Hashtag Convergence: #Georgefloyd and #Covid19

In Press (Forthcoming in First Monday)

Hashtag Convergence: #GeorgeFloyd and #Covid19

Jenny L. Davis The Australian National University School of Sociology [email protected]

Tony P. Love University of Kentucky Department of Sociology & The Australian National University School of Sociology [email protected]

The authors would like to thank Dr. Claire Benn, Professor Toni Erskine, and Dr. Sarah Logan for their feedback and support in developing this paper.

Bios:

Jenny L. Davis (@Jenny_L_Davis) is a sociologist at the Australian National University, 2019-2022 DECRA Fellow, and a Chief Investigator on ANU’s Humanising Machine Intelligence Project.

Tony P. Love (@tonyplove) is a sociologist at the University of Kentucky with an honorary appointment in the School of Sociology at the Australian National University.

Jenny and Tony co-lead the Role-Taking Project, an international collaboration that aims to learn about and improve the ways people understand others’ experiences, worldviews, emotions, and behaviors (https://www.role-taking.com/).

1 Hashtag Convergence: #GeorgeFloyd and #Covid19

Abstract

In the late spring of 2020, amid a global pandemic, George Floyd was killed by police in Minneapolis, Minnesota, triggering mass protests under the banner of the Black Lives Matter movement. We take this moment of coinciding crises as our point of analysis observed through the lens of concurrent hashtags on Twitter. Social media content both reflect and construct the social meanings of topics and events. We thus draw from social media to understand how George Floyd and Covid-19 inform and inflect each other, building a dataset from ~20,000 tweets that unite prevalent hashtags associated with each. Analyses reveal a repeating set of symbolic hooks—death, breath, masks, and voice—encompassing dense and competing narratives about justice and injustice, systemic inequality, degrading trust in institutions, and the changing identity of a nation. These narratives are anchored in the events under study and indexed through co-occurring social media registers. In addition to substantive findings, the study introduces and applies hashtag convergence, a novel methodological approach based on user-generated indexical pairings. . Keywords: Covid-19, Black Lives Matter, Race, George Floyd, Twitter, Hashtag, Hashtag Convergence

2 Since the onset of its global spread, Covid-19 has had an omnipresent character, overtaking news cycles and touching all elements of personal and public life. In this same period race relations loomed, animated in the United States by the Black Lives Matter (BLM) movement against state violence and systemic racial injustice—an ongoing project that predates the pandemic and continued to bubble as Covid-19 took hold. These dual contextual backdrops of a pandemic and a racial reckoning came sharply together in a single week of late spring,

2020. Just days after the United States surpassed 100,000 Covid-19 deaths (CDC, 2020) footage spread of Minneapolis police officer Derek Chauvin kneeling on George Floyd’s neck, killing the unarmed Black man who asphyxiated and expired. In the thick of pandemic shutdowns, BLM protestors took to the streets.

Although BLM and Covid-19 have indelibly shaped and informed each other

(Benjamin, 2020; Bonilla-Silva; 2020; Schachter, 2020), there has yet to be a systematic study of the social and cultural meanings produced and deployed through their entanglement.

In the present work, we probe this entanglement through an empirical study of concurrent hashtags on Twitter in the immediate aftermath of George Floyd’s death, as both racial justice mobilization and the Covid-19 pandemic swelled. Our study is driven by the following question: how have Black Lives Matter and Covid-19 symbolically converged, and what meanings are deployed through their pairing? We address this question with a targeted study of George Floyd and Covid-19 on Twitter, using a novel methodological approach—hashtag convergence—to build a dataset from tweets that indexically unite the two topics.

We collected ~20,000 tweets that join #GeorgeFloyd with #Covid-19 as dual hashtags, showing contested discourses of race, virality, mobilization, and national identity, all hung on four symbolic hooks: death, breath, masks, and voice. These data serve as a lens on meaning and action as they emerged in real time, forged at the crossroads of two historically transformative events. We use these data as an access point to BLM and Covid-

3 19, uncovering their sociological confluence via a communication medium that has been integral to each.

Intersecting Matters: BLM, George Floyd, and Covid-19

One of sociology’s earliest and most enduring insights is that social life is interconnected.

Across micro, macro, interpersonal, and institutional domains, social facts converge to co- produce material circumstances and symbolic meanings1. Such is the case with George

Floyd’s death and the Covid-19 pandemic, two intersecting matters that took shape through their relation. There is no intrinsic or obvious tie between a prolonged global pandemic that kills millions and a single police encounter that kills one. Yet, the connection between Covid-

19 and George Floyd is fundamental and profound. This relationship is based on several factors, not least of which is the temporal overlap between these two phenomena, coupled with a link between Floyd’s death and a larger ongoing movement for racial justice in the

United States—Black Lives Matter— that infused the cultural moment while Covid-19 surfaced and spread.

The Black Lives Matter movement dates back to 2013, when social activist Alicia

Garza wrote a Facebook post responding to the acquittal of George Zimmerman, a vigilante who shot and killed Trayvon Martin while the unarmed Black teenager walked home from a convenience store. Garza ended the post with the hashtagged statement #BlackLivesMatter.

The slogan and its hashtagged moniker picked up steam the following year during protests over Michael Brown, a 14-year-old who was killed by police outside Ferguson, Missouri.

With the Michael Brown protests, Black Lives Matter rose to prominence as a cultural icon, an organization (co-founded by Garza, Patrisse Cullors, and Opal Tometi), and a movement

1 That society is interconnected and co-constitutive is such an integral assumption of sociology that it would be strange to attribute the idea to any specific scholars or pieces of writing. However, some exemplary and canonical statements of this assumption include: Durkheim (1897/1952), Mills (1959), Berger and Luckmann (1966), and Collins (2002). 4 that now defines contemporary race politics in the United States (Bunting and Stamatel,

2019; Freelon, McIlwain and Clark, 2016; Garza, 2014).

