2.1.1 Face and Identity

The lack of exploration of intergroup communication within pragmatics/(im)politeness may be due to several factors. One of them being that, in general, most research has not necessarily focused on identity but on face. Face, the inspiration behind the core concept in Reference Brown and LevinsonBrown and Levinson’s (1987) framework, was defined by Goffman as “the positive social value a person effectively claims for himself by the line others assume he has taken during a particular contact. Face is an image of self delineated in terms of approved social attributes” (1955/1967: 5). As Reference Garcés-Conejos BlitvichGarcés-Conejos Blitvich (2013) argued, Goffman conceptualized face as being tied to a line (a role, an identity).Footnote ² However, Brown and Levinson’s construal of face altered its essence, as they separated face from line and presented it as a primarily cognitive construct possessed by a rational Model Person. This focus on individualism is one of the reasons why their politeness theory drew some of the earliest critiques from scholars from collectivism-oriented cultures (Reference MatsumotoMatsumoto, 1988; Reference NwoyeNwoye, 1992) and the argument behind taking into consideration discernment along with strategic politeness (Reference Jankovic and LudwigIde, 1989). It has also been a deterrent for advancing the study of intergroup communication.

Relatedly, another reason leading to a dearth of research on intergroup communication can be related to how face is further conceptualized, more specifically as to whether it is taken to be accrued over time, which means it can be attached to groups (Sifianou 2011; Reference Wang and Spencer-OateyWang & Spencer-Oatey, 2015) or is emergent in interaction (Reference ArundaleArundale, 1999; Reference HaughHaugh, 2007), in which case it cannot be shared by a group (since it is unlikely that all members would be interacting concurrently). Although this is a key conceptual point that needs to be tackled to theorize intergroup communication (see Reference Haugh, Kádár and MillsHaugh et al., 2013 discussion of relational histories), it has not received significant attention within (im)politeness research. Some exceptions are Reference Hatfield and HahnHatfield and Hahn (2014) and, more recently, O’Driscoll (2017: 104) who concluded that “it is again not clear how group face … is to be distinguished from (the self-evidently valid concept of) a group’s reputation / image / identity.” However, he went on to cite Reference SorellSpencer-Oatey’s (2005) “social identity face,” the desire for acknowledgment of one’s social identities or roles (e.g., group leader, valued customer), as appearing to have distinct meaning.

In my view, the problems that face, as understood in the literature discussed, poses in relation to groups can be bypassed by resorting, instead, to a closely related concept: identity or the social positioning of self and other (Reference Bucholtz and HallBucholtz & Hall, 2005: 586). Although face and identity are both significantly related to the construction and presentation of self (Reference GoffmanGoffman, 1955/1967), they have – for the most part – been the cornerstones of different research traditions. One fundamental difference between (im)politeness and identity scholarship is that the latter has devoted considerable thought to whether identity is accrued or emergent in interaction. There seems to be quite universal consensus in post-structural/discursive approaches to identity in this respect: identity is discursive, intersubjectively co-constructed, emergent in interaction (for instance, it involves those very ephemeral subject positions that we occupy in unfolding discourse and that make up stance), but also durable (Reference Bucholtz and HallBucholtz & Hall, 2005).

Identities are durable, not because individuals have essential or primal identities, but because dialogically constructed identities are recreated in multiple local practices where they make sense and create meaning (Reference Holland and LaveHolland & Lave, 2001). Reference ButlerButler (1990) refers to a process of sedimentation whereby people repeatedly draw on resources that gradually build up an appearance of fixed identities. Whereas it is true that certain aspects of person, role and social identities are more stable, even those are always overlapping with each other and constantly changing (Reference Burke and StetsBurke & Stets, 2009). Barker and Galasinski (2001: 31) sum up this well when they state: “Identities are both unstable and temporarily stabilized by social practice and regular, predictable behaviour.” From this unstable/stabilized, emergent/durable perspective, it is conceptually quite unproblematic to tie identity (and face, as we will see later in this section) to groups.

In the mid-to-late 2000s, identity started to be presented alongside face in the definitions of (im)politeness. Scholars also began enquiring about its relationship with face and introducing (im)politeness studies to the frameworks developed for the analysis of identity, arguing for their relevance to (im)politeness research (see among others Reference Spencer-OateySpencer-Oatey, 2007; Locher, 2008; Reference Garcés-Conejos BlitvichGarcés-Conejos Blitvich, 2009, 2013; Garcés-Conejos Blitvich & Sifianou, 2017; Garcés-Conejos Blitvich & Georgakopoulou, 2021). Those scholars who argued for a more multi-disciplinary approach to the study of (im)politeness phenomena saw identity models as a means to achieve that goal, due to the close relationship between face, identity, and self-presentation. Proposed ways of advancing the field can be summarized as follows:

1. a. (Im)politeness manifestations/assessments can be tied to identity (co)construction, not just to face.

2. b. Identity and face are inseparable, as they co-constitute each other.

3. c. (Im)politeness can be analyzed as an index in identity construction.

4. d. Models developed for the analysis of identity construction can be fruitfully applied to the study of (im)politeness (Garcés-Conejos Blitvich & Sifianou, 2017: 238).

Since then, identity has either replaced face as a core concept in (im)politeness research or has been seen as essential to grasp what face entails, as face needs vary depending on the identity being co-constructed by an agent engaged in a specific practice: that is, an agent’s face needs as a mother differ substantially from those of the same agent’s face as a surgeon, or as a bridge player. In this practice-based approach, face and identity are viewed as difficult to tease apart in interaction and as co-constituting each other (Reference JosephJoseph, 2013; Reference MillerMiller, 2013). Crucially, agents engaged in practice may claim or be attributed either an individual or a social identity and its concomitant face needs and thus group (social) identity is seen, from this perspective, as unproblematically tied to face. Indeed, the concept of self-esteem so closely connected to face is foremost in the conceptualization of social identity as discussed by Reference SwalesTajfel (1979) who proposed that the groups (e.g., social class, family, and football team) which people belonged to are an important source of pride and self-worth. Relatedly, Reference GoffmanGoffman (1959: 85) reconceptualized his framework in making the team, rather than the individual, the basic unit of the interactional order. The goal of the team is to help teammates maintain the line they have selected. In this sense, teams/groups provide us with a sense of belonging to the social world.

2.1.2 Collective Intentionality and Action

Yet, another motive behind the inattention toward intergroup communication in pragmatics/(im)politeness may have to do with the major role that intentionality has played in pragmatic understandings of the conveyance and interpretation of meaning (especially in the Cognitive-Philosophical foundational models and theories of the field, which primed speaker-meaning, such as Speech Act Theory, Grice’s Cooperative Principle, etc.).

Although its role as an a priori concept has been questioned (Reference HaughHaugh, 2008a), most pragmatic scholars still see intention as playing a (varying) influential role in communication (Culpeper, 2011; different contributions to Journal of Pragmatics 179, 2021). As Haugh (2008b) discussed, traditional views relate intentionality to individual utterances/speakers, but a higher-order intention would need to be attached to larger stretches of discourse. In addition, if intentionality is understood as co-constructed among participants in interaction, it may be appropriate to consider a we-intention as also being applicable (Reference HaughHaugh, 2008a; Haugh & Jaszczolt, 2012). However, Haugh (2008b) saw we-intention, as described in the then extant literature, as static and not adequately capturing the emergent nature of inferential work underpinning cooperative activities, such as conversation.

