"Rowback!"

"Where?"

"Right there! Didn't you see it?"

"No, are you sure?"

"I think so. It sure looked like a rowback, but I'll have to go back and check the archives to make sure."

There's a good chance you've never heard about rowbacks before, but I'm almost certain that you've witnessed at least one rowback with your very own eyes, even if you weren't quite sure what it was when you saw it. Rowbacks are usually pretty small, and they can be so quick and so slick that they zip right by without you ever noticing a thing.

Well, I spotted a really fine one just the other day. For a rowback, it was huge, yet it was so quick that I would have missed it if I hadn't been watching closely when it first popped up. Right there in front of my eyes, it morphed into something else and then blended into its surroundings so well that I never would have been able to pick it out if I hadn't caught sight of it before it morphed.

That's what a rowback does. It morphs itself into something else. And after a rowback morphs, you wonder if your eyes were deceiving you about what you saw the first time around. Unless you get a good hard look at a rowback, when you do happen across one, the experience feels like deja vu, except that instead of having a vague sense that something is exactly the same as it was in the past, you have a vague sense that something has changed, but you can't put your finger on what it was. If you don't see a rowback change right in front of your eyes, you won't have much way of knowing that it really was a rowback. So like I said, I'm pretty sure you've seen one before, even if you didn't know it was a rowback.

Rowbacks aren't that uncommon. The only thing unusual about the rowback I spotted the other day was that I reported it to the New York Times. I rarely report my rowback sightings to the media because the media conspiracy doesn't want the public to know rowbacks exist. And it’s hard to nail down solid proof because rowbacks are so slick. But I reported this one to the New York Times. After all, it was their rowback, it was big one, and I figured they would want to get it under control before it morphed again.

But I’m getting ahead of myself. First I need to explain more clearly what rowbacks are and why they matter.

A rowback is not some strange mythical beast. A rowback is a journalistic creation that a newspaper uses surreptitiously to correct a mistake in a previous story. A rowback comes into being when a newspaper prints

"a story that attempts to correct a previous story without indicating that the prior story had been in error or without taking responsibility for the error." A less charitable definition might read, "a way that a newspaper can cover its butt without admitting it was ever exposed."

When a newspaper uses a later story to "fix" a mistake in an earlier story, but fails to tell its readers about the mistake in the first story, most readers will never notice either the mistake or the "correction." If the rowback is slick enough, most readers never realize that they've just had a rowback experience.

The definition of "rowback" that I quoted above comes from a New York Times article titled "Setting the Record Straight (but Who Can Find the Record?)" by Daniel Okrent. Okrent was appointed as the Times' first ever Public Editor last autumn in response to the Jayson Blair fiasco. According to Okrent, his function as the Times Public Editor is "publicly evaluating, criticizing and otherwise commenting on the paper's integrity." (At other newspapers, the job Okrent has at the Times is titled "Omsbudsman" or "Reader Representative.") Integral to the Public Editor's function as watchdog of the paper's integrity is his role as "reader advocate." He's the one who's supposed to make sure readers' legitimate concerns, including rowback sightings, aren't ignored.

When Okrent first took on the job of Public Editor, he wrote an article introducing himself to Times readers, describing his role as reader advocate, and explaining his standards of journalistic integrity:

Journalistic misfeasance [i.e., the factual error] that results from what one might broadly consider working conditions may be explainable, but it isn't excusable. And misfeasance becomes felony when the presentation of news is corrupted by bias, willful manipulation of evidence, unacknowledged conflict of interest--or a self-protective unwillingness to admit error. That's where you and I come in. (bold emphasis added)

Under the journalistic standards Okrent espouses, a rowback--an attempt by a newspaper to "cover its butt without admitting it was ever exposed"-- is a journalistic felony because "the presentation of news is corrupted . . . by a self-protective unwillingness to admit error." The rowback "corrects" the story, but doesn't tell the reader what has been corrected or even that a correction has been made. The rowback takes the place of a proper correction, but prevents the reader from easily realizing that something he read earlier was incorrect and is now being corrected. A rowback leaves it to the reader to try to figure out on his own whether the two versions were actually contradictory and, if so, which version is the correct one. Indeed, the very process of creating a rowback can bias the reporter to "correct" an erroneous story in a way that makes his error less noticeable: If the original story reported "Z," but the true fact is "A," a writer hiding behind a rowback will be biased toward producing a story that reports something that looks like "N." In his next story, he can rowback to "G," and then a couple of stories later maybe he gets around to reporting "A." Even if a rowback morphs straight from "Z" to "A," by creating two contradictory, yet unretracted, versions of the same story, the rowback confuses the record. It corrupts the news. It's a journalistic felony.

