English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
The present study examined the Colombian people's positions on forgiving perpetrators of offenses against women during the armed conflict that ravaged the country for more than a half a century. Two studies conducted in Bogotá, Colombia, before the peace accord (López López, Pineda Marín, Murcia León, Perilla Garzón, & Mullet, 2013, López López, et al., 2018) showed that a majority of participants (67% in the 2013 study and 52% in the 2018 study) were of the opinion that forgiveness should not be granted to former perpetrators of violence, irrespective of circumstances; that is, even if these people had not committed very severe crimes or had expressed true repentance. The present study was similar in spirit to these previous studies, but it examined specifically the willingness to forgive perpetrators of crimes against women, and it explored the relationship between willingness to forgive perpetrators and attitudes towards the peace process (e.g., attitude to peace dialogs in La Habana).
In Colombia, a long-running armed conflict had, since 1960, opposed left-wing insurgents (mostly members of the Fuerzas Armadas Revolucionarias de Colombia, FARC), the governmental army, and right-wing paramilitaries. As the insurgents secured more and more resources from the illegal drug trade that they partly managed, the conflict expanded nationally in the 80s (see the reports from Colombia's National Centre for Historical Memory, 2018). The violence receded, however, at the beginning of the current century, due to the implementation of the democratic security policy of Uribe's government. The insurgents gradually lost control of most of their territorial gains, the Colombian army secured the recovered areas, and the paramilitary groups were dismantled (Granada, Restrepo, & Vargas, 2009). From 1960 to 2016, millions of people have been displaced, have been killed, or have disappeared. As of today, 8,910,526 persons are officially been registered on the Colombian National Register of Victims. Among them, 165,271 were victims of homicides, were killed through other circumstances, or have disappeared (Registro Único de Víctimas, 2019). Severe human rights violations have been perpetrated by all parts implicated in the conflict (Human Rights Watch, 2017).
The peace accord that was signed in 2016 provided adequate compensation of all the victims, full disclosure of information, and protection against possible aggression from all participants in the transitional justice process. It also stipulated that measures should be implemented in order to secure, as far as possible, coexistence and reconciliation between all citizens, in order to prevent the resumption of conflicts. Following the signing, the press reported several meetings between victims and perpetrators as, for example, in the small town of Bojayá in which, in 2002, the FARCs killed or severely hurt more than 200 villagers, among them many women: Ex-members of this group, who admitted to being the authors of the massacre, met the victims and the victims' families, expressed repentance, and begged for forgiveness (New York Times, 2016; Semana, 2016a).
Positive relationships among people living and working in the same areas is a condition of productive collaboration. After a civil conflict, a country must not only rebuild its material infrastructure in all the damaged areas, but it must also rebuild trust and cooperation among all its citizens (Espinosa, et al., 2016; Mukashema & Mullet, 2013; Páez, Martin Beristain, Gonzalez, Basabe, De Rivera, 2011). Although the rebuilding of trust and cooperation may seem to be an unattainable objective, it is nevertheless an inescapable one (Kaufman, 2006). Enduring resentment and feelings of revenge can only generate lasting unhappiness among affected parties. Promoting reconciliation at the national level (e.g., through Truth Commissions) can create the necessary conditions for a change in people's attitudes, but these collective efforts may have drawbacks (Cilliers, Dube, & Siddiqi, 2016; Twose, & Mahoney, 2015; Kpanake & Mullet, 2011). As a result, examining how former opponents in countries which have experienced bloody internal conflicts are able to personally forgive and to feel forgiven is not a futile exercise (Cárdenas, Páez, Arnoso, & Rimé, 2017; Gibson, 2007, López López, León Rincón, Pineda Marín, & Mullet, 2018).
Many studies assessing the willingness of citizens to forgive have been conducted in post-conflict settings (e.g., Azar, Mullet, & Vinsonneau 1999, Bayer, Klasen, & Adam, 2007, Cárdenas, Páez, Rimé, & Arnoso, 2015; Moeschberger, Dixon, Niens, & Cairns, 2005, Pham, Weinstein, & Longman, 2004). Overall, most of their findings were consistent with Fehr, Gelfand, & Nag's (2010) suggestion that circumstantial factors such as the offenders' apologies for the harm they caused accounted for most of the explained variance in willingness to forgive.
