Corrections in the Web Age: The Case of the New York Times' Terror Error

For all the Times' dedication to its corrections page, however, the Times policy on correcting older errors is deeply inconsistent, if not outright whimsical. The paper doesn't correct hoary old errors -- except when it does.

In 2009, for example, the paper corrected a 1906 story about a secret inscription in Abraham Lincoln's watch. And in another case that the paper itself had celebrated two days before KQED's Lindh broadcast, the Times corrected an obituary from 1899 that had, among other errors, misspelled the deceased's first name. That 1899 article is only available online in the form of a scanned image, so there's no way the Times could update the original text with corrected information. Instead, the correction took the form of a column by James Barron, chronicling an elusive hunt to verify the details of Lt. M.K. Schwenk's life story.

In Schwenk's case, the Times was willing to re-report facts from over a century ago. As Barron declared, "It is never too late to set the record straight. If journalism is indeed the first rough draft of history, there is always time to revise, polish and perfect."

But is there? Speaking for the Times, standards editor Corbett says there's a limit to how much we should expect of the paper in the set-the-record-straight department.

"We have not established hard and fast rules on handling old corrections like this," Corbett says. "But as a practical matter, we need to devote our limited time and resources primarily to fixing and correcting today's or yesterday's mistakes -- and ideally, to preventing tomorrow's mistakes. It can be very difficult to devote the time and effort to chasing down or re-reporting possible errors from five, ten or fifty years ago."

That's indisputable, as Barron's tale illustrates. But it raises a dilemma for any paper, like the Times, that aims to be conscientious about the factual record. When the Times fixes "today's or yesterday's mistakes," it does update the text to reflect the correction. At some point between "yesterday" and "five, ten or fifty years ago," the paper stops doing so -- it leaves mistakes like Lewis's Lindh error in circulation.

Corbett admits that the Times has not drawn any "bright line" to set an expiration date on correctability in its archives, but prioritizes more recent errors. "In the case of the Lindh correction," he says, "we made an exception because the error had arisen currently in a public, newsworthy context."

You can sympathize with Corbett, and newsroom managers everywhere shuffling limited staff and resources, as they contemplate opening up countless stories in decades-deep archives to fine-tooth-comb reappraisal. Unlocking such a Pandora's box of incertitude is daunting. It demands bravery and vision.

But it can't be dodged. The Web, and its search tools, have made it inevitable. The dust of the old paper-clipping morgue, the wind-and-rewind crawl of the microfilm reels, and the costly metered searches of the Lexis-Nexis era are gone, and decades-old information is as easily accessible online as today's headline.

In this world, where the noise and pace of news keeps accelerating, an archive like the Times' is more widely relied upon, and more valuable, than ever before. "We do try to do as much original reporting as possible, but for doing multiple topics a day, we often rely on sources like the New York Times," says KQED's Dan Zoll. "It's unusual for that to backfire on us."

Such a resource can no longer be treated as a static repository of established fact. As its errors continue to surface, its guardians must accept the responsibility of repairing them effectively. Otherwise, they're telling us there's a statute of limitations on their commitment to truth.

* * *

Journalists tend to be confident that most of their errors are efficiently caught and corrected. As New York Times managing editor and incoming executive editor Jill Abramson recently put it, "In the online world, the chances of a serious error in The Times going unnoticed or uncorrected are pretty slim."

The Lindh story suggests otherwise. Unfortunately, the best research we have on the matter paints a different picture, too. Scott Maier, a professor at the University of Oregon, put the work of 22 newspapers under the microscope and found that "59 percent of local news and feature stories were found by news sources to have at least one error." A followup study found that only two percent of those stories were ever corrected.

You can argue these numbers down a bit by quarreling with Maier's approach, but it's hard to avoid concluding that there are far more uncorrected errors in the press than most journalists believe. And you can buttress that conclusion by asking yourself how many errors you found the last time you -- or your company, your neighborhood, your profession, anything that you're deeply knowledgeable about -- got covered by the media. While many of these errors are indeed trivial, you can never be certain today what sort of error will prove non-trivial tomorrow (as the U.S. State Department found when it failed to apprehend Christmas bombing suspect Umar Farouk Abdulmutallab because his name had been misspelled).

It's reasonable, then, to assume that the New York Times' Lindh error, far from being a rare and isolated case, represents an iceberg-tip of inaccuracy. What, if anything, can editors do to steer clear of this disaster-in-waiting?

For starters, they can demonstrate that they invite and welcome pointers to potential errors from readers. Today's stance -- "we're willing to consider fixing our mistakes if you hunt down our contact information and pester us" -- is too passive. But it can be upgraded to an active pursuit of truth via some simple interface changes on a news website. Put a button on every page of every story, current and archived, that says "Report an error" -- and then follow up on what the public tells you. Think of this as a customer service feedback loop, or a quality-control system. That's how the most efficient large businesses that produce software or cars run their operations today; surely the information we use to run our society deserves at least as much care.

At this point, newsroom veterans are surely rolling their eyes: There's no money in our budget for covering the state house, and you're talking about fixing old errors?

Certainly, it's hard to plan foundation work when the roof is about to cave in. But you can't put the work off forever, either. Public trust in the product of journalism has been sinking for decades. If there's any hope of reversing the trend, it could begin with an active commitment to finding and fixing errors like the Times' Lindh mistake, no matter how old they are -- and before they cause more damage.

Silverman's back-of-the-napkin guess is that such an expansion of the corrections mandate might consume 20 hours a week of a New York Times employee's work. It could well be more than that -- costly, yes, but neither infinite nor unmanageable. Any such effort would bear steady returns in bolstered public confidence. If that isn't priceless, then why stay in the news business at all?

Image: Alexis Madrigal.

« Previous 12 3Single Page | Next »