March 14, 2003
Follow me on Twitter

Doing The Work

When you need to figure out whether a given Web page is trustworthy or not, there's no substitute for actually doing the work -- reading the page, researching the background, and forming a critical judgment. Relying on production values or GoogleRank can save time, but in the end it's bound to get you in trouble.

Adrian Miles, who is teaching students how to read and write weblogs, gives an example (that Jill Walker picks up for her class). A Google search turns up Rebecca Blood's history of blogging; can you rely on it?

We also discussed the domain (an eponymous .net) and what that might indicate, the design and writing values of the content, the extent of content available, and how none of these alone might mean much but when judged in relation to each other can tell you quite a lot. Finally, I pointed out that the site had a Google rank of 1 and that because of how Google determines page rank this is an index of authoritative rank, and possibly more authoritative than any of the other indicators.

The problem is, none of this turns up the fact that Blood's essay, though popular, is controversial. For example, Dave Winer (Scripting News, Radio Userland) wrote:

Her original history was all wrong, and while she's made some corrections, she still doesn't understand the medium, or even tell the story of how weblogs came to be with any accuracy. The appearance of Blood's book sparked a good deal of discussion -- much of it contentious and uncivil; Roger Cadenhead provides lead links.

Similar issues crop up all the time. Histories of hypertext always start with Vannevar Bush. The evidence for Bush's direct influence is slim, but you'd never know that without close reading. People are still talking about the hypertext navigation problem; we've known it wasn't a problem for a decade, but so many people say it that people suppose it must be true. The fourth-ranked site on "theory of evolution" at Google is crazy, a high school teacher's "Top Evidences" that the Christian creation myth is history, Just try a Google search for something that actively attracts flakes, like UFO or alien abduction .

From the surface -- design, domain, GoogleRank -- you can perhaps reach a preliminary opinion on whether the essay is crank or crackpot, but you cannot know whether it's wrong. It's entirely possible to be widely cited and wrong; that's why research hasn't been replaced by the one true Encyclopedia.