ROBERTO POSADA GILÈDE
Universidad Nacional de Colombia, Bogotá

Correspondence concerning this article should be addressed to Roberto Posada, e-mail: roposadagi@unal.edu.co.
Department of Psychology, Universidad Nacional de Colombia, Cr. 30 No. 45-03, Building 212, Bogotá, Colombia.

SCIENTIFIC RESEARCH ARTICLE
RECEIVED: 2 MARCH 2012 - ACCEPTED: 29 OCTOBER 2012

Abstract

Relations between exposure to violence and moral reasoning were explored. Ninety-six participants aged 6-16 years evaluated the acceptability of stealing and causing physical harm in a situation of vengeance. Self-reports of exposure to violence were collected from each participant. Findings indicated that previous exposure to violence is related to moral reasoning in contexts of vengeance. Participants who reported having witnessed more violence, especially against family members, evaluated causing physical harm more positively, provided justifications entailing retaliation more frequently, and offered fewer reasons related to the conventional and personal realms, in a context of vengeance. Moreover, witnessing particular violent events was found to be positively correlated with judgments justified with reasons involving retaliation. These results suggest that participants think of vengeance as a way to restore justice (a moral issue) through expiatory sanction.

Keywords: vengeance, exposure to violence, moral reasoning.

Resumen

Se exploraron las relaciones entre la exposición a la violencia y razonamiento moral en 96 participantes de 6 a 16 años de edad, quienes evaluaron la aceptabilidad de robar y causar daño físico en una situación de venganza. Autorreportes de exposición a la violencia fueron recolectados para cada participante. Los resultados indicaron que la exposición previa a la violencia está relacionada con el razonamiento moral en contextos de venganza. Los participantes que reportaron haber sido testigos de más violencia, especialmente contra miembros de su familia, evaluaron causar daño físico más positivamente, proveyeron justificaciones que abarcaban retaliación más frecuentemente y ofrecieron menos razones relacionadas con los ámbitos convencional y personal, en contextos de venganza. Además, se encontró que ser testigo de eventos particulares de violencia está correlacionado positivamente con justificaciones de juicios que abarcan razones de retaliación. Estos resultados sugieren que los participantes piensan sobre la venganza como una forma de restaurar justicia (un asunto moral) a través de sanciones expiatorias.

Palabras clave: venganza, exposición a violencia, razonamiento moral.

Resumo

Exploram-se as relações entre a exposição à violência e o racionamento moral em 96 participantes de 6 a 16 anos de idade, os que avaliaram a aceitabilidade de roubar e causar dano físico em uma situação de vingança. Autorrelatórios de exposição à violência foram coletados para cada participante. Os resultados indicaram que a exposição prévia à violência está relacionada com o racionamento moral em contextos de vingança. Os participantes que relataram ter sido testemunhas de mais violência, especialmente contra membros de sua família, avaliaram causar dano físico mais positivamente, deram justificativas que abrangiam retaliação mais frequentemente e ofereceram menos razões relacionadas com os âmbitos convencional e pessoal, em contextos de vingança. Além disso, constatou-se que ser testemunha de eventos particulares de violência está correlacionado positivamente com justificativas de juízos que abrangem razões de retaliação. Esses resultados sugerem que os participantes pensam sobre a vingança como uma forma de restaurar justiça (um assunto moral) por meio de sanções expiatórias.

Palavras-chave: vingança, exposição à violência, racionamento moral.

MILLIONS OF people are currently living in contexts of violence. The ubiquity of internal armed conflicts around the world has forced a great number of people to leave their homes and towns in order to protect themselves from the numerous violations of human rights. The United Nations (UN) reported that there were 33,924,476 uprooted people by the end of 2010; of those, 14,697,804 were internally displaced persons (UNHCR, 2011a). Many of these people have witnessed abductions, torture, and killings perpetrated on members of their families and their communities. They usually live in miserable conditions since they had to abandon all that they had and most of them have minimal support, if any, from their government. Also, the UN stated that until May of 2011, the Colombian government had reported 3.7 million displaced persons. This number was challenged by a renowned Non-Governmental Organization (CODHES) that stated that there were over 5 million individuals in a situation of forced displacement (ACNUR, 2011b).

Although few would disagree that this, as well as other kinds of violence (e.g., crime, gang activity, chronic discrimination), has an effect on people's psychological well-being (for comprehensive reviews on this topic see Fremont, 2004; Horn & Trickett, 1998; Jenkins & Bell, 1997), we know relatively little about the social cognitive development of children growing up in violent contexts. One aspect of their development that is likely to be vulnerable under conditions of chronic violence is their moral understandings.

It has often been shown that people living in the midst of violence are at a higher risk of displaying aggressive behaviors (e.g., Barkin, Kreiter, & DuRant, 2001; Brook et al., 2003; Farver & Frosch, 1996; Guerra, Huesmann, & Spindler, 2003; Miller, Wasserman, Neugebauer, Gorman-Smith, & Kamboukos, 1999). However, the role moral reasoning plays in this relation is less known. Therefore, our objective is to begin to understand in more specific and detailed ways how people's experiences of violence are related to what they think about conflict resolution, how they coordinate their needs and emotions with their concepts about fairness, rights, and their concerns with others' well-being. The present study is another step in that direction. This study expands the results of a broader project (Posada & Wainryb, 2008) whose main goals were to look into whether children growing up in the midst of violence developed moral concepts and how they applied concepts bearing on justice and welfare in situations concerning survival and revenge. The goal of this study was to explore whether there is a relation between identifiable aspects of exposure to violence and moral evaluations in a context of vengeance. More specifically, the idea was to explore whether having been chronically exposed to violence and, particularly, having witnessed violent events is related to positive evaluations of stealing and causing physical harm in contexts of vengeance. The study also explores the relations between closeness to people that have been victims of violence and/or that have victimized the participant (i.e., close individuals like family members, or other non-close individuals) and justifications entailing retaliatory components for act evaluations.

