Liminal Stigma and the Interference With the Positive Construction of Face Masks: The Case of Online Reactions to an Anti-Mask Rally

Personal Reflexive Statement

I have multiple levels of interest in the cultural dispute over mask use in public settings during the COVID-19 pandemic. On a professional level, this conflict is like other cases I have studied where actors construct reality in contentious situations. Mask use like vaccines, genetic engineering, creationism, Donald Trump’s political career, immigration regulations, and sports teams’ use of Native American imagery all involve situations where the same situation is defined in markedly different ways by many different people. Accordingly, it allows me to continue to examine ways that people make sense of their world. Unlike these other conflicts, one’s position on masks, especially in places where use was not mandatory, is dramaturgically realized in face-to-face interactions. One need not express their opinion on immigration or genetic engineering. Not so with protective face coverings. Beyond the theoretical intrigue, is my personal experience with shifting informal and formal regulations of masks in different times and places. During the height of the pandemic, I would go from my campus where there was much support for masks to my exurban grocery store where use was far more contested. This paper is my attempt to understand how to make sense of this emotionally charged but anomic situation.

Introduction

While humans have struggled with pandemics for millennia, such a problem has not been experienced in the lifetimes of most Americans. Accordingly, new norms needed to be established for coordinating response to COVID-19. Arriving at an agreed definition of the situation required to establish such norms was complicated by political polarization. Factions aligned with former President Trump tended to eschew measures to impede the transmission of the virus (Altheide 2020; Best 2020; Gonzalez et al., 2021; Janning et al., 2021; Whitehead and Perry 2020; Zhao et al., 2020) with Republican counties coming to surpass Democratic counties in death rates (Albrecht 2022; Chen and Karim 2021). While we know that the situation is organized along familiar cleavages, there is more to learn about the processes by which efforts to coordinate a response were impeded. Further understanding the suboptimal response to leadership during the pandemic could help to improve coordination during the next public health emergency.

As the pandemic spread, the use of face coverings or masks emerged as one of the means of mitigating the physical danger posed by the disease. These measures also include hand washing, surface cleaning, avoidance of touching one’s face, social distancing, and eventually vaccines and ventilation. To be effective, these practices required widespread participation. Such coordination requires that actors from multiple backgrounds come to the same interpretation of reality (Blumer 1969). After some initial hesitation, the Center for Disease Control (CDC) on April 3, 2020 sought to persuade the public to wear masks (Megerian et al., 2020; Netburn 2021). Their efforts at doing so did increase mask use (Brenan 2020; Goldberg et al., 2020). The response was effective but uneven, with party affiliation but also vulnerability to COVID being correlated with mask use (Katz, Sanger-Katz, and Quealy 2020). Mask use also proved to be politically divisive (Rojas 2020). This challenge was exacerbated by the fact that mask use is not without social costs (Collins 2020). While predictors of mask use have been identified (e.g., Haischer et al., 2020), there is more to be learned about the process by which pro-mask messaging was undermined. By examining how mask use was defined in public discourse, we might learn more about both the COVID-19 pandemic specifically, but also obstacles to mobilizing a divided population more generally.

One challenge for this study is the fact that the construction of COVID-19 occurs in many times and places (Best 2020). A single article cannot hope to provide a complete accounting of the reality construction that occurred. The online reaction to a protest of COVID-19 mitigation efforts, however, should provide a useful entry point. Protests cause a dramatic scene that motivates bystanders to actively consider a situation and the character of those who are involved with it (Benford and Hunt 1992). Accordingly, an anti-mask rally in Idaho in March 2021 dramatized the otherwise diffuse conflicts over masks and, more broadly, the response of the United States to the COVID-19 pandemic. At this rally, approximately 150 people deposited masks into a contained fire to protest local coronavirus restrictions. Video of this event was widely reported on by social and legacy media (Beer 2021; Quinlan 2021). The theoretical value of the online reaction to this protest is that it allows for the observation and analysis of how masks were constructed in numerous online arenas. Given the ever-increasing importance in online social interaction (Cetina 2009), it is imperative to understand how meaning making occurs in online spaces. Indeed, much of the public reaction to this nationally covered protest occurred online. Moreover, the reaction to this event is likely a representative case of the discourse on COVID-19 mitigation that occurred in many places. While there may be differences between the dynamics of the online discussions of masks that occurred here and elsewhere, there should be basic similarities, as well.

