The Language of Derogation and Hate: Functions,Consequences,and Reappropriation

The Language of Discrimination: Definition, Contexts, and Targets

For the purposes of this review, we define derogatory language as any disparaging statement referring to a social category as a whole or to its members. Such language may consist of entire sentences or specific labels, often referring to physical traits (“nigger,” “darkie,” “cunt”), to cultural habits (“holy roller,” “kraut”), or to the animal world (“cockroach,” “bitch”). Defined broadly, derogatory language includes any type of insinuation and allegation about members of a given social category that explicitly ridicules or insults them. As defined here, derogatory language embraces numerous phenomena that have been studied under more specific labels (e.g., ethnophaulisms, ethnic or racial slurs, sexist language, homophobic epithets, verbal bullying, etc.). It also includes hate speech, an extreme form of derogatory language, which constitutes a line of research on its own and which involves the expression of hate and/or the encouragement of violence against others based on their real or assumed membership in a given category. Given the increasing prevalence of hate speech, its specific collocation in online communication, and the legal problems it poses, we will discuss this type of language discrimination in a separate section.

Which social identities may be targeted by explicit derogatory language? Theoretically any social category is at risk, including those in power (e.g., the “elite,” the “establishment”), but the risk seems inversely related to a group’s social status and size. Starting from Mullen’s (Mullen & Smyth, 2004; Mullen et al., 2001) groundbreaking work on immigrant groups, it has been shown that, everything else being equal, derogatory labels are more numerous, more frequently used and more offensive, the smaller and the less familiar the social group is, and/or the lower its status in society. For instance, in Western countries, derogatory language is more likely to be used against religious minorities (e.g., Jews and Muslims) than against religious majority groups (e.g., Christians; for an overview see Nadim & Fladmoe, 2016). Moreover, recent studies suggest that those at the intersection of two or more minority categories, such as Muslim women or Black gay men, over-proportionally attract offensive or hateful comments (e.g., Morales et al., 2019).

Although these minorities may not pose concrete threats over limited resources such as jobs, they may be seen as posing symbolic threats to cultural norms and values (Stephan & Stephan, 2000).

The Social Functions of Derogatory Language

Why do people use derogatory language and hateful speech and what psychological and social functions does it serve? Researchers have identified the psychological antecedents of disparaging language and hate speech (for an overview of relevant psychological theories see Walters et al., 2016). These include (a) personality variables such as right-wing authoritarian personality, social dominance orientation, and the motivation to express prejudice (Forscher, 2015), (b) affective or emotional states specifically directed at the target such as hate, contempt, disgust, and feelings of intergroup threat (Gerstenfeld, 2002), and (c) more general (target-unspecific) motivations, such as thrill seeking (McDevitt et al., 2002) and self-esteem maintenance (Golec de Zavala et al., 2020). Moreover, psychological research has identified likely social functions served by derogatory language that can be grouped into five main categories: prejudice perpetuation; maintenance of status hierarchies; legitimization of violence against outgroups; norm and role compliance; and ingroup cohesion. We are not claiming that people engage in language-based discrimination with these aims in mind but, rather, that these are the primary social functions and consequences of disparaging language, which may well operate outside of people’s awareness and intention (see Burford-Rice & Augoustinos, 2017). To illustrate these functions, we will briefly and selectively review the relevant literature.

First, discriminatory language is one way to activate and perpetuate prejudice against minority groups. This mechanism is illustrated by a set of studies in which people were exposed to, or subliminally primed with, homophobic labels, without being personally the target of the slur (e.g., Bianchi et al., 2019; Fasoli, Maass et al., 2015). These and related studies compared homophobic labels (e.g., “fag”) with neutral category labels describing the same category (e.g., “gay”). This line of research coherently shows that both derogatory (e.g., “fag”) and neutral category labels (e.g., “gay”) activate stereotypes and they do so to approximately equal degrees. However, derogatory labels are much more likely than neutral category labels to trigger a strongly negative affective and evaluative reaction. Similarly, Bilewicz and Soral (2020) found that digital media users (i.e., the media users more exposed to hate speech) reported higher levels of Islamophobia, a relationship explained by greater acceptance of hate speech. Thus, derogatory labels seem to specifically affect prejudice, producing a more negative explicit and implicit evaluation of the group, but they do not activate stereotypes any more than neutral category labels.

