Just the facts, please
Sen. Elizabeth Warren (D-Mass.) speaks in the first of two Democratic presidential primary debates in July. Sen. Bernie Sanders (I-Vt.), left, and former Texas Rep. Beto O'Rourke listen. (AP Photo/Paul Sancya)
The Democratic presidential hopefuls will take part in another
debate tonight. Most of us will watch and then declare a winner based on who
had the most clever one-liners or snarkiest attacks.
But while we’re looking at style, fact-checkers will be looking
at substance. They will sift through the promises and attacks, the claims and
defenses. And the fact-checkers will be there to not only find out whose pants
are on fire, but to help us make us sense of what we’re watching.
I asked PolitiFact editor Angie Holan what she and PolitiFact,
which is owned by Poynter, look for during a debate.
“We definitely look for moments when the candidates challenge
each other on the facts or tell each other they’re wrong,” Holan said. “We also
look for attack lines with factual claims that either aren’t widely known or
easily recalled. Finally, we’re always on the lookout for comments that would
make a regular person say, ‘Hmm, I wonder if that’s true?’”
Tonight’s debate will be the first time all the major candidates
will share the stage together. That’s because only 10 have qualified for the
debate. And there’s a showdown that everyone is waiting for.
“For this debate, we’re watching to see how Joe Biden and
Elizabeth Warren interact,” Holan said. “They’re two candidates with
significant policy differences on health care, financial regulations and their
own campaign finance practices.”
New terms for the misinformation trade
The
language surrounding misinformation seems to change as fast as the tactics used
by the people who spread it. Terms that once meant one thing — “fake
news,” for example — now mean something else, or are used so differently by
different people that they have lost a common meaning.
For
people like us, who write about misinformation as a profession, it’s a little
hard to keep up.
(Dictionary.com,
by the way, defines misinformation — its
2018 word of the year — as “false information that is spread, regardless of
whether there is intent to mislead.” The word often works for us in this
newsletter as a catchall because we can’t always be certain something is
intended to mislead. If we are sure it’s “disinformation,” though, we will call
it that.)
Another
example of an ambiguous, evolving term is “troll” or “trolling.” In her report
last year for Data & Society, “The
Oxygen of Amplification,” Whitney Phillips, an assistant professor of
communications and rhetoric at Syracuse University, noted that “trolling” now
encompasses a broad range of online behaviors, “rendering the term so slippery
it has become almost meaningless.”
Fortunately,
several experts in this space are trying to keep up with the changes and sort
out the terminology.
Claire
Wardle, the U.S. director of First Draft, a nonprofit focused on ways to
address misinformation, last year published a
helpful glossary of frequently used — and commonly misunderstood — words
and phrases. An earlier Data & Society report, “Lexicon
of Lies,” laid out some of the basics.
Now
there are some new additions to add to the nomenclature, from a report (also
from Data & Society) called “Source
Hacking: Media Manipulation in Practice” by misinformation experts Joan
Donovan and Brian Friedberg.
Donovan
and Friedberg identify and give terms to ways that deceptive actors try to get
journalists and influential public figures to pick up falsehoods. Donovan has talked
about source hacking before, but this new report helps people in the
misinformation space zero in on specific techniques used by the
manipulators.
She
and Friedberg came up with four terms, and included case studies for each to
show how they have been employed in real life.
- Viral sloganeering, where short slogans are repackaged for social media and press amplification. One of the case studies in this example was the viral slogan “jobs not mobs.”
- Leak forgery, in which forged documents are staged as “leaks” in an effort to win media attention.
- Evidence collaging, where image files featuring a series of screenshots and text are arranged in a way that make them shareable.
- Keyword squatting, where social media accounts or specific terms are used to capture and direct search traffic. They give as an example the proliferation of fake Antifa accounts in 2017.
Will
these terms be added to the vocabulary of those who study misinformation?
Maybe. Phrases catch on, or don’t, for lots of reasons. But we like the effort
to parse for precision. The information system is so disordered right now that
the least we can do is agree on the tactics used to spread falsehoods, and what
to call them.
. . . technology
- Facebook is making its own deepfakes — highly realistic fake videos featuring actors doing and saying routine things. Why? The clips, reported the MIT Technology Review, will serve as a data set “for testing and benchmarking deepfake detection tools.”
