Online News Stories that Change Behind Your Back

Major news Web sites routinely rewrite stories after they are published, sometimes so heavily that they only bear a glancing resemblance to what was posted earlier. This CNN/Money article about the penalty phase of the Microsoft trial is a prime example. What you see at the other end of the link is quite different from the story that first appeared at that URL. Even the headline and byline have changed. But CNN/Money managing editor Allen Wastler says there is nothing wrong with this practice, even though there is no indication on the site that the article was heavily modified after it first appeared.

To see how radically this story was changed after Slashdot linked to it, check this snapshot of the original, provided by Slashdot reader John Harrold.

The second iteration was more favorable -- or at least less unfavorable -- to Microsoft than the original, but Wastler denies any Microsoft involvement in the change. "Advertisers do not interfere with our content," he says, and notes that neither he nor any other CNN/Money editors were contacted by Microsoft about this story. He does say, though, that the later version was "more balanced" than the earlier one.

In my experience, Microsoft PR people are not capable of reacting to anything as quickly as this story changed, so the chance of a conspiracy here is about zero. As for Wastler's "more balanced" comment, that is his judgement, and you are free to agree or disagree with it. (I'm sure some Slashdot readers will say he is correct, and others will say he is not. Editorial decisions never please everyone.)

"Writethroughs" are Routine in Online News

In the news business, stories that change after the originals run are called "writethroughs." This practice originated with wire services like UPI, AP, and Reuters, who might send subscribing editors a story with the headline, "Office building on fire in downtown Cleveland," followed by one or two paragraphs of copy, with progressively longer versions of the same story coming through the wire, hour by hour, as reporters on the scene gather more information.

Wastler says CNN/Money readers look at his site "like a wire service" and expect stories to change over the course of a day. As an example, during our phone conversation he pointed me to a recently posted CNN/Money story with the headline, U.S. productivity soars, and noted that this story might be updated and expanded several times, so that "by the end of the day, it might become a magazine length feature."

Online News Association President Bruce Koon says, via email, "Writethroughs are very common nowadays among news sites, from MSNBC to CBSMarketWatch to CNN. Pretty standard practice nowadays to freshen headlines and leads as new developments occur. Some sites have labels such as 'update' or 'breaking news' but it varies. For top stories, I don't see that kind of labeling." In his day job, Koon is Executive News Editor for Knight Ridder Digital, so he ought to know.

I was not aware that this practice was routine in the online news business until a few days ago. Old-style wire service writethroughs were as specific as a rigorously kept programmer's changelog, right down to paragraph and line number. Maybe I'm naive, but if I am going to trust a news source, I expect that same level of care in story updates, or at least something like News.com's corrections page, which lets readers know what changes, if any, have been made to published stories before they are archived.

What's the Difference Between an Update and a Correction?

I doubt that most news site readers know the story they are seeing at the moment they read it is not necessarily the same as the story that was published earlier at the same URL -- unless we tell them. We run the risk of getting into the habit of "getting it first" at the expense of "getting it right" if we start thinking, "Well, we can fix it later, so let's go with what we have now even if it's not confirmed as carefully as we'd really like."

This is not the same as running a story that begins by saying something like, "An unconfirmed statement by...," followed by a later story that either confirms or denies the original statement, and it is not the same as an Update notice added to the original story when it is expanded or corrected. At CNN/Money, when a story is updated it gets a fresh time/date stamp, and Wastler says that's plenty. The problem with this is that someone reading the latest version who didn't see the previous one has no way to know that an earlier -- possibly incorrect -- version ever existed.

Columbia University journalism professor Sreenath Sreenivasan (AKA Sree) says, "You really need to make it clear to your readers if your stories have been changed or updated." He makes his students do that on Columbia's Web sites, even though some of them complain that commercial news sites, where many of them hope to work after graduation, wouldn't necessarily make them take this extra step.

Sree feels strongly that if a Web site changes a news story, for whatever reason, it should put, "'last updated at' or something like that" along with the original publication time and date.