The interactionist literature on aligning activity (Stokes and Hewitt 1976) provides a lens for examining this topic. This body of scholarship suggests that analysis of this situation should attend to both the definition of mask wearing but also the evaluation of those who take a position on masks. It also has identified the forms of communication that actors use to defend both themselves and their preferred course of action. It thus facilitates the conceptualization of the specific communicative processes by which the construction of masks as a means of threat reduction was disrupted. The literature on aligning activity has tended to focus on situations where there is normative consensus, but there has been some movement towards examining situations where norms are contested (Silva forthcoming). This analysis has the potential to build on this literature by examining how aligning activity occurs in situations where new norms are being created. Responding to a crisis can necessitate the formation of new norms (e.g., masking). This study identifies some ways that aligning activity can be used to impede the establishment of new norms in response to physical dangers.

In what follows, I discuss the literature on resistance to masking and elaborate on the concept of aligning activity that guides this study. I then describe the online comments from YouTube, Twitter, and Reddit that I collected as data and the steps I took in analyzing them. I will show how the construction of mask use as a mitigation of threat became disrupted via four types of alignment and disalignment. This paper concludes with a consideration of the contributions and limitations of this study.

Resistance to Masks

Masks emerged as an effective option for responding to the COVID-19 pandemic (Brooks and Butler 2021; Howard et al., 2021). Mask use became quite common following the CDC’s April 3, 2020 recommendation to do so. Between August 2020 and April 2021, greater than 80 percent of Americans reported wearing a mask in the past week (Saad 2021). The use of masks, however, became a source of cultural and political conflict (Rojas 2020) and mask wearing as a norm varied throughout the United States (Katz, Sanger-Katz, and Queally 2020). Due to the importance of managing the spread of the disease, it is important to understand this resistance to mask use. Research has identified factors that are correlated with mask avoidance including location in counties and states with higher levels of Evangelicalism and political conservatism (Gonzalzes et al., 2021; Rains et al., 2022; Sanchez, Vargas, and Dominguez 2023), personal tendencies towards authoritarianism, machiavellianism, narcissism, psychopathy (Prichard and Turner 2023), being white (Sanchez, Vargas, and Dominguez 2023) male, younger, and in a rural area (Haischer, et al., 2020) and unvaccinated (Floyd et al., 2022). Mask use was associated with belief in the severity and susceptibility to COVID-19, the effectiveness of masks and one’s ability to wear one, fearfulness of COVID-19, the belief that mask wearing is a social norm (Chen and Lei 2022), the type of political discourse to which one was exposed (Howard 2022). Experimental research has identified potential mechanisms related to mask use. For instance, pro-mask messaging that aligns with audience values could be effective (Demora et al., 2021). Moreover, uncivil comments in response to pro-mask messages could provoke feelings of anger and negativity towards masks (Lu, Sun, and Oktavianus 2023). Finally, mask usage could be a source of identity that leads to bias against out-group members (Powdthavee et al., 2021). Taken together this research identifies some of the correlates and potential causes of mask resistance. These factors, however, exist within a cultural context that warrants examination.