These studies have also shown that the relationship between forgiveness and mental health is not as straightforward as one would assume. In a study conducted in Rwanda, for example, Mukashema and Mullet (2013) have shown that if a positive relationship between mental health and forgiveness exists, this relationship is mediated by the feeling of being personally reconciled with the harm-doer(s). In addition, (a) only one type of reconciliation sentiment, the one corresponding to a renewed capacity to listen to each other, to forge compromises, to work together, and to live together is positively associated with mental health, and (b) this type of reconciliation sentiment seems to be present, above all, among individuals who feel disposed to unconditionally forgive (see also, Mukashema, Bugay, & Mullet, 2017 and Lopes Cardozo, Kaiser, Gotway & Agani, 2003).
Colombian People's Attitudes towards Forgiving Former Actors in the Armed Conflict
As already stated, two studies on willingness to forgive perpetrators of violence have recently been conducted in Colombia. In the first of these studies (López López et al., 2013), 400 adults living in Bogotá were presented with a set of realistic vignettes in which a former perpetrator of violence (e.g., a member of a guerrilla group) asked for forgiveness from a victim's family. Three factors were manipulated: the degree of the perpetrators' responsibility (e.g., mere agent executing orders), the type of crime committed (e.g., destruction of property), and whether apologies were offered or not. No information regarding whether or not the perpetrator had previously been sentenced or amnestied was given. Through cluster analysis, four personal positions regarding forgiveness were found. For 38%, mostly from the wealthier segments of society, forgiveness should never be granted. In addition, 39% were hesitant: Their position was similar to the previous one, but they were less blatantly hostile to forgiveness. In contrast, 15%, mostly from the poorest segment of society, considered that forgiveness should always be granted, and 18%, mostly from the poor segments of society, considered that, each time the former perpetrators have expressed true repentance, forgiveness could be granted.
In the second study (López López et al., 2018), 550 adults, also living in Bogotá, were presented with a largely similar set of realistic vignettes in which information regarding whether or not the perpetrator had been sentenced or amnestied was, this time, indicated. The same positions were found: Never forgive (29%), Hesitant (23%), Always forgive (18%), and Depends on circumstances (13%). In addition, 17% of the sample was totally undetermined, which López López (2018) related to recent statements issued in the national press showing that as the peace process developed, some people may have become indecisive (e.g., El Espectador, 2013). In both studies, a striking finding was that very few participants adopted positions showing that at least minimal deliberation (weighing the pros and the cons) had taken place before responding.
The Present Study
The present study complemented the studies reported above in two ways. Firstly, it considered a range of offenses that were more varied than the ones considered in previous studies. In López López et al.'s (2013, 2018) studies, offenses ranged from theft and destruction of property to kidnapping and murder. In the present study two a priori less severe offenses were introduced: verbal assault (without any physical brutality) and manhandling. Also, rape was introduced, in addition to murder, because we wanted to know the extent to which rape might be considered more forgivable than murder. Given that rape was introduced as a type of offense, in all scenarios, the victim was always a woman. We expected that, as a result of the introduction of this larger range of crimes, the effect of the type of crime factor on willingness to forgive would be stronger than in both previous studies.
Secondly, the present study aimed at directly relating positions regarding forgiveness and the political context in the country. Several socio-political issues were debated in the press at the time of data gathering: Are the current peace dialogs that are taking place in La Habana really necessary for stopping the violence in the country? Does citizens' security depend on the peace agreements that are to be signed? Would it be acceptable to allow members from former illegal armed groups to participate in political life? (BBC News, 2016). As stated above, unforgiveness and concomitant feelings of revenge among most citizens; that is, damaged psychological infrastructure, can be a serious impediment to the rebuilding of other infrastructures, including political life. In an atmosphere of suspicion and resentment, the civil society cannot fully develop or redevelop, and the return to political normality may, as a result, be delayed.
