From Social Representations (SR), it is possible to know how individuals understand and position themselves in relation to phenomena (Camargo et al., 2018; Rouquette, 1998). And the way people see the world is a contribution of their belonging to certain social groups (Tajfel, 1981), since such belonging implies relatively common experiences and, consequently, shared opinions, beliefs and norms, which lead to a similar way of acting on the world. Thus, some ideas/representations only make sense when assessed from specific knowledge and values of certain social groups (Doise, 1992). SRs, as a form of socially constructed and shared knowledge, vary according to the context of social relations in which individuals are placed, giving meaning to the reality of groups and influencing their worldviews (Jodelet, 2001).

SRs can be a means for groups to assert their particularities, which demarcates their importance in the analysis of intergroup dynamics (Deschamps & Moliner, 2009). The sharing of SRs gives a status to different groups in society and supports the criteria that underlie hierarchies and social comparison (Moscovici, 2011). In the present study, we use the approach proposed by Jean-Claude Abric (2003), which considers SRs to be organized and structured sets of information, beliefs, opinions and attitudes, composed of two subsystems, central and peripheral, each of which has a specific and complementary role. More specifically, we intend to learn about the SRs of right-wing and left-wing groups, one of the other, by understanding that they interfere in intra- and inter-group relations in the context of political polarization in the country addressed in this study.

The number of people accessing the news via social media has increased on a global scale (Newman et al., 2019). Regarding political polarization on the social media, the SR theory can help in understanding selectivity in the search for information. Serge Moscovici (1981), when conceptualizing the changes caused by communication in societies in previous centuries, stated that while organization transforms spontaneous masses into organized masses, communication transforms the latter into publics, which, for him, would be true masses at home. Virtual communications are crafted through specific mechanisms of operation because, by being configured from anonymous and distance communication, they put in check the discursive and expressive capacity for another communication based on linguistic affirmation and modes of systematic mass operationalization (Camino et al., 2019).

The meanings of right and left have acquired different shades throughout history and have served as fuel for the most diverse political regimes in the world. According to Norberto Bobbio (2017), for more than two centuries, the concepts of right and left have existed, and being antagonistic, they are intended to diverge conceptions, ideologies, and political practices. The representation of the terms right and left is linked to the way of thinking by dyads in the sphere of knowledge.

These elements have undergone changes in meaning, and, in some cases, they are crossed by the phenomenon of prejudice, used by individuals to describe social groups that are different from their own (Rodrigues et al., 2015). It is noteworthy that, in a simplified manner, historically, the leftist group has been represented by fighting for the rights of working people and lower social classes, with the aim of promoting collective welfare and popular participation in social movements. Economically, it would seek to promote a more egalitarian society, with greater income distribution and interference by the State. On the other hand, being right-wing would represent a more conservative vision, tied to tradition, with the intention of keeping power centralized in oneself and in individual prosperity. In the economic sphere it defends market initiative with a free economy, without State intervention (Barros, 2019).

In recent years, a negative evaluation of left-wing parties has been perceived in Brazil, probably due to the corruption scandals in which the Workers’ Party (PT) was involved, in addition to Dilma Rousseff’s impeachment and the imprisonment of former president Lula, which led to the party’s tarred image and an increase in the derogatory tone of the left-wing parties (Mariano, 2019). In contrast, the right wing has also been unfavorably stereotyped since the military regime. This perception, however, has undergone changes since right-wing representatives were elected in the 2018 elections, with Jair Bolsonaro, who used an anti-party and conservative discourse in the field of customs in his electoral campaign. Moreover, the political growth of the “anti-PT sentiment” accompanied the rise of religious conservatism (Costa, 2019). Therefore, it is inferred that stereotypes (Rodrigues et al., 2015) regarding being “right-wing” or “left-wing” have become popularized in Brazil and have become stronger since the 2018 elections, potentiating political polarization in the country. This distinction of sides also results in conflicts of interests and values associated with social groups, as well as can reverberate in how social justice is conceived.

The belief in a just world (BJW) is a construct that is utilized to think about justice from a psychosocial perspective, and it can help in understanding the polarization process. Geared towards understanding a range of social phenomena, research has focused on investigating the phenomena and psychological processes that enable an understanding of how people judge issues related to justice. The basis of this theory is that individuals are motivated to perceive the world as fair, where people have what they deserve and deserve what they have (Lerner, 1998; Maes, 1998). There would be a tendency in individuals to resort to strategies in order to eliminate any kind of threat to BJW, such as injustice and the victim’s suffering. Thus, injustice would not be a congruent concept with this theory, since the misfortunes that occasionally happen to people are ascribed to their own responsibility, through the thought that bad things happen to bad people (Lerner, 1998).

