When Bernardo Huberman looks at the chaos of millions of people online, he is often amazed at the order that emerges.

And few destinations on the internet encompass this duality quite like the online encyclopedia Wikipedia, where almost five million volunteer editors have contributed over 50 million edits to close to 1.7 million articles with almost no supervision.

"It's this strange, haphazard system, but it forms a kind of collective intelligence," said Huberman, the director of the Information Dynamics Lab at Hewlett Packard Laboratories and author of The Laws of the Web: Patterns in Ecology of Information .

"People don't get paid and don't even get credit for the work they are doing," Huberman said to CBC News Online. "But somehow it has managed to thrive because of the value people place on knowledge."

On one level, Wikipedia has done more than bring knowledge to the masses: it has given them the ability to shape and decide what that knowledge could be.

Wikipedia and Wikimedia founder Jimmy Wales sees the online encyclopedia as a forum where people can fairly and openly discuss and explain things to each other. (Lisa Poole/Associated Press)

Wikipedia founder Jimmy Wales isn't a fan of the term "collective intelligence" himself, instead choosing to think of the intelligence as being individual, but shared.

"I think the most important thing is that people come together with love and reason to discuss things and try and explain things in a fair way," he wrote in an e-mail to CBC News Online.

So in this respect, Wales thinks Wikipedia is a good model for a more open internet.

But allowing anyone with access to the site to add, edit or remove information is not without its own perils, and as Wikipedia has grown in popularity it has also attracted attention to its warts: a sprawling, difficult-to-manage system that exposes the site's entries to vandalism, factual errors, uneven writing and even allegations of plagiarism and misrepresentation among its mostly pseudonymous editors.

Knowledge for the people is one thing, but knowledge by the people may still be a work in progress.

Wikipedia launched its English edition in 2001, incorporating the format of an encyclopedia with the technology of wikis: websites that allow visitors to edit and write the content on its pages.

Six years later, the site is now among the most popular on the web. Wikipedia is the ninth most popular destination for Canadians and the 11th most popular site worldwide, according to Alexa web search.

At the heart of its popularity is the speed with which entries are updated, the depth of the entries (Wikipedia entries can run well over 6,000 words) and the breadth of topics: on few other forums will entries on the Boer Wars and the World of Warcraft rub shoulders.

Tempting target

The Wikipedia biography of Sinbad, whose real name is David Adkins, incorrectly stated the U.S. comedian had died of a heart attack in March. 'It's gonna be more commonplace as the internet opens up more and more. It's not that strange,' the Los Angeles-based entertainer said. A Wikipedia spokesperson blamed the incident on vandals. (Danny Moloshok/Associated Press)

But with that popularity, Wikipedia has become a tempting target for vandalism, which itself ranges from the serious to the silly.

Some of these changes last for a long time on the site, but most often volunteer editors are able to quickly fix the entries.

For four months in 2005, journalist John Seigenthaler's entry suggested he might have been involved in the assassination of John F. Kennedy. For a day in March 2007, the comedian Sinbad was pronounced dead. For about five minutes on April 11, 2007, Canadian basketball star Steve Nash's height was listed as two centimetres. (He is actually six foot three).

The Seigenthaler incident was the first black eye for Wikipedia, but it hasn't been the only one. In the fall of 2006, critic Daniel Brandt, who is credited with tracking down the source of the Seigenthaler vandalism, cited 142 examples of plagiarism on Wikipedia on his website wikipediawatch. In December, golfer Fuzzy Zoeller sued the owner of an IP address cited as the alleged origin of defamatory comments written in his Wikipedia entry.

And last month it suffered another hit to its reputation after it was revealed an editor who had represented himself as a university professor was in fact a 24-year-old from Kentucky named Ryan Jordan with no higher-education credentials.

Jordan, who went by the name EssJay, told the New Yorker magazine last year he was a tenured professor of religion at a private university with "a PhD in theology and a degree in canon law." But Jordan was exposed after he accepted a job at Wikia, the internet company run by Wales. Wikipedia has since put forth a policy stating that those contributors who wish to list credentials will need to provide proof of their claims.

Mob rule?

