Who Invented G.S.?

Michelle Delio is a journalist based in New York City. That much we do know. Whether her articles are accurate or not has become unclear.

Two of Delio’s articles for the website of the Massachusetts Institute of Technology's magazine, Technology Review, were disputed by their subject, Hewlett-Packard, which claimed it had found that an anonymous source she described within the company couldn't exist. (Boston Globe ) The Technology Review has since removed those two articles from its website and replaced them with retractions; it's also removed seven other articles Delio wrote for the publication. (You can see the retractions here and here.)

Both the retracted articles concern the recent departure of H-P chief executive Carly Fiorinia. For Delio's second article, "Carly's Way," published on March 4, she refers to her sole source as G.S. According to the article, he fled Hungary in 1973 during communism, and subsequently working for H-P for 28 years. (The article is no longer online, but you can see snippets at the blogs FreeRepublic.com and Bob Embry's Time-life navigation.)

H-P told the Boston Globe that an internal investigation failed to uncover Delio's source, and so the company contacted Tech Review to air its suspicions.

Jason Pontin, editor-in-chief of the Technology Review, told Gelf that since then, he repeatedly has asked Delio to verify her source. “The story has varied,” he told Gelf. “She said that they only communicated by email but later told me that he phoned her.”

Delio disagreed with Pontin’s assessment. “It’s been consistent all the way through,” she told Gelf. Delio met G.S. eight years ago at the now-defunct Las Vegas tech show Comdex, she said. Over the next few years, he emailed her two or three times a year, using a Yahoo account. The one time she spoke to him over the phone, she said, was right before the initial Tech Review article was published last month; from her caller ID, it seemed to Delio that G.S. was calling from overseas.

Pontin listed reasons why he thinks G.S. doesn't exist: No one of his description has ever worked at H-P, the company claims. The professional societies Delio says he worked for have never heard of him. Scientific journals Delio claimed he wrote for have never heard of him. Pontin told Gelf that Delio claims she is not at fault. “While she concedes that clearly G.S. does not exist in the way described, she insists she was defrauded,” Pontin said.

“The story was specific enough,” Delio said. “It was not constructed to deceive.” Delio said that when she got back in touch with G.S. after the revelation, he refused to participate with the investigation. She said she hasn't heard from him since.

“What I must assume is that G.S. is a fiction,” said Pontin. “But is it a fiction created by a journalist or by a source?”

How to answer that question? With G.S. unreachable, it's instructive to turn to Delio's previous work. Delio is a prolific writer. Over the past five years, she has written over 700 articles for the tech site Wired News, which lists her as a correspondent. (Wired News is owned and operated separately from Wired Magazine, for which I worked until recently and still contribute articles.)

Within the last five months, Delio has written at least one Wired News story containing an apparent misquote, and at least one more with a quote of reasonable accuracy.

In “Patron Saint of the Nerds,” she quoted Rev. Michael Amesse, pastor of Our Lady of Guadalupe Chapel in New Orleans. The Reverend is quoted as saying, "I don't know why they say Expedite is the computer saint. St. Isidore is the saint of technology and the internet. Yet these people insist on praying to Expedite. Like all things that concern this saint, it is a mystery." Yet Amesse told Gelf that he never talked to Delio. “I never said that,” he said, adding, “I didn’t know that then and I don’t know it now.”

And in “Taming the Wild River", Delio quoted Michael Logue, chief of public affairs for the Army Corps of Engineers. But Logue told Gelf he recalled thinking he had been misquoted when he saw the article: "I don't remember what happened exactly. I think there was something in there attributed to me that I didn't say, but in the grand scheme of things, the article was pretty good."

Of course, these aren't smoking guns. Sources interviewed months after articles appeared might have forgotten some details. There's also some truth to the conventional wisdom that newspaper articles are accurate, except for the ones you know something about. (One of the bizarre lessons of the Jayson Blair scandal at the New York Times is that people who knew his articles were invented or inaccurate didn't bother calling the newspaper, which, as the Columbia Journalism Review noted, doesn't speak well for any journalists.)

