The July 7 open letter published in Harper’s Magazine and signed by 153 writers, artists, and intellectuals, was widely understood to be a brief in favor of “free speech.” The letter warned that the “free exchange of information and ideas … Is daily becoming more constricted.” But legions of critics including many other writers, artists, and intellectuals, asserted that the opposite is true. More people than ever, they countered, now have access to public platforms for their speech. In their view, recent examples of people being fired or forced to resign due to social media campaigns are not “cancel culture,” but proof of expanded expressive freedom. Now, average people who witness injustice or disagree with an influential New York Times op-ed writer have the means to call out the injustice and contest those ideas in a public forum like Twitter.
It has the appearance of a pitched debate but, in fact, the letter writers and their critics complement each other as two sides of the same digital coin. The new voice given to those previously shut out of debate and celebrated by critics of the letter is the direct result of social media platforms reengineering the American political environment. At the same time, the “stifling atmosphere” deplored by the more traditional liberals who signed the letter and tend to criticize “cancel culture,” is also a climate created by platforms like Twitter. Regardless of which side wins any particular battle in the recurring speech wars, both parties to the conflict end up reinforcing the power of the overall system in which the drama is enacted. And so a pattern emerges that is larger and more consequential than the specifics of the latest political flare-up. It is not the arguments or ideas of any political group, but the structure of the digital platforms that sets the tone of the culture as a whole.
And what is the structure? It is an arena for perpetual conflict driven by an accumulation of grievances collected in a mass program of decentralized surveillance. We are incentivized, by the coded logic of the social media platforms where public engagement now takes place, to find reasons to hate each other. The algorithms that encourage and reward particular behaviors on Twitter and Facebook play on our deepest human instincts and desires to create spectacles of symbolic violence and sacrifice. Much of the time, the violence and spectacle has the appearance of a game or a light amusement. To take it too seriously, therefore, is to risk being an alarmist, and likely of the reactionary sort. But it is precisely the gamelike aspect of the platforms that keeps us playing. Playing and paying because the point, finally, is profit.
The conflicts taking place over freedom, justice, and other noble ideas are captive performances in the most technologically advanced human cockfighting enterprise ever designed—one that has converted the essence of human struggle into a sure-win bet for the tech platforms who play the house.
Along with the expanded freedom provided by the new platforms for speech and self-expression has come a constant vigilance by potentially hostile parties on the lookout for transgressions. This dynamic is neither specific to older cultural institutions nor to the online forums that now seem to dictate their decisions. Recent decades have witnessed a broad shift toward open networks controlled by distributed surveillance. For instance, prisons can be downsized because former inmates can be tracked with GPS-tagged ankle bracelets. The punch card has given way to diffuse workplace monitoring, enabling the growth of remote work. As schools and colleges move instruction online, students’ liberation from institutional enclosure in a physical location will go hand in hand with technological systems that in some ways circumscribe their lives far more thoroughly.
Social media freedom likewise comes at the cost of continuous supervision. Tagging posts with location metadata facilitates the monitoring of physical movement. In parallel to the physical location tracking, digital platforms function as a social and ideological positioning system. This makes them useful tools for the state to keep tabs on citizens, but equally useful for citizens to keep tabs on each other. Social media recruits users to act as agents of spontaneous social and ideological enforcement, in a system of distributed surveillance just as impactful in its own way as the state-run version.
The horizontal policing criticized in the Harper’s letter spreads far beyond the elite cultural milieu that concerns its authors. Certain more obscure “cancellation” incidents that have received coverage lately show that average civilians with no public profile are meeting the same retaliatory fate as prominent media figures. A Mexican American employee of the San Diego Gas & Electric company got into a minor traffic dispute and was subsequently accused on Twitter of making a white supremacist hand gesture. The person who recorded the gesture later acknowledged that they may have misinterpreted the gesture but nevertheless, the worker’s employer fired him under pressure from outraged enforcers on social media. A small business owner in Minneapolis was evicted and lost most of his clients when it was revealed his teenage daughter had made offensive posts on Twitter and Instagram. Both the expanded expressive freedom and the “peer-to-peer” punitive functions of social media are on full display in these cases.
