Feb 3, 2022 9:00 AM

It's Not Your Fault You're a Jerk on Twitter

The design of the internet lets you harass and harm people without ever once interacting with them directly. Even if you're trying to defend them.

Photo-Illustration: Sam Whitney; Getty Images

I’m used to comics breaking my heart; I’m a Wicked and Divine fan, after all. But I felt the particularly painful crack of an old fissure reopening as I read cartoonist Mallorie Udischas-Trojan’s “Piled On” at The Nib. It describes the pain of being harassed over a jokey comic she wrote in which a character shoplifts supplies from an art store. The 10th panel, depicting her own heartbreak at friends and allies joining the dogpile—“I … I really thought we were rooting for each other”—just about broke me. I’ve been there, and I have many friends and colleagues who’ve experienced their own versions of this sort of lateral abuse.

Many seem to imagine that a greater focus on personal responsibility can detoxify the internet; perhaps Udischas-Trojan’s progressive harassers, her erstwhile friends, were just bad people all along, surely? I myself promoted the idea of digital ethical education as part of the solution to problems like these. But the tragic reality is that our good intentions—in whichever direction they may run—are so easily twisted online, corrupted from their purpose due to forces beyond our control.

The internet is a distributed network where you can harass and harm people without ever once interacting with them directly. Your intent is irrelevant. Even, crucially, if it’s an attempt to defend the person being harassed.

In the end, online harassment campaigns twist questions of personal responsibility and virtue into strange shapes, often rendering them immaterial. A campaign needs individuals to participate, but, as with Agatha Christie’s murderers on the Orient Express, no one knows who in the crowd strikes the fatal blow.

A harassment campaign, as described in my research, is marked by three qualities that social media is designed to cultivate almost automatically: crowdsourcing, organization, and longevity. The campaign against Udischas-Trojan was organized and marshaled by certain individuals, but the platform did most of the work, rapidly crowdsourcing like-minded people to like, retweet, and contribute their two cents to the drama. And, of course, it lasted. It went on for days and weeks at a time.

What sustains it is a structure that can best be visualized as an inverted pyramid, bearing down on an often helpless individual target, in descending degrees of severity. What I term first and second order harassment is the abuse you’re most likely familiar with, from the violence of swatting someone to the casual cruelty of abusing someone through a tweet or email or TikTok directed at them.

But it’s the third order that gets at what the rest of this essay is about. This third order is not hacking the target or engaging with them to spew abuse—rather, it describes the simple act of commenting on the situation. It is the enabling, apologism, and justificatory discourse about the target that ensures most people participating feel as if they’re doing the right thing and makes more overt and intense forms of harassment possible.

Crucially, third-order behavior exceeds private intentions. Most people in this tier don’t intend to do harm. From high-minded pseudo-philosophizing to a mocking TikTok to ironizing on Twitter, any discourse about the target keeps them aforemost the hivemind. If there’s no such thing as bad publicity, there’s no such thing as good social media discourse when you become a “main character.”

Subtweets can also constitute third-order harassment, specifically by providing a micro-forum for reinforcing a perception of a person as an acceptable target; the point of subtweeting is, specifically, to avoid commenting on the target directly. It can even be perceived as a kindness. It seems a painless way to comment on a Main Character without contributing to a flood in their mentions, and without the risk of incurring backlash from their defenders, while enabling one to just vent their spleen harmlessly into the ether. But it still often provides moral reinforcement for the more devoted, less scrupulous attackers.

In Udischas-Trojan’s case, the moral justifications for attacking her—people who argued that her comic was a “bad take” or that it painted artists in a bad light or that shoplifting “actually hurts workers”—weren’t directed at her; they just served to justify further discourse and rationalize the furor over a completely innocent comic. The target rarely sees third-order discourse, but they’ll certainly feel its effects.

“Where there’s smoke, there’s fire,” we often say, to justify suspicion and a presumption of guilt. Third-order harassment is a smoke machine.

Indeed, this is how the sorry saga of “I Sexually Identify as an Attack Helicopter” unfolded. The sci-fi short story, written by a transgender woman named Isabel Fall as a tongue-in-cheek exploration of gender based on a concept usually used to slur us, was spun up in Twitter’s violent accelerator until mild critiques and rebukes turned into existential campaigns.

Fall was harassed out of spec-fic writing, social media, and perhaps even her very identity by a small but motivated number of angry left-of-center social media users who had convinced themselves she was secretly a Nazi or other right-wing malefactor who conspired to plant a blatantly anti-trans story on Clarkesworld, the online magazine where the story was published. In the augmented reality game of social media, investigation is a common enough pastime—consider QAnon’s fervor for interpreting “breadcrumbs” and the like—and Fall’s bio on Clarkesworld listing her birth year as 1988 raised eyebrows as a possible Nazi signal (88 corresponding to HH or Heil Hitler).

But it turned out that the story, which was not in any meaningful sense transphobic, was just a story. Fall withdrew it hastily and disappeared; she would suffer a psychological breakdown and teeter on the edge of suicide, forced to check herself into a mental hospital.

Outraged at how Fall was treated, other Twitter users sought a scapegoat to call out and punish, finding one in Neon Yang, a different trans spec-fic writer who was supposedly a ringleader against Fall at the height of the crisis. Except they weren’t. They were singled out because they were one of the few to apologize for what was, in the end, a small role—and their apology, which seemed to excuse some of the worst attacks against Fall by others, rankled some of her defenders. Sustained attacks against Yang continued for weeks on end.

