Inglês (pdf)
Artigo em XML
Referências do artigo
Tradução automática
Permalink
Humour encompasses the cognitive and emotional processes of creating, perceiving and enjoying a funny stimulus. In psychosocial terms, humour can be divided into four main components: (1) social context; (2) cognitive process; (3) emotional response; and (4) vocal-behavioural expression of laughter (Martin, 2010). The social context and cognitive process dimensions assume that humour arises from social comparison and can serve as a means of affirming feelings of superiority over members of other groups. The emotional (pleasure) and behavioural (laughter) responses thus promote the belief that the individual is worth more than an outgroup member who is judged on the basis of negative stereotypes, thereby promoting and legitimising discriminatory and oppressive practices (Moreira, 2019).
Humour not only triggers a sense of pleasure, but also satisfies other emotional needs, particularly the need for positive distinctiveness from others, creating a sense of solidarity among ingroup members (Venkatesan et al., 2023). Humour is associated with racism to convey a sense of white supremacy (Anderson, 2015). Note that there are studies highlighting the positive side of racist humour and showing that its strategic use can enable the transformation of feelings of subordination and humiliation into forms of resistance, interracial relief and catharsis (Hylton, 2018). Nevertheless, our interest lies in the negative or pejorative dimension of racist humour.
By analysing humour, we can improve our understanding of the ways in which racism continues to be reproduced in society. Data from racist incidents show that racially motivated fun and humour play an important role in new expressions of racism in societies where more “serious” or overt forms are largely rejected (Pérez, 2017). Although racist jokes may seem harmless to some people, they are microaggressions that serve as discursive tools to dehumanize social minorities, configuring “amused racial contempt” (Pérez, 2022). Racist humour has long been used to promote social cohesion among whites. According to Lippard (2023), racist humour “represents nefarious wordplay that thinly veils their real intention with nervous laughter” (p. e2). The literature indicates that derogatory humour has social consequences, as it can create or reinforce negative stereotypes or prejudiced attitudes (Horisk, 2024, pp. 1-14; Hylton, 2018; Pérez, 2022).
The “racist laugh” has played a mediating role in a reality where strict laws against overt racism are in force and negative feelings and beliefs towards non-whites still exist (Kawakami et al., 2009). Faced with the pressure of anti-racist norms, humour serves as safe territory for racist expression (Pérez, 2017). There are different types of manifestations of racist humour, such as jokes, wordplay, ambiguity, and personification (Anderson, 2015). In this article, we analyse the manifestation of racist humour through jokes. There are also different approaches to distinguish when a joke is racist and when it is racial, one of which is the action - or harm-centred approach. From this perspective, regardless of the beliefs or attitudes of the person telling it, a joke is racist if it conveys and spreads racist opinions with the aim of: (i) actually harming, (ii) hoping to harm or even (iii) causing undue harm.
The actor-centred approach assumes that a racial joke is racist if the speaker has a cruel, malicious or disrespectful attitude towards members of the racial group targeted by the joke (Pérez, 2022). It is also necessary to consider the audience-centred approach to racist jokes, which assumes that in order to find a joke funny, one must approve of the negative stereotypes it expresses. Sharing laughter at the expense of an outgroup promotes a stronger social affiliation within the ingroup and simultaneously creates and/or increases social distance from the target of ridicule and offence (Billig, 2005). Finally, an integrative approach asserts that racial humour is racist when it unduly hurts the target because of his/her racial affiliation or when the speaker is motivated by a malicious or disrespectful attitude (Anderson, 2015). Thus, anyone who does not find racist humour funny should act by criticising instead of accepting its expressions.
Pérez (2017) conducted a study on the prevalence, popularity, re-emergence and persistence of racist jokes in different time periods and social settings in the United States. By analysing joke books that emerged in the 1980s, he found that although such books were considered “offensive to all” and marketed to “the not so easily offended”, they used their circulation and popularity to support the conservative right’s assault on Black achievement through the civil rights movement. Examining the reappearance of racist jokes in books on the internet, particularly on white supremacist websites such as ‘Tightrope.com’, ‘Niggermania.com’ and ‘chimpout.com’, Pérez showed that the internet plays an important role in the dissemination and perpetuation of racist humour and ideology. The author concluded that: (1) racist humour functions as a cultural tool in the racialisation, dehumanisation and criminalisation of Black people; (2) shared racist humour strengthens social bonds, cooperation and group identity among white people; and (3) forms of public expression of racist humour adapt to social, political and cultural changes, allowing racist ideologies and emotions to move from “behind the scenes” to public arenas.
