Larry Sanger, a co-founder of online encyclopaedia Wikipedia, now questions the validity of sites that reject expertise or authority.

Larry Sanger is an unusual combination: a philosophical technologist. Or a technological philosopher. You can place the emphasis either way, but what you end up with is a person who thinks in some very interesting and provocative ways about the internet and its many communities.

A philosophy PhD holder and former Ohio State University lecturer, his roots in technology and the internet are well established. Sanger is a co-founder of the well-known (some, including him, would now say notorious) Wikipedia, the online encyclopaedia that anyone can go in and edit.

In fine philosophical tradition, he has had to publicly debate and defend the extent of his role in Wikipedia with doubters, the most prominent being his other co-founder, Jimmy Wales. Sanger notes that Wikipedia press releases listed him as co-founder for two years while he was there, so what is the problem now?

Almost certainly, one of the issues is that he is a vocal critic of many aspects of what Wikipedia has become. The original idea to use "wiki", a type of webpage that can be mass-edited, as the basis for a free online encyclopaedia for the people, was a daring experiment in harnessing what might be called the creative commons - the energy and (one hopes) intelligence of the net's general population.

While many love Wikipedia's democratic approach to knowledge, based on an assumption that the wider audience often has more expertise and knowledge to contribute on a topic than a designated "expert", Sanger believes Wikipedia now proves that concept isn't workable.

Many entries have turned into examples of how a democracy can veer into a mob, especially as edits can be made anonymously.

Deliberate inaccuracies and invective and political tussles mean no entry can ever be trusted.

Yet, as he told an audience recently at the Irish Institute of European Affairs, students regularly cite Wikipedia articles in their research, and the encyclopaedia is generally taken as a trustworthy authority.

His former project thus becomes Exhibit A as he poses the question: "Why haven't governments tried to regulate online communities more?" He is quick to emphasise that he doesn't ask the question because he believes governments should; he asks because he fears they will, unless online communities become more adept at regulating themselves.

He refers to such communities as often-powerful "cyber-polities" and believes they engage in what he calls "a politics of knowledge".

"Knowledge is taken to come from those who have authority. But I think authorities can be, and frequently are, wrong," he says.

His concern is that there is little to no oversight over what is said and whether it has any truth or authority at all.

Yet, because Wikipedia itself has become so pervasive as a source of information, the very fact that an entry is in Wikipedia makes it "authoritative".

"When Wikipedia speaks, like it or not, people listen," he says.

He argues that online communities need clear policies for their internal governance and that this is closely tied to, and influences, what Sanger calls "their affairs at large".

He also claims that those internal policies and governance issues "are closely connected to real world politics"? In other words, he asks, does it matter what policies Wikipedia has?

"Actually, it does," he says. Decisions to allow anonymous edits without oversight mean the latest edits become "the truth" until the next edit comes along. And such edits can have real impact on individual people. He cites the well-known case of the prankster who edited an entry on longtime journalist John Seigenthaler Sr, to state that he was involved in the assassination of Bobby Kennedy (whom he once worked for) and John F Kennedy as well. The entry remained in this version for four months before Wikipedia agreed to take it down, to Seigenthaler's frustration.

Sanger says he thinks it entirely plausible that if such things continue to happen, governments could step in to take editing control, create new laws to regulate the internet in new ways, and become the web's de facto police. While admitting that yes, it would be very hard to totally control the net, he feels governments could indeed make the web a far more controlled space for the majority of users.

He proposes that cyberpolities will operate in a more transparent and responsible way when they are republics "with some sort of elected board and representatives" than a democratic free for all. He doesn't like benign dictatorships either - a term often used to describe the way some online communities are run by the original founder of the site with moderators chosen by the founder. Maybe for small online communities, where there aren't enough people who would want to run for a board, he says, but not for larger communities whose actions can have larger ramifications.

Seigenthaler directly contacted Sanger - by then, no longer involved with Wikipedia - and voiced his disappointment that Wikipedia had turned into a forum for such inaccuracies. "That really hit home bad," Sanger acknowledges, during an interview after his provocative lecture.

His more considered response was to think about how the same general idea of an encyclopaedia - that could be written and edited by the public - could be modified to get rid of what he sees as the inherent problems with open, anonymous access and lack of internal governance.

His answer: the Citizendium, or Citizen's Compendium, www.citizendium.org.

This is Sanger's "new knowledge society" where the general public and experts collaborate.

Contributors must use their real names, and articles are edited and overseen by volunteer experts in various subjects. He believes so strongly in a cyber republic that he has already prepared for his own obsolescence - he will step down as head of the Citizendium in the next two years and the torch will be passed.

Doesn't he worry that the Citizendium will invariably be seen as a way of getting back at his former colleague and Wikipedia co-founder Jimmy Wales?

He shrugs. "I don't worry too much about that," he says. However, whatever the politics, he is clearly passionate about Citizendium, both as a knowledge resource and as a more ideal form for an online cyberpolity.

Won't it be hard to pose any real challenge to something as big - some might say, such a force of nature - as Wikipedia? He acknowledges Citizendium is small right now, but adds: "I don't really compare ourselves to Wikipedia, but to Wikipedia's first year." Citizendium is growing at pretty much the same rate as Wikipedia which, back then, had seemed an unlikely future success story.

He believes fully that people ultimately will turn to an edited encyclopaedia with named contributors and editors.

But why does the world even need a free online encyclopaedia of any sort? If people want accountability and reliability, why not just opt for tradition and pay for the online Encyclopaedia Britannica? Sanger smiles. "It's bigger, and it's free."

Larry Sanger was the inaugural speaker at the Irish Institute of European Affair's lecture series "Digital Futures".

His full talk can be heard at: www.iiea.com/audio/larrysanger.mp3