Although BLM has had a sustained presence, it surges and recedes as high-profile incidents of racial violence emerge, flow through media cycles, and fizzle out or resolve. In the winter of 2020, mobilization was relatively subdued due to collective attention on Covid-

19 and the practical matter of people staying out of public and in their homes. However, tensions began to rebuild in February when White vigilantes killed Ahmaud Arbery, an unarmed Black man jogging in Georgia. These tensions increased in March, when Louisville police shot and killed Breonna Taylor while executing a wrongfully issued no-knock warrant based on faulty information. On May 25, when video footage of George Floyd’s death spread through media channels, the movement reignited with protests in Minnesota, then across the country, and across the world, marking what some would call a “turning point” in U.S. race relations (Guardian, 2020).

The Floyd protests were spurred by police violence but largely defined by their activation during pandemic shutdowns. To avoid viral spread, many cities and states kept people out of school and work, closed bars and restaurants, and enacted a range of measures to enforce physical distancing. These measures were politically divisive, with compliance and resistance falling along progressive and conservative lines, respectively (Gollwitzer et al.,

2020; Kushner Gadarian, Goodman and Pepinsky, 2020). It was thus the case that many who advocated for and adhered to physical distancing also supported and/or participated in racial justice protests which had people pressed together in the streets (Cobbina et al., 2021), while many who contested shutdown measures now cast the protests as public health hazards and the protesters as hypocritical actors (Nguyen, 2020). As we will show with the data, these circumstances of protest-amid-pandemic became at once a signal of the protests’ power and

5 also, a means for conservative critics to undermine both the protests themselves and the progressive position on public health mandates.

Just as the Floyd protests were defined vis-à-vis pandemic life, the pandemic too was inflected by BLM. Even before Floyd, the BLM movement had sensitized Americans to the breadth of systemic racism and this likely factored into public discussions and research priorities that focused on racial disparities in virus-related health and economic outcomes

(Craven, 2020; Devakumar et al., 2020; Leicht et al., 2020; Marshburn et al., 2021). With this foundation already set, Floyd’s highly public and violent death channeled and concentrated anti-racist energy, cracking through color-blind pandemic narratives (Bonilla-Silva, 2020) and breaking open a space to address and account for the diffuse effects of institutional racial oppression (Thelwall and Thelwall, 2021).

Social media have been integral to the matters under study and to their intersection, figuring into the “new normal” of remote contact amid pandemic shutdowns (Ferrara, 2020;

Fischer, 2020; Pei and Mehta, 2020; Singh et al., 2020; Wicke and Bolognesi, 2020) and, from its onset, central to the Black Lives Matter movement (Bonilla and Rosa, 2015; Bunting and Stamatel, 2019; Freelon, McIlwain and Clark, 2016; Ince et al., 2017; Kuo, 2018; Ray et al., 2017; Recuber, 2021). More generally, social media are sites of both communication and production, through which communicative acts create a living record of empirical events, what they mean, and how they matter (Couldry and Hepp, 2018; Davis and Love, 2018;

Davis and Love, 2019; Halford and Savage, 2017; Hepp et al., 2018; Sloan and Quan-Haase,

2017). We thus draw on social media data to understand the convergence George Floyd and

Covid-19 and the symbolic meanings that develop through their coupling.

METHODS

Hashtag Convergence: Developing a Dataset

6 To study the symbolic confluence of BLM, George Floyd, and Covid-19 we devised hashtag convergence as a data collection method. This approach builds a dataset from user- generated indexical pairings, probing these pairings for the meanings that emerge. Hashtag convergence represents a subtle but significant pivot from traditional hashtag studies, which have become a lynchpin of social media research (Bonilla and Rosa, 2015:5; Brock, 2012;

Bruns et al., 2016; Bruns and Burgess, 2015; Zappavigna, 2018).

Most hashtag analyses take three forms: single topic analyses, competing frame analyses, and multi-topic analyses via combined datasets. Single topic hashtag analyses focus on a singular tag and/or related set of registers, such as studies of #BlackLivesMatter, (e.g.,

Freelon, McIlwain and Clark, 2016) or studies of #Covid19 (e.g., Ferrera, 2020; Wicke and

Bolognesi, 2020). Researchers also examine competing hashtags about the same topic, recording contrasting frames such as those represented by #BlackLivesMatter vs.

#BlueLivesMatter in statements about race and police violence (e.g., Gallagher et al., 2018;

Ince et al., 2017) or #Covid19 vs. #ChineseVirus in framing the pandemic (e.g., Hswen et al.,

2021). Finally, when researchers do analytically combine distinct topics, it has thus far been through manufactured datasets in which hashtag pairings are exogenously constructed by the researcher, such as Thelwall and Thelwall’s (2021) repository of BLM and Covid related tweets.

In contrast to extant methods, hashtag convergence collects social media posts that join topically distinct hashtags within a single utterance (e.g., #GeorgeFloyd and #Covid19).

This approach treats dual topics as the unit of analysis, derived from user-generated pairings.

Such an approach, which we take in the present work, is ideally suited to the study of co- constitutive phenomena, attending to the ways discursive publics merge these phenomena into a sui generis whole.

7 On a technical level, we enacted hashtag convergence by scraping publicly available tweets that combine both George Floyd and Covid-19 in a single post, indexed by concurrent hashtags, and shared in the immediate aftermath of Floyd’s death. Tweets in our dataset thus meet the following criteria: they were tweeted between May 25, 2020 and June 26, 2020, and they contain the hashtag #GeorgeFloyd plus at least one variant of #Covid19 or #Coronavirus

(not case sensitive), such that all tweets in our dataset make indexical reference to both topics of interest. These dates represent the month following Floyd’s death, during which time social media activity around Floyd and BLM peaked (Rothschild, 2020). Our selected hashtags are the most prevalent for Floyd and Covid-19 on Twitter as calculated by hashtagify.me, a website that documents hashtag popularity and usage.

To make the data both tractable and precise, we did not scrape all BLM content.

Instead, #GeorgeFloyd served as a conduit for the BLM movement during coronavirus. Just over 1/3 of the tweets within the sample thus contain #BlackLivesMatter and its variants, but all included tweets are indexed by the hashtag #GeorgeFloyd + at least one version of the

#Covid19 hashtag. The dataset includes tweets that contain the designated hashtags directly, as well as tweets that have been quote-retweeted using the hashtags2. We scraped the data from the Twitter advanced search page using webscraper io, an add-on to Google’s Chrome browser, generating a total of 20,017 tweets and retweets.