An important, for this Element, related question is whether we should appeal to an ontology of group agents, i.e., when a team plays, is the team just made up of individuals coordinating in search of a common goal or is the team itself that plays? This question has been tackled mostly from the point of view of philosophy and resulted into two opposing camps: individualism and collectivism. Individualism understands and explains both individual action and social ontology resorting to individuals and their relations and interactions. For its part, collectivism argues that there are genuinely emergent social phenomena, such as social objects (i.e., groups), states, facts, events, and processes (Reference TuomelaTuomela, 2013).

When dealing with shared intentions, more specifically, two opposing views are upheld: (i) those who spouse reductive views of shared intention, that is, attributing an intention to us – as in a group – can be undertaken by resorting to concepts related to individual action/intention, and (ii) those who disagree with this position and hold non-reductive views and understand we-intentions as involving an irreducible we-mode. The we-mode approach (Reference TuomelaTuomela, 2007, 2011) is predicated on the crucial distinction between acting as a group member guided by the group ethos versus acting as a private person.

In more recent work, Tuomela (2017: 16) reviewed some of the central accounts of non-reductive views, namely those espoused by Reference GilbertGilbert (1989), Reference SearleSearle (2010) and himself (Reference TuomelaTuomela, 2013), and described them along the following lines:

Despite some commonalities, Tuomela criticized Gilbert’s plural subject theory arguing that it is circular and Searle’s argumentation of we-intentions as underdeveloped. His own proposal – based on I-mode/we-mode of sociability that takes joint intention, interdependent member intentions (we-intentions) are expressible by “we will do x together” – was, in turn, also criticized as circular (Schweikard & Bernhard Schmid, 2021). Further, even among those who agree that collectives can be the subject of intentional state ascription, there is manifest dissent: Reference GilbertGilbert (1989) and Reference ThorntonTollefsen (2002, 2015) argue that it is appropriate to attribute a range of intentional states including beliefs to groups. Others, Reference TuomelaTuomela (2004) and Reference WrayWray (2001), disagree with this position as groups, in their view, can accept propositions but cannot be believers.

From this discussion, it sounds as if all proposals that stem from philosophical explanations perhaps leave us with more questions than answers (Reference Watson-Jones and LegareWeir, 2014), which may also apply to general approaches to individualistic intention within pragmatics, mostly stemming from a philosophy of language tradition. The role of intentionality in interpersonal communication has proven to be difficult to ascertain or agree upon, and group intentionality is certainly not an unproblematic notion either. However, the post-hoc reconstruction of intentionality via inferences and attributions is key in communication, be it interpersonal or intergroup in nature. Therefore, understanding groupness and the attribution of goals and intentions to groups are essential for humans and, not surprisingly, they have played a major role in the development of the species’ social cognition (Reference Tomasello and RakoczyTomasello & Rakoczy, 2003).

Certainly, knowledge about groups carries important and, sometimes, vital information and impacts attitudes and behaviors (ingroup bias, implicit bias, moral obligations); therefore, it seems essential for humans to be able to categorize others as non/members. Apparently, this is something that children as young as 3 can easily do, using mutual intentions (i.e., the general agreement of individuals that they belong to a group, Reference NorrisNoyes & Dunham, 2017: 34) as a guiding principle, closely connected to perceived common goals (Reference Straka, Stanaland, Tomasello and GaitherStraka et al., 2021). Once groups are constituted socially, children seem strongly inclined to believe that groupness sanctions specific ways of relating and carries with it patterns of behavior, association, and moral obligations (Reference Noyes and DunhamNoyes & Dunham, 2017: 141).

Thus, how people are transformed from a random collection of individuals into a group, a social unit, oriented to joint action is, according to Jankovic and Ludwig (2017: 1–2) the way they think about what they are doing together; this way of thinking constitutes the focus of study of collective intentionalityFootnote ³ (CI) which can be defined as:

The authors see CI as crucial to understanding the nature and structure of social reality. In this they coincide, among many others, with Reference Tomasello and RakoczyTomasello and Rakoczy (2003), Reference Tomasello, Melis, Tennie, Wyman and HerrmannTomasello et al. (2012), and Reference Straka, Stanaland, Tomasello and GaitherStraka et al. (2021) who claim that CI is what makes human cognition unique. More specifically, Tomasello and Rakoczy (2003: 123, 143) argue that CI related skills underlie children’s cultural learning and linguistic communication and enable the comprehension of cultural institutions based on collective beliefs and practices. This is important as CI is, therefore, seen as the foundation for social norms, which emanate not from individual but group opinion. As a result, conforming to social norms displays group identity. In the same way not abiding by them signals disdain for the group. That is why, as is the case with CC, “punishment for the laggard needs to be by the group as whole – so that when an individual enforces a social norm, she is doing so, in effect, as an emissary of the group as whole” (Reference Tomasello, Melis, Tennie, Wyman and HerrmannTomasello et al., 2012: 683). The norms that define the group and its goals and assign roles to its members, that is, the group’s CI, are in turn internalized by those members, according to Reference TomaselloTomasello (2018: 74) as “‘an objective morality’Footnote ⁴ in which everyone knew immediately the difference between right and wrong as determined by the group’s set of cultural practices.”

These claims are consequential, not only for this Element and the understanding of cancelation practices, but also for pragmatics. Norms, the quintessential pillars of research in many pragmatic subfields ((im)politeness among them) are the outcome of groups’ collective intentionality. Therefore, a pragmatics/(im)politeness of intergroup communication needs to incorporate (social) identity and collective intentionality as key theoretical tenets. Social identity – an individual’s cognitive, moral, and emotional connection with a broader community, category, practice, and institution (Reference Polletta and JasperPolletta & Jasper, 2001) – is essentially connected to collective intentionality and both are fundamental to understanding the type of collective mobilization involved in cancelation.

This incorporation is not complicated if one takes a discursive approach to pragmatics/(im)politeness). Indeed, as the field underwent a discursive turn (Barron & Schneider, 2014; Reference Locher and WattsLocher & Watts, 2005), identity gained prominence. This is unsurprising as, according to discourse theorists (Reference FaircloughFairclough, 2003; Reference GeeGee, 2005), the construction of individual/social(group) identity is at the core of discourses, and the latter fundamentally involves collective intentionality. From this perspective, there are no apparent obstacles for pragmatics to deal with interpersonal and intergroup communication. In addition, as will be discussed, discourse pragmatics pays close attention to the meso-level, the level of group practices.

2.2.1 Genres and the Meso-Level

As emphasized throughout this Element, meso-level practices are essentially related to groups and, thus, merge in well with the intergroup perspective advocated here. In taking this step, as pragmatics becomes discourse-centered, pragmaticians of different persuasions can draw interdisciplinarily from linguistic and rhetoric fields that have made meso-level practices one of their main objects of research, such as Systemic Functional Linguistics, English for Specific Purposes (ESP), and Rhetorical Genre Studies.

