Moral struggles in social media discussion: the case of sexist aggression. Francesca D’Errico Giorgia Saleri Fil.Co.Spe Department Psychology Faculty Marinella Paciello Roma Tre University Uninettuno University Psychology Faculty Rome, Italy Rome, Italy Uninettuno University francesca.derrico@uniroma3.it g.saleri@students.uninettunouniversity. Rome, Italy net m.paciello@uninettunouniversity.net l l l ​ The online interactions have become increasingly Abstract— of aggression is relatively new, the phenomena of online negative and aggressive, in particular when stereotyped topics aggression can be understood based on the traditional are discussed. Based on extensive literature on aggression, the models developed in the extensive literature on aggression. present study investigates the forms, the frequencies of the negative comments and the related both emotive and cognitive First based on social cognitive theory [6], the hypothesis processes in response an offensive post toward Carola Rackete. such as the presence of environmental cues that can increase A total of 1.249 comments on Twitter were analyzed by coding aggression through priming mechanisms ​[​7; 8] and more basic position toward the tweet, moral argumentation, specific literature on media and aggression [9], the online underlying moral process, emotions and their intensity. Overall the results showed that online communication is aggressive phenomena can be understood if we consider the largely denoted by negative tone and aggressive forms, even characteristics of online environment. The online when a commenter would support the target of an aggressive interactions occur in a context in which some technological communication. In particular, the gender aggression is carried affordance hinder the possibility to empathize with a out by the same women who became aggressive and much less potential victim (e.g. lack of emotional cues) and the sense prosocial than what one would expect of personal responsibility with respect one’s own action (e.g. anonymity), and at same time foster the toxic online Keywords—discrimination, sexist aggression, prosocial, disinhibition [10]. Moreover, the selective media attention emotion, cognitive processes on aggressive situations that could backfire by providing an I. INTRODUCTION advantageous comparison point to reduce the severity Over the last years the online interactions have become perception of one’s own online aggressive behaviors [11]. In increasingly negative and aggressive (Pew Research Center, these situations, aggressive behavior can be trigger and 2017). It is always more frequent to observe negative online aggressive social modeling can exacerbate negative online phenomena such as hate speech, flaming and trolls in manifestations [11; 12]. Indeed, by observational learning particular related to sensible ethical topics that can polarize processes the exposure to violence leads to the acquisition the public opinion. For instance, some of the most online of cognitive beliefs (es. aggressive behavioral scripts) and discussed topics concern immigration and stereotypes that processes (e.g. cognitive desensitization” to violence) that could lead to explicit forms of cyber-racism and facilitate the adoption of aggressive behaviour. Secondly, discriminations [1; 2]. Moreover, some groups are most based on the F-A hypothesis [13], the transfer of excitation likely to became target of online aggression, as in the case [14], and the dissipation of aggression over time [15], the of girls and woman [3]. exposure to violence can be considered a case of frustrating condition that increase negative emotional arousal If this framework is typically applied to racists encouraging reactive aggressive response. Hostile emotions discussions, recent studies have also shown that regardless such as distress, frustration, anger, contempt, disgust and the ethical positions on discussions on ethical topics can be hate potentially promote aggressive behaviours [15; 16] characterized by a high degree of hostile emotions [4]. because they feature medium/high arousal that, in the Specifically, also people who support potential victims of presence of other aversive stimuli, can progressively discriminatory and offensive can react with the same increase. In addition, hostile emotions, such as anger, could aggressive modalities. Moreover, the spread of negative increase the attention towards provoking events and hostile online emotion is strictly related to cognitive processes that interpretation of neutral