A historical analysis of humorous productions in Brazil shows that they have generally reproduced derogatory ideas about racial minorities, conveying stereotypes that justify judgements about the places or social positions that members of racialised groups should occupy (Moreira, 2019). In contrast to bi-racial societies, racist expressions in Brazil have long been subtle or veiled (Lima, 2019). The Brazilian racism was influenced by a number of factors, including the myth of racial democracy, the informal and friendly style of social interactions and the ideology of whiteness. These elements contributed to the emergence of a specific form of racism in the country known as “cordial racism” (Turra & Venturi, 1995.
“Cordial racism” is characterised by discrimination against Blacks manifested in superficial politeness that conceals hostile attitudes and behaviours, often expressed in racial jokes in interpersonal relationships (Turra & Venturi, 1995). The laughter, ridicule and ambivalent approach to expressing attitudes and asserting identities allow “cordial racists” to see themselves as “racial democrats” and morally righteous people (Dahia, 2010). This produces a racism “without intention” or “without racists” (Bonilla-Silva, 2021) in a society historically characterised by social inequality, paternalistic relations, clientelism (Schwarcz, 1998), and formally “prejudiced against prejudice” (Fernandes, 1966, p. 33).
Despite its strategic nature for understanding the manifestations of white supremacy and analysing the links between the most overt and subtle expressions of racism, relatively little social psychology research has been conducted on racist humour. This article is the first on humour and racism to use research scenarios in the country with the largest Black population outside Africa, which has long characterised its racism as relaxed or cordial (Turra & Venturi, 1995). Although most publications on racist humour come from sociology, even in this field the topic remains ignored as a mean of racism analysis (Lu, 2023; Pérez, 2017; Pérez & Kuipers, 2024).
Moreover, to our knowledge, no study has systematically analysed the impact of laughter on racism in online contexts of social interaction between Whites and Blacks. The Kawakami et al. (2009) study is an important exception. The authors analysed the discrepancies between people’s predictions of their feelings and behaviour and their actual feelings and behaviour when witnessing a racist situation. In two studies, they observed that people’s predictions about their emotional distress and behaviour in response to an episode of explicit racism differed from their actual reactions. Participants who imagined the situation expected to be very upset, while those who experienced the event did not differ from control participants who were not exposed to a racist situation. The authors concluded that “despite current egalitarian cultural norms and apparent good intentions, one reason why racism and discrimination remain so prevalent in society may be that people do not respond to overt acts of racism in the way that they anticipate” (Kawakami et al., 2009, p. 278).
The aim of this research is to analyse the relations between racism and humour. The effects of the behaviour of a White or Black confederate who laughs or does not laugh in a social context of the broadcast of a racist joke on the emotional and behavioural responses (supportive or indifferent reaction) of White and non-white participants are taken into consideration. An important set of questions that can be asked is: How do witnesses of racist jokes react when confronted with them in real-life situations? How would they react in hypothetical situations involving racist humour? What role does the laughter of a White or Black person play in witnesses’ reactions to racist humour? Finally, what is the relationship between participants’ racial prejudices and their reactions to racist humour? This article attempts to answer these questions.
Overview of the studies and hypotheses
Study 1 examined the effects of the skin colour of the confederate delivering a racist joke and the context of the presentation (smiling or not smiling) on participants’ predictions of their behavioural and emotional responses. In Study 2, the two confederates (a White man and a Black man) depicted in the video from Study 1 interacted with the participant in a real-life encounter. While in Study 1 we try to understand participants’ judgement of their own behaviour (how they imagine they would behave) if they were in the scene, in Study 2 we investigate what actually happens when participants are present in the racist scene and can actively intervene in it.
In this context, the following hypotheses were formulated: H1 - participants who see a video in which a racist joke is told in the presence of a Black person indicate they would feel very uncomfortable and support the Black person hearing the joke; H2 - the racist joke delivered by a smiling fellow participant elicits less anti-racist reactions than when delivered in the non-smiling condition; H3 - the racist joke, when delivered by the Black confederate, arouses fewer anti-racist responses than when delivered by the White one; H4 - the greater the participants’ implicit racial prejudice, the less they will be bothered and the less will react negatively to the racist humour situation; and H5 - participants who see a real scene in which a racist joke is delivered in the presence of a Black person will be little bothered by the situation and will not support the Black person hearing the joke. Hypotheses 1 to 4 were tested in Study 1. Hypothesis 5 was tested in Study 2.