The study focuses on situations of vengeance in order to explore those possible relationships, because studies have shown that children exposed to violence tend to endorse harm in contexts of revenge (Ardila-Rey, Killen, & Brenick, 2008; Posada & Wainryb, 2008). In addition, it is important to specify what aspects of the exposure to violence are the most salient for moral evaluations, and, consequently, what particular aspects might be informing moral reasoning and behavior.

Furthermore, findings from several studies have shown that children's exposure to violence was related to aggressiveness, and antisocial behavior (e.g., Barkin et al., 2001; Farver & Frosch, 1996; Miller et al., 1999). Interestingly, high rates of maltreatment by family members have been found among children who reported higher levels of exposure to violence in their communities, which is related to the presence of externalizing behaviors (Lynch & Cicchetti, 1998). This finding suggests that children with high rates of exposure to violence are more likely to be maltreated by relatives and that these children usually display aggressive or delinquent behaviors later on. In this sense, one could expect many children and adolescents who have experienced violence chronically to be more prone to justify the use of aggressive means for solving conflicts and hence to approve retaliation.

Although aggression, antisocial behavior, and externalizing behaviors in general are not the same as the socio-moral evaluation of vengeance, and cannot be reduced to it, they are highly likely to be related since intentions to use violence are commonly associated with trying to restore justice or re-establish respect (Barkin et al., 2001; Cooney, 1998; Gilligan, 1997; Tedeschi & Felson, 1994). As an illustration, in a recent study with sixth-grade students from a violent area in Georgia, Barkin and colleagues (2001) found that children's exposure to violence was strongly related to their intentions to engage in what they called "moralistic violence," which is defined as "an expression of moral grievance, an attempt to right a wrong, or to protect an honor code" (p. 777). Although the results of that study showed that intentions to use violence were related to previous exposure to violence, the reasons for doing so were not investigated, nor were their possible relations to specific aspects of exposure to violence, such as the kind of violence experienced and the closeness to the victim or the perpetrator of the violent events experienced. The present study addresses that question in a sample from a country at war, and also explores the relations between specific aspects of the violent events experienced (i.e., exposure to violence as witness or as victim, closeness to the victim or to the perpetrator, and specific violent events) and the reasons for justifying the use of violence in a context of vengeance.

Another characteristic of children growing up in contexts of violence and, more specifically, in contexts of armed conflict is that they are full of memories about negative events that happened when they experienced violent events, many times committed against people they knew and/or by people known to them and their families (Wessells, 1998). Those memories often go along with feelings of extreme fear and/or hate and with desires for vengeance, which families might strengthen by accounts passed down across generations.

Witnessing violence toward family members has been found to be a particular factor associated with psychological well-being. Goldstein, Wampler, and Wise (1997) conducted a study with displaced children in Bosnia (ages 6-12) and found that the type of experience that consistently and dramatically affected these children was witnessing violence against members of the nuclear family. However, it is still uncertain if a relation exists between exposure to violence (or traumatic war-related events) and social thinking in contexts where dire need and feelings of hatred and fear are commonly present.

In one of the few studies conducted with people living in a country at war, Angel, Hjern, and Ingleby (2001) interviewed Bosnian refugee children and their parents, finding that there was a relation between exposure to violence and children's aggressiveness. In the present study, the purpose was to explore whether there is a relation between having experienced different types of violence and the approval of retaliatory acts in a context of vengeance against a past aggressor.

Research on moral development has shown that both aggressive and non-aggressive children think that it is morally wrong to behave in violent or hurtful ways in the absence of provocation (Astor, 1994; Astor & Behre, 1997). However, aggressive children justify violent and harmful responses in situations that involve provocation. Yet, research on the moral development of children growing up in violent countries has been scarce. In a number of studies, it has been found that children and adolescents that grow up in violent contexts show concerns about situations of injustice and lack of well-being, and, in general, think about moral issues (e.g., Ardila-Rey et al., 2008; Cairns, 1987; Lorenc & Branthwaite, 1986; Posada & Wainryb, 2008; Punamäki, 1996).

Colombian displaced children and adolescents who have grown up in the midst of continuous violence constitute a particular kind of population living in non-normative conditions. The research reviewed in the previous section indicates that they are likely to suffer from post-traumatic stress, depression, and externalizing symptoms. From a broader study (Posada & Wainryb, 2008), we now know that these children have also developed moral concepts bearing on justice and well-being. Furthermore, we found that the application of their moral concepts is not homogeneous across contexts (e.g., survival and revenge). In a few words, almost all children and adolescents stated that it was wrong to steal or cause physical harm in conditions of dire need and when no consideration was present (baseline condition). However, in the revenge situation, a sizable minority stated that it was all right to steal and cause physical harm, and only a 27.5% of those who made negative evaluations justified them with concepts bearing on justice, rights, or well-being (i.e., moral reasons). Therefore, in this context concerns with retaliation against those who had harmed them and their families seem to take priority; children and adolescents' reasoning in this regard suggests that in those cases, stealing and inflicting harm are experienced as ways to effect retribution and restore justice. An additional finding was that adolescents evaluated acts of revenge more positively than children. One of the questions raised in the project was whether these children and adolescents' moral evaluations of stealing and causing physical harm in situations of vengeance are related to specific features of their exposure to violence.