Beyond individual masking behavior, analyses of public discourse on masks have yielded insights into the cultural dimensions of the topic. ¹ Al-Ramahi and colleagues (2021) analyzed 51,170 tweets on masks and identified the following topics in anti-mask tweets “(1) constitutional rights and freedom of choice; (2) conspiracy theory, population control, and big pharma; (3) fake news, fake numbers, fake pandemic, and lies; (4) unhealthy, low oxygen, carbon dioxide, lung infections, and weakened immune system; (5) political, fear, and control people; (6) masks ineffective and cannot block tiny particles; (7) mental health and suicide; (8) herd immunity and dependency on the immune system; (9) child abuse and dehumanization; and (10) virus-related statistics (high recovery rates and low mortality rates).” They also found that a rising volume of anti-mask tweets tended to precede rises in cases. Scoville and colleagues’ (2022) analysis of mainstream media, conservative media, and Twitter showed that there were significantly more pro-masking messages in 2020 than anti-masking ones. Additionally, they argue that masking became politicized by the vilification of anti-maskers. Lang, Erickson, and Jing-Schmidt (2021) also found that Twitter discourse was predominantly pro-mask.

This research shows that mask use was broadly normative but there was politicized resistance. This resistance had identity implications and could be found in social media. This study adds to this work by examining the processes by which positive definitions of masks as efficacious were disrupted on social media. To do so, I draw on the theoretical work on aligning activity.

Aligning Activity and the Construction of Masks

Efforts from both the CDC and elements of civil society (e.g., online commentors) to encourage Americans to wear masks was an attempt to establish what Blumer (1969) conceptualized as a joint action -- wherein actors collaboratively direct their conduct in relation to one another. Given the existing cultural divisions in the United States, it is not surprising that such fitting together of action resulted in considerable disagreement. Nonetheless, it is worthwhile to examine how the establishment of this norm was stymied. The symbolic interactionist literature on aligning activity provides useful conceptual tools for doing so.

Aligning activity conceptualizes the processes by which actors construct reality during such strife. Stokes and Hewitt (1976) define aligning activity as explanations one gives to define oneself as upstanding despite having been accused of questionable conduct. These explanations maintain social relations by aligning dubious conduct with cultural standards. Aligning activity incorporates several sociological concepts (see Scott 2015), including motive talk and vocabularies of motive (Mills 1940), neutralizations (Sykes and Matza 1957), and accounts (Scott and Lyman 1968). These concepts “sensitize” (Blumer 1969) the analyst to the specific ways that actors answer for themselves. Of particular importance here is the concept of justification where people claim that their ostensibly problematic conduct was, in fact, morally sound (Scott and Lyman 1968). ¹

The decision of whether to follow the CDC’s recommendation to wear masks required Americans to interpret whether mask use was a necessary and effective way of managing the physical threat from the pandemic. Actors also had to consider how others might respond to them if they did or did not comply with these requests. The protest over masks provided an opportunity for actors to justify their position on masking. The literature on aligning activity suggests two objects that we should expect actors to construct in reaction to the mask protest: conduct and identities.

One focus of aligning activity will be on framing the definition of the conduct in question. One classic form of alignment occurs when the accused uses a type of justification called a “denial injury” by claiming that no one was harmed by the act in question (Scott and Lyman 1968; Sykes and Matza 1957). For example, shoplifters will redefine their theft by arguing that the businesses that they took from are “easily able to afford the loss” (Cromwell and Thurman 2003:543). Likewise, some convicted rapists will redefine their crime as consensual (Scully and Marolla 1984:535-536). These disparate cases involve a redefinition of the act for which they are accused. Petty larceny and rape are transformed into acceptable actions. At least according to the one offering these claims. Similarly, we can expect that aligning activity in response to the mask protest will include interpretations of the actions of protesters specifically but also more generally mask use and nonuse. This literature will be useful in sensitizing the analysis of the discourse to the specific ways that actors define conduct.