situations [17]. Third, aggression maintain and exacerbate the circle of online incivility, such can be resulted by dysfunctional coping strategies [18; 19]. as in the case of dehumanization and attribution of blame to The strategies used by defenders of potential victims range potential victims for their situation [4; 5]. Despite this form from an empathic support victim to angry attack toward a Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) perpetrator [20]. Thus, also in the case a prosocial position, another woman on Facebook to insult Carola Rackete by it is possible to find hostile emotions, in particular, the making assumptions about her sexual habits. On the so-called third-party anger [21; 22] or empathetic anger contrary, the text expressed indignation towards use of [23]. This type of anger is caused not by personal concern sexist hate speech online (with the aim of stigmatize this but by the desire to avoid injustice and restore social equity. sexist conduct): Moreover, the study of ‘Donne che se la ridono condividendo ‘sta roba. Ho stimulus-thought-emotion-reaction sequence for helping esaurito le parole.’ versus aggressive response could be particularly interesting ‘Women who laugh at sharing this stuff. I've run out of in the case of stereotyped topic, as in the case of woman. words.’ Several studies have attested that member of stereotypes and prejudice can became more easily target of hate-motivated aggression [3; 24]. In this case the aggression, usually acted B. Data and Procedure by multiple perpetrators, has be motivated not only by the intent to hurt one victim but all member of victim’s social A total of 1.249 comments made in response to category. Previous findings stressed the importance of Selvaggia Lucarelli’s post, distributed from the 2nd to the considering social category as well the gender of target and 3rd of July 2019, were extracted through Twitter API. agent of discriminatory and stereotyped forms of aggression Authors can privately share extracted data, by guaranteeing [25]. Literature has suggested that the online verbal commenters’ anonymity, in the interest of repeatability of aggression against girl and woman is related to primary analysis. All comments were manually filtered in order to intention of intimidate and control them [3; 26], especially better understand users’ point of view and exclude tweets in the case of traditional ‘male domains’ such as publicly that no longer exist. First of all, coders identified a level of relevant contexts. Based on these premises, the present agreement or disagreement with Lucarelli’s statement. In research questions are: addition to this coding, the valence and intensity of emotions were examined considering both semantic and 1. which are the emotional features of prosocial and formal elements. Data coding carried out in this way led to proself stance toward an exhibition of a clear attack the identification of basic position toward the tweet (pro and to woman image? contro), moral argumentation/reasoning (aggressive, proself 2. Are the prosocial stances toward a woman public and prosocial), underlying moral process based on Bandura aggression featured by a third-party anger? and Darley and Latanè models [27; 28] (see Table 1) and four emotions (joy, calm, anger and sadness) and their intensity (low vs high). Another control element in this Specifically, in the present study we investigate the investigation was the presence or absence of gender forms, the frequencies of the negative comments in response stereotypes in analyzed tweets. a post, and one of its comment, that intends to TABLE I. PROSOCIAL AND PROSELF ARGUMENTS declare/exhibit an offensive message toward a specific woman who actively defend the right of a minority group Cognitive Processes Loci (immigrants), and the women more in general. Based on Prosocial Processes Proself Processes above theoretical approaches and empirical studies, we Victim Humanization Dehumanization expected to find aggressive comments also from Value Attribution Blame Attribution commenters in defense of the attacked woman. We expected Behaviour Principle Recalling Moral Justification that both prosocial and proself stance will be featured by high levels of negative emotional activation. Realistic Labelling Euphemistic Labelling Structured Advantageous II. METHOD Argumentation Comparison Consequences Positive Consequences A. Online Communicative Scenario Disconfirm of Negative Consequences Distortion The online communicative scenario relevant to this