Running head: DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 1

Dehumanization Effects on Agency Attributions, Retributive Justice judgments, and

Resocialization Attitudes towards Former Perpetrators in Post-conflict

Cristhian A. Martínez1

Juan D. Leongómez2

1&2 Faculty of Psychology, El Bosque University. Bogotá, Carrera. 9 # 131a- 02,

Colombia.

Corresponding author: Cristhian A. Martínez, Faculty of Psychology, El Bosque

University, Bogotá, Carrera 9 # 131a -02, Colombia. E-mail address: [email protected].

(This paper has not been published yet, please do not reproduce any portion without citation)

Acknowledgments

We would like to thank Isabella Lopez, Karen Casas, and Leandra Cuestas for their comments and assistance in collecting the data. Furthermore, we are grateful to Jan-Willem van

Prooijen for their valuable suggestions and useful comments on previous drafts of this paper and to all our participants.

DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 2

Abstract

Previous research has evidenced the effects of dehumanization on observers’ attributions, social categorizations, and retributive judgments, towards different kinds of perpetrators.

Colombia is currently facing a peace process with the FARC, the largest guerrilla force in the country, and this post-war context becomes ideal for studying these effects. The current study evaluated the effects of a media dehumanizing discourse towards former perpetrators on observers’ agency attributions, retributive judgments, and resocialization attitudes towards them.

Each participant read two different web news articles about the capture of a FARC member which were manipulated to be framed in humanizing and dehumanizing speech respectively.

Then, they answered an agency attribution scale, retributive justice and resocialization attitudes measures, and a social distance scale about the news’ main character. Results show that when

FARC members were framed in a dehumanizing context, observers significantly attributed them less agency, endorsed more severe retributive judgments, showed more negative attitudes towards their resocialization, and showed greater social distance. The implications of these findings for a Disarmament, Demobilization, and Reincorporation (DDR) process in a post- conflict scenario are discussed.

Keywords: framing, dehumanization, agency, retributive justice, resocialization, post-conflict.

DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 3

Introduction

Colombian armed conflict is recognized as one of the most prolonged and severe in Latin

America’s contemporary history. It is calculated that from 1958 to 2012, the internal conflict caused approximately 30.000 kidnappings, 25.000 disappeared individuals, and 220.000 deaths between civilians and combatants (Grupo de Memoria Histórica, 2013). 6.8 million people have been forcibly displaced due to violence, generating the world’s second largest internally displaced population after Syria (“Human Rights Watch, World Report 2017: Colombia,” 2017).

The recent peace agreement with The Revolutionary Armed Forces of Colombia (FARC-EP for its Spanish acronym), the largest guerrilla group in the country, and the ongoing post-conflict situation in which former perpetrators are being reincorporated to civic society, provides a unique opportunity to test the effects of inter-group dehumanization on observers´ perceptions of former violent group members after conflict.

Within inter-group dehumanization literature, criminals, social norm transgressors, and terrorism suspects have been identified as targets of dehumanization. It has also been evidenced that dehumanization increase observer’s attributions of lack of civility, self-control, cognitive abilities, emotionality, culture, interpersonal warmth, depth and agency towards the dehumanized targets (Haslam, 2006).

The present study is intended to make a contribution to the general understanding of observers` perceptions of perpetrators in post-conflict settings by means of integrating and extending interdisciplinary knowledge from communication science and cognitive, political and social psychology. The specific aim of our study is to evaluate the effect of former perpetrators` dehumanization (through media discourse) on observer’s attributions of their agency, retributive justice judgments towards them, and attitudes towards their resocialization. We predict that DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 4 dehumanization will elicit in observers fewer agency attributions, more severe blame attributions and retributive judgments, more negative attitudes towards their resocialization and more social distance. To test our predictions, we use a within-subjects design and unlike previous research we use dehumanization as an independent variable by experimentally manipulating it.

Dehumanization

Dehumanization is defined as “the act of perceiving or treating people as if they are less than fully human” (Haslam & Stratemeyer, 2016, p. 25), and has been widely studied by social psychology (e.g., Haslam, 2006; Haslam & Loughnan, 2014; Haslam & Stratemeyer, 2016).

Two general kinds of dehumanization have been identified in the general model proposed by Haslam (2006): the mechanistic (i.e., object-like) and the animalistic (i.e., animal-like) dehumanization. According to the model, the undermining of human nature (HN) core traits such as emotional responsiveness, interpersonal warmth, cognitive openness, and individuality promotes the mechanistic form of dehumanization. On the other hand, the undermining of unique human (UH) traits such as moral sensibility, rationality or maturity, give rise to the animalistic dehumanization.

Both kinds of dehumanization are mediated by social factors such as political orientation, and there is evidence suggesting that it facilitates support for punitive forms of counter-terrorism, unfair and discriminatory treatment, insensibility to other’s pain and suffering (Haslam &

Stratemeyer, 2016), and moral disengagement in order to legitimize harsh punishments or aggressive actions against perpetrators (Khamitov, Rotman, & Piazza, 2016). The influence of dehumanization on interpersonal perceptions has been previously investigated in warfare contexts (Tajfel, 1981). In a recent study for example, it has been evidenced that dehumanization can validate torture as a way to fight terrorism (Lindén, Björklund, & Bäckström, 2016). DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 5

Dehumanization also increases the elicitation of visceral moral emotions like anger or disgust and influences retributive judgments and perceived harm caused by perpetrators. For example, it has been evidenced that when offenders are divested of their humanity, observers attribute that they are less sensitive to pain, more dangerous, uncontrollable, feel less guilty, less willing to restore the harm they caused, and that they are more easily morally disengaged

(Bastian, Denson, & Haslam, 2013).

Human agency refers to the self-reflective or metacognitive capacity that individuals have for recognizing themselves as causal beings, with an identity, and with autonomy over their own desires, intentions, beliefs, emotions, responsibilities, and moral sense (Bandura, 1999; Narvaez

& Lapsley, 2009). Recent research has evidenced that dehumanized offenders are granted with a decreased sense of agency (Bastian, Laham, Wilson, Haslam, & Koval, 2011; Khamitov et al.,

2016).

It has been also evidenced that dehumanization has negative effects on social connectedness with the dehumanized out-groups by means of increasing social and psychological distance (Haslam, 2006). Social distance is the reduction of the degree of understanding and intimacy characteristic of social relations between individuals and social groups (Wark & Galliher, 2007). When people are regarded as socially distant, they are perceived in a simple, abstract, and impoverished way by ignoring specific traits like motivations, beliefs, and intentions, and this facilitates observers’ cold cognition-based judgments (Haslam, 2006).

