A Tiny Outrage Machine, Sucking the Exhaust From a Giant One

You can’t even fight a social network without a social network.

Close-up of a woman's face, her eyes replaced by Facebook logos
Lambert/Getty ; Adam Maida / The Atlantic

Frances Haugen, a former Facebook data scientist, copied thousands of pages of internal documents and webpages before she left the company. Then she shared those materials with The Wall Street Journal, which began publishing stories about them last month under the heading “The Facebook Files.” Weeks later, she began to parcel the materials out to a consortium of news organizations, including The Atlantic. In that context, the files have come to be known as “The Facebook Papers,” drawing a lineage back to revelations about the U.S. military intervention in Vietnam 50 years ago—the Pentagon Papers.

But the differences between the Facebook Papers and their Cold War precursor are more relevant than their similarities. The entire mechanism for producing, storing, and disseminating knowledge has changed completely over the past half century. The Pentagon Papers were, in sum, a single narrative: a 3,000-page history of U.S. involvement in Vietnam, annotating 4,000 more pages of primary-source material, and written up to be a coherent historical record for internal use. Journalists had access to the Pentagon’s own, secret history of intervention in Vietnam, which became the basis for reporting on the prior two decades of U.S. activity in Southeast Asia.

The Facebook Papers amount to something different. They comprise, in large part, smartphone pictures Haugen took of her computer screen displaying internal research and communications  on Workplace, a social network for Facebook’s employees. The materials reveal, in places, what the company did or failed to do—it foresaw the January 6 insurrection, for example; it facilitated atrocities overseas—but, just as revealing, they disclose the chatter, the talk, the discourse that Facebook’s own employees have generated about the operation of its products.

In that sense, the Facebook Papers reveal as much about how social media has altered human knowledge as they do about the company’s specific accomplishments or errors. They don’t tell an organization’s story about itself, like the Pentagon Papers do; they show how people inside a company that makes social networks behave on a social network. (Facebook’s desperate name change yesterday, to Meta, seems pretty apt.) In other words, the Facebook Papers cede a victory to Facebook over the control of public knowledge: What happened inside the company, and what it means to the rest of us, has been defined by social media.


Like “nuggs” or “J-Lo,” “the Pentagon Papers” is a nickname. Its official title is less seductive: “Report of the Office of the Secretary of Defense Vietnam Task Force.” Daniel Ellsberg and others, working at the charge of Secretary of Defense Robert McNamara, had been asked to compile Defense Department documents related to Vietnam. But the group, which worked on the task for a year and a half from 1967 to 1969, did much more than that, producing a complete prehistory of the conflict.

Ellsberg, working for the RAND Corporation, copied the papers on a Xerox machine and released the report to The New York Times in 1971. The Times published a portion of the papers, but only as excerpts to buttress its reporting. If you read the Pentagon Papers today (they were declassified and released in full in 2011), you can, well, read them. They contain prose arguments and explanations; they are organized in volumes with a deliberate structure. The report’s authors called it a “history based solely on documents—checked and rechecked with ant-like diligence.”

The report’s nickname wasn’t quite deliberate. When the Times published its first article about the materials, its headline referred to a “Pentagon study.” The phrase Pentagon papers appeared below, used descriptively and in order to avoid saying study over and over. They were, after all, papers from the Pentagon. It’s not clear when the materials evolved into the Pentagon capital-P Papers, but that name did appear on a Time cover story two weeks later.

When the widespread use of computers made storage and dissemination easy, access to secret materials became less unusual. Leaks during the internet era have donned the cape of “Papers,” both in reference to, and aspiration for, the Pentagon Papers’ example. Like the Watergate scandal that would soon follow (which spawned its own naming conventions), the Pentagon Papers represented a high point in American investigative journalism. Through dogged sleuthing and faithful protection of sources, journalists shone light on the shadows, revealing devious scurries.

As journalism consolidated and declined, reporters longed to recapture the magic of newsmaking’s glory days. Putting out “Papers” became one way to do so. In 2016, more than 11 million leaked documents cataloging offshore tax avoidance were dubbed the Panama Papers. Later, similar dumps of similar quantities of kindred materials became the Paradise Papers and the Pandora Papers. In 2019, The Washington Post published a series of interviews on the Afghan War, obtained from the special inspector general for Afghanistan reconstruction, under the name the Afghanistan Papers. These events all took place during the age of fast internet, big data, and cheap storage; holding, let alone distributing (or acquiring by Freedom of Information Act request), terabytes of documents was by then a trivial affair. Papers were documents, and documents were files, and the more the better.