Alignments are also a place where the selves of the participants in the situations are defined (see Scott 2015:67-69). For example, some convicted rapists will “negotiate a non-rapist identity by painting an image of themselves as a ‘nice guy’” … “someone who had made a serious mistake but, in every other respect, was a decent person.” (Scully & Marolla, 1984:540-541). Not only will alignment involve defining the self, but also the construction of alter, as well. The classic justification, “denial of victim,” redefines identities by claiming that the supposed victim was deserving of the alleged crime (Scott and Lyman 1968; Sykes and Matza 1957). For example, shoplifters will claim that the entities they stole from “overcharged consumers and thus deserved the pay back from shoplifting losses” (Cromwell and Thurman 2003:543). Likewise, some serial killers claim that their victims “deserved their fate” (James and Gossett 2018:1,128). Both contrasting examples of denying victimhood seek to redefine the self in positive terms by disparaging another person. The conceptual scaffolding of aligning activity, then, could sensitize this analysis to ways that both mask wearers and non-wearers are constructed through alignment with and disalignment from others.

While the bulk of the work on aligning activity has focused on how people seek to maintain social interaction in the face of normative challenges (see e.g., Hunter 1984), recent work has extended the concept of aligning activity to the defense of the actions of collective conduct that occurs in the face of a conflicted or unclear normative order (Silva forthcoming). In a recent review of the literature, Silva (forthcoming) has identified several features of motive talk and aligning activity in situations where norms are not hegemonically or even widely agreed upon. In anomic situations actors are freer to justify their actions as generally acceptable rather than merely plead that they represent an exception to the rule. Additionally, aligning activity in times of cultural ambiguity leads to debates over not only what counts as acceptable conduct, but who has the moral authority or technical expertise to define reality. Unlike those who stand accused of breaking widely held norms, those who defend and oppose mask use will know that they are joined by numerous others. Likewise, the formation of a collective identity can facilitate the rejection of contested norms. Moreover, “this research demonstrates the importance of social media in dramatizing and concretizing cultural differences” (Silva forthcoming).

Mask wearing was a norm, but not a hegemonically unquestioned one. Unlike those norms children learn uncritically during primary socialization (Berger and Luckman 1967), masking was not simply natural for most American adults. Although the absence of normative unity empowers one to reject norms, one can still face sanctions for violating even emergent norms. The messaging from public health agencies, mandates to wear masks, and the high rates of mask usage likely put some social pressure on those who did not wish to wear masks (e.g., Rojas 2020). Recent developments in the aligning activity research on disalignment (Dellwing 2012, 2015) provides some insights about how actors might mitigate the power of emergent norms. Dellwing (2015) relates that much of the work on aligning activity has assumed that actors will seek to reduce conflict in problematic situations. But he contends, “Those whose behavior is challenged, “called out,” need not align their actions; they can resist others’ challenges and make counter challenges, leading to sustained conflicts over alignment” (p. 160). I would argue that such disalignment is facilitated by situations where norms lack hegemonic consensus. Silva and Flynn (2020) built on Dellwing’s work to argue that disaligning actions are an element of online cultural disputes. They found that the disaligning activity that occurred during online discussion of the family separation crisis at the U.S. border in 2018 presented a situation where the selves of all participants were threatened as they were subjected to liminal stigma. While the liminal stigma that occurs online does not carry the material consequences of stigma (see e.g., Hannem 2021), it does present participants in the discourse with evidence that they are despised by some uncertain number of others. Such a reflection, in turn, motivates a vigorous defense of the self. It is reasonable to expect, therefore, that liminal stigma could be produced in the online reaction to the mask protest. This contentious discourse might also produce validation of identities as commenters align themselves with likeminded individuals. This study can build on these recent developments in the literature on aligning actions by demonstrating how aligning activity could potentially impede the establishment of new norms -- in this case the norm that individuals should wear masks during an epidemic. I now turn to how the observations of this aligning activity were collected and analyzed.

Data and Method

The data for this project come from discussion of the protest that occurred on Twitter, YouTube, and Reddit. Despite their differences, these three forums can be included in a single study as they share some fundamental commonalities. Each website enjoys an active user base and provides a venue for individuals to participate in the actual if not idealized public sphere (Adut 2018). These data do not tell us how people feel about virus mitigation efforts, but it will show us the culturally acceptable reasons or “vocabulary of motives” (Mills 1940) for one’s opinion. Each forum provides an organically occurring, nonreactive record of a moment in time where competing interpretations of mask use were being offered. The 924 comments contain much variation and thus allow for a detailed analysis of the varieties of alignments that actors use when defining this situation.

