The truth about Truthy has become a scarce commodity.
Truthy is the name of an academic research project that analyzes and models the diffusion of information on Twitter. Launched in 2010 by researchers at Indiana University (IU), Bloomington, the project has recently also become a case study in how the Internet can accelerate the flow of misinformation.
In late August, Truthy began to draw scathing criticism from political conservatives in the media and government, who claim it is really part of an attempt by the Obama administration to monitor and stifle free speech. Truthy’s leaders steadfastly reject that claim. But along the way, the Indiana scientists have learned a hard lesson: The media are equally good at promoting new knowledge and spreading falsehoods.
The attacks are “not simply a misunderstanding of our research,” says Truthy’s lead investigator, computer scientist Filippo Menczer, but rather “a deliberate attempt to distort what we have done.”
In 2011, the National Science Foundation (NSF) awarded Menczer, the director of IU’s Center for Complex Networks and Systems Research, a 4-year, $920,000 grant to study how so-called memes—ideas, issues, and events—are spread across the Internet. In addition to exploring what the structure and diffusion of tweets say about how a society functions, the researchers have also looked at ways to tell whether messages are coming from real people or from computer programs, called bots.
Bots have become a periodic, and controversial, method used by some political advocates, companies, and others to simulate a groundswell of public interest in an issue or product and, thus, potentially shape public attitudes. (Twitter officially forbids such techniques.)
Menczer’s work, which is also supported by the military’s Defense Advanced Research Projects Agency and by the private James S. McDonnell Foundation, is part of a growing field that examines what are called complex, nonlinear feedback systems. “It has become a very hot topic of research,” he notes. The NSF program funding his work was created to foster interdisciplinary research on these systems. Menczer, for example, has a Ph.D. in cognitive and computer sciences, while other members of his team have backgrounds in the social and behavioral sciences.
The research supported by the NSF grant, which came after the group had already done some preliminary analyses of tweets, is quite broad in scope. It has ranged from assessing the degree of polarization in U.S. political discourse—to nobody’s surprise, it’s quite dramatic, he says—to exploring the dynamics of the antigovernment protests in Turkey’s Taksim Gezi Park in Istanbul in the spring of 2013.
The work has attracted considerable attention from the mainstream media. Truthy’s website includes a list of mostly favorable stories stretching back to 2010. A lengthy 2011 article in The Wall Street Journal, titled “Decoding Our Chatter,” explains that “[r]esearchers have found a bonanza of real-time data in the torrential flow of Twitter feeds” to “monitor an earthquake, track political activity [and] predict the ups and downs of the stock market.”
The value of that bonanza continues to grow. Twitter has begun awarding grants to academics seeking to drink from its daily fire hose of more than half a billion tweets, and this week it teamed up with IBM so the company can glean customer attitudes toward its products. The value of such information is obvious, said IBM’s CEO, Ginni Rometty: “Twitter provides a powerful new lens through which to look at the world—as both a platform for hundreds of millions of consumers and business professionals, and as a synthesizer of trends.”
The backlash begins
But Twitter’s huge popularity also makes anything connected with it potentially newsworthy. And in the past few months, the media worm has turned on Truthy. Menczer’s work is now the object of shrill attacks rather than praise.
The first negative piece appeared on 25 August in The Washington Free Beacon, a conservative online news website. Its headline proclaimed that the U.S. government is “Creating [a] Database to Track ‘Hate Speech’ on Twitter.” The next day, Fox News television picked up the theme. “So some bureaucrat decides whether you are being hateful or misinforming people—what could possibly go wrong?” the reporter asks sarcastically.
Both stories quote selectively from the grant’s abstract on NSF’s website to make their case that the research is another example of the Obama administration’s targeting of conservatives. The Free Beacon story, for example, says NSF “is financing the creation of a web service that will monitor ‘suspicious memes’ and what it considers ‘false and misleading ideas’ with a major focus on political activity online.” (The words in the abstract actually apply to an online platform that was created for the public to comment on anything being tweeted.)
