Ads via The Deck

Fact-Checking

John R. Holmes for The Morning News

Here There Be Monsters

A professor teaches his students skepticism by instructing them to create hoaxes with the web as their laboratory.

On June 14, 2010, Wikipedia published its weekly edition of the Signpost. In this particular edition, the online newsletter—launched in 2005 to track developments related to Wikipedia—announced the promotion of one new administrator and the completion of a May copy-editing campaign that improved more than 700 articles.

In the same issue, the Signpost mentioned two hoaxes that penetrated the Wikipedia community. The first hoax, perpetrated on the French Wikipedia site, also fooled a French presidential candidate, who referred in a speech to the subject, a man named “Léon-Robert de L’Astran.” Know the name? You ought not. Although a three-year-old Wikipedia entry praised de L’Astran for his anti-slavery sentiments during the 18th century, the man himself never existed.

The second hoax involved a more familiar name: Gladiator star Russell Crowe. According to Signpost, internet rumors of Crowe’s death prompted a contributor to note the actor’s passing on his Wikipedia page. The “death announcement” was swiftly deleted. Are you not entertained? Presumably, the actor was not.

Between those two briefs, Signpost linked to an interview with Wikipedia co-founder Jimmy Wales about the site’s use as an educational tool. During the interview, Wales responded to questions about a class project arranged by George Mason University professor Mills Kelly.

Two years earlier, Kelly’s students had designed and launched a hoax that made its way into Wikipedia’s pages. Wales’s comments hinged largely on the importance of what he called “media competency,” and he didn’t let his own site off the hook. However, he also said efforts like those undertaken by Kelly’s students “really, really, really annoy me.”

No warnings pre-empted the Russell Crowe death rumor or the false biography of L’Astran. Mills Kelly’s case was different. Mills Kelly had warned Jimmy Wales. He’d warned USA Today, which went on to feature the hoax as fact on its pop culture blog. Kelly had warned Wikipedia, and the internet at large. And, on June 28, 2011, he warned them all again.

 

The first warning came in 2008. Kelly, a history professor at George Mason, had launched a new course called “Lying About the Past.” For two months, his students studied “the history of historical hoaxes”—photographs of the Loch Ness monster, forged Hitler diaries, Orson Welles’s War of the Worlds broadcast. Then, they created their own hoax—a deception Kelly previewed on the first day of the semester on his blog.

“We will work together as a group to create an online historical hoax that we will then turn loose on the internet to see if we can actually fool anyone,” wrote Kelly. He signed off: “You have been warned.”

Once Owens’s background was complete, the students lowered him delicately into history, like optimistically introducing a Chinese mystery snail into an Appalachian lake.

Kelly’s students decided on a project they called “The Last American Pirate.” They invented a man named Edward Owens. He was a Virginian. He lost his money and job after the Panic of 1873. Desperate, he turned to robbing boats on the Chesapeake Bay to regain his lost wealth. And with that, once Owens’s background was complete, the students lowered him delicately into history, like optimistically introducing a Chinese mystery snail into an Appalachian lake.

The students filmed interviews with fake experts. They created a student named Jane Browning and a website, LastAmericanPirate.net, where Browning posted updates as she “discovered” more about Owens. Someone emailed Owens’s story to USA Today, where Browning’s work was featured on the newspaper’s Pop Candy blog.

“We expected some plans to fail [or] fall short, as some did,” Clint May, one of the students in the 2008 class, told me via email. “But some exceeded our expectations.”

All told, more than 1,000 people visited the Last American Pirate website. Some came from USA Today. Others came from Wikipedia, where Kelly’s class had created an entry for Edward Owens.

“I am happy to announce that my Wikipedia entry made it through the night and the first major round of editing,” wrote Browning on her blog in mid-November, roughly three months into the course. She added: “I thought it would never make it through.”

The day after George Mason students completed their final exams, the class sank the Edward Owens hoax. On Dec. 18, Kelly posted a confession on his blog. One hour later, his students posted a final entry on the Last American Pirate website.

“Jane Browning does not exist,” they wrote. “There was a person named Edward Owens who lived in Virginia in the 19th century, but as far as we know, he never engaged in piracy.” The post includes a photo of the pranksters and their teacher, all wearing eye patches.

