How can journalists stop providing oxygen for trolls and extremists?

Are journalists partly to blame for the rise of the alt-right and the outcome of the 2016 election? A report from the research institute Data & Society looks at the ways in which journalists help to popularize extremist views, in some cases accidentally. The paper—written by Whitney Phillips, an assistant professor of communications at Syracuse University, and entitled The Oxygen of Amplification—argues that alt-right and other groups were aided and abetted by the media, which helped expose their ideas to new audiences. Says Phillips:

The takeaway for establishment journalists is stark, and starkly distressing: just by showing up for work and doing their jobs as assigned, journalists covering the far-right fringe – which subsumed everything from professional conspiracy theorists to proTrump social media shit-posters to actual Nazis – played directly into these groups’ public relations interests. In the process, this coverage added not just oxygen, but rocket fuel to an already-smoldering fire.

Among the 50 interviews Phillips did with journalists were many who agreed their work covering the alt-right and other extremist groups provided them with more publicity, and in some sense may have helped create the movement itself. “Without journalists reporting on them, there’s no way they would have gotten the attention they did,” said HuffPost reporter Ashley Feinberg. “At this point we have built the world they told us existed.”

The mechanisms by which this happened are complex, as Phillips describes at some length in the report, and many of them are not easy to change because they are built into the very structure of journalism itself. But the author also suggests ways of mitigating the damage—steps that journalists can take to ensure that the coverage they are providing of such groups is not only justified, but reduces the amplification problem.

So what can journalists or media organizations do to avoid this kind of problem? One thing they can do, Phillips suggests, is to not send reporters or writers who are unfamiliar with online behavior like trolling to write about those kinds of individuals, since it makes them unprepared for many of the tactics and strategies online natives are used to. But there are other tips and tests that can be used as well, she says, including:

  • Has the story in question expanded beyond the interests of a specific online community or subset of a community, to the point where it is being shared or discussed more widely? This is what Claire Wardle of First Draft News and others call “the tipping point.” If a story hasn’t reached this point, then writing about it will almost certainly push it past that and give it more oxygen and legitimacy.
  • Does the story have some larger positive social benefit, in the sense that it will add to an existing conversation about solutions to a problem, or open up a new conversation about an important topic? This is similar to the kind of test that journalists and media outlets use when reporting on other forms of social behavior such as suicide, violent crime etc.
  • Will reporting on the story produce some kind of harm to those involved, including embarrassment, re-traumatization or professional damage, or could the audience for the story use the information in it to cause harm, whether it’s attacking sources or imitating crimes? That last point is similar to the test some media outlets use when reporting on computer hacking and other similar incidents.

One thing that makes it difficult to see these kinds of tests as a solution to the problem is that they all involve subjective appraisals of the social landscape, and assumptions about potential outcomes, which means that they are open to debate. On top of that, as Phillips points out elsewhere in the report, some media organizations simply don’t care about the potential damage a story could cause, so long as writing and publishing it brings in revenue or attracts attention.

“Many reporters also acknowledged that some journalists are themselves manipulators, cynically repeating lines they know are false and misleading because it will get them clicks,” the author writes. Max Read, editor of New York magazine’s technology blog Select All, tells Phillips that “there are so-called journalists more than happy to embrace the fucked-upness and make a buck off it.” And that is a problem that won’t be solved until the entire revenue model for news is changed.

Whitney Phillips; incoming assistant professor of communications at Syracuse University and author of the 2016 book This Is Why We Can’t Have Nice Things: Mapping the Relationship between Online Trolling and Mainstream Culture PhD 2012, English with an emphasis on folklore, University of Oregon

In their analysis of mainstream media coverage and Twitter linking patterns during the 2016 US presidential election, Faris, Roberts, and Etling (et al.), in collaboration with the Media Cloud, Berkman Klein, and Center for Civic Media, conclude that far-right media, from small extremist blogs to larger outlets like Breitbart (a dragnet that certainly included its fair share of “trolls,” depending on how someone was using that term), did in fact set the mainstream agenda. But not without help. As influential as these far-right media may have been within a certain user base, they simply didn’t have enough clout to shift the national conversation themselves, and certainly didn’t have enough votes to win an election.

