Dr. Dobb's Journal June 2000
I f only there were some sort of reality bit, analogous to a parity bit, that could be attached to every byte of every news story published online. Then you could sum the reality bits to get a check on the story's accord with the truth.
Another useful innovation would be a semantic bit that would show internal inconsistencies in news stories. If the semantic parity bits add up, then maybe the story does, too.
There are plenty of errors in print news publications, too, but there are at least some error-correcting mechanisms in place in print. There really do exist such things as journalistic standards and practices, and it is the job of a publication to report on its own errors. Maybe the errors get printed in inch-high letters on the front page of a newspaper and the correction appears in small type on page 23, but that's just because journalists are as modest as the next person when it comes to admitting error. Still, when the publication is reputable and the error is noticed, corrections do get published.
Online journalism is another matter. Online journalism, even when practiced by print media operations, often falls below the standards of print journalism.
One temptation that online news sources seem to be falling into is a tendency to correct or remove the story rather than publish a correction.
On the face of it, this seems like a good idea. In print, all you can do is publish a correction, but in electronic publishing you can go out there and change the story. Why clutter up the Web with the erroneous version and a correction when you can just fix the error, leaving nothing behind but the truth?
The problem with this is that some people will already have read the erroneous story. The news source has a responsibility to let those people know that the information they got was in error. Just changing the story also lets the news source off the hook too easily: It should admit its errors publicly, so that people can judge its accuracy. If you're not willing to have your accuracy examined, you don't deserve public trust.
I should cite some examples of these corrected or removed stories, but by their very nature, they're hard to document. I do have reports from The Register and Ars Technica about two CNet stories, one on the unauthorized release of Windows 2001 code online and the other on the Napster/Wapster video-swapping hack, that were allegedly changed or removed. But I have not confirmed either report. Even if the reports are true, please note, this doesn't mean that CNet has done anything unethical or violated any existing online publishing standards.
The claim regarding the Napster piece was that it linked to the MPAA, the RIAA, and the RIAA's antiNapster page, and also linked to the Napster page when first posted, but that the Napster link was later removed, leaving links to one side of the issue only.
Some would argue that there is only one side to the issue, but this nevertheless raises the linking-fairness issue: In the interest of not linking to an illegal site, news sources may introduce a bias in their links. There is absolutely no established ethical standard on this, but I wonder if the same standards that apply to content shouldn't apply to links. That is, news sources should be as unbiased in deciding what to link to as they are in deciding what to say.
That's certainly not the standard under which many online news sources operate today. As it stands, often the only link in a news story is to some commercial site, which is arguably an inappropriate promotion of a commercial interest.
Of course, you may not care about traditional news sources online. A phenomenon called "Weblogs" is, for some people, replacing online journalism. I'll give you a link on Weblogs and you can decide for yourself if you think they measure up: http://www.weblogs.com/about/. If I knew of an antiWeblog site, I'd link to that, too, but I can't find one.
Okay, here are two corrections to a recent column of mine: Rebol, Inc. is really Rebol Technologies; and saying that Rebol runs without recompilation on a lot of platforms is true enough but misleading, given that Rebol is interpreted, not compiled.
Michael Swaine
editor-at-large
mswaine@swaine.com