This study intends to understand how the SRs of one group regarding the other (left-wing and right-wing individuals) may be influencing the relationships between those groups and the conflicts between them, as well as to understand how the political polarization phenomenon being experienced in Brazil may be affecting the daily relationships between groups of people, friends, and family members, considering the psychosocial aspects associated with the phenomenon, such as BJW.

An online survey was developed, consisting of both open- and closed-ended items, based on previous research (Gerber et al., 2017; Iyengar et al., 2019; Pimentel et al., 2010; Wisneski & Skitka, 2017). The questionnaire comprised 54 questions divided into 8 sections, addressing the following topics:

a) the Global Scale of Beliefs in a Just World, composed of seven items to evaluate BJW (Pimentel et al., 2010); b) a question on the perception of the country’s situation regarding political polarization; c) a close-ended question concerning political positions, and two open-ended questions on intergroup relations, in which the participants were asked, after stating their political position, to describe their own group (ingroup) using three words, and then to describe the opposing group (outgroup) using three more words; d) questions on the conflicts arising from the political theme and, in case of affirmative answers, with which groups of people, in what situations/locations they occurred, and how the situation developed with certain groups of people (12 closed questions); e) Questions on feelings associated with political polarization and violence (15 questions, of which 14 were closed- and one was open-ended); f) Questions on which means of communication were used to obtain information on politics (5 questions); g) Questions on trust in the means of communication and on the current Brazilian political system (4 questions); h) at the end of the questionnaire, six socio-demographic questions were included, such as those concerning gender, region of the Brazilian state where the participants lived, their age, income, socioeconomic level, religion, and schooling for further analysis.

The free-consent form was presented in the first section of the questionnaire. The study was approved by CEPSH – Ethics Committee for Research Involving Human Beings, according to Report no. 3.590.056. Participants were contacted through social media, WhatsApp and Snowball Sampling, in which each individual participating in the social medium creates connections from it with other members of his/her circle and disseminates the access to the questionnaire (Baltar & Brunet, 2012).

Descriptive and relational statistical analysis was performed on the closed-ended questions using the Statistical Package for the Social Sciences (SPSS-17.0) software. The answers to the word evocation questions were analyzed using the IRaMuTeQ software for textual data analysis (Ratinaud & Marchand, 2015). Prototypical analysis was performed, so that the material collected in the evocation test was analyzed considering the individual (frequency) and collective (mean evocation order) realms, thus identifying the centrality of the elements in the representations (Wachelke et al., 2016).

In the prototypical analysis, the constitution of the SR structure in the ingroup and outgroup descriptions appear through elements with a higher frequency, and those evoked first were more likely to belong to the central core, while those with a lower frequency and evoked later represented the representation periphery (Wachelke et al., 2016). In Tables 1 and 2, the elements that the left-wing and right-wing groups used to describe themselves (ingroup) can be observed. For this analysis, 2019 evocations were obtained, of which 658 were different words, and those with a higher frequency than 1 were considered in the analyses. In the right-wing group, we obtained an intermediate word frequency (f) of 7.31, and a mean evocation order (MEO) of 1.92. The left-wing group showed f = 6.25 and MEO = 1.89.

In the upper left quadrant of each of the diagrams, the elements that constituted the possible central core of the left-wing and right-wing groups’ SRs of themselves are listed. In the sample of individuals who claimed to be leftists, the words that were prominent to describe the ingroup and had a higher evocation frequency were: “equality” (62), “empathetic” (47) and “justice” (40). The latter element obtained higher activation as compared to the first two mentioned, with MEO = 1.5. Other elements such as “freedom”, “just”, “democracy”, “social justice” and “human rights” were also important for the group, showing an SR linked to the leftist ideological model. The word “conscious” was the one that obtained the highest activation (MEO = 1.2), despite its lower frequency of evocation (15).

In comparison, in the right-wing group, the words with higher evocation frequency had a higher activation rate than those in the left-wing group, despite having a lower frequency when compared to the same group: “freedom” (f = 49, MEO = 1.6); “conservative” (f = 34, MEO = 1.5); and “justice” (f = 25, MEO = 1.4). It is noteworthy, that, in both groups, the element “justice” obtained a prominent place, as it remained among the words with the highest frequency and highest activation rate to describe themselves. Other core elements for the right wing were: “intelligent”, “liberal”, “liberalism” and “realist”. The elements “freedom”, “liberal” and “liberalism”, occupying this first quadrant, also showed the SR linked to the right-wing group’s conception of an economic model.

