was approved by the Research Ethics Committee of a Brazilian university (CAE 66357516.9.0000.5188)and complies with the Ethics Guidelines of the National Health Council in Brazil (Resolution 466/12).
Participants volunteered and were informed that they should answer the questions anonymously.We also made the data of all studies publicly available through the Open Science Framework (https://osf.io/wmrzs/).

Data analysis
We analyzed the data using SPSS version 25, applying a repeated-measures factorial ANOVA to design 3 (types of blame: victim blame, systemic violence, and perpetrator blame) x 4 (target of violence: control, black people, women, and gays), with the first factor varying within participants and the second varying between participants.
The dependent variable was the number of evocations as a reason for violence by participants.

Results
We found strong consistency across the three types of explanations for violence present in the four contexts analyzed: violence as systemic, victim blaming and offender blaming.The contents of the systemic violence category characterized violence as widespread throughout society due to the usual way in which Brazilian culture works (e.g., "a hierarchical society, where class, race and gender differences still predominate"; "this is because there is still a social and political hegemony that praises the White race"); the victim blaming content tended to blame the victims for the violence they suffered (e.g., "they accept verbal violence as something normal"; "victims go through issues related to fear of dying, fear for the family and it also happens because they believe in love and end up accepting violence.");and the perpetrator blaming referred mainly to the attribution of guilt for the violence to the person who commits the violence (e.g., "the aggressors who still have a colonized mentality"; "blame must be placed on the aggressor").The complete list of evocations of these categories in each target group analyzed can be found in the OSF.

Discussion
In this study, we predicted that participants would evoke content for three types of explanations to justify violence.The results were consistent with our predictions, as we categorized participants' evocations into three types of reasons for violence: Perceiving violence as systemic, blaming the victim, and blaming the perpetrator.In addition, participants tended to perceive violence as systemic in society more than the other two types of social actors.Overall, the results suggested that people tend to assign blame according to the victim membership group.However, because of the anti-prejudice norm in Western societies prescribing that "good people are not prejudiced and demonstrate a positive view of minorities", it is possible that the content of the answers may have been inhibited by the participants' desire to present themselves as more sensitive to the suffering of minorities (see Abad-Merino et al., 2018).To address this limitation, we conducted a second study in which half of respondents were asked to report their opinion about the reasons for the occurrence of violence, while the other half described society's opinion about this phenomenon.

Study 2
The purpose of this study was to replicate Study 1 and analyze how people attribute society's perception of the violence against minoritized groups.Similar to study 1, we aimed to test the influence of the target victim of violence in the content of individual's evocations not only as personal opinions, but also regarding what society thinks about the occurrence of this violence.Our hypotheses were like those in Study 1: individuals would explain violence by evoking elements that justify the system, so that the content of evocations would vary in accordance with the target group of the violence.
However, we considered that the intensity of responses to the types of blame would vary not only in relation to the target group of violence, but also in relation to the type of opinion (personal vs. societal opinion).The hypothesis was that people responding as a society are not pressured by the anti-prejudice norm to suppress negative opinions, and as a result, would be more prone to blame the victims.This effect should occur because the perception of victim blaming is condemned by the anti-prejudice norm (Tavares et al., 2022).However, when people react according to general societal opinion, they do not feel personally responsible for expressing prejudiced opinions and attitudes (Lima et al., 2018), reasoning that this is society's way of thinking and not their own.

Participants
A convenience sample of 220 university students from the city of João Pessoa, Brazil (MAge = 21.20;SD = 5.92; 61% male).Three participants were excluded because they did not answer the dependent variable (i.e., missing value), which reduced the sample to 217 participants.A sensitivity power analysis for interaction effects in ANOVA showed that we have 80% power of detecting an effect at least of f = 0.25.We used an experimental design in which participants were randomly assigned to one of eight conditions, corresponding to a 2 (opinion focus: personal opinion vs. society opinion) X

Experimental Manipulation and Procedures
As in Study 1, the participants received a sheet containing a vignette addressing the problem of violence in Brazil.Depending on the experimental condition, the vignette addressed violence without specifying a target group (control condition) or addressing this problem in relation to black people, women or gay individuals.The difference from the first study was the manipulation of the role played by the participants.We asked half of them to respond according to "their personal opinion".For the other half, we asked them to express not their personal opinion, but the opinion of society in light of the presented case of violence-

