Trust me, this is the last word. No, this is

Is Wikipedia an open online forum that leads to accuracy or an anti-elitist free-for-all in which people 'vote for the truth'? Chris Bunting investigates

In the four years since it was set up, the internet encyclopaedia Wikipedia has become the largest encyclopaedia in the world.

It boasts more than 1.4 million articles in more than 195 languages and is adding about 1 million words a day.

The English-language edition alone has more than 510,000 articles; its rival, the online Encyclopaedia Britannica , has about 120,000. This extraordinary outpouring of human effort has all been done without a penny being paid for its contributions or an ounce of kudos lent to its contributors. This is because Wikipedia is based on a simple but radical concept: anybody can write or edit any of its articles without having to register or offer credentials.

From the days of Diderot, encyclopaedias have spoken with authority. They are founts of knowledge, produced by relatively small groups of people who earn the right to tell us what is true and what is not by drawing their credibility from formal positions within academic hierarchies. Wikipedia, taking its name from a type of webpage that can be altered by any visitor, turns that philosophy on its head.

For some, it is an abomination. Robert McHenry, a former editor-in-chief of Britannica , says the time-tested methods needed to ensure accuracy have been ditched for a "completely irrelevant principle - openness". "It is simply not possible to produce a reliable work of reference without having some sort of process in place that will encourage, if not guarantee, better rather than poorer content."

Andrew Orlowski, an internet journalist, calls Wikipedia enthusiasts "the Khmer Rouge in diapers" who casually reject established institutions of learning for the sake of messianic ideas about the "mystical power" of the internet. He feels that its exponents fail to explain how a system can work in which any article is only as good as the last know-nothing who decided he was qualified to edit it. "You can't vote for the truth," he says.

There is certainly plenty of wrong information in Wikipedia's pages, admits Jimmy Wales, the former futures and options trader who founded the project. But he notes: "There are tons of errors in Britannica . The average quality on Wikipedia is actually very high. But it is also true that any particular page might be rubbish at the moment you happen to look at it."

He stresses that efforts are made to maintain standards. "There is a lot of anarchy, but there are rules and community structures that are there to help us create a high-quality encyclopaedia." Contributors who abuse the site can be banned, while controversial articles that become the subjects of "editing wars" between irreconcilable contributors can be locked so that further changes cannot be made. More typically, though, the administrators try to encourage debate on linked discussion boards to resolve disagreements.

Wikipedia's main argument against attacks on its lack of authority is that free discussion and co-operation will tend, over time, to create more accurate and comprehensive articles. McHenry, though, says it is just as likely that there will be a tendency towards the mediocre.

The Wikipedia model does have one major strength: the freedom of its coverage. There is something liberating about browsing its entries compared with those of its more orthodox competitors. The ease with which content can be created also means it is often far ahead in covering recent events.

The challenge for Wikipedia will be to retain that speed and lightness of touch while ensuring authority and reliability. Larry Sanger, who worked with Wales at the birth of Wikipedia but left three years ago to become a philosophy lecturer at Ohio University, has criticised its "anti-elitism".

He says: "We need experts contributing and, generally speaking, you are not going to get many experts if they are not in control. On many subjects, such as very technical areas, you cannot write credibly without expertise."

He believes there needs to be a decisive shift back towards a more traditional methodology: perhaps appointing experts with authority to commission and lock articles to oversee parts of the site.

Wales, though, is unequivocal: "My basic response is that he (Sanger) is completely wrong about everything. We need to find ways of acknowledging expertise, but we need to move carefully. It is one thing to say we need to acknowledge expertise. It is another to see how to make it compatible with what Wikipedia is."