Consequences (of hosting) study is related to a message posted by Selvaggia Lucarelli, Exagerate Negative a popular Italian journalist and TV commentator, on July Consequences Diffusion of 2nd, 2019. The tweet was intended to condemn cruel and Agency Agentic Trigger Responsibility rude criticisms among women after the Sea-Watch case: at Assumption of Displacement of the end of June, the civil rescue vessel Sea-Watch 3, Responsability Responsibility captained by Carola Rackete, entered the port of Lampedusa without permission with 42 migrants on board. (https://www.theguardian.com/world/2019/jun/29/sea-watch III. RESULTS -captain-carola-rackete-arrested-italian-blockade) These A Chi Square analysis ​[χ​ (​1296​)​= 87,4; p< .001​] pointed 2​ facts have sparked heated debates in public opinion. The out how commenters who openly support the main tweet are tweet was composed of text and an image. The image was a featured by negative emotions, mainly anger (49,3%) and screenshot representing a dark humor meme shared by sadness (35,1%), while commenters against the main tweet Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) express their positions in an angry (66,5%) or happy way nave della guardia di finanza è un’azione giustificata; (19.9%). Transl: ‘​But what are you saying ??? Now ramming a finance guard ship is justified​’). The difference between pro and agianst position are significant to the chi square analysis TABLE II. POSITION TOWARD THE MAIN [χ​2​(​697​)​= 164; p< .001​]​. The difference between Pro and TWEET*EMOTIONS Against the denouncing tweet of Lucarelli is very close to Emotions the gender difference, in the sense that women tend to Anger Sadness Joy Calm dehumanize or attribute guilt or on the other hand assuming Pro 49.3% 35.1% 12.8% 2.8% their responsibilities’ or recall to a legal principle. Agains 66.5% 6.8% 19.9% 6.8% t TABLE IV. PRO VS AGAINST POSITION*TYPE OF Neutral 16.7% 50.0% 16.7% 16.7% PROCESSES Proself Processes Rarely commenters in support of the journalist Lucarelli Pro Against expressed their opinion by using a prosocial argumentation Dehumanization 26.0% 9.9% but they are mainly aggressive (63,5%); while people Blame Attribution 21.8% 25.7% against the act of public stigmatization express their Advantageous comment with a proself (56,3%) or aggressive act (33,5%) Comparison 0.8% 12.9% but in a very less extent compared to the pro position Diffusion of 4.6% 2.9% [χ​2​(​1296​)​= 162; p< .001​]​. Responsibility Displacement of 10.7% 6.4% Responsibility TABLE III. POSITION TOWARD THE MAIN Consequences 9.2% 7.6% TWEET*MORAL STANCE Distortion Euphemistic 2.7% 16.4% Moral Stance Labelling Aggressive Proself Prosocial Moral 0.0% 1.2% Justification Pro 63.8% 17.6% 18.6% Prosocial Agains Processes Pro Against 33.5% 56.3% 10.2% t Principle Neutral 57.5% 25.4% 17.0% 6.5% 5.8% Recalling Structured 4.2% 9.9% Argumentation In particular when we consider the processes emerging Assumption of 6.9% 0.0% from the commenters words we unexpectedly found that Responsability people who support Lucarelli act of public indignation for Value Attribution 0.6% 0.6% the sexist tweet report a greater percentage of Disconfirm of dehumanization (26%)- more than the unsupportive Negative 0.8% 0.0% Consequences commenters (9,9%) - and guilt attribution (21,8%) toward Exagerate the woman who make fun of Carola. In this sense the Negative 1.1% 0.0% woman became another victim in the ‘pro’ commenters. Consequences Another frequent process activated by the Lucarelli Realistic 1.7% 0.0% supporters is the responsibility displacement (10,7%), when Labelling commenters tend to attribute the responsibility of this Agentic Trigger 1.7% 0.0% ‘cultural level’ to politicians as in the case ‘Il fatto che il Humanization 1.0% 0.6% ministro del food blogging non condanni mai questa violenza verbale e scritta, di fatto, la sdogana’; on the other side commenters in favour of Lucarelli assume to When we consider the expressed emotions with respect themselves the shame of belonging to the female genre (Mi to the moral stances we can report several significant vergogno di essere donna...’: ‘​I am ashamed of being a differences ​[χ​2​(​1296​)​= 189; p< .001​] in the sense that the woman​’) A minimal part of the coded processes are devoted aggressive one as we could expect is featured by anger to principle recalling, women who ask if it is possible to (59,8%) similarly to the proself one (49,1%) and differently denounce. Differently the commenters against the Lucarelli from part of prosocial