Would You Trust Joe Isuzu’s Blog?

"I read it on the Web, so it must be true." OK, stop laughing. In fact, the only time I've heard that said recently is in an ironic tone of voice. But seriously folks, how can you trust what you read or see on the Web?

Brian Walski, a photojournalist covering the war in Iraq for the Los Angeles Times, submitted a photo in April 2003 that turned out to be a composite of two images taken moments apart. The manipulation was detected because he failed to eliminate a duplicate image of a bystander in the crowd. The paper published a correction and Walski resigned, but this incident has caused some readers to wonder whether other news photos have been altered to heighten their impact, particularly given the popularity of digital cameras.

And thinking about evaluating content that you cannot directly verify, last summer, it seemed that a number of journalists discovered Wikipedia (www.wikipedia.org), the free online encyclopedia built by contributors. Yes, anyone can add or change articles within Wikipedia, but changes are monitored and can be removed. One columnist, Al Fasoldt of the Syracuse Post-Standard, wrote a scathing article about Wikipedia, called it untrustworthy, and he was backed up by a local school librarian who told him that it "is not an authoritative source." Based on these comments, Alex Halavais, an assistant professor at the School of Informatics, SUNY Buffalo, decided to test the ability of Wikipedia to "heal" itself. He made 13 factual changes to the site—which he later described as vandalism—and planned to leave them there for two weeks. He expected to prove that Wikipedia is unreliable and un-vetted. Instead, all the changes were detected and fixed within a couple of hours, and Halavais conceded that he was impressed with Wikipedia's self-correcting nature. Of course, he made fairly obvious factual changes; if they'd been more subtle, and if no one was monitoring the entry to see when changes were made, it is possible that his test would have not been detected.

What this demonstrates is that we info pros may need to develop new ways of evaluating and testing the validity of Web-based resources that our clients and patrons are already comfortable with, such as Wikipedia. We are accustomed to looking for a brand name—we will trust a Web-based encyclopedia if it has a print equivalent, for example—but it is more difficult to judge the value of content that is built and maintained by thousands of volunteers. Clearly, a number of info pros are still reluctant to put our seal of approval on an emerging reference source.

Perhaps "trust, but verify" should be our mantra. As I write this column, there is still a debate raging as to the authenticity of a collection of documents relating to George W. Bush's service in the National Guard, obtained by CBS' Sixty Minutes. Within hours after they were released, bloggers had commented on the suspicious nature of the format, content, and especially the superscripts in documents apparently typed on typewriters in 1972 and 1973, most of which did not have the variable type needed to generate the raised "th" in "187th". There were discussions throughout the blogosphere: what typewriters were readily available, what size paper the Air Force used at the time, whether the signatures on different memos matched, and so on. And sure enough, within a day, journalists had picked up on the blogs and were reporting doubts about the documents.

This is, of course, one of those politically charged debates, and accusations are flying as to the motivations of all participants. But it does challenge the notion that information from an established news source is more reliable than information found on a blog. In this case, it was the bloggers who flushed out the bad (or at least questionable) information through their own analysis techniques.

And speaking of bloggers driving the news, recently I read an article in my local paper about the ease with which Kryptonite bicycle locks can be opened with the hollow shaft of a ballpoint pen. In bicycle-obsessed Boulder, this is front-page news. However, the original posting of the vulnerability of U-locks with cylinder keys appeared in bikeforums.net almost a week before. The news percolated through the blogosphere until it was finally picked up by the traditional press. One hopes that bike thieves are not particularly tech-savvy and had missed this news until now. (And I find it amusing that a bike mechanic quoted in one article referred to Kryptonite as "the Microsoft of locks." Any monoculture—whether in the forest, in email clients, or bike locks—is much more vulnerable to attack than a more diverse environment.) This suggests that print journalists are finally truly paying attention to the world of blogs, a remarkable change from a year ago.

The challenge will be for us to develop a new set of tools for monitoring these new information channels and new criteria for evaluating the authority and authenticity of the information we find.