To investigate this issue, some of the different assessments employed in that broader study (Posada & Wainryb, 2008) were used.  The study explored the relation between assessments of children's evaluation about harmful or unjust acts in hypothetical situations where the possibility of retaliation is explicit (vengeance) and self reports of exposure to violence from each participant (My ETV, Selner-O'Hagan, Kindlond, Buka, Raudenbush, & Earls, 1998). In general, it was expected to find a relation between exposure to violence and act evaluations in the situation of vengeance since the results of our previous study showed that the variability centered mainly on this condition. That is, almost all of the participants considered stealing or causing harm was wrong in the survival and the baseline conditions, but several made positive or mixed evaluations in the revenge condition.

In addition and more specifically, it was expected to find that those who reported exposure to more violence in general (i.e., as victims and witnesses), and those who reported having witnessed more violence, in particular, would provide justifications to their socio-moral evaluations involving retaliatory reasons more often than those who reported having been less exposed. This hypothesis was based on the previous findings presented above (e.g., Buka, Stichick, Birdthistle, & Earls, 2001; Lynch & Cicchetti, 1998; Richters & Martinez, 1993) showing that children who have witnessed violence chronically display high levels of externalizing behaviors that are thought to be related to retaliatory intentions. It was also expected that participants who have been victimized by family members would provide more justifications to their act evaluations entailing retaliatory reasons than those who have been victimized by non-family members. This hypothesis was based on the work of Lynch and Cicchetti (1998), according to which children who reported more exposure to violence in their communities were also found have been maltreated by family members and displayed more externalizing behaviors.

Method

Participants

Participants in this study were 48 Colombian children (ranging from 6.0 to 9.8) and 48 adolescents (ranging from 13-16.9), equally distributed by gender, and in a situation of geographical displacement. All participants were from a working-class background and live in different neighborhoods in Bogotá (most of them in Usme) where a large number of displaced families are usually found.

Instruments

Socio-moral evaluation. For the purpose of assessing participants' moral evaluations in a context of vengeance, a hypothetical condition was designed with the assistance of local workers, in order to make the scenarios meaningful to the participants. In the vengeance condition, the main character recognizes the other character as the member of the group responsible for her/his family's situation of displacement (for the full vignette see Posada & Wainryb, 2008).

Participants were told that the main character engaged in (a) stealing and (b) causing physical harm to the other character. For each of the alternatives participants were asked to evaluate the main character's behavior. To allow for generalizability across content areas and to maintain the participants' interest, three comparable examples of the stimuli were designed (dealing with a bike, a jacket, and a boom-box).

Exposure to violence. For a more precise and detailed view of the kind and amount of violence that participants had experienced, a short form of a structured interview developed in the United States to assess exposure to violence among urban youth (Selner-O'Hagan et al., 1998) was used. My Exposure to Violence (My ETV-SF) assesses the type and extent of violence experienced as witness and as victim, as well as who the victim was when the respondent was a witness, and who the perpetrator was when the participant was the victim. My ETV-SF (version 3) includes questions about 24 violent events (e.g., being beaten up, being attacked with a weapon, seeing someone killed) from which we intentionally omitted 7 events (serious accidents, sexual abuse -omitted originally by the scale authors- natural disasters, and suicide. The first three types of events had questions as victim and as witness), since these were not relevant for the original study's purposes. As reported by Selner-O'Hagan and colleagues (1998), test-retest reliability ranged from .75 to .94, and internal consistency indices as measured by Cronbach alphas ranged from .68 to .93; in previous research, My ETV scores were linked to age, gender, race/ethnicity, violent offending, and neighborhood crime in theoretically predicted ways.

Procedure

IRB (Institutional Review Board) guidelines for working with vulnerable populations were strictly followed and IRB approval was obtained. The director of the Cooperemos D.P.S. agency (a non-governmental agency in Bogotá that assisted displaced people), a well known and trusted figure among displaced children and adolescents assisted by this agency, informed them about the study, introduced the author, and, together with a local social worker, who served as research assistant, invited them to participate. Families that accepted children's participation dropped by the main office of the Agency at pre-arranged times. After signed consent and verbal assent procedures were carried out, participants were interviewed. The first part of the interviews (i.e., socio-moral evaluations) was tape recorded and later transcribed for analysis. The social worker interviewed participants with the My ETV-SF.

Scoring and Reliability

Socio-moral evaluation. Act evaluations were first scored on a 3-point scale, where okay=1, mixed=2, and not okay=3. Next, justifications were scored using categories adapted from previous scoring systems in moral development research (Davidson, Turiel, & Black, 1983) such as: welfare (e.g., "because the other person would feel really bad"); justice and rights (e.g., "because that boom box does not belong to her and that is not fair"); rules and punishment (e.g., "because then he will be caught by the police"); prudential (e.g., "because if Inés hurts that girl then she will find her later and will do something worse to Maria"); and retaliatory (e.g., "because he caused harm to Julio's family and now he has to pay in the same way"). For the purposes of this project, these categories were then collapsed into three broader categories: moral (includes welfare, and justice and rights), non-moral (includes rules and punishment, and prudential), and retaliatory (includes retribution).