These media, instead, depended on the signal-boosting power provided by center-left establishment publications like The New York Times, The Washington Post, and CNN.com to ensure that their messages would spread to a national, or even global, audience.9 That’s how Pepe the Frog lept onto the public stage. That’s how Donald Trump Jr.’s Instagram post became a national news story, and ultimately, a talking point in two presidential candidates’ campaigns. That’s how many Americans first heard the term “alt-right.”

The takeaway for establishment journalists is stark, and starkly distressing: just by showing up for work and doing their jobs as assigned, journalists covering the far-right fringe – which subsumed everything from professional conspiracy theorists to proTrump social media shitposters to actual Nazis – played directly into these groups’ public relations interests. In the process, this coverage added not just oxygen, but rocket fuel to an already-smoldering fire.

“Without journalists reporting on them, there’s no way [far-right elements] would have gotten the attention they did,” Ashley Feinberg, now a senior reporter at HuffPost, told Phillips. “We’re setting the tone for them by covering them that way…at this point we have built the world they told us existed. We are the reason that these people are getting actual legitimate platforms now.” The Atlantic’s Emma Green, meanwhile, said wall-to-wall coverage of far-right elements “creates a cycle where the line between a constructed reality that is amplified by the mainstream media then flips into something that’s covered more, that people attend, that then has real life consequences . . . it becomes hard to delineate the boundaries between what’s constructed and what actually would have existed without that kind of media attention.”

Within these conversations, many reporters also acknowledged that some journalists are themselves manipulators, cynically repeating lines they know are false and misleading because it will get them clicks. Max Read, editor of New York Magazine’s technology blog Select All, stated, “There are so-called journalists more than happy to embrace the fucked-upness and make a buck off it.”

One Gawker reporter admitted feeling torn; she’s not sure what she could have done differently. And yet, she admitted, “Every once in a while I’ll look back and see something that I wrote a year and a half ago and the pit of my stomach falls, because either I was joking about these trolls, or making light of the fact, joking about Trump becoming president. It makes me physically sick to read [the articles] now.”

However critically it might have been framed, however necessary it may have been to expose, coverage of these extremists and manipulators gifted participants with a level of visibility and legitimacy that even they could scarcely believe, as nationalist and supremacist ideology metastasized from culturally peripheral to culturally principal in just a few short months.

The choices reporters and editors make about what to cover and how to cover it play a key part in regulating the amount of oxygen supplied to the falsehoods, antagonisms, and manipulations that threaten to overrun the contemporary media ecosystem—and, simultaneously, threaten to undermine democratic discourse more broadly. This context demands that journalists and the newsrooms that support them examine with greater scrutiny how these actors and movements endeavor to subvert journalism norms, practices, and objectives. More importantly, journalists, editors, and publishers must determine how the journalistic rule set must be strengthened and fortified against this newest form of journalistic manipulation—in some cases through the rigorous upholding of long-standing journalistic principles, and in others, by recognizing which practices and structural limitations make reporters particularly vulnerable to manipulation.

Analyzing the relationship between journalism and the amplification of harmful, polluted, or false information before, during, and after the election requires understanding the influence of earlier online subcultures on the journalists involved. In particular, the subculture that organized around 4chan during the previous decade had a direct impact on the ways many journalists conceptualized and reported on the emerging alt-right narrative. This impact hinged on the changing use of the term “troll” across that decade, as well as an underexamined division among journalists who were troll-trained (through previous exposure to and familiarity with the subculture) and those who were not troll-trained, or who simply rejected the category of trolling.