It is the last two approaches I mostly draw from in my own understanding of genre. Regarding the first, in its foundational paper, Reference MillerMiller (1984) approached genre as social action and defined genres as “typified rhetorical actions” that respond to recurring situations and become instantiated in groups’ behaviors. Miller argued that “a rhetorical sound definition of genre must be centered not on the substance or the form of discourse, but on the action, it is used to accomplish” (151), which – in Miller’s view – makes them pragmatic in nature. Bazerman (1997: 19), a key contributor to this approach, summarizes it well:

Further, genres are stabilized-for-now but constantly-in-flux sites of social and ideological action (Reference Berkenkotter and HuckinBerkenkotter & Huckin, 1993) and thus recognizable by members of communities. As Bazerman (1994: 100) pointed out, “[w]ithout a shared sense of genre, others would not know what kind of thing we are doing.” Importantly, as the meso-level realization of discourses, genres open certain positions, roles, and relationships, since identity is always at the core of discourses/genres (Bazerman, 1994; Blommaert, 2017a; Reference FaircloughFairclough, 2003; Reference GeeGee, 2005). Relevant to our discussion, knowledge of genres through socialization practices includes rhetorical expertise, dynamic knowledge of processes and audience’s expectations (regarding norms of behavior, for instance) (Reference TardyTardy, 2009)

Reference MillerMiller (2015), when pondering on the influence of her foundational views, argued that genre continues to be a useful concept because it connects our experiences to our sense of the past and the future, makes recurrent patterns significant, and very importantly for this Element, it characterizes communities by offering modes of engagement through joint action and uptake.Footnote ⁶

Within ESP, one of the most influential definitions of genre was provided by Reference SwalesSwales (1990: 58): “A genre comprises a class of communicative events, the members of which share some set of communicative purposes. These purposes are recognized by the expert members of the parent discourse community, and thereby constitute the rationale for the genre.” Swales’ model is referred to as the three-level genre model and provides a useful template for the actual analysis of genres.

Communicative purpose/goal emerges as the main rationale for the genre and triggers a move structure. Moves are understood as discoursal or rhetorical units that perform a coherent communicative function in a written or spoken discourse (Reference SwalesSwales, 2004: 228), where some moves are essential/obligatory and others are nonobligatory. For instance, in resumes, one is expected to include a part (move) on education and another on professional experience, whereas a move on hobbies is not necessarily expected. For their part, rhetorical strategies refer to the different semiotic modes (lexico-grammatical structures, jargon, layout, font, register, among others) used at the micro-level to realize each move and carry out the general purpose of the genre, and are, therefore, audience-designed. To continue with the same example, a resume follows a well-established and recognizable layout and uses professional-looking fonts; it should be formal in register, use field-related jargon, be short (circa two pages for corporate jobs), and so on In this respect, as mentioned, genres are stabilized for now and socio-rhetorical; therefore, move structure and rhetorical strategies are also in flux as communities’ conventions and expectations change or influential trailblazers fundamentally alter the genre in a bottom-up fashion.

Figure 2 Swales’ model.

Relatedly, the advent of digital technologies partly changed, among many others, our conceptualization of texts and genres, as traversals facilitated by hyperlinks created significantly different expectations regarding the finiteness of texts and the high hybridization of genres (Reference BhatiaBhatia, 2015; Reference LemkeLemke, 2002). Although mentioned in relation to digital mediation, as Reference BhatiaBhatia (2015) argued, hybridization is a constant across genres, despite scholarship having tended to focus on “pure” genres. For instance, antitobacco bills merge and enact legal and health discourses. However, genres do not only merge and influence one another, they relate to each other in a myriad of other ways, as it is not often the case that a given activity can be carried out by just one single genre. Note how a political campaign or a college application involve many genres practices coming together to achieve a specific goal: being (re)elected and being (re)admitted, respectively; those genre practices include, to name a few, from phone calls and letters to potential voters/donors, flyers, rallies, fundraisers, and so on, in the case of campaigns, and application forms, statements of purpose, official transcripts, letters of recommendation, meetings with academic advisors and finance counselors, and so on in the case of applications.

As a result, scholars have devised conceptualizations regarding how genres come together, such as in genre sets (full range of the kind of texts that one person uses to fill out one side of a multiple-person interactions, Bazerman, 1994), genre systems (“interrelated genres that interact with each other in specific settings,” Bazerman, 1994: 97), genre repertoires (“The set of genres that are routinely enacted by the members of a community,” Reference Orlikowski and YatesOrlikowski & Yates, 1994: 542), genre sequences (a similar concept to genre systems, Reference SwalesSwales, 2004), genre colonies (a term developed to describe how genres move from one activity system to another to create new clusters of genres, Bhatia, 2002), and genre ecologies (a contingent, interacting, interdependent system or network of genres, as distinct from a series of genres with a specifically sequential series of uptake, Reference Spinuzzi and ZachrySpinuzzi & Zachry, 2000).Footnote ⁷

It is the last concept, genre ecologies, that I find particularly useful. Applying this model to the data, I will argue that cancelation is carried out by complex genre ecologies (Section 5.2).

4.3.1 Nichols

In August 2021. Nichols parted ways with ESPN (Entertainment and Sports Programming NetworkFootnote ¹²) and her basketball program The Jump was canceled. This happened after the New York TimesFootnote ¹³ reported, on July 4 of that same year, on derogatory comments Nichols had made regarding one of her colleagues, Maria Taylor (who is African American). In a 2020 conversation with a colleague from her room at the NBA’s Walt Disney World bubble – unaware that her video camera was on, and she was being recorded to an ESPN server – Nichols said that Taylor had been selected to host the 2020 NBA finals instead of her because ESPN was feeling pressure on diversity.

This conversation, available to those ESPN employees with access to the server, was leaked to the Times a year later, although ESPN had long known about it, which set in motion hers and her show’s cancelation. “I wish Maria Taylor all the success in the world – she covers football, she covers basketball,” Nichols is heard saying on the tape. “If you need to give her more things to do because you are feeling pressure about your crappy longtime record on diversity – which, by the way, I know personally from the female side of it – like, go for it. Just find it somewhere else. You are not going to find it from me or taking my things away.” Nichols later apologized on The Jump, saying she respected her ESPN colleagues and noting that she was “deeply, deeply sorry” for “disappointing those I hurt, particularly Maria Taylor.”

4.3.2 DeGeneres

Nichols’s cancelation trigger is quite straightforward, DeGeneres’, however, is more multilayered, being attributed to various factors. Many point to Dakota Johnson’s two awkward interviews with DeGeneres as the beginning of the end. In 2018, DeGeneres told Johnson she was offended because the latter had not invited her to her birthday party, even though DeGeneres had invited Johnson and her boyfriend (Chris Martin) to hers. When Johnson returned to the show in 2019, DeGeneres brought up not having been invited to Johnson’s most recent birthday party, but Johnson promptly retorted that DeGeneres’s statement was untrue, which was confirmed by one of the show’s producers: DeGeneres had indeed been invited but had not attended. It soon came to light that, the day after the party, Louisiana-born DeGeneres – who is openly gay – had been seen cozying up to former President George W. Bush at a Dallas Cowboys’ game; this caused a big uproar due to Bush’s anti-LGBTQ policies and perceived failure to act during hurricane Katrina aftermath – which mostly affected the poor, amongst those many people of color – in DeGeneres home state.