- In Argentina, Chequeado used a forensic tool developed by The Laboratory of Sensory Research to fact-check a WhatsApp audio file. Forensia can work — though not for free — in Saxon and Romance languages.
- Speaking of WhatsApp, a report from New York University found that it, along with Instagram, may pose bigger challenges for 2020 election misinformation in the U.S. than Facebook or Twitter.
. . . politics
- Democrats are divided over what policies to pursue to prevent disinformation in the 2020 campaign, Politico reported. The disagreement is over the Democratic National Committee’s reluctance to bring to a vote a pledge to rule out illicit tactics such as deepfakes. Among the party’s presidential candidates, only former Vice President Joe Biden has signed it so far.
- Democrat Beto O’Rourke’s campaign has called on the big social media platforms to better police disinformation after a Twitter user falsely said that the gunman responsible for the Aug. 31 mass shooting in Odessa, Texas, had a sticker supporting O’Rourke’s presidential bid on his truck.
- A new University of Calgary study says the Russians are likely to interfere in the country’s campaigns to serve an interest by the Kremlin in competing against Canada in the Arctic. Here’s a report from the Canadian Press.
. . . the future of news
- The German Marshall Fund analyzed 13 startups that aim to use artificial intelligence to fight misinformation. Among its topline findings: natural language processing alone can’t identify all forms of fakery, and such technology would likely hit several hurdles before ever being implemented.
- Reuters, the BBC and other news outlets have partnered with tech companies like Facebook and Twitter to fight online misinformation. The partnership’s early plans include an early warning system for potential disinformation attempts and a joint civic education effort.
- A U.S. Department of Defense contractor tasked with creating manipulated images and videos to learn more about how to combat them wrote a piece for Nieman Lab. In it, she talked about how it doesn’t take a very convincing fake to dupe someone — it just needs to confirm their preconceived beliefs.
This
could actually be the fact check of the month.
Between
Aug. 5 and Sept. 5, the Indian fact-checking platform Boom Live spotted 49
misleading messages about the Kashmir crisis on Facebook, Twitter and WhatsApp
(more than one per day). Fifteen
of those messages were posted by Pakistan-based accounts — not by
ordinary social media users, but by prominent journalists, ministers
and also leading political
parties.
Indians
are currently divided into two narratives related to the Kashmir crisis.
According to the first one, everything is fine and Kashmiris have welcomed the
revocation of Article 370, which used to maintain the region's autonomy.
According to the second narrative, Kashmiris are being subjected to brutal
atrocities. Pakistanis usually fall for the latter – but have been using old
photos and videos to argue it.
What
we liked: It is
wonderful to see that fact-checkers find the time to publish single fact
checks, build monthly databases and even dig up the origins of misleading data.
Boom’s high volume of fact checks is an impressive counterweight to
misinformation about Kashmir, and its exposure of misleading claims from
mainstream actors should be applauded.
- Rolling Stone had a good catch this week. It reported that on Sunday, more than 100 people displaced from the Bahamas were instructed to disembark a rescue ferry bound for Florida because they didn’t possess visas. Bahamians, however, are not required to have visas when traveling to the U.S.
- In India, FactChecker.in keeps a national database about attacks linked to child-lifting rumours. Since 2012, 45 people have been killed and at least 150 injured in 98 cases — 20 of them happened this year.
- Aos Fatos, in Brazil, counted 300 false and/or misleading claims made by President Jair Bolsonaro in 294 days.
- Malaysian police have filed a police report against several social media users who have been accused of spreading misinformation about the government granting citizenship to Chinese nationals.
- Violence-prone regions are the places where Facebook has the hardest time finding fact-checkers to work with to debunk disinformation, company executives told Yahoo News in an interview.
- People in poverty are more susceptible to disinformation, the Canadian Broadcasting Corporation reported, sparking efforts to help them get quality information ahead of Canada’s upcoming election.
- Politico wrote about how Democrats are pushing back against American fact-checking.
- A new poll from Gallup found that 92% of U.S. adults want reporters to use social media to correct false or misleading statements made by politicians.
- The Washington Post nominated Glenn Kessler and his team at The Fact Checker for a Pulitzer Prize, but they weren’t chosen as a finalist.
- In partnership with The Maharat Foundation, the IFCN will offer a free fact-checking workshop in Beirut on Oct. 19. Read about it here.
