Sergei Karpukhin / Reuters

Let’s get this out of the way first: The Internet is the real world.  

What you say online is still you saying something—even if you’re shielded by an anonymous account; even if you’re saying it just to be provocative, or performative, or God only knows why else. You on the web is still you, just like you on the telephone is you. Technology doesn’t magically make a person’s behavior inauthentic, or pretend, or inconsequential.

In an essay last week, the writer Stephen Marche set out to explore a Reddit-hosted community that has a reputation for being one of the most misogynistic swamps on the Internet. Marche’s puzzling conclusion was that the participants in this group are pathetic and afraid, their fear fueled in part by a desire for cultural clarity, not by mere hatred of women. He dismissed them, implying they weren’t threatening in any substantial way, and, in the end, suggested they read more classic literature. All this, it seemed, stemmed from Marche’s central (and misguided) question: “Are we our real selves on the internet, or are we not?”

The answer, of course, is that we are our real selves online as much as we are our real selves anywhere else. The Internet is the real world! This stuff should be easy. But it gets harder from here.

* * *

Harassment has been a serious problem online since the dawn of the web. In 1984, scientists puzzled over the “surprising prevalence of rudeness, profanity, exultation, and other emotional outbursts,” that seemed to characterize computer-based communications, The New York Times reported that year.

Today, the majority of Internet users have witnessed name-calling and attempts at humiliating someone online, according to a 2014 Pew survey—and 40 percent of those surveyed said they’d experienced such treatment (or more severe forms of harassment) themselves. Among those who have been harassed, many episodes went beyond name calling to include physical threats, stalking, sexual harassment, or sustained attacks over time. Men (44 percent) were more likely than women (37 percent) to experience online harassment of any kind, but much of the worst harassment is disproportionately targeted at women—and young women, in particular.

And yet, despite close attention from scholars, and a seemingly endless stream of terrifying anecdotes, it has remained difficult to quantify or otherwise analyze harassment on any given platform. Which means it’s even harder to gauge whether things are getting better or worse.

“At the moment, it’s impossible to know if or when any online platform has actually improved in terms of the harassment people receive or their response to harassment,” said Nathan Matias, a Ph.D. candidate at MIT who studies online harassment. “The Guardian is literally the first platform in history, as far as I can tell—and I’ve looked very, very carefully—to ever do that.”

Matias is referring to a major analysis of The Guardian’s comment section, published by the news organization last week—notable because of its findings, but also because of how rare such an examination is. It found that eight of The Guardian’s 10 regular writers who got the most abuse from commenters online were women (four white and four non-white) and two were black men. (The 10 regular writers who got the least abuse were all men.)

The Guardian also found that, over a five year period, articles written by women consistently elicited more abusive responses than articles written by men. That was the case across almost all sections of the website—though women received particularly egregious treatment compared with their male peers in sections that were otherwise dominated by men (like sports and technology). Certain topics prompted more abusive responses, too. “Conversations about crosswords, cricket, horse racing and jazz were respectful,” The Guardian wrote. “Articles about feminism attracted very high levels of blocked comments. And so did rape.” The message to women is clear: You’re permitted to speak publicly about word puzzles and saxophone solos, but examine gender politics at your own risk.

This sort of vitriol isn’t just aimed at journalists. Many women report experiencing abuse on dating sites, too. “Men also sometimes call me a whore if I don’t write back,” a friend of mine, Frannie Steinlage, wrote in a recent essay. “They abruptly sexually proposition me. They tell me they’d rape me. They send me unsolicited photographs of their penis.”

Her experience is typical, as is demonstrated across many horrifying websites where screenshots of such harassment are swapped. One women recently turned her collection of unsolicited penis-photos into an art exhibit. “It’s not about sex. It’s about power,” Whitney Bell told Vice. “It’s like screaming at a woman from a car. You’re just doing this because you can, and because the world has taught you that that’s OK.”

Tales of online harassment are as predictable as they are frightening—it happens on Tinder, on Twitter, on Facebook, on Tumblr, and all across the web. “None of those companies release data overall, or publish the results of any of their efforts to address problems like harassment,” Matias said. “And that’s a real problem.”