In the three years after Kelly’s students wrapped up their hoax, the Edward Owens Wikipedia entry became a chronicle of the class project. PC World magazine named it one of the “10 biggest hoaxes in Wikipedia’s first 10 years.” Jimmy Wales got really, really, really annoyed. Russell Crowe didn’t die.

During the same three years, two men claimed they had found the body of a Sasquatch in Georgia, then later admitted the “corpse” was a costume. Bill Cosby spoke with CNN anchor Kyra Phillips on Larry King Live to debunk rumors of his death after it became a trending topic on Twitter. (The comedian said it was his “fourth time being reported [dead].”) One day, for a few hours, news outlets tracked a runaway balloon thought to be carrying a six-year-old boy named Falcon. He was later found at home.

Then Kelly revived his course.

“For those who care, I will be teaching ‘Lying about the Past’ again in the spring 2012 semester,” Kelly posted on his blog on June 28, 2011. “So, as I wrote in August 2008, you have been warned.”

Kelly’s goal is to grow his students’ engagement and skepticism by asking them to make ethical choices instead of giving them an ethics lecture. “I wanted my students to have better critical analysis skills when it came to online sources,” Kelly told me in an interview. “And I wanted them to have more fun.” Each hoax lasts only as long as the semester because, as Kelly put it, “We don’t want zombie facts out there.” His approach has the support of at least a few colleagues. George Mason University President Angel Cabrera posted a link to the coverage of Kelly’s 2012 class on Twitter, and remarked, “Educational lies. Fascinating.”

Of course, one could argue that we’re better off with as few zombie facts as possible, or that Kelly’s classes play a zero-sum game—that they simply kill off the monsters they invent and leave others to mangle history. However, Kelly and his students could likely pose the same arguments to those fact-checkers, voluntary contributors, reporters, and editors who, despite their best intentions, occasionally miss a zombie fact or label a hoax “history.”

Should Kelly be praised or condemned? Those who do the latter feel Kelly and his students will undermine online communities of trust and sabotage efforts to provide accurate information about the world we share. Those who do the former, however, know that those communities are impressionable, susceptible, and fallible. They are only as factually accurate as their contributors permit them to be.

 

“I’ve been getting a lot of hate mail,” Kelly told me during a phone interview.

It was May 17, and Kelly was between convocation events at George Mason University. Since his course had wrapped up May 9—“Coming clean, alas,” he wrote on the syllabus—he had spent a considerable amount of time sifting through emails and online comments that either praised his methods or assaulted his character.

“This professor and his brood are pond scum,” a commenter wrote this spring on the Atlantic’s website, where Kelly’s most recent class was covered after the hoaxers came clean. “They are worse than parasites and deserve all the scorn society has at its disposal.”

In Kelly’s second “Lying About the Past” course, his students had launched two short-lived hoaxes. The first detailed the discovery of a 200-year-old beer recipe. It was briefly reported by a Washington, DC, radio station but otherwise received little attention or scorn.

The second hoax centered on a New York man named Joe Scafe and a modern-day distant relative, Lisa Quinn, who suspected that Scafe “was connected, involved, and had intimate knowledge of at least four murdered women between 1895 and 1897.” Neither Scafe nor Quinn exist.

Understandably, the second hoax generated a bit more attention. Using Quinn’s name, Kelly’s students submitted the story to Reddit, where readers dismissed the tale in a matter of minutes.

Members of the Reddit community “noted the timing and IP addresses of the author/editors of the entries and declared them the result of someone engaged in ‘sock puppetry,’” Kelly later wrote on his blog. “The irony here is that the students were all real people...and the entries were entirely true and based on extensive historical research.”

The hoax and its undoing caught the attention of Yoni Appelbaum, who published a story on the class in the Atlantic. That story attracted Kelly’s critics.

He wanted his students to become more skeptical readers, better fact-checkers, and astute historians. He wanted them to understand the systems by which people create and preserve fact—and, yes, the flaws in some of those systems

“My older son said today, ‘Have a great day at work, pond scum,’” said an amused Kelly when we spoke in May.

A reader posting on the Atlantic’s site speculated that websites like Wikipedia “would become useless” if all college history programs offered such a course. “Wikipedia is useless,” replied another. “It can be a good source of information but not a reliable source of information.” The 236 comments on the story largely fall into one of those two camps.

