Abstract
We examined majority group members’ perceptions of racial slurs, compared to what we have labeled as combination terms. These combination terms possess the same semantic and pragmatic linguistic functions as racial slurs, functioning to express negative emotion toward, and to describe, a target. Across three studies (total N = 943) racial slurs were not perceived as significantly different from combination terms. We then examined whether participants higher in social dominance beliefs reported greater perceived justification for using combination terms over racial slurs because of their lack of historical denigration of marginalized groups that racial slurs have. Participants, even those higher in socially dominant attitudes, did not perceive greater justification for the use of combination terms than racial slurs. Indeed, an important implication is that race-marking, an understudied area of social psychology, paired with general derogative terms produces terms which may function similarly to racial slurs, but, fortunately, are also similarly vilified in modern society.
Pejoratives are a range of terms that intend to disparage and put down others. These terms may include slurs, expressive and other derogative terms (Corredor, 2014; Hom, 2008). Among these terms, the most powerful in terms of disparaging their target and creating/reinforcing status hierarchies are slurs, with the specific focus of our article being racial slurs. Racial slurs are terms that, in typical use, function as a downward, divergent communication strategy meant to create social distance between groups (see Giles, 2016). In this way, racial slurs reinforce traditional status hierarchies pushing one group downward while propelling another upward. Indeed, slurs have extreme negative impacts on their targets, including decreased self-esteem, anxiety and/or stress from repeated victimization by racial slurs, and lower career advancements from negative perceptions created or reinforced by racial slurs (Gabriel, 1998; Greenberg & Pyszcynski, 1985; Mullen, 2001, 2004). Recent empirical research has shown that racial slurs are generally perceived as highly offensive by participant samples (e.g., O’Dea et al., 2015; O’Dea & Saucier, 2017), suggesting that people in modern society generally perceive their use as unjustified. In recent decades, society has shifted somewhat away from these more overt forms of prejudice toward more covert expressions of prejudice in recent decades that carry lower rates of suppression, but racial slurs do continue to be used toward minority group members (e.g., Dovidio & Gaertner, 2000; McConahay et al., 1981; Nail et al., 2003; Saucier et al., 2005; Sydell & Nelson, 2000).
Motivations for the Use of Racial Slurs
Social Dominance Theory (SDT: Sidanius & Pratto, 1999) provides a broad understanding of individuals’ motivations for the use of racial slurs. SDT states that social hierarchies are reinforced through three types of behaviors: institutional discrimination (i.e., discriminatory behaviors perpetuated by institutions within society; see Sidanius et al., 1994), behavioral asymmetry (i.e., behaviors perpetuated by the dominant and subordinate groups to maintain hierarchies; Sidanius & Pratto, 1999), and aggregated individual discrimination (i.e., discriminatory behaviors perpetuated by individuals of one social group toward individuals of another social group; Sidanius & Pratto, 1999).
While institutional discrimination is a major component of SDT, we will focus on aggregated individual discrimination and ideological asymmetry (i.e., when individuals endorse certain ideologies or beliefs that maintain the social status hierarchy; Mitchell & Sidanius, 1993; Sidanius, 1993; Sidanius et al., 1994), which is a specific type of behavioral asymmetry. This is because institutional discrimination (e.g., harsher court rulings for minority than majority group members) generally involves laws and rulings that maintain status hierarchies, rather than the use of derogative language to maintain hierarchies. While important, institutional discrimination is less likely to occur regarding the use of racial slurs (though this has happened; e.g., the controversy around Webster’s dictionary defining the slur, “nigger,” as “a Black person,” Henderson, 2003; Himma, 2002).
When racial slurs are used to perpetuate ideological asymmetry, they function to convince majority group members that minority group members deserve to be discriminated against. This is because when a racial slur is used toward an individual, the slur can elicit negative stereotypical perceptions of the individual because of their racial group membership (i.e., stereotype threat; Devine, 1989; Greenburg & Pyszcynski, 1985; Steele, 1997). This allows the perpetrator to justify their discriminatory behavior by believing in the stereotypes attributed to the racial group to which the target belongs. Interestingly, research has found that when a racial slur targets a member of a racial minority group, negative stereotypical perceptions about the target are not only cognitively activated in the mind of the perpetrator, but also for the target and third-party observers (Jeshion, 2013; Merskin, 2010). These negative stereotypical perceptions may convince minorities to believe that they deserve to be discriminated against, even to the point that they favor racial majority group members over their own racial group (i.e., systematic outgroup favoritism, Sidanius & Pratto, 1999). Minority group members who experience outgroup favoritism will comply more with the existing unequal social hierarchy because they believe that they should have positive attitudes toward the overarching system to which they belong (see System Justification Theory: Jost & van der Toorn, 2012). Further, when these stereotypes are activated, they may influence the target to act in ways that reinforce the stereotypes (see self-fulfilling prophecies; Madon et al., 2018).
Individual discrimination, on the other hand, can be perpetuated through a variety of expressions, including humor (see Rappoport, 2005; Saucier et al., 2016, 2018), microaggressions, stereotypes, hate crimes, and racial slurs (see also Saucier et al., 2008). Recent research by Crandall et al. (2018) has shown an increase in recent years in the use of blatant hate speech, such as racial slurs and derogatory language, typically gets expressed when there is ambiguity or justification for individuals to claim that the expression is non-racially/prejudicially motivated (e.g., Saucier et al., 2017). With racial slurs and other overt expressions of prejudice, there is seldom room for ambiguity. However, below, we describe situations in which certain terms that function linguistically similarly (i.e., that have the capacity to express negative emotion toward, and describe targets) to racial slurs may provide greater room for ambiguity, possibly giving individuals, especially those higher in socially dominant and prejudicial beliefs, greater opportunity to justify their use.
