BuzzFeed Now Looking to Institutional Dems to Police a Phantom Surge of Lefty Fake News

One of my many concerns about the fake fake news scare is that it provides a way to discredit alternative voices, as the PropOrNot effort tried to discredit a number of superb outlets that don’t happen to share PropOrNot’s Neocon approach to Syria. BuzzFeed, in its seemingly unquenchable desire to generate buzz by inflating the threat of fake news, takes that a step further by turning to institutional Democratic outlets — outlets whose credibility got damaged by Hillary’s catastrophic loss — to police an alleged surge of fake news on the left.

First, consider its evidence for a surge in Democrats embracing fake news.

There are new cases daily. Suspicions about his 2020 reelection filing. Theories about the “regime’s” plan for a “coup d’état against the United States” (complete with Day After Tomorrow imagery of New York City buried in snow). Stories based on an unverified Twitter account offering supposed “secrets” from “rogue” White House staffers (followed by more than 650,000 people). Even theories about the Twitter account (“Russian disinformation”).

Since the election, the debunking website Snopes has monitored a growing list of fake news articles aimed at liberals, shooting down stories about a new law to charge protesters with terrorism, a plan to turn the USS Enterprise into a floating casino, and a claim that Vice President Mike Pence put himself through gay conversion therapy.

[snip]

Panicky liberal memes have cascaded across the internet in recent weeks, like an Instagram post regarding Steve Bannon’s powers on the National Security Council shared by a celebrity stylist and actress. Some trolls have even found success making fake news specifically aimed at tricking conservatives.

Let’s take the purported “fake news” story BuzzFeed bases its argument on, one by one:

  • debunking of a Twitter thread (not a finished news piece) of the conclusions about a discovery that Trump, very unusually for a President, filed for reelection immediately after inauguration. There’s no debunking that Trump filed his candidacy, nor that it is unusual, nor, even, that Trump is fundraising off it. That’s not fake news. It’s an attempt to figure out why Trump is doing something unusual, with a fact-checking process happening in the Twitter discussion.
  • An admittedly overblown Medium post about some of the shady things Trump has done, as well as the much rumored claim that the reported sale of 19% of Rosneft confirms the Trump dossier claim that Carter Page would get part of Rosneft if he could arrange the lifting of US sanctions on Russia. The story’s treatment — and especially it’s use of the word “coup” — is silly, but the underlying question of whether Trump will instruct agencies to ignore the law, as already happened in limited form at Dulles over the first weekend of the Muslim ban, as well as the question of how Trump intends to target people of color, is a real one.
  • A story basically talking about the formation of the RoguePotusStaff Twitter account that notes prominently that “there’s no way to verify the authenticity of the newly minted Twitter channel.” BuzzFeed provided no evidence this was being preferentially shared by people on the left.
  • A Twitter thread speculating, based off linguistic analysis, that the RoguePotusStaff account might be Russian disinformation. Again, BuzzFeed made no claims about who was responding to this thread.
  • A debunking of a claim posted in November on a conservative fake news site claiming that protestors would get charged with terrorism.
  • A “debunking” of a satirical story from November posted in the Duffel Blog claiming Trump was going to repurpose an aircraft carrier.
  • A debunking of a fake news story from November claim that Mike Pence had put himself through gay conversion therapy that notes Pence did, indeed, push gay conversation therapy.
  • A liberal trolling effort aimed at conservatives, which started in December, claimed that Trump had removed symbols of Islam from the White House.
  • An instagram post that (BuzzFeed snottily notes) got shared by an actress and a stylist reporting the true fact that Bannon had been added to the National Security Council and noting the arguably true fact that the NSC reviews the kill list including the possibility of targeting Americans (technically, the targeted killing review team installed by Obama is not coincident with the NSC, but it does overlap significantly, and Anwar al-Awlaki was targeted by that process).

Most of these things are not news! Most are not pretending to be news! The only single thing included among BuzzFeed’s “proof” that lefties are resorting to fake news that would support that claim is the Mike Pence story. And to get there, BuzzFeed has to pretend that the Duffel Blog is not explicitly satire, that multiple cases of conservative fake news are lefty fake news, that well-considered discussions on Twitter are fake news, and that we all have to stop following RoguePotusStaff because we don’t know whether its writers are really Rogue POTUS staffers or not.

It’s a shoddy series of claims that BuzzFeed should be embarrassed about making. Effectively, it is calling discussion and satire — including correction — fake news.

To BuzzFeed’s credit, after months of mis-stating what a poll it did revealed — BuzzFeed had been claiming that 75% of people believe fake news, but in reality the poll showed that 75% of those who recall fake news believe it — BuzzFeed finally got that, at least, correct. Bravo BuzzFeed!

But other than that, they’ve got almost nothing here.

Believe it or not, that’s not the most offensive part of this story. Having invented a lefty fake news problem out of satire and Twitter discussions, BuzzFeed then decided it’s important what official Democratic sources thing about it. While one Bernie source said it was best to ignore these things (another said it was a real problem), BuzzFeed framed other responses in terms of left protests of elected officials.

Democratic operatives and staffers at left-leaning media outlets predict that viral anti-Trump conspiracy theories will ultimately distract from real reporting about the administration, undermining legitimate causes for outrage on the left over what the administration is actually doing.

Still, for now, it’s a conversation that exists almost entirely outside the political class itself. Elected officials are not hawking phony stories as true, like Trump’s calls to investigate widespread voter fraud during the election. But that remove poses its own problems for leaders with no obvious way to dismantle widely shared false stories.

“It exists on the left and that’s a problem because it misinforms people,” said Judd Legum, editor in chief of progressive news site ThinkProgress. “That’s harmful in other ways because the time you’re spending talking about that, you could spend talking about other stuff.”

“It contributes to a broader environment of distrust, and it sort of accelerates the post-factual nature of our times,” said Teddy Goff, co-founder of Precision Strategies and a former senior aide to Barack Obama and Hillary Clinton. “Fake news is pretty damaging no matter who it benefits politically. No one on the left should think we ought to be replicating the fake news tactics on the right.”

[snip]

The online energy also raises questions about the party’s relationship with its base. In recent weeks, progressives have pressured lawmakers to adopt a tougher stance toward Trump and join ranks with the millions of protesters who marched over inauguration weekend.

The two top-ranking Democrats in Washington, Chuck Schumer in the Senate and Nancy Pelosi in the House, have both signaled an openness to working on legislation with Trump. Last week, protests formed outside Schumer’s home in Brooklyn. And among progressive activists online, Pelosi was met with vehement push-back after saying the party has a “responsibility to the American people to find our common ground.”

“Elected Democrats are stuck struggling to keep ahead of the anger that the base is feeling right now,” said [Jim] Manley, the former Reid adviser. “It’s very palpable.”

First, BuzzFeed is wrong in saying elected officials are not hawking phony stories as true. One reason the claim that Wikileaks doctored Democratic emails got so much traction is because Dems repeatedly made that claim (and as I’ve noted, Hillary quickly escalated the Alfa News story that most media outlets rejected as problematic).

Worse, BuzzFeed deems Democratic operatives and staffers as somehow chosen to decide what are “legitimate causes for outrage on the left over what the administration is actually doing.” It further suggests there’s a connection between people protesting elected leaders and fake news.

Finally, BuzzFeed shows absolutely no self-awareness about the people it seeks about and the stories they’ve pitched. Consider: Manley is in the very immediate vicinity of the people who got the WaPo to push the claim that CIA had decided Russia hacked the DNC in order to get Trump elected, a conclusion that — we’ve subsequently learned — is the single one any agency in the IC (in this case, the NSA) expressed less confidence in. Moreover, we know that Harry Reid spent months trying to get the FBI to reveal details included in the Trump dossier that no one has been able to confirm. And when the dossier was released, Judd Legum magnified it himself, in much the same way the Medium post did the Rosneft claim.