Larry Sanger, a former philosophy professor and co-founder of Wikipedia is critical of the lack of expertise in many Wikipedia entries. He started at alternative, called Citizendium.com, in March 2007. (Kiichiro Sato/Associated Press)

Larry Sanger, Wikipedia's one-time de facto editor-in-chief, believes the series of scandals have exposed the limitations of the online reference tool, which he argues embraces the ideal of equality over a real-world need for guidance from a community of experts.

"There's a deep commitment to egalitarianism within the Wikipedia community," said Sanger, who resigned in 2002 and launched a rival encyclopedia called Citizendium in March 2007. "It makes them uncomfortable with the idea of any special role for experts or anyone with any sort of real editorial power."

Citizendium has 1,490 articles available at this point.

Sanger is also fascinated by the changing and elusive question of what it means to "know" something and thinks experts play a vital role in helping to shape knowledge.

"The basic wiki model is sound and I'm a great believer in it," Sanger said. "But I think experts in particular fields can play a role in releasing more stable entries, in the same way that open source software relies on a core group of programmers to release versions others can work from."

Without the guiding hand of experts, Sanger argues, the rules governing Wikipedia are left to the mob or — when an issue becomes particularly polarizing — a small group of core administrators. The result, he says, is that real experts are staying away from the site for fear their voices will be drowned out.

Wales doesn't buy this line of reasoning, however, arguing real academics aren't afraid to get their hands dirty in the court of public opinion.

"Our experience is exactly the opposite: the best academics think that interacting with the general public in the open marketplace of ideas is core to what it means to be an intellectual," he wrote.

A new field of study

In an effort to understand the effectiveness of the marketplace of ideas, HP researcher Dennis Wilkinson, with help from Huberman, conducted a study of 1,211 "featured articles" on Wikipedia — those deemed by users to be particularly useful.

The two researchers found that, on average, featured articles had more edits and contributors than entries deemed less accurate or thorough on the site.

"Wikipedia article quality continues to increase, on average, as the number of collaborators and the number of edits increases," said their report, published in February 2007 on Cornell University's arXiv website.

But Sanger is skeptical of the results of the study, saying it is a leap to suggest the popularity of articles translates to quality.

"If it were established that featured articles were indeed of higher quality, then I'd be willing to accept the wisdom of crowds," he said.

"But I suspect popularity and quality do not necessarily match up. There are articles that have been worked on extensively that are not very good."

It's not the first study to lend an air of authority to Wikipedia. A 2005 study in the journal Nature found Wikipedia was about as accurate in covering scientific topics as Encyclopedia Britannica. Based on 42 articles reviewed by experts, the average scientific entry in Wikipedia contained four errors or omissions, while Britannica had three.

Inconsistency of writing

But when it comes to consistency and comprehensibility of writing, Wikipedia's team writing and editing approach is not without its drawbacks. The introduction to the specialized theory of physics known as quantum chromodynamics, for example, eschews easy language in sentences such as "QCD is a quantum field theory of a special kind called a non-abelian gauge theory," instead offering 11 hyperlinks in the first paragraph alone to explain difficult terms.

By comparison, an entry on the same topic in the Britannica concise edition uses simple language to explain a complex subject.

Huberman says the popularity of certain subjects lead to a winner-take-all approach for Wikipedia entries, leading to articles on the fringes of popular knowledge having more uneven levels of quality.

"I think the more specialized and dense subject matters don't invite as many edits, and so if there is a problem in it, it may take longer to find," he said.

Wales agrees some entries could be better written, but says the subject matter often determines the language used.

"What we normally do is allow the nature of the subject matter at hand to help determine the level of the article," he wrote.

"An article on Paris, France, ought to be accessible to just about any adult of normal intelligence. An article on physics, too. But on a particular obscure [to the layperson] concept, it is acceptable to write at a higher level, so long as the links are given to background materials."

"We hope to achieve Britannica or better quality, and in some areas we do, but for the most part we currently do not," said Wales. "We take our moral responsibilities very seriously and deeply regret every error."

But Wales said that while he sometimes feels put upon by the media every time an error creeps up on the site, he remains undeterred in his basic vision for the wiki model of collaboration.

"We have this global tool for sharing knowledge, so?let's use it for sharing knowledge."