Coding reliability was assessed through the independent recoding of 20% of the protocols by a second judge. Interjudge agreement ranged from 88.06% for justifications (kappa=.86) to 99.16% for act evaluation (kappa=.95).

Exposure to violence. My ETV (Selner-O'Hagan et al., 1998) scores for witnessing, (including having heard about violent events against known people) victimization, and total exposure to violence were calculated by counting the items endorsed for each subscale and estimating the total exposure to violent events as a sum of the frequency of occurrence for each item. Identification of the victim and of the perpetrator was assessed using the following categories: family member, someone known and close, someone known but not close, and someone unknown.

Results

The main goal of this project was to explore whether exposure to violence was related to children and adolescents' socio-moral evaluations of stealing and causing physical harm in a context of vengeance. In addition, the relation between participants' justifications for their act evaluations in a hypothetical context of vengeance and their self-reports on the kind of violent events experienced (as witness or as victim) was examined. The third issue investigated was the relation between participants' justifications for their positive evaluations of harm in a context of vengeance and exposure to violence, that is, whether their reasons for the approval of stealing and causing physical harm were related to their exposure to violence as victims of closely related individuals and as witnesses of violent acts against closely related individuals.

Some of the general results about exposure to violence, which were originally presented elsewhere (Posada & Wainryb, 2008), are presented below  in order to provide the reader with a context regarding the kind of violence reported as experienced by this sample. These are followed by more specific results not used in the broader study and entailing information about the kind of relationship existing between the participant and the victim of the act witnessed, and the kind of relationship existing between the participant and the perpetrator of the violent acts experienced, as well as the frequency of violent acts experienced. Participants' moral evaluations of the courses of action presented in the hypothetical situation are then presented, followed by the result of the main analyses that entail the relation between participants' exposure to violence and their moral evaluations.

Participants' self-reports on exposure to violence as victim obtained low scores (children M=.90; adolescents M=1.40) and exhibited restricted variance (see Posada & Wainryb, 2008). The most common score for participants' self-reports about victimization was zero, out of a possible maximum score of 5 (children 42% and adolescents 35%), although almost half of adolescents (46%), in contrast to 25% of children, reported having been victims of two or more kinds of violent acts. With respect to exposure to violence as witness the picture is different. Its mean is higher (children M=4.94; adolescents M=6.50) and its variance is not as restricted as in the previous case; the most common scores were 4 (19%) and 5 (19%) for children and adolescents respectively (11 was the maximum possible score). Regarding the total score for exposure to violence, the difference between children and adolescents is more notorious and is more clearly shown by the mode (for children=5, for adolescents=9). Although the total scores for victimization, witnessing, and general exposure to violence provide a useful description of the exposure to violence that this sample had experienced, it is also relevant to provide a more detailed description regarding the specific violent events experienced by participants.

The violent acts to which participants have been most exposed are: having heard gunfire nearby (92%); having seen someone beaten up (83%); having seen and/or heard about someone getting killed (74%); having seen someone be chased (63%); and having seen someone be seriously threatened (61%). Although with a lower percentage, the results for other types of violent experiences are also important. For example, participants who have witnessed very severe acts of violence such as seeing and/or heard about someone getting shot (49%); finding a dead body (48%); seeing someone attacked with a weapon (39%), and seeing someone be shot at but not wounded (31%) (See Posada & Wainryb, 2008). Adolescents reported to have seen someone get beaten up more often than children (F(1, 92)=14.56, p<.001) and females more often than males (F(1, 92)=6.98, p<.01). In addition, adolescents reported having seen someone be shot at more often than children (F(1, 92)=14.5, p<.001); male adolescents were the participants most exposed to this particular violent act (F(1, 92)=5.55, p<.021).

The acts most commonly experienced by participants as victims were having been hit, slapped, or punched (46%); having been chased (29%); and having been seriously threatened (28%). Four participants reported having been shot at (without having been wounded), and one having been shot. Males reported having been chased more often than females (F(1, 92)=7.67, p<.007), and male adolescents reported having been attacked with a weapon more often than children and female adolescents (F(1, 92)=4.39, p<.04). More adolescents (81%) than children (48%) reported having been told that someone known had been killed (F(1, 92)=12.86, p<.001), and also more adolescents than children reported having been told that someone known had been shot (F(1, 92)=9.23, p<.003).

Table 1 shows percentages of participants reporting who the victim of witnessed violent acts was and who the perpetrator was when they were the victims. With respect to the self-report about who the victim of witnessed events was, most violent acts that participants witnessed had occurred to strangers. It is interesting, however, to notice that 52% of those who reported having seen someone get beaten up (41 participants) said that this had happened to someone considered close to them (i.e., family member or close friend), 42% (30 participants) reported having seen and/or heard that someone close to them had been killed, 45% (26 participants) reported having seen someone close being seriously threatened, 51% (24 participants) reported having seen and/or heard someone close getting shot, and 37% (23 participants) reported having seen someone close being chased.

Regarding who the perpetrator was when violence was experienced as victim, more than  half (71%) of the participants who reported having been beaten up, and 30% of those who were seriously threatened said that they were victimized by people they considered close to them (see Table 1).