Oh, and as a reminder: BuzzFeed was the entity that decided it was a good idea to publish an unverified intelligence dossier in the first place!

I mean, if the institutional Dems that BuzzFeed has deemed the arbiters of what is “legitimate” to talk about think the unproven Russian dossier counts, then BuzzFeed has even less in its claim about fake news.

Nevertheless, it thought it was a good idea to assign two journalists to make thinly substantiated claims about a lefty news problem that it then used to police whether lefty protestors are doing the right thing.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

The Three Most Believed Fake News Stories of the Election (Tested by Stanford) Favored Hillary

In a piece repeating erroneous BuzzFeed reporting, the Atlantic expresses concern that the left is now sharing fake news stories just like the right shared them during the election.

If progressives are looking to be shocked, terrified, or incensed, they have plenty of options. Yet in the past two weeks, many have turned to a different avenue: They have shared “fake news,” online stories that look like real journalism but are full of fables and falsehoods.It’s a funny reversal of the situation from November. In the weeks after the election, the press chastised conservative Facebook users for sharing stories that had nothing to do with reality. Hundreds of thousands of people shared stories asserting incorrectly that President Obama had banned the pledge of allegiance in public schools, that Pope Francis had endorsed Donald Trump, and that Trump had dispatched his personal plane to save 200 starving marines.
The phenomenon seemed to confirm theorists’ worst fears about the internet. Given the choice, democratic citizens will not seek out news that challenges their beliefs;  instead, they will opt for content that confirms their suspicions. A BuzzFeed News investigation found that more people shared these fake stories than shared real news in the three months before the election. A follow-up survey suggested that most Americans believed fake news after seeing it on Facebook. When held to the laissez faire editorial standards of Facebook, the market of ideas fails.

As I laid out, BuzzFeed’s claim that most Americans believe fake news was not what BuzzFeed’s poll actually showed; rather, it showed that those who remember fake stories believe them, but that works out to be a small fraction of the people who see the story. And this piece is one of many that points out some methodological problems with BuzzFeed’s count of fake news sharing.

The Atlantic then goes onto cite stuff (like the @AltNatParSer and @RoguePOTUSStaff) that is not verified but might be true but in any case is critique as the left’s new habit of fake news.

All that said, the Atlantic is right that the left can be sucked in by not-true news — but that was true during the election, too. Consider this Stanford study that, generally, found that fake news wasn’t as impactful as often claimed.

We estimate that in order for fake news to have changed the election result, the average fake story would need to have f ≈ 0.0073, making it about as persuasive as 36 television campaign ads.

Buried deep inside the story is a detail one or two people have noted, but not mentioned prominently. Among the fake news stories studied by the authors (which were limited to stories debunked at places like Snopes, which is a significant limit to the study), two stories favorable to Hillary were the most believed.

Blue here is the percentage of the US adult population that believed a story and red is being “not sure.” Both if you aggregate those two categories and if you take only those who affirmatively say they believed something, this story — claiming Congressman Jeff Denham helped broker Trump’s deal for the Trump Hotel in DC — and this story — repeating Kurt Eichenwald’s claim that he had proof WikiLeaks led all the fake stories Stanford tested, with close to 30% definitely believing both (see my post on that story). This story claiming Clinton paid Beyonce for a campaign appearance was the most-believed anti-Hillary story, which came after a third Hillary-friendly story claiming Trump was going to deport Lin Manuel-Miranda (note, as also shown in other studies, the fake news stories weren’t recalled or believed at the same rates as the true ones, though in the aggregate, the Denham story rivaled “small true” stories).

Note, the Stanford study did not test this story, which also claimed Wikileaks had doctored emails. It appeared on the same Clinton site three days earlier, which was itself based off a fake news created by a Hillary supporter (with some spooky ties), and magnified by Malcolm Nance and Joy Reid. Those two stories likely reinforced each other.

I’m interested in both of these stories — in part, because the reality about Trump’s corruption and his ties to Russia are both bad enough, without Democratic operatives inventing stories about it. But obviously, I’m particularly interested in the latter, in part because so even in spite of the real evidence implicating Russia in the hack of the DNC, Democrats tend to believe anything involving Russia without evidence.

That’s ironic, given that the risk of fake news is supposed to stem from Putin poisoning our airwaves.

Update: I’ve added “three” to the title because a number of people said it would make it more clear. Thanks to those who suggested it.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

The Latest “Fake” News Tizzy: Garbage In, Garbage Out

It seems that if you label something “fake news” and add some pretty charts, a certain class of people will share it around like others share pictures of big breasted women or stories accusing Hillary of murder.

Consider this study published at CJR, which purports to be the “calculated look at fake news’s reach” that has been “missing from the conversation” (suggesting its author may be unfamiliar with this Stanford study or even the latest BuzzFeed poll).

It has pretty pictures everyone is passing around on Twitter (the two charts look sort of like boobies, don’t you think?).

Wowee, that must be significant, huh? That nearly 30% of what it calls “fake news” traffic comes from Facebook, but only 8% of “real news” traffic does?

I’ve seen none of the people passing around these charts mention this editor’s note, which appears at the very end of the piece.

Editor’s note: We are reporting the study findings listed here as a service to our audience, but we do not endorse the study authors’ baseline assumptions about what constitutes a “fake news” outlet. Drudge Report, for instance, draws from and links to reports in dozens of credible, mainstream outlets. 

Drudge may be a lot of things (one of which is far older than the phenomenon that people like to label as fake news). But it’s not, actually, fake news.

Moreover, one of the points that becomes clear if you look at the pretty pictures in this study closely is that Drudge skews all the results, which is unsurprising given the traffic that it gets. That’s important for several reasons, one of which being that Drudge is in no way what most people consider fake news, and if it were, then the outlets that fall over themselves to get linked there would themselves be fake news and we could just agree that our journalism, more generally, is crap. More importantly, though, the centrality of the Drudge skew in the study — along with the editor’s note — ought to alert anyone giving the study a marginally critical review that the underlying categorization is seriously problematic.

To understand whether these charts mean anything, let’s consider how the author, Jacob Nelson, defines “fake news.”

As has become increasingly clear, “fake news” is neither straightforward nor easy to define. And so when we set out on this project, we referred to a list compiled by Melissa Zimdars, a media professor at Merrimack College in Massachusetts. The news sites on this list fall on a spectrum, which means that while some of the sites we examined* publish obviously inaccurate news (e.g., abcnews.com.co), others exist in a more ambiguous space, wherein they might publish some accurate information buried beneath misleading or distorted headlines (e.g., Drudge Report, Red State). Then there are intentionally satirical news sources, like The Onionand Clickhole. Our sample included examples of all of these types of fake news.

We also analyzed metrics for real news sites. That list represents a mix of 24 newspapers, broadcast, and digital-first publishers (e.g., Yahoo-ABC News, CNN, The New York Times, The Washington Post, Fox News, and BuzzFeed).

Nelson actually doesn’t link through to the list. He links through to a story on the list, which at one point had been removed from public view but which is now available. The professor’s valuation, by itself, has problems, not least that she judges these sites against what she claims is “traditional” journalism (itself a sign of either limited knowledge and/or bias about journalism as a whole).

But as becomes clear, you see that she has simply listed a bunch of sites and categorized them, including “political” and “credible.”

*Political (tag political): Sources that provide generally verifiable information in support of certain points of view or political orientations.  