Still regarding the contrast-zone quadrant, in the right-wing group, the elements that were more frequent and that also similarly converged to a more stabilized behavior representation were: “dialogue” (6), “ethical” (6) and “moderate” (5). Despite its low frequency, the latter word was the most readily evoked in relation to the former, with MEO = 1.4. Other elements such as “free market”, “order” and “libertarian” showed prompt evocation (1.4) and expressed a representation of a political model based on patriotism/conservatism as well as the neoliberal market model.

In the upper right quadrant, the left-wing group described themselves with: “equity” (24), “right” (19), “humanist” (19), and “struggle” (16), which again expressed political ideals permeated by social and economic movements in the socialist realm. The right-wing group, on the other hand, used the words “honest” (f = 32), “just” (f = 21) and “workers” (f = 17) with high frequencies of evocations, and reiterated upright and correct behavior. Still in the same group, the elements “family”, “patriot” and “progress” showed representations permeated by conservative ideals in which they favored the same group, both in family and nation.

In the second periphery, or lower right quadrant, the most highlighted element used by the left-wing group to describe themselves was “diversity” (f = 7 and MEO = 2), which pointed to the primacy of freedom and the plurality of ways of existing. Other elements that converged with this idea also appeared: “resistance”, “inclusion” and “sensitive”. Likewise, the elements “health” and “justice” were also highlighted. In the right-wing group, elements involving concern with the future appeared, such as: “right”, “values”, “future”, “studious”, and “growth”. Finally, other elements such as “Christians” and “God” signaled importance to the religious realm.

Tables 3 and 4 shows the elements listed for the outgroups of the left-wing and right-wing groups. For this analysis, 1985 evocations were obtained and, of these, 760 were different words, and those with a frequency higher than 1 were analyzed. In the left-wing group, we obtained f = 6.57, and MEO = 1.89. The right-wing group showed f = .97 and MEO = 1.85.

In the upper left quadrant, the left-wing groups’ possibly central elements concerning right-wingers were “selfish”, “ignorant”, “individualistic”, “intolerant”, “conservative”, “extremist”, “conservatism”, “fascist” and “unfair”. Despite their somewhat lower evocation frequency, the elements “extremist”

(24) and “fascist” (17) had both higher activation (MEO = 1.5) in comparison to the other words. They refer to ideological and political stereotypes. The right-wing group showed elements that also referred to stereotypes. The most evoked words to describe the outgroup were: “corruption” (32) and “intolerant” (22). The element “extremist” had high frequency and activation (f = 21, MEO = 1.3). Other elements like “utopian”, “communist”, “authoritarian”, “deluded” also gained importance. The words “socialism” and “socialist” came next. In the lower left quadrant, or contrast zone, it was observed that the leftist group listed the item “uninformation” (f = 6, MEO = 1.3) as the best description for the other group, achieving a good frequency and better activation compared to the following others: “arrogant”, “backwardness”, “rigid”, “aggressive”. The elements “capitalist” (f = 5, MEO = 1) and “racist” (f = 5, MEO = 1.4) appeared next as the most readily evoked items and indicated importance for describing the outgroup, as the former revealed a stereotype of an economic model, and the latter one of prejudice. Regarding the right-wing group, the words that were highlighted were “aggressive” and “unfair”. The words “welfarism”, “innocent”, “dependent”, “communism”, “revolutionary”, and “idolater” dialogued in stereotyped ideologies of the left-wing group.

In the upper right quadrant, the left-wing group listed the item “meritocracy” (f = 30, MEO = 2) to describe the outgroup, which shows a group’s representative conception that has the “belief in a just world” as its foundation. The words “prejudiced” and “prejudice” appeared frequently, and by being synonymous, they also showed importance for the group. Other elements, such as “inequality”, “alienated”, “capitalism”, “money” and “elitist” were highlighted as well and reaffirmed the opposing group’s conception of representation ideas that were primarily economic. Regarding the right-wing group, elements such as “radical” (19), “hypocritical” (18) and “alienated” (17), “ignorant” (16), “selfish” (14), “naïve” (11), “dumb” (11), “dishonest” (11), and “fanatic” (10) frequently gained strength in elements, showing the outgroup’s negative conceptions.

Finally, in the second periphery, or lower right quadrant, the left-wing group listed the element “fanatic” as the best description for the outgroup (f = 6, MEO = 2), and then: “utopian”, “religion”, “uninformed” and “indifferent”, which indicates a representative idea of alienation about the world. There were also elements linked to prejudice: “homophobic” and again “racism”. Finally, still regarding affections, the right-wing group also elected elements that were close to alienation to describe the outgroup, such as: “uninformed”, “envious”, and “cattle”. This group also made use of elements such as “hunger”, “inequality”, and “indoctrination” to designate a representation of the left-wing group’s lack of economic support and alienation from the social realm.