These data about the victim and the perpetrator were also categorized differently based on the assumption that violence against, or perpetrated by, family members is likely to be more impacting than that perpetrated by friends, and hence could provide some information on its own (rather than combined with that about friends) when investigating its relation to socio-moral evaluations. Therefore, this variable was also categorized in two groups: (a) exposure to violence against, or by family members, and (b) exposure to violence against or by non-family members. Furthermore, in order to test one of the hypotheses for the present study (i.e., whether being exposed to violence against, or perpetrated by, family members is related to justifying socio-moral evaluations with reasons entailing retaliation) this categorization was necessary. Although 40 participants reported not having witnessed any kind of violence against family members, 56 participants reported having experienced at least one type of violent act against someone in their family. Twenty-seven participants reported having witnessed one type of violent act (e.g., seeing someone from their family be threatened, or seeing a family member get beaten up), and 18 reported having witnessed three or more types of violent acts against family members (see Figure 1-A). With respect to witnessing violence against non-family members, only two participants reported not having experienced any violence of this type. Twenty-one participants reported having experienced two types of violent acts, 19 participants reported having witnessed three types, and 36 participants reported four or more types of violent acts against non-family members (see Figure 1-B). Regarding who the perpetrator of violent acts was when participants were the victims, 25 participants reported having been victimized by family members at least once (see Figure 1-C), whereas 49 reported having been victimized by others different from family members, and 22 of them in at least two different types of events (see Figure 1-D). It is important to note that no participant reported having been victimized by a family member for more than two different violent acts, and only five of them reported having experienced two different kinds of violent acts by family members.

With respect to the socio-moral evaluations, it is important to note that in the hypothetical condition entailing vengeance, 18 participants (19%) said it was okay to steal and an additional ten (10%) gave mixed judgments. As far as causing physical harm is concerned, 20 participants (21%) said it would be okay and another three gave mixed evaluations. It is also worth noting that 16.7% of adolescents said that it would be all right to cause harm in the vengeance condition, compared to only 7.3% of children (F(2, 188)=2.43, p<.09).With respect to justifications for socio-moral evaluations, Table 2 shows the distribution of participants who justified their evaluations with moral (i.e., justice, rights and/or welfare), non-moral (i.e., social conventions, avoidance of punishment, and prudential), and retaliatory (i.e., retaliation or vengeance) reasons. It is important to mention that in the vengeance condition there was not a single unelaborated justification, as could be observed if the numbers of the correspondent cells were added (e.g., children's justifications for stealing in the vengeance condition: 48+29+23=100).

Regarding the main analyses, first, correlations were computed to examine the relationship between evaluation of stealing and causing physical harm in a context of vengeance, and exposure to violence. Table 3 shows the correlations between exposure to violence as witness, victim, and in general (total) and the act evaluations in that condition.  The analyses yielded two significant correlations between act evaluations and exposure to violence in a vengeance condition when the course of action referred to physical harm. Analyses showed that there is a significant relation between exposure to violence as witness and the evaluation of causing physical harm (r(94)=-.25, p<.007, 1-tail). That is, the higher the score in exposure to violence as witness, the lower the negative evaluation of physical harm in a condition of vengeance. In other words, participants who have witnessed more types of violence tended to evaluate causing physical harm in a context of revenge as positive. The other significant correlation found was that between physical harm and the total score of exposure to violence (r(94)=-.25, p<.007, 1-tail). The higher the total score on exposure to violence the lower the negative evaluation of causing physical harm in a context of vengeance; participants who reported having experienced a larger variety of violent events also evaluated positively acts of physical harm in a context of vengeance. Point biserial correlations were computed between justifications for the act evaluations in a vengeance condition and exposure to violence. As shown in Table 4, the analyses yielded four significant correlations. One was between exposure to violence as witness and the justification of physical harm entailing revenge components (r(94)=.23, p<.013, 1-tail). Those who reported more exposure to violence as witness also tended to provide justifications for their socio-moral evaluations entailing retaliatory reasons. Exposure to violence as witness was significantly correlated with justifying causing physical harm with non-moral components (r(94)=-.24, p<.009, 1-tail). That is, the higher the exposure to violence as witnesses the lower the tendency to justify their act evaluations with non-moral reasons. In addition, total exposure to violence was significantly correlated with justifications of physical harm entailing retaliatory components. Those that had experienced a broader variety of violent events tended to provide more justifications comprising retaliatory components in their evaluations of causing physical harm in a condition of vengeance. Finally, total exposure to violence was significantly associated with justifications for act evaluations based on non-moral reasons. No significant correlations were found between evaluations.

Subsequently, point biserial correlations and also chi square tests for independence were computed to explore the relations between closeness to the victim and to the perpetrator, when witnessing violence and when being victimized, respectively, and the kind of justifications participants provided in a vengeance condition when evaluating stealing and causing physical harm.