Framing

It is relevant to ask about which factors could influence observers’ (i.e., common citizens) attitudes, perceptions, and attributions towards former perpetrators of violence. Taking into DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 6 account that armed conflicts are often experienced directly by certain portion of the population but also indirectly through mass media by most of the people, it is pertinent to address how media depictions of the actors could influence observers´ perceptions. According to Sánchez,

López, & Barreto (2013), media directly influence the process of construction and diffusion of reality. By means of agenda settings, and cognitive framing of messages, media could shape the ideological consume through which individuals acquire representations, beliefs, and meanings around the social order and their participation in it (Sánchez et al., 2013). Regarding the specific media framing of peace processes, Sheafer & Dvir-Gvirsman (2010) suggests that news reports provide citizens with important clues to form a public opinion around a peace process, but also, that media coverage is often destructive due to the contradiction between a peace process requirements and the drama/sensation journalistic norm. They found for instance that a negative bias in the news coverage about the Israeli-Palestinian Oslo peace process facilitated a stronger public negative response and had a negative effect on government efforts to publicly promote the peace agreement.

Framing theory predicts that the form in which an issue is characterized or presented has an influence in the way it is understood by an audience through their cognitive schemas for information processing (Scheufele & Tewksbury, 2007). Framing has two levels, a sender-level focused on the construction and presentation of messages, and a receiver-level focused on impression formation in the receptor. Regarding the former, some research has explored how specific forces or groups within society try to shape public discourse by establishing predominant labels to the issues being communicated (Gamson & Modigliani, 1989). Regarding the latter, it has been found that repetition of frames has an impact in less informed individuals, DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 7 whom at the same time pay more attention to peripheral clues in the messages to form their opinions (Scheufele & Tewksbury, 2007).

Some studies within linguistics and discourse analysis approaches have shown biases in the framing of Colombian armed conflict actors in mass media. For example, Gutierrez (2007) found that when demobilized members (both from FARC and paramilitary groups) starred news, they were mainly associated with crime and judicial penalties, and 47.4% of the journalistic notes in the sample, emphasized demobilized members as guilty perpetrators of crimes, excluding other individual characteristics. Relatedly, Castellanos (2014) evidenced how ex- president Álvaro Uribe, a major opinion leader within the country, build a warlike rhetoric towards guerrilla which markedly differentiate between “them” (i.e., FARC) and “us” (i.e., him, his team, and civilian society).This differentiation has been enforced by means of using specific linguistic formulas such as nominating FARC members as “terrorists” or “murderers”, comparing them to the Nazis and Al Qaeda, and associating his politic agenda with values such as security, trust and progress. According to the in-group love/out-group hate hypothesis, this clear-cut way of differentiating between groups is one of the main motivations to engage in inter- group conflict (Weisel & Böhm, 2015).

In the same line, after having analysed more than 500 news articles from the principal

Colombian journals about violent events perpetrated by the FARC and the paramilitaries between 1998 and 2006, García-Marrugo (2013) found that press uses differential linguistic strategies such as the passive voice, or euphemisms (e.g., dead in combat vs. beheaded) to lessen the perception of responsibility of paramilitary groups, and highlight FARC responsibilities via dehumanization and demonization. Within the 200 adjectives used in the news to denominate

FARC members, are for instance: “savage”, “satanic”, “sanguinary”, or “damned” among others. DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 8

Despite the doubtless responsibility of former perpetrators in past violent events, these forms of dehumanizing discourse have serious effects on observers’ attitudes and perceptions towards their agency, blame, and acceptance as future members of the in-group, which are crucial issues for a successful DDR process.

Disarmament, demobilization, and reintegration

A disarmament, demobilization, and reintegration (DDR) process is a great responsibility since it directly involves the security guarantees for victims and civilians (e.g., no repetition, no recidivism in crime), and the survival of ex-combatants and their relatives, considering their reception by society. Protocols for peace negotiations define reincorporation as an economic, social, and political issue (Nussio, 2013; Theidon, 2009). One of the primary factors for a successful DDR process is citizens’ attitudes and perceptions towards the transgressors who are about to enter civilian life. Most of the times, communities receive ex-perpetrators with resentment as they have suffered violence without any governmental support (Özerdem, Podder,

O’Callaghan, & Pantuliano, 2008); For instance in the Colombian specific case, some people have criticized the unequal benefits for ex-combatants and victims, the limitations of transitional justice, and the new outbreaks of violence following the demobilization of paramilitary groups in

2005 (Nussio, 2013). Therefore, any DDR process has the difficult task of balancing the social inclusion of ex-combatants with the demands and acceptance of the community (Annan &

Cutter, 2009). If not carried out properly, DDR processes could intensify prejudice, facilitating the segregation between civilians and demobilized members (Bøås & Hatløy, 2008; Jennings,

2007), and hinder their psychological reintegration (Hangman & Nielsen, 2002).

DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 9

Colombian post-conflict context

FARC guerrilla is one of the oldest existing guerrilla groups in the world, the largest in the country, and are one of the main actors in the conflict. Established in 1964 as a leftist guerrilla, the FARC started a direct confrontation with the conservative government of the time and, later, with right-wing paramilitary groups. In 1966 they stipulated the expansion of their offensive operation to the entire country and were determined to bring the conflict to major urban locations (Sanchez, Diaz, & Fornisano, 2003). The FARC are responsible for having killed, abducted, kidnapped, forcibly recruited children, and displaced civilians, amongst other transgressions.

Several attempts to negotiate peace between FARC and government have failed.

Attempts between 1982-1987, 1991-1992, 1998-2002 did not work mainly due to procedural errors, a lack of a multidisciplinary approach, a lack of institutionalization, and a weak support from the civilian society (García-Duran, 2004). On August 2016, FARC finally reached a peace agreement with the Colombian government and announced a definitive ceasefire, ending a more than 50 years of war. However, a popular voting to validate the agreement was held in October

2016, and 50.21% of the voters did not agree with endorsing the peace treaty. In the midst of huge controversies with well-established right-wing sectors who opposed the peace agreement, and with a completely polarized society, additional negotiations took place, and a final corrected version of the document was signed by the parts in November 2016. Finally, on July 2017 FARC handed over their weapons to a UN verifying commission. The government estimates that approximately 17.500 demobilized members of FARC will be incorporated into civilian life

(“Gobierno calcula que 17.500 miembros de las Farc se desmovilizarán con la paz,” 2016). DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 10

Despite the historic achievement, approximately half of the population remains skeptic about transitional justice and resocialization.

Methods

Participants

A total of 96 subjects (47 men, 49 women) took part in our within-subjects design study.

They were recruited in Bogotá, Colombia, using a non-probabilistic convenience sampling. Of these, 32 were students from 11th grade from a private high school (age: M ± SD = 17.5 ± 1.4 years old), 32 college students from different careers from a private university (M ± SD = 21.1 ±

1.5), and 32 adults with different occupations (M ± SD = 36.5 ± 6.0). All of them participated voluntarily and were Colombian citizens. No participant was a direct victim of the armed conflict, nor reported any physical or psychological dysfunction that impedes to answer the instruments.

Instruments

Two fictional web news articles about the capture of a FARC member were designed based on several real news about this issue previously compiled from the web. The news articles were piloted and then manipulated to test the effects of dehumanization. In a dehumanizing condition (D), the article was framed using dehumanizing speech towards the ex-perpetrator.

