Recently, there has been a fair amount of attention given to the fact-checking responsibilities of social media: Facebook has resisted calls to be an arbiter of fact when it comes to President Trump’s frequent lies (even in the face of protests by its own employees), but Snap has stopped promoting Trump’s Snapchat account because it “incite[s] racial violence and injustice.” Twitter made waves when it appended a label to (but otherwise left up) a Trump tweet that referenced shooting looters but took no action against another Trump tweet which baselessly suggested a 75-year-old demonstrator injured by police could be an “ANTIFA provocateur” because, it said, that particular kind of falsehood did not fall within its narrow “authenticity” rules.
As those stories show, The New York Times has been vigorously covering the debate over the duty of technology companies to prevent the spread of untruths. Gone is any examination of the Times’s own policies on accuracy. Since I joined CJR’s Public Editor project, Times readers have frequently asked me about its policy on corrections. At the root of these frustrations isn’t a belief that the Times doesn’t care about getting the facts right. It’s puzzlement over why, when presented with what outsiders think are glaring errors, the Times can be so slow to issue certain corrections.
THE MEDIA TODAY: Crowd size is not a moral metric
The Times, like any reputable news organization, regularly admits its errors. The Times’s Corrections column is a parade of punctiliousness. “He is Damian Lewis, not Damien.” “She is Alyona Getmanchuk, not Gemantchuk.” “It is the Kissimmee River, not Kissimee.” “They were crab apples, not cranberries.” The overall impression created by these daily self-rebukes is that the Times is an institution whose dedication to accuracy knows no bounds. And, as Times standards editor Philip B. Corbett (that’s one L, two T’s) tells me, that’s exactly the point. “Pretty much across the board, we say if we get something factually wrong, we should do a correction. And as you probably know from reading a lot of our corrections, a lot of times it is, I don’t want to call them small things, but very specific details,” he says.
“But I will say that we hear from readers that they sweat the small things and that if we get a small detail wrong, it risks undercutting how much readers trust the rest of the story and the rest of our work.” He adds, “The simple statement of the policy is if it’s an error of fact we should fix it and do a correction.”
What about when the error of fact is more complicated than a misspelled word? It is in those more difficult cases, often involving questions of news judgment and fairness, that the Times can be slower to grapple with its errors. Corbett acknowledges that these cases—not calling somebody for a response, for instance, or leaving readers with the wrong impression—are hard to fit into the corrections rubric. “That’s the kind of thing that we might typically handle with an editor’s note,” he says.
But unlike corrections, it can often take days or even months for an editor’s note to appear. It was an editor’s note, for example, that was used to explain why the Times came to believe two days after it printed an op-ed by Sen. Tom Cotton that the piece “fell short of our standards and should not have been published.” (I wrote about that controversy in my previous column.) It was also an editor’s note that New York Times Magazine editor Jake Silverstein used to explain, in March, the publication’s response to historians who objected to elements of Nikole Hannah-Jones’ introductory essay for the 1619 Project, published seven months earlier. An editor’s note also explained why an unfounded rumor of a young doctor dying of COVID-19 was removed from a first-person essay by an ER doctor six weeks after it ran in the magazine in April.
For outsiders, the distinctions between ‘corrections’ and ‘editor’s notes’ are not obvious, and there are plenty of signs that they are misunderstood.
Even as fact-checking came to greater prominence following the 2016 election, the Times’s handling of its own mistakes still causes frustration, including from other journalistic truth squads. Take, for instance, an error uncovered by Glenn Kessler, the head of the Washington Post’s Fact Checker unit. On October 2, Kessler published a column about a widely repeated falsehood about Joe Biden and Ukraine that had appeared on the front page of the Times on May 1: “One of his most memorable performances came on a trip to Kiev in March 2016, when he threatened to withhold $1 billion in United States loan guarantees if Ukraine’s leaders did not dismiss the country’s top prosecutor.”
This story about Biden’s trip to Ukraine, which has been pushed by Trump and other Republicans and was central to the President’s impeachment, is based on a boast Biden made in 2018. As the quote, used since in numerous online attack ads, goes: “I said, I’m telling you, you’re not getting the billion dollars. … I looked at them and said: I’m leaving in six hours. If the prosecutor is not fired, you’re not getting the money. Well, son of a bitch. He got fired.”