These incidents and the higher-profile ones referenced in the Harper’s letter follow a pattern so predictable it’s as if it has become algorithmic, and in a sense, it has. Many have discussed this sort of online crowd behavior using the archaic language of “witch hunts,” “inquisitions,” and so on. Such terms may seem hyperbolic but they aptly intuit that online conflict tends toward mob attacks culminating in a sacrificial climax. Social media users of all sorts, in many situations, gravitate toward group aggression against individuals. What predisposes the digitally mediated social sphere to these standardized dynamics?
In our polarized environment, analysis of this question too often focuses on the political culture of a particular group. Thus, on the center and right, criticisms of these dynamics fixate on “cancel culture”—a term not used but implied by the Harper’s letter authors. Many to their left downplay the problem of cancel culture by either denying or defending the practice—or, as is often the case, both denying and defending it depending on the context—while expressing concern about the similar phenomenon of online harassment. Although the content and severity of the attacks vary, what’s at stake is the same basic mechanism: individuals’ or small groups’ exploitation of the herd behavior of a larger mass of users to shame and castigate individuals for some real or imagined offense.
Critics of these tendencies, regardless of political orientation, often see them as separate from the functioning of the networks on which they unfold. Hence, critics of cancel culture try to ascribe it to so-called Social Justice Warrior (SJW) psychology. When examining incidents of online harassment, those on the left likewise attribute them to the intolerance and bias of their ideological enemies, and argue that if the platforms enforced anti-harassment policies, such behavior could be prevented.
But these responses underestimate the extent to which “harassment,” “cancellation,” or however you describe the tendency to herd users into attacks on individuals, is a feature of the platforms, not a bug. They result from the way social media incentivizes behavior with quantitative metrics. A few critics of the letter indirectly acknowledged this reality when, after rejecting the signatories’ claims about the oppressive impact of social media, they reported being harassed on social media for doing so. Harassment of this sort can be the cost of participation, not necessarily a penalty for any specific offense.
Any social media post, however trivial, attempts to win “points” in the form of likes, favorites, retweets, new followers, and so on. The reward for “winning” in this way is a minor but profound sensation of exultant victory, a physical state often associated with activity in the brain’s dopamine receptors. The egalitarianism of the network is a part of what draws us in: Any user can theoretically rack up points and victories in this manner, regardless of their status “in real life.” But this nominal egalitarianism is in tension with the platforms’ de facto inequality. The average follower count on Twitter is 707; the average count on Instagram is 150. The elite tier of accounts have millions of followers. The platforms thus hold out the promise of infinite free dopamine stimulation by showing us users with more followers who seem to be obtaining that promised dividend. But that level of success remains elusive for the majority of users.
No wonder, then, that the dominant online mood is one of resentment that spills over into bitter conflict. The game is signaling to most of us that we are losing. Crucially, rewards accrue to a profile that functions as a digital representation of an individual. The status-jockeying therefore does not reward discrete skills or activities so much as it seems to validate (or invalidate) a person.
This constant thwarting of desire does not make most users give up in despair. Instead, it makes the rewards seem more desirable. Since the points that accrue to our posts have no intrinsic value beyond whatever fleeting rush they provide, our desire for them is mimetic in the sense delineated by the social theorist René Girard. We want likes, retweets, and followers because the platform primes us to perceive other people getting them as proof of their desirability. The fantasy of reaching the fulfillment we project onto high follower count users holds out a hope of future success, but infuses the experience with envy and frustration.
The more of us that are transfixed by spectacles of victimization, the greater the revenue the platform brings in. Like a bloodthirsty god, the platform business feeds off of sacrifice.
But there is also another path to the satisfaction of the desires the platforms instill in us that has a lower bar to entry. This is joining one of the regular waves of enthusiasm around hashtags and other indicators of shared purpose. The “black square” phenomenon on Instagram offered a visualization of the melting of identity into a collective that occurs at these moments. When we shift in this way from seeking rewards via individualized competition to seeking them via fusion with a crowd, we flip what psychologist Jonathan Haidt calls “the hive switch.” This switch can fleetingly relegate the envy and conflict of competition between individuals to the background.