The second function concerns the maintenance of existing group hierarchies. Derogatory language is intrinsically linked to social hierarchies, both in its use and its effect. Compared to members of socially subordinate groups, those of dominant groups are more likely to use racial slurs, are less likely to be at the receiving end of such language discrimination and, importantly, are less likely to speak up when overhearing others use racial slurs (Rosette et al., 2013; for exceptions see later section on confronting). At the same time, slurs are considered more offensive when directed against low status groups because people anticipate a stronger emotional reaction in this case (Henry et al., 2014). Thus, the asymmetrical use of disparaging language not only reflects the existing social stratification, but it also contributes to maintaining the power differential between dominant and subordinate groups. This is evident in the case of heterosexual participants showing more outgroup discrimination in resource allocation after exposure to homophobic labels (Fasoli, Maass et al., 2015). This may also explain why people of dominant groups (e.g., White people), especially when high in social dominance orientation, react negatively against slur reappropriation by the target group (e.g., Black people: O’Dea, 2019).

The fact that slurs are mainly used against members of low status groups, therefore, reinforces existing hierarchies. To quote Simpson (2013, p. 7), “the best way to make sense of the claim that hate speech in general inflicts harm on people, is to think of hate speech as something that contributes to (identity-based) social hierarchies.” Slurs not only keep social minorities in their subordinate position, but also assure the privileged position of the dominant group.

Third, derogatory language and hate speech are tools to legitimize violence and social exclusion, either by dehumanizing outgroups or by portraying them as a threat to the ingroup. Dehumanization generally involves the most extreme language forms, such as hate speech, the use of animal or object metaphors (Haslam, 2006) and, more generally, slurs of low complexity (Leader et al., 2009; Simon & Greenberg, 1996). These language forms signal that the target person is excluded from humankind or that those belonging to the same category cannot be considered full members of the human species, as they lack some of its essential features (e.g., intelligence, verbal abilities, specifically human emotions). Dehumanizing verbal attacks generally address people “for who they are, not for what they have done” (Fischer et al., 2018, p. 315). That is, they refer to “essential” internal qualities (e.g., ape-ness) rather than to specific actions, which makes defense very difficult. Verbal attacks of this sort are often accompanied by equally dehumanizing images as in the case of animal metaphors (see Haslam et al., 2011).

The dehumanizing narratives and images have been documented, among others, for homophobia (Fasoli et al., 2016), anti-Semitism (Volpato et al., 2010) and racism (Goff et al., 2008). For instance, exposing participants to homophobic labels (e.g., “faggot”), rather than to neutral category labels (e.g., “gay”) or generic insults (e.g., “asshole”), increased their implicit dehumanization of gay men as a social category. Fasoli et al. (2016) conclude that “homophobic epithets foster dehumanization in ways that other insults or labels do not” (p. 237). Different authors have argued that language- and image-based denial of humanity ultimately serves to free people from moral restraints and to legitimize violence (Bar-Tal & Hammack, 2012). History is replete with examples of how hateful discourse generates a fertile breeding ground for intergroup violence and at the most serious end, genocide. Hostile and persistent anti-Semitic propaganda by the Nazi regime exploited longstanding folklore narratives that depicted Jews as “vermin” and less than human, comparing them to “parasites” or “rats” (Volpato, 2017). Likewise, Indigenous peoples in the Americas, Africa, and Australia were subject to massacres and inhumane government policies, all of which were predicated on the colonizers’ beliefs of their racial superiority (see Volpato, 2017, for a historical overview). Such examples are not limited to the distant past: as described by Donohue (2018), the genocide in Rwanda in 1994 was precipitated and fueled by media calls to exterminate the Tutsi (labeled Invenzi, cockroaches) and as recently as 2017, the Rohingya minority in Myanmar has been persecuted by state sanctioned violence fueled by Facebook posts depicting them as dogs and pigs (Dussich, 2018).

A different language strategy to justify hostile acts against outgroups is to portray them as a threat, against which the ingroup can legitimately defend itself. How hate can be morally justified on the basis of threat, is at the heart of Reicher et al.’s (2008) social identity model of collective hate. The model identifies five steps through which speakers justify and legitimate group hate: (a) defining a cohesive ingroup with clear definitional boundaries (who are we and who is included?); (b) exclusion (who does not belong?); (c) constructing the outgroup as a threat to the ingroup (how do they threaten us, our way of life?); (d) representing the ingroup as virtuous (we are better, superior to them); and (e) celebrating the ingroup’s right to defend itself from the outgroup (we have the right to protect and defend who we are).

