Wikipedia is a real-life Hitchhiker's Guide: huge, nerdy, and imprecise.
By Paul Boutin Posted Tuesday, May 3, 2005, at 2:37 PM PT
It's too bad Douglas Adams wasn't able to see his vision brought to life. I don't mean the so-so movie version of The Hitchhiker's Guide to the Galaxy. I'm talking about Wikipedia, the Web's own don't-panic guide to everything.
The parallels between The Hitchhiker's Guide (as found in Adams' original BBC radio series and novels) and Wikipedia are so striking, it's a wonder that the author's rabid fans don't think he invented time travel. Since its editor was perennially out to lunch, the Guide was amended "by any passing stranger who happened to wander into the empty offices on an afternoon and saw something worth doing." This anonymous group effort ends up outselling Encyclopedia Galactica even though "it has many omissions and contains much that is apocryphal, or at least wildly inaccurate."
Adams actually launched his own onlin guide before he died in 2001, but it was, he wrote, "still a little like the fossil record in that it consists almost entirely of gaps." Wikipedia is a colossal improvement—it's just like the fictional Hitchhiker's Guide, only nerdier. Wikipedia is the Web fetishist's ideal data structure: It's free, it's open-source, and it features a 4,000-word exegesis of Dune.
For decades, software-makers competed to build complex collaboration systems. These high-end tools, like Lotus Notes, let companies specify who can edit which documents and establish complex approval procedures for changes. In 1995, software researcher Ward Cunningham destroyed the hierarchies by designing a site, the WikiWikiWeb, that anyone could dit. (Wiki-wiki means "quick quick" in Hawaiian. Cunningham saw it on a Honolulu Airport bus.)
Wikipedia, with more than 1 million entries in at least 10 languages, is the mother of all wikis, but there are also wikis devoted to quotations, the city of Seattle, and Irish politics. (Check out this wiki of wikis, which lists more than 1,000 sites.) Instead of enforcing rules, wikis trust that groups can behave. Anyone can edit or reorganize their contents. If you realize something's missing, incomplete, or incorrect, you can fix it yourself without asking permission. "People told me that the experience changed their lives," Cunningham said via e-mail.
Don't expect Wikipedia to change your life, though, unless you've secretly longed to be an encyclopedia editor. Just because you give everyone read and write permissions doesn't mean everyone will use them. Wiki lovers argue that they are collaborative, self-correcting, living documents that evolve to hold the sum of all the knowledge of their users. But, like blogging, editing the Net's encyclopedia appeals to a small, enthusiastic demographic.
Like the Guide's lengthy entries on drinking, Wikipedia mirrors the interests of its writers rather than its readers. You'll find more on Slashdot than The New Yorker. The entry for Cory Doctorow is three times as long as the one for E.L. Doctorow. Film buffs have yet to post a page on Through a Glass Darkly; they're too busy tweaking the seven-part entry onTron.
But excessive nerdiness isn't what's keeping Wikipedia from becoming the Net's killer resource. Accuracy is. In a Wired feature story, Daniel Pink (kind of) praised the hulking encyclopedia by saying you can "[l]ook up any topic you know something about and you'll probably find that the Wikipedia entry is, if not perfect, not bad." But don't people use encyclopedias to look up stuff they don't know anything about? Even if a reference tool is 98 percent right, it's not useful if you don't know which 2 percent is wrong. The entry for Slate, for instance, claims that several freelance writers are "columnists on staff" and still lists Cyrus Krohn as publisher months after the Washington Post Co.'s Cliff Sloan took over.
Just because the Wikipedia elves will probably fix those errors by the time you read this article doesn't mean that the system is inherently self-healing. Not everyone who uses a wiki wants to hit from both sides of the plate. The subset of enthusiastic writers and editors is orders of magnitude smaller than the group of passive readers who'll never get around to contributing anything.
Bashing Wikipedia is nearly as risky as bashing Scientology. I know that I'm going to get barraged by the Wikivangelists—"If an entry's wrong," they'll say, "stop complaining about it and fix it." But if I were truly conscientious, I'd have to stop and edit something almost every time I use Wikipedia. Most people are like Douglas Adams' characters—we resolvefirmly to stay and fix it after work then forget the whole episode by lunchtime. Wikipedia is a good first stop to get the basics in a hurry, especially for tech and pop culture topics that probably won't ever make it into Britannica. I'm just careful not to use it to settle bar bets or as source material for an article. I made that mistake exactly once.
Wikis are a great way to collect group knowledge, but not every reference book in the galaxy will turn into one. A couple of weeks ago, online reports claimed that Microsoft's Encarta decided to wikify its paltry 42,000 entries. Encarta's Editorial Director Gary Alt told me that the truth is prosaic. Readers will be able to submit suggested corrections or improvements to existing entries, but Encarta is not looking for new entries, and the editors will still decide what's worth including.
An elitist encyclopedia like Encarta will never be able to match the breadth or speed of a user-edited reference library, but it's smart to coax readers into helping stretch its inherent advantage—reliability. Alt told me he's hiring all of six people to review and research reader submissions. Unlike the editor of The Hitchhiker's Guide, they'll probably be eating lunch at their desks.
Related in SlateLast year, Clive Thompson asked if an online crowd could write a novel. In January, Jack Shafer said that blogs need a little time to catch up to the hype.
Paul Boutin is a Silicon Valley writer who spent 15 years as a software engineer and manager.