Study 1
This study aimed to analyse the impact of exposure to racist humour and smiling on participants’ expressions of racism (the “forecaster” context). Following Kawakami et al. (2009), we hypothesized that people who hear a racist joke presented in a video would have strong reactions against the racist humour and a greater preference to interact with a Black confederate than with a white confederate.
Method
Participants
The study consisted of a 2 x 2 factorial design: skin colour of the accomplice who delivers a racist joke (White or Black) and racist humour context (the accomplice smiles or does not smile when delivering the joke). Initially, there were 138 participants, but 13 were excluded for not completing all phases of the experiment. Among the 125 remaining participants, 66 (44.3%) defined themselves as Brown and 59 (39.6%) as White. We did not include any Black participants because the experimental situation would be emotionally unpleasant to them. Participants were randomly allocated to each of the four experimental conditions. The majority were female (82 or 65.6%) and aged between 16 and 62 years (M = 23.41; SD= 7.27). Regarding education, 76% were undergraduate students.
Material
(1) The participants’ prediction about their behaviour in the presented scenario - each participant watched one of the four experimental context videos, then had to answer the following questions: “Imagine that you are Participant X in the presented scene. How would you react in this situation? How do you imagine you would feel? Which of the two participants would you invite to help you with a later problem-solving task?” All questions were open, with the exception of the last one, which contained the fictitious names of the accomplices. The answers to the open questions were categorised according to the postulates of content analysis
(2) Implicit racial prejudice - assessed by the racial version of the Implicit Association Test (IAT). We used the Portuguese translated version by Santos et al. (2024). A standardised difference score (D-score) data-cleaning and scoring algorithm (Carpenter et al., 2019) was calculated for each participant, indicating in which condition (compatible vs. incompatible) they were faster (Greenwald et al., 2003). The rate of rejected trials in the present study was low (< 0.01%). No participants were excluded for responding too quickly. Errors were replaced by the average latency for correct trials in that block, supplemented by a penalty time of 600 milliseconds (i.e., D 600 procedure) (Greenwald et al., 2003). The error rate was 7%. The effect size was evaluated using Cohen’s d method with d = 0.52. Positive values indicated a faster association when recognising negative features of the Black target. The results indicated an IAT effect: scores tended to associate Black with negative, MD score = 0.291 SD = 0.45, which was significantly different from 0.14, t(123) = 3.76, p < 0.001. This result indicated the presence of mild implicit racial prejudice in the sample studied (Maina et al., 2018).
Procedure
Data Collection. Data collection was done remotely, using the Qualtrics platform. The independent variables were manipulated in experimental scenarios expressed in videos recorded by previously trained actors. The actors played the roles of a Black and a White confederate or accomplice. The researcher, a Black woman, acted as herself. Participants were invited via social media to collaborate in the study. In the data collection session, in a remote room, the researcher explained the objective of the study. Then, the video was presented to the participant, who had to watch the video and imagine being the person who heard the racist joke in the scene presented. The videos, the skin colour of the accomplices and the researcher were pre-tested by judges and validated, as presented in the supplementary material.
The experimental conditions were: a) the Black confederate telling a racist joke without smiling; b) the Black confederate telling a racist joke while smiling; c) the White confederate telling a racist joke without smiling; and d) the white confederate telling a racist joke while smiling. Participants were randomly distributed across the four experimental conditions, in a between-participants design.
Data Analysis. Data were analysed with use of the SPPS (Statistical Package for the Social Sciences), descriptive and inferential statistics. The analyses were carried out using chi-square tests and bivariate correlation tests.
Ethical Considerations
All data and materials of the studies reported here are publicly available from the Open Science Framework. For all the studies we report all the conditions and measures. The studies were conducted with Brazilian individuals, and were registered and approved by the Ethics Committee (Protocol number 3.741.249).