*Credible (tag reliable): Sources that circulate news and information in a manner consistent with traditional and ethical practices in journalism (Remember: even credible sources sometimes rely on clickbait-style headlines or occasionally make mistakes. No news organization is perfect, which is why a healthy news diet consists of multiple sources of information).

[snip]

Note: Tags like political and credible are being used for two reasons: 1.) they were suggested by viewers of the document or OpenSources and circulate news 2.) the credibility of information and of organizations exists on a continuum, which this project aims to demonstrate. For now, mainstream news organizations are not included because they are well known to a vast majority of readers.

She actually includes (but misspells, with an initial cap) emptywheel in her list, which she considers “political” but not “credible.”

The point, however, is that sites will only be on this list if they are considered non-mainstream. And the list is in no way considered a list of “fake news,” but instead, a categorization of a bunch of kinds of news. In fact, the list doesn’t even consider Drudge “fake,” but instead qualifies it as “bias.”

So you’ve got the study using a list that is, itself, pretty problematic, which then treats the list as something that it’s not. Both levels of analysis, however, set up  false dichotomies (between “fake” and “real” in the study, and between “traditional” or “mainstream” and something else (which remains undefined but is probably something like “online outlets a biased media professor was not familiar with”) in the list. Ultimately, though, the study and the list end up distinguishing minor-web-based publications, of a range of types, from “major” (ignoring that Drudge and a number of other huge sites are in the list) outlets, some but not all of which actually engage in so-called “traditional” journalism.

You can see already why aiming to learn something about how people access these sites is effectively tautological, as the real distinction here is not between “fake” and “real” at all, but between “web-dependent” and something else.

Surprise, surprise, having substituted a web-based/other distinction for a “fake”/real one, you get charts that show the “fake” news (which is really web-based) rely more on web-based distribution methods.

Nope. Those charts don’t actually mean anything.

That doesn’t mean the rest of the study is worthless (though it is, repeatedly, prone to the kind of poor analysis exhibited by misunderstanding your basic data set). The study shows that not many people read “fake” news (excluding Drudge), and that people who read “fake” news (remember, this includes emptywheel!) also read “real” news.

Ultimately, the entire study is just a mess.

Which makes it a far more interesting example — as so much of the “fake news” panic does — of how precisely the people suggesting that “fake news” is a sign of poor critical thinking skills that will harm our democracy themselves also may lack critical thinking skills.

Update: This post raises some methodology questions about the way BuzzFeed defines “fake news.”

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

BuzzFeed Discovers We’re Not the Rubes It Has Claimed, But Insists We Still Have a Fake News Problem

Back in December, I called out BuzzFeed for a bogus news story about fake news. Based on a poll it commissioned, it claimed that 75% of people believe fake news. That’s not what the poll showed. Rather, it showed few people recalled headlines BuzzFeed had IDed as fake, but those who did believed it. It also showed that people recalled and believed “real” news more than they did fake.

[T]he poll showed that of the people who remember a given headline, 75% believed it. But only about 20% remembered any of these headlines (which had been shared months earlier). For example, 72% of the people who remembered the claim that an FBI Agent had been found dead believed it, but only 22% actually remembered it; so just 16% of those surveyed remembered and believed it. The recall rate is worse for the stories with higher belief rates. Just 12% of respondents remembered and believed the claim that Trump sent his own plane to rescue stranded marines. Just 8% remembered and believed the story that Jim Comey had a Trump sign in his front yard, and that made up just 123 people out of a sample of 1809 surveyed.

Furthermore, with just one exception, people recalled the real news stories tested more than they did the fake and with one laudable exception (that Trump would protect LGBTQ citizens; it is “true” that he said it but likely “false” that he means it), people believed real news at rates higher than they did fake. The most people — 22% — recalled the fake story about the FBI Agent, comparable to the 23% who believed some real story about girl-on-girl pictures involving Melania. But 34% remembered Trump would “absolutely” register Muslims and 57% remembered Trump’s claim he wasn’t going to take a salary.

BuzzFeed is back with another poll. Here’s what Craig Silverman claims this poll shows.

The online survey of 1,007 American adults found roughly the same percentage of American adults said they consumed news in the past the month on Facebook (55%) as on broadcast TV (56%). Those were by far the most popular sources of news, followed by print newspapers (39%), cable news (38%), “social media (generally)” (33%), and newspapers’ websites (33%).

But a significant gap emerged when people were asked how much they trust the news they get from these sources. Broadcast TV once again scored the highest, with 59% of respondents saying they trust news from that source all or most of the time.

In contrast, only 18% of respondents trust news on Facebook all or most of the time — and 44% said they rarely or almost never trust news on Facebook.

And unlike the last time Silverman read a poll, that is what the poll actually shows: people say they get “news” from Facebook but don’t really trust it, in contradistinction from the “news” they get from broadcast TV. In case you’re wondering (because BuzzFeed didn’t include this in its narrative of whether people read and trust news), 23% of people get “news” from online only publications like this one and like BuzzFeed; 35% trust it most or all of the time. Given how much BuzzFeed has been claiming that we have a “fake news crisis” driven by Facebook, you’d think they’d find low trust rates for Facebook to be great news. “Golly, people aren’t the rubes we’ve been getting clicks telling you they are, sorry.”

But BuzzFeed doesn’t do that. Instead, it returns to overstating what its last poll showed (though not quite as badly, this time), to suggest that people may believe Facebook, even if they don’t trust it .

But other research suggests that trust is not the same as belief — and that beliefs can be shaped even by distrusted sources. A recent online poll conducted by Ipsos for BuzzFeed News found that on average about 75% of American adults believed fake news headlines about the election when they recalled seeing them. The headlines tested were among those that received the highest overall engagement (shares, reactions, comments) on Facebook during the election, which means they received a large amount of exposure on the platform. A research paper published this week also found that just over half of people who recalled seeing viral fake election news headlines believed them to be true.

So while American adults say they don’t trust a lot of the news they see on Facebook, that apparently doesn’t stop many of them from believing it.

As a threshold matter, note that BuzzFeed’s pollster, Ipsos, appears not to have defined “news,” “trust,” or (the last time) “belief.” I guess it is unsurprising that someone claiming we have a “fake news” crisis believes there are essential terms, because fetishizing “news” is a key part of pitching “fake news” as something new.

But even though BuzzFeed introduced both apples and oranges in its bowl of fruit, I do find the numbers taken in conjunction instructive. Had BuzzFeed actually done the math I did last time — showing that recall rates for fake news are actually low so the poll did not, in fact, show that “many” people believe fake news — it might consider the possibility that when people read stuff on Facebook they don’t find to be credible, they don’t retain it. BuzzFeed measured different things, but the 18% of people who believe everything they read on Facebook is not far off from the 8-16% of people who remembered and believed fake news headlines from Twitter.

If you want a crisis, I’d say look to cable news, which is where 38% of people say they got their news, with half trusting it most or all the time. It appears more disinformation gets disseminated and retained that way than via Facebook. But as a very old problem, I guess that wouldn’t give BuzzFeed the same clicks as the Facebook fake news panic does.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

On “Fake News”

I’ve been getting into multiple Twitter fights about the term “fake news” of late, a topic about which I feel strongly but which I don’t have time to reargue over and over. So here are the reasons I find the term “fake news” to be counterproductive, even aside from the way Washington Post magnified it with the PropOrNot campaign amidst a series of badly reported articles on Russia that failed WaPo’s own standards of “fake news.”

Most people who use the term “fake news” seem to be fetishizing something they call “news.” By that, they usually mean the pursuit of “the truth” within an editor-and-reporter system of “professional” news reporting. Even in 2017, they treat that term “news” as if it escapes all biases, with some still endorsing the idea that “objectivity” is the best route to “truth,” even in spite of the way “objectivity” has increasingly imposed a kind of both-sides false equivalence that the right has used to move the Overton window in recent years.