The Linguistic Properties of Racial Slurs
Recent literature has debated the origins of the extreme negative connotations that racial slurs possess. Most simply, this explanation could be due to the history that these pejoratives have (Croom, 2011, 2014; Kennedy, 2002; Rappoport, 2005). However, recent theoretical discussion on the linguistic functions of racial slurs suggests their power may come, at least partly, from their linguistic properties. While no one debates the historical use of racial slurs to disparage their targets, it does seem the linguistic properties of racial slurs may contribute to the negatively expressive perceptions they evoke. Indeed, racial slurs and derogative terms are both used to insult and disparage others. However, they are functionally different, and thus, should be perceived as functionally different by the group that perpetrates them. Derogative terms (e.g., “asshole,” “motherfucker”) express negative emotion toward the target without derogating the target’s group identity (e.g., not referencing the target’s race, gender, etc.; Hom, 2008; Potts, 2007). As such, majority group members should not perceive, in typical use, the use of a derogative term toward a minority target to be as functionally offensive as a racial slur because derogative terms are not based on racial differences, but are instead a negative expression toward someone who happens to be a minority target. Thus, we predict that majority group members may perceive non-racial derogative terms to be less negatively expressive and racist than are racial slurs.
Similarly, a racial descriptor (e.g., “Black”) should not necessarily be perceived by a White individual to be an indication of disrespect in traditional uses. This is because the term “Black” (as well as other descriptive terms) carries no inherent derogation (although they may be used unnecessarily to race-mark which can carry negative connotations; see Anderson & Lepore, 2013; Hedger, 2012; Holt, 1995). That said, while we would expect that unnecessary race-marking in negative situations could evoke some feelings of derogation, these feelings would not be nearly as functionally offensive as a racial slur (likely being perceived similarly to a microaggression; see Pierce et al., 1978).
On the contrary, research has suggested that racial slurs function as a hybrid of both descriptors and expressives in that they have the potential to both describe targets and express derogation toward them (Croom, 2011, 2014; O’Dea & Saucier, 2017). It is this hybrid ability that separates racial slurs linguistically from truly descriptive and derogative terms which each serve predominantly one function, to describe or to express negative emotion, respectively (with the caveat that situational features likely moderate the ability of descriptive and expressive terms to solely function using their predominant function; O’Dea & Saucier, 2017). Because racial slurs possess both descriptive and expressive functions, they should be perceived by majority group members to be necessarily descriptive and negatively expressive toward targets (Croom, 2011, 2014; O’Dea & Saucier, 2017). Further, previous research has shown that the offensive capabilities of these terms may be inextricably rooted in the ability of these slurs to both express negative emotion toward their target as well as to describe the target (see O’Dea & Saucier, 2017).
Current Studies Overview
The focus of our current studies is to examine how majority group members perceive the use of pejoratives used toward minority group members with a specific focus on understanding how different features of these terms impact majority group members’ perceptions of the offensiveness of the terms. While recent research has examined perceptions of how racial slurs are perceived and the justification behind the use of a racial slur (O’Dea et al., 2015; O’Dea & Saucier, 2017), no research has empirically examined majority group members’ perceptions of racial descriptors, non-racial derogatives, racial slurs, and what we have labeled as combination terms (which have not been examined in the extant literature). Combination terms include a combination of race-marking descriptives (e.g., “Black”) and non-racial derogatives (e.g., “asshole,” “motherfucker”). It is in this way that the combination terms may disambiguate the potential ambiguity of either term being used separately to either describe or offend the target. Admittedly, each of these terms are slurring terms in that they express negative emotion toward their target and do so on the basis of his/her/their group membership. As such, the combination terms could also be described as “slurs.” However, they differ from traditional slurs, because traditional slurs belong to an intense lexical category of slurring terms, while combination terms function similarly to traditional slurs without the same intense lexical category labeling. As such, for the duration of this article, we will refer to traditional slurs (see Table 1) using the “slur” labeling, and the terms which combine the descriptive and expressive functions of racial slurs as “combination terms.” We also predict these combination terms may not be similarly suppressed by majority group members due to a lack of historical denigration that racial slurs possess.
Terms Used in Our Studies by a White Perpetrator Toward a Black Target.
These possibilities form the basis of our hypothesis, which we labeled our Taboo Nature of Slurs hypothesis. This Taboo Nature of Slurs hypothesis predicted that racial slurs would be perceived as more negatively expressive and racist than would the combined terms (i.e., ethnic descriptors combined with derogative terms) due to the intense and taboo history that racial slurs carry with little room for ambiguity in their typical usage. This would indicate that, despite combination terms having the potential for extreme derogation of people belonging to marginalized groups, majority group members may perceive greater justification for their use in modern society because they lack the intense and instantaneous taboo nature of racial slurs. This could be especially problematic in a society that generally purports to vilify overt expressions of prejudice (albeit potentially not as strongly in recent years; see the Trump Effect on Prejudice; Crandall et al., 2018).
