We are not the cultural consumers we used to be. Data, streaming, and Web 2.0 have remade how we read and how we watch. Platforms are the new publishers. But although we consume culture differently now, much of how we talk about and study it remains lodged in the analog world of the 20th century. It’s time for our methods to catch up with our objects. Born digital culture requires a born digital approach. Hacking the Culture Industries showcases the power of data-driven cultural criticism, and reinvigorates cultural studies for the 21st century. These five new essays move between book culture, streaming TV, social media, and online writing platforms: Squid Game and streaming hits; Goodreads and romance fiction; Twitter and hive-critique; Tik Tok and cultural attention; and who gets to decide who wins book prizes in the age of social networks. This series takes up a call we issued a year ago: to hack the culture industries. To challenge their dominance by using their data to study them and their stranglehold on cultural production. To tell new stories about culture in a time of ubiquitous data.
—Laura B. McGrath, Dan Sinykin, and Richard Jean So
The most tweeted-about show of the decade is HBO’s teen drama Euphoria. During the Gen Z fever dream’s second season, Twitter (now X) users sent out 30 million Euphoria-related tweets. While airing, live reactions were so abundant that one user tweeted: “watching euphoria! (scrolling through twitter).” Indeed, weekly episode viewings became entangled with the live Twitter response to the show. Viewers—with eyes on both screens at once—participated through viral memes and baiting reactions, as scenes ticked away.
Call it hive critique: a collective interpretation of streaming fiction. Viewers can read thousands of opinions on a scene even as it’s airing. It’s a newly participatory form of watching, and a new forum for narrative discussion.
So, what do viewers do within this forum? Studying these bite-size responses to TV gives us a window into how narrative is discussed in this hyperactive mode. Euphoria, like much of contemporary media, transcends its original format. Its narrative is repurposed by viewers and reworked into media bits across platforms, what Henry Jenkins calls “spreadable media.” Among the proliferating memes, a cursory glance at popular Euphoria tweets may not seem to hold much substantive narrative commentary.
Yet, amid the frenzy on Twitter, I get a sense of the narrative elements that prompt response: the characters users claim to feel close to, the moments that shock them, the plot twists they wish had turned out differently. If we examine responses systematically using natural language processing techniques, what do these millions of tweets teach us about hive critique? Through a computational analysis, I uncover how viewers gossip about Euphoria’s teenage characters: emotionally cleaving to them, while also virtue signaling through their tribulations.1
Euphoria’s high schoolers give viewers a lot of material to work with. It aims to be progressive and transgressive, featuring frank portrayals of drug abuse, multiple queer relationships, sex work, and a transgender actor. It is narrated by pillhead Rue Bennett, played by Zendaya. Rue guides viewers through her friends’ twisting, illicit plotlines. The glittery scenes of messy drama are hard to look away from, with their continuously swelling and crashing waves of hijinks. 19 million viewers, primarily between ages 18 and 34, tuned into the second season in 2022. Of HBO shows, only Game of Thrones exceeded Euphoria’s viewership while on air.
Much of the media attention paid to Euphoria’s Twitterverse has focused on popular memes born from the show. Early in the second season, a widely reworked meme template compared the sexy outfits of Euphoria’s teens to viewers’ own high school wardrobes.
In iterating on memes based on screen stills, viewers fold the characters into the networked world, finding points of relatability to their own lives or to characters in other diegetic universes. Characters become potent and portable, extracted from their original narrative bounds.
While many of these responses are lighthearted, the emotional investment in the show is exaggerated, as the language of fandom commonly is. Alongside meme making, Euphoria tweet-and-watchers are “crying, screaming, and throwing up” over the plot’s chaotic turns. Viewers affect anxiety-induced mania, or pillory writer-director Sam Levinson for writing out a character.
To find out what Euphoria’s tweets can teach us about hive critique, I focus on the set sent during the live broadcast, those hurriedly typed reactions that showcase audience affinities and dispositions. HBO broadcast Euphoria during its coveted Sunday evening time slot, 6:00 p.m. Pacific, 9:00 p.m. Eastern. I pull a corpus of tweets sent during each episode in the second season. This provides me with a total of one million tweets, posted by about 400,000 users.