Results
Regarding the first hypothesis; participants who see a scene of a racist joke being told in the presence of a Black person will feel very uncomfortable and support the Black person hearing the joke, the results confirmed the hypothesis: 64 participants (53.3%) said they would reprimand the person who told the joke, another eight (6.4%) said they would be concerned about the Black person in the scene, 34 (28.3%) said they would not react, 11 ( 8.8%) said they would be curious or interested to hear the joke, and only three participants (2.4%) found it funny, X 2(4) = 106.92; p < 0.001 (Figure 1). In the same vein, when asked how they would feel in this situation, 82.9% said they would have unpleasant feelings, 7.3% said they would have pleasant feelings, 6.5% would be neutral and 3.3% of participants did not know how they would feel, X 2(3) = 220.58; p < 0.001 (Figure 2). Regarding the choice of one of the actors for a subsequent interaction, the Black actor was chosen more often than the White actor [65.6% vs. 34.4%, X 2(1) = 12.17; p < 0.001].
Figure 1 Percentage of responses to the question “What would you do if you were in the situation described?”
Figure 2 Percentage of responses to the question “How do you think you would feel if you were in the situation described” (n = 125)
The second hypothesis, which stated that a racist joke told by a smiling actor would elicit fewer anti-racist reactions than a joke told in the non-smiling situation, was not confirmed. There was no significant effect of the actor’s reaction (smiling x non-smiling) on the participant’s reactions, X 2(4) = 1.60; p= 0,81. Similarly, there was no effect on emotions (what would you feel?), X 2(3) = 3.39; p = 0,34, or on the Black participant’s choice of subsequent activity, X 2(1) = 0.15; p =0,70.
The third hypothesis stated that a racist joke delivered by a Black accomplice would elicit fewer anti-racist reactions than when delivered by a White accomplice. The results showed that the actor’s skin colour had no effect on participants’ reactions, in terms of what they would do, X 2(4) = 1.38; p = 0,85, or in terms of what they would feel, X 2(3) = 6.16; p = 0.10, or also in terms of who they would interact with, X 2(1) = 0.06; p= 0,94. However, in a one-tailed test, the difference found in the question about how they would feel was significant. An analysis of the adjusted residuals indicated that a difference between the colour of the actor conveying the joke occurred only in the “I wouldn’t know what to do” response condition. In this scenario, all participants did not know how to react to the joke conveyed by the Black actor (adjusted residual > 2.0). This result partially confirmed our third hypothesis.
The fourth hypothesis predicted that the greater the participant’s implicit racial prejudice, the less disturbed he/she would be and the less negatively he/she would react to the racist humour situation. For the present analysis, we converted the categories of the two questions into an interval scale. Thus, responses to the question of what they would do were coded as 1 (they found it funny), 2 (they showed interest/curiosity), 3 (they did not react), 4 (they felt emotionally uncomfortable), 5 (they showed concern for the Black person) and 6 (they blamed the person who told the joke). The answers to the question of how they would feel if they were in the situation were coded as follows: 1 (I would have pleasant feelings), 2 (I would have neutral feelings) and 3 (I would have unpleasant feelings). The answer “I do not know” was excluded from these analyses. First, we measured the participant’s IAT level. The test indicates that participants who associate White faces with positive words more quickly and with fewer errors than Black faces have an implicit pro-white bias. The IAT from 0 to 0.14 indicates the absence of racial prejudice; 0.15 to 0.34, slight pro-white bias; 0.35 to 0.64, moderate pro-white bias, and the D-IAT score > 0.65, indicates strong pro-white bias. Negative scores of the same degree indicate similar categories of pro-Black prejudice (Maina et al., 2018). An analysis of bivariate correlations revealed that the relationship between the D-IAT score and what participants would do was negative in the expected direction (r = -0.146; p = 0.113). The greater the implicit prejudice, the more the participants felt pleasant emotions in the situation of racist humour, now in a slightly significant way (r = -0.172; p= 0.062). In other words, hypothesis four was only confirmed in a one-tailed test, showing that implicit racist prejudice was weakly related to reactions to racist humour.
Discussion
The aim of this study was to analyse the effects of the skin colour of the accomplice presenting or relaying a racist joke and the social context (smiling or not smiling) on participants’ predictions of their behavioural and emotional reactions. The relationship between these reactions and implicit racial prejudice was also analysed. The results found confirmed the hypotheses about the emotional reactions of displeasure and behavioural support towards Black people reported by participants. Kawakami et al. (2009) found similar results in their “forecaster” condition. The hypotheses about the specific effects of the racist humour context (smiling/not smiling) and the skin colour of the actor delivering the joke were not confirmed. The psychology of humour states that the purpose of laughter is not only to convey a state of joy, but also to evoke this state in others. However, laughter also fits into the normative social contexts of behavioural reinforcement (“laughing with”) or social sanction (“laughing at”) (Martin, 2010). It can therefore be assumed that the accomplices’ laughter did not influence the participants’ behaviour in the direction expected by the strength of the normative context depicted in the scenario, which simultaneously implied “laughing with” and “laughing at”. Overall, these results allow us to confirm that the participants’ anti-racist reaction proved to be unconditional and not influenced by elements of humour, that is, who tells a joke and how it is told.
