The Big Question: Do we need a more reliable online encyclopedia than Wikipedia?

The Independent, UK/October 18, 2006
By Paul Vallely

So what is Wikipedia?

Wikipedia has been around since 2001, which gives it whiskers in internet terms. It is now the largest encyclopedia in the world with articles on more than 5 million subjects in 229 languages - and an average of 1,515 new articles posted every day this month. It's also one of the most popular research tools on the web; last month 33 million people used it.

Not bad for an organisation with just one member of staff. The rest of the work is done by thousands of nerdy anorak devotees - self-confessed wikimaniacs - dedicated to the hubristic project of trying to assemble "the sum of all human knowledge". It reckons to have 3,800 hardcore users making more than 100 edits a month, another 18,000 who make at least five and countless others dipping in as the mood takes them. It's one of the closest expressions ever seen of genuine anarchy - a "self-regulating cooperative of free thinkers acting voluntarily for a greater common good". Their motto is "out of mediocrity, excellence."

So why would we need a new one?

You've had the clue in the word mediocrity. Wikipedia has no editor-in-chief, no academic peer review, no university imprimatur. Anyone can edit any entry. Which is why this week one of its founders, Larry Sanger, has broken away and announced he's launching a spin-off called Citizendium. Again, anyone can submit material, but he's got financial backing to hire experts to review all the submissions. This will, he says, make it far more reliable than Wikipedia.

How unreliable is it?

How long is a string of clichés? That's how a lot of Wikipedia entries read. But then others read as if they were written by people who know what they're talking about. The problem is with all the stuff in between, which looks reliable, but you never know. Using it is like asking questions of a bloke you met in the pub. He might be a nuclear physicist. Or he might be a fruitcake. Wikipedia acknowledges this. Lots of articles are prefaced with a rubric such as: "To meet Wikipedia's quality standards, this article may require clean-up." Which often is a considerable under-statement.

Robert McHenry, the former editor-in-chief of Encyclopædia Britannica, has been rather ruder. Visiting Wikipedia is like using a public lavatory, he has said. You don't know who has used the facilities before you.

But Wikipedia-use now dwarfs that of Britannica, primarily because the former is free and the latter has to be paid for. And with a free encyclopedia you get what you pay for.

But what is clear is the bias evident in some Wikipedia entries. Look up an entry like "Islam" or "Israel" and you will detect the hands of dedicated contributors with idiosyncratic beliefs whose views are there because no one has the time and energy to counteract the bias.

Some pages seem to have been taken over by fanatics and special interest groups (try the Scientology page). When others try to correct their pages the dedicatees "revert" the contributions of new contributors.

Are there other problems?

You have to watch out for self-aggrandisement, or for PR people adding entries for their clients. Then there are the politicians. Earlier this year it was revealed that staff in the US House of Representatives had embarked on a campaign to clean up the biogs of their political masters, and insert negative remarks about opponents, some of it pretty puerile, such as that the Congressman from Virginia "smells of cow dung". Most difficult to stop is disproportionate emphasis.Earlier in the year the entry on Coronation Street was twice as long as the article on Tony Blair. "The holes in Wikipedia are big enough to throw an old-fashioned printing press through," lamented one critic, frustrated at Wikipedia's negligible entries on Australian literature. If it didn't call itself an encyclopedia, said another frustrated user, but something like "Jimbo's Big Bag O'Trivia", it wouldn't be the problem it has become.

So why does anyone bother with it?

Because it has lots of good material on it. A report by the scientific journal, Nature, found that on 42 randomly selected science articles Wikipedia came close to Britannica in terms of accuracy. (The average Wikipedia article contained four errors or omissions; the average Britannica article, three).

Where Wikipedia is useful is as a pointer to more reliable sources - it is diligent about giving a reference to most key points. What that doesn't help with is the hoaxes.

What are the hoaxes and could they be stamped out?

There are leg-pulls like the now-deleted entry on the Funerary Violin - a little-known genre of music condemned by the Catholic church and almost wiped out by the Great Funerary Purges of the 1830s. (The entry proved to be entirely fictitious).

Then there is vandalism, much of it anonymous mischief, but occasionally surfacing publicly, as when the US comedian Stephen Colbert asked viewers to vandalise the page on elephants by inserting the bogus claim that the elephant population has tripled in the past six months.

Then there are the malicious hoaxes - as when someone created a page on a US politician called John Seigenthaler suggesting - utterly without foundation - that he might have been involved in the assassinations of both Robert F. Kennedy and John F. Kennedy. The page went undetected for months.

Wikipedia claims to have improved its response time in dealing with such errors. The organisation's founder Jimmy "Jimbo" Wales says he has introduced a "three revert rule", which refuses access to anyone who has made the same changes to an article more than three times in a 24-hour period. (Which is tough if they are right, and it is their opponent who is wrong). A recent study by IBM found that most vandalism on Wikipedia is reverted in about five minutes.

Mind you, I got that fact off Wikipedia, so you can't be sure. Any new version might be an improvement. But it could also be a lot more boring.

Can we trust the information on Wikipedia?

Yes...

* It's the world's largest encyclopedia covering 5 million topics - on the law of averages a lot of that must be right

* Last month alone 33 million people used it - 33 million people can't be wrong

* Wikipedia's founder, Jimbo Wales, says he has tightened up the rules on policing dodgy entries

No...

* Wikipedia has no editor-in-chief, no academic peer review, no university backing.

* Some pages get taken over by fanatics and special-interest groups

* Wikipedia's premise - that continuous improvement will lead to perfection - is completely unproven