For the benefit of reporters and editors who slept through Ethics in Journalism 101, Public Editor Okrent explains that when a fact has been incorrectly reported, there are lots of easy ways to avoid doing a rowback:

Online and in archives, connect the second version of a story to the first. In print, take care to insert the words "as reported in The Times yesterday" when the cross-reference is germane. When appropriate, the insertion of "mistakenly" or "erroneously" between "as" and "reported" wouldn't be such a bad thing either.

Okrent leaves the impression that he despises rowbacks and that he considers it is his duty as Public Editor to point out rowbacks publicly when readers bring them to his attention. "That's where you and I come in," he said.

So when I spotted a rowback in a story the Times covered a couple of weeks ago, I sent Okrent the following email:

From: [Adeimantus]
To: public@nytimes.com
Subject: Why the Times Flubbed the Allawi Story
Date: Sat, 29 May 2004

Dear Mr. Orkent:

The New York Times has completely changed its story on how Allawi was tapped as interim Iraqi prime minister.

A certain NY Times article now reads:

The decision to name Dr. Allawi was made with the approval of Lakhdar Brahimi, the United Nations envoy, though it was unclear how enthusiastic his support was. At United Nations headquarters in New York, officials contended that they were caught unawares by the announcement but said that they endorsed the choice."

But the version of the same NY Times article earlier in the day read:

The decision to name Dr. Allawi was made by Lakhdar Brahimi, the United Nations envoy, and the governing council was then summoned to be informed of the choice. The council more or less showed its approval, some officials said, with one member saying the decision was unanimous. But other people said a vote did not really take place, because the decision had already been made.

Simply put, the NY Times writers changed their story 180 degrees,

from:

"Brahimi chose PM, and the Governing Council acquiesced,"

to:

"the GC chose PM, and Brahimi acquiesced."

After pointing out this 180 degree rowback, my email to Mr. Okrent went on to point out the bias that I believed caused the Times reporter, Dexter Filkins, to get the story wrong the first time around:

This is all-too-typical of NY Times' recent reporting on the Iraq handover, where the desire to achieve the "right" spin seems to muck up the reportage. In this case the Times writers' "right" spin was to have Brahimi and the UN appear to be dominating the process of selecting the interim Iraqi PM, with the GC hinting displeasure at the way he was selected, and the whole thing being riven with chaos and dissent. That original spin didn't last long, because it was immediately completely contradicted by everybody under the sun, so the NY Times has rewritten the article.

The NY Times' new spin, (in the lede of the . . . newer article on the same topic) is:

After turning to the United Nations to shore up its failing effort to fashion a new government in Baghdad, the United States ended up Friday with a choice for prime minister certain to be seen more as an American candidate than one of the United Nations or the Iraqis themselves." (The italics are mine; the passive voice is Times' writers'.)

Using the passively voiced "seen as" to veil an expression of their own views and preferences, the NY Times writers posit a false dichotomy with the "bad" side being an "American candidate," which is set up in opposition to the "good" side, which is a "candidate of the UN or the Iraqis themselves." The grouping of that dichotomy further insinuates that (1) there is no difference worth noting between "a candidate of the UN" and "a candidate of the Iraqis themselves," and (2) a candidate of the Americans could not at the same time be "more" a candidate of the Iraqis themselves than he is a candidate of the Americans. But the relevant dichotomy is not "an American candidate" versus "a candidate of the UN or the Iraqis themselves," which your writers want to have us accept as the ground for analysis. The relevant dichotomy is "more a candidate of the Iraqis themselves" versus "less a candidate of the Iraqis themselves."

With the truly relevant dichotomy in mind, ask yourself who--Brahimi or the GC--is in a better position to choose a PM who would not only be "seen more as" a candidate of the Iraqis themselves, but would actually be more of a candidate of the Iraqis themselves (so far as that is possible under the present circumstances)? Why do the Times' writers want to insinuate that a man chosen by Brahimi (who is not an Iraqi himself and who was appointed by the UN Secy General, also a non-Iraqi) would have been "seen more as a candidate of the Iraqis themselves" than a man chosen unanimously by the all-Iraqi Governing Council, a body whose extraordinarily diverse membership includes at least one representative of every group with any possible standing within Iraq? Try to name one group with a halfway legitimate claim to participate in Iraqi government that is not ardently represented on the GC. You can't do it! Now name one group that is not represented by the UN's Mr. Brahimi. The Shias! The GC voted for Allawi unanimously, but your writers imply that the GC's unanimous choice will have less legitimacy with "the Iraqis themselves" than would a candidate chosen by Brahimi.

Even Brahimi seems willing to concede (albeit grudgingly) what the NY Times writers still won't concede: If the GC unanimously wants Allawi, then Brahimi is in no position to question that choice, because Brahimi knows that the GC is more representative of the "Iraqis themselves" than Brahimi ever could be!

