Illustration by Jackie Ferrentino

The Junk Cycle

December 4, 2019

At 1:15pm Eastern on Wednesday, May 22, a Trump supporter named Shawn Brooks posted a video on Facebook called “Is Nancy Pelosi drunk?” It suggested that Pelosi, the Speaker of the House, was slurring her words. Later that day, another video featuring Pelosi, slowed so she sounded drunk or incapacitated, appeared on two pro-Trump Facebook pages; later, it moved to Twitter and YouTube. 

In its first few hours online, the doctored video was viewed about three hundred times on Twitter. On Facebook, it was shared 18,881 times—which put it on the same scale as a recent video of a mysterious sea creature someone caught in Alaska. The American far right, fixated on the notion of strength, often attacks its enemies with the idea that they are weak or sick, and the Pelosi video seemed destined mostly to provide whoever happened to be browsing for hatred that day a temporary sense of affirmation. Were it not for subsequent developments, it would likely have slipped back into obscurity among posts targeting Bette Midler for her online activism and accusing Rep. Ilhan Omar of spreading Islamic law in America. 

It is, roughly, the business model of pornography. 

But the next day, the video caught the attention of a new breed of digital journalist, expert in trawling social networks and using Google’s analytical tools to look for attention spikes on a variety of topics—weather, celebrities, terrifying diseases, politics. Their aim is to ride a wave of fascination or outrage with a quickly cobbled-together post that uses only information available online and is headlined to appeal to the algorithms that push stories to the top of readers’ screens. 

Success means that more people click. It is, roughly, the business model of pornography. Every day, hundreds, if not thousands, of digital journalists seek the high that comes from posting first. In aggregate it means you can have anything you want from the news media in 2019, except less. 

The first major outlet to have published the Pelosi video appears to have been the Washington Post, which in February hired its first “competition analyst” to “monitor competitor sites and use digital and social analytics tools” to decide what to cover “right now” (the Post’s italics, in a press release). 

At 6:29pm on May 23, the Post published “Pelosi videos manipulated to make her appear drunk are being shared on social media.” Lou Dobbs’s Fox Business show, on at 7pm, featured a video showing a supposedly drunk Pelosi. At 8:39 that evening President Donald Trump, who is both a Dobbs devotee and acutely aware that desperate journalists will scramble after any morsel he throws, tweeted a reference to the segment. Rudy Giuliani, Trump’s attorney, tweeted a Pelosi video link, too. 

Sign up for CJR's daily email

The strange business of covering what people say on the Web has distinct stages. The first is: there is this thing on social media. That completed, newsrooms could now move on to the second: this famous person has said something about this thing on social media. Every publication and network you can think of—even those, like the New York Times, that publicly decry a cheapened information economy—joined in. A graph provided by Facebook shows views for the video spiking as a direct result. The number of daily shares rose from eighteen thousand to fifty thousand on Facebook alone. It is not recorded how many clicks the publications, most of which hosted their own versions of the video, collectively received. But it must be in the millions.


American newspaper readers of the nineteenth century might have struggled to understand what a Facebook or a Pelosi was, but they’d have been intimately familiar with the dynamics at play. News in that era, and into the twentieth century, according to James L. Baughman, who was a historian at the University of Wisconsin, amounted to a free-for-all, often with the stated aim of manufacturing or coloring facts to serve an agenda. 

The best interests of the public were someone else’s responsibility. The truth, Baughman reported, was not suppressed but could not all be found in any one place, because every editor thought it his responsibility to push a partisan agenda. When Grover Cleveland won the presidency in 1884, the Los Angeles Times, which bitterly opposed him, maintained for eleven days that his rival, James G. Blaine, whom the paper preferred, was president-elect. 

For decades the situation seemed immovable. Simply the way things were. But over time, according to a speech by Baughman, most publications moved toward what we mistily think of as the modern news media—attempting an honest, accurate, and fair representation of events. 

We don’t have to breathlessly detail every video or salacious rumor on our screens.

The reason was not just an outbreak of morals, or a sudden concern for the greater good, but economics. Straight reporting presented a competitive advantage. Gradually, a set of ethics and methods emerged for sorting the screaming chaos of the real world—every bit as ridiculous and messy and tempting as the darkest corners of the Web—into the calm and precise slots of the printed page. The Associated Press—whose general manager at the turn of the twentieth century, Melville Stone, helped define an ethos of responsible reporting—characterized the new way as an abhorrence for inaccuracy, carelessness, bias, and distortion; a refusal to plagiarize; and a determination to verify and seek comment before publication. 

Though most respectable publications still adhere to these rules for their flagship reported stories, much of what they produce in 2019 is Web aggregation. And there, Stone’s ethos is once again the exception rather than the rule. 


In the days after the Pelosi video appeared, the American news media moved on to stage three: an attempt to cast blame anywhere except at itself. Stories like “Facebook refuses to delete fake Pelosi video spread by Trump supporters,” in The Guardian, shifted the debate to the irresponsibility of social media for allowing the video, and other disinformation, to propagate for profit. 

It was followed by stage four: the literary or aesthetic analysis that allows upscale publications to garner clicks while seeming to rise above the whole thing (such as New York magazine’s “Nancy Pelosi Slowed Down 800 Percent Is Hauntingly Beautiful”). As May ended and June began, the story entered its fifth and final stage: breaking down what it all means (as seen in ABC’s “Manipulated Nancy Pelosi video may be a sign of things to come in the 2020 election cycle”). 

The Post, which launched the video to prominence with its first story, published an online guide to spotting manipulated footage. Reuters manufactured its own fake videos to see if its journalists could spot evidence of doctoring. (In my experience, most people can’t.) A Politico story citing the head of Instagram as well as academics from Harvard, the University of California, Berkeley, and elsewhere positioned “deepfakes” created using new technology as an insurmountable hazard for an unfairly victimized media. 

Nobody mentioned a solution that’s possible to achieve already. A solution that would cripple both partisan political disinformation like the Pelosi video and more nefarious online espionage attempts by Russia and others. One that would hamper “populist” politicians seeking to divide and conquer by stoking outrage while simultaneously restoring faith in the news media. Namely, applying the same reporting methods and ethics to the modern free-for-all that news organizations applied to the last free-for-all. 

False videos cannot spread if reporters insist on verifying them before publishing. No matter how convincing a fraud is, it cannot stand up to the bravery of an editor who refuses to cover it until it is confirmed and comment has been sought from everyone in it, no matter how long that takes or who else publishes in the meantime. 

Facebook is a toxic town square. Social media is Christmas for fraudsters. But we don’t have to breathlessly detail every video or salacious rumor on our screens in a fog of adrenaline and fear. The best way of sifting the information we find online is, simply, to report before we publish, as we would offline. 

That would make news outlets once more where readers go to find the most reliable account of what happened over the course of a day, not just what everyone said happened that day, or what arguments they had about it. It’s a service, in fact, that people might even be willing to pay for.

Ravi Somaiya has been a correspondent for HBO’s Vice News Tonight and covered the media and breaking news for the New York Times. He is currently working on a historical nonfiction thriller for Twelve Books.