To test our hypothesis, we conducted three studies comparing a racial slur to combination terms, a racially descriptive term, a non-racial derogative term, and a control term. Each of the terms we used in the current studies are presented in Table 1. In Study 1, we examined White participants’ perceptions of terms which combined a racially descriptive term with a non-racial derogative term compared to their perceptions of a traditional racial slur. Our Taboo Nature of Slurs hypothesis predicted that the racial slur would be perceived as more offensive than the combination terms. In Study 2, we replicated and extended the findings from Study 1 by examining how participants’ perceptions of the racial slur and combination terms compared to their perceptions of the racially descriptive term, non-racially derogative term, and a control term used in isolation. Our Taboo Nature of Slurs hypothesis, again, predicted that the racial slur would be perceived as more negatively expressive and racist than the combination terms, but also that the racial slur and combination terms would be perceived as significantly more negatively expressive and racist than the racially descriptive, non-racially derogative, and control terms in isolation. In Study 3, we then examined if participants’ perceptions of each of the terms were related to participants’ levels of socially dominant attitudes. We predicted that participants higher in social dominance beliefs would perceive less negatively expressive and racist perceptions of all the terms. However, building on recent research on the shift of prejudice to more covert expressions in modern society, we predicted that these negative relationships would be stronger for the more ambiguous terms (i.e., the combination terms, non-racial derogative term, descriptive term) than for the overtly racist slur term. We believe this examination of slur terms compared to terms sharing similar linguistic properties will help understand the justification and suppression that majority group members experience for the use of prejudicial language.
Study 1
In Study 1 we presented participants with a vignette in which a White individual used a racial slur, combination term 1, or combination term 2 toward a Black target and examined participants’ perceptions of the terms as negatively expressive. We also examined participants’ perceptions of the perpetrator as positive, likable, and racist. Our Taboo Nature of Slurs hypothesis predicted that, because of the historical use of the racial slur to disparage Black individuals, the slur would be perceived as significantly more negatively expressive, and that the perpetrator of the slur would be perceived as less positive and more racist, than the combination terms, indicating less perceived justification for the use of racial slurs compared to combination terms in modern society by majority group members.
Method
Participants
Participants for Study 1 were recruited via CloudResearch software (Litman et al., 2016). A power analysis for a MANOVA with one predictor with three levels, and four response variables with a power of .80, α = .05, and f2 = .0625 yielded a necessary sample of 196. Thus, we attempted to recruit at least 327 participants to account for a 40% loss in participants who either did not report their race, reported their race as something other than White, or who failed the attention checks. 439 participants accessed the study on Qualtrics. Twenty-nine participants did not complete the full study, thereby not reporting their ethnicity and were removed from data analysis. An additional 122 participants who reported their ethnicity as something other than White were removed them from data analysis because we were interested in how White individuals perceive combination terms compared to racial slurs and whether they perceive greater justification for the use of combination terms compared to racial slurs. Although it would be interesting to examine minority group members’ perceptions of these terms in future studies on this topic, our current sample did not allow us to make these comparisons. Nine more were removed for failing our attention check. This left 279 participants for data analysis which exceeded the necessary sample size requirements for the above MANOVA and subsequent ANOVA probings of the MANOVA (necessary sample = 158). There were 96 participants who reported their sex as male, 182 as female, and one as other. The average age of participants was 40.55 (SD = 13.41).
Materials
Vignette
We manipulated whether a slurring term used by a White individual to refer to a Black individual was a racial slur or one of two combination terms (see Table 1). In the interest of limiting the inclusion of these offensive terms in this article, we have omitted these terms throughout our discussion of these study methods, results, and discussion. We refer to them explicitly only as often as is necessary to clarify our methods. Participants saw the explicit terms used in the table and vignette below (brackets denote alternate conditions). Additionally, while stereotypic names of White individuals (i.e., Mason and Sam) were chosen, race was not explicitly mentioned or shown through images. This limitation is addressed in Study 2: Mason and Sam are friends that are having lunch one afternoon. After welcoming each other and ordering their food, Mason and Sam begin catching up. Halfway through lunch, Mason remembers what happened to him at the grocery store. Mason begins the story with, “the other day this nigger [Black motherfucker] [motherfucking Black guy] was standing in front of me in line at the grocery store.” Mason proceeds to tell Sam how the man could not make up his mind up on which candy bar to buy before checking out his groceries. Mason expressed how annoyed he was with having to wait for the man to make up his mind.
Perceptions of the slur use
We used O’Dea and Saucier’s (2017) positively and negatively expressive measures. However, these were combined and modified for the current study because four of the items (e.g., “This term was used by Mason to bond with DeShawn”) referenced a direct interpersonal encounter between the White perpetrator and Black target of the slur. However, the current study was not a direct encounter, but instead a story told by a White individual to another White individual. This yielded a measure with 4 items (e.g., “This phrase was used to express negative emotion toward the individual in Mason’s story”) which were measured on 1 (strongly disagree) to 9 (strongly agree) scales. Participants’ responses to each of the items were averaged after reverse-scoring the two positively expressive items to create composite scores with higher scores indicating more negatively expressive perceptions of the term. Additionally, we included 1 item with this measure assessing perceptions of the perpetrator of the slur as racist (“Mason is racist”) that was analyzed separately.