But, as Kessler wrote, Biden himself was not giving a factual account of the events: linking the loan guarantees to anti-corruption reforms was part of a broader, months-long diplomatic push by the US, the European Union, and the International Monetary Fund, not an on-the-fly ultimatum from Biden; when Biden did convey the collective effort’s message that the loan guarantees were contingent on removing the prosecutor during a trip to Kyiv, the prosecutor was not removed until months after Biden had returned to Washington; and that Biden visit to Kyiv was in December 2015, not March 2016.
Kessler concluded in his piece: “We’ve corrected our mistake, and we urge other news organizations to do so as well.” The day the story went up, Kessler tweeted, “The first paragraph of the May 1 @nytimes article that started the Biden-Ukraine story contained a significant factual error, later picked up by many other reporters. Countdown on how soon it’s corrected.”
No one from the Times seems to have noticed. (In fairness, neither did I, even though I was at the time working on a column about the same Times story.) Over a month after his column, on November 11, Kessler tweeted again, this time including Ken Vogel, the lead reporter on the story: “It’s been more than a month but the @nytimes and @kenvogel have still not corrected the error in the first paragraph of this May 1 article about when Biden issued the threat on the loan.”
Five days later, Vogel was profiled by the Post and quoted as saying, “Not a single fact in either story has been successfully challenged.” Kessler took issue with that and posted a tweet thread with more pointed criticism of the Times. “The NYT has refused to correct this significant mistake, so other news organizations keep repeating it,” Kessler wrote. “Surely a ‘newspaper of record’ should want to acknowledge this error, given the high stakes of the impeachment inquiry. It’s a mystery why a correction has not been made.”
Unlike Kessler’s previous tweets, his November 16 tweetstorm went viral. Seven hours later, a narrow correction was issued — changing the date in the lead from March 2016 to December 2015, and noting that the mistake had been “brought to the attention of The Times by a tweet from Glenn Kessler.”
Asked about the incident and the six-week delay between Kessler’s initial column and his viral tweetstorm, a spokesperson for the Times tells me, “We corrected the May 1 story within hours of learning about Glenn Kessler’s [November 16] tweet, publicized the correction, and acknowledged that Glenn had pointed out the error earlier. As far as we know, no one here knew that Glenn had raised the issue earlier.”
It was not the first time Kessler had to wait for a correction from the Times. In December 2015, he debunked a false narrative from US Senator Marco Rubio, who was pushing a story in his presidential campaign ads and speeches about how he was personally responsible for a piece of legislation meant to hamper Obamacare when, in fact, it had been other Republican senators who had pushed the provision through. A front-page Times story on December 9 had led with Rubio’s version of events; Kessler published his fact-check on December 23; and three weeks later, on January 14, the Times issued a correction.
Obviously it’s easier to determine we said ‘Wednesday’ when we should’ve said ‘Tuesday,’ or we said ‘ten’ when we should have said ‘fifteen.’
IF KESSLER, WHO HAS A COLUMN in a national newspaper as well as a Twitter account with more than 127,000 followers, is having trouble getting notice from the Times, what about people with smaller social-media footprints?
One such example is an October 30 story by Rukmini Callimachi, who covers terrorist groups including ISIS and Al Qaeda and hosted the Times’s popular Caliphate podcast, offered the first explanation for how American Special Forces were able to locate and kill ISIS leader Abu Bakr al-Baghdadi while he was hiding in Syria. According to Callimachi’s story, it was “because he was paying protection money” to a rival group of jihadists to hide him, based on what Callimachi reported were ISIS receipts that “showed that the group paid at least $67,000 to members of Hurras al Din, an unofficial affiliate of Al Qaeda and an enemy of the Islamic State.”
The day after the story was published, researchers who specialize in tracking jihadist groups voiced skepticism about her reporting. They pointed to intricacies of rivalries and alliances between various jihadist factions which would make such an arrangement extremely unlikely. But questions were also raised about the authenticity of the receipts. In a series of tweets expressing his doubts over Callimachi’s story, Charles Lister, the director of the Countering Terrorism & Extremism program at the Middle East Institute, noted that some predated February 2018, when Hurras al Din was founded, and added, “that’s a **massive** red flag in terms of their unreliability.” Hassan Hassan, director of the Non-State Actors Program at the Center for Global Policy, tweeted a lengthy thread noting problems with the report and concluded, “The story is not credible and doesn’t stand scrutiny.”