So why do your writers spin it the way they do? Because they are disappointed that the UN does not appear to be so vitally relevant to the process. When the UN gets cut out of the process a little, it offends your writers' internationalism, even if it is the best existing representatives of the "Iraqis themselves" who are cutting the UN out. Given the choice between having the "Iraqis themselves" decide their own fate or having a UN muckabout decide the Iraqis' fate, your writers can't help but prefer the latter. But they still feel uncomfortable with that preference, so they want to make it appear that it is not the "Iraqis themselves" exercising their own choice, but that the Iraqi choice is really an American choice (or will be "seen more as" the American choice, which to them amounts to the same thing).

The revision of the Times' original article reflects how badly the Times writers bungled this story from the start. More remarkable is the extent to which the NY Times writers continue to spin the story to convince its readers that the handover process should not be "seen as" legitimate unless controlled by the UN. The real story (or a least a major part of the real story) is the story of how the GC, by the astoundingly bold and statesmanlike act of unifying around Allawi, gave both the US and the UN no choice but to accept that decision, thereby establishing Iraqi independence from both of those entities and expressing the Iraqi people's right and intention to govern themselves. That's a big story, maybe the biggest story yet of the Iraq saga, but instead of writing it, your guys opted for the predictable, conventional spin.

[Adeimantus]
Houston, Texas

I wrote that email about two weeks ago and had pretty much given up on receiving a response, when, lo and behold, the Times favored me with a reply, not from Public Editor Daniel Okrent, but from his right hand man,

From: public@nytimes.com
To: [Adeimantus]
Sent: Thursday, June 10, 2004
Subject: 5/29 Changing Passage in article

Dear Mr. [Adeimantus],

Thank you for your message and your patience-- I apologize for the delayed response.

Breaking news articles change during the period when they first appear on NYTimes.com, to when they appear in print the next day as a result of further reporting and more time to check up on facts and follow up on thing.

That said, I will inquire further with the foreign desk regarding your concerns.

Sincerely,
Arthur Bovino
Office of the Public Editor

Hmmm? Bovino seemed to be implying that the Times doesn't feel obliged to make public corrections of errors in "breaking news articles" when the errors "appear on NYTimes.com." In other words, rowbacks are an acceptable way to "correct" errors in a "breaking news article" the Times puts on the internet, just as long as the erroneous story never makes it into print the next day? But that didn't strike me as consistent with Okrent's description of the easy ways to avoid rowbacks in online articles: "There are ways to correct this. Online and in archives, connect the second version of a story to the first." No, the Times didn't connect (i.e., link) the second version of the Allawi story to the first. Instead, the Times removed the first story and replaced it with the second story, which was 180 degrees different than the first one. So when I clicked on the link the first time, I read about how Brahimi had selected Allawi over the Governing Council's objections. But when I clicked on the same link later that day, the story was that the Governing Council had selected Allawi and Brahimi had acquiesced in their choice.

(I should mention that I was not the only blogger to notice the Times' rowback of the Allawi story. Several other, including Josh Marshall's TalkingPointsMemo have been wondering what the heck the Times was doing with the Allawi story.)

This kind of internet rowback is actually worse than a print rowback. When a printed story gets rowbacked in the next day's paper, you can go back and read yesterday's paper. But when an internet story gets replaced by a rowbacked version, the first version disappears altogether. In this instance, although the Times deleted the erroneous story and replaced it with a rowbacked version tied to the same link, I was able to track it down on the Lexington Herald-Leader, which picked up the story off the Times wire service before the Times rowbacked it. And that brings up another reason that a paper like the Times that runs a wire service should scrupulously avoid rowbacks: The poor Lexington Herald-Leader is still running the Times original version as gospel truth because the Times never informed them [or us] that it was inaccurate.

Anyway, I figured Bovino's unsatisfactory reply would be all I would get out of my disappointing correspondence with the office of the Times Public Editor. But, lo and behold, two days ago I received another email from Bovino:

From: public@nytimes.com
To: [Adeimantus]
Sent: Monday, June 14, 2004
Subject: 5/29 Changing Passage in article 6/10 Bronner

Dear Mr. [Adeimantus],

I raised your concern with Mr. Okrent and Mr. Bronner, deputy editor of the foreign desk.

Mr. Okrent instructed me to provide you with Mr. Bronner's response below:

When ever a story is breaking, early versions tend to prove false and get updated. Think of the first dispatches out of Madrid when the trains blew up in March -- Reuters said ETA had done it. As you get it right, you move forward.

I hope this helps.

Sincerely,
Arthur Bovino
Office of the Public Editor

Wow! This was getting interesting. You have to understand that I have sent countless letters for publication in the Times Letter's section, and never once have they printed a single word I've written. (Okay, maybe my letters are too damn long!) But now little ol' me had commanded the attention of not just Daniel Okrent's right hand man Arthur Bovino, but of the Times' Public Editor himself. And not just the Public Editor, but also the Times' "deputy editor of the foreign desk." Three big shots at the Times worrying about what I think!