First, point biserial correlations were computed between the types of justifications for act evaluations (i.e., moral, non-moral, and retaliatory) and the number of different violent events that each participant had witnessed against (a) a family member or (b) a non-family member. The analysis yielded a positive significant correlation (r=.30) between having witnessed violence against family members and justifications comprising retaliation for the evaluation of physical harm in the vengeance condition (see Table 5). The higher the number of types of violent events experienced against family members, the higher the likelihood of providing justifications referring to retaliation in a context of vengeance. In addition, a significant correlation (r=.28) was found between a broader variety of violent events witnessed against family members and non-moral justifications. The higher the number of types of violent events witnessed against family members, the lower the probability of providing a justification comprising non-moral components (i.e., justifications based on avoiding punishment, following conventional rules, or avoiding later negative consequences). No significant correlations were found between having witnessed violent events against non-family members and justifications for act evaluations, or between having been a victim of violence by family or non-family members and the kind of justifications offered.

Finally, a chi square test for independence was computed to explore if there was a relation between the kinds of justifications participants provided for their act evaluations in a context of vengeance and their exposure to specific violent events perpetrated by or against close individuals (i.e., closeness). Violence could be experienced (a) only as witness (against people considered close to the participant), or (b) only as victim (by people considered close to the participant), or (c) as both witness and victim, or (d) they might not have experienced the event. Justifications categories were grouped as (a) moral, (b) non-moral, and (c) retaliatory. The types of violent events were combined as follows: (a) having been chased or having seen someone be chased, beaten up, and threatened and (b) having been or seen someone attacked with a weapon (two items, one referring to fire arms and the other to cold weapons), get shot, and seeing someone killed. Note that the categories were formed taking into account the distribution for all the items' responses and the level of seriousness of violent acts. Both distribution of responses and seriousness of violence were similar within each category. These analyses did not yield any significant relationship between the kind of justifications for the act evaluations and the specific kind of exposure to violence against people considered close to the participants. It is highly likely that the results were affected by the low statistical power.

Discussion

The results of the present study suggest that experiences of violence are related to socio-moral reasoning, and more specifically to the way in which children and adolescents who have been victims of displacement in Colombia evaluate causing physical harm to past aggressors in a hypothetical situation. Findings indicate that there are some specific aspects of their previous experiences of violence that seem to be informing participants' act evaluations more than others. First, analyses showed that having been exposed to several types of violent events in general, and as witness in particular, is related to a more positive evaluation of causing physical harm to someone who belongs to the group responsible for a family displacement in a hypothetical situation. That is, participants who have experienced more types of violent events (as witness and as victims), as well as those that have witnessed more types of violent events, tended to judge causing physical harm to others in a context of vengeance as all right. The expected finding about the relation between being a witness of violence and a positive evaluation of retaliation is in line with results from other studies indicating that aggressive behaviors displayed by children who have been exposed to violence are usually related to retaliatory intentions (e.g., Barkin et al., 2001; Buka et al., 2001; Farver & Frosch, 1996; Miller et al., 1999). It is worth highlighting again that although aggressive behaviors are not the same as moral thinking and that aggressive behaviors are not always necessarily related to retaliatory reasons (e.g., as in instrumental aggression), it is reasonable to expect retaliatory wishes and positive evaluations of aggressive acts against past aggressors in order to make the situation fair, at least in terms of making the other feel the same thing they had felt (as some of the participants stated), on the part of people who have chronically witnessed violent events. We know that many of these events were committed against others they knew and were close to them, and that those violent events are not usually punished by the country's legal system. It is important to note that there was no significant relation between being exposed to violence and stealing.

This finding, in addition to the findings of the broader project (Posada & Wainryb, 2008) about judging it wrong to hurt or steal from other people in situations of extreme need or just to fulfill a desire, supports the idea that approval of causing physical harm seems to be highly informed by socio-moral reasons, instead of being a notion that is generalizable across contexts and/or a habit that has been learned. These children who have experienced extreme violence mainly as witnesses do not accept violence and do not evaluate aggressive behaviors as something correct in all or most situations. The results of the present study strongly suggest that witnessing extreme violence does not make people amoral or immoral. The fact that participants did not evaluate stealing in the revenge condition as okay indicates that moral concepts support their evaluations. This is so even when they can also benefit by gaining something material (in the case of stealing) in a context of miserable living conditions.

Thus, analyses on whether exposure to violence is related to this moral logic (reasons or justifications for socio-moral evaluations) show that witnessing violent events is associated with the kind of justifications that participants provided. Participants who had witnessed more violent events evaluated causing physical harm to a previous aggressor for retaliatory reasons more positively. Furthermore, participants who had witnessed more violent events tended to use fewer non-moral reasons (i.e., reasons related to personal jurisdiction, social conventions and pragmatic issues) to justify their evaluation about causing physical harm in the revenge condition. This suggests that participants take vengeance as a moral issue, since a justice component is involved when they justify their evaluations with the idea of paying back in order to restore justice. In other words, when these children were facing that complex situation (i.e., causing physical harm in a revenge condition), they prioritized components of justice and fairness (in a morally paradoxical way since they are accepting to hurt another person) over conventional and prudential aspects, insinuating that vengeance is understood as a moral issue. Those components are related to the search for equality and not for equity in most of the cases. That is, they seek to make the situation as similar as possible to their own experience. In particular, they want the victim to "feel" the same thing that the main character felt when he/she was displaced and similar to what participants have felt since their own displacement (as a number of participants expressed it). What makes this reasoning paradoxical and contradictory in a way is that although vengeance can have some components of equity for those participants that accepted vengeance as a course of action, the reason set forth was not related to repairing the social bond but merely to making the other suffer similarly to the way they did.