The day after the story appeared, Callimachi posted a thread acknowledging criticism and an error: “(Correction – thanks @Charles_Lister and @ajaltamimi) Hurras is publicly announced in Feb 2018, not 2017.” But no correction was issued by the Times, which still reads, “The book contains eight receipts dated from early 2017 to mid-2018 showing payments by ISIS to members of Hurras al Din.”
Callimachi had received the documents from Asaad Almohammad, a researcher affiliated with the Program on Extremism at George Washington University, which also maintains the ISIS Files, a collection of documents that Callimachi collected in Iraq and used as the basis of her reporting which was recognized as a Pulitzer Prize finalist in 2019. Almohammad had received the documents from his own sources, so Callimachi asked another researcher, Aymenn Jawad al-Tamimi, to assess whether they were authentic. And her story originally reported, “He agreed to review the eight receipts and concluded that they did not appear to have been forged.”
But after it was published, and after questions were raised about the documents, al-Tamimi posted on his own blog that he had not reviewed eight receipts; he said he was given one set of four documents, which he concluded were forgeries, and another set of four receipts “which superficially showed no obvious problems.” Once he learned that there were four other receipts, including some with dates that predated the founding of Hurras al Din, he wrote, “it raises a problem regarding authenticity.” A week and a half later, he updated his post to say he had since reviewed all eight receipts and “I have come to the conclusion they are not authentic.”
Several days later, on November 14, the Times published a new story by Callimachi, which summarized al-Tamimi’s conclusions but quoted Almohammad as standing by the documents under the headline “Experts Divided on Authenticity of Islamic State Receipts.” On the original October 30 story, an editor’s note was added to the top: “After this article was published, questions were raised about the authenticity of the documents upon which it was based.” On the story itself, only one correction was made: “An earlier version of this article said that Aymenn Jawad al-Tamimi, an independent Syria researcher, had reviewed eight receipts purporting to be from the Islamic State. He was provided with, and reviewed, only four receipts, not eight.” And while the new story said al-Tamimi “believes they are forgeries,” to this day, the original story still includes the (as corrected) line “He agreed to review four of the receipts and concluded that they did not appear to have been forged.”
I asked why the Times decided to run a new story rather than recast the original to reflect the criticism and skepticism it faced from ISIS experts. A spokesperson for the Times told me, “Given the complexity of the questions that arose after publication, editors decided the clearest and best way to explain would be to put an Editors’ Note at the top of the original story alerting readers that questions had been raised and directing them to the new story, which walked through the dispute in detail. We think this gave readers the fullest account possible.”
After the correction was added to the story, I asked al-Tamimi what he thought of it. “It seems to me more like the paper wanted to save face,” he said. He emphasized that he would not have concluded that the receipts were authentic if he had originally seen all eight. Asked if he thought the story’s description of his views was now correct, he said, “On the day before they published it, it was. Not subsequently.” He added, “In retrospect, I don’t think the story should have been published.”
I asked Corbett if he felt that the Times was quicker to correct small errors than bigger matters. “Obviously it’s easier to determine,” he tells me, “we said ‘Wednesday’ when we should’ve said ‘Tuesday,’ or we said ‘ten’ when we should have said ‘fifteen.’ Those kinds of errors are usually pretty easy to figure out and do a correction. If it’s something much more complicated than that, it might require more thought, but also it may be that you’re getting into areas that aren’t really about errors of fact. Sometimes we hear from readers saying, ‘You should correct this story, it’s all wrong.’ And you sort of think, ‘Well, all wrong? What, you don’t like the story? You think it’s a stupid premise? What specifically do you have in mind?’”
And he was clear on another matter: the best way to alert the Times to an error that requires a fix, whether a correction or editor’s note, is to email firstname.lastname@example.org, which Corbett and Rogene Jacquette, the Times’s correction editor, personally check. “That’s the best and quickest avenue for people to report errors and all the emails that come in through that email address are monitored.” He adds, “If a thoughtful reader raises a legitimate concern about the premise of the story, the whole approach of a story, whether a story is fair or properly framed, that’s certainly something we would take seriously.”