Analysis

Our analysis was guided by the “small data” tradition in social media research, prioritizing depth and richness through qualitative interpretivist techniques (Brock, 2018;

Latzko-Toth, Bonneau and Millette, 2017; Sloan and Quan-Haase, 2017), while maintaining a broad lens on the holistic dataset. We drew in particular on the abductive method, a top- down, bottom-up approach in which existing theoretical and empirical insights are brought to

2 This means that some of our example tweets do not display the hashtags. Tweets that do not display the hashtags have been retweeted with both #GeorgeFloyd and a variant of #Covid19, thus meeting the inclusion criteria. 8 bear on new data, while the data work back to advance and elucidate established understandings using grounded methodological practices (Tavory and Timmermans, 2014;

Timmermans and Tavory, 2012). We thus read the data to let themes emerge, triangulating those themes with theoretical insights from the fields of symbolic interaction, social media, social movements, and critical race studies, with insights form each woven throughout the findings.

The authors read the data independently and then together, establishing initial impressions and orienting patterns. This process revealed a notable trend: a refrain of symbolic markers that were present across tweets, serving as vehicles for claims and contestations about the specific events under study and about broader social issues, anchored and amplified by those events. We distilled these down to four symbolic hooks—death, breath, masks, and voice—which guided subsequent analyses led by the first author and spot- checked by the second. This process included keyword searches for thematic terms, hand coding of all relevant tweets, parsing each theme into relevant subthemes, and identifying exemplar tweets for each theme and subtheme3.

Table 1 represents the presence of each symbolic hook within the dataset. This includes a reference-level and tweet-level accounting. The reference column shows the total number of times the symbolic hook was mentioned in the dataset, allowing for multiple references within a single tweet. At the tweet level, we show the number of whole tweets that draw on the theme, including the percentage of each theme in relation to the full data corpus, allowing tweets to be counted in multiple categories. Although some of the thematic percentages are relatively small, each theme was referenced over 1,000 times and featured in more than 600 tweets, figuring into a patterned story about life and death at the intersection

3 Thematic coding was based on a holistic reading of text, links, and attached images for each tweet. For the exemplar tweets presented in the Findings section we include the original text and links, describing relevant imagery in accompanying brackets. 9 of Covid-19 and racial (in)justice. We organize the findings around these four symbolic hooks, evinced with exemplar tweets from the dataset.

FINDINGS

The Covid-19 pandemic and George Floyd protests came together as part of a socially transformative moment, marked by ideological shifts and deep social divisions. Within this context, our data reveal complex narratives about justice and injustice, systemic inequality, degrading trust in institutions, and the changing identity of a nation. These abstract issues take diverse and often divergent form, articulated through a repeating set of symbolic hooks: death, breath, masks, and voice. The boundaries between each hook remain porous, with discourses from each seeping into, shaping, and informing the others. We separate them here for analytic clarity while allowing their overlaps to remain evident.

Table 1 About Here

Death

The story of Covid-19 and the story of George Floyd are each steeped in death. The former threatens to kill those afflicted and has done as much to over three million worldwide at the time of this writing. The latter, a single death in late spring 2020, represents centuries of continuing state violence against Black people in America and the fatal devaluation of Black lives. The intersection of these events is thus marked with a distinctly morbid character, as represented in our dataset with over 8,000 death-related references across 5,980 tweets.

Invoking death, the tweets we analyzed frame Covid-19 and racism as dual, deadly pandemics, read Covid-19 mortality rates through a racial prism, and tie these realities to the conditions of systemic racial inequality.

Dual Pandemics

The BLM movement is premised on the social fact that Blackness can be fatal.

Arising after high-profile extrajudicial killings of Black Americans by vigilantes and police,

10 BLM works to highlight and dismantle the systemic forces that make Blackness deadly.

Within the dataset, racism in general, and Black mortality in particular, were situated alongside Covid-19 as dual pandemics, framed as a matter of racialized public health, with

Covid-19 operating as both a metaphor and a scale for expressions of racial violence.

We mobilized the planet to combat the #COVID19 virus. What are we doing to combat the disease of rampant racism? We must combat and eliminate this far more insidious disease. #GeorgeFloyd #RacismMustEnd [Image of Floyd pinned to the ground]

While our attention has been on #coronavirus, and rightfully so... there is another type of contagion that still keeps some of White America paralyzed: Fear of black men in public spaces. https://t.co/h5OTD7b8NU #GeorgeFloyd #AhmaudArbery #FranklinTempleton #tooman[y]tohashtag

Although Floyd is a central figure, his name was often catalogued alongside a string of names and incidents, joining Floyd’s death with others who have suffered similar fates and highlighting the systemic—rather than isolated—nature of Floyd’s murder. This linking of

Floyd with other victims, along with reference to pervasive racial bias (e.g., “fear of black men in public spaces”), position racism vis-à-vis the pandemic as insidious, pervasive, and in need of fundamental redress.

Racial Health Disparities

The death and infection rates of Covid-19 are disproportionately higher among Black and Latinx Americans as compared with other groups (Hooper, Nápoles and Pérez-Stable,

2020). These health outcomes follow existing trends in racial health disparities that show significantly reduced life expectancies for people of color in the United States (National

Center for Health Statistics, 2016). Over the course of the movement, BLM has trained national attention on the expanse of racial inequalities, including those of health and wellness. Thus, even before Floyd’s death, racial disparities in Covid-19 outcomes were already part of the national conversation. Although it is impossible to know how salient these racial disparities would have been without the BLM movement, it is near certain that BLM’s

11 racial justice messaging increased social sensitivity to race related variables in analyses of coronavirus and its effects (Craven, 2020; Devakumar et al., 2020; Leicht et al., 2020).

With the death of Floyd and re-centralization of BLM concerns, Covid-19 racial health disparities surged into the spotlight (Thelwall and Thelwall, 2021). Within our data, this became part of a bigger story about death, Blackness, and the consequences of systemic, institutionalized White supremacy.