Positive perceptions of the perpetrator
To measure participants’ perceptions of the perpetrator of the slur, we used seven positive and five negative traits (e.g., smart, capable, superior, inferior, incompetent) to which participants responded to a prompt asking them to rate the perpetrator of the term on 1 (strongly disagree) to 9 (strongly agree) scales. In the interest of clear communication about all included items, three additional traits were included but were not analyzed due to their being ambiguous about whether they are measuring positive traits about the person (disadvantaged, privileged, equal) and were removed a priori. Composite scores were calculated by reverse scoring negative traits and then averaging participants’ responses to each of the items with higher scores indicating more positive perceptions of the perpetrator.
Likability
We then measured participants’ perceptions of the perpetrator of the slur as likable using the 5-item subscale from McCroskey and McCain (1974)’s interpersonal attraction scale (e.g., “I think he could be a friend of mine”). Participants responded on 1 (strongly disagree) to 9 (strongly agree) scales and responses were averaged to create a composite score with higher scores indicating greater social attraction toward the perpetrator of the term.
Procedure
Participants were recruited via the CloudResearch software (Litman et al., 2016) software. After providing informed consent, participants read the vignette. Participants reported their negatively expressive perceptions of the term, their positive perceptions of the perpetrator of the term, and their likability perceptions of the perpetrator of the term. Participants also responded to one item assessing the extent to which they perceived the perpetrator of the term as racist for the use of this language. On a separate page, participants also indicated who they believed was being insulted (the Black person, Black people generally, White people generally, the person Mason was telling the story to). 1 Participants then provided demographic information and were debriefed and thanked for their participation. We report all measures, manipulations, and exclusions in these studies.
Results
To test our Taboo Nature of Slurs hypothesis, we conducted a one-way between-groups multivariate analysis of variance. The term used by the White individual about the Black individual was entered as the independent variable, while the negatively expressive perceptions, perceptions of the perpetrator as racist, positive perceptions of the perpetrator, and likability of the perpetrator were entered as the dependent variables. The omnibus test yielded a significant main effect of slur predicting the set of dependent variables, Wilks’ λ = .90, F(8, 546) = 3.57, p < .001, η2 partial = .05. Upon examination of the univariate one-way between-groups ANOVAs, our results yielded a significant main effect of slur predicting participants’ perceptions of the slur as negatively expressive, F(2, 276) = 3.28, p = .039, η2 partial = .02; and perceptions of the perpetrator as racist, F(2, 276) = 5.20, p = .006, η2 partial = .04; but not participants’ positive perceptions, F(2, 276) = 0.17, p = .846, η2 partial < .01; or likable perceptions, F(2, 276) = 0.83, p = .437, η2 partial = .01; of the perpetrator of the slur.
We probed the two significant ANOVAs using Bonferroni pairwise comparisons which showed that participants perceived combination term 1 (M = 8.40; SD = 1.13) as significantly more negatively expressive than the racial slur (M = 7.91; SD = 1.27; p = .034), but neither was significantly different from combination term 2 (M = 8.12, SD = 1.27; ps > .403). Further, the racial slur (M = 7.49, SD = 1.94) was perceived as significantly more racist than combination term 2 (M = 6.48, SD = 2.38; p = .005), but neither were significantly different from combination term 1 (M = 6.82, SD = 2.14; ps > .117). These findings are shown in Figure 1. These results generally did not support our Taboo Nature of Slurs hypothesis. Instead, combination terms were generally not perceived differently than the racial slur by majority group members, except that the racial slur was perceived as slightly more racist than combination term 2. These findings have important implications for the theoretical linguistic functions of racial slurs. Specifically, while racial slurs have undoubtedly garnered negativity from their historical oppression, our racial slur did not significantly differ from terms possessing the same linguistic properties suggesting that slurs may garner at least some of their power from their linguistic abilities to disparage on the basis of group membership.
Perceptions of the slurs used in Study 1.
Study 2
In Study 2, we replicated the findings of Study 1 by examining how a racial slur compared to the combination terms that possess the same linguistic abilities to express negative emotion toward, and describe, people of another social group. We again measured White participants’ perceived justification for the terms which we operationalized using participants’ perceptions of the terms as negatively expressive and their perceptions of perpetrators of these terms as positive, likable, and racist. We then extended these findings by examining how these terms compared to the non-racial derogative term and racially descriptive term (see Table 1). We also compared terms to a control term, “guy.” Consistent with our findings in Study 1, in which our Taboo Nature of Slurs hypothesis was generally not supported, with the racial slur not being perceived as significantly more negatively expressive and racist than combination terms, we did not expect these terms to significantly differ. That being said, we predicted that the slur and combination terms would be perceived more negatively expressive, and perpetrators of these terms would be perceived as less positive and racist than either the descriptive term or non-racial derogative term in isolation. We also predicted that the non-racial derogative term and racial descriptor would be perceived as significantly more negatively expressive and that perpetrators of these terms would be perceived as less positive, likable, and more racist, than perpetrator of the control term.
Method
Participants
A power analysis for a MANOVA with one independent variable with six groups, four response variables, .80 power, α = .05, and an f2 = .0625 yielded a necessary sample size of 196 participants. We recruited participants via the CloudResearch software (Litman et al., 2016) and, because a typical sample results in 60% White individuals who complete all portions of the study, we attempted to recruit at least 327 participants to account for participants who were not White and participants who did not complete the full study. About 415 participants accessed the study on Qualtrics online survey software. We removed 135 participants who either did not answer or who indicated their ethnicity as something other than White because we were, again, interested in White individuals’ perceptions of these terms and their perceived justification for the use of these terms. This left 280 participants for data analysis which exceeded the necessary requirements for the above MANOVA and subsequent ANOVA probing of the effects (necessary sample size of 211). Of these participants, 74 self-reported as male, 205 as female, and 1 as other. The average age of participants was 40.92 (SD = 13.38).