I’ve got news (heh) for you America. What we call “news” is one temporally and geographically contingent genre of what gets packaged as “news.” Much of the world doesn’t produce the kind of news we do, and for good parts of our own history, we didn’t either. Objectivity was invented as a marketing ploy. It is true that during a period of elite consensus, news that we treated as objective succeeded in creating a unifying national narrative of what most white people believed to be true, and that narrative was tremendously valuable to ensure the working of our democracy. But even there, “objectivity” had a way of enforcing centrism. It excluded most women and people of color and often excluded working class people. It excluded the “truth” of what the US did overseas. It thrived in a world of limited broadcast news outlets. In that sense, the golden age of objective news depended on a great deal of limits to the marketplace of ideas, largely chosen by the gatekeeping function of white male elitism.

And, probably starting at the moment Walter Cronkite figured out the Vietnam War was a big myth, that elite narrative started developing cracks.

But several things have disrupted what we fetishize as news since them. Importantly, news outlets started demanding major profits, which changed both the emphasis on reporting and the measure of success. Cable news, starting especially with Fox but definitely extending to MSNBC, aspired to achieve buzz, and even explicitly political outcomes, bringing US news much closer to what a lot of advanced democracies have — politicized news.

And all that’s before 2002, what I regard as a key year in this history. Not only was traditional news struggling in the face of heightened profit expectations even as the Internet undercut the press’ traditional revenue model. But at a time of crisis in the financial model of the “news,” the press catastrophically blew the Iraq War, and did so at a time when people like me were able to write “news” outside of the strictures of the reporter-and-editor arrangement.

I actually think, in an earlier era, the government would have been able to get away with its Iraq War lies, because there wouldn’t be outlets documenting the errors, and there wouldn’t have been ready alternatives to a model that proved susceptible to manipulation. There might eventually have been a Cronkite moment in the Iraq War, too, but it would have been about the conduct of the war, not also about the gaming of the “news” process to create the war. But because there was competition, we saw the Iraq War as a journalistic failure when we didn’t see earlier journalistic complicity in American foreign policy as such.

Since then, of course, the underlying market has continued to change. Optimistically, new outlets have arisen. Some of them — perhaps most notably HuffPo and BuzzFeed and Gawker before Peter Thiel killed it — have catered to the financial opportunities of the Internet, paying for real journalism in part with clickbait stories that draw traffic (which is just a different kind of subsidy than the family-owned project that traditional newspapers often relied on, and these outlets also rely on other subsidies). I’m pretty excited by some of the journalism BuzzFeed is doing right now, but it’s worth reflecting their very name nods to clickbait.

More importantly, the “center” of our national — indeed, global — discourse shifted from elite reporter-and-editor newspapers to social media, and various companies — almost entirely American — came to occupy dominant positions in that economy. That comes with the good and the bad. It permits the formulation of broader networks; it permits crisis on the other side of the globe to become news over here, in some but not all spaces, it permits women and people of color to engage on an equal footing with people previously deemed the elite (though very urgent digital divide issues still leave billions outside this discussion). It allows our spooks to access information that Russia needs to hack to get with a few clicks of a button. It also means the former elite narrative has to compete with other bubbles, most of which are not healthy and many of which are downright destructive. It fosters abuse.

But the really important thing is that the elite reporter-and-editor oligopoly was replaced with a marketplace driven by a perverse marriage of our human psychology and data manipulation (and often, secret algorithms). Even assuming net neutrality, most existing discourse exists in that marketplace. That reality has negative effects on everything, from financially strapped reporter-and-editor outlets increasingly chasing clicks to Macedonian teenagers inventing stories to make money to attention spans that no longer get trained for long reads and critical thinking.

The other thing to remember about this historical narrative is that there have always been stories pretending to present the real world that were not in fact the real world. Always. Always always always. Indeed, there are academic arguments that our concept of “fiction” actually arises out of a necessary legal classification for what gets published in the newspaper. “Facts” were insults of the king you could go to prison for. “Fiction” was stories about kings that weren’t true and therefore wouldn’t get you prison time (obviously, really authoritarian regimes don’t honor this distinction, which is an important lesson in their contingency). I have been told that fact/fiction moment didn’t happen in all countries, and it happened at different times in different countries (roughly tied, in my opinion, to the moment when the government had to sustain legitimacy via the press).

But even after that fact/fiction moment, you would always see factual stories intermingling with stuff so sensational that we would never regard it as true. But such sensational not-true stories definitely helped to sell newspapers. Most people don’t know this because we generally learn a story via which our fetishized objective news is the end result of a process of earlier news, but news outlets — at least in the absence of heavy state censorship — have always been very heterogeneous.

As many of you know, a big part of my dissertation covered actual fiction in newspapers. The Count of Monte-Cristo, for example, was published in France’s then equivalent of the WSJ. It wasn’t the only story about an all powerful figure with ties to Napoleon Bonaparte that delivered justice that appeared in newspapers of the day. Every newspaper offered competing versions, and those sold newspapers at a moment of increasing industrialization of the press in France. But even at a time when the “news” section of the newspaper presented largely curations of parliamentary debates, everything else ran the gamut from “fiction,” to sensational stuff (often reporting on technology or colonies), to columns to advertisements pretending to be news.

After 1848 and 1851, the literary establishment put out alarmed calls to discipline the literary sphere, which led to changes that made such narratives less accessible to the kind of people who might overthrow a king. That was the “fictional narrative” panic of the time, one justified by events of 1848.

Anyway, if you don’t believe me that there has always been fake news, just go to a checkout line and read the National Enquirer, which sometimes does cover people like Hillary Clinton or Angela Merkel. “But people know that’s fake news!” people say. Not all, and not all care. It turns out, some people like to consume fictional narratives (I have actually yet to see analysis of how many people don’t realize or care that today’s Internet fake news is not true). In fact, everyone likes to consume fictional narratives — it’s a fundamental part of what makes us human — but some of us believe there are norms about whether fictional narratives should be allowed to influence how we engage in politics.

Not that that has ever stopped people from letting religion — a largely fictional narrative — dictate political decisions.

So to sum up this part of my argument: First, the history of journalism is about the history of certain market conditions, conditions which always get at least influenced by the state, but which in so-called capitalist countries also tend to produce bottle necks of power. In the 50s, it was the elite. Now it’s Silicon Valley. And that’s true not just here! The bottle-neck of power for much of the world is Silicon Valley. To understand what dictates the kinds of stories you get from a particular media environment, you need to understand where the bottle-necks are. Today’s bottle-neck has created both what people like to call “fake news” and a whole bunch of other toxins.

But also, there has never been a time in media where not-true stories didn’t comingle with true stories, and at many times in history the lines between them were not clear to many consumers. Plus, not-true stories, of a variety of types, can often have a more powerful influence than true ones (think about how much our national security state likes series like 24). Humans are wired for narrative, not for true or false narrative.

Which brings us to what some people are calling “fake news” — as if both “fake” and “news” aren’t just contingent terms across the span of media — and insisting it has never existed before. These people suggest the advent of deliberately false narratives, produced both by partisans, entrepreneurs gaming ad networks, as well as state actors trying to influence our politics, narratives that feed on human proclivity for sensationalism (though stories from this year showed Trump supporters had more of this than Hillary supporters) served via the Internet, are a new and unique threat, and possibly the biggest threat in our media environment right now.