First, a straightforward finding. During broadcast time, 70 percent of tweets center on character, and viewers tweet about those characters who are actively onscreen as the show progresses, giving their take on a character’s actions in real time. Rather than engage in debates that depart dramatically from the narrative, or offer aesthetic judgments, viewers primarily respond to character.
That viewers focus on character tells us what we already know from literary theorists like John Frow: character drives much of our interest in narrative. The tweets show that viewers strongly identify with these eight Euphoria high schoolers: they praise their good vibes, critique their misbehavior, empathize with them during difficult moments. Their responses, like the tweet below, highlight how seamlessly we discuss characters as if they are real, while still knowing that they are not, as Toril Moi has put it.
Leaning on Rita Felski’s work on identification, I dig into one major vector of character discussion online: allegiance, or felt affiliation and solidarity, and its opposite, a kind of moral distancing. In the latter, dissociation from a character generates a viewer’s attachment.
I focus on allegiance because, in this corpus, interest in the ethical lives of characters drives response. When a character makes a decision that is deemed ethically questionable, like engaging in drug use or cheating on a partner, Euphoria tweet counts spike on average by 30 percent. Some of these responses simply express surprise or shock; but these scenes also present an opportunity for the viewer to provide ethical critique or stand in allegiance.
In her book Simple Passion, Annie Ernaux writes that fiction should induce a suspension of moral judgment. Euphoria twin-screen viewers, at large, do not oblige, even while character critiques are made alongside aesthetic judgments. Critics have recently lamented an overindexing on moral judgment in narrative reception. Quickening moral judgments, they say, can diminish the joy of the aesthetic encounter, the willingness to sit with ambiguity, the practice of tolerance. While narrative inevitably engages our moral pulse, that viewers feel quickly compelled to share their judgments is shaped by a social platform’s affordances.
After all, what drives viewers to post about TV—which is to say, when they tweet, most of the time, about characters? Tweeting while watching is an instinct driven by some promise from the platform. Yes, active fans may seek digital company via a massive watch party (though half of these tweets have no active engagement; that is, they aren’t liked or retweeted by other users). At the same time, posting works to cultivate a public narrative about oneself. When a user shares their shifting identifications, they share something about who they are—who they are most similar to, who they adore, who they condemn. A moral judgment may offer a satisfying signal about the self.
To understand the ethical dissection of character online, I ran the Twitter corpus through a deep learning network trained to detect moral outrage in tweets. The most discussed character in Euphoria, Cassie, is also one half of a duo that drives major ethical analysis in the second season. Over a third of tweets sent out about Cassie are classified as portraying moral outrage. On Twitter, rather than showcase moral tolerance, viewers jump to appraise her.
What is it about Cassie that draws this attention? Cassie begins the second season by entering into a secret, terribly tumultuous relationship with her best friend’s boyfriend, Nate. Euphoria is chock-full of this type of scandal. Before season two begins, Cassie is largely supported by the online audience. Season one tweets show that viewers most often claim they love Cassie, feel bad for her, and that she deserves better. By the middle of the second season, users most commonly describe Cassie as embarrassing, pathetic, and annoying. In the show, Cassie’s character betrays her best friend and spends her time perfecting outfits and skin-care routines in order to attract Nate’s attention. Cassie falls into a stereotype: a woman tragically risking it all for the false promise of love. These primary descriptors indicate that many viewers distance themselves from her.
Not everyone turns against Cassie, though. When one user sends out the above tweet after Cassie and Nate’s infidelity begins, over a thousand people respond critiquing this response and portraying allegiance with Cassie: “How are you going to blame the emotionally vulnerable girl who’s struggling to break a toxic relationship. …” These viewers use Cassie’s difficult upbringing and the popular language of trauma response, or Nate’s manipulation of her, to defend her character. In allegiance with or against Cassie, virtue signaling seeps into discourse. But what is the goal of virtue signaling?