More Analysis of the CNN/Money Story Example

Andrew Nachison, of the American Press Institute's Media Center, took a close look at our original CNN/Money example and gave us this analysis:

The Microsoft trial story on CNN looks like a typical write-thru of an earlier story, with new information from afternoon events. The morning's top news, that a Microsoft witness had trouble answering some questions, got bumped lower in the story as other witnesses testified later in the day. On its face, no big deal.

However, CNN did a disservice to its audience - especially the audience paying close attention to that particular story - by failing to explain the changes. A brief note would have helped, or a link to a journal of update notes for the story, so users - like newspaper wire editors - could, in a glance, understand how the story had changed from previous versions.

Something else would have helped CNN's audience: if CNN had an obvious, standard policy for publishing update notes that the audience expected and was used to.

What's most remarkable to me is that we're well into the digital publishing era but most digital news providers have yet to develop clear standards for how to handle updates and notes about updates so users are better informed. Publishers need to do this for two reasons: first, to better serve their audiences (which should translate into credibility with the audience) and second, to promote expectations and standards that audiences can come to expect of all credible news providers.

Errors that require corrections add a whole different level of challenge to digital publishing. Today it's virtually impossible to erase a mistake once it's published online. Web browsers call up cached versions stored on hard drives, some sites intentionally archive Web sites for historical research, and Internet service providers like AOL cache popular pages to speed service to customers. So AOL customers may hit a cached version of a story that contains errors corrected in a subsequent version that has yet to be cached by the AOL servers.

If online news publishers truly have their audience's best interests in mind then they should go out of their way to alert the audience to corrections and to make it clear when an update corrects previously published errors. They need to set the record straight.

University of Florida journalism professor Mindy McAdams has also looked at our example story. She says:

Updating the story in real time without noting that it has been changed: That's okay by me, in principle. But in this case, it's really very different.

I would be inclined to believe the Money.CNN folks who told you it's no big deal -- for them. In other words, I do NOT believe it's sneaky or anything like that.

But for the rest of the world (non-journalists), this MUST be very confusing!

I asked Wastler if CNN/Money had ever thought about archiving older story versions as new ones appeared, and linking from the new versions to the older, archived ones. He said, "The name of the game is speed, getting [stories] up on the site." He talked of the sheer number of stories a site like his publishes daily, and how loading any more work on his editorial staff, like moving old story versions to an archive, "would bog things down." I pointed out that this was something a simple script could do with a single "replace story/move old story to archive" click from an editor, and his reply was, "Well, I am not as technical as you... I don't know about that."

(This was not a hostile conversation. Wastler reads Slashdot now and then and likes it, and says, "My tech guys love Slashdot." Perhaps one of you Slashdot-reading CNN tech guys could talk to Wastler and other CNN editors about automatic story versioning. Wastler said that because of syndication deals and inbound links, his main concern was keeping a stable URL for each story even if went through a series of updates. This should not be hard to arrange.)

Future Directions for Online News

In a followup email, Bruce Koon said the idea of constant story updates on the Internet should not surprise anyone. His exact words:

How is the model different from TV or radio broadcast news? As news gets reported as it's happening, facts are going to change, new developments are happening. If anything, we've been trying to get newspapers away from this notion that they print once. The Internet is about continuous updates and reporting.

Also, unlike Slashdot or other new forms of information gathering and reporting, news audiences only go to a news site a few times a day to read what the latest news is. Most seem to know that the version of the story they're reading now is different from what they read before, just as they know the top of the hour report on the radio news may be different from what they heard two hours earlier.

Based on Koon's statement, the long term question seems to be whether Internet news evolution should be based on a broadcast model, with broadcast-style immediacy as its most important goal, or whether it should be based on a print model that assumes we are writing the "first rough draft of history" so that what we say today has archival significance tomorrow.

I think the two patterns are going to coexist, and rather than "convergence" we are going to see a gradual divergence between the two as "Internet news" simply becomes "news" instead of being seen as different or separate from other media. Watching how readers (viewers?) react to this change (assuming they notice it at all) over the next decade or so is going to be interesting.

