Fake news is everywhere. Or is it? And what do we even mean when we say fake news? How is it fake? Who believes it?

These and other questions framed the conversation about journalism for the public sphere at the inaugural NULab conference last month. The panel, which included Lisa Friedland (Post Doctoral Research Fellow in the Lazer Lab), John Wihbey (Assistant Professor, Journalism), David Smith (Assistant Professor, College of Computer and Information Science), and Ryan Cordell (Assistant Professor, English), featured a diverse set of perspectives representative of the interdisciplinary nature of the NULab.

In her talk, titled “How Fake News Spreads,” (slides) Lisa Friedland detailed research she and her colleagues in the Lazer Lab are conducting using Twitter to determine the ways and rate at which fake news—a category she acknowledged is difficult to define—go viral online. While the research is still in progress, Friedland noted that though it might seem like fake news is everywhere, the reality is that it is coming from a fairly small number of websites, shared by a small number of people, who just happen to share a lot.

John Wihbey (slides) picked up the conversation from there, offering some context regarding the media climate we are currently experiencing in his talk “Journalism, Fragmentation, and Digital Publics: The Democratic Role of News in an Age of Networks.” Wihbey described a lack of confidence in the press that, contrary to the prevailing narrative, predates the advent of cable news, but is perhaps at an apex today. Among the markers of this lack of confidence, people are increasingly skeptical of the press and they tend to trust their own news sources, according to Wihbey.

If part of the problem that leads to the spread of fake news is the proliferation of misinformation, it might be helpful to be able trace the way texts change over time and how one text might propagate into several others. This is the focus of in-progress work by David Smith and his colleagues Shaobin Xu and Ansel MacLaughlin, as discussed in Smith’s talk “Dubious Sources: Probabilistic Models of Text Cascades.” Smith, who is also a PI on the Viral Text Project, is studying information cascades in nineteenth century newspapers, among other sources.

Finally, Smith’s colleague on the Viral Text Project, Ryan Cordell, rounded out the discussion with a historical perspective on fake news in his talk, “A Pre-History of Fake News.” Cordell looks back to the nineteenth newspaper hoax as perhaps one precursor to fake news from the time, as Cordell says, before the professionalization of journalism. He further complicates the question, however, by introducing the vignette, another popular nineteenth century newspaper genre that exists in the space between fact and fiction, as shown via computational classification work conducted by the Viral Texts team. Ultimately, Cordell says, when talking about fake news we must recognize the power of emotional truths in shaping how we perceive what is real or fake.