When talking about character, Twitter users gossip or manage the reputations of characters and react to their decisions. As has been outlined by Blakey Vermeule, gossiping through fiction provides moral comfort; when viewers condemn Cassie on Twitter, they distance themselves from her, as if to assure others that they are not like her, that they wouldn’t find themselves in the bathroom with Nate. Alignment with narrative reflects our beliefs about people; and tweeting about character identification involves self-cultivation, a narrative crafting. Those who protect Cassie from scorn perhaps mean to convey empathy but also their own nuanced understanding of what Cassie has been through. In gossiping, our reading of another reveals information about ourselves.
That fiction can guide us toward self-understanding is a major pillar in Euphoria’s character development, whether that be an online persona or through imagined intervention. The Euphoria teens develop fictional versions of themselves to facilitate their self-knowledge or, inversely, to provide a comforting alternative reality to the one that they embody. “Real life is always such a letdown … It’s just, like, easier to talk to people online,” Jules says to a therapist in a special episode. While Jules later does acknowledge online performativity, that doesn’t reduce its draw. The internet offers a respite, where the characters can exercise control over their narrative and feel a sense of belonging with others.
It may be satisfying to resound in an affective echo chamber, to belong to a community of feeling; even if that community is We hate Nate Jacobs.
In the second season, two episodes are devoted to showcasing a play that Cassie’s younger sister, Lexi, writes, retelling the story of Euphoria from her own perspective. Once a “sidekick,” she now writes herself as the main character. Viewers find Lexi most relatable: one viewer writes “I think we’re all Lexi”; as if to say, aren’t all of us here trying to find a way into the story?
The plot showcases the kind of self-cultivation through molding of character that appears in response to the show on Twitter, in both the aforementioned meme template and viewers’ professed ethical stances. In front of a constant audience, posters are continuously crafting; watching just provides another opportunity.
What happens when, as viewers manage their own response to character, they simultaneously watch as characters come alive in other minds online? We should not diminish audience member autonomy. Still, on Twitter, actively ingesting and sharing identifications with character is inevitably shaped by the live response of the rest of the collective audience. We performatively remake ourselves in the image of the social network.
To understand the claims characters place on the audience, we also need to understand their emotional effects. While data-driven sentiment models are imperfect, they nonetheless offer us some insights into emotion embedded in textual response online. Applying the latest deep learning sentiment models to the corpus shows that the public emotional response to character is consistent; as in, when a character is onscreen, the emotions present in the audience’s tweets are highly predictable. Audiences tend to adopt a common emotional response to a character’s presence. For example, when Nate is onscreen, we will see tweets that portray disgust (he makes me feel sick).
A sample distribution of the emotional categories computationally assigned to tweets that reference a given character in the corpus.
That emotional response to character on Twitter is largely consistent does not indicate that viewers feel the same at home on their couch. Posting online involves showcasing oneself for a public audience. Simultaneously, actively reading tweets while viewing affects the judgments viewers share. Just as we gossip in the real world, what others say about someone is sure to impact our own understanding of them. It may be satisfying to resound in an affective echo chamber, to belong to a community of feeling; even if that community is We hate Nate Jacobs.
Watching and posting becomes, in part, a bonding exercise around judging character. These affective responses are enmeshed in the ethical opinion, reflecting who the community deems acceptable. A show’s extreme virality could be tied to characters that viewers can reflect themselves off of; a set of characters that are believable, stereotypes that viewers have encountered before, that viewers are discerning how to respond to again.
Here, I have shown just a small slice of the ecosystem of television consumption online. But, these results suggest that social media’s affordances shape viewer response to Euphoria. As hive critique absorbs the show into the social world, the characters are brought into the tribunal of our media landscape, and, I suggest, used as demonstrative tools.
Social psychology researchers found that political tweets conveying a moral message are more likely to go viral. The evaluative landscape of social media has pervaded response to narrative as well, and twin-screen viewers are particularly attuned to moral issues within and around a work.
The moral work of narrative, Garth Greenwell writes, “is to teach us how much richer and more capacious our engagement with others can be.” Agree with the claim or not, it contains something fundamental: narrative fiction necessitates interiority plus processing time. Yet both requirements are curbed by social media and its impulse toward immediacy, as the narrative is scrapped for parts.
Perhaps the real hazard isn’t viewers’ moral capaciousness; after all, they may wrestle with complexities offline. What is more worrisome is that content studios will optimize storytelling for social media engagement. 