Several other awkward interviews soon started to surface, which further pointed to a mismatch between DeGeneres’s public “queen of nice” façade versus her “true” not so nice off-air personality. And then, in early 2020, comedian Kevin T. Porter tweeted a message mocking DeGeneres’s purported kindness and inviting people to share their unkind stories with her, whom he described as one of the meanest people on earth, further undermining DeGeneres’s public image. Porter pledged to match each story with $2 and to donate the full amount to a food bank. With over 2,600 replies at data collection point, Porter later said he had received about 300 stories detailing DeGeneres perceived meanness and ended up donating $600. Amid this climate, current and former employees of DeGeneres’s show made public their complaints about the show’s toxic atmosphere – for instance, a Black woman complained she had been the target of many “microaggressions” – and unfair treatment of employees during the COVID-19 pandemic.Footnote ¹⁴ Although related complaints were mostly directed at the show’s producers, there was quite widespread consensus that DeGeneres was ultimately responsible for what happened on her show. She apologized both in a letter addressed to her staff and on air, to her audience, during the first show of the 18th season of her show. Despite her seemingly heartfelt apologies, after 19 seasons, the final episode of DeGeneres’ daytime talk show aired on Thursday, May 26, 2022, on NBC.

4.3.3 Cheney

Liz Cheney, a Republican, served as Representative for the state of Wyoming. Cheney’s voting record shows her to have been one of the most stalwart conservative members of the House; in this, she closely followed in the footsteps of her own father, Dick B. Cheney, who served as President George W. Bush’s Vice President for two consecutive terms (2001–2009) and is often viewed as the most powerful Vice President in American history. Congresswoman Cheney was not doing too badly herself. In 2016, she was chosen House Republican Conference Chair, which effectively made her the number 3 Republican in the chamber.

All this was to change, however. In the aftermath of the January 6, 2021 insurrection and during the second impeachment trial of former President Donald Trump, Cheney was one of ten Republican representatives who voted along with Democrats to impeach him. The dissidents would come to face dire consequences. Even before the impeachment formally ended, Wyoming Republicans overwhelmingly censured Cheney for her vote against Trump. A few months later, on May 12, 2021, in a voice vote behind closed doors, the House Republicans ousted Cheney from her leadership position. As he left the House, right after the vote, Congressman Ken Buck (Republican, Colorado) stated “Liz Cheney was canceled today for speaking her mind and disagreeing with the narrative that President Trump has put forth.”Footnote ¹⁵

Cheney’s position within the GOP did not improve when she agreed to serve, as one of two minority members, on the Select Committee to investigate the January 6th attack on the US Capitol. Concerted efforts, ostensibly led by Trump, were put in place to derail her reelection bid, held in August 2022. Despite repeatedly denouncing the Big Lie (Trump’s contention that he, not Biden, had won the presidential elections) and, thus, antagonizing her party’s majority, Congresswoman Cheney was nonetheless hopeful that her strong conservative record would matter more than anything else for Wyoming voters (a state where Trump obtained 70% of the vote, a higher share than anywhere else in the country). Her hopes did not materialize, and – after a fierce campaign – Cheney lost to Trump-endorsed Harriet Hageman in the primaries. Hageman went on to secure Wyoming’s only congressional seat.

Far from apologizing, Cheney has made her rebuke of Trumpism her brand and the platform on which she is rumored to perhaps run for President. Relatedly, in her concession speech, after losing the 2022 primaries, Cheney stated: “Two years ago, I won with 73% of the vote. I could easily have done so again. But it would have required that I go along with Trump’s lies about the 2020 election. That was a path I would not take. No House seat is more important than the principles we are all sworn to protect.”Footnote ¹⁶

5.1.1 Data Sampling

The understanding of the genre practices involved in cancelation, used as background 31 cases of cancelation included in the reference CC Corpus (309,246 words and 256 videos) out of which three cases – described earlier, that is, Cheney, DeGeneres, and Nichols – were selected for detailed analysis.

All three cases and corresponding documentation were freely accessible in the public domain. Currently, public discourse refers to political or social debates found in newspapers, magazines, television, films, radio, music, and web-mediated forums (Reference MarlowMarlow, 2017). Therefore, the research design based on related data did not need Institutional Review Approval.

5.1.2 Theoretical Framework and Procedure of Analysis

To present a nuanced description of cancelation, this level of the analysis applied the tenets of, among others, Rhetorical Genre Theory that views genre as social action (Reference MillerMiller, 1984; Bazerman, 1997, among others) and ESP (Reference Straka, Stanaland, Tomasello and GaitherSwales, 1990) to the data. More specifically, I draw from Reference Spinuzzi and ZachrySpinuzzi and Zachry’s (2000) conceptualization of genre ecologies as an open-system framework to tackle the complexity of cancelation practices. A genre ecology includes an “interrelated group of genres (artifact types and the interpretive habits that have developed around them) used to jointly accomplish complex objectives” (172). Although prior research has shown that genres, as discussed, are not static but dynamic and fuzzy, the authors argue that to account for variation “a more robust ecological perspective is required, one that accounts for the dynamism and interconnectedness of genres” (173).

Genre ecologies are thus characterized by contingency, decentralization, and stability. By contingency, the authors refer to uncertainty in the sense that the relationship among genres in each ecology involves, “complex, opportunistic, sometimes risky coordination among genres that are made by people who are trying to accomplish certain things” (Reference Spinuzzi and ZachrySpinuzzi & Zachry, 2000: 173). Decentralization, for its part, refers to usability, design, and intention being spread across the ecology of genres, not self-contained within individual genres. Finally, stability is seen as the tendency to make the interconnections among the different genres in the ecology conventional and official; therefore, “just as genres themselves are relatively stable while still being dynamic … genre ecologies achieve relative stability – a dynamic equilibrium – over time. In relatively stable genre ecologies, certain connections among genres become commonplace: groups of users tend to use and interconnect genres in quite similar ways” (175). What is also interesting about the genre ecology conceptualization is that it is considered open-ended as users can import genres into the ecology in an ad hoc manner to help achieve its overarching goals (180).

Regarding procedure, the meso-level analysis involved the tracking, both on legacy and social media, of the specific individual’s cancelation from its inception to its outcome and thus considering and carefully documenting all the genre practices (both on and offline) needed to achieve each cancelation.