Many of Kelly’s harshest critics assail him and defend Wikipedia in the same breath. Contributions to the online encyclopedia, launched in 2001, are submitted and edited by volunteers. The site’s terms of use prohibit “engaging in false statements, impersonation, or fraud.”

After Wales expressed his frustration over the Edward Owens hoax in 2010, Kelly took to his blog and urged the Wikipedia co-founder to understand his methods and his students’ efforts. He wanted his students to become more skeptical readers, better fact-checkers, and astute historians. He wanted them to understand the systems by which people create and preserve fact—and, yes, the flaws in some of those systems. He did not set out to discount Wikipedia.

“Perhaps...it is time to simply accept crowdsourced information as a category of information with its own attributes and move on,” wrote Kelly.

A commenter called “Jimmy Wales” responded.

“Imagine if we heard about a community where, interestingly enough, the citizens all chip in to keep the streets clean, and [we] decided to test the system by going in the middle of the night to dump trash around, to see what they would do,” read the response. The commenter added that the act would be considered juvenile, “yet, an experimenter doing the same thing to Wikipedia manages to hide behind pseudo-intellectual pretense.”

Kelly’s students didn’t dump trash on Wales’s lawn this year. Rather, they wrote three factually accurate entries to the website. Two pages—for Alice Walsh and Florence “Diamond Flossie” Reilly, each described by a moderator as a “non-notable prostitute”—were deleted according to the site’s “notability” guidelines, which prohibit (as any good editor should) information printed without third-party sources. The remaining entry, for Baltimore-based Brown’s Brewery, survived an “Articles for Deletion” hearing by Wikipedia editors. Other contributors have since added to the brewery article.

Despite his students’ contributions, some of Kelly’s critics at Wikipedia decided they’d had enough. After the Atlantic story, a group of volunteer site administrators debated in an online forum whether to ban Kelly’s students individually or block George Mason University entirely.

“Can we preemptively block these folks?” one administrator asked. “A controversial all-institution block on the school?”

Another dissented: “We’d be better off to monitor any names created as part of the experiment, revert vandalism as it appears, and block users who do it repeatedly.”

Wikipedia exhaustively documents such arbitration hearings. Its terms of use are clear, even if the costs of transgression are less than crystalline.

The spirit of crowdsourced information, however, can trump its quality. On Wikipedia’s “About” page, there is a link to the “five pillars,” described as “the fundamental principles by which Wikipedia operates.” They include notes on balanced perspective, civil commentary, and fair use. A fourth states that Wikipedia “is an encyclopedia,” then moves on to an exhaustive list of what Wikipedia is not:

“Wikipedia is not a soapbox, an advertising platform, a vanity press, an experiment in anarchy or democracy, an indiscriminate collection of information, or a web directory. It is not a dictionary, a newspaper, or a collection of source documents.”

The fifth pillar is more direct and vanquishes its predecessors: “Wikipedia does not have firm rules.”

Being a historian, says Kelly, requires that he “assess[es] how much information is real or fanciful, or just making a better story out of it.” It’s the same charge for reporters and editors, judges and jurors. But when we confine these expectations to certain jobs, we cut ourselves a sliver of slack. Kelly deprives his students, and many others, of that comfort, so they might earn it back. For now, all they get is a warning.

 

Before they created Edward Owens and Joe Scafe, Kelly’s classes studied a curated history of hoaxes. The reading is “especially lighthearted fare,” says May, Kelly’s former student, “[compared to] the historian’s usual diet of war, famine, pestilence, death, and revolution.”

War, death, and revolution have their places, however. In addition to texts like The Book of General Ignorance and A Treasury of Deception, students read a 2008 GQ feature by John Jeremiah Sullivan. The story, “Violence of the Lambs,” is a seven-page collection of animal-on-human attack stories, bound together by expert testimony from an academic named Marcus Livengood, who claims that animals are attacking people with greater frequency.

In the final paragraphs of Sullivan’s story, however, the author reveals that Livengood is not a real person.

“Big parts of this piece I made up,” writes Sullivan. “I didn’t want to say that, but the editors are making me, because of certain scandals in the past with made-up stories, and because they want to distance themselves from me.”