A big part of the change is going to be figuring out how to maintain audience trust when it is so easy to digitally morph stories, pictures and almost anything else into states that are far different from their original ones. As Nachison points out, despite the apparently transitory nature of online news, nothing on the Internet ever quite goes away. It is all archived or cached somewhere once it gets into digital form, whether it was originally prepared for delivery on the Internet, on printed pages or for cable or over-the-air broadcast.

Professor Sreenivasan says, "We're all in the early days of this business. We need to evolve standards."

That we do. But is the "we" who evolves standards going to be the people who read (or view) the news or is "we" going to be the people who produce it? And that leads to another question: Where will we draw the line between reporters and readers/viewers, or will we even bother to differentiate between them, when PDAs with broadband wireless connections and built-in digital video cameras become common, everyday consumer items?

Well...

[JimPooley]

And this, kiddies, is why traditional media is best. You can't go back and change yesterday's newspapers.

the only past is the past we tell you

[Vodak]

story changing constantly without making note of it... sounds hella like 1984 to me.

Re:This is not a newspaper

[thaigan]

I don't think that news that is proven incorrect should just be changed. If it's news that's being reported, I think it should stay as is, but with corrections added. Authors should at least note that the original story has been modified.

This is fairly amusing...

[gamorck]

Considering thats its being posted on a site that routinely engages in actions similar to these. I mean how many times have we caught the editors modding down hundreds of posts in single threads to -1 just because they were critical of the way things were down here at slashdot?

How many times have we seen articles mystically updated and changed here without any mention of the revision on the actual article? Now some of you may attempt to argue that slashdot isn't a real news site and isn't subject to the same standards as the likes of CNN and Foxnews - but I contend that this is not the case.

Slashdot provides "news" and information to hundreds of thousands of eager eyes on a daily basis. To deny this is to simply deny the effect that slashdot has on many members of the tech community. By default they are subject to the same standards no matter their origin. Taco can scream and whine all day about how this is "just his hobby" but as long as (a) hes getting paid for it and (b) society believes that people are responsible for their own actions - he is just as open for examination as everybody else.

Thanks for the laugh Roblimo - I guess you havent kept up with the slashdots frontpage lately huh? I mean they actually posted 6 Anti Microsoft stories in a SINGLE day on Monday. This is truly pathetic.

J

TV vs Newspaper

[jefferson]

I can understand why CNN thinks this is no big deal. CNN was (and is) primarily a TV news station. On TV news, there is no archive or changelog for writethroughs: the copy gets rewritten, and the reporter or anchor reads it on the air. The only way you notice the changes is if you happened to see a previous version of the story earlier in the day.

CNN obviously sees the web as a translation of their TV news business, rather than as a translation of a print-news wire service business, so to them it seems fine! To them the web is a transient medium, like TV, not a fixed medium like print.

Of course, at first glance this seems fine, until linking of stories factors into the equation.

Of course, there are technological solutions to this, but getting CNN to adopt them could be a challenge, because it means converting them from a TV mindset to a print mindset.

So what's the problem, again?

[Patman]

I read both versions. The first was skewed heavily towards the performance of one witness in the trial.
The second was a much more well-reasoned discussion of the case as a whole vs. one tiny piece of it.

So what's the problem? The second story seems to be better-written and easier to read, and contains more information.

It's not like they changed the facts of the story; just the scope and the level of detail.

As an aside, does anyone else find it funny that a site that claims to be "News for Nerds", yet claims they shouldn't be handle to any journalistic standards, thinks that they have the right to call other news services on minor issues like this? At least those folks are trying.

More disturbing...