Despite the nebulousness of the “troll” framing, many within the news media, on social media, and even in some academic circles have credited some combination of “trolls” with shifting the norms of acceptable public discourse and contributing directly to Trump’s electoral victory in 2016. This narrative is problematic because it lumps too many individuals and actions into an imprecise category, in turn providing violent bigots, antagonists, and manipulators a built-in defense of plausible deniability, summarized by the justification “I was just trolling.”

The nonstop coverage devoted to alt-right antagonists operating under the banner of trolling illustrates the fundamental ambivalence of amplification. However critically it might have been framed, however necessary it may have been to expose, coverage of these extremists and manipulators gifted bad actors a level of visibility and legitimacy that even they could scarcely believe, as nationalist and supremacist ideology metastasized from culturally peripheral to culturally principal in just a few short months.

The Tyranny of Analytics: In the social media age, the measurability and commoditization of content, in the form of traffic, clicks, and likes, has tethered editorial strategy to analytics like never before. The emphasis on quantifiable metrics stacks the news cycle with stories most likely to generate the highest level of engagement possible, across as many platforms as possible. Things traveling too far, too fast, with too much emotional urgency, is exactly the point, but these are also the conditions that can create harm

According to respondents, two primary factors complicating the information imperative, particularly in digital environments, are the prevalence of “iterative reporting” and the frequent inclusion of false equivalencies in news reports, particularly in the US. Iterative reporting is the expectation that journalists should report on what other journalists are already covering. The inclusion of false equivalencies in news reports represents the journalistic norm of reporting on both sides of a story (described by several reporters as “both sides-ism”) on steroids, as positions that are false, manipulative, dehumanizing, and in many cases not worth reporting at all, are given an equal platform to positions that are factually true, relevant to the public interest, and unquestionably newsworthy.

One alt-right beat reporter for a national outlet, who rejected the “troll” frame for bad actors (I introduced this reporter in Part One) perfectly summarized the amplification tension when he noted that the institution of journalism is synonymous with amplification. “There’s no way around that,” he said. Nor is there any way around the fact that “there’s bad people in the world, and there are poisonous ideologies in the world, and at a certain point you have to realize that you’re promoting them to a . . . [long pause] not promoting them, but you’re getting those ideas out to a wider audience.” For him, the goal of getting those ideas out to a wider audience is targeted resistance; that people can’t push back against the monsters they don’t know are there. But in shining that spotlight, bigots’ messages spread even further, with the potential for further recruitment, further unpredictable engagement, and further radicalization. Both options are just as likely, and just as vexing, in every case.

Not every piece of information is worth reporting. In all cases, for all stories, journalists must assess what is newsworthy and what is not. To assess newsworthiness – a particularly important task when the story contains manipulative elements posted to social media – First Draft News’ Claire Wardle encourages reporters to ask whether or not the story has extended beyond the interests of the community being discussed. In the case of online memetic content, for example, the question would be whether a particular meme has been broadly shared by anyone outside the core group of participants.

This is the “tipping point” criterion (Moschella and Watts 2017): if the story hasn’t yet reached that point, all reporting will do is provide oxygen, increasing the likelihood that it will reach the tipping point. When presented with a story pitch that will take a small issue and make it much bigger through amplification, former senior editor at Refinery29 Laura Norkin asks herself, “If we didn’t cover this, and it didn’t get covered elsewhere, would it just go away?” If the answer is probably yes, and the coverage would have no social benefit otherwise, her policy is to pass on the story.

The question of “social benefit” is critical for April Glaser, technology writer at Slate. When weighing the question of newsworthiness, she considers whether the reporting will have a positive social benefit, if it will open up a new conversation, and/or if it will add weight and exemplars to an existing conversation. If the answer to these questions is yes, the story is likely worth reporting. But, Glaser also emphasizes that the quest for knowledge must be balanced with careful consideration of the harm – embarrassment, retraumatization, professional damage – that this knowledge might cause.