The level of implicit racial prejudice had a weaker influence on predicted responses to racist humour than hypothesised. However, note that the sample studied had low levels of implicit racial prejudice. In samples with higher D-IAT scores, the relationship between prejudice and approval of racist humour could be more pronounced. Others studies claim that the most subtle or less controlled expressions of racial prejudice in Brazil are mediated by humour (Moreira, 2019; Turra & Venturi, 1995). The results showed we are facing a new and promising scenario for the analysis of the relationship between emotional and behavioural responses in the most subtle expressions of racism in real-life social contexts. However, it is worth asking whether participants were actually unconditional anti-racists or simply followed the politically correct normative logic of “do as I say”. In Study 2, we analysed whether participants’ actions aligned with the second part of the popular saying (“not as I do”), specifically in a real-life interaction with Black and White people where racist humour was used.
Study 2
This study aimed to analyse what actually happens when the participants were present in a racist humour situation and could immediately intervene in it (the “experiencer” context). In accordance with Kawakami et al. (2009), negative feelings towards Black people often occur in the context of more spontaneous reactions, as is the case in real-life scenarios. We hypothesised that participants would feel little emotional discomfort, and would neither express an anti-racist reaction, nor offer support to the Black person victimized by the joke (H5).
Method
Participants
The study involved 35 participants, all of whom self-identified as white. We decided to study only White participants to avoid exposing Black participants to racist humour. Ages ranged from 19 to 55 years (M = 25.16; SD = 8.55), and the majority were female (67.6%). In terms of education, 51.4% were students and 34.3% had a university degree. Participants were recruited via social networks in the same way as in Study 1. In this study, the experimental design considered the skin colour of the accomplice who told the racist joke (White or Black) as an independent variable. We did not use the smiling context in this study because our aim was to assess participants’ reactions to racist humour. The experimental conditions were: (a) a Black confederate delivering a racist joke; and (b) a White confederate delivering a racist joke. Participants were randomly assigned to one of the two conditions.
Instrument
The instruments used in this study were similar to those used in Study 1 and were hosted and administered via the Qualtrics platform. We used the same racist joke from Study 1 as stimulus material. The experimental scenario is described below.
Procedures
The procedures were similar to those of the previous study. In this study, during the data collection session in a remote Google Meet room, the Black and white confederates entered the room (same actors as in Study 1) seconds after the participant. After this brief rapport interval, the interviewer asked accomplices and participants to wait a few moments while she searched for a file on her computer. At that moment (with the camera open to everyone) one of the accomplices reported that he had received an audio file via a messaging application and presented it to the others. The audio had the same racist joke of Study 1. After that, the interviewer informs that the experiment is starting. The behavioural reactions of participants after the presentation of the racist joke were observed and recorded independently by the researcher and accomplices, and were classified using the same categories as in Study 1.
Data collection. The Implicit Positive and Negative Affect Test (IPANAT) was used in this study to assess the emotional reactions of participants immediately after the presentation of the racist joke. The IPANAT assumes that affects function according to the principle of projection, meaning they influence the judgement of objects unrelated to the affective experience in question (Hernández et al., 2020). Participants were instructed to rank the degree of correspondence between six nonsense words (SAFME, VIKES, TUNBA, TALEP, BELNI and SUKOV) and the following affects: happy, enthusiastic, active, helpless, tense and inhibited. The resulting 36 items (six nonsense words associated with six mood states) were ranked on a 4-point scale (0 = does not fit to 3 = fits very well). The estimates of internal consistency yielded good values for the positive affect (happy, enthusiastic and active) and Cronbach’s alpha was 0.85, while for the negative affect (helpless, tense and inhibited), alpha was 0.84.
As in the previous study, the IAT data were processed using the D-score algorithm for data cleaning and analysis (Carpenter et al., 2019). The rate of rejected trials (trials > 10 s) was low (< 0.002%). No participants were excluded for responding too quickly. The error rate was 6%. Errors were then replaced by the average latency (D-600) for correct trials in this block.