6.1.1 Data Sampling and Selection

An analytic corpus comprising circa 80,000 words, 2,220 comments was obtained from the reference CC corpus for this part of the analysis. These data were extracted from online interactions triggered by articles, social media videos, and posts related to the three cases described (roughly 740 comments per case). The first 300 comments posted (150 from the original publication and 150 from its reposting to Facebook in those cases where this applies; Table 1), responses, and metadata were manually downloaded with the “sort by” settings set to the “All comments” option. The analytic corpus was then cleaned, and comments were anonymized and numbered. Some were discarded for being off-topic or posted more than once. That brought the number to the circa 740 comments per cancelation case mentioned.

The main criteria for data selection, inspired by polymedia (Reference AndroutsopoulosAndroutsopoulos, 2021) involved visibility and balancing of different ideological positions on the issue at hand. Therefore, Facebook pages were given preference. At the time of writing, Facebook is the top social media platform (2.99 billion monthly active users, followed by YouTube with 2.2 billion) and has the most spread out generational and gender demographics of all social media. For instance, Twitter has only 436 million monthly active users many of those being in the 25–34-year bracket (38.5%) and mostly maleFootnote ¹⁷ (56.4%). When the original posts were shared by news and advocacy organizations, those that represented the breadth of the political ideological spectrum as well as more personal sources such as the cancelees’ profile pages were selected (Table 1).

Regarding ethical considerations, all sources included in the corpus had their privacy settings set to public, which means information posted to them is visible to anyone who accesses the page and is thus in the public domain. Therefore, the data and research design were exempt from Institutional Review Board approval. Nonetheless, and in compliance with the Association of Internet Researchers’ ethical recommendations for best practices,Footnote ¹⁸ all potential personal identifiers were removed. For the most part, comments will be discussed in aggregate form – that is, results will be compiled and summarized – and those that are used as illustrations may also be slightly rephrased to further avoid personal identification. In the discussion, comments will appear in their original form, including any spelling, grammar, and punctuation idiosyncrasies.

6.1.2 Theoretical Framework and Procedure of Analysis

The general approach to the data is netnographic – and thus qualitative and non-media centric – and aims to offer a thick description (Reference GeertzGeertz, 1973) of the online cancelation practices of light groupings. More specifically, it constitutes a descriptive netnography (Reference Addeo, Delli Paoli, Esposito and BolcatoAddeo et al., 2019) which means the analysis is theoretically driven with the aim of filling a gap in the literature and providing detailed information about a less researched phenomenon. My role within the light group was that of a lurker,Footnote ¹⁹ as participants were not aware of my presence or research aims. This covert form of netnography allows not only for an unintrusive, invisible exploration and collection of natural data during which members do not modify their behavior in ways related to the observer’s paradox but also for the unpredictability of results associated with this type of analysis.

In consonance with the tenets of netnography, the analytic corpus underwent a thematic analysis with the help of NVivo 1.6.1, designed for qualitative researchers working with very rich text-based and/or multimedia information, where deep levels of analysis are required. Further, thematic analysis (Reference Fereday and Muir-CochraneFereday & Muir-Cochrane, 2006) offers an accessible and theoretically flexible approach to analyzing data. It involves recording or identifying passages of text or images that are linked by a common theme or idea. Themes are patterns across data sets that are important to the description of a phenomenon and are associated with (a) specific research question(s). In addition, and importantly for this study as described below, NVivo’s matrix coding query option facilitates relevant data linking, which provides further insights into networks of information.

The coding of the corpus was carried out by the author and her research assistant and took a second-order approach (Reference Wang and Spencer-OateyWatts, 2003), in the sense that it reflects the analysts’ views regarding the phenomena scrutinized. Nonetheless, and due to its ethnographic and discursive ethos, the analysis also strove to reflect participants’ own assessments of the interactions. The multilayered nature of the analysis, from a theoretical standpoint, and the fact that each cancelation, although similar in many respects, required the development of case-specific codes led to lengthy discussions (over ten hours) during which we first familiarized ourselves with the data, assigned and labeled preliminary, parent codes, which were selected based on having the most analytic value and were sometimes later modified, applying a hermeneutic circle, as we developed a deeper understanding of the data set. A hierarchal structure was then created by adding clusters of child codes under each parent code as patterns/themes in the corpus emerged; this process continued until coding saturation was achieved. Although parent codes remained quite stable during the coding process, some were eliminated at a later stage due to lesser relevance. Often, more than one recurrent theme came up in a comment, in which case said comment was coded for different themes. However, each comment was coded just once for each theme, even if the theme appeared multiple times within the same comment. For instance, every comment was coded for different types of (im)politeness present; however, each type was only coded once, even if there were more cases of the same (im)politeness type present. Detailed notes about our field observations were taken and shared via NVivo’s annotations, and weekly meetings were held to compare the coding results and to resolve any intercoder differences. Intercoder reliability remained high (above 96.5%) during the entire process.

Regarding the creation of the coding scheme, and to address the third research question, the analysis drew from a range of interdisciplinary sources. Notions of identity construction, claims, attributions, and (non)verification were central and applied following the frameworks and conceptualizations proposed by Reference Bucholtz and HallBucholtz and Hall (2005) and Reference JosephJoseph (2013). Since cancelation had been critically related to morality (Section 4.2), the influential work by social psychologist Jonathan Haidt (and colleagues) underlaid the interpretation of the identity co-construction process. Crucial to this understanding was Moral Foundations Theory (Reference HaidtHaidt, 2012). Drawing from evolutionary psychology to explain their origin, the theory proposes that several innate and universally available psychological systems are the foundations of “intuitive ethics.” These five foundations are summarized in Figure 7.

Figure 7 Moral foundations,

In addition, Haidt et al.Footnote ²⁰ believe that there are several other good candidates for “foundationhood,” such as:

Also due to the very close connections established between CC and emotions (Section 4.2), the analysis explored the data to identify patterns of moral emotions expressed when reacting to and evaluating the cancelee’s or the ultimate canceler’s behavior. According to Haidt (2003: 853), moral emotions are those “that are linked to the interests or welfare either of society as a whole or at least of persons other than the judge or agent.” Applying Haidt’s classification, both parent and child codes were created for these moral emotions “families” (Figure 8):

Figure 8 Moral emotions

In the case of suffering emotions and following Haidt (2003) who argued that empathy is not a moral emotion and, for their part, compassion and sympathy are hard to tease apart, all suffering emotions were merged into a single category: compassion/sympathy.

CC is often portrayed as highly uncivil. As discussed, online incivility includes – among others – many different forms of language aggression. Within pragmatics, language aggression has mostly been studied from the point of view of impoliteness models, where impoliteness is generally understood as behavior that causes offense. Impoliteness is always situated, that is, it is an attitude toward specific behaviors occurring in specific contexts when there is a mismatch between these behaviors and how one would expect/want/think they ought to be (Culpeper, 2011: 254). The analysis of aggression was based on Culpeper’s (2016: 441) classification of impoliteness, and child codes were created accordingly. Additionally, in a bottom-up fashion, the following impoliteness manifestation not included in the extant taxonomy emerged in the analysis and were added as child codes to the codebook:

Importantly, impoliteness is functional and can be used to achieve aims such as forcing (attempting to force) others to succumb to one’s will, what Culpeper (2011: 226) describes as coercive impoliteness; impoliteness can also be used as the targeted display of heightened emotion, mostly anger, implying that the target is to blame for producing such emotional state, that is, affective impoliteness (Culpeper, 2011: 252). These two functions of impoliteness are salient in the corpus, as results show.