The ways that forum specific features might shape the construction of reality goes beyond the scope of this paper. Instead, this analysis will focus on the communication on social media that occurs in online forums. An important warrant for this approach is that processes that occur in multiple forums are not simply an artifact of any specific forum. While these forums also have idiosyncratic norms, moderation policies, organization, and user bases that make them distinct from each other, by focusing on alignments that occur in more than one forum, we might expect to find general processes. Additionally, because these forums are embedded within a wider cultural struggle of COVID-19, it is reasonable to assume that the aligning activities that are observed in the discourse of these online forums also occurred in other contexts.

From YouTube, I used the search terms mask burning to find a CNN report on the March 6 Idaho protest. This video has received 225,880 views and 5,231 comments as of December 27, 2021. I selected all comments from the top 16 threads (268 comments). From Twitter, I selected the thread of tweets and responses to it (165 tweets) posted by Sergio Olmos who reported on the protest and provided a video that had 13.3 million views, about 10,310 retweets, 29,386 quote re-tweets, and 32,995 likes as of December 27, 2021. From Reddit, I selected discussions of the protest that occurred on 10 subreddits (491 comments). I selected comments by sorting the comments by top comments and taking all that were displayed by the website (without collecting obscured comments). While the majority of the comments come from Reddit, the logic of the analysis is that processes observed in more than one forum are potentially of a general nature and not specific to Reddit.

While the literature on aligning activity directs analysis towards an examination of how conduct and identities are defined in online discussion (e.g., Silva and Flynn 2020), it does not call for a specific coding schedule that one would use in a quantitative content analysis (¹Altheide 1987). Rather, this study was centered on what Altheide (1987¹:68) calls ethnographic content analysis which consists of reflexive movement between concept development, sampling, data collection, data coding, data analysis and interpretation.” Importantly, “although categories and ‘variables’ [such as aligning activity] initially guide the study, others are allowed and expected to emerge throughout the study.” Following the analytic guidelines outlined by Lofland et al. (2006) this ethnographic content analysis was conducted in multiple rounds moving between open ended and focused coding. In the first round of the analysis, I engaged in open coding of the material that generated 60 codes and 46 pages of coded material. The analysis was guided by the literature on aligning activities. When analyzing the comments, I coded for the alignments commenters used to “justify” (Scott and Lyman 1968) themselves. In the second round, based on a review of the open codes and references to the literature on alignment, I coded for whether statements aligned with mask proponents, mask opponents, partisans, and interlocutors. Using the NVivo software I combined these codes into anti-mask discourse, pro-mask discourse, and extra-mask discourse (91 pages of material). In the third round of the analysis, I coded the previously coded material for whether it referenced threat to physical body, threat to social self, or validation to social self. With the fourth round, I open coded the material from round two to examine the ways that commenters aligned and disaligned from mask proponents and opponents. In the fifth round, upon review of previous rounds, I noted that there seemed to be many versions of threats to the physical self. I then open coded material that had been coded as threats to the physical self to identify the types of physical threat that commenters drew upon. Following the literature on aligning activity, the discussion of threats to the physical body related to how actors evaluated conduct. In the final round, I reviewed the numerous ways that the claim that masks were a protection against threat to the physical self was mitigated in the discourse. Again, following the aligning activity literature, the threats to and validation of selves involved the construction of identities. Although I will provide the frequencies for each process observed in the data, these categories are not discrete, and these frequencies are not generalizable.

Findings

The analysis described above revealed four processes by which efforts to construct a norm of mask use was impeded: fragmentation, neutralization, dilution, and displacement. The first two processes occur as people’s conduct was evaluated in terms of the types of physical threats that it exacerbated or mitigated. It was here that the advisability of masking norm was constructed. The second two processes involve the construction of identities. While Twitter, Reddit, and YouTube have distinct features, each of these processes occurred in all three forums and each type of example was found in a minimum of two forums.