Data Analysis. Data were analysed with use of the SPPS (Statistical Package for the Social Sciences), descriptive and inferential statistics. Statistical analyses were performed using chi-square tests and bivariate correlations.
Results
Figure 3 Percentage of behavioural reactions observed in the participants in the real scene (n = 35)
This study tested the hypothesis that White participants who see a real-life racist humour scene would feel little experience emotional discomfort with the situation, and would neither react in an anti-racist manner, nor offer support to the Black person hearing the joke (H5). To test this hypothesis, participants’ behavioural and emotional reactions were recorded. In this study, only three types of reactions occurred: “reprimanding the participant who told the joke”, “no reaction” and “finding the joke funny (smiling)”. The analyses showed that the most common reaction was “no reaction”, which pertained to 33 of the 35 participants (94.3%). One participant reprimanded the accomplice who had spread the joke (via the chat room of the platform) and another found it funny and smiled (Figure 3). Regarding the choice of one of the actors for a subsequent interaction, as in the first study, the Black actor was chosen more often than the White actor [67.6% vs. 32.4%, X 2(1) = 4.23; p = 0.040].
The affective reactions measured with the IPANAT showed that the values for positive emotions were 1.78 (SD= 0.48), and for negative emotions they were 1.68 (SD= 0.46). Thus, participants tended to have a stronger activation of positive affect than negative affect, t(34) = 1.85; p = 0.074, a result that persisted even when simulated via bootstrap for a thousand samples. Participants’ emotional response was not influenced by the skin colour of the accomplice who told the racist joke, or for the expression of positive or negative affect, F(1, 34) < 1; n.s.
Implicit racial prejudice was similar to that found in Study 1 (MD score = 0.250, SD = 0.56), but not different from 0.14 (indicator of racial prejudice absence), t(34) = 1.17, p = 0.24. The IAT score did not correlate significantly with the choice of a White or Black partner for a subsequent activity, r(34) = - 0.21, p = 0.21, nor with the projection of positive affect, r(34) = - 0.22, p = 0.20, and negative affect, r(34) = - 0.17, p = 0.34, as analysed by the IPANAT. This means the participants’ behavioural and emotional reactions to racist humour were not influenced by their level of implicit prejudice. In this study, it was not possible to analyse the relationship between the D-IAT score and behavioural response because it hardly varied between participants (94.3% did not react to the racist scene).
Discussion
This study investigated the emotional and behavioural responses of participants confronted with a real-life situation of recreational racism. The results confirmed the findings by Kawakami et al. (2009). In addition, the main hypothesis was also confirmed; that White participants would experience little emotional discomfort in the situation of racist humour and would not react to it, as predicted in Study 1. The vast majority of them (94.3%) showed no behavioural reaction, and if an emotional reaction occurred, it was positive (pleasant) rather than negative (unpleasant). As in Study 1, the level of implicit racial prejudice in the sample was low, which may have affected the relationship with the response to racist humour. However, a very important issue is that even non-prejudiced participants at an implicit level did not react in an anti-racist way in situations of racial offence. Kongnetiman (2022, p. 17) suggests that anti-racism is structured in five levels: Compliance (i.e., “Deny systemic racism is a problem”), Complacency (i.e., “Recognize systemic racism is present and a current problem”), Awareness (i.e., “Actively seek ways to educate employees about racial equity”), Integration (i.e., “Actively promote and advocate anti-racist actions”) and, finally, Anti-Racist (i.e., “Actively participate in combating systemic racism”). It can be assumed that independently of their prejudice level, our sample is predominantly at the phases of compliance and complacency with respect to the racist norm.
General discussion
In the two studies, we examined the effects of the confederate’s skin colour and racial humour on participants’ behaviour and emotions, including their implicit racial biases. In Study 1, we sought to understand how people would imagine, behave and feel when confronted with a racist humour scene. In Study 2, we investigated how people actually behaved and felt in such a context. Consistent with Kawakami et al. (2009), our results suggest that although people predicted they would be very upset by a racist act depicted in a joke, when they actually experienced this event, they showed little or no emotional discomfort. Furthermore, people overestimated the extent to which a racist joke would elicit hostile behaviour from their interlocutor.