Let me make clear: I do think it’s a threat, especially in an era where local trusted news is largely defunct. I think it is especially concerning because powers of the far right are using it to great effect. But I think pretending this is a unique moment in history — aside from the characteristics of the marketplace — obscures the areas (aside from funding basic education and otherwise fostering critical thinking) that can most effectively combat it. I especially encourage doing what we can to disrupt the bottle-neck — one that happens to be in Silicon Valley — that plays on human nature. Google, Facebook, and Germany have all taken initial steps which may limit the toxins that get spread via a very American bottle-neck.

I’m actually more worried about the manipulation of which stories get fed by big data. Trump claims to have used it to drive down turnout; and the first he worked with is part of a larger information management company. The far right is probably achieving more with these tailored messages than Vladimir Putin is with his paid trolls.

The thing is: the antidote to both of these problems is to fix the bottle-neck.

But I also think that the most damaging non-true news story of the year was Bret Baier’s claim that Hillary was going to be indicted, as even after it was retracted it magnified the damage of Jim Comey’s interventions. I always raise that in Twitter debates, and people tell me oh that’s just bad journalism not fake news. It was a deliberate manipulation of the news delivery system (presumably by FBI Agents) in the same way the manipulation of Facebooks algorithms feeds so-called fake news. But it had more impact because more people saw it and people may retain news delivered as news more. It remains a cinch to manipulate the reporter-and-editor news process (particularly in an era driven by clicks and sensationalism and scoops), and that is at least as major a threat to democracy as non-elites consuming made up stories about the Pope.

I’ll add that there are special categories of non-factual news that deserve notice. Much stock market reporting, especially in the age of financialization, is just made up hocus pocus designed to keep the schlubs whom the elite profit off of in the market. And much reporting on our secret foreign policy deliberately reports stuff the reporter knows not to be true. David Sanger’s recent amnesia of his own reporting on StuxNet is a hilarious example of this, as is all the Syria reporting that pretends we haven’t intervened there. Frankly, even aside from the more famous failures, a lot of Russian coverage obscures reality, which discredits reports on what is a serious issue. I raise these special categories because they are the kind of non-true news that elites endorse, and as such don’t raise the alarm that Macedonian teenagers making a buck do.

The latest panic about “fake news” — Trump’s labeling of CNN and Buzzfeed as such for disseminating the dossier that media outlets chose not to disseminate during the election — suffers from some of the same characteristics, largely because parts of it remain shrouded in clandestine networks (and because the provenance remains unclear). If American power relies (as it increasingly does) on secrets and even outright lies, who’s to blame the proles for inventing their own narratives, just like the elite do?

Two final points.

First, underlying most of this argument is an argument about what happens when you subject the telling of true stories to certain conditions of capitalism. There is often a tension in this process, as capitalism may make “news” (and therefore full participation in democracy) available to more people, but to popularize that news, businesses do things that taint the elite’s idealized notion of what true story telling in a democracy should be. Furthermore, at no moment in history I’m aware of has there been a true “open” market for news. It is always limited by the scarcity of outlets and bandwidth, by laws, by media ownership patterns, and by the historically contingent bottle-necks that dictate what kind of news may be delivered most profitably. One reason I loathe the term “fake news” is because its users think the answer lies in non-elite consumers or in producers and not in the marketplace itself, a marketplace created in and largely still benefitting the US. If “fake news” is a problem, then it’s a condemnation of the marketplace of ideas largely created by the US and elites in the US need to attend to that.

Finally, one reason there is such a panic about “fake news” is because the western ideology of neoliberalism has failed. It has led to increased authoritarianism, decreased quality of life in developed countries (but not parts of Africa and other developing nations), and it has led to serial destabilizing wars along with the refugee crises that further destabilize Europe. It has failed in the same way that communism failed before it, but the elites backing it haven’t figured this out yet. I’ll write more on this (Ian Welsh has been doing good work here). All details of the media environment aside, this has disrupted the value-laden system in which “truth” exists, creating a great deal of panic and confusion among the elite that expects itself to lead the way out of this morass. Part of what we’re seeing in “fake news” panic stems from that, as well as a continued disinterest in accountability for the underlying policies — the Iraq War and the Wall Street crash and aftermath especially — enabled by failures in our elite media environment. But our media environment is likely to be contested until such time as a viable ideology forms to replace failed neoliberalism. Sadly, that ideology will be Trumpism unless the elite starts making the world a better place for average folks. Instead, the elite is policing discourse-making by claiming other things — the bad true and false narratives it, itself, doesn’t propagate — as illegitimate.

“Fake news” is a problem. But it is a minor problem compared to our other discursive problems.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

The Bible Still Outperforms Facebook in Delivering Fake News

screen-shot-2016-12-07-at-8-30-48-am

We’ve reached the stage where articles about fake news themselves engage in fake news tactics.

Buzzfeed’s Craig Silverman — who has written many of the stories on fake news in recent weeks — had Ipsos do a poll querying whether or not people believed some of the real and fake news headlines that got shared around during the election. He presented the results, in both tweets and his BuzzFeed article on the results, this way:

screen-shot-2016-12-07-at-7-25-37-am

But that’s not actually what the poll showed, though a number of people — even some of the people who are the most dedicated serious commentators on fake news — seemed to believe the headline without reading the article closely (that is, they treated it precisely like fake news consumers might, including sharing it before they had evaluated it critically).

Rather, the poll showed that of the people who remember a given headline, 75% believed it. But only about 20% remembered any of these headlines (which had been shared months earlier). For example, 72% of the people who remembered the claim that an FBI Agent had been found dead believed it, but only 22% actually remembered it; so just 16% of those surveyed remembered and believed it. The recall rate is worse for the stories with higher belief rates. Just 12% of respondents remembered and believed the claim that Trump sent his own plane to rescue stranded marines. Just 8% remembered and believed the story that Jim Comey had a Trump sign in his front yard, and that made up just 123 people out of a sample of 1809 surveyed.

Furthermore, with just one exception, people recalled the real news stories tested more than they did the fake and with one laudable exception (that Trump would protect LGBTQ citizens; it is “true” that he said it but likely “false” that he means it), people believed real news at rates higher than they did fake. The most people — 22% — recalled the fake story about the FBI Agent, comparable to the 23% who believed some real story about girl-on-girl pictures involving Melania. But 34% remembered Trump would “absolutely” register Muslims and 57% remembered Trump’s claim he wasn’t going to take a salary.

The exception should be an exception, because Buzzfeed shouldn’t have treated it as news anyway. Just 11% recalled Mike Morell’s endorsement, titled “I ran the CIA. Now I’m endorsing Hillary Clinton,” which appeared in NYT’s opinion section. All endorsements should be considered opinion, and this one happens to be from a proven liar with a history of torture apology, so for the rare people who knew anything about Morell, I would hope his opinion would carry limited weight.

What all of this shows is that the fake news headline claims Buzzfeed made last month, that “Fake Election News Stories Outperformed Real News On Facebook,” should be revised. What that clickbait story actually showed was that the top fake stories received more “engagement” — shares, reactions, and comments — on Facebook than the top real news. But the last paragraph of the article admitted that might not be the same as actual consumption or even non-Facebook moderated engagement.

It’s important to note that Facebook engagement does not necessarily translate into traffic. This analysis was focused on how the best-performing fake news about the election compared with real news from major outlets on Facebook. It’s entirely possible — and likely — that the mainstream sites received more traffic to their top-performing Facebook content than the fake news sites did. As as the Facebook spokesman noted, large news sites overall see more engagement on Facebook than fake news sites.

What this newly reported poll at least suggests (one would need to do a more scientific study to test this hypothesis) is that even the most shared fake news was not really retained, whereas more of the real news was. And that’s true even in spite of the fact that Buzzfeed/Ipsos did not test the most popular real news (in reality this, too, is an opinion piece), “Trump’s history of corruption is mind-boggling. So why is Clinton supposedly the corrupt one?” That’s a pity, because it’d be interesting to see how many and what kind of people remembered and believed that one.