It is hard to hear "we are all in this together" when POC [people of color] are dying at higher rates due to #COVID19 + being murdered by white men #GeorgeFloyd #AhmaudArbery - Can we be in this together + challenge systemic racism? That's my "together" #BlackLivesMatter https://t.co/Y1Gc0JUrEr [Link to article about race, surveillance, and COVID-19]

Folks are now asking which is worse for Black people: #COVID19 or racism. This false dichotomy misses the larger, structural reality. Covid harms Black people disproportionately BECAUSE OF racism. This is all connected, folks. The tragic reality of our Blackness. #GeorgeFloyd

This focus on disparate Covid-19 outcomes and the connection between racialized health disparities and state violence against people of color remains tied to the broader framing of dual pandemics (coronavirus and racism) which together, make the prospect of Black death apparent and acute.

Breath

Physical breath is a focal element of both Covid-19 and of Floyd’s death, featuring 2,705 times in our data across 1,292 tweets. The novel coronavirus is a respiratory disease that causes inflammation in the lungs and low blood oxygen. Critical cases can require intubation onto ventilators and a great many who die from the virus do so because they can no longer breathe. Like coronavirus victims, Floyd too was deprived of breath, which he indicated to officers as they applied lethal force to his body4. At the same time that literal breath factors into Covid-19 and systemic racial oppression, these dual crises have also had socially

4 The phrase “I can’t breathe” was already central to the BLM movement prior to Floyd’s death, arising with the case of Eric Garner who was killed by police in 2014 for allegedly selling loose cigarettes. He died in a chokehold while telling officers he could not breathe. 12 suffocating effects, overwhelming daily life and upending a taken-for-granted sense of normalcy. In both physiology and metaphor, breath thus presented in the data as a defining thread.

Physical Breath

George Floyd’s refrain—“I can’t breathe”—communicated the literal loss of breath from his body at the hands (and knees) of police. This proclamation of breathlessness resonates with the experience of Covid-19 patients, whose most notable symptom is respiratory distress. On Twitter, subjects highlighted this connection.

George Floyd's heartbreaking final words "I can't breathe" are a wake-up call for the world and perhaps the definitive epitaph of the year 2020 #GeorgeFloyd #JusticeForFloyd #Coronavirus #COVID #Riots2020 #ProtestFromHome https://t.co/ 522Rmz0v7q [Image of Floyd, an animated COVID-19 infected lung, and animated face masks]

The last time I heard the words “I can’t breathe” was when I took care of a nurse who had #covid19. She was in tears. These were the words that #GeorgeFloyd said before he was killed by a police officer who suffocated him. Haunting. #icantbreathe

This protestor's mask captures the pain of 401 years of systemic, silent (silenced) racism in the USA: 1619-2020 And conjures pain of last breaths taken by the majority of #COVID19 fatalities: Hispanic, Asian, African American & the poor & uninsured #icantbreathe #GeorgeFloyd [Image of woman of color wearing a surgical mask that reads ‘Please, I can’t breathe’].

Interwoven in these statements of breathlessness are acknowledgments of raced and classed

Covid-19 mortality rates, as discussed in the previous section (Death), along with allusions to encompassing forms of subjugation, to which we now turn.

Breath as Metaphor

Both Covid-19 and state-based violence against people of color have physical, embodied consequences. These phenomena also have flow-on effects that pervade the lived experiences of those affected in ways big, small, obvious, and invisibilized. These flow-on effects present as layers of pain that wreak havoc on the body, on society, and on the self, culminating at the intersection of Covid-19 and George Floyd’s death. Along with accounts

13 of exhaustion in the face of persistent struggle, tweeters in our dataset attached their laments to Floyd’s three simple words.

The last words #GeorgeFloyd is remembered by is "I can't breathe" amid the #COVID19 pandemic that's literally taking people's breath away. Floyd's case is a symbolic image of the US wretched by police brutality and the respiratory disease

#Breathing: the ultimate #rhythm that defines whether or not we are alive & how we live. Between #covid19 respiratory issues & the tragic death of #GeorgeFloyd, it seems like deep breath is what humanity is missing now. It will remain for me the marking #metaphor to evoke #2020

In these tweets, we observe how the pandemic and state violence have both come to represent suffocating circumstances, with compounding, racialized effects.

Masks

Masks have been a fraught issue in the United States since the early days of Covid-19.

Plagued by inconsistent advice from health experts and political leaders and complicated by supply shortages for frontline workers, the question of whether or not to wear a mask was unclear for a period of time. Today, we know that masks effectively curb the spread of the virus. However, mask wearing remains contentious, and was even more so in the period under study. Exacerbated by then U.S. president Donald Trump’s protracted resistance to face coverings, masks took on political significance. Masks also became critical technology in protest participation. The physical co-presence of BLM activists posed a risk of infectious spread, including among communities of color that were (and remain) disproportionately affected. Masks thus featured in the data as a point of political division and as a tool of political mobilization (see also (Lupton et al., 2021). The dataset includes 1,061 references to masks through 691 total tweets.

Masks as Political Signifiers

Science shows that masks are a simple and effective way to contain respiratory droplets, thereby reducing the spread of coronavirus (Chu et al., 2020; Eikenberry et al.,

14 2020). However, conservative politicians, pundits, and activists have resisted mask mandates and a range of social distancing regulations on the grounds that these measures ostensibly suppress individual liberties, overstate the seriousness of Covid-19, and directly and indirectly hurt the economy. This resistance has played out in dramatic fashion with armed protestors converging on public buildings—mask-less and in close physical proximity—to assert their demands. These anti-regulation protests and in particular, resistance to masks, have come to signify a conservative political position at-odds with the racially conscious, collectively oriented message of the BLM movement, and the alignment of this movement with progressive positions on public health mandates. Masks thus serve as both a symbol of progressivism and, as we show here, an artifact through which critiques of White conservativism materialize.