Materials
Vignette
We presented participants with a vignette similar to that used in Study 1, and we manipulated the slur that was used by a White individual in reference to a Black individual as a racial slur, one of two combination terms, an expressive term, a descriptive term, or a control term. Participants read about a White individual describing an interaction with a Black individual. Specifically, participants read that “Mason and Sam are on their way to lunch. After arriving at the restaurant, Mason and Sam are seated in a booth. After their food arrives Mason sees someone that he recognizes sitting in an adjacent booth.” Images of Mason and Sam (both White individuals) were included. Participants also saw an image of a Black man as the “person that Mason recognizes.” The rest of the short story was then manipulated according to explicit slurring term condition and is provided below (the brackets denote alternate conditions): Mason decides to tell Sam about the person he recognizes. Mason says to Sam, “Hey, see that guy [Black guy, motherfucker, Black motherfucker, motherfucking Black guy, nigger] over there? The other day I was in line at the grocery store and he couldn’t make up his mind on what candy bar he wanted. He made me late to work because he took over 5 minutes deciding which one he wanted.”
Dependent measures
Each of the measures, except the likability measure (due to this measure largely being redundant in Study 1 with the positive perceptions measure) used in Study 1 were, again, used in Study 2. These measures include perceptions of the terms as negatively expressive, the perceptions of the perpetrator as racist, and general positive perceptions of the perpetrator.
Procedure
We recruited participants via the CloudResearch (Litman et al., 2016) software and they completed the current study on Qualtrics online survey software. After providing informed consent, participants read the vignette. They then reported their perceptions of the manipulated term as negatively expressive, their perceptions of the perpetrator as positive, and their perceptions of the perpetrator as racist as in Study 1. We did not use the likability measure used in Study 1 because these findings were generally redundant with the other measures. On a separate page, participants again indicated who they believed was being insulted (the Black person, Black people generally, White people generally, the person Mason was telling the story to). 2 Participants then completed the demographic information and were debriefed and thanked for their time. We report all measures, manipulations, and exclusions in these studies.
Results
To test our Taboo Nature of Slurs hypothesis, we conducted a one-way between-groups MANOVA. Slur was entered as the independent variable and how negatively expressive participants perceived the slur, how racist participants perceived the perpetrator of the slur, and how positive participants perceived the perpetrator of the slur were entered as the dependent variables. The MANOVA yielded a significant effect of slur predicting the set of dependent variables, Wilks λ = .28, F(15, 751.27) = 29.34, p < .001, partial η2 = .35. Upon examination of the univariate effects, there was a significant effect of slur predicting participants’ perceptions of the slur as negatively expressive, F(5, 274) = 85.17, p < .001, partial η2 = .61; of the perpetrator as racist, F(5, 274) = 49.80, p < .001, partial η2 = .48; and of the perpetrator as positive, F(5, 274) = 18.80, p < .001, partial η2 = .26.
We probed these effects using Bonferroni pairwise comparisons. These comparisons are presented in Table 2 and Figure 2. Consistent with the findings of Study 1 but not our Taboo Nature of Slurs hypothesis, the slur was generally not perceived to be significantly different from the combination terms (albeit was perceived as more racist than combination term 2). Consistent with our predictions, however, the combination terms and the racial slur were perceived as significantly more racist, and the perpetrator was generally perceived less positively after using combination terms and slur toward the Black individual in his story than the non-racial derogative term, the descriptive term, or the control term. Interestingly, the non-racial derogative term was not perceived by White individuals to be significantly different from the racial slur and combination terms in participants’ negatively expressive perceptions. Further, the racial descriptor was perceived as more racist than the control term. This indicates that, while it is not perceived as substantially racist (3.5 on a 1–9 scale), the act of indicating someone’s race (i.e., race-marking) in a situation in which it was unnecessary leads to greater perceptions that the negative statement was at least somewhat prejudicially motivated. The descriptive term was not perceived as different from the non-racial derogative term except that it was perceived as less negatively expressive than the non-racial derogative term.
Means, Standard Deviations, and Bonferroni Pairwise Comparisons Between Slur Conditions in Study 2.
Note. Means in a row that do not share a common subscript are significantly different from one another.
Perceptions of the slurs used in Study 2.
These results replicate and extend our previous study by again failing to support our Taboo Nature of Slurs hypothesis that the combination terms would be perceived as less negatively expressive and racist compared to a racial slur. Thus, it appears that, while racial slurs are rooted in the historical denigration of marginalized groups, their vilification by majority group members in modern society may be largely a result of their linguistic functions which give the slurs the ability to describe and derogate the targeted individual.