Hypotheses
Our first hypothesis was that the four positions reported in both previous studies would be found in the current study: Never forgive (irrespective of circumstances), Hesitant, Depends on circumstances, and Always forgive. We also expected that the Depends on circumstances position would include severity of offense as an important circumstance, in addition to the presence or absence of apologies.
Our second hypothesis, which was closely related to the first one, was that, among participants sharing the Depends on circumstances position, rape would not be considered as more forgivable than murder. This hypothesis was based on previous work showing that, in countries where the death penalty exists, most people consider that rape deserved the death penalty to the same extent as murder (see, Kamble & Mullet, 2016).
Our third hypothesis was that positions regarding forgiveness would be associated with attitudes regarding socio-political issues currently debated in Colombia (Correa, García-Chitiva & García-Vargas, 2018) In particular we expected that participants asserting the always forgive position would be more favorable to the peace dialogs in La Habana, and would be more convinced that Colombian's security strongly depends on the signature of a peace accord than participants asserting the never forgive or hesitant positions.
Method
Participants
Participants were 230 adults (43% males) aged 18-76 years old (M = 30.72, SD = 11.93) who lived in Bogotá. Forty-two per cent of them were from a low economic background, 36% from an intermediate economic background and 22% from a high economic background. Eight per cent had primary school education, 46% had secondary education, and 46% had a university degree. The participation rate was 58%. The main motive expressed for not participating was a lack of time. The study conformed to the ethical recommendations of the Colombian Society of Psychology; that is, full anonymity was respected and informed consent was obtained from all participants.
Material
The material consisted of 72 cards describing situations in which a male offender (e.g., a member of a guerrilla group) committed a violent act (e.g., rape) against a non-combatant woman. Each scenario contained four items of information, in the following order: (a) the identity of the perpetrator (a member of a paramilitary group, a guerrilla group, or the military), (b) the severity of the violent act (killing, rape, manhandling, or verbal assault), (c) the degree of responsibility for the act (organizer of the violence or mere executor of orders), and (d) the offender's subsequent behaviour (neither repented nor begged for forgiveness, met the family and begged for forgiveness, or begged for forgiveness and offered compensation for the harm done).The scenarios were obtained by orthogonally crossing these four factors. The design was Identity x Severity x Responsibility x Subsequent Behaviour, 3 x 4 x 2 x 3.
An example of a scenario is the following: "Enrique Calderón is a former member of the guerrilla. He has committed acts of violence during the civil conflict. Among these acts was the killing of Belén González, a non-combatant woman who lived in a village. These violent acts were planned by his superiors. Enrique just followed orders. Currently, Enrique wishes to reintegrate into civil society. He spoke in person to the González family. He explained what happened to Belén. He begged for forgiveness. He was apparently very emotional. If you were a member of the González family, to what extent would you be willing to forgive Enrique Calderón for what he did to Belen?" Responses were given on 11-point scales raging from Not at all (0) to Fully agree to forgive (10).
The material also comprised one question about the participant's level of religiosity and five additional questions related to the current political context in Colombia: To what extent do you think that (a) the current peace dialogs in La Habana are necessary for stopping the violence in the country, (b) the citizens' security depends on the agreements that are to be signed in La Habana, (c) the political participation of members from former illegal armed groups would be acceptable, (d) every former member of illegal armed groups who wish to be reintegrated into society should, first of all, be judged and incarcerated, and (e) people who subject themselves to the reintegration process are not fully trustworthy. Responses were given on 5-point scales raging from Not at all (0) to Fully agree (5).
Procedure
Data collection took place in 2015 and 2016, before the referendum; that is, at a time where the peace process in Colombia was still running and the cease fire was about to be signed. The procedure followed Anderson's (2008, 2018) recommendations for this kind of study (see also Mukashema & Mullet, 2015). Each person was tested individually, and data collection was completed in a quiet room, usually in the participant's home. Participants took 35 to 40 minutes to complete the ratings. No participant voiced complaints about the number of vignettes or about the credibility of the proposed situations. The additional items were presented at the end of the sessions.