In a flight of truly pathetic megalomania, I imagined Bovino, Okrent, and Bronner had spent two whole weeks meeting late at the office worrying about how to respond to my brilliant critique. Flattered by the attention, I began to think that the New York Times really did want to be a shining beacon of truth for the Western world.

But then I noticed that the Times foreign editor was trying to excuse the rowback by pointing out an even more glaring error Reuters had made. Pardon me for saying so, but my momma taught me that it is craven chickensh*t to try to avoid responsibility for a screwup by pointing your finger at someone who's an even bigger screwup. Besides which, no self-respecting news organization would ever, ever, ever compare itself with Reuters.

And then I noticed that the Times foreign editor never clearly admitted that its first story was wildly mistaken. Instead he dismissively fell back on a generic statement that, "early versions tend to prove false." Starting tomorrow, the Times should print a disclaimer in bold above its nameplate.



I'm sure that would cement the Times' reputation.

And then I realized that the upshot of the foreign editor's response was an implication that rowback is an acceptable journalistic technique, not just for breaking internet stories, but for all breaking stories. The main thing is, just get something into print. It can always be "updated" later. And after it's "updated," don't bother with a correction. Just "move forward" to the next "update."

So now I've written the Public Editor's office another email about this Allawi story, this time a short one:

From: [Adeimantus]
To: public@nytimes.com
Sent: Monday, June 14, 2004
Subject: Re: 5/29 Changing Passage in article 6/10 Bronner

Dear Mr. Bovino,

Thanks for your reply relaying Mr. Bronner's explanation for The Times' erroneous reporting on the story of how Allawi was chosen as the interim Iraqi PM. Yet I'm still wondering, did the Times "move forward," or did the Times "rowback?"

. . . a classic example of the rowback. The one definition I could find for this ancient technique, from journalism educator Melvin Mencher, describes a rowback as "a story that attempts to correct a previous story without indicating that the prior story had been in error or without taking responsibility for the error." A less charitable definition might read, "a way that a newspaper can cover its butt without admitting it was ever exposed." (Daniel Okrent, New York Times, March 14, 2004)

[Adeimantus]
Houston, Texas

I'll let you know if I hear back from him.

UPDATE (June 23, 2004, 2:15 p.m.):

Okrent has honored me with a response to my last email:

From: public@nytimes.com
To: [Adeimantus]
Sent: Tuesday, June 22, 2004 11:57 AM
Subject: Changing Passage in article 6/10 Bronner

Dear Mr. [Adeimantus],

This is not a rowback -- stories get updated from edition to edition all the time, and I'm glad they do. It's the final edition that is the official one and that stays in the archives.

Yours sincerely,
Daniel Okrent
Public Editor

Impressed by the quantity of hooey Okrent packed into such a brief message, I emailed him back:

From: [Adeimantus]
To: Public
Sent: Tuesday, June 22, 2004 7:10 PM
Subject: Re: Changing Passage in article 6/10 Bronner

Dear Mr. Okrent,

Thanks for your reply explaining that the second version of the Allawi story was not a "rowback," but was merely an "update."

Let me see if I've got this straight:

1. If a factual error appears in an early edition of the Times, but does not make it into the final edition, the error is not considered official.

2. Unless an error becomes official, there is no need to make any sort of explicit public correction.

3. In such cases, a rowback is an acceptable journalistic technique for dealing with an unofficial error, and is in fact not truly a rowback, but is more properly classified as an update.

But now it occurs to me that consistency in terminology suggests that an update, when employed surreptitiously to correct an unofficial error, should more precisely be termed an unofficial rowback. On the other hand, I can see how the use of the term unofficial rowback would tend to engender even more confusion inasmuch as all rowbacks are, by their very nature, unofficial.

Alas, it's so hard to know just what is the right name to call things. Contemplating your distinction between official and unofficial, I sense the misty presence of one of those "obfuscating cloud formations that befog modern journalism."

[Adeimantus]

That poetic phrase ("obfuscating cloud formations that befog modern journalism") comes straight from a column Okrent himself wrote. How aptly that metaphor describes the substance of Okrent's email! To avoid acknowledging that the Times did a rowback of the undeniably obvious error in the first Allawi story, Okrent relies on a distinction between an official final edition and an unofficial earlier edition. But that distinction is completely irrelevant to the question of whether the factual error in the first Allwai story needed a proper (i.e., conspicuous) correction. The words from a different piece of Okrent's own writings dispel the fog:

"Because its voice is loud and far-reaching," the paper's stylebook says, "The Times recognizes an ethical responsibility to correct all its factual errors, large and small (even misspellings of names), promptly and in a prominent reserved space in the paper."

I'll let you know if Okrent honors me with a further response.