After GQ published Sullivan’s story, complete with a photo of a lamb with blood-stained lips, editor Jim Nelson told Washington Post reporter Peter Carlson that Sullivan had the magazine’s permission to create his source.

“You have to stay with it,” Nelson told Carlson in an interview. “We were betting that the readers would stay with it until the end.” Carlson writes that Sullivan had his editor’s permission to create his source, to stitch together factual reports with fictitious thread. “It’s baloney,” explained Carlson. “It’s a fake, a fraud, a hoax, a prank.”

Sullivan’s prank isn’t without its predecessors. In November 1874, the New York Herald reported that a horde of animals—lions, rhinoceros, tigers—had escaped the Central Park Zoo. That hoax gets its own chapter in The Martians Have Landed: A History of Media-Driven Panics and Hoaxes.

“In their shock and disgust, many people failed to read the entire story,” write Robert Bartholomew and Benjamin Radford. “If they had, they would have discovered that the story was a hoax intended to jolt New Yorkers out of their apathy to do something about the poor state of the Central Park Zoo before a real disaster occurred, or so readers were told in the last paragraph.”

The Herald was assembled by reporters, reviewed by editors, and distributed to its audience. For a website like Wikipedia, any member of the audience may be both reporter and editor. They will be watched, but imperfectly.

And, if the Central Park Zoo hoax is any indication, credibility and readership don’t suffer in equal measure. According to Alex Boese’s Museum of Hoaxes website, “the Herald did not report any drop in circulation as a result of the hoax.”

Bartholomew and Radford discuss the merits of the deception: “Is a newspaper ever justified in creating hoaxes?” They reason that “perhaps the greatest risk is to the newspaper itself, in losing the trust of its readers.”

Now, however, the information ecosystem is different. The Herald was assembled by reporters, reviewed by editors, and distributed to its audience. For a website like Wikipedia, any member of the audience may be both reporter and editor. They will be watched, but imperfectly.

The nature of such sites raises questions of “transparency and credibility,” said Andie Tucher, a historian and journalist, during a recent interview. Tucher, who directs Columbia University’s doctoral program in communications is executive secretary of the Society of American Historians, said she uses Wikipedia as a quick reference, but would no sooner consult it as a solitary source than she would the Encyclopedia Britannica.

“I’m not dismissing that crowdsourcing can be full of perfectly smart people,” said Tucher. “But we just don’t know it, and we can’t judge it.”

Warning or no warning, Tucher was also critical of Kelly’s methods.

“I don’t think that it helps the already tattered credibility of the internet to carry out these kinds of hoaxes,” said Tucher, who mentioned instances of online hotel reviews written by employees posing as customers and the debunked “Gay Girl in Damascus” blog, actually written by a middle-aged American man. Efforts like those undertaken by Kelly’s classes “may teach 15 students something, but they do nothing for the comfort or enlightenment of the millions of people who already struggle to make decisions about what’s true or untrue on an Internet that can be a sort of Wild West.”

One wonders how Clifford Irving might have fared as an online publisher. On Dec. 7, 1971, McGraw-Hill announced plans to publish an autobiography of reclusive billionaire Howard Hughes. The book was co-authored by Irving, a novelist whose previous works include Fake! The Story of Elmyr de Hory, the Greatest Art Forger of Our Time.

At the time of the publisher’s announcement, Hughes had reportedly not been seen in more than a decade. To gain their confidence, Irving presented his publishers with letters he had purportedly received from Hughes. When it came time for the billionaire to sign his publishing contract, Irving said he would transport the document to Hughes for his signature. Irving’s wife opened a bank account under the name “Helga R. Hughes.” Checks made out to “H. R. Hughes” were deposited in her account.

Irving’s hoax hit the news before his autobiography hit bookstores. On Jan. 7, 1972, Howard Hughes coordinated a telephone conference with reporters and denied he had met Irving. The next week, Irving appeared on 60 Minutes to defend his book.

“Is the autobiography genuine?” asked host Mike Wallace. “I can only say that it is laced with detail that one would think that only Howard Hughes could know. Did Clifford Irving meet with Hughes and tape 30 hours of conversation with him? Or did he come upon the material, the transcripts from Hughes memoirs, from some other source? I just don’t know.”