In addition, using the classification of strategies provided by Reference Brown and LevinsonBrown and Levinson (1987), politeness manifestations were also analyzed. My point of departure, as mentioned, is aligned with views of (im)politeness as situated (Reference Davies, Merrison and HaughDavies et al., 2011). That clearly diverges from Brown and Levinson’s initial understanding of some speech acts as intrinsically face-threatening. Situated impoliteness means that it is speakers in certain contexts, not linguistic expressions per se, that are assessed as (im)polite. Relatedly, although there has been some discussion regarding the need to do away with the distinction between negative and positive face (Reference BousfieldBousfield, 2008), I agree with Culpeper (2016) and have long argued (Reference Garcés-Conejos BlitvichGarcés-Conejos Blitvich, 2010b) that this distinction (with caveats) remains theoretically useful and reflects central behaviors observed in human interaction. Accordingly, negative face is understood as “the want of very ‘competent adult member’ that his [sic] actions be unimpeded by others” whereas positive face is seen as “the want of every member that his [sic] wants be desirable to at least some others” (Reference Brown and LevinsonBrown & Levinson, 1987: 61).

Codes were also created for the targets of (im)politeness and the expression of moral emotions and how they related to the poster’s in/outgroup(s). Positionings regarding the out/ingroup are highly situated, and vary frequently, and coding decisions need to rely necessarily on sequentiality.

Furthermore, and due to their key role in light group formation, parent/child codes were created for George and Leidner’s (2019) social media affordances and activism classification (Digital Spectator activities [including Metavoicing, Clicktivism/Assertion], Transitional activities [including Political consumerism; Digital petitions; Botivism; E-funding] and Gladiatorial activities [including Data activism; Exposure; Hacktivism]). Finally, notions of geo-semiotics (Scollon & Wong Scollon, 2003) and emotional geographies (Reference Davidson, Bondi and SmithDavidson et al., 2005) were also applied for a deeper understanding of the data.

Once the coding was completed, and with the help of NVivo, several main matrix coding queries were run. These included select parent and child codes and aimed to examine relevant coding intersections. The results of these queries helped unveil further patterns in the data and provided access to the content in which those patterns occurred, which added nuance to the results and informed the discussion.

6.2.1 Identity Claims/Attributions/(Non)verification: Morality and Degradation Ceremonies

Concerning CC generally, the micro-level analysis revealed a more complex picture of cancelation than previously described in the literature, which just refers to two waves of CC led by liberal and conservative groups, respectively. The data showed the simultaneous enacting of both waves, but not necessarily attached to political persuasion, and it further uncovered combinations of cancelation targets.

Three distinct cancelation processes emerged:

1. 1. Straightforward cancelation of cancelee (Cheney, DeGeneres, Nichols)

2. 2. Showing dissent with cancelation/supporting the cancelee: the ultimate canceler (the GOP, NBC/the general public, ESPN) is subject to cancelation (by not watching its shows, discontinuing membership to channels, not voting its candidates/actively supporting opponents, etc.)

3. 3. Both the cancelee and the ultimate canceler are subject to cancelation

Cheney’s and Nichols’ cases involved, to varying degrees, the three cancelation types described above. In DeGeneres’, just the first two were identified. Having been numbered 1–3 does not indicate sequentiality (as they are simultaneous) but the order in which they have been accounted for.

What these three cancelation types share are the centrality of identity construction processes, regarding both self-positioning and attribution (Reference JonesJones, 2017). Engaging in cancelation implies assuming the identity canceler and, thereby, displaying the attributes and collective intentionality associated with the ecology, that is, to expose/denounce cancelees and/or the ultimate cancelers, make them accountable, and deter third parties. Identities are relational (Reference Bucholtz and HallBucholtz & Hall, 2005: 598), so alterity, or otherness, is key here (as subject positions are constructed by qualifying others).

Importantly, in relation to the different cancelation types that emerged, results showed that in 762 occurrences cancelees (Cheney, DeGeneres, and Nichols) were attributed a negative social identity. However, in 783 different occurrences, other cancelers showed dissent by associating cancelees instead with positive aspects but attributing a negative social identity to the ultimate cancelers (GOP, NBC/General Public, ESPN). By doing so, the first groups of cancelers relationally constructed opposing subject positions (Reference AuAu, 2017) for themselves: that is, nonracist, authentic, constitutionalist, pro-free speech. Other cancelers, for their part, by not ratifying the negative identity attributions and showing dissent, ascribed positive characteristics either to the cancelee and/or the ultimate canceler, thus, justifying not engaging in their specific cancelation. This shows that – far from being homogeneous – cancelation processes are fractured and involve a multiplicity of out/ingroups, as illustrated below. In all cases, the negative identities attributed, as predicted by prior research, were strongly related to perceived moral failings (Reference MorganMueller, 2021; Reference NgNorris, 2023).

Regarding Nichols, see Table 2, she and her ingroup undergo cancelation mostly for being considered racist, as shown in Example 1. Racism is here related to the not upholding of the Harm/Care foundation (Reference HaidtHaidt, 2012) regarding the ability to feel the pain of others, especially more vulnerable populations.

On a second process, she is constructed as having positive characteristics: non-racist, truthful, etc., see Example 2:

It is her former employer and ultimate canceler (ESPN) that undergoes cancelation for wokeness, i.e., caving to liberal pressures, as illustrated in Example 3.

ESPN is thus assessed as not upholding the Fairness/Cheating and the Loyalty/Betrayal foundations, for firing Nichols unfairly and not showing loyalty to an outstanding employee.

The third cancelation-type positions both Nichols and ESPN as deserving of cancelation. Nichols for being woke, liberal, a hypocrite. ESPN, in turn, for terminating someone’s employment who just expressed an opinion, wrongly perceived as racist, thus also displaying wokeness, as pointed out in Example 4. Here, we observe a reaction – related to the Liberty/Oppression foundation – to caving to liberal and woke forces that restrict liberty of expression and action:

Occurrences related to cancelation processes 2 and 3 were mostly found in the Fox News Webpage comments, a conservative outlet, and are thus closely aligned with accounts of the second wave of CC.

Cheney’s cancelation also shows complexity and can be ascribed to political persuasion more clearly than Nichols’ and certainly Degeneres’ (Table 3). A first type – related to the first wave of CC albeit not led by liberals, as claimed (Section 4.2) – we find Republicans who attribute to Cheney and her ingroup RINO (Republican in name only), traitor, and disloyal social identities (see Example 5); this group wholeheartedly supports Cheney’s ousting from her leadership position and encourages Wyoming voters not to reelect her (Example 6). Cheney and her ingroup are perceived as having undermined the Loyalty/Betrayal foundation by not supporting Trump, the GOP’s leader, and the Authority/Subversion by not respecting the wishes of those who elected and put her in a leadership position:

The second type here involves both conservatives, independents, and liberals. Conservatives consider themselves, and Cheney, to be constitutionalists and flatly reject Trumpism: it is the Trumpist GOP who should pay dearly for not voting in favor of Trump’s second impeachment and mistreating such a staunch conservative, as argued in Example 7,

Many democrats/independents in this group, who share no ideological common ground with her, admire Cheney’s courage and agree it is the GOP that is at fault and should be penalized accordingly (see Example 8 and Example 9).