Dependent Variable Measurement
The main measure was participants' responses to the open-ended question about reasons for violence.As in Study 1, we classified the participants' responses into three categories: Perception of violence as systemic (e.g., "Brazilian society uses it as an excuse, mainly as something cultural, as a sexist society, and with a failed political system"; "There are still many prejudices in society in general"); victim blaming (e.g.., "The reason why these murders are mainly committed by blacks is because they are more involved in drug trafficking than whites"); and offender blaming ("There are many aggressive people, in some cases these people commit such atrocities"; "aggressive partner, very jealous partner to the point of assaulting his partner").The full list of invocations of these categories in each target group can be found in the online materials (https://osf.io/wmrzs/).We then asked four experts who did not know the hypotheses to validate the categories of invocations according to the three predicted categories.They indicated how many category descriptors were elicited by each participant.We found the following inter-raters' internal consistencies: victim blaming (α = 0.87); systemic violence (α = 0.77); perpetrator blaming (α = 0.71).We used the number of evocations as the reason for violence in each category as the dependent variable.

Data analysis
As with the first study, we analyzed data using SPSS version 25, using a factorial repeated measures ANOVA with design 3 (types of blame: victim blame, systemic violence, and perpetrator blame) X 4 (target of violence: control, black people, women, and gays) X 2 (opinion focus: personal opinion vs. society opinion), with the first factor varying within participants and the second varying between participants.The dependent variable was the number of evocations as a reason for violence by participants.

Results
Descriptive statistics are presented on Table 2.The results revealed a significant main effect of type of blame, F (2, 418) = 121.13,p = .001,η² = .were no significant differences between the other targets of violence.
The interaction between type of blame and participants' focus was significant, F (2, 418) = 4.46, p = .01,η² = 0.02).Participants who responded according to personal opinion perceived more systemic violence than to the victim (b = 0.90, SE = 0.12, p = .001,d = 1.14) and than offender blaming (b = 0.94, SE = 0.09, p = .001,d = 1.41).In the condition in which participants responded about society's opinion, the violence was also perceived as systemic more than as victim blaming (b = 0.73, SE = 0.12, p = .001,d = 0.94) and than offender blaming (b = 1.18,SE = 0.09, p = .001,d = 1.74).Comparison between experimental conditions indicated that the difference was significant only for offender blaming so that participants in the society condition blamed the offender less than those in personal condition (b = 0.33, SE = 0.06, p = .001,d = -0.70).
The interaction between type of blame and target was also significant, F (6, 418) = 16.76,p = .001,η² = .19).In the control condition, participants perceived more systemic violence than offender blaming (b = 1.14, SE 0.13, p= .001,d = 1.75) and victim blaming (b = 1.23,SE = 0.17, p= .001,d =1.59).In the context of violence against black people, systemic violence was greater than the blaming the victim (b = 1.08,SE = 0.17, p = .001,Analyzing this interaction from another perspective, we observed that blaming victims in the condition of violence against women was greater than in the other situations.More blame was attributed when the target of violence was females than in the blaming was marginally lower when the target was women than in the control condition (b = -0.15,SE = 0.09, p = .09,d = 0.34.Finally, the three-way interaction between type of blame, target and focus of participants' opinion was not significant, F(6, 418) = 1.17, p = .32,η² = 0.02.

Discussion
Based on the results of Study 1, the current study predicted that individuals would invoke elements that justified the system, so the content of the evocations would vary depending on the target group of violence.In addition, we examined whether the intensity of the invocations varied not only with respect to the target group of violence, but also with respect to the type of opinion (personal vs. societal opinion).Overall, the results confirmed the findings of Study 1. Participants perceived violence as systemic rather than victim-blaming and blamed the victim more than the perpetrator of violence.In addition, Study 2 went further by showing that participants explain violence in accordance with the target group of the violence and with the focus of their opinion (personal vs. societal).
They blamed women more for their victimization than the perpetrators and tended not to blame the perpetrators of violence against black people.When participants responded to what society thinks, they blamed perpetrators of violence against black individuals less.
However, the results showed a weak effect on whether participants responded according to their own opinion (versus society's opinion), as manipulating this aspect did not affect the pattern of invocations related to the system and victim blaming.This last result may indicate some convergence between participants' personal and societal opinions on the portrayal of violence, especially against women and gay individuals.