Discussion and Conclusion

The online reaction to the mask burning protest held in Idaho in March of 2021 provides an opportunity to examine how attempts to construct mask use as a means of mitigating the threat of COVID-19 were impeded in the United States. This qualitative analysis of comments found on Twitter, Reddit, and YouTube demonstrates four processes that contributed to this disruption: the physical threat masking is supposed to be protecting is discussed alongside other types of threat related to COVID-19 and people’s reaction to it. Each of these separate fragments is then neutralized in the ensuing sub debates about the acceptability of various conduct taken in relation to the threat. The discussion of the physical threats posed by COVID-19 is also diluted and displaced by aligning activity that evaluates the moral worth of the actors in the situation.

The mask burning protest and the online reaction to it presumably have non-generalizable elements. Nonetheless, it is reasonable to assume that the reaction to this event is not entirely idiosyncratic. At least some of these processes have likely occurred in other discussions of masks or other mitigation efforts that depend on a shared definition of the situation. The protest itself was not a discrete event but was embedded within ongoing discourse on masking and COVID-19 precautions. Accordingly, some of the alignments we see here are likely to be reenactments of discursive patterns.

This study has implications for understanding mask resistance specifically and aligning action more generally. Regarding mask resistance, these findings call attention to how constructions of masks and COVID-19 were intertwined with constructions of selves. A result of this identity construction is that the production of liminal stigma, can override the discussion of threats to the physical body. The aligning activity found in the comments section served to validate or stigmatize the protestors themselves, but also those who support and oppose the protestors. Some discussion focused on the specifics of the protests (e.g., the problems caused by fumes from the burning masks) but much of the discourse settled on more general issues about mask use. The existing literature on mask resistance reveals how conflicts over mask use are exacerbated by existing political polarization (e.g.,Gonzalzes et al., 2021; Scoville et al., 2022) and that mask wearing carried identity implications (e.g., Powdthavee et al., 2021). This study confirms those findings by demonstrating the salience of identity in the cultural opposition to mask use. It further adds to the research on social discourse on masks (e.g., Al-Ramahi et al., 2021) by documenting the texture of these arguments about masking. In looking at the discussion responding to an anti-mask protest we can see the role of human agency in dramatizing these cleavages. Polarization is not something that simply exists, it is realized through human activity. Both the protesters in Idaho and the reaction to protest online created an opportunity to publicly contest a burgeoning masking norm that might otherwise go uncontested in everyday life. One product of this dramatization of the conflict over masks is that the selves of both supporters and opponents of mask use became contaminated with a liminal stigma. While liminal stigma is not as severe as stigma as traditionally understood (Hannem 2021), it does provide concrete evidence that one is despised for their position on masks (Silva and Flynn 2020). Justifying the use or nonuse of masks, then, became justification for one’s identity in online social interaction. Accordingly, this research further illuminates some of the cultural obstacles facing those who wish to promote masks during the next epidemic.

This study has also has implications for the literature on aligning activity. While much of the research on aligning activity has focused on situations where actors redefine conduct that is problematic according to longstanding norms, this analysis builds on research efforts to show aligning activity is also used in situations where actors defend against norms that have not yet achieved hegemonic acceptance (Silva forthcoming). This study supports recent literature arguing that aligning activity can be a mechanism for creating division (Dellwing 2012, 2015; Silva and Flynn 2020). It extends this literature by identifying specific processes (fragmentation, neutralization, dilution, and displacement) by which aligning activity can undermine a construction of a new norm, even in a time of crisis. These processes could prove to be sensitizing concepts for scholars in this tradition who study how non-hegemonic norms are resisted.