All in all, the Trumpist GOP is seen as not sustaining the Loyalty/Betrayal foundation, and having rejected patriotism by putting party, instead of country, first.

The third cancelation-type involves democrats and independents who profoundly dislike/disagree with Cheney’s very conservative voting record – perceived as not upholding the Harm/Care foundation, since many policies she has supported may have hurt women and other minorities – and liken her to her father, Vice-President Cheney. This group thinks, as argued in Example 10, that her recent anti-Trump stance is not enough to redeem Cheney, while also deploring and exposing the actions of the Trumpist GOP, which is assessed (also Example 7) as not maintaining the Loyalty/Betrayal foundation by not abiding by the USA’s constitution:

For DeGeneres (Table 4), a first cancelation-type involves exposing her as fake or inauthentic – that is, putting up a sham nice facade to maintain her popularity, but truly being mean, cruel, fake, sadistic – and (at least) partially responsible for the toxic climate on her show, as argued in Example 11:

This view surmises DeGeneres’ behavior as not upholding either the Fairness/Cheating foundation, by pretending to be different from whom she really is in order to be liked, and the Loyalty/Betrayal for betraying the audience’s confidence. Further, by her alleged mistreatment of employees, DeGeneres is perceived as not supporting the Care/Harm foundation.

In a second cancelation-type, more difficult to relate to a distinct political ideology (although DeGeneres is openly gay and liberal), she is constructed as either genuine or not genuine (but posters argue that one should separate the person from the host) and, importantly, as a victim of societal misogyny, as discussed in Example 12).

Further, as illustrated by Example 13, Ellen is portrayed as human and, as such, prone to making mistakes. Her apologizing publicly and privately (to her employees) garnered her further support, as some may see obtaining an apology as the goal of CC (Reference MorganMueller, 2021).

Here, those who are exposed are CC and the public for believing unsubstantiated information and showering hate on her and NBC for ultimately canceling DeGeneres’ talk show. The former is seen as not sustaining the Harm/Care foundation for destroying DeGeneres’ career based on unproven allegations or by setting irrational standards for her. The ultimate canceler, NBC, is negatively assessed for caving to the authority of the mob and not supporting such an outstanding, highly recognized show (which had won, among others, 34 Daytime Emmy Awards) and, therefore, not upholding the Liberty/Oppression foundation and the Loyalty/Betrayal foundation, respectively.

As shown, in the three cancelation types, the cancelee/ultimate canceler and the group they are associated with are subject to radical identity reduction, oversimplification, and stigmatization (Reference GoffmanGoffman, 1963a), that is, they are defined by a single, negatively valenced attribute: fake, racist, RINO, unconstitutionalist, woke, and so on. This is reminiscent of a degradation ceremony, that is, the communicative work that transforms a person’s entire status and identity into something lower (Reference GarfinkelGarfinkel, 1956). This ceremony is, essentially, quite an aggressive, and necessarily collective, process whereby an individual/group identity is debased and is tied to moral outrage/indignation and public denunciation (21). Cancelers feel justified in publicly shaming, socially excluding, and ostracizing cancelees/ultimate cancelers, because the latter and their ingroup were reduced to mere immoral beings/entities. In this respect, it has been argued that “[o]nline settings reduce empathic distress by representing other people as two-dimensional icons whose suffering is not readily visible” (Reference CrockettCrockett, 2017: 770). This is where, as we will discuss below, the dark side of morality emerges (Reference MillsMonroe & Ashby Plant, 2019). Here, as well, we see in action the three processes described by Tajfel and Turner (1979/2004) involved in the qualifying of others: social categorization, social identification, and social comparison. The authors argued regarding the third that, in being compared, one’s ingroup needed to come out as being better or having some sort of higher status than the outgroup. In the processes under scrutiny, groups compete for higher moral ground and most respected objective morality.

Identity claims/attributions/reduction, (non)verification, and concomitant processes of othering versus affiliation are key to the process of gelling online light groups and the first step toward achieving group agency during digital degradation ceremonies. Results show, and this is an important finding, that online spaces reflect and co-construct offline ideologies, as there is more than one group and more than one ideology at work although both share a similar collective intentionality, cancelation albeit of different targets. Results also show that conservatives can lead cancelation and that cancelation is a more complex process than initially described and that there is more than one “mob” involved in cancelation processes.

6.2.2 (Im)politeness Manifestations

As seen, the processes of social identity claims/ attributions and (non)verification play an essential role in in/outgroup formation. We need to further probe how these very light online out/in groups quickly gel and gain enough entitativity and agency to join or encourage other light/thick groups related to the vast array of contingent genres involved in each cancelation ecology to accomplish its collective intentionality. These key queries were also answered through a detailed micro-analysis of the digital data.

To that end, and due to the key role that (im)politeness plays in ratifying, modulating, and modifying the social relationships associated with certain social identities (Reference JonesJones, 2017), the next part of the analysis focused on (im)politeness manifestations in the corpus. Importantly, CC and related practices, such as online public shaming, are inevitably portrayed as a punitive and aggressive phenomenon that seeks to disqualify cancelees and associated groups for perceived moral failings (Reference GoldmanGoldman, 2015). Further, as described, the corpus is made up of online comments a genre that follows what Thorne and Reinhardt (2008: 562) labeled as “vernacular digital language conventions” and can be often aggressive and use impolite and profane language (Reference ChenChen, 2020); therefore, the presence of impoliteness in the corpus was anticipated. This proved to be the case as the coding process unveiled 4,288 occurrences – 51.7% of impoliteness. Not as anticipated, however, was the almost equally significant presence of politeness with 3,999 occurrences – 48.3% (total occurrences 8,287 – 100%).

To further probe these unexpected results, a coding matrix query was run to check for meaningful intersections between two sets of parent codes [(im)politeness and out/ingroup]. As shown in Tables 5 and 6, a clear pattern emerged: impoliteness was mostly directed intergroup whereas politeness was directed intragroup. In this case, each comment was coded only once for either intergroup (im)politeness or intragroup (im)politeness (totaling 3, 112occurrences – 100%). This means (considering the total occurrences of (im)politeness) that roughly 900 comments – 41% (out of 2,220 – 100% in analytic corpus) were coded for more than one combination of the above, which is not uncommon due to the polylogal nature of online commentary (Reference Bou-Franch and Garcés-Conejos BlitvichBou-Franch & Garcés-Conejos Blitvich, 2014), as it addresses a multiplicity of audiences, due to context collapse (Reference Márquez-Reiter and PlacenciaMarwick & boyd, 2011).