New York lawyer Edward Hayes, whom Delio has quoted frequently, speaks highly of her work. Hayes told Gelf, "Whenever she's quoted me, she's always talked to me. In my dealings with her, she was always very thorough. ... I'm surprised" about the Technology Review retractions. Rutgers neuroscientist Wise Young, another recurring Delio source, told Gelf in an email, "To my knowledge, I have been accurately quoted by Michelle in all the articles that I know of." (This paragraph was updated to reflect Dr. Young's comments, received after publication.)

There's a more-troubling thread common to many of Delio's articles, including over 70 Wired News stories and the Tech Review pieces: She repeatedly interviews Hungarian sources in her tech articles. This isn't surprising in and of itself, as her husband, Laszlo Pataki, is Hungarian. “We go there,” she said. “We wander around and meet people.” Delio met Pataki when they both worked at Outlaw Biker magazine.

However, Gelf hasn't been able to verify any of her Hungarian sources. She has referred to Yanos Kovas, Janos Kovacs, Kovacs Janos (“I can't see me balancing all 278 pounds of my middle-aged body on a scooter”), and many others, though Google searches of their names turn up little or nothing besides their Wired News appearances. (Google searches are only a rough way of investigating sources.)

She also once quoted her future husband in Wired News.

“Taltos” was the pseudonym of one Hungarian source whom Delio referred to as a "cracker" or "hacker" and quoted several times between February 1 and June 4 in 2001. On a virus.org message board that same year, one post questioned Taltos's credentials, noting he had never seen that name before. “Taltos” the hacker can't be found online apart from Delio articles or quotes from them, according to Google.

Many other articles by Delio feature quotes and anecdotes by sources (including Americans) who do not show up on the internet outside of the Wired News articles and sites linking to those articles.

This demonstrates only that Delio's sources weren't known to other journalists, not that they're fictitious. Too many reporters call "experts" they've found in others' stories rather than seeking new voices. When a reporter ventures from the beaten path, she shouldn't automatically be doubted.

But editors who encounter such newly discovered sources often regard them with appropriate skepticism. That attitude would have helped Tech Review avoid publishing the second retracted article, which was based entirely on one anonymous source who seemed too good to be true.

Pontin, the Technology Review editor, told Gelf in an email that when his magazine has published online-only articles, they haven't undergone the same level of scrutiny and independent fact-checking as do the articles in the print magazine—calling all interviewees and verifying quotes. "I believe that if you asked most Web editors, they would say they do not fact check stories to print magazine standards—and that something like this could have happened to them," Pontin said. (Fact-checking doesn't necessarily suffice to catch reportorial invention, as New Republic learned in the Stephen Glass scandal, exposed in Forbes Digital and dramatized in the 2003 film Shattered Glass.)

However, once-bitten, Tech Review now plans to crack down more harshly on anonymous sources in online articles, Pontin said. "Going forward, all unnamed sources at Technology Review, whether online or in print, will be held to the standards we use with print fact-checking. And in general, we will only use unnamed sources when there is no other choice." It's debatable, though, whether the Tech Review had no choice but to publish an anonymous screed against a deposed CEO; one wonders if anti-anonymice activists Jack Shafer of Slate and New York Times public editor Daniel Okrent would agree.

Delio said she couldn't answer specific questions about sources for articles other than her Tech Review articles, saying that she kept lists of her sources for only one year. But asked whether she expected any more of her stories to be called into doubt, Delio said, “I can’t imagine why they would.” She didn't respond to several follow-up requests for comment, including specific questions about recent articles.

Wired News’s managing editor, Marty Cortinas, told Gelf that Wired is currently checking Delio’s past articles, but added that the website has no plans to take her articles down. “I plan on running a Michelle Delio[-authored] story tomorrow,” Cortinas added.

--Carl Bialik contributed to this article.

Email this entry to:

Your email address:

Message (optional):

Email this entry to:

Your email address:

Message (optional):