What prompts these moments of collective effusion is often outrage. This may mean taking a stand against injustice, as with the black squares. But it can also involve taking a stand in favor of violence, or at least verbal violence, against a target perceived as deserving it. It’s plausible to speculate that the evolutionary origins of the human “hive switch” entailed the activation of a group’s self-defense against predators or enemies. It would make sense, then, that in-group cohesion we witness today often takes the form of hostility directed at those outside of the group. However, anyone who observes online pile-ons will notice that the victim is often not someone from a totally different social and political world but is rather, far more often, an in-group member who has transgressed a group norm.
There is a way of thinking about the “hive switch” that helps account for this pattern. René Girard argues that groups riven with competition and rivalry turn to scapegoating as an ad hoc form of pacification. The steady grind of envy, resentment, and reciprocal hostility that erupts between individuals and groups can be overcome if they redirect their aggression to a “surrogate victim.” When the divided community reunites in unanimous acts of violence against this victim a catharsis can be achieved. Mimetic impulses, which caused divisions between individuals to proliferate, are realigned as internal antagonisms dissolve into common hostility against the scapegoat.
This victim will ideally be in some sense part of the group, because the efficacy of scapegoating requires conflictual emotions within the group to be temporarily discharged from it through the victim’s expulsion. But the latter should also be marked to some extent as an outsider, since this is what allows their repudiation to be accepted within the group as just and legitimate. Hence, the ideal victim is an insider who has violated one of the group’s taboos or shibboleths.
Traditional societies, Girard argues, formalized and controlled this mechanism in the form of sacrifice. These cultures used ritualized violence to maintain social unity. He attributes the decline of sacrificial cults in the modern world to various factors, including the rise of the modern judicial system, which attempts to resolve intragroup conflicts through neutral mediation mechanisms. But his account suggests that a social space like the internet, which lacks both the checks and balances of the justice system and the ritualized mechanisms of traditional religious societies, will witness both multiplying reciprocal hostility and temporary pacifications through the spontaneous scapegoat mechanism. In other words, a constant cycle of outrage mobs, pile-ons, and cancellation, which, after a brief respite, provide the fuel for the next outrage—precisely what has occurred.
It’s not simply that digital platforms have removed the brakes on these dynamics. Their gamified point systems encourage them to unfold by attenuating one of the main natural restraints against mob aggression. This is what Girard calls the “first stone” problem. Because people are mimetic, meaning they imitate both the behaviors and desires they observe in others, they can be prompted most easily to perform an action by others doing so. As a result, Girard says, “the first stone is the most difficult to throw ... Because it is the only one without a model.” Contagious violence can be forestalled if the crowd lacks a model to follow.
“Threshold models” developed by sociologists complement Girard’s “first stone” theory. For instance, Mark Granovetter concludes that for a riot to occur, there must be one or more individuals in a crowd with low “thresholds” for engaging in risky or antisocial behavior. Once they start the process, individuals with higher thresholds, who are less predisposed to such acts, will be more likely to join in. The mimetic chain goes from the lowest-threshold actors, who are least in need of a model, to the highest-threshold ones who will only join the riot once many others are already participating. Just a few individuals with low thresholds suffice.
Social media lowers the average user’s threshold for throwing the “first stone” by attaching rewards to this act when it is performed successfully. Compare the situation of physical “first-stone throwers” to those online. In historical cases, the person to incite the violence of a group can escape once the riot gathers steam, vanishing into the crowd and gaining nothing beyond whatever frisson the act itself generates. In the attention economy, in contrast, the equivalent act can reap the dividends of viral fame. Hence, digital platforms’ point systems counteract the social values that might keep the threshold for casting the first stone high and limit the outbreak of riots.
Players of the online status game have good reason to throw the first stone. Since users easily come together around shared objects of moral indignation, a negative post about a person who can serve as some group’s scapegoat can be a predictable way to reap a good harvest of likes and followers. It’s unsurprising, then, that some users are trying to make a name for themselves on the basis of first-stone throwing. The career of Karlos Dillard, who is now infamous for concocting dubious viral racism videos to grow his brand, exemplifies this tendency.