Results from Tables 5 and 6 also show that it was only in a few instances that intergroup impoliteness was potentially mitigated by the concurrent use of politeness; therefore, in 921 comments – 42% circa half of those included in the analytic corpus, contained straightforward impoliteness toward the outgroup, whereas intragroup impoliteness was only present in 301 comments – 14% (in analytic corpus). Due to space constraints, only patterns of (im)politeness with a circa 10% (or higher) of incidence in the corpus will be discussed and illustrated in detail.

6.2.3 Moral Emotions

One of the key functions of (im)politeness is the expression of emotions (see Culpeper, 2011; Reference Garcés-Conejos BlitvichGarcés-Conejos Blitvich, 2009, 2013). For instance, individuals may resort to impolite language or gestures to vent their emotions and express dissatisfaction or irritation. Conversely, feeling grateful or empathetic can lead to expressions of politeness and appreciation. In addition, expressing negative/positive emotions may be assessed contextually as im/polite behavior. In their in-depth review of (im)politeness and emotions, Langlotz and Locher (2017) propose three areas of research regarding the intersection between pragmatics and emotions. Most relevant to this Element is sociality and emotions, that is, “[w]hat is at stake when emotions come into play (e.g., identity construction and relationship negotiations)?”

Cancelees/ultimate cancelers/associated ingroups/outgroups’ perceived im/moral behavior may be the trigger of positive/negative (moral) emotions which are, in turn, communicated (non)linguistically. This points to morality and (im)politeness being key in social identity construction.Footnote ²¹ Indeed, emotions have been claimed to play a critical role in the processes whereby individuals assume cultural identities (Reference Keltner and HaidtKeltner & Haidt, 1999; Reference LakoffLawler, 1992; Reference PiresPolletta & Jasper, 2001) and in collective intentionality (Reference Jankovic and LudwigJankovic & Ludwig, 2017). More specifically, also regarding groups, shared negative emotion has been linked to group conflict, whereas collective positive emotion is seen as eliciting cooperation and higher performance (Reference Guillory, Spiegel, Drislane, Grinter, Aoki, Cutrell, Jeffries and OlsonGuillory et al., 2011).

To further probe the data, the intersections between moral emotion and in/outgroup(s) were scrutinized through several related matrix coding queries. In this respect, the corpus was coded following Haidt’s (2003) moral emotions taxonomy and considering how expression of said emotions were related to in/outgroup(s).

6.2.3.1 Other-Condemning Emotions

Results showed that other-condemning emotions were mostly expressed in relation to the outgroup(s), with “Contempt” being the one more frequently found in the corpus, 1,104 occurrences – 74.6%, in relation to the intergroup and 283 – 19.1% in relation to the intragroup (Table 11). This makes sense in the cancelation context since contempt is defined as the feeling of (usually moral) dislike for and superiority over another person, group of people, and/or their actions. Further, contempt may be used to reduce interaction with individuals perceived as not contributing in meaningful ways to the group, thus judged to be lower or less capable than the self (Reference Hutcherson and GrossHutcherson & Gross, 2011).

Table 11 Other-condemning emotions and in/outgroups.

	Other-condemning emotions
	Intragroup		Intergroup		Percentage of Total
Anger	54	3.7%	33	2.2%	87	5.9%
Contempt	1,104	74.6%	283	19.1%	1,387	93.8%
Disgust	3	0.2%	2	0.1%	5	0.3%
	1,161	78.5%	318	21.5%	1,479	100.0%

Regarding how it is realized, definitions of contempt further detail that treating others with disrespect and mocking them with sarcasm and condescension are forms of contempt, as are hostile humor, name-calling, mimicking, and body language such as eye-rolling and sneering. Based on this interpretation, it would be expected that contempt would be mostly realized using impoliteness formulae. Related to this context, impoliteness and negatively valenced emotions may be expressed to construct individuals and groups as nonconformist, rebellious, or unique and to position group members very clearly in opposition to the outgroup(s).

To verify these assumptions, another matrix coding query was run which showed a strong correlation between the expression of other-condemning emotions and impoliteness, more concretely “Criticism/complaints” (1,223 occurrences – 29.3%), “Condescension” (767 occurrences – 18.4%), “Disagreement” (504 occurrences – 12.1%), and “Insults” (437 occurrences – 10.5%) (Table 12). In this respect, in Examples 30, 31, and 32, “Contempt” is expressed along with “Criticism” (Republican lie, deny, deflect, and vilify; the public is gullible; ESPN treated Nichols unfairly), “Insults” and “Condescension” (despicable, RepubliCONS, undiversified, hateful, PURE BS), and “Disagreements” (it is not Democrats but Republicans who are fascists; Nichols should not have been fired) among others.

Example 30 And they’re calling Democrats fascists?! The default strategy of RepubliCONS, in every situation, is to lie, deny, deflect and vilify. How can anyone vote for these despicable people?

Example 31 It’s fascinating how their own hatefulness is OK because Ellen must be evil because the media and the mob have told them so. I’m no fan of Ellen but I hate this trial by social media and the mob.

Example 32 **THIS IS PURE B.S.!!! RACHEL**

NICHOLS WAS THE BEST THING E.S.P.N. HAD ON THEIR NETWORK EVERY SINGLE DAY REGARDING THE N.B.A., IN MY OPINION!! NOTHING THAT SHE SAID WAS OFFENSIVE TO OR ABOUT MARIA TAYLOR!! IF ANYTHING SHE CALLED OUT ESPN FOR THEIR LACK OF DIVERSITY!! EVERYONE SHOULD BOYCOTT ESPN NBA COVERAGE THAT IS SLOTTED IN TO REPLACE RACHEL NICHOLS’ & “THE JUMP!”

In view of the results, it seems important to try to account for why further offense is considered a moral and fair retribution to perceived offense, a kind of reactive aggression (Allen & Anderson, 2017). In a highly moralizing context, such as cancelation, expressing other-condemning emotions may be a way for groups to signal their moral quality to others. Another way to explain it is by recourse to “the dark side of morality” (Reference MillsMonroe & Ashby Plant, 2019; Reference Reichl, Akhgar, Bayerl and LeventakisRempala et al., 2020). This concept refers to instances in which our deeply held moral convictions serve as justification for actions that are usually deemed morally impermissible, such as engaging in collective violence with the goal to regulate social relationships, as in the case with cancelation. Moral convictions do not only regulate the self, but also what others ought to do. As Workman et al. (2020: 2) claim: “When the morally convicted are confronted with societal attitudes out of sync with their moral values, some may find this sufficiently intolerable to justify violence against those who challenge their beliefs.” This violence may range in scope from the display of aggressive communicative behavior (as we have observed in the corpus) to the bombing of abortion clinics. In the present case, cancelees’, ultimate cancelers’ behavior is seen as morally reprehensible; the morally convicted see themselves, therefore, justified in exposing them to show who they “really” are, make them face consequences for their actions (legal, financial) and – also aiming at social regulation – try to dissuade others from displaying similar behavior or risk being similarly exposed. In this sense, the dark side of morality is closely connected to degradation ceremonies (Section 6.2.1).

Table 12 Impoliteness formulae and other-condemning emotions.