First, he was nominated by his alias, and with qualifiers such as ‘terrorist’, ‘soulless bandit’,

‘misfit’ and ‘criminal’. Secondly, character was divested of specific human traits such as emotional and mental states by omitting any reference or information about it. On the other hand, in a humanizing condition (H) the article was framed in a way that preserved perpetrator humanness. First, he was nominated with qualifiers such as ‘person’, ‘man’, ‘individual’ or DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 11

‘citizen´, and personal information (e.g. name, age, and origin) was included. Additionally, references to his emotional and mental states (i.e. describing his current feelings and giving him voice) were also included (Appendix S1, news articles texts).

Both news articles featured realistic references to geographical settings, army and police units, but fictional character names were used. Both had the layout and organization to make them look like authentic articles from a news website. Apart from the manipulations, news articles had approximately the same content and extension. In both scenarios, characters were apprehended due to the same misdeeds (i.e. criminal conspiracy, rebellion, and illegal possession of weapons). In order to keep a within-subjects design in which each participant read the two versions of the news, minor variations in paragraph structure and story details such as location, characters or contextual information were made to avoid reactivity and testing effects

(Christensen, Johnson, & Turner, 2015).

For measuring agency attributions, a cognitive trait scale was adapted from previous studies (Gray, Gray, & Wegner, 2007; Khamitov et al., 2016). Two additional items about self- recognition (i.e. ‘this person appears to be capable of recognizing himself as a subject’) and responsibility (‘this person appears to be capable of taking responsibility for their actions’) were included for a total of nine items (α = .81); After reading both news articles, participants were asked to judge the main character in each story by rating their agreement with the nine items on a seven-point scale (1= totally disagree, 7= totally agree).. (Table S2, agency and social distance scales).

To evaluate retributive justice judgments, blame attributions and punishment severity measures were adapted from Bastian et al., (2013). Attributed blame, was measured by asking participants “to what extent do you think the main character of the new (i.e., FARC member) DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 12 should be blamed for their actions?” Then, they rated their answer on a seven-point scale (0= not at all, 6= totally). Punishment severity was measured by asking participants: “if the penalty were imprisonment, how many years in prison should receive the main character of the new?” and then they were asked to enter a value from 0 to 60 which is the maximum penalty in Colombia.

To evaluate attitudes towards resocialization participants rated how suitable they considered the FARC member for a resocialization program on a seven-point scale (0 = not suitable at all, 6= very suitable). Additionally, a social distance scale (Wark & Galliher, 2007) was designed and adapted for the context of this study. Participants rated on a five-point scale (0

= nothing at all, 4 = very much), their agreement with an ascendant list of seven items which increase in the level of social closeness with FARC members. The scale showed high reliability levels (α = .95) (Table S2, agency and social distance scales). A single questionnaire including all the measures described above was designed.

Procedure

A within-subjects design was employed. Each participant read both manipulated news articles and answered the questionnaire at the end of each one. Order of presentation of news articles was counterbalanced so that half of the participants read the dehumanizing (D) version first, and then the humanizing one (H), and the remaining participants read the articles in the opposite order. Stimuli and questionnaires were administered individually, in quiet settings away from distractions, and each application lasted 25 minutes on average.

The procedure was approved by the Bosque University psychology faculty ethics committee. Written informed consent was obtained from each participant (or their parents in case of minors) before application and every participant (including minors) verbally agreed to participate before the application began. Then, manipulated news articles were presented one by DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 13 one on a laptop, and participants were asked to read both of them and at the end of each one to answer the questionnaire in paper and pencil format, specifying that questions were about the

FARC member who starred each news piece. It was highlighted that there were no wrong answers and that it was important to give an honest and free personal opinion, remembering to participants the anonymity and confidentiality conditions. It was also indicated that there was no time limit for reading the news articles and answering the questionnaire, that browsing within each news article was permitted if needed, as well as correcting answers if considered. Finally, when both questionnaires were completed, participants were debriefed and thanked for their participation.

Data analysis

To measure the effect of dehumanizing speech on agency attributions, retributive justice judgments (attributed blame, allocated years in prison), and resocialization attitudes (suitableness for resocialization, social distance), five separate linear mixed-effect models were conducted, one for each dependent variable. Each model was fitted using the lmer function from the package lmerTest (Kuznetsova, Brockhoff, & Christensen, 2017: https://www.rdocumentation.org/packages/lmerTest) in R version 3.5.1 (R Core Team, 2018).

Each model included the humanization/dehumanization condition (H/D) as a within- subject factor, and sex (x2) and group (x3) (high school, college, and adults) as between-subject fixed factors. Subject was included as a random factor, with random intercepts. To measure interactions between Condition, Gender and Group, each formula was entered in lmerTest as:

y = Group * Gender * Condition * (1 | Subject)

Although allowing slopes to vary randomly is recommended (Barr, Levy, Scheepers, &

Tily, 2013), we only included random intercepts in the models because there is only one data- DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 14 point per subject, on each humanization/dehumanization condition (H/D). To visually inspect the residual distribution of each model, both quantile-quantile (QQ) plots and histograms were produced (Appendix S2).

In addition, to determine the best predictors for each model and confirm whether the experimental manipulation (humanization/dehumanization condition) had an effect on Agency

Attributions, Retributive Justice judgments, and Resocialization Attitudes, we used the step

(https://www.rdocumentation.org/packages/lmerTest/versions/3.01/topics/step.lmerModLmerTes t) function to choose a parsimonious model in a stepwise (backward) elimination of effects algorithm. In all cases, the random factor (subject) was maintained in the models, but in most cases most fixed effects and interactions were eliminated. To report this, type III ANOVA tables were created, showing each fixed main or interaction effect. In lmerTest, ANOVA p-values are calculated using Satterthwaite’s method for the degrees of freedom (for detailed information, see the package description; https://cran.r-project.org/web/packages/lmerTest/lmerTest.pdf), as recommended (Luke, 2017).

To ensure openness and reproducibility, an R markdown (*Rmd file) containing the R code for all LME models and figures, as well as the output (in HTML format) that contains all figures and tables, are provided in the supplementary materials, available online

(https://osf.io/jc6te/).

Results

Correlational Analysis

In the first place, we conducted repeated measures GLM with the five dependent variables to test between-subjects effects by order of presentation and no significant effects were found by this factor for any variable. Then, we conducted a correlational analysis with all the DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 15

dependent variables in both conditions to corroborate expected associations between them (e.g.

positive associations between social distance and harsher retributive judgments or negative

associations between social distance and agency attributions). As expected for both conditions,

results indicate significant positive correlations between social distance, punishment allocation

(attributed years of imprisonment) and attributed blame; and between agency attribution and

suitableness for resocialization; on the other hand, significant negative correlations between

suitableness for resocialization and punishment allocation, social distance and agency attribution,

and social distance and suitableness for resocialization were found (Table 1).

Table 1. Correlations between agency, retributive justice, and resocialization measures by condition.

1 (H). 2 (H). 3 (H). 4 (H). 5 (H). 6 (D). 7 (D). 8 (D). 9 (D).