In both cases, third-order harassment—discourse about the targets—kept it going. One prominent author tweeted at the time that she was glad the story was taken down by Fall, because "not all art is good art. Sometimes art causes harm.” In response to critics, she even suggested that the story might aggravate trans people’s PTSD, claiming that she’d heard this from trans folks themselves (the author herself is not trans). Such claims lay a moral foundation for more overtly abusive behavior by others actually willing to engage more forcefully. And, ironically, the author freely admitted she never even read the story. What she had read was other discourse about the story, demonstrating how propulsive this sort of thing can be.

But later, this same author, along with Yang, was subject to a great deal of abuse by people who blamed them disproportionately for what had happened. When an article by Emily VanDerWerff about the saga came out on Vox, the author wrote an apology that seemed to soft-pedal some of her earlier remarks. “She is only doing this to try and save face. She doesn't actually care,” according to one Reddit comment. From the same thread, “That's honestly why I'm going from ‘She's low on the list of books to read’ to ‘Nope.’ Same goes for Yang, who I've just learned about...” Emphasis mine.

This combustion is fueled by the conversion of a complex individual into a concept that represents a suite of other social problems. The rage at the helicopter story wasn’t about the story or Fall as a person, but was driven largely by progressive, cisgender allies of trans people attacking something they perceived as transphobic. The attacks on Yang were really about the deep-seated anxiety among trans women that our notional allies, even among other trans people, are secretly waiting to cut us down at the first opportunity. The attacks directed at an individual are a metacommunicative shorthand—“I hate Neon Yang” isn’t about Yang, it’s about a suite of ideas that they discursively represent; you can’t @ an idea on Twitter, only a person.

This is why even the numerous attempts at “constructive” callouts or criticism in the helicopter story saga, directed at both the original story and Neon Yang in later months, merely added to the pain and fury. The sheer weight and volume of so many people bearing down on an individual all at once becomes powerfully destructive, even if many of those people are being “nice.”

How did this come to be? The answer is twofold: design and dissociation.

Road design in countries like the Netherlands promotes what is known as “traffic calming,” reducing pedestrian deaths and car accidents; by contrast, road design in North America promotes high-speed driving, passively nudging drivers to step on the gas, giving them less time to stop, even in crowded areas. Understood this way, you can get away from solely individualist narratives about accidents—about bad drivers or “pedestrians who weren’t looking”—and focus on how design encourages broad outcomes not attributable to any one actor.

Similarly, social media is designed in a way that agitates, rather than calms, its traffic. It leans into, rather than curbs, the augmented reality aspects that arise from computer use—tricking you into believing you’re somewhere other than reality.

You see, nearly all internet use is fundamentally dissociative, subtly divorcing us from the consequences of our words and deeds—what psychologist John Suler dubbed the “dissociative imagination.” In my own research, I came to this conclusion from the other way around, arguing that in online gaming spaces, the magic circle conceit of video gaming enabled people to extend the game’s unreality to their own words and actions. But I eventually realized that it wasn’t just games that had this effect. It was the entire online space, disinhibiting and ludic all at once.

There is a seductive quality to posting into the void, a Möbius strip sense that you’re the voyeur who no one can see, and the exhibitionist who everyone must see.

If it is so easy for good intentions to be perverted by the platform, then perhaps the fault lies in the binary stars rather than ourselves. Like most structural problems, from the ongoing pandemic to climate change to the rampant inequalities that worsen the devastation of both, we cannot delude ourselves into believing that additive individual virtue will be enough to overcome the problem—especially when people commenting on every side feel virtuous.

That doesn’t mean that you have no responsibility when using social media, of course. In both the helicopter story and shoplifting comic sagas, for instance, many progressives who went on the attack were picking up on rumors started by right-wing harassers spreading deliberate disinformation. Abandoning the “where there’s smoke there’s fire” mentality we all have about internet callouts can only serve us well.

But sometimes there’s simply nothing to say that won’t cause trouble, which is one of many reasons why the periodic appearance of think pieces about Twitter discourse masquerading as dispatches from the front line of “The Culture War'' are risible. The idea that there is some cabal of lefty Twitter power users convening at a summit to set the right level of ironic vitriol for their tweets, as if setting oil production quotas at OPEC, has always been absurd.

The problem isn’t that Twitter’s users are toxic, it’s that the platform makes toxicity the path of least resistance, and turns even intentionally positive commentary into just more fuel for the fire.

This is what the road design of social media has wrought; like a wide, straight road encouraging dangerous speeding, the limitless void of social media encourages pained shouting, and its financially backed incentive structures make an addicting game out of the whole affair.

Solutions are few and far between. One can avoid participating in the discourse of the day, but that is tantamount to self-censorship, and it deprives one of the social networks these platforms readily provide. For some people, the socializing they can do on a platform like Twitter is a lifeline, after all. What about at the structural level? Involving the police will merely redouble other structural oppressions; paeans to “inclusiveness” will come to nothing; and technical solutions merely tinker around the edges or dull our ability to hear the crowd, rather than calm them.

If we had built the internet around the idea of our embodied humanity, rather than the naive idea that we should be liberated from it, dissociation might not have come so easily.

No, it’s not your fault you get twisted to nastiness on social media. But if that corruption is inexorable, it has uncomfortable implications. Not the least of which is that these platforms, supposedly all about you and allowing you to broadcast your truest self to the world, don’t give a damn about your good intentions.