Reicher et al. (2008, p. 1327) argue that social psychology has focused almost exclusively on how outgroups are perceived and treated yet how “we define the ingroup is as crucial, if not more crucial . . . in generating hatred.” Hate speech flourishes in communicative environments where the audience identifies strongly with the common ingroup identity provided by the speaker, perceives this shared ingroup identity to be under threat—threatening our values, culture and “way of life”—and thus feels justified to not only reject the outgroup, but to actively protect and defend the ingroup from these perceived threats. In extreme cases such as genocide, eliminating the outgroup is seen as virtuous, indeed, a moral duty, to preserve and maintain the ingroup’s existence. In this way, atrocious acts become celebrated as the right thing to do. Indeed, when far-right leaders have been prosecuted for hate speech, they have projected themselves as martyrs who have stood up for freedom of speech and the rights of “ordinary people” against dangerous “others” (Pettersson, 2019). Paradoxically, Jacobs and van Spanje (2020) have found that prosecuting anti-immigration European leaders for hate speech can increase their electoral support. Together, portraying outgroups as either subhuman, and hence as being outside of the sphere of human morality, or as threating constitute two opposite, but complementary strategies to justify hostile actions, including atrocities, against outgroups.

The above social functions of derogatory language (perpetuation of prejudice, maintenance of social hierarchies, and legitimization of hostile actions) all involve the outgroup as the primary target. However, derogatory language—frequently used in intragroup discourse—may also serve ingroup goals, quite independently of the effect it has on outgroups. Two social functions figure prominently among these, namely the establishment and maintenance of ingroup norms and the increase in ingroup cohesion.

In intragroup communication, derogatory language serves to establish and maintain norms and regulate role behaviors that are considered socially appropriate. A prominent example are homophobic epithets such as “fag” or “poofter,” that are regularly used to insult gender non-conforming men and that children, in many cultures, use as a generic insult long before they have any notion of sexual orientation. According to Plummer (2001), these epithets are initially used with very broad meanings (being feminine, weak, soft, acting like a girl, being studious and artistic, being a loser, not confirming to peer expectations, showing low peer group loyalty) that define male gender norms and that vividly discourage deviation from these norms. Given that they are experienced as particularly offensive, such homophobic epithets (different from neutral category labels) set the stage for subsequent compliance to masculinity norms (Carnaghi et al., 2011).

Finally, disparaging language is part of a collective narrative that fosters ingroup cohesion and that is used as a sign of comradeship and shared group membership. It connects like-minded people in defense of a common identity and in the pursuit of a common cause (Douglas, 2007). Together, the literature suggests that derogatory and hateful language serves multiple purposes and fulfill distinct social functions in the construction and maintenance of intergroup relations.

The Consequences of Derogatory Language and Hate Speech

The effects of derogatory language and hate speech are often disregarded as “only words.” Although spreading hate does not always lead to violence, Allport (1954/1979, p. 57) emphasized how violence is always preceded by verbal hostility. Allport argues this eloquently in the quote below:

Although most barking (antilocution) does not lead to biting, yet there is never a bite without previous barking. Fully seventy years of political anti-Semitism of the verbal order preceded the discriminatory Nuremberg Laws passed by the Hitler regime. Soon after these laws were passed the violent program of extermination began. Here we see the not infrequent progression: antilocution > discrimination >. . .violence.

There is much to be concerned, therefore, about the increasing proliferation of uncivil language and, in particular, of hate speech on digital/social media. Narratives of conspiracy and hate can spread easily and quickly to all levels of society, including everyday discussions in online forums (e.g., Nortio et al., 2020). The increasing use of social media platforms has coincided with increases in the reported incidence of hate crimes throughout the world. For example, in 2018/19 there were over 103,379 hate crimes recorded by the police in England and Wales, an increase of 10% on the previous year (British Home Office, 2019). Extremist websites have also proliferated: the first to be launched was stormfront.org in 1995 by White nationalist and former Ku Klux Klan member Donald Blac; by 2010, there was an estimated 8,000 hate sites. Such sites enable the recruitment of members, facilitate the spread of hate and the building of extremist subcultures through private message boards, listservs, and email (Banks, 2010). Hate sites also target other minorities, such as women and LGBTQIA+ people, and have become vehicles through which misogynist and heterosexist discourse are disseminated. It therefore becomes essential to analyze the effects of derogatory language and hate speech both on targets as the direct victims and on bystanders. ²