Just remember. The same people who will say #AllLivesMatter in response to #GeorgeFloyd are the same ones who don’t want to wear a mask; says the #coronavirus is just the flu and believe the deaths of 100k Americans to #COVID19 shouldn’t interfere with their ‘freedoms’

The same white men who refuse to wear a mask around other people also deny that white privilege exists. #GeorgeFloyd #BlackLivesMatter #minneapolisriots #whiteprivilege #mask #coronavirus

Mask refusal is positioned here in alignment with White privilege, colorblind racism (e.g.,

“#AllLivesMatter), and related indifference to the loss of life caused by viral spread, drawing on and mobilizing the political meaning of these small material implements and the ideological currents they have come to represent.

Masks as Technologies of Resistance

Over the past decade, social movements have been marked by their use of social media to organize, broadcast, and galvanize action (Boulianne, 2019; Christensen, 2011; Earl and Kimport, 2011; Schradie, 2019; Tufekci, 2017). In contrast, the most striking feature of the BLM protests following Floyd’s death was activists’ physical co-presence. These protests took place while the virus surged, and while whole cities were shut down to minimize person-

15 person contact and mitigate viral spread. Cognizant of this tension, protesters were vocal about safety measures and placed a particular emphasis on masks. Masks thus became a technology of risk control that made protest participation possible, with activists issuing regular reminders about the norm and the need, to mask-up.

WEAR YOU MASKS! Protest to picnics. Wherever you are WEAR YOUR MASK! We still in a pandemic. #protests #pandemic #COVID19 #love #BlackLivesMatter #BeWell #BeSafe #GeorgeFloydProtests #GeorgeFloyd

It's a really scary time to be black. It's a really scary time to be a POC [person of color]. It's a really scary time to be.. well, anyone in these protests who are standing up for #GeorgeFloyd and #BlackLivesMatter Stay safe. Wear your mask. We need everyone strong. #coronavirus #LAProtests

With masks operating as symbolic signifiers of life under the pandemic and political signifiers of partisan ideology, they were also a vital material tool for protest participation, enabling protestors to coalesce while staving off the dangers of close physical proximity5.

Voice

We situate voice as an umbrella category that captures the contested meanings of street protests following Floyd’s death. Because mobilization was so central to the Floyd incident, it is unsurprising that it blankets the dataset, with 16,421 thematic references over 9,261 tweets. With these data, we find competing narratives about the protests themselves, whose voices are heard, and whose are stifled. More specifically, we analyze claims about the

(ir)responsibility of physically protesting during a pandemic, and narratives about how BLM protests and anti-lockdown protests have been variously supported and condemned by institutional authorities.

Social (Ir)Responsibility

As discussed in the previous section about masking as a practicality of protest participation, the physical co-presence of BLM activists during Covid-19 created a potential

5 Masks have long played a multifaceted role in protest movements, protecting activists from both tear gas and surveillance. This is highlighted most recently by the pro-democracy protests in Hong Kong in the months just before Covid-19 hit, during which government forces enacted mask bans. 16 health risk. Protestors cited this risk as not only acceptable, but necessary in the context of systemic racial violence, while detractors in the dataset pointed to the irresponsibility of these actions and the hypocrisy these actions imply. We can thus observe competing claims about the riskiness of bodies in shared physical space and the relationship between this risk and broader social imperatives.

BLM activists and allies cited the risk of physical co-presence as evidence of protestors’ commitment to the cause and positioned this risk as less troublesome than the systemic violence enacted against communities of color. Tufekci (2017) points out that protests can be measured by the power they signal, indicated by how effectively they rise up despite external forces that should stifle mobilization. Physically gathering during a pandemic thus showed the urgency of the BLM cause and the power and tenacity of protest participants.

White supremacy and anti-Blackness are public health threats, b/c of police violence and failed pandemic response. We *have* to protest because BLM. Cities should spend $$ on testing and community health & support, not policing. https://t.co/Sg24VIOj1Z [Link to article about protesting during a pandemic]

They showed up during #Covid19 This speaks volumes! All of these ppl are risking a lot to fight for black people and #GeorgeFloyd who lost his life. This shyt is beautiful! https://t.co/lTx0Da1r0M [Four images of BLM protestors]

People have decided that potentially exposing themselves to #COVID19 is less important than seeking justice for #GeorgeFloyd and so many others. If you are not in awe, you are ignoring their agony

In contrast, detractors accused protestors of being hypocrites for their vocal support of physical distancing and subsequent assembly in public squares and city streets. These claims tied-in with doubts about the true contagion of Covid-19 and antagonism towards public health measures. They also worked to discredit the voices of protestors and undercut the validity of the BLM cause.

If you were wondering what would trigger the 2nd wave of #covid19, now you know. #GeorgeFloyd https://t.co/mPOLFrloGd [Image of crowded street protests]

17 If you disapprove of people protesting the #Michigan lockdown than surly you must disapprove of #BlackLivesMatter protesting the #GeorgeFloyd case. Either protests spread #Coronavirus or they don't, stop being hypocrites.

We observe in these tweets Covid-19 serving competing rhetorical ends vis-à-vis street protests. For BLM supporters, the dangers of physical co-presence signal urgency and collective power. For detractors, physical protests during the pandemic undercut protestors’ credibility and undermine the mainstream public health messaging that has become an ideological fracture point.

Institutional Suppression

Many of the Floyd protests were met with militarized force, including tear gas, rubber bullets, and deadly gunfire. Activists and allies in the dataset juxtaposed this treatment to the

(non)response of law enforcement during anti-lockdown protests occurring just before and continuing concurrently with, BLM mobilizations. At the same time, critics juxtaposed the sympathetic response by the media and some public officials to the BLM movement as compared with harsh admonishments against those who flout and/or actively oppose physical distancing measures.

BLM activists and allies on Twitter contrasted their suppression with the noninterventionist treatment of anti-lockdown protestors. Tweets making this comparison often highlighted the racial demographics of each respective group, thus generating a meta- statement on systemic racism and its manifestation through authoritarian violence.