As stated above, the finding on the relation between exposure to violence and act evaluations (i.e., judgments about stealing and harm) might involve a salient component: the level of closeness (i.e., kind of social relationship) with the victim of the violent events. As expected, analyses showed that witnessing violence against family members, in particular, is related to thinking about retaliation as the right thing to do and at the same time that conventional and prudential concerns are less important in this specific condition (vengeance). This finding suggests that the target of previous violent events experienced as witness also informs participants' evaluations of causing physical harm in a context of vengeance. However, the results of additional analyses involving the specific kinds of events experienced, where the level of closeness was divided in a different way (having witnessed violence against close individuals such as family members and other known individuals considered close, and having witnessed violence against known individuals but not considered close and strangers) did not yield any significant relation. This could mean that the level of closeness that informs the kind of socio-moral evaluation is specifically the one involving family members, rather than family members and other people considered close to the participants. Alternatively, it is possible that the kind of violent event is a key factor when informing an individual's thinking; however, there was not enough statistical power to examine this possibility. Regardless, we still can conclude that what seems to be informing participants' socio-moral evaluations is not only witnessing violence, in general, but also whether the victim of those violent acts was a family member or not.

For future studies, it would be important to explore whether some specific violent events are more salient than others when informing people's socio-moral thinking in specific contexts. For instance, having heard about violent events perpetrated against close individuals (a different type of witnessing violence) seems to be related to justifying retaliation. Wessells (1998) suggested that when hearing about tragic events intentionally committed against family members, stronger feelings involving fear, hatred, and shame usually appear and guide aggressive and violent behaviors prolonging cycles of violence. Exploratory analyses were conducted on the relationship between having heard about a close individual being shot or killed and justifications for act evaluation; results interestingly suggest that a relationship between these two variables exists (r(94)=.27, p<.008, 2-tails).

The second hypothesis tested in this project about the relationship between being victimized by family members and the approval of revenge against others was not supported. The hypothesis was based on the work of Lynch and Cicchetti (1998) and on research conducted by war journalists in Colombia (Bustos, 2000; Gonzalez, 2002) who reported that child maltreatment was associated to externalizing behaviors and to continuing cycles of violence. Although child maltreatment was not assessed directly, it was assumed that knowing the perpetrator when they were victimized, as reported by participants in My ETV, might be an indicator of child maltreatment. Even though the scores on self-reports of victimization were very low in general, being beaten up by family members was a common event reported by participants (43%). This may be an indicator of family maltreatment. The data used, however, might have provided an underestimation either of child maltreatment, or of reporting about it since participants could be scared about being separated from their parents if they reported those issues. Also, the results about the inexistence of a significant relation between victimization and moral reasoning might be an artifact of a restricted range for the scores on frequency of victimization. Other than the items regarding being chased, beaten up, and threatened, no other events related to victimization were reported and even fewer events committed by family members were reported. All these limitations could affect the results obtained. However, this finding might also be taken as consistent with most research on the general effects of exposure to violence, where it has been found that being a victim, as opposed to a witness, of violent events is associated with depression and internalizing rather than externalizing behaviors (Fitzpatrick, 1993; Martinez & Richters, 1993; Palacio, Sabatier, Abello, & Madariaga, 2001).

In conclusion, the results of the present study suggest that when these children and adolescents accept violence against others, they do it looking for justice, since the suffering they have gone through, due to a political conflict, is not seen as a reasonable motive for such damage. A significant number of displaced Colombian children evaluated causing physical harm in a context of vengeance as okay, thinking of it as a way to make the situation equal and to restore justice. They thought in moral terms when evaluating acts of vengeance. Importantly, the results show that witnessing violent events is negatively related to non-moral reasons. This suggests that vengeance is understood as a moral issue.

However, it is important to bear in mind that the results presented refer to correlational analysis and therefore are not suggesting causality. Different experiences in the lives of these people, their ways of coping with stress, their family relationships, their social support, among others, might interact and thus inform individuals' social evaluations too. In addition, exposure to violence is a multifaceted phenomenon with many components, such as the role of the individuals involved (i.e., victim, witness, and/or perpetrator), the frequency, the variety (i.e., many or few different violent events), and the kind of events. Although these experiences are part of their social knowledge, displaced people who have experienced extreme violence construct and apply moral concepts when evaluating social situations involving injustices and physical harm. Based on the finding about their acceptance of retaliatory acts against past aggressors, it is likely that they would also display aggressive behaviors, at least to restore justice.

References

ACNUR. (2011a). La historia de los refugiados en estadísticas. Retrieved from: http://www.acnur.org/t3/recursos/estadisticas/         [ Links ]

ACNUR. (2011b). Desplazamiento interno en Colombia. Retrieved from http://www.acnur.org/t3/operaciones/situacion-colombia/desplazamiento-internoen-colombia/         [ Links ]

Angel, B., Hjern, A., & Ingleby, D. (2001). Effects of war and organized violence on children: A study of Bosnian refugees in Sweden. American Journal of Orthopsychiatry, 71, 4-15.         [ Links ]

Ardila-Rey, A., Killen, M., & Brenick, A. (2008). Moral reasoning in violent contexts: Displaced and nondisplaced Colombian children's evaluations of moral transgressions, retaliation, and reconciliation. Social Development, 18, 181-209.         [ Links ]

Astor, R. (1994). Children's moral reasoning about family and peer violence: The role of provocation and retribution. Child Development, 65, 1054-1067.         [ Links ]