General Discussion
In two studies we manipulated the target group of violence and asked Brazilian participants to indicate the reasons for violence in Brazil.The results were consistent between the two experiments, as the participants evoked elements featuring violence as systemic, followed by blaming victims and, offender blaming.Perceiving systemic violence was therefore the most strongly evoked to explain violence against each victim category.This suggested that the participants mainly believed that violence is normative, at least at the descriptive level.
Looking more closely at differences in the relevance of each type of reason for violence, we found that, in the first study, there were no significant differences between victim and aggressor accountability, but only between these two categories and the responsibility attributed to the system.In the second study, in which we added the condition that participants should answer according to the society's opinion, there were significant differences between the three categories mentioned, with the system's accountability being the most important.Interestingly, and in contrast to the results observed in Study 1, participants attributed less responsibility to the aggressor than to the victim.These results may suggest that participants not only perceive violence as something "normal" in Brazilian society because it is systemic (see Adorno and Pasinato, 2009), but that they indirectly legitimize the system by attributing less blame to the aggressor, which is a strong reinforcement of impunity, mainly when victims belong to minorities.
This possibility is consistent with the sociocultural characteristics and history of the context in which we conducted the studies (Miguel, 2015).Brazil is socially a slave culture in economic and organizational relations at the state level, patriarchal in family relations, sexist in gender relations, and heterosexist in social relations (Tiburi, 2021).
Accordingly, the pattern of our findings must be understood considering this broader frame of reference.Although the participants were college students with higher levels of education, who would be expected to think more critically about socially sensitive issues such as violence, the reasons they gave directly reflect the pattern that Brazilians have historically used (Camino et al., 2001): They acknowledge the existence of the problem in a more abstract and intangible sense when they perceive it as widespread in a society where they do not feel responsible for the issues that affect them; they tend to blame the victims more than the perpetrators , thus contributing to keeping the system as it is (Silva Pereira, 2022).This means that the participants consistently followed the normative ideologies reinforcing the social system's maintenance.It is as if it was violence without a causative actor, but only a permissible context and a victim, where the participants are merely observers of such a situation.Indeed, recent research on tolerance of violence in Brazil shows that discourses on violence are consistent with this interpretation (see Rollero & Tartaglia, 2019;Paiva & Pereira, 2021).
On the one hand, there is the proliferation of discourses against violence carried out by various institutions (Sousa, 2015); on the other hand, there are discourses consisting of ideas that justify and promote violence, especially when the victims of violence are black, gay, or women (Lima et al., 2019;Paiva & Pereira, 2021;Souza et al., 2021).As we predicted, participants gave reasons for violence depending on the victim group.Specifically, they gave women more responsibility for the violence they suffer and blamed the offender less when the victim was a black person, while they did not distinguish a specific pattern of explanation when the victim was a gay individual.This trend occurred most intensely in Study 2, especially when the participants were asked to respond as society thought.In summary, the participants mainly considered that violence in Brazil as a characteristic phenomenon in Brazilian society, which probably works as a way of perceiving this as inevitable.In the specific situation of women, they blamed the victim herself.When it was against blacks, the responsibility was not attributed to the offenders.
Our interpretation is that violence is not directed indiscriminately against individuals, but against specific targets, usually perceived as minorities.This interpretation follows the literature, which describes violence as a dramatic form of discrimination (Paiva & Pereira, 2021;Pereira et al., 2010).In fact, in addition to the high rate of violence directed at black people, women and gay individuals, the way society perceives and treats this violence may favor its perpetuation.The results of the two studies suggest that one of these forms of perpetuation may be reflected in the participants' evocations of the reasons for violence, which can be understood as different ways of justifying the system and maintaining the status quo (see Jost & Banaji, 1994).
Accordingly, justifications of violence can also be understood as examples of the legitimation of domination, as a naturalizing view of violence may reflect a need to see group-based dominance as justified (e.g., Paiva & Pereira, 2021).For example, Kunst et al. (2017) found in several countries that an orientation toward social dominance plays a central role in the relationship between social inequality and violence against social minorities.
Our findings have several interesting theoretical and practical implications for studies of the factors that legitimize social inequalities, particularly for research and theory on people's motives for legitimizing the way society is organized (e.g., Jost & Banaji, 1994).First, we contextualized the results within social psychological approaches to social inequalities (Costa-Lopes et al., 2013).Our results provide the first experimental evidence that people elicit elements that justify violence, describing it primarily as a naturalizing functioning of society.Second, the results also offer new insights into secondary victimization research (Tavares et al., 2022).Indeed, some social groups not only suffer violence against themselves, but are also visibly blamed for their own suffering.This study shows that people do this openly, especially against women.