stance that is featured mainly by public act are mainly focused to the guilt attribution to the sadness (63,2%) but also anger (29,2%). But a tendency that journalist or to Carola (25,7%) or to labelling characterized the aggressive and the proself is the presence euphemistically the indignation act as a political strategy of good percentages of joy (18,8% and 12,3% respectively) toward the vice minister Matteo Salvini or comparing to the that signal how the aggression and not supportive toward deviant act of the woman to the Carola ‘aggression‘ to women (against the aggression) is expressed by derision, Italian law (Ma che cosa stai dicendo??? Ora speronare una Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) irony and ridicule that can be considered an indirect act of increasingly leave people "protected" behind their screen aggression [29]. and in peace with their conscience despite the words written (aggressive, sarcastic, violent). It is conceivable that the modeling processes are working and the presence of pervasive aggressive cues support these dynamics that people might not necessarily be aware of. Future studies should focus on the possible moderators of these dynamics and the role of awareness in a process that seems above all supported by unruly emotions and processes that denote a loss of moral control. REFERENCES [1] Navarrete, C. D., McDonald, M. M., Molina, L. E., & Sidanius, J. (2010). Prejudice at the nexus of race and gender: an outgroup male target hypothesis. Journal of personality and social psychology, 98(6), Fig. 1. Moral stance*Emotions 933. [2] Faulkner, N., & Bliuc, A. M. (2016). ‘It’s okay to be racist’: moral disengagement in online discussions of racist incidents in Australia. Ethnic and Racial Studies, 39(14), 2545-2563. IV. DISCUSSION [3] Pacilli, M. G., & Mannarini, T. (2019). Are women welcome on The present study confirmed how online communication facebook? a study of facebook profiles of italian female and male public figures. Tpm: Testing, Psychometrics, Methodology in can be largely denoted by negative tone and aggressive Applied Psychology, 26(2). forms in most of the case, even when a commenter would [4] D’Errico, F., & Paciello, M. (2019). Online moral struggles in hosting support the target of aggressive communication. The immigrant’s discourses: The underlying role of expressed anger and denounce of an aggressive behavior towards a victim only socio-cognitive processes. Journal of Language Aggression and partly stimulates users’ sorrow or other deactivated Conflict. 7(2); 10.1075/jlac.00024.der emotional states that could lead to a supporting scenario for [5] D’Errico, F., & Paciello, M. (2018). Online moral disengagement and hostile emotions in discussions on hosting immigrants. Internet the victim of the aggression (Carola Rackete). Instead both Research, 28(5), 1313-1335. supporters and unsupporters of the journalist’s denounce [6] Bandura, A. (1978). Social learning theory of aggression. Journal of commented being mainly angry. Anger, that, as in other communication, 28(3), 12-29. studies, was associated with a series of aggressive [7] Engelhardt, C. R., & Bartholow, B. D. (2013). Effects of situational cognitions that feed the vicious cycle of aggression. In cues on aggressive behavior. Social and Personality Psychology particular, we saw that users who support the sexist Compass, 7(10), 762-774. denounce on their side attribute the blame and dehumanize, [8] Berkowitz, L., & LePage, A. (1967). Weapons as aggression-eliciting and they are more aggressive and less prosocial than stimuli. Journal of Personality and Social Psychology, 7(2p1), 202. expectations. Pro sexists denounce people even get more [9] Anderson, C. A., & Bushman, B. J. (2001). Effects of violent video games on aggressive behavior, aggressive cognition, aggressive enraged than against ones, in which, on the contrary, affect, physiological arousal, and prosocial behavior: A meta-analytic positive emotions emerged accompanied by acts of ridicule review of the scientific literature. Psychological science, 12(5), or derision, presumably deriving from the attempt to use 353-359. humor to diminish and defend one's own social image [29; [10] Lapidot-Lefler, N., & Barak, A. (2012). Effects of anonymity, 30; 31]. The proself unexpectedly dehumanize less and invisibility, and lack of eye-contact on toxic online disinhibition. Computers in human behavior, 28(2), 434-443. mostly focus on behavior with ‘euphemistic labeling’ and [11] Runions, K. C., & Bak, M. (2015). Online