Though harassment affects a huge number of people who use the web, journalists are presumably uniquely positioned to do something about it. If the people who have prominent platforms—heavily-trafficked websites or lots of followers on social media—fight back publicly against harassers, it could help demonstrate that such attacks are unacceptable. Or it could encourage more bad behavior, incentivized by the possibility of a high-profile reply. It’s that age-old question: Should we feed the trolls, or does that just embolden them?

“The problem is that the word ‘troll’—we’ve burdened it with so many different meanings, that anything I tell you will be unhelpful and potentially dangerous for the person whose situation completely mismatches the advice I give,” Matias told me. “My top piece of advice is find a friend. I think it’s always helpful to have someone you trust to help you make sense of—and decide how to respond to—a frustrating or potentially dangerous situation. Find someone to help you.”

People seem to do that already. Last year, Twitter gave Matias—along with several colleagues working with the group Women, Action, and the Media—access to data on how Twitter harassment is reported. “In our research on Twitter harassment, we found that most of the people who submitted reports … were friends and other bystanders” who had witnessed the abuse, but not the people who experienced it firsthand. This behavior seems to reflect common sense: Ignore the troll or bully, and enlist someone you trust who can help you decide what to do. I see this firsthand all the time among journalists—both men and women—who face gross treatment online. Much of the time, ignoring is the obvious way to go.

Sometimes, though, people decide to respond.

It’s a dicey move, but talking back to a troll can yield positive results. In 2015, in a story for This American Life, the writer Lindy West told the moving story of how she confronted a man who had cruelly harassed her, assuming the identity of her deceased father on Twitter and using that account to send her insulting messages. First, she decided to write about the harassment in an article for the website Jezebel, which prompted the man to email her to apologize, which prompted West to ask him if he’d talk to her about why he’d harassed her, which led to one of the most remarkable pieces of radio I’ve ever heard. (Please listen for yourself. It’s really that good.)

West, along with several other high-profile women, has long been a target for harassment online. She is, as Cosmopolitan put it last year, a Famous Internet Feminist, a designation that pretty much guarantees an army of trolls to follow. The blogger and video-game designer Brianna Wu occupies a similar cultural position. As an outspoken critic of sexism in the gaming world, Wu has received dozens of death threats and a barrage of harassment online. “The organized effort to harass people has never been more serious,” Wu told me. “The misogynist effort to harass, bully, and threaten us—it’s a very dangerous time to be a woman online.”

And yet there’s very little recourse for those facing harassment. “When I report a serious death threat to the police, this is what happens,” Wu said. “Invariably, a local cop comes to my house and instructs me to stay off social media. I cannot have a career without that online presence. One of the most insidious forms of sexism in 2016 is that my friends who have children stay off social media because they’re terrified of being targeted.”

“There is a mental overhead cost to this,” Wu added. “It’s an overhead that men in our fields don’t have. It’s the same thing with being held to quadruple standards on how we dress, it’s the same with not being taken as seriously as men, it’s the same with women being given the burden of childcare, and it’s frankly exhausting.”

Or, as the writer Jessica Valenti recently put it, “It’s a workplace harassment issue that doesn’t stop at the workplace.” Valenti writes frequently about feminism, too, and is The Guardian’s most targeted writer.

Which brings us to the cost—in time, in energy, potentially in personal safety—of responding to a troll at all. Or even just trying to elevate a level of discourse to basic civility, captial-T trolls notwithstanding. Even if you can make someone understand why what they’re saying is harmful or inappropriate, even if they apologize in dramatic fashion, why should the victim of harassment or condescension take on the burden of explaining systematic cultural and social forces that are well documented in decades of academic literature?

Often, responding just isn’t worth the energy. When colleagues come to me lamenting the belittling or rude emails, tweets, and comments they receive from strangers online, I like to recommend a website I have bookmarked for such occasions. (It’s a looping gif of Bender, from Futurama, and it’s perfect.) If it’s not a serious threat, often it’s best to ignore the person and move on. Which brings us back to the question of what people mean when they talk about “trolls” in the first place.

I’m used to having people call me stupid, and telling me that my reporting is a useless waste of time. These kinds of comments aren’t threatening, but they are discouraging. Seeking out criticism from those who disagree with me is one of the reasons I bother with comment sections at all. I want to engage with people who have ideas that are different than mine, perspectives that might improve my thinking; these sorts of exchanges are one of the great privileges of being a journalist. But often the criticism I encounter isn’t of my work, but rather of my gender, or of entire races and socioeconomic classes of people I write about.