Exploratory Analyses
We also examined how slur condition impacted participants’ perceptions that the slur was insulting toward the Black individual in the story, to Black people generally, to White people generally, and to the friend that the White individual was telling the story to. We expected that participants would perceive the slurs and combination terms as being quite insulting toward the Black target and Black people generally. That said, we expected that the non-racial insult would only be derogative toward the target because it does not make reference to the person’s race. We did not make specific predictions about how the slurs would be perceived as offensive to White individuals and to the friend the White person was speaking to. Consistent with these expectations, there was a significant effect of race predicting the set of items, Wilks λ = .48, F(20, 899.76) = 11.19, p < .001, η2p = .17. There was a significant effect of slur predicting each of the outcomes, Fs(5, 274) > 9.80, ps < .001, η2ps > .15. We probed these effects using Bonferroni pairwise comparisons. These findings are shown in supplemental Table 2 (which can be seen in the online version of this article). Consistent with the above findings (but inconsistent with our Taboo Nature of Slurs hypothesis), there were no significant differences between the slur and combination terms predicting participants’ perceptions of the terms as disparaging toward the Black individual in the story or disparaging toward Black people generally. Interestingly, none of these were significantly different from the non-racial derogative term either in participants’ perceptions of the terms as disparaging toward the Black individual, but the non-racial derogative term was perceived as significantly less derogative toward Black people generally—indicating the slurs and combination terms are perceived to disparage more than just individuals, implying the entire social group is despicable on the basis of their racial membership. All of these were perceived as more disparaging than the descriptive and control term toward the Black target. However, the non-racial derogative was not perceived significantly different from either the descriptive or control term in how disparaging it is to Black people generally. Interestingly, it was the descriptive term that was perceived as significantly more offensive to Black people generally than the control term suggesting support for the microaggression hypotheses described above. Interestingly, the racial slur and combination term 1 were both perceived as significantly more insulting toward White people generally and toward the White friend the White individual was talking to in the vignette than the control term and the racial descriptor which may indicate that participants actually saw the overt prejudice being exhibited as potentially damaging toward the reputation of White individuals as a whole. This implication should be tested in future studies but is tangential to the research question we are addressing and will not be discussed further.
Study 3
Studies 1 and 2 generally did not support our Taboo Nature of Slurs hypothesis. This may indicate that slurs are perceived by majority group members to garner much of their power to offend (especially entire groups of people) from their linguistic properties functioning to not just derogate the target, but to derogate the target on the basis of their race. While Studies 1 and 2 did not show that combination terms differed from the racial slur in terms of participants’ perceptions of the slur and the perpetrator of the slur, it is possible that people higher in socially dominant attitudes may perceive lower negative perceptions of terms, and people who use terms, which combine racial descriptors with derogative terms due to the potential ambiguity surrounding these combination terms. This is because in modern society, overt expressions of prejudice are generally vilified. It does appear that people generally perceive that combination terms should similarly suppressed in comparison to overt racial slurs. However, it could also be that people higher in socially dominant attitudes may perceive greater justification for combination terms, essentially making statements like, “It isn’t a racial slur. All I did was say his race and insult him. That’s not racist!” To test this possibility, we presented participants with the six terms used in Study 2 and examined the relationships between participants’ socially dominant attitudes and their perceptions of the terms. We predicted that people higher in social dominance beliefs would generally perceive all the terms in reference to the Black individual as less negative and less racist (consistent with O’Dea et al., 2015), but that these relationships would be weakest for the slur condition due to this being the most obviously derogative and prejudicially motivated, and thus least able for its expression to be rationalized.
Method
Participants
A power analysis for linear multiple regression fixed model R2 increase with 11 tested predictors (six main effects, five two-way interactions) with .80 power, and .05 alpha with a f2 = .0625 yielded 279 participants necessary. Because a typical sample results in approximately 60% of White individuals who complete the entire study, we attempted to recruit at least 465 participants to account for participants who were not White and participants who did not complete the study for a final sample of at least 279 White participants. A total of 562 participants accessed the current study, and 178 participants were removed who either did not respond to or indicated their ethnicity as something other than White, again due to our interest in examining White individuals’ perceptions of justification for these terms. This left 384 participants for data analysis. Of these participants, 117 self-reported as male, 265 as female, and 2 as other. The average age of participants was 39.36 (SD = 12.82).
Materials and Procedure
We recruited participants via the CloudResearch software (Litman et al., 2016) who completed the study on Qualtrics. After providing informed consent, they completed the 16-item Social Dominance Orientation (SDO) scale (Pratto et al., 1994) which is a widely-used measure of social dominance beliefs with items such as, “Some groups are simply inferior to other groups” (participants responded on 1[strongly disagree] to 9[strongly agree] scales). Participants were then presented with one of the six randomly assigned terms used in Study 2. It is important to note that there was no context provided for the term in the current study, but participants were asked to respond to how they perceive the term in typical use. This methodological change allowed us to test participants’ perceptions of the terms beyond specific situations, instead testing their perceptions of the terms as they are perceived being most often used. For the term that they were randomly assigned, participants responded to three items assessing their perceptions of the term as negatively expressive (“This language is typically used to express negative emotion toward Black individuals”), descriptive (“This language is typically used to describe Black individuals”), and racist (“This language is racist”) on 1(strongly disagree) to 9(strongly agree) scales. Participants then completed demographic information and were debriefed and thanked for their participation. We report all measures, manipulations, and exclusions.
Results
To test our hypotheses, we conducted a series of multiple regression analyses. Participants’ levels of SDO, the slur term, and their interaction were entered as the independent variables and participants’ perceptions of the terms as negatively expressive, descriptive, and racist were tested as the dependent variable in three regression analyses.
