The article about Eric Raymond in Wikipedia certainly paints an intriguing portrait, even if some of the facts seem to be in dispute. A US technology visionary, Raymond is described as both an anarcho-capitalist (defined elsewhere in the online encyclopedia as a libertarian who believes in the completely unfettered workings of the market) and a neopagan “initiate witch and coven leader”. He is said to be an “enthusiastic amateur musician”, a strong opponent of gun control, and a controversial commentator who has argued that blacks are responsible for a larger proportion of crimes than whites because they have lower IQs. His claim to a black belt in taekwondo is unsubstantiated, and his self-professed leading role in the Linux software development project is said to be in dispute.

If these incidentals in Raymond’s life seem to get undue prominence, the article also finds room for his main claim to fame. As the author of The Cathedral and the Bazaar, an influential book about the workings of a new approach to software development known as open source, Raymond is considered one of the early proponents of an intellectual movement that has swept to prominence with the rise of the internet.

Its essence is simple: that ad hoc groups of like-minded people, working in loosely organised online communities (the bazaar of the title) are capable of producing feats of intellectual achievement that until now had been thought within reach only of highly organised teams of specialised experts (the builders of the cathedral.)

That makes him just the kind of person likely to lend his intellectual weight to a project like Wikipedia itself. Open to anyone with access to a browser, the collaborative encyclopedia has become one of the most remarkable phenomena of the online world since its launch in 2001. Its English language version claims some 1.4 million articles, with another 3.5 million in 229 other languages. It ranks as the 15th most popular site on the Web, and its listings often appear near the top in Google search results - the modern stamp of public approval, since it shows that many other websites consider its results authoritative enough to link to.

For Raymond, though, mixing with the masses on Wikipedia has left a lot to be desired. His contributions on the subject of science fiction have frequently been rewritten by others, and his attempts to reinstate his own ideas shouted down.

“It left a bad taste in my mouth,” he says. His unflattering conclusion about the quality of intellectual debate on Wikipedia is: “You tend to find that articles are taken over by moonbats.” Asked what he thought of the Wikipedia entry about him, whether it was accurate, fair, balanced, he replied: “Last I checked, it was neither accurate, nor fair, nor balanced.”

Raymond has now retreated into the sort of intellectual cocoon that might seem surprising for an arch proponent of open online collaboration. He points with pride to his contributions to Encyclopaedia Britannica - a distinctly closed realm where the traditional virtues of scholarly excellence and expert editing are still valued. Ironically, those contributions will soon include an article on the open source movement, whose aspirations seem so antithetical to those of Encyclopaedia Britannica itself.

According to critics such as Raymond, there is a crisis in Wikipedia-land, an idea that is starting to show its fatal flaws. “It’s not a process issue: it’s a fundamental philosophical issue,” says Raymond, who argues that a loosely organised group can never agree on something as important and as subjective as this: how to present the sum of human knowledge.

Others are trying to do something about these supposed weaknesses. Larry Sanger, one of the two co-founders of Wikipedia and its first editor, is working on a rival project that would take all of the online encyclopedia’s existing entries, then feed them through a filter comprised of expert specialists. Known as Citizendium, it would in effect impose a traditional layer of scholarly order on the unruly masses.

The very idea sends Wikipedia purists into paroxysms. Clay Shirky, a technology writer and teacher who is one of the project’s most vociferous supporters, says that trying to impose a more traditional hierarchical system on top of an open collaborative one would kill the idea stone dead. “The distributed systems that are succeeding in the world today are doing so not because they are the best possible systems, but because they work,” he says.

The outcome of this intellectual tussle is about more than just Wikipedia. The online encyclopedia is the most visible and ambitious example of a new cultural movement that is forming around the internet. It is based on the belief that the collective actions of millions of individuals, properly co-ordinated, can yield powerful results. If Wikipedia can overcome its shortcomings and create a stable model for mass collaboration, it will have gone a long way towards breaking down the entrenched power of society’s experts.

The idea that a specially trained group holds unique insight and power has been prevalent “ever since the formation of guilds”, says Ross Mayfield, founder of Socialtext, a company that sells “social software” tools like wikis - webpages that are open to a group of people to write and edit. “Things are different today: we have an alternative, and it is more open.”

No wonder the experts, through projects like Citizendium, are planning a counter-attack.

