behind the news

To Catch A Plagiarist

There are tools to catch plagiarists in action. Why don’t news outlets use them?
February 19, 2010

As the general manager of the iThenticate plagiarism detection service, Robert Creutz has unique insight into the recent Gerald Posner plagiarism flap at The Daily Beast.

Posner’s theft was first identified by Slate’s Jack Shafer, which caused Beast editors to begin looking into Posner’s previous work. (Meanwhile, Shafer published another column that revealed additional examples of Posner plagiarism.) At the same time that was going on, someone at the Beast started paying $50 a pop to run articles through iThenticate’s pay-per-use plagiarism check. (Most of the company’s customers pay an annual licensing fee, combined with a per-scan price.)

“I don’t know what their need would be for the service going forward, but at the rate they were using it, it might be worth it for them to sign up [for a subscription],” Creutz said.

If recent experience is any indication, that’s not going to happen. iThenticate, which is the largest of the many plagiarism detection services out there, has yet to make inroads with the journalism industry. It has relationships with seventy-five publishers of scientific journals, but it seems newspaper and magazine companies aren’t interested in checking their work for plagiarism. That’s worth examining given that over the past two weeks, The Daily Beast and The New York Times discovered they had plagiarists on staff.

Creutz said his company initially thought publishers would be a natural fit for the service.

“We tried to target media outlets but they can’t seem to substantiate the expenditure, or they’re just not particularly concerned [about plagiarism],” Creutz said. “I’ve pitched the service what seems like over 100 times over to the New York Times and the last time was after Maureen Dowd copied [part of] that column. They basically told me, ‘Maureen is going to be fine – this will blow over’.”

Sign up for CJR's daily email

Times business reporter Zachery Kouwe, however, wasn’t fine. He resigned earlier this week after an internal review, which was set off by a complaint from The Wall Street Journal, found he had plagiarized from a variety of sources.

“We’ve had a lot conversation with media outlets, particularly after a major issue comes up, but the conversation is ultimately what is the cost and whatever cost I give them, they think it’s nuts,” Creutz said.

With iThenticate, a news organization would pay between $5,000 and $10,000 per year to perform a series of random checks on articles. (Checking every article would obviously be more expensive for a large newspaper.) In today’s economic climate, that’s an expense of note, though not a crippling one. iThenticate is also probably the most expensive service out there. Competitors such as Copyscape and Docoloc charge far less, though they focus more on searching the Web, whereas iThenticate combines Web searches with access to a range of content databases.

The issue of cost cuts both ways. Having a plagiarist on staff is bad for your brand, bad for business, and a morale killer. Those too are costs. Then there’s the fact that plagiarism is proscribed in every statement of principles and ethical code you can find in journalism. Plagiarism is covered in every journalism program in operation. Journalists often talk and write about how bad it is, which makes me wonder why we find it acceptable to do almost nothing to prevent it? The general attitude towards plagiarism checking seems to be, “Let’s save our money and let readers and other people catch the culprits.”

During the course of over five years of tracking incidents of plagiarism in journalism, I can’t recall an incident of plagiarism that was exposed by the news organization that employed the perpetrator. Posner was busted because an alert reader put Shafer on the case. Jayson Blair was caught because he stole from a journalist at another paper. Ditto for Kouwe.

Along with fabrication, plagiarism is perhaps the worst sin in all of journalism, and yet we leave it up to readers, competitors, and other outside parties to let us know when we have a thief or fabulist on staff.
That would be understandable if it weren’t so easy to perform a basic authorship check on written work.

Journalism has had brief flirtations with these services. For example, in 2004, the American Journalism Review examined plagiarism detection tools and reported that the Hartford Courant had signed up for iThenticate. (A contributor to the op-ed page had been busted stealing words.)

“It’s worth the cost,” Carolyn Lumsden, the paper’s commentary editor, told the magazine. “It doesn’t catch absolutely everything, but it catches enough that you’re alerted if there’s a problem.”
Lumsden told me this week that the paper no longer uses the service.

“We had a really good deal with them when we first signed up because we were the first newspaper to do so,” she said via e-mail. “And they were a great service. But then iThenticate wanted us to sign a complicated multipage legal contract. Maybe that’s what they do with universities (their primary client). But we weren’t interested in such entanglements. We haven’t used it since, and we haven’t had any problems with op-eds lately. Hope our luck continues.”

The paper caused itself some very bad luck last year when it was caught repeatedly stealing articles from competitors and slapping Courant bylines on the work. It was sued as a result. (Lumsden’s section was not involved in this incident.)

So, aside from academic and scientific journals, is anyone in publishing checking for plagiarism before an article runs? One big iThenticate client is Demand Media. Yes, the company that has been derided as a “content farm” checks every single article for plagiarism. That’s around 1 million checks per year. Demand has even integrated iThenticate into its content management system.

It seems the organizations that turn their noses up at the quality of work produced by Demand Media can’t be bothered to spend a few thousand dollars a year in order to perform random checks. (Note that a policy of random checks helps foster a more careful approach by reporters, since they never know when their stuff will get checked.)

For Demand Media, a plagiarism check is an essential part of being an online publisher hell bent on dominating search results.

“With an online content publisher like Demand Media, duplication of content is going to kill your page rank with search engines,” Creutz said. “So Google will punish Demand if they have fifty of the same article.”

They also don’t want to pay a writer for an article that has already appeared somewhere else. Perhaps their reasons for using a plagiarism detection service aren’t the same as for The New York Times, but they’re no less valid. At its core, the service is part of Demand’s quality control.

I’m not suggesting a subscription to one of these services is a foolproof strategy. It’s not. A great start, an essential tool to have, a good deterrent—yes, all of those things. Important things. Along with utilizing technology, editors should be trained to spot the telltale signs of a thief or fabulist. I’ll offer some actionable advice for editors and writers in next week’s column, so please share any relevant tips, links or advice in the comments or via e-mail at craig at

For now, though, anyone considering making plagiarism detection a part of their quality control procedure should read the 2007 research by Dr. Debora Weber-Wulff, a professor of Media and Computing at Hochschule für Technik und Wirtschaft Berlin. She tested several services and offers an assessment of their strengths and weaknesses. No doubt things have changed since then, but her research, which is also summarized here, is a good starting point.

Certainly, additional advice and tools can be brought to bear in order to root out plagiarists before they publish. But if we’re not willing to invest time, money, and attention to prevent one of the great sins of our profession—something that can tar an organization for years—then perhaps we should stop acting as if having a plagiarist on staff is such a shocking turn of events.

Correction of the Week

“Regarding a phrase in our obituary of Dick Francis (15 February, page 36), a reader writes: ‘The concept of an ‘unauthorised autobiography’ is an interesting one!’” – The Guardian

CORRECTION: The original version of this column stated the Daily Beast used iThenticate to check Gerald Posner’s articles for plagiarism. Robert Creutz says he is unaware of the specific nature of the material the Beast was checking with the service. The lead has be changed to reflect this. A quote from him also suggested he had contacted the Beast to recommend they move to the subscription option, but Creutz says he was speaking in general and did not make a recommendation. The quote has been corrected. Finally, Associated Content was cited as an iThenticate customer. Creutz says his references to Associated Content were in the context of noting that the company is similar to Demand Media. He did not mean to imply that they were a customer. We regret the errors.

Craig Silverman is currently BuzzFeed's media editor, and formerly a fellow at the Tow Center for Digital Journalism.