As for the belief in the Brazilian political system, the statement: “In general, the Brazilian political system works as it should” had a very low general mean (M = 1.72, SD = 1.06) on the six-point scale, with right-wing groups showing a higher mean (M = 1.92, SD = 1.23) than left-wing groups (M = 1.52, SD = .8). The difference between the means was 0.40 [t(629.45)= 5.27; p<0.001]. And regarding the assertion: “The functioning of most public policies serves the population’s general interest”, the overall mean was also quite low (M=2.0; SD = 1.06), with no difference between the groups, according to Table 5.

From the question: “Would you change your political position if a government with a different position from yours brought improvements to the country?”, we obtained an overall mean of 3.60 with SD = 1.81, with the left-wing groups showing a lower mean (M = 3.46; SD = 1.67) than the right-wing groups (M = 3.75; SD = 1.93). The difference between the means was 0.28 [t(719.35) 2.12; p < 0.05].

Regarding the participants’ trust in the Brazilian communication media, the sample’s overall mean was slightly below the scale’s midpoint (3.0) (M = 2.53; SD = 1.34). Left-wing groups had slightly higher means (M = 2.82, SD = 1.28) than did the right-wing groups (M = 2.24, SD = 1.35), with a statistically significant difference [t(734) = 5.96; p < 0.001].

And regarding the means for the BJW scale, there were differences between the groups’ means, with the left-wing group showing a lower mean (M = 2.15; SD = .8) than the right-wing group (M = 3.37; SD = 1.0). The difference between the means was 1.22 [t(704.22) = 17.81; p < 0.001].

In terms of having suffered any forms of political violence, 48% of the participants reported having suffered verbal violence in person, and 5% reported having suffered some form of physical violence. It is noteworthy that 77% of the participants reported having suffered violence through social media, such as Facebook, Twitter, YouTube, Instagram, and 68% exclusively through WhatsApp. Moreover, 72% reported knowing someone who had suffered political violence in the last two years, most frequently through the social media.

On a six-point scale, where one corresponds to not once and six corresponds to many times, when comparing the means of the groups with different political orientations about having suffered violence for political issues, it was observed that the means were similar for both groups through Facebook, Twitter, YouTube, and Instagram (M = 3.05; SD = 1.9). The means for WhatsApp, on the other hand, were 1.97 SD = 1.66, for the right-wing and 2.37 (SD = 1.55) for the left-wing groups, showing that the right-wing groups reported having suffered less violence through WhatsApp [t(730) = 3.34; p < 0.01], according to Table 6. On having experienced verbal violence in person, there was a marginally significant difference between right-wingers (M = 1.96; SD = 1.42) and left-wingers (M = 2.16; SD = 1.44) [t(734) = 1.96; p < 0.001]).

As for having caused violence in opposing groups, 42% reported having caused it on social media (Twitter/YouTube/Instagram); 30% via WhatsApp; 24% reported having caused verbal violence in person, and 2% physical violence. Regarding the differences between the groups, on a six-point scale, the left-wing group showed a slightly higher mean (M = 1.71, SD = 1.21) for having caused it through WhatsApp than the right-wing group (M = 1.50, SD = 1.13) [t(730) = 2.45; p < 0.05]. The scenario reverses regarding having caused physical violence, where the right-wing participants’ mean (M = 1.09, SD = 0.54) was slightly higher than that of left-wingers (M = 1.01, SD = 0.09) [t (3870 = 2.95; p < 0.005].

Concerning conflicts arising from political disagreements, measured on a six-point scale (where 1 = never and 6 = always), family and friends had the highest value (M = 3.13; SD = 1.42); acquaintances (M = 2.99; SD = 1.44), and strangers (M = 2.58; SD = 1.53). Regarding relatives/friends [t (730) = 6.27; p < 0.001] and acquaintances [t (731) = 3.17; p<0.005], the mean for the frequency of conflict occurrence was significantly higher in the left-wing group.

As for having experienced conflicts and disagreements with friends and family about politics, the feelings accessed by the participant at the time of the experience were questioned. The most accessed feeling was Disappointment sometimes felt by 27.9%, and often or always felt by 50% of the participants. Sadness was sometimes felt by 27.7%, often or always by 37.6%, while Anger was felt by 31.8% in some political discussions, often or always by 34.2% of the participants. Comparing the frequency of occurrence, from the political belonging, it was observed that left-wing individuals reported feeling more anger [t(608.9) = 6.37; p < 001]; more sadness [t(632) = 6.34; p < 001] and more disappointment [t(606.78) = 4.86; p < 001].