1. Agency (H) - * 2. Blame (H) ,240 - *** 3. Prison (H) -,154 ,392 - 4. Resocialization (H) ,214* -,153 -,307** - 5. Social Distance (H) *** * ** * -,358 ,216 ,343 -,222 - *** * 6. Agency (D) ,500 ,085 -,039 ,022 -,252 - *** ** *** 7. Blame (D) -,001 ,446 ,345 -,068 ,387 -,034 - 8. Prison (D) -,158 ,308** ,835*** -,287** ,353*** -,110 ,452*** - 9. Resocialization (D) ,290** -,128 -,343** ,518*** -,405*** ,438*** -,156 -,395*** - 10. Social Distance (D) -,390*** ,116 ,308** -,136 ,873*** -,402*** ,380*** ,383*** -,499*** Note. *p < .05, **p < .01, ***p < .001 (2-tailed). Significant correlations are in bolds, variables in humanization condition are indicated with (H) and in dehumanization condition with (D).

Agency attributions

Consistent with predictions, results revealed a main effect for agency attributions by

condition (H/D) (Table 2, Fig. 1A), which indicate that participants attributed less agency to

FARC member in the dehumanization condition (M ± SD = 4.42 ± 1.28), as compared with the

humanization condition (5.19 ± 0.94). Also, a main effect by group was found, in which 11th DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 16 grade students attributed less agency (4.38 ± 0.17) to FARC members than college students (5.0

± 0.16) and adults (5.0 ± 0.16). These two main fixed effects, but no interactions, were maintained in the most parsimonious mixed-effect model.

Participants’ scores for each variable were positively associated across conditions (i.e. between their corresponding pairs in H and D), and some of the previous positive correlations such as between imprisonment, social distance and blame; and negative correlations such as between social distance and agency, and imprisonment and resocialization attributions, were also evidenced across conditions showing consistency in participants` evaluations.

Table 2. Full mixed-effect model for agency attributions. Agency Fixed effects df F value p Group 2, 86 5.42 0.006 Gender 1, 87 0.118 0.732 Condition 1, 87 38.884 <0.001 Group*Gender 2, 87 0.733 0.483 Group*Condition 2, 87 2.187 0.118 Gender*Condition 1, 87 1.938 0.167 Group*Gender*Condition 2, 87 0.496 0.61 Random effect (subject) σ2 0.60

τ00 0.61 ICC 0.51 Marginal R2 / Conditional R2 0.182 / 0.595 2 2 Note: σ = variance; τ00 = random intercept variance; ICC = intraclass correlation coefficient; Marginal R = variance explained by fixed factors; Conditional R2 = variance explained by both fixed and random factors. These values were obtained using the sjt.lmer function from the package sjPlot (Lüdecke, 2018). Retributive justice judgments

For blame attributions, we did not find a main effect of condition (H/D), nor group or gender. However, there was a significant three-way interaction between group, gender and condition (Table 3, Fig. 1B), in which all participant groups and genders tended to attribute more blame in the dehumanizing compared to the humanizing condition, except male school-age DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 17 participants, who surprisingly attributed significantly more blame in the humanizing than in the dehumanizing condition (Fig. 1B). Fixed effects or interactions were excluded during the backward elimination of effects.

Table 3. Full mixed-effect model for retributive justice judgments. Retributive justice judgments Blame Prison Fixed effects df F value p df F value p Group 2, 87 2.751 0.069 2, 86 1.664 0.195 Gender 1, 87 1.851 0.177 1, 86 0.207 0.650 Condition 1, 86 0.463 0.498 1, 87 39.548 <0.001 Group*Gender 2, 87 0.279 0.757 2, 86 1.423 0.247 Group*Condition 2, 86 1.695 0.19 2, 87 0.757 0.472 Gender*Condition 1, 86 3.752 0.056 1, 87 0.992 0.322 Group*Gender*Condition 2, 86 5.325 0.007 2, 87 0.386 0.681 Random effect (subject) σ2 0.70 61.02

τ00 0.48 291.33 ICC 0.41 0.83 Marginal R2 / Conditional R2 0.100 / 0.468 0.097 / 0.844 2 2 Note: σ = variance; τ00 = random intercept variance; ICC = intraclass correlation coefficient; Marginal R = variance explained by fixed factors; Conditional R2 = variance explained by both fixed and random factors. These values were obtained using the sjt.lmer function from the package sjPlot (Lüdecke, 2018).

On the other hand, the punishment judgment was measured by the allocated years in prison. We found a main effect for punishment allocation by condition (H/D) (Table 3, Fig. 1C).

Results indicate that participants considered that FARC member in the dehumanized condition should receive a longer sentence in prison (37.2 ± 18.9 years) as compared with the member in the humanized condition (30.0 ± 18.9 years). All participants from all groups and gender combinations allocated longer prison sentences in the dehumanizing than in the humanizing condition, and this difference reached significance for all participants, except adult males. No further effects or interactions by sex or group were found, so that the most parsimonious mixed- effect model included only condition as a fixed predictor. DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 18

Attitudes towards resocialization

Consistent with predictions, the analysis yielded a main effect by condition (H/D) (Table

4, Fig. 1D). Results indicate that participants attributed less suitableness for resocialization to

FARC member in the dehumanization condition (3.11 ± 1.9) as compared with the humanization condition (4.31 ± 1.5). No further effects or interactions including sex or group were found, and the backward elimination of fixed effects revealed that the most parsimonious mixed-effect model included only condition as a fixed predictor.

Table 4. Full mixed-effect model for Attitudes towards resocialization. Attitudes Towards Resocializacion Resocialization Expect. Social Distance Fixed effects df F value p df F value p Group 2, 86 0.443 0.644 2, 87 1.064 0.350 Gender 1, 86 1.307 0.256 1, 87 3.642 0.060 Condition 1, 87 42.933 <0.001 1, 86 27.46 <0.001 Group*Gender 2, 86 1.004 0.371 2, 87 2.187 0.118 Group*Condition 2, 87 0.002 0.998 2, 86 0.182 0.834 Gender*Condition 1, 87 0.53 0.468 1, 86 0.587 0.446 Group*Gender*Condition 2, 87 0.898 0.411 2, 86 0.766 0.468 Random effect (subject) σ2 1.67 0.19

τ00 1.61 1.13 ICC 0.49 0.86 Marginal R2 / Conditional R2 0.134 / 0.559 0.114 / 0.875 2 2 Note: σ = variance; τ00 = random intercept variance; ICC = intraclass correlation coefficient; Marginal R = variance explained by fixed factors; Conditional R2 = variance explained by both fixed and random factors. These values were obtained using the sjt.lmer function from the package sjPlot (Lüdecke, 2018).

DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 19

Figure 1. Dehumanization effects on agency attributions, retributive justice judgments, and attitudes towards resocialization. A. Agency; B. Attributed blame; C. Allocated years in prison; D. Positive resocialization expectations; E. Social distance. Results are split by condition (H = Humanizing, D = Dehumanizing) and gender (red = female, blue = male). Data distribution is represented by boxplots, with superimposed data points. The experimental manipulation (humanizing/dehumanizing condition) effect, is represented as a trend line with error, for each group. For pairwise comparisons and estimated marginal means, see the supplementary materials, available online (https://osf.io/jc6te/) DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 20

Discussion

The aim of the present study was to test the effect of former violence perpetrators’ dehumanization on observers’ retributive judgments, attributions of their agency, and attitudes towards their resocialization. The results generally support our main predictions: When FARC member was framed in dehumanizing speech, observers significantly attributed him less agency and more blame, allocated more severe punishment, deemed him as less suitable for a resocialization program, and perceived the FARC member as more socially distant in comparison to the humanization condition.

The ways in which the FARC member was dehumanized in the (D) condition in the present study, comprised features both of the mechanistic and animalistic kinds of dehumanization proposed by Haslam (2006). Omitting references to the character`s name, age, origin, mental states, and emotional states pointed to the denial of human nature traits (HN) which facilitates the mechanistic variant of dehumanization in which targets are perceived as inert, cold and passive beings. On the other hand, referring to the character as a ‘terrorist’,

‘soulless bandit’, ‘criminal’ or by his alias ‘scorpion’ was intended to deny unique human traits

(UH) facilitating the animalistic form in which targets are perceived as instinctual and irrational beings.

Animalistic dehumanization has been employed in contexts such as genocide, racial stereotyping, or immigration. Animal metaphors are often used for referring to out-group members and are intended to associate intergroup differences with interspecific differences

(Haslam, 2006). These animal-like dehumanizing qualifiers are commonly used to publicly nominate FARC members in Colombian mass media (e.g., García-Marrugo, 2013) as well as in the public speech by opinion leaders and lay people. As a straightforward example of animalistic DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 21 dehumanization, a 2008 headline in the main newspaper in Colombia states for instance ‘FARC are animals and barbarians without humanity’ (Gómez, 2008).

Our findings support the negative effects of dehumanization on perpetrator`s reduced agency attributions evidenced in previous research (Khamitov et al., 2016) and extended this effect to specific instances of media dehumanization of ex-perpetrators in post-conflict contexts.

As predicted, agency attributions were undermined due to dehumanization, giving support to a dehumanizing account of perpetrators over a moral typecasting perspective in which offenders are attributed with more agency (Khamitov et al., 2016). As particularly relevant, a consequence of the perception of less human agency via dehumanization is that observers stripe perpetrators of basic rights and protection generally granted to all humans (Khamitov et al., 2016). Since former perpetrators in post-conflicts often face legal transitional justice processes, divesting them of rights and protection contradicts the mechanism itself and has a direct influence on observers’ retributive judgments.

Another consequence of attributing less human agency to dehumanized perpetrators is the negative effect on observers` identification and empathy. Perceiving perpetrators as lacking traits typically granted to oneself like self-recognition, meta-cognition, or a sense of responsibility facilitates the perception of ex-perpetrators as fundamentally different beings. Previous research has evidenced that observers tend to attribute more humanness to themselves than to others

(Haslam & Bain, 2007; Haslam, Bain, Douge, Lee, & Bastian, 2005) and that this self- humanizing bias is associated with a reduced perspective taking and empathy (Park, Haslam,

Kashima, & Norasakkunkit, 2016). A failure to empathize is rooted in the perception of the other as shallow and emotionally impoverished, and this lack of empathy is predicted by mechanistic dehumanization (Haslam, 2006). Our results corroborate the effect of mechanistic DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 22 dehumanization on diminished agency perceptions, but not the direct effect on empathy which is a promising variable to explore in the future. Since empathy has been identified as a strong positive predictor of intergroup forgiveness in other post-conflict societies such as Chile and

Northern Ireland (Noor, Brown, Gonzalez, Manzi, & Lewis, 2008; Tam et al., 2008), the humanization of enemies is of particular importance to solve intractable conflicts and overcome strong feelings like hate (Staub, 2005), and in the case of ex-perpetrators is essential for their reincorporation to society.

The results of the present study also corroborate the strengthening effect of dehumanization on retributive judgments. Participants on average attributed more blame in the dehumanization condition and allocated 7.2 more years of prison, which is significant considering that the transitional justice systems in post-conflict societies contemplate punishments that sometimes do not even involve a prison sentence. Previous research evidenced that dehumanization predicts greater blame and harsher punishment attributions across different crimes (e.g. child molestation, violent and white collar crimes) and across the severity of these crimes (Bastian et al., 2013). Our results extend these findings to transgressors known to have committed severe crimes and in the process of being reincorporated to society.

This relationship between dehumanization and harsher retributive judgments could also be explained from a moral disengagement perspective. Previous research suggests that people tend to infra-humanize out-groups whose suffering or immoral treatment is the responsibility of the in-group, and by these means, justify violent behavior towards them and disengage from self- sanctions (Castano & Giner-Sorolla, 2006). While violent behavior or aggression for punishing the out-group predicts its dehumanization, the opposite direction could also be quite reasonable, DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 23 and dehumanization may predict less tolerance and harsher punishments towards perpetrators as we found.

Regarding blame, our results also corroborate previous positive associations between the dehumanization of perpetrators and higher blame attributions. For example, Bastian et al., (2013) found that participants judged dehumanized criminals as more blameworthy than non- dehumanized ones across different crimes (i.e. violent crimes, child molestation, and white collar crimes). Another study revealed that animalistic dehumanization predicts perceptions of incapacity of the criminal and internal causal attributions of the crime, and led to harsher blame attributions and sentence decisions (Tsukamoto & Karasawa, 2015).

Although blame and punishment attributions are inevitable justice motivations in observers’ judgments towards perpetrators of war crimes, it is worth highlighting the strong influence that even the way of referring to them has on the severity of these judgments. Another objective and subjective criteria such as intentionality, crime type, severity, or utilitarian motives, seems to be overshadowed by implicit dehumanizing clues like the ones tested in the present study. Relatedly, Zebel, Zimmermann, Viki, & Doosje (2008) found that when Muslim

Serbs were dehumanized, reparation policies favoring them were less supported by observers in the context of the Bosnian conflict. In a transitional justice context, this severity of punishment assignment by observers could clash against restorative and special justice conditions for perpetrators and could increase observers’ nonconformity and impunity beliefs.

In line with our predictions, participants attributed less suitableness for a resocialization program to the ex-perpetrator in the dehumanization condition. This particular finding is congruent with a self-regulatory moral perspective (Bandura, 2002; Hoffman, 1998). When targets are granted qualities such as civility, refinement, or rationality, they are also attributed DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 24 inhibitive moral agency, which is the capacity of moral restraint and self-control. Moreover, when individuals are perceived as possessing warmth and emotionality, they are viewed as more morally worthy, and suitable for rehabilitation (Bastian et al., 2011). Positive perceptions of inhibitive moral agency and moral worthiness are central for a DDR process, taking into account the no repetition and recidivism guarantees demanded by society as well as the perceived potential for resocialization of perpetrators. Further studies should explore in depth this effect of dehumanization on inhibitive moral agency attributions in post-conflict contexts, using, for example, more specific measures of attributed self-control and moral restraint in hypothetical recidivism scenarios.