In another words, although most participants in Study 1 predicted they would reprimand the actor (53.3%) or worry about the Black person who heard the joke (6.4%), in the actual situation depicted in Study 2, more than 94% remained unresponsive. A statistical comparison of these behavioural responses from Studies 1 and 2 shows the enormous gap between what participants said they would do and what they actually did when witnessing a racist scene, X 2(2) = 37.03; p < 0.001; η2= 0.49, with a large effect size (Metsämuuronen, 2024).
These results confirm the postulates of the theory of aversive racism (Dovidio & Gaertner, 1998) regarding the ambivalence between racist beliefs and emotions with the pressure of the self-image of an egalitarian, unprejudiced person. When participants were asked to predict how they would feel emotionally in the racist humour scene, 81.6% in Study 1 indicated they would feel unpleasant emotions. However, an assessment of more automatic or uncontrolled emotional activation measured with the IPANAT in Study 2 indicated that emotional activation in the racist scene would be more positive than negative.
These findings suggest a possible link between affective and behavioural responses to racism and contribute to theorising about the role of emotions in studies of prejudice and discrimination. Indeed, contemporary race relations embody an apparent paradox: blatant racism is strongly condemned but persists and has intensified in Brazil in recent years (Lima, 2019). Kawakami et al. (2009) stated that one of the reasons for this apparent inconsistency is that people misunderstand how they would feel and behave after witnessing racism. Another reason may be that humour acts as a shelter for racism, especially in a society characterised by informality in social and racial relations (Schwarcz, 1998), by the prejudice of having prejudice (Fernandes, 1966) and by the belief that racial prejudice is always someone else’s problem.
Participants’ levels of implicit racism and the socio-demographic similarities between the samples of both studies contributed to the understanding of this gap between what is said and what is done. Even though the sample in Study 2 included only white participants and Study 1 included more Brown participants, researches on humour and racism with a representative sample of the country showed that the cordial or “friendly” racism of whites and Browns did not differ (Turra & Venturi, 1995). Kawakami et al. (2009) found the same pattern of results, stating the following: “In addition, we posit that participants may have mispredicted their emotional responses to witnessing a racist comment because of their own ambivalent racial attitudes” (p. 277).
In the two studies, we examined the relationship between implicit racial prejudice and participants’ behavioural and emotional reactions. In Study 1, we found that although the correlation was in the expected direction, it was small and only significant in a one-tailed test. In Study 2, there was no correlation. These results could indicate that humour remains a space that is less accessible to the anti-racist norm. It is accepted even by people with low levels of racial prejudice because it allows them to express their real feelings while remaining protected.
As a limitation of the present study, we point out that an important restriction on the generalisation of our conclusions is that the samples were mainly university students. The online data collection and experimental design increased the difficulty of recruiting participants from the community at large. In any case, these factors did not prevent the analysis and comparison between the studies. It is also noteworthy that the interaction under investigation took place in a remote fashion. It is worth asking what the reactions to racism are in face-to-face interaction situations. Another limitation concerns the small number of participants in Study 2. The research scenario, which required four people to have a schedule at the same time, had a negative impact on the sample size.
Notwithstanding such limitations, the current research makes an important conceptual contribution to the field of racist humour, an area that remains very under-researched in social psychology. The results contribute to the understanding of the relationship between emotions, supposed reactions and actual behaviours in studies of prejudice and discrimination. Future studies could further explore these nuances (e.g., face-to-face data collection strategies) and promote effective strategies to reduce implicit bias and cultural sensitivity. It is also important to explore the relationships between racist humour and expressions of explicit and implicit bias in greater depth in new research, taking into account more diverse samples.
The research also makes the important and immediate practical contribution of exposing the racial hypocrisy of ‘do as I say, not as I do’. Our results demonstrate that despite the pressure of the anti-racist norm, which is particularly pronounced in the type of ‘WEIRD’ sample examined in both studies, racism and discrimination persist in Brazilian society, manifested either through inaction or indifference to the suffering of racialised groups, whether or not due to the lack of emotional mobilisation of white people who experience racist situations. Our findings provide important information about how racism operates in real-world contexts without the influence of societal norms of equality and justice. This may enable the development of better informed and more effective anti-racist interventions.
Data Availability
The entire dataset supporting the results of this study is available at https://drive.google.com/drive/folders/1xkGDjJhNjDFpfK6QlUxCtTKKMx6JxZIp?usp=sharing