We no longer need to engage in hypothetical "what ifs." White protestors, armed with rifles, aggressively confront the police to defend their right to a speedy haircut. Nothing. Unarmed Black protestors congregating at the site of a police murder. Tear gas and riot police

Protesters of Black causes and anti-police brutality, met with military gear&assault. #GeorgeFloyd #RIPGeorgeFloyd #GeorgeFloydWasMurdered. Below are angry White thugs protesting the rights to infect their fellow citizens with #coronavirus https://t.co/TJfUsVUIWd [Side-by-side image of anti-lockdown protests and Floyd protests]

18 In turn, anti-lockdown activists and BLM detractors critiqued the media, law enforcement, and key political figures, alluding to a larger narrative about liberal elitism and the alleged progressive biases of mainstream news.

@Julio_Rosas11 Hey, @GavinNewsom and @MayorOfLA you gonna go on TV and scold these people about social distancing? If you can't sit on a beach peacefully should you be able to riot, burn, loot and assault? #COVID19 #minneapolisriots #GeorgeFloyd

For past month, media attacked peaceful anti-lockdown protesters nationwide for "endangering" public. Silicon Valley de-platformed them. Many jailed. Now, Minnesota rioters are burning buildings, robbing stores & hurling incendiary devices at firemen & cops. How many arrests??? https://t.co/IaBxgrc8pX [Video of BLM protests with buildings on fire in the background]

Both BLM activists and anti-lockdown advocates made claims that their voices were stifled by powerful institutions. For BLM, the suppression came from federal authorities and police. To those opposing shutdowns, the alleged suppression came from mainstream media outlets and select political figures. For each, the injustice of their own suppression was positioned against concessions for the other.

CONCLUSIONS

The Covid-19 pandemic and Black Lives Matter movement are two distinct phenomena that converged acutely with the death of George Floyd at the end of May 2020, just as the United

States was surpassing 100,000 virus related deaths. Taking this juncture as our point of analysis, we used Twitter data to understand how a virus and a movement came to mutually shape each other.

Social media both reflect and affect social realities. Examining Twitter data with simultaneous hashtags pertaining to George Floyd and Covid-19 provides a lens on a significant historical moment through the prism of these two highly salient events. More broadly, the data illuminate the interlocking nature of personal troubles and public issues

(Mills, 1959), and of public issues and each other, forged through the connection between the

19 death of a man and death rates of millions; in the multifaceted ways race informs corporeal vulnerability; in the imbuing of medical masks with both political significance and instrumentality; and in the dialogue between indirectly related protest movements and their competing claims about institutional responses to individual and collective action.

Findings show these complex issues navigated through a repeating set of symbolic hooks: death, breath, masks, and voice. Moored by these symbolic hooks, the tweets in our dataset reveal a cultural moment poised for transformation, hastened and intensified by the events under study. Saturated with morbidity and sharp with political divides, these data capture meaning-in-process as everyday people and public figures collectively and combatively construct the defining narratives of this time. Though brief and ephemeral, each digitally mediated micro-statement became part of a living record, making up the material— however partial—from which history will be sculpted. In the preceding text, we documented that material and parsed its sociological significance in terms of race, mobilization, institutions, and ideology.

In addition to substantive findings, our study demonstrates a novel methodological approach to social media data—hashtag convergence— using co-occurring hashtags to understand their sociological confluence. This approach is valuable for the study of social life via social media, as empirical events never occur in a void, but always in conjunction with myriad other happenings. History is full of intersecting matters—e.g., the Vietnam war and racial integration, women’s suffrage movements and prohibition, eugenics and the birth control pill. Historians have carefully demonstrated the co-constitution of these phenomena, weaving together disparate documents, stories, and physical artifacts. Today, social media offers a new kind of analyzable artifact, recording the words and claims of mass publics in real time, organized and framed through hashtag registers. Hashtag convergence, which captures users’ organic pairings of otherwise discrete tags, works to identify and parse

20 sociohistorical couplings, creating a repository from which researchers can draw grounded, sociological insights.

As with all studies, ours has qualifying factors that should be considered when inferring from the findings. First, our use of Twitter data is not generalizable to the overall population, nor even generalizable to the entirety of Twitter users. The ~20,000 tweets we analyzed represent a fractional snapshot of publicly available, English language content, shared on one particular platform. This does not undermine the value of the data as a lens on a specific historical moment, but it does act as a reminder that our analysis is a lens, rather than the full holistic picture. A second qualifying factor is our use of only two co-occurring hashtags, rather than multiple. In the course of scraping data, we picked up on myriad other issues (e.g., global economic downturns, a failed space launch, an upcoming presidential election, democracy protests in Hong Kong), all of which contributed to the broader historical context. Although we decided on a two-topic analysis for purposes of scope, including more than two hashtags would have generated differently nuanced findings.

Qualifying factors necessarily create bounds around the applicability of the work, while providing opportunities for future research. Empirically, researchers might take our initial findings and build a fuller dataset that incorporates multiple social media platforms, investigating how the affordances and dynamics of different services and communities compare, contrast, and intertwine to capture and create this particular cultural story. Future research may also add additional hashtags to the ones we selected, exploring a complex ecology of events. Finally, our study represents the first instance of the hashtag convergence method. This approach can be further developed in conversation with other methodological advances in big data, small data, and social media studies and applied to myriad cases in which distinct phenomena intersect.

21 Table 1: Thematic Hooks

Category Total Total References Tweets and % of Corpus* Death 8,059 5,980 (30%)

Breath 2,705 1,292 (6%)

Masks 1,160 691 (4%)

Voice 16,421 9,261 (46%)

*Percentages equal less than 100 because the dataset contains tweets that do not adhere to any discernable pattern (e.g., idiosyncratic communications between individual twitter users) and/

22 or were not codable in the context of this study (e.g., shared links with no textual additions from the tweeter, advertisements appended with the hashtags).