Finally, participants showed higher social distance towards FARC members in the dehumanization condition, giving support to previous findings regarding the relationship between dehumanization and psychological distance (Haslam, 2006). Social distance scales for measuring the level of understanding and intimacy with others have been widely used worldwide to evidenced prejudice attitudes towards a variety of social groups including immigrants, homosexuals, disabled and ill people, and specific racial, religious, ethnic, nationality and occupational groups (Wark & Galliher, 2007; Adewuya & Makanjuola, 2008; Ouellette-Kuntz,

Burge, Brown, & Arsenault, 2010). Typically, greater social distance is related to the perception of out-group members as abstract beings lacking psychological richness, and thus, hindering their acceptance as members of the in-group. For example, it has been evidenced that stigmatizing attitudes towards people with intellectual disabilities increase observers’ social distance towards them and therefore their social inclusion (Ouellette-Kuntz et al., 2010).

The social distance product of dehumanization evidenced in this study could be also explained on the basis of the relationship between more hardwired disgust reactions and social DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 25 attitudes. Some research, for example, has provided evidence suggesting that dehumanization of the out-group mediates the association between interpersonal disgust/revulsion and negative intergroup attitudes (Hodson & Costello, 2017). In the same line, disease avoidance has been associated with xenophobic (Faulkner, Schaller, Park, & Duncan, 2004) and ethnocentric attitudes (Navarrete & Fessler, 2006), and disgust reactions have been linked to purity and social order protection (Rozin, Lowery, Imada, & Haidt, 1999). It is likely that former perpetrators are perceived as a threat to social order or values, and in a more implicit way as carriers of ‘vices’ and thus targets of moral disgust. Although it is rather natural to feel certain levels of social distance with the ex-perpetrators, it is worth underscoring the powerful influence that even a dehumanizing discourse has on how this distance is increased, and how this could difficult a

DDR process.

Perceiving other’s similarity to oneself is important for judging their actions. Research suggests that greater levels of perceived interpersonal similarity (i.e. reduced social distance) increase the weight that people give to subordinate information in judging others’ actions (e.g. in end-related terms more than means-related terms, feasible more than desirability concerns, and secondary features of others decisions) (Liviatan, Trope, & Liberman, 2008). Perceiving a sub- group of the society like the former perpetrators as socially distant entities negatively affect their acceptance and resocialization to civilian life. Although such endeavour pose a challenge to any society, dehumanizing frames in news media increases social distance as evidenced in our results and do not facilitate the process.

Strengths and limitations

The present study faces some limitations. First, there could be some non-controlled features between news articles across conditions such as references to geographical settings, the DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 26 extension of the text, or certain narrative aspects that were included to avoid making participants aware of the manipulation. Although the methodological advantages of a within-subjects design as the one we employed imply this cost, and we strived to minimize it, future studies could consider using a between-subjects design or a time gap between conditions to reduce variations between stimuli. Although these differences could be argued to explain some of the effects we found, we believe this is unlikely, as the results show that the manipulation had its intended effects, and the effects are consistent with all predictions. Second, the present study focused on a specific post-conflict context which has its own historical, cultural, and social characteristics so an avenue of future research could be to test these effects in other post-war contexts to generalize the results. Finally, including a third control condition with neutral speech could be useful to confirm the effects.

Despite these limitations, the present study has also several strengths. Our sample is balanced in terms of sex, age and occupation and was not taken from a western, educated, industrialized, rich and democratic (WEIRD) society, which benefits external validity. We also experimentally manipulated dehumanization unlike previous research which only measured it as a dependent variable or a covariate. In addition, we used a strong statistical procedure (LME), with random intercepts, that allows us to keep into account individual differences for example in how harsh are participants when judging the hypothetical FARC members giving us a better picture of the effect of our manipulation. Our study is also timely and grounded in relevant socio-political issues currently happening, and applicable to similar dehumanization speech used in other post-conflict societies facing similar challenges. Finally, the results of our study could be informative for diverse disciplines including communication sciences, criminology, justice research, and political science. DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 27

Conclusions

Observers attributed less agency, displayed more severe retributive justice judgments, attributed less suitableness for a resocialization program, and showed more social distance towards former perpetrators when they were presented in a dehumanizing discourse frame.

Despite the undeniable responsibility of former violent groups in war-related crimes, the dehumanization of its members could pose more obstacles than advantages to overcome war in a post-conflict scenario. As evidenced in this study, just the use of a humanizing speech towards ex-perpetrators, such as mentioning their name, age, origin, emotional or mental states, is enough for eliciting more agency attributions, social closeness, and positive attitudes towards their reincorporation to civil society. It is natural for observers to condemn and judge the perpetrators of violence in the aftermath of an armed conflict, but it is also important to recognize the subtle ways in which public discourse could shape observers’ attributions, judgments, and attitudes towards perpetrators, which are paramount for the success of a DDR process. The present study could be useful for understanding how dehumanization discourses towards perpetrators could affect observers’ judgments and attitudes in post-conflict societies, and for raise awareness within journalist, editors, and policymakers about these effects to promote a debate about its effects.

DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 28

References

Adewuya, A. O., & Makanjuola, R. O. A. (2008). Social Distance Towards People with Mental

Illness in Southwestern Nigeria. Australian & New Zealand Journal of Psychiatry, 42(5),

389–395. https://doi.org/10.1080/00048670801961115

Annan, J., & Cutter, A. (2009). Critical issues and lessons in social reintegration: Balancing

justice, psychological well being, and community reconciliation. Retrieved from

http://cartagenaddr.org/literature_press/ART_21.pdf

Bandura, A. (1999). Social Cognitive Theory: An Agentic Perspective. Asian Journal of Social

Psychology, 2(1), 21–41. https://doi.org/10.1111/1467-839X.00024

Bandura, A. (2002). Selective Moral Disengagement in the Exercise of Moral Agency. Journal

of Moral Education, 31(2), 101–119. https://doi.org/10.1080/0305724022014322

Barr, D. J., Levy, R., Scheepers, C., & Tily, H. J. (2013). Random effects structure for

confirmatory hypothesis testing: Keep it maximal. Journal of Memory and Language, 68(3),

255–278. https://doi.org/10.1016/j.jml.2012.11.001

Bastian, B., Denson, T. F., & Haslam, N. (2013). The Roles of Dehumanization and Moral

Outrage in Retributive Justice. PLoS ONE, 8(4).

https://doi.org/10.1371/journal.pone.0061842

Bastian, B., Laham, S. M., Wilson, S., Haslam, N., & Koval, P. (2011). Blaming, praising, and

protecting our humanity: The implications of everyday dehumanization for judgments of

moral status. British Journal of Social Psychology, 50(3), 469–483.

https://doi.org/10.1348/014466610X521383

Bøås, M., & Hatløy, A. (2008). “Getting in, Getting out”: Militia Membership and Prospects for