Effects on Targets

Even though targets of hate speech recount feelings of pain and the knowledge of the potential long-term consequences of these episodes, in their accounts a feeling of resignation is also present, as if the attacks were inevitable; so much so that victims may trivialize the event and not acknowledge it as discrimination (Browne et al., 2011; Greenland et al., 2018). Nonetheless, the toll of verbal derogation is still present on the mental and physical health of victims. Diary studies by Swim and colleagues, for example, evidenced negative affects (fear, anxiety, and/or lower self-esteem) in victims of sexist, racist and heterosexist language and prejudice (e.g., Swim et al., 2009). Research focusing on the experiences of LGBTQIA+ students shows that being a target of heterosexist language and microaggressions is linked to problems within the academic setting as well as physical and mental health issues (see Table 1 for examples).

OPEN IN VIEWER

Table 1.

Effects of Heterosexist Language and Microaggressions on LGBTQIA+ Students: Examples of Correlational Studies.

Type of effect	Effects	References
Relationships within the academic setting	Social exclusion	Woodford et al. (2012)
	Less acceptance and respect	Silverschanz et al. (2008)
	Poorer relationships with faculty
	School avoidance
Academic profit	Academic disengagement	Woodford and Kulick (2015)
	Academic stress	Mathies et al. (2019)
	Lower GPA	Mathies et al. (2019); Woodford and Kulick (2015)
Physical and mental health	Physical symptoms	Woodford et al. (2012)
	Distress	Poteat and Espelage (2007)
	Withdrawal	Poteat and Espelage (2007)
	Anxiety and/or depression	Poteat and Espelage (2007); Silverschanz et al. (2008); Slåtten et al. (2015); Woodford et al. (2018)

Interestingly, it appears that both heterosexual and non-heterosexual victims respond similarly to being subject to homophobic language (Silverschanz et al., 2008) and that, among peers, these responses are linked to the target’s awareness of the harmful, rather than harmless, intentions of the speaker (Slåtten et al., 2015). Derogatory language is also associated with risk behaviors, such as smoking (Ylioja et al., 2018), substance abuse (Silverschanz et al., 2008), and even attempted suicide (Woodford et al., 2018). “Hate speech kills,” write Mullen and Smyth (2004, p. 346), who show that the probability of suicide is greater among immigrant groups targeted by ethnopaulisms.

With regards to other-related emotions, anger is common in targets (e.g., Browne et al., 2011; Swim et al., 2009). Anger toward sexist language and comments, for example, predicts intentions to protest (Paladino et al., 2014), but only when sexist comments are recognized as such. Subtler forms of sexism, such as modern and benevolent sexism, elicit less anger and subsequently less inclination to engage in collective action (Ellemers & Barreto, 2009). Interestingly, in one study, when women were asked to recall episodes of sexist comments, there was no difference between hostile and benevolent sexism in terms of affective reactions and time taken to recover from the event (Bosson et al., 2010). This suggests that blatant hostility only makes a difference when the target is not the direct victim of the discriminatory event.

Reacting to Language Discrimination: Confrontation and Reappropriation

How can targeted minorities cope with verbal insults and hateful comments and react to them? As of today, recent literature has identified two potential strategies of facing verbal discrimination, namely confrontation and reappropriation.

Future Directions

Most Western liberal democracies have enacted legislation against hate speech, the most notable exception being the United States where hate speech is protected under the First Amendment Right to free speech in the American Constitution. In contrast, under European Union Law, hate speech is illegal and concerted efforts have been implemented to regulate such speech across European signatories (Clark & Grech, 2017; European Commission, 2020). Nonetheless, curtailing free speech has become a contested political issue and one that generates heated public debate. Free speech advocates routinely complain that “political correctness” is constraining what people are allowed to say in public and that this constitutes unacceptable censorship (Maass et al., 2013). Competing democratic values are invoked in these debates, around the rights of minorities to be protected from fear and violence, so that they can participate freely in a democratic society, versus individual rights to freedom of expression, with far-right leaders depicting themselves as defending the rights of the majority to speak frankly about their concerns, and anxieties.