References

Ruha Benjamin, 2020. “Black skin, white masks: Racism, vulnerability & refuting black pathology.” Lecture for Department of African American Studies, Princeton University, April 15, 2020. https://aas. princeton. edu/news/black-skin-white-masks- racism-vulnerability-refuting-black-pathology. Peter L. Berger and Thomas Luckmann, 1966. The social construction of reality: A treatise in the sociology of knowledge. New York: Anchor. Yarimar Bonilla and Jonathan Rosa, 2015. “Fergusion: Digital protest, hashtag ethnography, and the racial politics of social media in the united states,” American Ethnologist, volume 42, number 1, pp. 4-17. Eduardo Bonilla-Silva, 2020. “Color-blind racism in pandemic times,” Sociology of Race and Ethnicity Doi: 2332649220941024. André Brock, 2018. “Critical technocultural discourse analysis, New Media & Society, volume 20, number 3, pp. 1012-1030.

23 AndréBrock, 2012. “From the blackhand side: Twitter as a cultural conversation, Journal of Broadcasting & Electronic Media, volume 56, number 4, pp. 529-49. Shelley Boulianne, 2019. “Revolution in the making? Social media effects across the globe,” Information, Communication & Society, volume 22, number 1, pp. 39-54. Amanda M. Bunting and Janet Stamatel, 2019. “Exploring geospatial characteristics of hashtag activism in Ferguson, Missoura: An application of social disorganization theory, Geoforum, volume 104, number 1, pp. 55-62. Axel Bruns and Jean Burgess, 2015. “Twitter hashtags from ad hoc to calculated publics,” In: N. Rambukkana (editor). Hashtag Publics: The power and politics of discursive networks. New York: Peter Lang Publishing, pp. 13-27. Axel Bruns, Brenda Moon, Avijit Paul, and Felix Münch, 2016. “Towards a typology of hashtag publics: A large-scale comparative study of user engagement across trending topics,” Communication Research and Practice, volume 2, number 1, pp. 20-46. CDC, 2020. "United States coronavirus death toll surpasses 100,000,” Washington, DC: Centers for Disease Control and Prevention, https://www.cdc.gov/ media/releases/2020/s0528-coronavirus-death-toll.html#:~:text=United%20States %20Coronavirus%20(COVID%2D19)%20Death%20Toll%20Surpasses %20100%2C000,-Media%20Statement&text=Today%20the%20number%20of %20people,been%20lost%20to%20this%20disease Christian Christensen, 2011. “Twitter revolution? Addressing social media and dissent,” The Communication Review, volume 14, number 3, pp. 155-57. Derek K. Chu, Elie A. Akl, Stephanie Duda, Karla Solo, Sally Yaacoub, Holger J. Schünemann, Amena El-harakeh, Antonio Bognanni, Tamara Lotfi and Mark Loeb, 2020. “Physical distancing, face masks, and eye protection to prevent person-to- person transmission of Sars-Cov-2 and Covid-19: A systematic review and meta- analysis,” The Lancet, volume 395, number 10242, pp. 1973-1987. Jennifer Cobbina and Ashleigh LaCourse, Erika J. Brooke, and Soma Chaudhuri, 2021. “Protesting during a pandemic: Narratives on risk taking and motivation to participating in the 2020 March on Washington,” Crime & Delinquency Doi: 0011128721999333. Patricia Hill Collins, 2002. Black feminist thought: Knowledge, consciousness, and the politics of empowerment. New York: Routledge. Julia Craven, 2020. “How racial health disparities will play out in the pandemic,” Slate, https://slate.com/news-and-politics/2020/03/how-racial-health-disparities-will-play- out-in-the-coronavirus-pandemic.html Nick Couldry and Andreas Hepp, 2018. The mediated construction of reality. New York: John Wiley & Sons. Jenny L. Davis and Tony P. Love, 2018. “Women who stay: A morality work perspective,” Social Problems volume 65, number 2, pp. 251-265. Jenny L. Davis and Tony P. Love, 2019. “Generalizing from social media data: A formal theory approach,” Information, Communication & Society, volume 22, number 5, pp. 637-647. Delan Devakumar, Geordan Shannon, Sunil S Bhopal, and Ibrahim Abubakar, 2020. "Racism and discrimination in Covid-19 responses,” The Lancet, volume 395, number 10231, pp.1194. Emile Durkheim, 1897/1952. Suicide: A study in sociology. London: Routledge. Jennifer Earl and Katrina Kimport, 2011. Digitally enabled social change: Activism in the internet age. Cambridge, MA: MIT Press. Steffen E. Eikenberry, Marina Mancuso, Enahoro Iboi, Tin Phan, Keenan Eikenberry, Yang Kuang, Eric Kostelich, and Abba B. Gumel, 2020. “To mask or not to mask:

24 Modeling the potential for face mask use by the general public to curtail the Covid-19 pandemic.” Infectious Disease Modelling, volume 5, number 2020, pp. 293-308. Emilio Ferrara, 2020. “#Covid-19 on Twitter: Bots, conspiracies, and social media activism,” arXiv preprint arXiv:2004.09531. Sara Fischer, 2020. “Social media use spikes during pandemic,” Axios, https://www.axios.com/social-media-overuse-spikes-in-coronavirus-pandemic- 764b384d-a0ee-4787-bd19-7e7297f6d6ec.html Deen Freelon, Charlton D. McIlain and Meredith Clark, 2016. “Beyond the hashtags: #Ferguson, #Blacklivesmatter and the online struggle for offline justice,” Center for Media & Social Impact, American University, https://cmsimpact.org/resource/beyond-hashtags-ferguson-blacklivesmatter-online- struggle-offline-justice/ Ryan J. Gallagher, Andrew J. Reagan, Christopher M. Danforth, and Peter Sheridan Dodds, 2018. “Divergent discourse between protests and counter-protests: #Blacklivesmatter and #Alllivesmatter. PLOS ONE, volume 13, number 4, Doi: https://doi.org/10.1371/journal.pone.0195644 Alicia Garza, 2014. “A herstory of the #Blacklivesmatter movement,” In: J Hobson (editor). Are all the women still white? Rethinking race, expanding feminisms. New York: SUNY Press, pp. 23-27. Guardian, 2020. “The Guardian view on the death of George Floyd: A turning point? The Guardian, https://www.theguardian.com/commentisfree/2020/jun/01/the-guardian- view-on-the-death-of-george-floyd-a-turning-point Anton Gollwitzer, Cameron Martel, William J. Brady, Philip Pärnamets, Isaac G. Freedman, Eric D. Knowles and Jay J. Van Bavel, 2020. “Partisan differences in physical distancing are linked to health outcomes during the Covid-19 pandemic,” Nautre Human Behavior, volume 4, number 11, pp. 1186-1197. Susan Halford and Mike Savage, 2017. “Speaking sociologically with big data: Symphonic social science and the future of big data research,” Sociology, volume 51, number 6, pp. 1132-1148. Andreas Hepp, Andreas Breiter, and Thomas N. Friemel, 2018. “Digital traces in context—an introduction,” International Journal of Communication, volume 12, number 2018, pp. 439-449. Yulin Hswen, Xiang Xu, Anna Hing, Jared B. Hawkins, John S. Brownstein and Gilbert C. Gee, 2021. “Association of #COVID19 versus #Chinesevirus and anti-Asian sentiments on Twitter: March 9-23,” American Journal of Public Health, volume 11, number 5, pp. 956-964. Monica Hooper, Anna María Nápoles and Eliseo J Pérez-Stable, 2020. "Covid-19 and racial/ethnic disparities,” JAMA, volume 323, number 24, pp. 2466-2467. Jelanie Ince, Fabio Rojas, and Clayton A. Davis, 2017, “The social media response to Black Lives Matter: How Twitter users interact with Black Lives Matter through hashtag use,” Ethnic and Racial Studies, volume 40, number 11, pp. 1814-1830. Rachel Kuo, 2018. “Racial justice activist hashtags: Counterpublics and discourse circulation,” New Media & Society, volume 30, number 2, pp. 495-514. Shana Kushner Gadarian, Sara Wallace Goodman, and Thomas B. Pepinsky, 2020. “Partisanship, health behavior, and policy attitudes in early stages of the Covid-19 pandemic,” SSRN https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3562796. Guillaume Latzko-Toh, Claudine Bonneau, and Melanie Millette, 2017. “Small data, thick data: Thickening strategies for trace-based social media research,” In: L. Sloan and A. Quan-Hasse (editors). The Sage Handbook of Social Media Research Methods. London: Sage, pp. 199-214.

25 Deborah Lupton, Clare Southerton, Marrianne Clark, and Ash Watson, 2021. The face mask in Covid times: A sociomaterial analysis. Berlin/Boson. Walter de Gruyter GmbH & Co KG. Christopher K. Marshburn, Abigail M. Folberg, Chelsea Crittle, and Keith B. Maddox, 2021. “Racial bias confrontation in the United States: What (if anything) has changed in the Covid-19 era, and where do we go from here?,” Group Processes & Intergroup Relations, volume 24, number 2, pp. 260-269. C. Wright Mills, 1959. The sociological imagination. Oxford, UK: Oxford University Press. National Center for Health Statistics, 2016. “Health, United States, 2015: With special feature on racial and ethnic health disparities,” https://pubmed.ncbi.nlm.nih.gov/27308685/ Tina Nguyen, 2020. “Conservatives charge liberals with social-distancing hypocrisy,” Politico, https://www.politico.com/news/2020/06/06/conservatives-charge-liberals- with-social-distancing-hypocrisy-304435 Xin Pei and Deval Mehta, 2020. “#Coronavirus or #Chinesevirus?!: Understaning the negative sentiment reflected in tweets with racist hashtags across the development of Covid-19,” arXiv preprint arXiv:2005.08224. Rashawn Ray, Melissa Brown, and Wendy Laybourn, 2017. “The evolution of #Blacklivesmatter on Twitter: Social movements, big data, and race,” Ethnic and Racial Studies, volume 40, number 11, pp. 1795-1796. Timothy Recuber, 2021. “Race, racism and mnemonic freedom in the digital afterlife,” Information, Communication & Society, volume 24, number 5, pp. 684-699. Neal Rothschild, 2020. “The massive power of the George Floyd protests,” Axios, https://www.axios.com/george-floyd-racial-inequality-protests-comparison- d7c32b17-3078-4c62-9d6fee70b6c37cb1.html? fbclid=IwAR0snelnaw78LQanNOSQwsXpZE2V2rppEaDKoaXXYRW30knRMoP6 B6SNCIk Madeliene Schachter, 2020. “Black Lives Matter and Covid-19,” The International of Journal of Information, Diversity, & Inclusion (IJIDI), volume 4, number 3/4, pp. 81- 86. Jen Schradie, 2019. The revolution that wasn’t: How digital activism favors conservatives. Cambridge, MA: Harvard University Press. Lisa Singh, Shweta Bansal, Leticia Bode, Ceren Budak, Guangqing Chi, Kornraphop Kawintiranon, Colton Padden, Rebecca Vanarsdall, Emily Vraga, and Yanchen Wang, 2020. “A firsr look at Covid-19 information and misinformation sharing on Twitter,” arXiv preprint arXiv:2003.13907. Luke Sloan and Anabel Quan-Hasse, 2017. “A retrospective on state of the art social media research methods: Ethical decisions, big-small data rivalries and the spectre of the 6vs,” In: L. Sloan and A. Quan-Hasse (editors). The Sage hanbook of social media research methods. London: Sage, pp. 662-672. Iddo Tavory and Stefan Timmermans, 2014. Abductive analysis: Theorizing qualitative research. Chicago: University of Chicago Press. Michael Thelwall and Saheeda Thelwall, 2021. “Twitter during Covid-19: George Floyd opening a space to address systematic and institutionalized racism? SSRN https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3764867 Stefan Timmermans and Iddo Tavory, 2012. “Theory construction in qualitative research: From grounded theory to abductive analysis,” Sociological Theory, volume 30, number 3, pp. 167-186. Zeynep Tufekci, 2017. Twitter and tear gas: The power and fragility of networked protest. New Haven: Yale University Press.

26 Philipp Wick and Marianna M. Bolognesi, 2020. “Framing Covid-19: How we conceptualize and discuss the pandemic on Twitter,” arXiv preprint arXiv:2004.06986. Michele Zappavigna, 2018. Searchable talk: Hashtags and social media metadiscourse. New York: Bloomsbury publishing.