Re-integration in Post-War Liberia. The Journal of Modern African Studies, 46, 33–55. DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 29

https://doi.org/10.2307/30224873

Castano, E., & Giner-Sorolla, R. (2006). Not quite human: Infrahumanization in response to

collective responsibility for intergroup killing. Journal of Personality and Social

Psychology, 90(5), 804–818. https://doi.org/10.1037/0022-3514.90.5.804

Castellanos, E. (2014). Discurso e ideología de Álvaro Uribe Vélez sobre las guerrillas

colombianas y su impacto en los procesos de paz en Colombia. Discurso y Sociedad, 8(2),

182–209. Retrieved from https://dialnet.unirioja.es/servlet/articulo?codigo=5372810

Chambon, V., Filevich, E., & Haggard, P. (2014). What is the Human Sense of Agency, and is it

Metacognitive? In The Cognitive Neuroscience of Metacognition (pp. 321–342). Berlin,

Heidelberg: Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-45190-4_14

Christensen, L. B., Johnson, B., & Turner, L. A. (2015). Research methods, design, and analysis

(12th ed.). Essex: Pearson Education. Retrieved from

https://books.google.com.co/books?id=eQaqBwAAQBAJ&dq=research+methods+Christen

sen&hl=es&sa=X&ved=0ahUKEwjq4Zy7qLDZAhWRjVkKHWPnBDIQ6AEIJjAA

Eisenberg, N., Cumberland, A., Guthrie, I. K., Murphy, B. C., & Shepard, S. A. (2005). Age

Changes in Prosocial Responding and Moral Reasoning in Adolescence and Early

Adulthood. Journal of Research on Adolescence, 15(3), 235–260.

https://doi.org/10.1111/j.1532-7795.2005.00095.x

Farc tienen 5.765 miembros en sus filas. (2016, September 28). El Tiempo. Retrieved from

http://www.eltiempo.com/politica/proceso-de-paz/numero-de-guerrilleros-de-las-farc-en-

sus-filas-33484

Faulkner, J., Schaller, M., Park, J. H., & Duncan, L. A. (2004). Evolved Disease-Avoidance

Mechanisms and Contemporary Xenophobic Attitudes. Group Processes & Intergroup DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 30

Relations, 7(4), 333–353. https://doi.org/10.1177/1368430204046142

Fiske, A. (2004). Four modes of constituting relationships: Consubstantial assimilation; space,

magnitude, time, and force; concrete procedures; abstract symbolism.itle. In N. Haslam

(Ed.), Relational models theory: A contemporary overview (pp. 61–146). Mahwah, NJ:

Lawrence Erlbaum Associates.

Gamson, W. A., & Modigliani, A. (1989). Media Discourse and Public Opinion on Nuclear

Power: A Constructionist Approach. American Journal of Sociology, 95(1), 1–37.

https://doi.org/10.1086/229213

García-Duran, M. (2004). Colombia: retos y dilemas en la búsqueda de la paz. In M. Garcia-

Duran (Ed.), Alternativas a la guerra, iniciativas y procesos de paz en Colombia (pp. 4–9).

Bogotá: CINEP.

García-Marrugo, A. (2013). What’s in a name? The representation of illegal actors in the internal

conflict in the Colombian press. Discourse & Society, 24(4), 421–445.

https://doi.org/10.1177/0957926513482063

Gobierno calcula que 17.500 miembros de las Farc se desmovilizarán con la paz. (2016, March

15). El Espectador. Retrieved from http://www.elespectador.com/noticias/paz/gobierno-

calcula-17500-miembros-de-farc-se-desmovilizar-articulo-622360

Gómez, S. (2008, July). ‘Las Farc son animales y bárbaros sin humanidad’ - Archivo Digital de

Noticias de Colombia y el Mundo desde 1.990 - eltiempo.com. El Tiempo. Retrieved from

http://www.eltiempo.com/archivo/documento/MAM-3011490

Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of Mind Perception. Science,

315(5812), 619–619. https://doi.org/10.1126/science.1134475

Gray, K., & Wegner, D. (2009). Moral typecasting: Divergent perceptions of moral agents and DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 31

moral patients. Journal of Personality and Social Psychology, 96(3), 505–520.

https://doi.org/10.1037/a0013748

Grupo de Memoria Histórica. (2013). ¡BASTA YA! Colombia: Memorias de guerra y dignidad.

Bogotá: Imprenta Nacional. Retrieved from

http://www.centrodememoriahistorica.gov.co/descargas/informes2013/bastaYa/basta-ya-

colombia-memorias-de-guerra-y-dignidad-2016.pdf

Gutierrez, L. (2007). La prensa como creadora de estereotipos sobre los reinsertados y el proceso

de paz en Colombia. Palabra Clave, 10(2), 11–25.

Hangman, L., & Nielsen, Z. (2002). A framework for lasting disarmament, demobilization, and

reintegration of former combatants in crisis situations. International Peace Academy, 12–

13.

Haslam, N. (2006). Dehumanization : An Integrative Review. Personality and Social Psychology

Review, 10(3), 252–264. https://doi.org/10.1207/s15327957pspr1003

Haslam, N., & Bain, P. (2007). Humanizing the Self: Moderators of the Attribution of Lesser

Humanness to Others. Personality and Social Psychology Bulletin, 33(1), 57–68.

https://doi.org/10.1177/0146167206293191

Haslam, N., Bain, P., Douge, L., Lee, M., & Bastian, B. (2005). More human than you:

Attributing humanness to self and others. Journal of Personality and Social Psychology,

89(6), 937–950. https://doi.org/10.1037/0022-3514.89.6.937

Haslam, N., & Loughnan, S. (2014). Dehumanization and Infrahumanization. Annual Review of

Psychology, 65(1), 399–423. https://doi.org/10.1146/annurev-psych-010213-115045

Haslam, N., & Stratemeyer, M. (2016). Recent research on dehumanization. Current Opinion in

Psychology, 11, 25–29. https://doi.org/10.1016/j.copsyc.2016.03.009 DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 32

Hodson, G., & Costello, K. (2017). Interpersonal Disgust , Ideological Orientations , and of Inter

group Attitudes Dehumanization as Predictors, 18(8), 691–698.

Hoffman, M. L. (1998). Varieties of Empathy-Based Guilt. In Guilt and Children (pp. 91–112).