“One of the areas that I cover is how government is spending taxpayers’ money, and I found this grant interesting,” explains Elizabeth Harrington, the Free Beacon reporter. “The whole premise of the project struck me as questionable, and I hadn’t seen any other coverage of this aspect of the research.” Last week, Harrington wrote a follow-up story about the project, headlined “Truthy Removes Part of Website That Monitored Conservative Hashtags.” (Indiana officials say the change was part of a routine archiving and updating of the site to feature the most current research findings.)
Within a few days of Harrington’s first article, several conservative bloggers and other media had jumped onto the anti-Truthy bandwagon. On 3 September, the Columbia Journalism Review took note of the allegations, describing “how shoddy information can quickly become an online narrative.” It also gave Menczer his first chance to present his side of the story.
“The headlines are saying something that is completely false and fabricated,” he told the magazine. “We are not defining hate speech. We are not tracking people. We don’t have a database.” He also emphasizes that the project’s findings are being posted and written up in regular academic journals—not funneled to any government overlord for some nefarious purpose.
Even so, on 17 October, The Washington Post ran an op-ed that sparked a new wave of attacks. In it, Ajit Pai, one of two Republicans on the five-member Federal Communications Commission (FCC), characterizes the project as an attempt “to squint for and squelch” what Americans are saying about the political process. “The federal government has no business spending your hard-earned money on a project to monitor political speech on Twitter,” Pai concludes.
An FCC staffer says Pai wrote the op-ed because he felt “this project should not have been funded” and that Pai sees it “as a threat to free speech and the First Amendment.” The aide, who asked for anonymity, says that Menczer’s rebuttal “is arguing against a straw man. I think it’s pretty clear that what we said is correct.”
Senior Republican lawmakers also regard Truthy as a threat. Representative Lamar Smith (R–TX), the chair of the science committee in the U.S. House of Representatives, and Representative Kevin McCarthy (R–CA), the body’s second-ranking Republican, have blasted the project. (McCarthy used the hashtag #KillingFreeSpeech in a blog post last week.)
Smith is already investigating several dozen NSF research grants that he believes are examples of wasteful spending. Four days after Pai’s op-ed appeared, Smith declared that “this one appears to be worse than a simple misuse of public funds.” The science committee, he said, “is investigating how this grant came to be awarded taxpayer dollars.”
‘Manipulation of the truth’
Truthy takes its name from “truthiness,” a word satirist Stephen Colbert invented for his television show, The Colbert Report. In a 2006 interview, Colbert explained that he was trying to capture a new phenomenon.
“It used to be, everyone was entitled to their own opinion, but not their own facts,” Colbert noted. “But that's not the case anymore. Facts matter not at all. Perception is everything. It's certainty [that counts].”
Colbert often used the term to skewer the policies of President George W. Bush and poke fun at his supporters for what Colbert saw as distorting reality. Menczer says critics of his research are guilty of similar distortions. “This is a complete manipulation of the truth,” he says. “It takes something and makes it look like it’s something else.”
Ironically, the NSF solicitation that led to Menczer’s grant noted that the growth of social media may aggravate social conflict because it offers a common platform to “parties who may not know or trust each other.” NSF also warned of possible “chain reactions” with unforeseen consequences.
Those “chain reactions” are exactly what Menczer and his colleagues have been experiencing since the first Free Beacon article, which appeared a few weeks after Menczer went on a year’s sabbatical. In response, the Truthy team has followed the criticism on its website and added explanatory posts about its work. Menczer and his colleague Alessandro Flammini also gave an extensive interview to the Monkey Cage, a blog by political scientists on The Washington Post’s website.
Menczer says NSF officials warned him that attempting to defend himself could backfire. “They said, ‘Be very careful. You don’t want to talk about politics. That’s not what you do.’ ”
Notwithstanding that warning, he’s decided to fight a limited war against his critics by talking to “bona fide reporters who want to know about my work.” Still, the experience has put his team on the defensive. “So if you ask me how we feel about it, we feel awful,” he says. “My work is being slandered, and our integrity is being questioned.”
Menczer hasn’t decided whether to apply for another NSF grant, saying that the team will spend the next year “delivering what we’ve already collected.” His more recent work doesn’t deal with U.S. political issues, he says, and he plans to keep it that way. If the team does submit another proposal, he says, “we will probably stay away from anything in politics.”