To fully understand how derogatory language, and in particular hate speech, has gained traction and social acceptance a more inclusive theoretical and empirical research agenda is required in social psychology, one that combines quantitative and experimental research with qualitative discursive approaches that analyze the situated and interactional use of derogatory language in naturalistic settings. Despite the epistemological and methodological differences of these traditions, several researchers have demonstrated how they can be brought together creatively to produce a nuanced and rich understanding of how inequality and power is produced and legitimated in language (e.g., Reicher et al., 2008).

In particular, we see a number of directions in which future research could evolve. First, it may be useful to develop a multi-dimensional classification of communication settings in which language discrimination occurs. Such a system may, among others, distinguish intra-, inter- versus mixed group settings, verbal versus written language production, private versus public settings, as well as the characteristics of specific media platforms, including the degree of anonymity they provide. Drawing on prior social psychological work on de-individuation, norm compliance and accountability, this may allow more precise predictions about the likelihood of uncivil communication. Such a classification system may also help to understand the role of technology in language production. The very format of social networking services may, to some degree, determine the kind of message that is likely to be produced by its users. For instance, twitter facilitates short and direct messages that are comparably simple, impulsive and often uncivil (Ott, 2017).

Second, given that people live in (social) media bubbles that expose them very selectively to specific types of messages (e.g., via social networks or via indexing through hashtags), we believe that greater attention ought to be paid to local norms and their development and change over time. We agree with Spears et al. (2007) that offensive computer-mediated messages are more likely to reflect group norms than de-individuation processes. Such norms not only define whether derogatory and hate messages are acceptable, but also what kinds of offenses are tolerable (e.g., derogatory vs. objectifying sexist slurs, Fasoli, Carnaghi et al., 2015). Considering the lively controversy surrounding legal interventions to limit hate speech, norms may provide a basis for alternative intervention strategies to counter hate speech, possibly as part of broader media and information literacy training programs (for an overview see Gagliardone et al., 2015).

Third, it may be beneficial to broaden the scope of research on derogatory and hateful language beyond minority groups. In keeping with the definitions of hate speech offered by international organizations such as the Council of Europe or the United Nations, most psychological research has focused on language discrimination of small and/or low status minority groups (e.g., immigrants, women, LGTBQIA+, religious minorities, people with disability). However, there is now increasing evidence that hate speech may also target individuals who, for personal or professional reasons, publicly express opinions and take stances on social, political or economic issues. For instance, journalists have become a prime target of verbal insults and threats, with severe consequences for media freedom (e.g., Clark & Grech, 2017). Just like hate crime (Craig, 2002), hate speech curtails the behavior and freedom of expression of victims (and their colleagues), often leading to self-censorship that poses a threat to the basic principles of democracy (Clark & Grech, 2017). Thus, derogatory or hate speech against public figures such as politicians and journalists deserve a space in social psychological research.

Fourth, future research may draw a clearer line between offensive or hateful language produced by single individuals or by organizations. There is now evidence that a portion of hate speech is strategically produced by organizations with specific political aims in mind, including the intimidation of political opponents, the creation of religious outrage (George, 2016), or the influence of election outcomes (George, 2020). We believe that such deliberate and well-orchestrated use of hate speech for sociopolitical reasons, often spread via political trolls, is qualitatively different from the spontaneous and often impulsive hate messages produced by individuals. Organized (vs. individual) hate speech deserves attention both under a psychological and legal perspective, similar to the way organized crime is seen as distinct from individually committed crimes.

Finally, given the multifaceted phenomenon and its frequent collocation in social media, interdisciplinary research efforts are needed that involve researchers from psychology, media studies and communication, sociolinguistics and law. The advent of big data has also made it increasingly imperative for social psychologists to collaborate with experts in IT and computer science, who are developing algorithms for identifying hate speech, to understand the process through which users are selectively exposed to certain messages, and responding to regulatory requirements to restrict such communication.

Conclusion

Our review suggests that derogatory and hateful language is a widespread, increasing phenomenon that targets mainly minorities and groups with less power (as it maintains social hierarchies and ingroup-outgroup dynamics) and that affects not only victims, but also bystanders and the society at large. We hope that this article paves the way for future multi-method work on derogatory language: how it proliferates, its psychological and social consequences, how victims and bystanders cope with the phenomenon, and how groups and society at large face up to its challenge.