Another staff writer covering cybersecurity reiterates Glaser’s point, and adds a further wrinkle. The question isn’t just what harm could be caused by published information, he says. The question is also what harm could the audience cause by using that information, for example by finding and attacking someone quoted in the story, or replicating the crimes the story chronicles. Put another way, to assess newsworthiness, one must also assess what weapons the story would hand to its audiences.

Journalists must always assess the newsworthiness of information, especially when it concerns potential manipulations. There are three broad criteria to assess:

? Tipping Point – has the story extended beyond the interests of the community being discussed (Moschella and Watts 2017)? In the case of online memetic content, this question would direct reporters to consider whether a particular meme has been broadly shared by anyone outside the core group of participants.

? Social Benefit – will the story have a positive social benefit, open up a new conversation, or add weight or exemplars to an existing conversation?

? Potential Harms – will the story produce harm (embarrassment, retraumatization, professional damage), or could an audience use the story to cause harm (attacking sources, imitating crimes)?

Treat violent antagonisms as inherently contagious, akin to coverage of suicide,3 mass shootings,4 and terrorism,5 all of which are known to inspire and even provide behavioral blueprints for future copycat attacks. Similarly, wall-to-wall coverage of online harassment and manipulation incentivizes future attacks by signaling that such behaviors will result in the desired outcome – non-stop attention for the attackers.

To the extent possible, stories should specify the number of participants in a particular online attack/campaign, rather than using vague mass nouns (i.e., trolls did this, the alt-right did that). Important contextualizing information includes the apparent number of online participants (based on observational data), why the reporter believes this count is accurate, and any unknown variables that might impact the readers’ understanding of the story. When describing media manipulation campaigns of any kind, stories and their headlines should employ the most precise language possible.

Reporters and their editorial teams should exercise an abundance of caution when reprinting memetic images used during particular attacks, especially when the images are dehumanizing and bigoted. When sharing an image is deemed necessary, editorial teams, along with members of the communications team, should consider including captions from the story and/or other contextualizing information within the image itself so it can’t be hijacked and spread by manipulators as easily

Reporters and editors should be aware of how strategic many groups of white supremacists and nationalists are in their communications and messaging, which is geared toward maximizing recruitment. Similarly, reporters and editors should be aware that extremist groups, along with other groups of media manipulators, are eager to use journalistic norms as a weapon against journalism.

Reporters and their editors should internalize the idea that social media does not constitute a “person on the street” scenario, nor is an embedded tweet or Facebook post akin to a pulled quote. Regardless of the kind of story being reported, reporters should avoid linking to a handful of social media posts and then attributing that perspective, positive or negative, to “the internet.”

This perspective aligns with the Council of Europe’s “Information Disorder” report (Wardle and Derakhshan 2017), which urges news organizations to exercise extreme caution when dealing with emerging hoaxes and other dis-, mis-, and mal-information. The report is particularly concerned with information that is demonstrably false, and which is intentionally designed to deceive and cause harm. Outlets should avoid publishing this information whenever possible, the report states, especially as part of preemptive debunking stories, which may seek to correct false, manipulative information, but may still spread that information before it has achieved any organic reach of its own. Further, the choice to engage with a false story – even in the effort to refute it – aligns with the interests of the manipulators, who see any form of amplification as a victory

Reporters and editors should be equally aware that extremist groups, along with other groups of media manipulators, are eager to use journalistic norms as a weapon against journalism. In order to spread their messages as far as possible, as quickly as possible, they will engage in strategies such as “source hacking,” as described by Data and Society’s Media Manipulation Initiative research lead Joan Donovan (Scarpelli 2017). This involves seeding false or misleading narratives with authoritative sources, in the hopes that other outlets pick up, amplify, and therefore reinforce the initial falsehood. More journalists on the story, in turn, means more opportunities for more misleading interviews, thus providing the manipulators increasing opportunities to hijack the news narrative.7

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s