[Archie Steel]

...are (admittedly controversial) articles that are posted on a major news web site, then taken off a few days later, like this one [1accesshost.com], or this other one [democrats.com]. This is a dangerous trend, and asks a sensitive question: why "remove" stories instead of putting out counter-arguments? Freedom of speech has it that you can say anything (almost: libel and slander are not acceptable), but anyone can challenge what you say by bringing their own arguments to the discussion. Too often, though, the american media silences alternative viewpoints by excluding them from the debate, so that the public doesn't even know they existe. Case in point: how come Chomsky hasn't been invited to present his views about the 9/11 events on television? If his arguments are so weak as the conservative pundits claim, why not simply try to prove him wrong on the air? Well, there's a good answer to that: they can't, and they know it. So they just ignore his existence altogether, and immediately try to discredit him (without ever challenging his arguments) whenever he is mentioned. Quite revealing...

Congratulations Roblimo and Thanks

[gamorck]

For finally showing slashdot what it takes to create a real news story. While I do find it amusing that slashdot engages in the same practices that you seem to rebel against here, I think its actually quite impressive that:

(1) Actual research was done by a slashdot employee for this article. Roblimo actually took the time to call a CNN employee and allow them to confirm/deny the allegations at hand.

(2) Roblimo doesnt appear to jump to any "off the wall" conspiracy conclusions as some editors here have been known to do. He leaves that for the comment posters to do :-)

(3) The article is very balanced all in all. I think Roblimo is attempting to present both sides of the story and give the reader a chance to make up his own mind. Now that is true journalism.

In short thank your Roblimo for helping to raise the bar here at /. I can only hope that the other editors learn from your example and attempt to follow suit.

J

Re:Oh really?

[Anonymous Coward]

Not to toot slashdot's horn, but when slashdot screws up, they tend to admit it.

When everyone's favorite really-pretty-nice-guy-who-played-an-annoying-as-H ell-character-on-TV pointed out that James Doohan was, in fact, in reasonably good health, an update was posted in the story blurb to that effect, and the original story was left up.

In other words, slashdot didn't just yank the story and hope no one noticed.

This is different; the story was clandestinely changed, presumably by a secret cabal of webninjas maintained for the purpose.

Webninjas are cool, and you won't catch me saying otherwise, but putting them to this kind of use is really sleazy. Their mighty HTML tags should be used only for good. To do otherwise would be double-plus ungood.

Re:eh?

[Dante_H]

Didn't you guys do the same thing with the James Doohan coma story?

If you look at the story that's still there :

ThreeHamsWillKillHim writes "Apparently, it's rumored that actor James Doohan, from Star Trek fame, is in a coma." The article notes that he's not likely to come out of it. James Doohan is 82 and is known best for his role as Engineer Lt. Commander Montgomery Scott on Star Trek.

Basically, the record of Slashdot's original comment is still there. They did however change the headline - which I presume was to stop thousands of people posting "No he's not." or "Oh my god!" unnecessarily. The line isn't exactly blurred in these matters : You have to keep a record that you were originally wrong, and then add an update. Changing the headline can be interpreted as dubious though, although in the case it's just confusing as the headline is contradicted by the story, and then the story is contradicted by the update. Personally I think the change of headline should be noted along with the update.

Re:What's the problem?

[Esgaroth]

The problem is after-news sites like Slashdot. The original was linked and the slashdot story [slashdot.org] mentioned the complete misunderstanding of what KDE and Gnome were by the witness. This wasn't mentioned at all in the updated article.

This is a big deal to after-news sites.

Re:1984 reference yet again

[iangoldby]

There is a huge difference though.

In 1984, Big Brother made up his own 'truth' as convenient for the moment. If was was expedient to change it, then the 'truth' changed.

In news reporting, an initial story may have inaccuracies. One hopes that with each revision, the reported story becomes closer and closer to the actual truth. It is fairly unlikely that the original story is better than the revised one.

Most consumers of news aren't interested in older and less-accurate versions of a story. It's quicker and easier to read the most-accurate-so-far version than to read the initial version and then mentally overlay all the updates.

I guess the latter approach appeals more to geek-types because we tend to be more interested in the mechanics of things. Irrelevant details matter to geeks 8-)

I fail to see what the big deal is.