Effectively, then, Buzzfeed was testing the most popular fake news (about the Pope endorsing Trump, with 960,000 engagements) against the third ranking real news (the Melania girl-on-girl story, with 531,000 engagements) and real news still performed better overall in terms of recall. Which would seem to suggest these Facebook engagements don’t actually track how much “news” — fake or not — people will consciously retain (I admit unconscious retention is probably an issue too).

Which is how I get to my claim that the Bible outperforms Facebook for spreading false news. After all, as recently as 2014, 42% of Americans believed in creationism, while just 19% believed in evolution. That number is changing quickly (importantly, as more purportedly fake news consuming youngsters who don’t consider themselves religious get asked). Nevertheless, a significantly larger chunk of the country believes that God plunked us down fully-formed into Eden than believe that an FBI Agent involved in the Clinton case died in a murder suicide.

We should expect more people to believe what they read in the Bible, because it is a story that gets reinforced week after week by people with some authority in the community. It also gets reinforced in institutions like the Creation Museum, where I took the picture of white Adam and Eve above. For people who believe in creationism, their religion is fundamentally tied to their self-identity in a way that politics might not be. It is precisely for that reason it provides important counterpoint to these fake news stories. Especially given the way that a preference for religious stories over scientific ones poisons so much of our ability to deal with crises like climate change.

Don’t get me wrong: algorithmically-delivered sensationalism is a problem (as are polls that get shared to make claims about headlines they don’t really support). But it is one of many problems with our politics, and the evidence from this poll actually suggests it isn’t yet the most urgent one.

Update: Pope Francis, who believes the notion of evolution can coexist with that of creation, just issued a statement calling those who spread shit news sinners.

Francis told the Belgian Catholic weekly “Tertio” that spreading disinformation was “probably the greatest damage that the media can do” and using communications for this rather than to educate the public amounted to a sin.

Using precise psychological terms, he said scandal-mongering media risked falling prey to coprophilia, or arousal from excrement, and consumers of these media risked coprophagia, or eating excrement.

[snip]

“I think the media have to be very clear, very transparent, and not fall into – no offence intended – the sickness of coprophilia, that is, always wanting to cover scandals, covering nasty things, even if they are true,” he said.

Update: Matthew Ingram covers this issue at Fortune.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

The #FakeNews about Iraqi WMD Got Hundreds of Thousands Killed

screen-shot-2016-12-06-at-2-12-32-pm

This morning, Dana Milbank — who used to have a podcast with Chris Cillizza on which he once suggested Hillary would choose to drink Mad Bitch beer —  wrote a piece warning of the dangers of fake news.

After writing about a threatening email he received, Milbank considered whether episodes like the attack on Comet Ping Pong — which he described as “the family pizza place in Northwest Washington I’ve been frequenting with my daughter ever since she was a toddler a decade ago” — were the new normal. Milbank described the role of Alex Jones in making a “bogus and bizarre accusation” against Hillary. Then he turned the attack on Comet Ping Pong, in part, into an attack on the media.

This would appear to be the new normal: Not only disagreeing with your opponent but accusing her of running a pedophilia ring, provoking such fury that somebody takes it upon himself to start shooting. Not only chafing when criticized in the press but stoking anti-media hysteria that leads some supporters to threaten to kill journalists.

The man whose “Mad Bitch beer” comment targeted Hillary ended his piece by scolding Trump for fueling rage against Hillary and those who support her.

If Trump were a different leader, he would declare that political violence is unacceptable in a free society. Perhaps he’d say it after eating a “Steel Wills” pie at Comet.

But instead he continues to fuel rage against his opponents and his critics.

On Twitter, Peter Singer — who wrote a very worthwhile book that uses fiction to lay out near term threats to the US  — RTed Milbank’s story with the comment, “stop winking and nodding” at fake news because it can get people killed.

screen-shot-2016-12-06-at-2-29-06-pm

Singer works at New America Foundation, but he used to work at Brookings Institution, which employs people like Michael O’Hanlon and Charles Lister to write propaganda, funded in part by Qatar, designed to generate support for endless wars in the Middle East.

In response to Singer’s tweet, I RTed it and pointed out that “The #fakenews about Iraqi WMD DID get hundreds of thousands killed.” That in turn led to some interesting discussions, most notably with Zeynep Tufekci, who claimed that by “conflating two very, very different types of failure” I was being unhelpful because those different kinds of fake news operated via different mechanisms.

Tufekci is right. The means by which an uncritical press — enthusiastically joined by the WaPo’s editorial page and many, but not all, of its reporters — parroted Dick Cheney’s lies about Iraqi WMD are different than the means by which millions of people sought out the most outrageous claims about Hillary. The means by which the financial press claimed the housing market would never collapse are different than the means by which millions of people sought out conspiracy theories about the people who didn’t prosecute the banksters. The means by which Dana Milbank got to insinuate the Secretary of State might choose Mad Bitch beer are even different than the means by which millions of people sought out news that called the former Secretary of State #CrookedHillary. The means by which the traditional press focused more attention on Hillary’s email server than on Trump’s fraudulent business practices are different than the means by which millions of people sought out claims that Hillary’s email server was going to get her indicted. All of those traditional news examples of fake news included an editorial process designed to prevent the retelling of fake news.

The means by which traditional news media shares fake news are different than social media’s algorithm driven means of sharing fake news.

Until you remember that a week before the election, Fox’s Bret Baier, who eight months earlier had moderated a GOP primary debate, reported that the investigations into Hillary “will continue to likely an indictment.” While Baier retracted the claim just over a day later, the claim was among the most damaging pieces of fake news from the campaign, not least because it confirmed some of what the most inflammatory social media claims were saying and magnified the damage of Jim Comey’s irresponsible announcement about finding new emails.

Baier got manipulated by his sources who knew how to game the means the press uses to avoid fake news. Baier got manipulated into sharing fake news that served the goals of his sources. It turns out Baier was not any more immune from the manipulation of his biases than your average news consumer is.

Now, the NYT (though not, I think, the WaPo) apologized for their WMD coverage and Milbank apologized for his Mad Bitch podcast and Baier apologized for his indictment scoop. No one has yet apologized for focusing more attention on Hillary’s email server than Trump’s own corruption, but I’m sure that’s coming. I’m not aware that the financial press apologized for the cheerleading that ultimately led to millions of Americans losing their homes to foreclosure, but then it also hasn’t stopped the same kind of fake news cheerleading that led to the crash.

Indeed, while it shows remorse after some of the worst cases, the traditional news media still lapses into the habit of reporting fake news, often in a tone of authority and using an elite discourse. Such lapses usually happen when a kind of herd instinct or a rush to get the news first sets in, leading news professionals to tell fake news stories.

And, now that social media has given average news consumers the ability (and after financialization has led to the disappearance of reliable local news), average news consumers increasingly bypass news professionals, listening instead to the stories they want to hear, told in a way that leads them to feel they are assuming a kind of self-control, told in a language and tone they might use themselves. At its worst — as in the case of PizzaGate — a kind of herd instinct sets in, with news consumers reinforcing each others’ biases. On Sunday, that almost got a lot of innocent people — families like Milbank’s own — killed.

Elite commentators may view the herd instincts of average news consumers to be more crude than the herd instincts of professional news tellers. Perhaps they are. Across history, both types of herd instincts have led to horrible outcomes, including to the deaths of hundreds of thousands, even millions of people.

But as we try to deal with our herd instincts and the mistakes we all make (myself very much included), we might do well to exhibit a little less arrogance about it. That certainly won’t eliminate the mistakes; we are, ultimately, herd animals. But it might provide a basis to rebuild some trust, without which leads all of us — the professionals and the average news consumers — further into our own bubbles.