Although the phenomenon of victim blaming occurs in all the contexts studied, it proved to be stronger in the context where the victims were women than in the other situations.
This tendency has been observed in other studies that have shown that people perceive women as culpable for their own suffering (e.g., Felson & Palmore, 2018;Bothamley & Tully, 2018;Canto et al., 2017).This stronger tendency to perceive women as responsible for their situation may be related to the sexist ideologies that permeate people's ideas about women, reflected in honor culture (e.g., Canto et al., 2017), rape myths (e.g., Shaw et al., 2017), gender roles (e.g., Bothamley & Tully, 2018), and the breakdown of established patriarchal patterns (e.g., Baldry et al., 2015).Our findings can also contribute to understanding the social implications of these representations by focusing on the need to consider people's tendency to legitimize the social situation of women in violent situations and to blame them for their victimization.
From a gender perspective (Bandeira, 2014), it is interesting to note that many participants view violence against women as an individual phenomenon, a victim problem.This phenomenon is consistent with recent research on domestic violence around the world.For example, analyzing 307 studies from 2000 to 2018, Sardinha et al. (2022) found that in 2018, up to 492 million women between the ages of 15 and 49 were victims of intimate partner violence at least once.In Brazil, the rate of this type of violence is also alarmingly high.In 2019, in addition to unreported cases, there were 1,246 homicides of women in households and 50,056 homicides of women between 2009 and 2019 (Atlas da Violência, 2021).Our findings can also be interpreted in the context of asymmetrical power relations, in which women are institutionally devalued relative to men.This devaluation anchors various structural relations of violence against women, thus perpetuating gender hierarchy in society (Pratto et al., 2006).In this perspective, victim blaming reinforces this hierarchy (see Paiva and Pereira, 2021;Ruggiero, 2020).
Perhaps the most dramatic result is the lower blame of the offender when the victim was black.As highlighted in Study 1, this result corroborates studies that have indicated that when the victim is black, the tolerance for police violence is pervasive and widely tolerated across several cultural contexts (Johnson & Lecci, 2020;Lima, et al., 2018;Silva et. al., 2018).Although we did not manipulate the offenders' group membership, this study elucidated this phenomenon by showing that the legitimation of violence against blacks can go beyond police contexts, reflecting a general motivation in people to minimize the offenders' responsibility for their abuse behavior.This motivation has already been described in studies of police violence against black people (e.g., Johnson & Lecci, 2020), but here we documented for the first time the presence of an inhibition in evocation that leads people to absolve the perpetrators of aggressive acts against black individuals.This may indicate not only tolerance of violence against black people, but also acceptance of the acquittal of offenders.
Although we could not verify any specific pattern of justification for violence against gay individuals, we found that the participants tended to explain this type of violence by evoking the same pattern of motives attributed to violence in general.As shown by Banhs and Branscombe (2011), the legitimacy of discrimination against gays increases the likelihood of heterosexual men engaging in verbal attacks, and the effect of legitimizing "gay bashing" is mediated by lower collective guilt.Furthermore, Bahns and Crandall (2013) questioned whether heterosexuals who endorse social inequality present greater discrimination against gay individuals when they perceive them as a threat to the social hierarchy.They found that the ideologies that legitimize the prevalence of the hierarchy of status, in which heterosexuals are at an advantage over gay people, allow the persistence of prejudice and discrimination against gay individuals.Furthermore, beliefs about social inequality were used by heterosexual participants to legitimize discrimination against gays when status positions between straight and gay individuals are threatened.
Despite the diverse contributions of these two studies, they have some limitations and suggest some future directions for further research.They are subject to the usual limitations inherent in research among university students, which may be a less pressing concern in the present case because of the social relevance of studying violence against participants' task was to write down the reasons they believed could explain the occurrence of the violence described in this situation.The purpose of this study was to analyze whether individuals gave reasons that could legitimize violence against minorities.Because individuals are motivated to justify society as it is (e.g.,Jost, 2019), we predicted that individuals would use at least three types of explanations to justify Revista Interamericana de Psicología/Interamerican Journal of Psychology 37. As in Study 1, participants tended to perceive violence as systemic more than as victim blaming (b = 0.82, SE = 0.09, p = .001,d = 1.00), and then offender blaming (b = 1.07,SE = 0.06, p = .001,d = 1.54).They blamed the offender less than the victim (b = 0.25, SE = 0.09, p =.001, d = -0.43).The main effect of the participants' focus was significant, F (1, 209) = 5.01, p = .03,η² = 0.02.Participants who were instructed to answer according to their own opinion gave more invocations (i.e., the sum of the average scores of invocations between categories) than those who answered according to the society's opinion.The main effect of the target was also significant, F (3, 209) = 3.01, p = .03,η² = .04.Evocations were lower in the condition of violence against blacks than in the control (b = 1.17,SE = 0.06, p = .01,d = -0.58),violence against women (b = 0.14, SE = 0.06, p = .02,d = -0.51),and violence against gays (b = 0.13, SE = 0.06, p = .03,d = -0.48).There
Within each section of line or column, means with distinct letters are significantly different at p <.05 (Least Significant Difference).

Table 2
Estimated marginal means and standard errors (in parenthesis) of the reasons for the violence in each experimental condition (Study 2)