Not long after, Irving confessed his deceit. He pled guilty to conspiracy and grand larceny charges and returned his publisher’s money. In 2000, Wallace arranged another interview with Irving.

“You wondered how I could lie so fluently to you,” said Irving to Wallace. “That’s because at some level, I believed everything I was telling you.”

In his 1973 film F for Fake, Orson Welles interviews Irving. Giddily, the filmmaker lists the layers of his subject’s deception: “The author of Fake, a book about a faker, was himself a faker, and the author of a fake to end all fakes that he must have been cooking up when we were filming him.”

There are additional, unmentioned layers of cunning. Welles, of course, found fame with his 1938 War of the Worlds broadcast, which attracted an audience of six million listeners. (In their book, Bartholomew and Radford write that more than one million believed the alien invasion was real.) At the beginning of the film, Welles performs sleight-of-hand tricks for an onlooker. By its conclusion, F for Fake proves itself to be as trustworthy as a magician’s sleeve.

Orson Welles knew how to exploit a medium, and Clifford Irving knew how to navigate an industry. In these respects, Kelly and his students are no different. And while Kelly’s methods are meant to teach inside the classroom, his lessons extend beyond it. As Welles remarks to his audience, “A faker...makes fools of the experts. So, who’s the expert? Who’s the faker?” As we tirelessly revise and re-read our shared history, the answer might be that we’re both.

 

During his 2010 commencement address to Williams College, author Jay McInerney recounted his 10-month tenure as a New Yorker fact-checker, which ended with his termination.

“In my defense, I tried citing the Talking Heads,” said McInerney. The lyric—a line from the Heads’ “Crosseyed & Painless”—ultimately found a place in his breakthrough novel Bright Lights, Big City: “Facts all come with points of view / Facts don’t do what I want them to.”

Like hoaxes, facts are created by their proponents, but sustained by their recipients. Properly designed and executed, the two should be indistinguishable. There are no “successful” hoaxes; there are simply untested facts: a comedian’s death, a costume in a cooler, a boy on a perilous balloon ride. And untested facts flatter neither their creators nor their believers.

There are no “successful” hoaxes; there are simply untested facts: a comedian’s death, a costume in a cooler, a boy on a perilous balloon ride. And untested facts flatter neither their creators nor their believers.

The Lifespan of a Fact, a novella-length correspondence between essayist John D’Agata and fact-checker Jim Fingal, offers a take on how we test those stories we receive. In Fingal’s introductory email to D’Agata, he writes, “I’m new at this, so bear with me.” We’re all new to this. Our stories change as we find new ways to tell them. As David Byrne put it, “facts continue to change their shape.”

As of Sept. 1, the English Wikipedia has more than four million articles and 28 million pages. More than 17 million registered users have combined for more than 550 million edits on those entries. The site is patrolled by 1,465 administrators. If administrators monitored an equal share of contributors, then each would struggle to keep pace with more than 11,000 people.

Of course, neither the Wikipedia community nor Mills Kelly’s classroom introduced error to history. Both communities can function as Reddit did. They can debunk and demystify. They can set the record straight and, with careful attention, preserve it, for their own benefit and for the education of others. They can replace cynicism with skepticism. But such work requires vigilance—and, in the case of these communities and others, collaboration.

Kelly and his colleagues also teach a course called “The Digital Past,” which one might view as an antidote to “Lying About the Past.” In his 2009 syllabus, Kelly writes that students will learn “basic principles of credibility testing for online sources, and will include opportunities to engage in the process of online encyclopedia contribution and editing.” In short, he encourages them to become the sort of fact-checkers from whom Wikipedia and countless other websites might benefit.

Have they? To Kelly’s knowledge, none of his students have gone on to careers in fact-checking—although, he notes, “one is a high school history teacher.” This year, George Mason University offers two sections of “The Digital Past.” One assignment asks students to “judge a Wikipedia article on a historical topic by looking at its sources, discussion, and history.”

During our interview, I asked Kelly whether he planned to teach “Lying About the Past” again. He said, “Oh yeah.”

So consider this a warning to pre-empt Kelly’s own. Here there be monsters. Today’s facts are only as good as tomorrow’s headlines. And without skeptical readers, the best hoaxes—the magician’s greatest tricks—will never be exposed.