Update: This Current Affairs piece treats WaPo’s peddling of fake news — including the PropOrNot story — well.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

Look Closer to Home: Russian Propaganda Depends on the American Structure of Social Media

The State Department’s Undersecretary for Public Diplomacy, Richard Stengel, wanted to talk about his efforts to counter Russian propaganda. So he called up David Ignatius, long a key cut-out the spook world uses to air their propaganda. Here’s how the column that resulted starts:

“In a global information war, how does the truth win?”

The very idea that the truth won’t be triumphant would, until recently, have been heresy to Stengel, a former managing editor of Time magazine. But in the nearly three years since he joined the State Department, Stengel has seen the rise of what he calls a “post-truth” world, where the facts are sometimes overwhelmed by propaganda from Russia and the Islamic State.

“We like to think that truth has to battle itself out in the marketplace of ideas. Well, it may be losing in that marketplace today,” Stengel warned in an interview. “Simply having fact-based messaging is not sufficient to win the information war.”

It troubles me that the former managing editor of Time either believes that the “post-truth” world just started in the last three years or that he never noticed it while at Time. I suppose that could explain a lot about the failures of both our “public diplomacy” efforts and traditional media.

Note that Stengel sees the propaganda war as a battle in the “marketplace of ideas.”

It’s not until 10 paragraphs later — after Stengel and Ignatius air the opinion that “social media give[s] everyone the opportunity to construct their own narrative of reality” and a whole bunch of inflamed claims about Russian propaganda — that Ignatius turns to the arbiters of that marketplace: the almost entirely US-based companies that provide the infrastructure of this “marketplace of ideas.” Even there, Ignatius doesn’t explicitly consider what it means that these are American companies.

The best hope may be the global companies that have created the social-media platforms. “They see this information war as an existential threat,” says Stengel. The tech companies have made a start: He says Twitter has removed more than 400,000 accounts, and YouTube daily deletes extremist videos.

The real challenge for global tech giants is to restore the currency of truth. Perhaps “machine learning” can identify falsehoods and expose every argument that uses them. Perhaps someday, a human-machine process will create what Stengel describes as a “global ombudsman for information.”

Watch this progression very closely: Stengel claims social media companies see this war as an existential threat. He then points to efforts — demanded by the US government under threat of legislation, though that goes unmentioned — that social media accounts remove “extremist videos,” with extremist videos generally defined as Islamic terrorist videos. Finally, Stengel puts hope on a machine learning global ombud for information to solve this problem.

Stengel’s description of the problem reflects several misunderstandings.

First, the social media companies don’t see this as an existential threat (though they may see government regulation as such). Even after Mark Zuckerberg got pressured into taking some steps to stem the fake news that had been key in this election — spread by Russia, right wing political parties, Macedonian teenagers, and US-based satirists — he sure didn’t sound like he saw any existential threat.

After the election, many people are asking whether fake news contributed to the result, and what our responsibility is to prevent fake news from spreading. These are very important questions and I care deeply about getting them right. I want to do my best to explain what we know here.

Of all the content on Facebook, more than 99% of what people see is authentic. Only a very small amount is fake news and hoaxes. The hoaxes that do exist are not limited to one partisan view, or even to politics. Overall, this makes it extremely unlikely hoaxes changed the outcome of this election in one direction or the other.

[snip]

This has been a historic election and it has been very painful for many people. Still, I think it’s important to try to understand the perspective of people on the other side. In my experience, people are good, and even if you may not feel that way today, believing in people leads to better results over the long term.

And that’s before you consider reports that Facebook delayed efforts to deal with this problem for fear of offending conservatives, or the way Zuckerberg’s posts seem to have been disappearing and reappearing like a magician’s bunny.

Stengel then turns to efforts to target two subsets of problematic content on social media: terrorism videos (definedin a way that did little or nothing to combat other kinds of hate speech) and fake news.

The problem with this whack-a-mole approach to social media toxins is that it ignores the underlying wiring, both of social media and of the people using social media. The problem seems to have more to do with how social media magnifies normal characteristics of humans and their tribalism.

[T]wo factors—the way that anger can spread over Facebook’s social networks, and how those networks can make individuals’ political identity more central to who they are—likely explain Facebook users’ inaccurate beliefs more effectively than the so-called filter bubble.

If this is true, then we have a serious challenge ahead of us. Facebook will likely be convinced to change its filtering algorithm to prioritize more accurate information. Google has already undertaken a similar endeavor. And recent reports suggest that Facebook may be taking the problem more seriously than Zuckerberg’s comments suggest.

But this does nothing to address the underlying forces that propagate and reinforce false information: emotions and the people in your social networks. Nor is it obvious that these characteristics of Facebook can or should be “corrected.” A social network devoid of emotion seems like a contradiction, and policing who individuals interact with is not something that our society should embrace.

And if that’s right — which would explain why fake or inflammatory news would be uniquely profitable for people who have no ideological stake in the outcome — then the wiring of social media needs to be changed at a far more basic level to neutralize the toxins (or social media consumers have to become far more savvy in a vacuum of training to get them to do so).

But let’s take a step back to the way Ignatius and Stengel define this. The entities struggling with social media include more than the US, with its efforts to combat Islamic terrorist and Russian propagandist content it hates. It includes authoritarian regimes that want to police content (America’s effort to combat content it hates in whack-a-mole fashion will only serve to legitimize those efforts). It also includes European countries, which hate Russian propaganda, but which also hate social media companies’ approach to filtering and data collection more generally.

European bureaucrats and activists, to just give one example, think social media’s refusal to stop hate speech is irresponsible. They see hate speech as a toxin just as much as Islamic terrorism or Russian propaganda. But the US, which is uniquely situated to pressure the US-based social media companies facilitating the spread of hate speech around the world, doesn’t much give a damn.

European bureaucrats and activists also think social media collect far too much information on its users; that information is one of the things that helps social media better serve users’ tribal instincts.

European bureaucrats also think American tech companies serve as a dangerous gateway monopolizing access to information. The dominance of Google’s ad network has been key to monetizing fake and other inflammatory news (though they started, post-election, to crack down on fake news sites advertising through Google).

The point is, if we’re going to talk about the toxins that poison the world via social media, we ought to consider the ways in which social media — enabled by characteristics of America’s regulatory regime — is structured to deliver toxins.

It may well be that the problem behind America’s failures to compete in the “marketplace of ideas” has everything to do with how America has fostered a certain kind of marketplace of ideas.

The anti-Russian crusade keeps warning that Russian propaganda might undermine our own democracy. But there’s a lot of reason to believe red-blooded American social media — the specific characteristics of the global marketplace of ideas created in Silicon Valley — is what has actually done that.

Update: In the UK, Labour’s Shadow Culture Secretary, Tom Watson, is starting a well-constructed inquiry into fake news. One question he asks is the role of Twitter and Facebook.

Update: Here’s a summary of fake news around the world, some of it quite serious, though without a systematic look at Facebook’s role in it.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

When Your Lived Reality becomes an Algorithm of the Popular

Yesterday, as people panicked about Twitter’s plan to present tweets to users using algorithmic calculations of what that person might like, I wrote this:

What if humans started to experience time as an algorithm of the popular rather than lived narratively ordered experience?

Especially since I work from home and in flyover country, Twitter is very much a lived conversation for me. And if Twitter alters the way it appears to me — basically choosing who it thinks I want to talk to rather than what the serendipity of the unique collection of people I follow presents in time-ordered fashion — it will be fairly dramatically altering my lived reality.