moral disengagement, ‘advantageous comparison’, dampening the force of the cyberbullying, and cyber-aggression. ​Cyberpsychology, Behavior, public denounce through lower emotional activation. On the and Social Networking​, ​18​(7), 400-405. social network the gender aggression is carried out by the [12] Zimmerman, A. G., & Ybarra, G. J. (2016). Online aggression: The same women who became aggressive and much less influences of anonymity and social modeling. Psychology of Popular prosocial than what one would expect. The high occurrence Media Culture, 5(2), 181. of dehumanization mechanisms could be explained by the [13] Dollard, J., Miller, N. E., Doob, L. W., Mowrer, O. H., & Sears, R. R. (1939). Frustration and aggression. fact that the Lucarelli’s post focused not only on sexist [14] Zillmann, D. (1979). Hostility and aggression. attack against Carola Rackete, but also on woman who [15] Konecni, V. J. (1975). The mediation of aggressive behavior: Arousal made that attack. Thus, it is possible that dehumanization is level versus anger and cognitive labeling. Journal of Personality and also related to the fact that this specific woman represents a Social Psychology, 32(4), 706. member of a "political" outgroup [32], that is the citizens [16] Berkowitz, L. (1989). Frustration-aggression hypothesis: Examination sustaining the political forces against the hosting toward and reformulation. Psychological bulletin, 106(1), 59. immigrants. The study confirmed that hate scenario can be [17] Zillmann, D. (1988). Cognition-excitation interdependences in spread on the web highlighting the affective and cognitive aggressive behavior. Aggressive Behavior, 14(1), 51-64. processes that can feed it. The reference to the real world is [18] Averill, J. R. (1983). Studies on anger and aggression: implications filtered by mechanisms of non-responsibility that for theories of emotion. American psychologist, 38(11), 1145. Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) [19] Lazarus, R. S. (1991). Progress on a cognitive-motivational-relational [27] Bandura, Albert. 2015. Moral Disengagement: How People Do Harm theory of emotion. American psychologist, 46(8), 819. and Live with Themselves. New York: Macmillan Higher Education. [20] Rudolph, U., Roesch, S., Greitemeyer, T., & Weiner, B. (2004). A [28] Darley, J. M., & Latané, B. (1968). Bystander intervention in meta-analytic review of help giving and aggression from an emergencies: Diffusion of responsibility. Journal of personality and attributional perspective: Contributions to a general theory of social psychology, 8(4p1), 377. motivation. Cognition and emotion, 18(6), 815-848. [29] D’Errico, F., & Poggi, I. (2016). “The Bitter Laughter”. When Parody [21] Van Doorn, J., Zeelenberg, M., & Breugelmans, S. M. (2014). Anger Is a Moral and Affective Priming in Political Persuasion. Frontiers in and prosocial behavior. Emotion Review, 6(3), 261-268. psychology, 7, 1144. [22] Vitaglione, G. D., & Barnett, M. A. (2003). Assessing a new [30] D’Errico, F., & Poggi, I. (2012). Blame the opponent! Effects of dimension of empathy: Empathic anger as a predictor of helping and multimodal discrediting moves in public debates. ​Cognitive punishing desires. Motivation and Emotion, 27(4), 301-325. Computation​, ​4​(4), 460-476. [23] Batson, C. D., Kennedy, C. L., Nord, L. A., Stocks, E. L., Fleming, D. [31] Hodson, G., & MacInnis, C. C. (2016). Derogating humor as a Y. A., Marzette, C. M., ... & Zerger, T. (2007). Anger at unfairness: Is delegitimization strategy in intergroup contexts. Translational Issues it moral outrage?. European Journal of Social Psychology, 37(6), in Psychological Science, 2(1), 63. 1272-1285. [32] Pacilli, M. G., Roccato, M., Pagliaro, S., & Russo, S. (2016). From [24] Craig, K. M. (2002). Examining hate-motivated aggression: A review political opponents to enemies? The role of perceived moral distance of the social psychological literature on hate crimes as a distinct form in the animalistic dehumanization of the political outgroup. Group of aggression. Aggression and Violent Behavior, 7(1), 85-101. Processes & Intergroup Relations, 19(3), 360-373. [25] Navarrete, C. D., McDonald, M. M., Molina, L. E., & Sidanius, J. (2010). Prejudice at the nexus of race and gender: an outgroup male target hypothesis. Journal of personality and social psychology, 98(6), 933. [26] Richardson​-​Self, L. (2018). Woman​-​Hating: On Misogyny, Sexism, and Hate Speech. Hypatia, 33(2), 256-272. Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)