Results
As very diverse positions were expected, a cluster analysis, using the K-means procedure (Hofmans & Mullet, 2013), was first applied in order to detect qualitatively different rating patterns. As each participant had given three series of 24 ratings on scenarios containing the same information (type of crime, level of responsibility, and presence/ absence of apologies) but differing regarding the perpetrator's identity (paramilitary, guerrilla, and military), the total number of profiles of ratings analyzed was 3 x 230 = 690. As four positions were expected a four-cluster solution was first applied, and then three-, five-, six-and seven-cluster solutions were subsequently examined. The six-cluster solution was retained because it provided the most interpretable patterns.
An overall ANOVA was conducted with a design of Cluster x Severity x Responsibility x Subsequent behaviour, 6 x 4 x 3 x 3. Owing to the great number of comparisons, the significance threshold was set at .001. Given that the cluster effect and all the two-way interactions involving cluster were significant, separate ANOVAs were conducted at the cluster level. Results are shown in Table 1. Five of the six clusters are shown in Figure 1.
Figure 1 Five of the six observed clusters. In each panel, the mean willingness to forgive judgments is on the y-axis, the levels of the apologies factor are on the x-axis, and the type of crime corresponds to the three curves.
The first cluster of profiles (N = 187, 27% of the profiles of ratings) was the expected Never forgive cluster. Most ratings were very low (M = 0.26). They were slightly higher (a) in cases of verbal assault (M = 0.37) than in cases of killing (M = 0.17), (b) when the offender was an executor (M = 0.30) rather than an organizer (M = 0.21), and (c) when the offender directly apologized to the family and offered compensation (M = 0.31) than when no apologies were offered (M = 0.20). As shown in Table 2, profiles of participants who were single were less often found in this cluster (22%) than profiles of participants who were married or who had been married (33%).
Table 2 Composition of the Clusters
Note: N = Total number of profiles. The percentages with the same subscript were significantly different, p < .05.
The second cluster (N = 236, 34% of the profiles) was the expected Hesitant cluster. Ratings were low (M = 2.21), although higher than in the previous cluster. Also, they were higher (a) in cases of verbal assault (M = 2.72) or manhandling (M = 2.40) than in cases of killing (M = 1.82) or rape (M = 1.89), (b) when the offender was an executor (M = 2.52) rather than an organizer (M = 1.90), and (c) when the offender directly apologized to the family (M = 2.33) and offered compensation (M = 2.79) than when no apologies were offered (M = 1.50). As shown in Table 2, profiles of male participants (28%), of 22-78 year olds participants (28%), of participants from either low (27%) or high background (30%), and of not-single participants (24%) were less often found in this cluster than profiles of females (39%), of very young participants (47%), of participants from intermediate economic background (45%), and of single participants (42%).
The third cluster (N = 127, 18%) was called Depends on crime because this factor was, by far, the one with the strongest impact. Ratings were clearly higher in cases of verbal assault (M = 5.67) or manhandling (M = 4.50) than in cases of killing (M = 1.82) or rape (M = 1.89). Also, ratings were higher when the offender was an executor (M = 3.65) rather than an organizer (M = 3.06), and when the offender directly apologized to the family (M = 3.70) and offered compensation (M = 9.98) than when no apologies were offered (M = 2.39). In addition the effect of the subsequent apologetic behaviour varied as a function of type of crime. In cases of verbal assault, the effect of the apologetic behaviour was stronger (6.52 - 4.25 = 2.27) than in cases of killing (1.91 - 1.04 = 0.87). In all four cases involving verbal assault and offer of compensation, ratings were clearly on the positive side of the response scale; that is higher than 6. As shown in Table 2, profiles of older participants (8%), and of participants with primary education (2%) were less often found in this cluster than profiles of participants aged 18-39 (22%), and participants with secondary or tertiary education (20%). Profiles corresponding to situations involving military personnel were less often found in this cluster (13%) than profiles corresponding to situations involving a paramilitary (24%).