The thought got even creepier for me as respondents to my tweet (author William Gibson was among those who retweeted it, so I got so really awesome responses) pointed out that associations divorced from lived time is much closer to dreaming than waking reality. Of course, as a shrink friend noted, in dreaming, we consciously and unconsciously select what those associations are, rather than having a computer do it for us.

So thought of as its almost most dystopian, Twitter wants to take the serendipitous global conversation we’ve been having and instead replace it with a living dream world chosen for us algorithmically.

But let me go one step more dystopian. As I noted yesterday, Google recently told the British Parliament that it is testing ways to show “positive” ad words and YouTubes when people look for hateful, potentially terrorist speech. Google’s announcement follows an earlier one from Facebook, stating it would do the same.

In other words, since the early January meeting in Silicon Valley, two of the big tech companies announced plans to rejigger their algorithms selectively for users the algos identify as expressing an interest in terrorism. For those interested in terrorism, Google and Facebook will create a waking dreamworld.

Thus far, Twitter has made no such announcement. Yesterday (that is, the same day this algorithm report came out) it did, however, announce how many perceived terrorists it has kicked off Twitter.

Like most people around the world, we are horrified by the atrocities perpetrated by extremist groups. We condemn the use of Twitter to promote terrorism and the Twitter Rules make it clear that this type of behavior, or any violent threat, is not permitted on our service. As the nature of the terrorist threat has changed, so has our ongoing work in this area. Since the middle of 2015 alone, we’ve suspended over 125,000 accounts for threatening or promoting terrorist acts, primarily related to ISIS.

The blog post making that announcement also addressed algorithms, admitting that they can’t really work, linking to this report from December (which discussed Facebook and Google) on another presentation to the UK Parliament.

As many experts and other companies have noted, there is no “magic algorithm” for identifying terrorist content on the internet, so global online platforms are forced to make challenging judgement calls based on very limited information and guidance. In spite of these challenges, we will continue to aggressively enforce our Rules in this area, and engage with authorities and other relevant organizations to find solutions to this critical issue and promote powerful counter-speech narratives.

All of which might have left the impression that Twitter, unlike its counterparts Google and Facebook, would not be fiddling with its algorithms in response to the request to magnify voices deemed to be positive targeted at those seeking terrorism content.

Except that at almost exactly the same time, came these reports that everyone would get (or would get the option of) the algorithmic treatment.

Now, having studied how, after 1848, the powers that be in Paris found ways to eliminate the growing newspaper public in the belief that it had led to that year’s revolution (that’s partly where the idea of high literature, as embodied in Madame Bovary, arose from), I’m pretty paranoid when I see the ways the elite would current neuter the voices that contributed to the Arab Spring, Black Lives Matter, and yes, terrorism.

But I have to say, I always get buggy when communications companies — like AT&T and Microsoft — don’t have an apparently robust business model, especially when I consider ways they could (and do, in the case of AT&T) profit handsomely off working for the government. Most people assume Twitter is doing this as a way to monetize, which it has thus far failed to do. And that may well be the case.

If so, who is the customer, and what is Twitter delivering?

Update: Jack Dorsey just had this to say on Twitter.

Hello Twitter! Regarding #RIPTwitter: I want you all to know we’re always listening. We never planned to reorder timelines next week.

Twitter is live. Twitter is real-time. Twitter is about who & what you follow. And Twitter is here to stay! By becoming more Twitter-y.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.

Does the NYT Publish “All the News That’s Fit to Print” Anymore?

NYT’s ombud, Margaret Sullivan, dedicated her column today to whether the NYT should have done a story on The Intercept’s drone package a few weeks back. She concludes that given the NYT’s extensive coverage of this issue, it’s reasonable they gave the story just a mention, though suggests maybe they should give it more than that going forward.

I’m particularly interested in this subject because it says so much that is troubling about how our government functions – and yes, kills — in secret and often without adequate oversight. I’ve written about aspects of it a number of times.

Times journalists have done plenty of worthy coverage of the drone program themselves, with one national security reporter, Scott Shane, writing a significant big-picture story last April, covering some of the same ground that the Intercept is exploring now. He and Jo Becker also wrote a stunning story in 2012 detailing the existence of the president’s “kill list.” Mr. Shane is the author of a well-regarded recent book on the subject, “Objective Troy: A Terrorist, a President, and the Rise of the Drone.”

Since The Times has done so much on this subject, it is understandable that only a brief mention of The Intercept’s scoop has been made so far. Still, given the revelations in the released documents — as well as the mere existence of a major intelligence leaker who is not Edward Snowden — Times journalists might have served readers well to do more on “The Drone Papers.”  They also could consider doing so in the future.

I suspect there are two other things going on. Shane seems to still be on book leave, and to cover the Intercept stuff — which in significant part confirms his earlier reporting, most importantly that the government treats males killed in drone attacks as military aged males appropriate for targeting — might be a bit awkward. I think some of the documents — such as the ones showing that JSOC’s targeting was bad because it relied on CIA’s SIGINT, might advance questions about why we decided to build a CIA drone base in Saudi Arabia in 2011. That might be appropriate follow-up reporting from other reporters like Mark Mazzetti (I have long suspected the Saudis were fiddling with the intelligence to force our hand on a drone base, since they had been trying for years to get drones from us), but that would take further time. So, too, would be a report on what these documents say about the CIA versus DOD debate on drones.

Still, underlying the whole question is whether the NYT publishes all the news that’s fit to print anymore.

There was a time when a NYT reader could expect, by reading the NYT, to know everything the elite of this country deemed worth knowing. It promised comprehensiveness, at least for those subjects that the NYT judged important, for better and worse.

Now, I think the NYT (which still plays that agenda setting function, and will still get fed stories to place items in the news agenda) often limits itself to items it can claim a scoop on (though far too often, it borrows these scoops from outlets obscure enough they’ll get away with it). As a result, when another outlet advances the news that’s fit to print in a publicly recognized scoop, or when news comes without an exclusivity agreement, the NYT may not always report it, until such time as it can own it in the future.

I think we’ll probably be better off when the NYT no longer serves as the agenda-setter for the country, in part because there are a lot of stories (like the Iraq War then, and now like anything pertaining to Israel or Ukraine) where other outlets are far more reliable, in part because the NYT’s official perspective is often so jingoistic as to disinform its readers (as with the report that Russia might cut cables into the Middle East, which includes no acknowledgment that this is a tactic we make ample use of). But we’re in a weird place now where the NYT doesn’t claim to be comprehensive, but readers still assume it is. Which means that until something shows up in the NYT it won’t be considered common knowledge, but the NYT will sometimes delay such reports until they can “own” it in some way. That, in turn, delays the time when something can be considered “official” and therefore worthy of debate.

I do expect the NYT to do more coverage on drones that reflects these documents, because both Shane and Mazzetti have already done so much.

But I’m at least as interested by this unacknowledged question about whether the NYT aspires to “print” all the news that’s fit to print anymore.

Marcy has been blogging full time since 2007. She’s known for her live-blogging of the Scooter Libby trial, her discovery of the number of times Khalid Sheikh Mohammed was waterboarded, and generally for her weedy analysis of document dumps.

Marcy Wheeler is an independent journalist writing about national security and civil liberties. She writes as emptywheel at her eponymous blog, publishes at outlets including the Guardian, Salon, and the Progressive, and appears frequently on television and radio. She is the author of Anatomy of Deceit, a primer on the CIA leak investigation, and liveblogged the Scooter Libby trial.

Marcy has a PhD from the University of Michigan, where she researched the “feuilleton,” a short conversational newspaper form that has proven important in times of heightened censorship. Before and after her time in academics, Marcy provided documentation consulting for corporations in the auto, tech, and energy industries. She lives with her spouse and dog in Grand Rapids, MI.