[saider]

The world changes and so does the news. If I read a story that is developing, I'd rather read a single cohesive document rather than an initial report followed by a truckload of corrections and additions. I read enough changelogs in my job, I really do not want to have to deal with it when I'm checking the market. Just give me the latest stuff. If I want an update, I can go back to my bookmark and get the update.

Re:This is fairly amusing...

[bwhaley]

Yes that certainly is unacceptable. However, it was a discussion about slashdot itself, not a discussion on an article..

Re:TV vs Newspaper

[jefferson]

There absolutely *is* and archive. All broadcast material is archived on tape Who cares? It's inaccessible to the viewers. Nobody links to a specific airing of a TV story and expects it to be the same the next time they see it.

Re:Congratulations Roblimo and Thanks

[selan]

I have to admit I don't understand why so many people consider /. to be "journalism".

The majority of /. stories are links to news, features, rumors, innuendo, etc. originating elsewhere on the web. Some links are to legitimate news stories and others are less so. The "editors" merely post links that they find interesting and add their own purely subjective opinions (they've never claimed to be objective). Then we all comment and discuss amongst ourselves.

The only /. stories that are actually original journalism are the features [slashdot.org], including this one by Roblimo and, yes, JonKatz's articles. So if it's real journalism you want, read JonKatz.

Re:Excellent story, Roblimo.

[banuaba]

Where did you get that he attempted to mislead the court? He 'fumbled.' Fumble is a term used in american football to describe the act of accidentally dropping the ball, giving the other team a chance to pick it up.
It is not a deliberate act, just like the MIT Prof in question's fumble wasn't deliberate, either. Simply put, this professor was a bad witness. Smart guy, bad witness. He got flustered and stumbled over his words. This does not make him a liar.

And if you're suggesting that someone not attend MIT just because one professor likes Microsoft, you're an ignorant git and should be hit by a bus, fall on a soup spoon, get cancer and die.

Updating the news

[T1girl]

You can't go back and change yesterday's newspapers.

Yes, but you can update the articles throughout the day as later editions roll off the press. I used to work for an afternoon newspaper, covering court trials. You would have to write one version of the story, perhaps speculating on what was going to happen, for the early editions that went out to outlying counties, then file another story with the morning highlights of the trial for the editions delivered to homes within the city, then try to get something sensational splashed across the front page for the final edition that was sold on the downtown streets. The focus of the story could change throughout the day, and often another reporter would be sent in to make sure you didn't miss anything while the first reporter was outside the courtroom filing a report (no laptop usage was allowed inside the courtroom).

Re:More disturbing...

[bluebomber]

In the NYT case (read the stuff at the democrats.com link in the parent), it seems like there was a legitimate case for not publishing the original story. Here's a summary for those who don't want to follow the link: 1) NYT publishes story on bin Laden on 9/8, 2) NYT yanks the story a day or two later because it didn't make it into the print edition, 3) terrorists linked to bin Laden attack the US on 9/11, 4) NYT revises the article to account for 9/11 and run the article on 9/12 in both print and online editions. It is a policy of the NYT not to run online stories that don't get into the print edition. Nothing scary or dangerous here, just keeping the print in sync with the online editions. Note that many stories don't make it into a given edition of a print paper, they have space issues, deadlines, timeliness, etc. to contend with so not everything gets printed.

Sure, in hindsight it looks like the bin Laden story on 9/8 was EXTREMELY important. But don't forget that 9/11 hadn't happened yet! We've known about bin Laden for years, we've known that he is capable of dastardly deeds. This didn't prevent the embassy bombings, it didn't prevent the attack on the USS Cole, it didn't prevent the 9/11 hijackings, and just knowing about certain terrorists existence won't prevent future attacks. So someday something bad will happen, and you'll point back to [insert a date here] when [insert FBI memo/news story/etc here] seems extremely prescient. But in fact it wasn't because there are dozens of other dates and memos that contained similar but inaccurate warnings.

Whew. Sorry, got a little offtopic there, but recent news stories have gotten me going. I'll stop now.