The fourth cluster (N = 54, 8%) was called Depends on apologetic behaviour because this factor was, by far, the one with the strongest impact. Ratings were clearly higher when the offender directly apologized to the family (M = 7.21) and offered compensation (M = 7.45) than when no apologies were offered (M = 0.93). Also, ratings were higher in cases of verbal assault (M = 5.59) or manhandling (M = 5.29) than in cases of killing (M = 4.95) or rape (M = 4.97). In all 12 cases involving the seeking of forgiveness, ratings were clearly on the positive side of the response scale. As shown in Table 2, profiles of female participants (5%), of very young participants (2%), of participants from either low or intermediate background (7%), and of participants with secondary and tertiary education (6%) were less often found in this cluster than profiles of males (11%), of older participants (11%), of participants from high economic background (15%), and of participants with primary education (18%).
The fifth cluster (N = 52, 8%) was the expected Always forgive cluster. Most ratings were high (M = 8.33). As shown in Table 2, profiles of participants aged 18-29 (1%), of participants from either high or intermediate background (5%), of single participants (4%), of participants with tertiary education (5%) were less often found in this cluster than profiles of older participants (15%), of participants from low economic background (12%), and of participants with primary education (17%). Finally, the sixth cluster (N = 34, 5%, not shown in Figure 1) was called Undetermined. Ratings were always in the middle range of the scale (M = 5.53), and no significant effect was detected.
An overall ANOVA was also conducted on the raw data with a design of Identity x Severity x Responsibility x Subsequent Behaviour, 3 x 4 x 2 x 3. As shown in Table 1, all four main effects were significant. Table 3 compares participant's positions as a function of the perpetrators' identity. Among 71% of participants, positions did not vary as a function of identity. Among the remaining 29%, change of position mostly reflects a more unforgiving attitude to the military than to other groups. For example, for ten participants, forgiveness was possible and mostly depended on the severity of the crime when perpetrators were members of a guerrilla group or a paramilitary group, but not when they were members of the military.
Figure 2 shows the relationship between the participant's positions when these positions were homogeneous (N = 169) and ratings given to the additional items. Overall, the association with religiosity was not significant, but post-hoc analyses using Fisher's LSD test showed that religiosity ratings were significantly higher among participants in the Always forgive cluster (M = 3.88) than among participants in the Never forgive cluster (M = 3.19), p < .05. The association with attitude to the peace dialogs in La Habana was significant, F(4, 158) = 4.64, p < .005, η2 p = .11. Post-hoc analyses showed that attitude to the peace dialogs was significantly more negative in the Never forgive cluster (M = 2.54) than in all the other clusters (M = 3.40), p < .02. Overall, the association with the belief that Colombian's security depends on the peace agreements was not significant, but post-hoc analyses showed that this belief regarding security was significantly higher among participants in the Always forgive cluster (M = 2.88) than among participants in the Never forgive and Depends on crime clusters (M = 2.06), p < .02. The association with attitude to participation in politics of members from former illegal armed groups was significant, F(4, 158) = 3.76, p < .01, η2 p = .09. Post-hoc analyses showed that attitude to participation in politics was significantly more negative in the Never forgive cluster (M = 1.96) than in the Hesitant, Depends on Behaviour and Always Forgive clusters (M = 2.81), p < .02.
Discussion
As hypothesized, several qualitatively different positions were found regarding the participants' willingness to forgive perpetrators of offenses against women during the armed conflict in Colombia. As in López López et al.'s (2013, 2018) studies, the majority of participants in the present study (27% + 34% = 61%) were quite unwilling to forgive. This percentage was higher than the one reported in the 2018 study (52%). This may be due to the fact that the situation in the vignettes specifically depicted violence against women: In the present study, female participants were more frequently hesitant or unwilling than male participants, which was not observed in previous studies.
Among participants who were not completely hostile to the idea of forgiveness, three positions were found. The Depends on circumstances position that had been found in previous studies and split up into two contrasted positions: one that stressed type of crime (18%) and one that stressed apologetic behaviour (8%). This finding was, therefore, consistent with our hypothesis that the type of crime factor would have a stronger effect on the current study than on the previous studies, but the way this strengthened effect manifested itself was not anticipated. Participants who focused on type of crime were essentially younger (less than 40 years old) and more educated. They also expressed a more unforgiving attitude to the military. Participants who focused on apologetic behaviour were essentially older (more than 22 years old) and less educated. In addition, 8% of participants expressed unconditional willingness to forgive - the expected Always forgive position. Overall, 34% of the sample was of the opinion that if the perpetrator has apologized and if the crime was not too severe, forgiveness was possible. A small group of participants (5%) who did not take any position was, as in the 2017 study, also found.