Consistent with Studies 1 and 2, the racial slur was not perceived differently compared to the combination terms (see Table 3), but each of these were perceived as significantly more racist, more negatively expressive toward Black individuals, and more racist than the racial descriptor, non-racial derogative term, and the control term. These findings again fail to support our Taboo Nature of Slurs hypothesis, instead suggesting that racial slurs and combination terms are not perceived as significantly different. These findings also highlight the generalizability of our findings by showing these effects even in situations in which no context is provided, thus allowing participants to indicate their perceptions of the terms in what they perceive to be typical use of each term.
Means, Standard Deviations, and Comparisons Between Slur Conditions in Study 3 From the Separate Regressions.
Note. Means in a row that do not share a common subscript are significantly different from one another.
We then examined the relationships between participants’ levels of SDO and their perceptions of each of the terms as negatively expressive toward Black individuals, descriptive of Black individuals, and racist. These relationships between participants’ levels of SDO and their perceptions of each term as negatively expressive, descriptive, and racist in typical use are presented in Supplemental Figures 1 to 3 (which can be accessed by viewing the online version of this article) respectively. Contrary to our hypotheses, participants’ levels of SDO generally did not predict or interact with perceptions of any of the categories of term. SDO was associated with significantly more descriptive perceptions of the non-racial derogative term, B = 0.43, p = .049; indicating that people higher in SDO may perceive this language as more normative than those lower in SDO. The only effect of SDO that was significant for the slur/combination terms was in the combination term 1 condition predicting participants’ perceptions of the term as racist, with greater levels of SDO being associated with significantly lower perceptions of this term as racist, B = −0.29, t = −2.00, p = .046 (see Supplemental Figure 3 in online version of this article). That said, this finding should not be over-interpreted as being consistent with our hypothesis because the slope of this line was quite similar to the slope of SDO in the combination term 2 (B = −0.28, p = .120) and racial slur (B = −0.24, p = .104) condition, indicating that SDO is weakly negatively associated with perceptions of each of these as racist (see Supplemental Figure 3 in online version of this article).
Taken together, our findings replicated the findings of Studies 1 and 2 providing evidence against our Taboo Nature of Slurs hypothesis; racial slurs were not perceived as significantly different in negativity compared to combination terms which possess similar linguistic properties to disparage individuals on the basis of their racial group membership. Indeed, it appears that, while racial slurs have been used for decades to disparage individuals belonging to other racial groups (thus having very little room for ambiguity in modern society), terms which combine the two features of racial slurs to describe the target and to derogate the target causes the terms to be perceived quite negatively, even to the point that they were not perceived significantly differently from racial slurs. Extending the findings of Studies 1 and 2, we examined whether individuals higher in social dominance beliefs might be motivated to perceive combination terms as less negatively expressive and racist compared to decrease the suppression of these potentially more ambiguous combination terms. Our results did not support this hypothesis. Instead, the current study provides some reason for optimism about the perceptions of racial slur use in society. It appears that even those with motivations to push down other social groups perceive slurs and combination terms to be highly offensive and perpetrators of these terms are perceived negatively regardless of one’s level of SDO.
General Discussion
Racial slurs are unique terms that have the ability both to indicate the racial group membership of the target and to derogate the target on the basis of that racial group membership. Thus, racial slurs differ from non-derogative descriptive terms and non-racial derogative terms in that they combine each of these functions, making them unique linguistic terms with an intense and taboo history to denigrate marginalized groups. In the current studies, we extended previous research by examining how participants’ perceptions of racial slurs compared to participants’ perceptions of combinations of terms which possessed the same linguistic properties as a racial slur, being both descriptive and derogative. We proposed our Taboo Nature of Slurs hypothesis which predicted that if racial slurs are vilified based on their historical uses, they would be perceived to be vilified at higher levels than combination terms in modern society.
Across three studies, we showed that the slur term and combination terms were generally not perceived as significantly different, suggesting that, in contemporary society, slurs likely garner a great deal of their power from their linguistic properties, thus allowing the racial slur to derogate the entire targeted social group rather than one specific targeted individual (a finding which was further supported in Study 2 in our exploratory analyses showing that slurs and combination terms did not significantly differ from a non-racial derogative term in how disparaging they were toward the target, but that they were significantly more disparaging toward the targeted group more broadly). Further, greater social dominance beliefs were unrelated with perceptions of these terms indicating that people, even those higher in social dominance beliefs, perceive both the racial slur and the combination terms to be highly offensive.
The current studies do have important limitations. These studies were conducted cross-sectionally with vignettes and used self-report data. The completion of the social dominance beliefs and prejudicial attitudes measures right before reading the vignettes and reporting their perceptions of the slur use may have impacted participants’ perceptions of the slurs in Study 3. That said, participants’ perceptions of the slurs across the studies seemed to be fairly consistent regardless of whether they completed the measure of social dominance beliefs before (Study 3) or not (Studies 1, 2). The effects we showed may also be different if participants were exposed to these slurs in a video or in real life. In our studies, these terms were used in private settings and among friends. In more realistic situations or in public situations, participants may be more able to gauge the emotions and intent of the perpetrator via non-verbal communication. Our statistical tests are also likely somewhat impacted by a ceiling effect for both the slurs and combination terms in terms of how negatively they were perceived. Indeed, different relationships/mean differences may emerge when this ceiling effect is addressed. Fortunately, it seems that people do perceive combination terms and racial slurs to be quite negatively expressive and racist, and while there may be more subtle differences if the distributions of these perceptions were more normal, we do not believe this to substantially impact our conclusions that both combination terms and racial slurs are damaging, and that the semantic and pragmatic features of each seem to be primarily what lead to these terms being so negatively perceived and vilified in modern society.
