At the end of last month, Shane Fitzgerald, a twenty-two-year-old student at University College Dublin in Ireland, performed an experiment for one of his classes. The goal was “to show that journalists use Wikipedia as a primary source and to demonstrate the power the internet has over newspaper reporting.”

After Fitzgerald learned that French composer Maurice Jarre had died, he immediately went to Jarre’s Wikipedia page, inserted some fake quotations, and waited to see if they would be picked up by news organizations. His experiment worked better than he ever imagined, as evidenced by this correction from The Guardian:

An obituary of Maurice Jarre (31 March, page 36) opened with a quotation which we are now advised had been invented as a hoax, and was never said by the composer: “My life has been one long soundtrack. Music was my life, music brought me to life.” The article closed with: “Music is how I will be remembered,” said Jarre. “When I die there will be a final waltz playing in my head and that only I can hear.” These quotes appear to have originated as a deliberate insertion in the composer’s Wikipedia entry in the wake of his death on 28 March, and from there were duplicated on various internet sites.

Fitzgerald wrote in the Irish Times this week that he was shocked to see his quotes used by so many reputable news organizations. “Quality newspapers in England, India, America and as far away as Australia had my words in their reports of Jarre’s death,” he wrote.

Fitzgerald’s experiment might sound familiar to espionage buffs. It was a variation on the “barium meal,” a term used by the British intelligence service MI5 to describe a process used to expose a leak or a mole: different versions of similar information would be fed to several sources and then you’d wait to see which version leaked, or ended up in enemy hands. Track it back and, voila, you’ve got the culprit. Tom Clancy called it the “canary test” in his novels—or at least that’s what a Wired.com journalist wrote after reading it on this Wikipedia page. (See how it works?)

Rather than planting multiple versions of his fake quotes, Fitzgerald stuck them in one high profile place and waited to see who’d take the bait. He didn’t have to wait long. I call the result of this experiment the Wikiback Effect. (Feel free to come up with your own catchy name. First person to see their term end up with its own Wikipedia entry wins.)

The Wikiback Effect is when errors contained in Wikipedia entries serve to expose the laziness and fallibility of professional journalists. Basically, if you’re a journalist and you crib from Wikipedia, it will get you back.

The Wikiback Effect is equally good at exposing hypocrisy, because some journalists love to chastise Wikipedia for its inaccuracies (often while making their own errors in the process). If nothing else, Wikipedia has become a great tool for highlighting subpar reporting.

The Wikiback Effect also occurred in February when someone added a name to the already lengthy moniker listed in the Wikipedia entry of the incoming German finance minister. The result was that the extra long name made it into scores of media reports. This too was something of an experiment, though not for course credit. The hoaxster in question later wrote a commentary for a German media site:

I asked myself if anyone would notice if I simply added one more entry to the long list of names. It turns out that no one noticed, and scores of online media, newspapers and television stations used my invention without verifying it.

One interesting side note is that The Guardian’s reader’s editor noted that Wikipedia editors exercised more skepticism about Fitzgerald’s fake quotes than the professional journalists:

Wikipedia editors were more skeptical about the unsourced quote. They deleted it twice on 30 March and when Fitzgerald added it the second time it lasted only six minutes on the page. His third attempt was more successful - the quote stayed on the site for around 25 hours before it was spotted and removed again.
The rule for using Wikipedia as a source is simple: it’s okay to read an entry for background, but it’s unacceptable to cite a Wikipedia entry as fact. Follow the links to external sources and confirm any facts with multiple sources. It’s a mantra that’s repeated over and over again, yet journalists, while often condemning the inaccuracies contained within the site, still turn to Wikipedia as a quick way to churn out an article.

This latest example also exposes another flaw inside newsrooms: most papers were unaware that they had used fake quotes until Fitzgerald himself contacted them to say a correction was necessary.

Give the kid an A.

Correction of the Week

“A week-long series of World factfile booklets appeared with the Guardian from 18 April to 25 April. They contained some non-facts.


“New Zealand’s prime minister should have been listed as John Key, not Helen Clark, his predecessor (23 April, page 15). Jerusalem was referred to as Israel’s de facto capital instead of as a disputed city claimed as capital by both Israelis and Palestinians (Sources panel, page 2, daily).


“Jamaica’s “living national icons” included the late Bob Marley (21 April, page 31). Bulgaria’s highest point, Musala peak, was listed under its defunct and short-lived name, Stalin peak (18 April, page 29). Poland was partitioned in the 1700s, not the 17th century (23 April, page 29). A map of Turkey included northern Cyprus, which Turkey occupies but does not claim (25 April, page 14).


  • 1
  • 2