The best-known test to have been carried out on Wikipedia’s accuracy was largely a red herring. Last year, the magazine Nature published the findings of a comparison between the open online encyclopedia and Britannica. The apparent conclusion: that there is really not much to show between the two.

Clay Shirky offers one reason for this: the way that Wikipedia is organised makes it relatively easy to defeat “vandalism”, the term that Wikipedians give to the work of malicious editors who try to undermine articles on the site. Returning an article to an earlier state is quicker to do than defacing it in the first place, he says - like making it easier to remove graffiti than it is for the miscreant to apply it.

A study by three IBM research scientists, carried out in 2004, concluded that Wikipedia has “remarkable resilience” to vandalism. Malicious edits “were often corrected rapidly, sometimes in a matter of minutes, by members of the community”, they found.

The Nature test, however, was hardly representative, or even one that touched on the most fundamental questions raised by the Wikipedia experiment. It turns out that a collaborative reference work like this is pretty good at explaining the basic facts (the focus of the Nature experiment). Even the encyclopedia’s critics concede this point.

Things start to go awry when it comes to less clear-cut issues - history, say, or politics. The so-called “edit wars” that are waged on Wikipedia over any issue that attracts strong ideological disagreement have forced the site to retreat from its own ideal of openness. The article about George W. Bush, for instance, has been “locked” since 2004, open only to a few people to edit. Roughly one out of every thousand pages is protected in this way.

This invites a philosophical question. If anyone can present his or her own view of the truth on a communal website like this, which should take precedence? From this perspective, the whole idea of a post-modernist encyclopedia seems ironic: if there are as many different views of the truth as there are contributors, can any one version claim enough authority to merit being placed in an encyclopedia in the first place?

Asked whether Wikipedia invites this kind of relativistic critique, Jimmy Wales, chair of the board of trustees of its parent group, the Wikimedia Foundation, instinctively recoils. “Yikes, I most certainly hope not,” he says. “Postmodernist/deconstructionist theories of truth are incoherent at best.”

Sanger, who left Wikipedia after its first year and has been at odds with Wales ever since, can at least agree with him on this. “As a philosophical point, the Wikipedia doesn’t add much to our understanding of truth or knowledge,” he says. “But it’s adding a lot to the debate about these as social topics.”

The real value of Wikipedia, in other words, is what it shows about the social interaction that yields an agreed view of the truth. That makes it primarily a study in politics, not philosophy. It is also, like many academic disputes that have gone before, highly personal.

Larry Sanger comes across as a man with a grudge. With a PhD in philosophy and an abiding personal interest in epistemology, he must have been an easy choice when Jimmy Wales was looking to hire an editor-in-chief for Nupedia, his first attempt at an open online reference work. That project quickly ran into the sand: the process for editing contributions took too long, with the result that few articles saw the light of day. In its place, Sanger proposed something altogether simpler: a reference work built around a “wiki”, an online document that is open to anyone who wants to write in or edit it.

Freed from Nupedia’s constraints, Wikipedia took off quickly. Yet to hear Sanger’s version of events, things started to go off the rails just months after it was launched. By the summer of 2001, he says, the new online community was being overrun by what he calls “trolls” and “anarchist-types” - people “opposed to the idea that anyone should have any kind of authority that others do not”.

Sanger’s response was to propose a stronger role for expert editors, people with the power to resolve disputes and lay down the law. Tired of fighting endless battles - and feeling he had little support from Wales - Sanger eventually left. Citizendium, the expert-edited version of Wikipedia that he says will be launched before the end of this year, is his latest attempt to rewrite this particular piece of history.

“He’s really out to right a wrong,” says Shirky. “He regards a lack of deference to expertise as wrong.” Certainly, Sanger is unconcerned about sounding unfashionably elitist. In the online community that he proposes, people with the trappings of expertise - such as an academic or professional qualification, or demonstrable experience in a particular field - will hold sway.

“The concept of expert I’m dealing with is a traditional one,” he says. The traditional course of specialised study that scholars undertake leaves them with “a reliability and rapidity of judgment” that self-made “experts” can’t hope to master. Also: “Scholars are trained to represent different sides in a debate.”

Wales, by contrast, has always taken a less interventionist stance. While he says that expertise certainly has a role in Wikipedia, he has not tried to enshrine it in the encyclopedia’s editorial processes. Asked how disagreements should best be resolved, Wales offers only this: “With strong support for individual rights, and respect for reason.”