If first-stone throwers are a necessary part of any public shaming incident process, so are the apologists who crop up during and after the event. Those who defend the behavior of their own group in such incidents frequently make two apparently contradictory points: First, they insist that the person attacked was guilty of a serious infraction. Second, they claim that the consequences suffered were not all that severe anyway, since people being mean to you on the internet isn’t the end of the world, after all. In other words, they assert the absolute moral seriousness of the enterprise, but also suggest that on some level, it is all just a game.
In a way, though, both propositions are true. Since the moral passions of a significant portion of the population are now almost completely oriented by social media platforms, what occurs there is as serious a moral matter as there is. But it is also a game in the simple sense that it is designed as one: It uses a point system to incentivize and disincentivize behaviors.
In the sacrificial rituals of ancient societies, the priestly class engineered moments of what sociologist Émile Durkheim called “collective effervescence” as a mode of social control. The priests harnessed the violent impulses that can erode social cohesion when left unchecked, and channeled them into shared experiences that unified the community. Today, the equivalent of these priests are the engineers who lure us into virtual Skinner boxes and use trivial rewards to induce us to mimic brutal ancient rites.
Like their archaic counterparts, these modern priests grasp that “collective effervescence” is a powerful force that needs to be channeled and circumscribed. But the essential goal of today’s priests is not control, although that is an effect. Nor is it social cohesion and the greater good of the community. It is profit. The gamified mechanisms that precipitate us toward indignation against enemies also drive our continued use of the platforms. The more of us that are transfixed by spectacles of victimization, the greater the revenue the platform brings in. Like a bloodthirsty god, the platform business feeds off of sacrifice.
What’s become apparent lately is that the offline world is increasingly subordinate to the ravenous appetites of this god. The recent wave of cancellations of figures both prominent and obscure reveals the degree to which the online logic of gamified sacrifice has taken hold of institutions, and not just cultural ones. The latter may be a leading indicator because their mostly deskbound denizens are more glued to screens than others, but the automated logic of sacrificial resolution is determining outcomes in spaces far less integrated with platforms. Supporters and critics of “cancel culture” alike would agree that journalism’s Twitter-facing culture was a key factor in the recent ouster of several newspaper editors, most prominently The New York Times op-ed page’s James Bennet. But the same mechanisms that empower media workers to take down their bosses also enable anonymous civilians to enact mob vengeance against other anonymous civilians, as the San Diego Gas & Electric case indicates.
The Harper’s letter signatories fear that horizontal policing will dampen debate, and that the chilling effects of the punitive digital environment will induce intellectual uniformity. But for the platform god, this would be an undesirable outcome. If the fear of mob vigilantism made everyone conform to the standards of their group, there would be no more sacrificial immolations to keep people captivated by their screens. The god needs such spectacles to live.
Our rolling digital auto-da-fé will never fully stamp out heresy. Rather than producing uniformity, it will perpetuate the symbiosis between deviance and conformity. Users more disposed to standing apart from the crowd will continue to violate taboos and become targets of mobs, yet will also sometimes be the first to call out apostates and whip up the mob’s hostility. (Given that both acts imply a disposition toward high-risk but potentially high-reward behavior, it’s unsurprising that first-stone throwers may also be more likely to end up on the receiving end of a mob’s aggression.) The risk-taking minority will continue to play these roles in the hope that the attention economy will reward them for doing so. In the process, they will provide the hypermimetic majority with the periodic flare-ups it needs to reconsolidate in-group cohesion.
Stifled debate may present an immediate political loss, but the far greater risk is the subjugation of all realms of human life to the sacrificial profit-driven logic dictated by the digital platform god and its priests. Many of those who defend participating in these cruel rites seem to regard them as a vehicle of progress. But sacrifice is cyclical, never progressive. It permits a temporary substitutive crisis resolution, in which individual victims stand in symbolically for vast, multifaceted problems. The punishment and expulsion of the victim occasions a cathartic transference of responsibility and an ephemeral relief of pressure. But the platform god and its priests will keep stoking the resentments that precipitated us toward this harsh resolution. That’s the game they need us to play.
Geoff Shullenberger is a writer and academic. He blogs at outsidertheory.com.