The buzzwords for social media editors at news outlets are conversation, curation and collaboration. But when using Twitter and its ilk to collect and disseminate news in real-time, another word is becoming just as important: corroboration.
During big, breaking events such as Hurricane Irene, the East Coast earthquake and uprisings in the Middle East, social media editors monitor Twitter, YouTube and Flickr. They ask people what they’re seeing and spread eyewitness accounts and images to a broader audience.
Yet they’re finding that it’s not enough simply to share accurate information. They also must try to stem the flow of inaccurate information.
They become debunking editors, real-time Snopes who cast a skeptical eye on the dramatic photo that’s making the rounds. Even if they decide that something is a hoax, simply declining to share it isn’t enough.
“I think there is a hunger out there for us to debunk misinformation when it’s out there,” said Liz Heron, social media editor for The New York Times.
Saturday night, as Irene approached New York, Reuters Social Media Editor Anthony De Rosa tweeted, “There’s an image going around of the East River cresting. It is fake.”
Lexi Mainland, one of De Rosa’s counterparts at the Times, retweeted it, adding, “Please don’t RT it.”
Later that night, the team at msnbc.com’s Breaking News alerted their followers about a different image making the rounds:
Another effort to fight misinformation occurred on Twitter, but behind the scenes. Late Saturday night, an employee of a company called Storyful posted this to the company’s protected Twitter account, accessible to the company’s subscribers:
CLIENTS: FYI … a number of fake Irene images are now circulating on the Web. We are running an image check on all images we see.
Storyful acts as a social media news agency, curating and verifying news-related information on social media and relaying them to clients such as Reuters and The New York Times.
“Fakery has caught up to real-time, in the same way that true content is real-time,” said David Clinch, editorial director at Storyful.
Misinformation among the noise
Hoaxes did not sprout up in 2006 when Twitter was invented. But social media has aided the spread and speed of bogus information.
After tornadoes ripped through Tuscaloosa, Ala., in May, the town was consumed with rumors. Many were about the number of victims – bodies discovered in a lake and a landfill, children killed in a Chuck E. Cheese, the corpses of hundreds more victims concealed by authorities.
“A lot of the rumors originated, and were perpetuated, on social media sites and our own news forum,” Tuscaloosa News reporter Stephanie Taylor told me via email. “I can’t tell you how many people tweeted outrageous ‘helpful’ pieces of information at me.”
Staff responded on social media and wrote a couple of stories about rumor-mongering. And the News did something else to directly combat the misinformation, creating a blog called “Storm Rumors” to verify or debunk what people were talking about. No, authorities were not stacking bodies at a school because they had run out of body bags. There was no evidence that a child survived the tornado by ducking into a refrigerator, only to be found alive days later.
Debunking in public
The Storm Rumors blog was unusual because the traditional approach has been to check out rumors in the newsroom, not in the newspaper. You don’t want to seem to substantiate something by publishing a story that discredits it.
But in the same way that Twitter has pushed into public some of the fundamental practices of journalism — such as source development, reporting and verification — it has pushed debunking into plain sight.
In Tuscaloosa, the misinformation was so widespread that “it would have been irresponsible not to” address it, Taylor said.
Stemming misinformation has become a key part of Andy Carvin’s reporting on the uprisings in the Middle East. A few days ago, for instance, protesters in Bahrain claimed they were being attacked by “nerve gas,” a banned chemical weapon, rather than tear gas.
Carvin, senior social media strategist at NPR, and a few other people challenged the tweets. “The purpose of this,” he told me by email, “wasn’t to get into an argument with the people making the claim — I mean, they were literally being gassed at that moment and had better things to do than argue with us — but to get the attention of lots of other people who might be likely to retweet those claims.”
It can be harder to fight misinformation when news outlets file stories based on rumors.
“Gaddafi’s son, Khamis, for example, has ‘died’ about eight or nine times over the last six months, and often you see the rumor spread on Twitter,” Carvin said:
But this isn’t a Twitter problem, per se, since numerous news outlets have filed stories about these reports, adding fuel to the Twitter rumor mill. … So every time Khamis has reportedly died, I ask people for pictures of the body. Just exhibiting that skepticism can be enough to slow down the rumor mill.
How much times does it take to knock down those rumors? Sometimes just long enough to post a couple of tweets, Carvin said:
It’s harder when there are stories that people want to believe, like a rumor several months ago that a kamikaze pilot killed Gaddafi in his compound. For many days in a row, I had to challenge this report in my tweets, because so many people hoped it was true that they kept circulating it, with no evidence to back it up.
Taylor noted a similar phenomenon in Tuscaloosa. “It’s interesting to me that our posts attempting to debunk rumors became a place for people to perpetuate even more of them,” she said.
When to call out, when to ignore
When Irene was bearing down on New York, the image that kept popping up as De Rosa monitored Twitter was a six-year-old photo showing water pouring down the steps of a subway station.
One of the challenges as a social media editor tasked with collecting and sharing information is when to debunk and when to ignore. By asking someone where he got an image, do you put that person on notice that you’re wise to their prank, or does it just encourage them?
“Sometimes I don’t even like to mention” questionable images, De Rosa said. “But sometimes I try to question the folks that are putting them out there to see if there’s some possibility that it could be something real.”
That photo of a shark swimming in the street (which appeared to be a pretty sophisticated Photoshop job, according to Storyful’s Clinch) probably wasn’t worth debunking, Heron told me.
But a photo that supposedly showed the East River cresting? Maybe so, Heron said, because so many people were passing it around “and felt like they were in imminent danger.”
In the end, Heron said, Times staff didn’t end up debunking that photo because they became consumed with other work. And while no one used @NYTLive, the Times’ new intensive, real-time news feed, to call out hoax photos, she’s interested in giving it a try.
De Rosa also said Reuters could use a branded Twitter account to call out hoaxes and misinformation, as long as it was a breaking news account similar to @NYTLive.
Fooled like anyone else
As vigilant and savvy as these journalists are, they can fall victim to misinformation like anyone else as they sift through the cacophony on Twitter. De Rosa and Storyful both tweeted information about Irene that turned out to be incorrect.
In Storyful’s case, a curator saw a tweet linking to the East River image and passed it on before putting it through the standard verification process. The error was picked up immediately, and Storyful told its clients not to use the image.
On Sunday, De Rosa retweeted a news account that New Jersey Gov. Chris Christie had said that a firefighter had died in a water rescue. De Rosa corrected himself 20 minutes later.
He told me he thought the information was reliable, considering the source. “It came out that the governor didn’t have his own facts straight.”