Even so, not all such revelations adopted or seemed to deserve the designation of Papers. Some are mere leaks: materials that would be loosed to everyone, not just traditional media. Papers would be reserved for formal journalistic practice, a way for professionals to distinguish their acts of access and synthesis from those of ordinary folk, or even their whistleblower sources. Reporters and their editors turned leaks into gilded Papers. In many cases, these materials traced what people did, usually in secret, and perhaps in the hopes that no one would see. But few Papers would match the original from 1971. Few would demonstrate what an organization knew, how it thought, and how it represented that knowledge to itself.


When the Haugen affair began, it looked a lot more like the Pentagon Papers: a trove of documents shared with a single, well-respected outlet. A trickle of stories combined to paint a picture of misdeeds and deceits. But narratives diverged when the materials spread to the consortium. A dozen outlets, handpicked by Haugen, established for themselves a set of terms, including an embargo that was lifted on Monday. Even now, Haugen’s team distributes a new and weighty dump of documents to this group—plus at least a dozen additional news outlets in America that asked to access the documents after the embargo—every day, which must be sorted through, interpreted, and turned into stories. The resulting barrage is both powerful and overwhelming.

The consortium’s shotgun approach was conceived by Haugen’s PR team—led by Bill Burton, a former Obama official. In a kind of mission statement for the Facebook Papers effort, The Washington Post quoted the Pulitzer Center executive editor Marina Walker Guevara, who said that the present era’s stories have become “so complex and so multilayered and global” that it would be impossible to report them without a large, global network. (The consortium itself dissolved the day after the embargo lifted, and did not coordinate on the substance of any particular stories.)

Guevara might be right. Enormous, fruitful work went into sifting Haugen’s raw materials, then synthesizing, writing, verifying, and publishing stories that tell us what they mean. How convenient, though, that this futuristic structure for investigative journalism would take the same form as the social internet itself: a worldwide array of characters all clamoring to retrieve and process the same information as instantly as they can in order to compete for a limited supply of globalized attention.

Haugen’s team must have thought this approach would produce the greatest quantity of pressure on Facebook, maybe inspiring intervention. This Hail Mary pass might yet be caught in the end zone—points could still be scored in Congress—but its launch has also reproduced the logic of the company it is meant to beat. The Facebook Papers are, in aggregate, a supersensory supply of materials about a social network, produced on its own, internal social network, prima facie assumed to have meaning whose depth exceeds their surface, and mustered as rapidly as possible to generate emotions. They’re a tiny outrage machine, sucking the exhaust from a much bigger one.

A consortium journalist reading the Facebook Papers, with all its earnest arguments and efforts to do better, might even start to wonder whether Facebook is so purely evil. At the very least, some of its employees clearly struggled with how to address the problems they knew the company had caused. For example, a Washington Post report on how Facebook chose to amplify posts that received emoji reactions—even angry ones—explains that the company eventually removed the same signals from the algorithm on account of their harm. (It did take a while.)

This dance between Facebook’s internal debates and journalists’ interpretation of them as withering revelations repeats the ritual that online debate has normalized: Posts beget discourse that begets ever more posts that take the place of action, let alone knowledge. Depth and surface become indistinguishable, always implying that there’s more to the story, only to recede back into the shadows moments later.

By 1972, after the Pentagon Papers and Watergate had broken, a “they’re out to get you” mentality had overtaken American culture. The internet—the real internet, not the cautionary version described in science fiction—kept all the 1970s paranoia but varnished it in 1980s neon. There was a time when internet utopians such as Clay Shirky celebrated the “collective intelligence” that emerges when thousands or millions of people come together online. Now every day is a new matryoshka doll of conspiracy thinking. We've learned to push back against a world run by angry emoji by posting more angry emoji. Even reality must be amplified, to rise above the noise. But every such move adds fresh clamor too, requiring even more outrage the next time to catch the algorithm’s wind and win attention.

In the end, the Facebook Papers stage a showdown between investigative reporting and conspiratorial paranoia. The Fourth Estate is mounting a last stand against the enemy that would, and largely has, undone it. But no gun even fires anymore unless it’s loaded with internet-tipped bullets.