The current studies are also limited in their testing of only one racial slur and terms which were used by a White individual to disparage a Black individual. As such, these findings may not generalize to slurs targeting other social groups or other races. Research should examine the replicability of our findings to slurs targeting other social groups. We expect the current findings will generalize to slurs targeting other social groups with slurs toward each of these groups (e.g., “bitch” targeting women) being perceived as similarly negative to terms which both disparage (e.g., “asshole,” “motherfucker”) and indicate group membership of the individuals (e.g., “woman”), thus possessing similar linguistic properties. Importantly, the effects of social dominance and prejudicial beliefs on participants’ perceptions of other slurs or slurs targeting other social groups may be stronger. If discrimination toward the targeted social group is perceived as more acceptable in society, there may be less of a ceiling effect for the terms, allowing for more variance to be accounted for by participants’ prejudicial attitudes toward the targeted social group(s).
These studies are among the first to empirically examine how people perceive the racial slurs compared to terms which possess similar linguistic properties, combining descriptive and derogative terms together. Racial slurs have extremely negative impacts on targets of the term ranging from the target(s) feeling devalued, having extreme negative emotions, can result in the target perpetrator, or bystanders experiencing stereotypic thinking or perceptions that the target is deserving of discrimination, and are used to reinforce existing status hierarchies (Gabriel, 1998; Greenberg & Pyszcynski, 1985; Jeshion, 2013; Merskin, 2010). As such, racial slurs have been increasingly vilified in modern society (although this shift in recent years may be tempered; see “the Trump effect on prejudice”; Crandall et al., 2018). That said, we predicted that terms which contain the same linguistic features as a racial slur may theoretically possess similarly derogative potential as racial slurs, while simultaneously being perceived as more justified (especially by individuals higher in social dominance beliefs) due to these terms not having the level of association with historical derogation that racial slurs have (see Crandall & Eshleman, 2003). Fortunately, our findings did not support these hypotheses. Instead, combination terms and racial slurs were generally not perceived as being significantly different, indicating that individuals, even those higher in prejudicial and/or socially dominant attitudes do not perceive justification for the use of these terms.
Future studies should address how to best reduce the negative use and impacts of racial slurs by majority group members toward minority group members. Rather fortunately, it appears that racial slurs and other terms which disparage individuals on the basis of their group membership are increasingly vilified in modern society. However, the terms are still used in modern society despite norms discouraging their use. As such, simply attempting to prohibit the use of specific racial slur terms (see Anderson & Lepore’s prohibitive strategy) may not be the most effective strategy because other terms which combine descriptive and derogative properties can emerge. Some recent theoretical discussion has been aimed at this objective with two competing theories emerging. The first, as noted above, is Anderson and Lepore’s (2013) prohibitive strategy suggesting the best way to get rid of derogative language is to bar its use in all situations. However, as some authors point out (e.g., Bianchi, 2014; Croom, 2011, 2014; Jeshion, 2013; Spotorno & Bianchi, 2015), an additional strategy that minority groups have and can use to reduce the negative impacts of racial slurs is what O’Dea and Saucier (2020) have labeled as the subversive perspective by which social groups reappropriate and use language previously used to disparage them in an intentional prosocial way to affiliate rather than disparage. That said, these two competing theories have yet to be fully empirically tested and future research should examine their efficacy at reducing the negative impacts of slurs.
These studies also highlight a need for additional research on race-marking as a way to subtly discriminate against individuals belonging to other social groups (see Holt, 1995). Our findings provide support for the negative effects of adding a racially descriptive term to a derogative term, and show that simply adding a racial descriptor to a neutral term (e.g., “Black guy”) increases negatively expressive and racist perceptions compared to a control term. Given the rise in covert expressions of prejudice and microaggressions that persist despite the general downward trend of overt expressions of discrimination in recent decades (see Dovidio & Gaertner, 2000; McConahay et al., 1981; Nail et al., 2003; Sydell & Nelson, 2000), research on subtle forms of prejudice, race marking, and their effects are important. To this point, the phenomenon of race-marking and the negative effects of race-marking on minority group members are understudied. Taken together, our studies present a novel examination of racial slurs that advances recent theoretical discussion on how racial slurs differ from other derogative and racially descriptive language. Specifically, it appears the semantic abilities of these terms to disparage individuals on the basis of their group membership, and not their taboo and intense history, may be primarily what leads to the modern-day discouragement of the use of racial slurs among majority group members, and that race-marking general derogative terms may produce dangerous linguistic weapons, akin to racial slurs, that allow the continued denigration of marginalized groups.
Supplemental Material
ODea_Smith_Saucier_Supplementary_Materials – Supplemental material for Referencing Race = Racist? Examining Perceptions of References to a Target’s Race as Offensive
Supplemental material, ODea_Smith_Saucier_Supplementary_Materials for Referencing Race = Racist? Examining Perceptions of References to a Target’s Race as Offensive by Conor J. O’Dea, Bayleigh N. Smith and Donald A. Saucier in Journal of Language and Social Psychology
Footnotes
Acknowledgements
We would like to thank two anonymous reviewers and the Editor for their immense help in restructuring and reframing the current manuscript. We believe the current manuscript is much better as a result of their suggestions and are excited to share it with others.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Part of this project was paid for using funds from a Kansas State University Arts and Sciences Undergraduate Research Award presented to Bayleigh Smith.
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