As also hypothesized, willingness to forgive ratings attributed to cases of rape were always close to ratings associated to murder. Overall, the difference was minimal (less than 0.10 point on a scale ranging from 0 to 10).
Finally, the association between participants' position regarding forgiveness and their attitudes to currently debated political issues was in the hypothesized direction. Participants who did not reject the possibility of forgiveness expressed more positive views regarding the current peace process than participants who expressed rejection.
Overall, the present findings were consistent with the previous ones (López López et al., 2013, 2018) and similar to those of Nadler and Liviatan (2006) among Jewish students living in Israel and of Staub, Pearlman, Gubin, & Hagengimana (2005), Mukashema and Mullet (2010), Heim and Schaal (2014) in Rwanda, and Arnoso Martinez, Cardenas Castro, and Páez Rovira, (2015) in Chile. In all cases, the damage caused was severe, but the political situation had objectively improved as a result of the demobilization and of the peace process). The level of willingness to forgive that was associated with these situations was usually intermediate.
To be successful, any attempt at reintegrating former perpetrators into society must be done with care; that is, on a voluntary basis. The present study, as well as similar previous studies (Pineda-Marín, Muñoz-Sastre & Mullet, 2018), has shown that a plurality of Colombian citizens is blatantly hostile to the idea of forgiving or reconciling with former perpetrators. This is perfectly understandable owing to the fact that being in the presence of former perpetrators on a daily basis, or just the fact of having to think about them, would lead most people -- particularly direct victims -- to vividly recall painful events. As shown by Bogliacino, Grimalda, Ortoleva, and Ring (2017) the simple recall of such events has, among victims who were exposed to intense violence, the capacity to severely affect their behaviour, particularly their short-term memory and cognitive control -- a very undesirable consequence.
Forgiving is difficult, particularly when the offense has lasting negative consequences and the harm-doers' apologies, when present, are not perceived as sincere. National policies can create conditions favoring victims' decisions to forgive and to reconcile with perpetrators (Allan, Allan, Kaminer & Stein, 2006). As suggested by Lichtenfeld, Buechner, Maier, & Fernández-Capo (2015), however, it is not enough to simply decide to forgive in order to be able to do it. Most of the time, forgiveness can only occur as the result of a long, and at time instable, emotional process (Worthington, 2006).
Some citizens, however, seem to already be open to the idea of forgiving and reconcile. In other words, a process of reintegration is, therefore, not impossible everywhere. The current findings, as well as the findings from previous studies suggest that such a process should, first, be attempted in areas where a majority of people share positive views regarding the reintegration of former perpetrators. Then, building on the experience gathered in these areas, the process could be progressively extended to areas in which views are less positive, although not uniformly negative.
Finally, the process of reintegration should ideally be a democratic process: Local consultations should always be carried out before any attempt at reintegrating perpetrators of violence into a determined area takes place.
Limitations
An important limitation of this study was that all the participants lived in Bogotá, the largest city in the country. Owing to the regional aspect of the conflict, the percentage of people holding one or another of the six positions presented above is, as suggested by the results of the October referendum (Semana, 2016b), likely to vary considerably from one department to the other. In Bogotá, 56% of voters favored the peace accords, but in Chocó it was 79%, however, in Meta it was only 36% (El País, 2016).
Another limitation of the study was that the participants' status as primary victims, secondary victims or tertiary victims (Govier, 2002) was not assessed. In other words, the personal impact of the conflict on each participant's wellbeing, including their mental health, was not known. Personal experiences (lasting suffering) may have important effects on willingness to reconcile (e.g., Stein et al., 2008). As a result, future studies should include measurements of personal victimization and of the physical and mental consequences of victimization, and relate them to positions regarding forgiveness of former perpetrators of violence.