Astor, R. & Behre, W. (1997). Violent and nonviolent children's and parents' reasoning about family and peer violence. Behavioral Disorders, 22, 231-245.         [ Links ]

Barkin, S., Kreiter, S., & DuRant, H. (2001). Exposure to violence and intentions to engage in moralistic violence during early adolescence. Journal of Adolescence, 24, 777-789.         [ Links ]

Brook, D., Brook, J., Rosen, Z., De la Rosa, M., Montoya, I., & Whiteman, M. (2003). Early risk factors for violence in Colombian adolescents. American Journal of Psychiatry, 160, 1470-1479.         [ Links ]

Buka, S. L., Stichick, T., Birdthistle, I., & Earls, F. (2001). Youth exposure to violence: Prevalence, risks and consequences. American Journal of Orthopsychiatry, 71, 298-310.         [ Links ]

Bustos, A. (2000). Los niños de la violencia. Bogotá, Colombia: Intermedio.         [ Links ]

Cairns, E. (1987). Caught in crossfire: Children and the Northern Ireland conflict. Belfast, Ireland: Appletree Press.         [ Links ]

Cooney, M. (1998). Warriors and peacemakers: How third parties shape violence. New York: New York University Press.         [ Links ]

Davidson, P., Turiel, E., & Black, A. (1983). The effect of stimulus familiarity on the use of criteria and justifications in children's social reasoning. British Journal of Developmental Psychology, 1, 49-65.         [ Links ]

Farver, J. A. & Frosch, D. L. (1996). L.A. stories: Aggression in preschoolers' spontaneous narratives after the riots of 1992. Child Development, 67, 19-32.         [ Links ]

Fitzpatrick, K. M. (1993). Exposure to violence and presence of depression among low-income African-American youth. Journal of Consulting and Clinical Psychology, 61, 528-531.         [ Links ]

Fremont, W. (2004). Childhood reactions to terrorisminduced trauma: A review of the past 10 years. Journal of the American Academy of Child & Adolescent Psychiatry, 43, 381-392.         [ Links ]

Gilligan, J. (1997). Violence: Reflections on a national epidemic. New York: Vintage.         [ Links ]

Goldstein, R. D., Wampler, N. S., & Wise, P. H. (1997). War experiences and distress symptoms of Bosnian children. Pediatrics, 100, 873-878.         [ Links ]

Gonzalez, G. (2002). Los niños de la guerra. Bogotá, Colombia: Planeta.         [ Links ]

Guerra, N. G., Huesmann, L. R., & Spindler, A. (2003). Community violence exposure, social cognition, and aggression among urban elementary school children. Child Development, 74, 1561-1576.         [ Links ]

Horn, J. & Trickett, P. (1998). Community violence and child development: A review of research. In P. Trickett & C. Schellenbach (Eds.), Violence against children in the family and the community (pp. 103-138). Washington, DC: American Psychological Association.         [ Links ]

Jenkins, E. & Bell, C. (1997). Exposure and response to community violence among children and adolescents. In J. D. Osofsky (Ed.), Children in a violent society (pp. 9-31). New York: The Guilford Press.         [ Links ]

Lorenc, L. & Branthwaite, A. (1986). Evaluations of political violence by English and Northern Irish school children. British Journal of Social Psychology, 25, 349-352.         [ Links ]

Lynch, M. & Cicchetti, D. (1998). An ecologicaltransactional analysis of children and contexts: The longitudinal interplay among child maltreatment, community violence, and children symptomatology. Development and Psychopathology, 10, 235-257.         [ Links ]

Martinez, P. & Richters, J. E. (1993). The NIMH community violence project: II. Children's distress symptoms associated with violence exposure. Psychiatry, 56, 22-35.         [ Links ]

Miller, L. S., Wasserman, G. A., Neugebauer, R., Gorman-Smith, D., & Kamboukos, D. (1999). Witnessed community violence and antisocial behavior in high-risk, urban boys. Journal of Clinical Child Psychology, 28, 2-11.         [ Links ]

Palacio, J., Sabatier, C., Abello, Ll. R., & Madariaga O. C. (2001, April). Impact of political violence and displacement on Colombian children. Poster session presented at the biennial meeting of the Society for Research in Child Development, Minnesota.         [ Links ]

Posada, R. & Wainryb, C. (2008). Moral development in a violent society: Colombian children's judgments in the context of survival and revenge. Child Development, 79, 882-898.         [ Links ]

Punamäki, R. L. (1996). Can ideological commitment protect children's psychosocial well-being in situations of political violence? Child Development, 67, 55-69.         [ Links ]

Richters, J. E. & Martinez, P. (1993). The NIMH community violence project: I. Children as victims of and witnesses to violence. Psychiatry, 56, 7-21.         [ Links ]

Selner-O'Hagan, M. B., Kindlond, D. J., Buka, S. L., Raudenbush, S. W., & Earls, F. J. (1998). Assessing exposure to violence in urban youth. Journal of Child Psychology and Psychiatry, 39, 215-224.         [ Links ]

Tedeschi, J. T. & Felson, R. B. (1994). Violence, aggression, and coercive actions. Washington, DC: American Psychological Association.         [ Links ]

Wessells, M. G. (1998). The changing nature of armed conflict and its implications for children: The Graca Machel / UN study. Journal of Peace Psychology, 4, 321-334.         [ Links ]