Sanger, for his part, is still clearly irritated by his former mentor’s unwillingness to intercede more in the early disputes that racked Wikipedia. “He [Wales] is the quintessential hands-off manager, and he applies that to the community,” he says.

This rivalry has spilled over into a very personal form of “edit war”. Sanger accuses Wales of airbrushing history: in press releases from the foundation that runs Wikipedia, he claims, his own involvement in the project mysteriously disappeared some time in 2004. “I was not even mentioned, and Jimmy Wales began to be represented as the singular founder of Wikipedia,” he complains.

Wales has also made changes to his own entry in the encyclopedia, even after a rule was adopted against this practice in 2002.

Sanger also admits to having edited his own entry in Wikipedia “in the early days, along with a lot of other people”, though he adds he stopped when, “at a certain point, it became clear that it was against the policy of Wikipedia”.

When it comes to the history of Wikipedia itself, that has left a curious form of impersonal stand-off between Sanger’s version of events and Wales’s.

“The people who have edited these articles weren’t actually there when these events were happening,” complains Sanger. “They make an attempt to state events neutrally: what that means is some midpoint between what Jimmy thinks and what I think.”

This highlights a delicious irony at the heart of Wikipedia. Every person is the world’s greatest expert on at least one subject: him- or herself. And yet how can you trust anyone to offer up a true rendition of something so personal?

Since Wikipedia is at heart a massive online social experiment, it is hardly surprising that so much of the discussion about it is shot through with the language of Utopianism. It also touches deep political nerves, which helps to explain why its proponents and critics can sound so strident.

On one side of this divide are the idealists who believe that individual expression and mass online collaboration, freed from the traditional constraints found in the offline world, can yield incalculable benefits. Traditional deference to expertise imposes a huge tax on society, argues Shirky - by concentrating perceived wisdom in a small number of people, it limits the flow of ideas. According to this view, the experts represent a ruling caste whose main interest is in protecting their own privileges. The internet offers a historic opportunity to undermine that authority.

The other side of this argument was outlined by technology author Jaron Lanier in an essay earlier this year titled “Digital Maoism”. Wikipedia, he argued, is part of a broader movement on the internet that aims to promote the collective view above individual judgment. The ability to collect and present the views of the many has given rise to the belief in “a new online collectivism that is nothing less than a resurgence of the idea that the collective is all-wise”.

Collectivist ideas like this have had “dreadful consequences when thrust upon us from the extreme right or the extreme left in various historical periods,” says Lanier. “The fact that it’s now being reintroduced today by prominent technologists and futurists, people who in many cases I know and like, doesn’t make it any less dangerous.”

The rivalry between these views points to what Sanger calls “the new politics of knowledge”. How a society’s collective knowledge is determined - what appears in its encyclopedias - depends more than anything on the processes at work. Like a political system, the design of these processes is all-important to the final result.

In Sanger’s view, Wikipedia has fallen into the trap that befalls all anarchistic societies: starting in a perfect state with no preconceived rules, they collapse into mob rule or dictatorship. Constitutional representative democracy, on the other hand, relies on a willing transfer of authority to people charged with representing the interests of the many: in other words, the experts.

Translated to the Utopian world of Citizendium, anyone with sufficient specialist knowledge and training will be able to volunteer to become an expert, and order will be maintained by a cadre of volunteer “constables” whose job it is to keep the peace. The ultimate arbiter will be known as the Chief Constable.

The critics say it simply isn’t possible to impose an authority structure like this on the sort of free-wheeling and open online communities that seem to thrive on the web - such as Wikipedia - without destroying the very thing that gave rise to them in the first place.

“The minute you create a society where one person has more authority than another, you have a policing problem,” says Shirky. He compares the debate about online social engineering represented by Citizendium to debating the number of “angels on the head of a pin - it’s just not going to work”. Rather than try to engineer a perfect social outcome, he says, better to leave it to the community to evolve its own processes.

A more recent study by the IBM research scientists may offer some hope that this could indeed lead eventually to a more ordered and stable online community. A debate has grown up around Wikipedia’s editing processes, producing an explosion in the number of pages on the site dedicated to airing, and resolving, disputes. One possible conclusion, say the IBM researchers, is that Wikipedia “is becoming less anarchic and more driven by policies and guidelines”.

The mob, it seems, may have the power to create order out of its own chaos: “Such trappings of bureaucracy are often seen as the result of the exertion of power from the top down, yet in Wikipedia they seem to emerge, to some degree, spontaneously.”