New York: Academic Press. https://doi.org/10.1016/B978-012148610-5/50005-9

Human Rights Watch, World Report 2017: Colombia. (2017). Retrieved June 29, 2017, from

https://www.hrw.org/world-report/2017/country-chapters/colombia

Jennings, K. M. (2007). The Struggle to Satisfy: DDR Through the Eyes of Ex-combatants in

Liberia. International Peacekeeping, 14(2), 204–218.

https://doi.org/10.1080/13533310601150800

Khamitov, M., Rotman, J. D., & Piazza, J. (2016). Perceiving the agency of harmful agents: A

test of dehumanization versus moral typecasting accounts. Cognition, 146, 33–47.

https://doi.org/10.1016/j.cognition.2015.09.009

Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). lmerTest Package: Tests in

Linear Mixed Effects Models. Journal of Statistical Software, 82(13), 1–26.

https://doi.org/10.18637/jss.v082.i13

Lindén, M., Björklund, F., & Bäckström, M. (2016). What makes authoritarian and socially

dominant people more positive to using torture in the war on terrorism? Personality and

Individual Differences, 91, 98–101. https://doi.org/10.1016/j.paid.2015.11.058

Liviatan, I., Trope, Y., & Liberman, N. (2008). Interpersonal similarity as a social distance

dimension: Implications for perception of others’ actions. Journal of Experimental Social

Psychology, 44(5), 1256–1269. https://doi.org/10.1016/j.jesp.2008.04.007

Lüdecke, D. (2018). sjPlot: Data Visualization for Statistics in Social Science.

https://doi.org/10.5281/zenodo.1308157 DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 33

Luke, S. G. (2017). Evaluating significance in linear mixed-effects models in R. Behavior

Research Methods, 49(4), 1494–1502. https://doi.org/10.3758/s13428-016-0809-y

Narvaez, D., & Lapsley, D. K. (2009). Moral Identity, Moral Functioning, and the Development

of Moral Character. In D. Bartels, C. Bauman, L. Skitka, & D. Medin (Eds.), The

psychology of learning and motivation (Vol. 50, pp. 237–274). Burlington: Elesvier Inc.

https://doi.org/10.1016/S0079-7421(08)00408-8

Navarrete, C. D., & Fessler, D. M. T. (2006). Disease avoidance and ethnocentrism: the effects

of disease vulnerability and disgust sensitivity on intergroup attitudes. Evolution and

Human Behavior, 27(4), 270–282.

https://doi.org/10.1016/J.EVOLHUMBEHAV.2005.12.001

Noor, M., Brown, R., Gonzalez, R., Manzi, J., & Lewis, C. A. (2008). On Positive Psychological

Outcomes: What Helps Groups With a History of Conflict to Forgive and Reconcile With

Each Other? Personality and Social Psychology Bulletin, 34(6), 819–832.

https://doi.org/10.1177/0146167208315555

Nussio, E. (2013). Desarme, desmovilización y reintegración de excombatientes: políticas y

actores del postconflicto. Colombia Internacional, 77, 9–16.

Ouellette-Kuntz, H., Burge, P., Brown, H. K., & Arsenault, E. (2010). Public Attitudes Towards

Individuals with Intellectual Disabilities as Measured by the Concept of Social Distance.

Journal of Applied Research in Intellectual Disabilities, 23(2), 132–142.

https://doi.org/10.1111/j.1468-3148.2009.00514.x

Özerdem, A., Podder, S., O’Callaghan, S., & Pantuliano, S. (2008). Reinsertion Assistance and

the Reintegration of Ex-Combatants in War to Peace Transitions.

Park, J., Haslam, N., Kashima, Y., & Norasakkunkit, V. (2016). Empathy, culture and self- DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 34

humanising: Empathising reduces the attribution of greater humanness to the self more in

Japan than Australia. International Journal of Psychology, 51(4), 301–306.

https://doi.org/10.1002/ijop.12164

R Core Team. (2018). R: A language and environment for statistical computing. Vienna, Austria:

R Foundation for Statistical Computing.

Rozin, P., Lowery, L., Imada, S., & Haidt, J. (1999). The CAD triad hypothesis: a mapping

between three moral emotions (contempt, anger, disgust) and three moral codes

(community, autonomy, divinity). Journal of Personality and Social Psychology, 76(4),

574–586. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10234846

Sanchez, F. J., Diaz, A. M., & Fornisano, M. (2003). Conflicto, violencia y actividad criminal en

Colombia: Un análisis espacial. Archivos de Economía. Retrieved from

http://econpapers.repec.org/paper/col000118/002185.htm

Sánchez, L. S., López, W. L., & Barreto, I. (2013). Enmarcamiento cognitivo de la cultura

política. un análisis desde las redes sociales en internet (twitter). Interamerican Journal of

Psychology, 47(3), 383–394.

Scheufele, D. A., & Tewksbury, D. (2007). Framing, agenda setting, and priming: The evolution

of three media effects models. Journal of Communication, 57(1), 9–20.

https://doi.org/10.1111/j.1460-2466.2006.00326.x

Sheafer, T., & Dvir-Gvirsman, S. (2010). The spoiler effect: Framing attitudes and expectations

toward peace. Journal of Peace Research. Sage Publications, Ltd.

https://doi.org/10.2307/25654556

Staub, E. (2005). The Origins and Evolution of Hate, With Notes on Prevention. In The

psychology of hate. (pp. 51–66). Washington: American Psychological Association. DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 35

https://doi.org/10.1037/10930-003

Tajfel, H. (1981). Human groups and social categories : studies in social psychology. Cambridge

University Press. Retrieved from

http://www.cambridge.org/co/academic/subjects/psychology/social-psychology/human-

groups-and-social-categories-studies-social-

psychology?format=HB&isbn=9780521280730#941zYSe6I4g4oDuX.97

Tam, T., Hewstone, M., Kenworthy, J. B., Cairns, E., Marinetti, C., Geddes, L., & Parkinson, B.

(2008). Postconflict Reconciliation: Intergroup Forgiveness and Implicit Biases in Northern

Ireland. Journal of Social Issues, 64(2), 303–320. https://doi.org/10.1111/j.1540-

4560.2008.00563.x

Theidon, K. (2009). Reconstrucción de la masculinidad y reintegración de excombatientes en

Colombia. Bogotá: Fundación ideas para la paz.

Tsukamoto, S., & Karasawa, M. (2015). Dehumanization in the Judicial System: The Effect of

Animalization and Mechanization of Defendants on Blame Attribution. In Proceeding of the

10th Asian Association of Social Psychology Biennial Conference (pp. 256–267).

Wark, C., & Galliher, J. F. (2007). Emory Bogardus and the Origins of the Social Distance Scale.

The American Sociologist, 38(4), 383–395. https://doi.org/10.1007/s12108-007-9023-9

Weisel, O., & Böhm, R. (2015). “Ingroup love” and “outgroup hate” in intergroup conflict

between natural groups. Journal of Experimental Social Psychology, 60, 110–120.

https://doi.org/10.1016/J.JESP.2015.04.008

Zebel, S., Zimmermann, A., Viki, T., & Doosje, B. (2008). Dehumanization and Guilt as Distinct

but Related Predictors of Support for Reparation Policies Author ( s ): Sven Zebel , Anja

Zimmermann , G . Tendayi Viki and Bertjan Doosje Source : Political Psychology , Vol . 29 DEHUMANIZATION & ATTRIBUTIONS TOWARDS PERPETRATORS 36

, No . 2 ( Apr ., 2008 ), pp . 193-219. Political Psychology, 29(2), 193–219.

Zimmer-Gembeck, M., & Collins, A. (2003). Autonomy Development during Adolescence. In

G. Adams & M. Berzonsky (Eds.), Blackwell Handbook of Adolescence (pp. 175–204).

Wiley-Blackwell.