This study shows two ways that aligning activity can impede the establishment of a new norm. First, this analysis shows how aligning activity can reduce the attention given to the construction of a new norm. When the topic of masks is divided into multiple issues about what constitutes a threat, is diluted with conversation about the identities of those involved, and other matters, there is less “discursive space” (“the total volume or amount of space available for talk, discussion, commentary, and/or reportage -- whatever it’s form … or context … -- that is or can be bounded in time, and often physical space as well” (¹Snow 2008:20) to consider the mask itself. Unlike norms that become natural through childhood socialization (Berger and Luckmann 1967), a new norm should presumably require some explanation to take hold. The reaction to the mask burning protest was just one instance of a larger struggle over the establishment of masking as a norm. If this threat fragmentation process occurred in other places where masking was discussed, it could impede the construction of this incipient norm. The establishment of a new norm, such as masking, must thus contend with existing cultural divisions and commitments that are brought to bear through the use of the aligning actions identified in this study.

Second, the aligning activity that occurs in reaction to a new norm might also produce a liminal stigma. Silva and Flynn (2020) observed liminal stigma being produced in online discussions of the family separation program of the Trump administration. It was generated in the online response to the Idaho mask burning rally observed here. The obdurate threat of COVID-19 results in over one million American lives lost, nonetheless, discussion of how to respond collectively became sidetracked by contests over the worthiness of certain identities. It is worth noting that the processes outlined in this study were co-produced by commenters who, ostensibly, supported the masking norm. It is possible that the conflict over a new norm can provide a sense of validation. If one sees that one has been stigmatized, even if only passingly, by online discourse, there could be validation to be found in identifying with others who also reject this same stigmatization. The reaction to the protest included many acts of alignment where individuals who were otherwise strangers could find commonality in their opposition to an out-group. This study does not show how those who read and write these comments actually feel about either the attacks on or defenses of their character. There is, however, much evidence of commenters offering validations of the selves of others. Multiple positions towards masks were celebrated in the reaction to the protest. If there are people in need of validation, such a conflict appears to be a reliable supply of it. Whether they would seek to avail themselves to it goes beyond what the data collected here can tell us.

This study, then, does offer a practical suggestion. Those who participate in civil society discourses in support of some change in social norms, might well consider the extent to which they themselves are helping to move a debate about a behavioral conduct (such as masking) into a question of which identities are most worthy of respect. Experimental research has suggested that messaging that aligns with audience values could improve support for masking (Demora et al., 2021) and that stigmatizing comments are likely to be ineffective (Lu, Sun, and Oktavianus 2023). This analysis of discourse that occurred in a natural setting confirms that many commenters were doing the opposite of what the experimental research suggests would be persuasive for those opposed to mask use. Those who seek to encourage others to adopt a shared definition of the situation might do well to avoid threatening the selves of their interlocutors. Both those who ostensibly supported and opposed mask use indulged in threatening the selves of their opponents. In turn, we see selves being validated in the alignments that defend against this liminal stigma. Such alignments are a part of the construction of identity. The importance being that to have a particular identity sets boundaries on what is and what is not permissible (e.g., Sharp 2009). It is worth considering the work that mask supporters contributed to the maintenance of these boundaries.

While this study has contributed to the literature on the construction of threat during COVID-19, there is much that it could not do. First, this project tells us little about subjective experiences and internal motivations to mask or not mask. This analysis has followed Mills’ (1940) argument that sociologists can only access stated motives, not the actual motivation of individuals. Perhaps, another logic of inquiry or method would allow for such a discovery. Second, the messaging actions of the CDC and other public health agencies are beyond the scope of this paper. A fuller understanding of challenges to establishing a norm of masking could be gained from such a project. Third, the definition of COVID-19 surely varies by context. The organizational features of online forums often encourage contentious interactions (Lin and Tian 2019; Phillips 2015). The communications on masking that occur in other settings could well be less emotionally charged. Future research should address the question of how mask use was constructed in other settings and at different stages of the pandemic. This research might well uncover other interactional processes by which efforts to establish new norms definitions were undermined.