(These kinds of comments prompt some kindness, at least: “I happened to take the time to read the comments, and I was horrified by the reactions," one reader recently emailed to say. “I read the comments and felt an overwhelming urge to write something encouraging to you,” another wrote. “Everyone was being so mean!”)

Sometimes, when a stranger tweets something sexist at me, I write a response and take a screenshot of it instead of hitting send. All the catharsis, none of the unnecessary engagement. Other times, I like to write back to comments that strike me as mean spirited—sometimes to see if a path toward civility is even possible.

A few months ago, for example, I wrote an article about my attempts to send a telegram in an age when telegrams are basically obsolete, and someone I don’t know tweeted a response to me (and my editor), indicating he didn’t like the story: “in this article a millennial shows off her gullibility/naivety as a consumer and regurgitates Wikipedia.” Now, as far as online vitriol goes, this is pretty innocuous stuff, but the accusation about Wikipedia irked me.

So I responded: “yikes!”

Here’s a screenshot of the exchange that followed:

A few minutes later, an email arrived in my inbox. The subject line: “all right, I was kind of a dick.” The message: “All right, I was kind of a dick and as penance I will discuss things with you if you want.”

So I wrote back and asked him what I wanted to know: “Did you think I would actually read your tweet?” In other words, was he trying to get me to respond—or was he venting into the Internet ether, the way you might shout at a football game?

“I was honestly really shocked when you replied, which is kind of odd now that I think about it, because that's the opposite of how I normally think about interactions online,” he wrote. “I generally am the person who reaches out to people all the time and gets responses.” This one time, he went on, “I had a question for Elon Musk, so I asked him, and he wrote back in minutes.”

“So I don't know why it didn't occur to me that of course you were going to see that and maybe respond.”

What is it, other than $14 billion, that made me so different in his eyes than Musk? I asked him: Was it my age? “Maybe? I definitely seem to find that people under 25 or so and over 45 or so seem to be a lot less critical and wary of their information sources.” (I’m 33.) Was it my gender? He didn’t respond to that one. (It wasn’t just my tweet that prompted him to talk to me about why he’d sent the initial message; he was called out by a friend who had seen the tweet. “[I was] also really embarrassed when a few seconds after my tweet he called me out on it in our irc channel.”)

Finally, I asked him if I should have even responded to him at all. His responses were fascinating to me, but was the exchange worthwhile for him?

“Absolutely,” he wrote. “I'm pretty embarrassed by how I acted and being called out on it was extremely helpful. I definitely need to more often step back and think a bit more before. I usually do; sometimes I fail.”

The exchange felt to me like a tiny victory for civility in the Internet age, but it probably doesn’t make sense to try to change the world one commenter at a time. Not only is that approach mentally exhausting, but it’s potentially dangerous. Before I even responded to the man who had insulted me, I scrolled through his tweets to be sure he didn’t seem unhinged. I chatted with coworkers who had seen his tweet. I proceeded with caution. And besides all that, the kinds of tweets and comments that can be considered harassment are in an entirely different, much scarier category than what I faced.

“When we think about problems of harassment or conflict, the overwhelming way we think is to consider ways to deal with that specific conflict, that specific troll and that specific person—and that often leads us to think about reactive responses, things that we can do after we spot something happening,” Matias, the MIT scholar told me. “But if we recognize that these problems are at least sometimes rooted in deeper social and cultural issues, there’s a huge scope for working on these problems at the deeper roots.“

“It may actually be,” Matias said, “that deeper social change that could have the larger impact.”

On one hand, this is an overwhelmingly disheartening conclusion—the idea that you can’t fix trolls until you fix all of the prejudice. Matias’s point, though, is that social change on a massive scale is possible. We’ve seen it over and over again. Addressing the problem of harassment online means, for example, thinking about how the design of various platforms might encourage such attacks, interrogating the usefulness of reporting and blocking features, and building communities where people can find and offer support.

Reacting to individual trolls, cretins, and bullies may well be an essential part of this larger process, but fighting harassment  in a lasting way will take a much larger and more coordinated effort—one that’s rooted in a shared belief of what the web should be like, which is a reflection of what the world should be like, too.  

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.