It was a classic storm in a teacup. Several newspaper obituaries of Ronnie Hazlehurst, the BBC's former light entertainment musical director and author of the theme music for many well-known comedy shows, claimed that he had been the co-author of the song 'Reach' by S Club 7. The only problem was that Mr Hazlehurst hadn't had anything to do with the song, as a casual search on Google reveals in seconds (it was written by Cathy Dennis and Andrew Todd).
Why did the newspaper obituarists converge on the same error? The answer seems to be that the obits were at least partially sourced from Wikipedia, the online open-source encyclopedia, which contained a short entry on Hazlehurst in which the spurious claim about the S Club 7 song was published.
The reference was deleted on 3 October, two days after Hazlehurst's death. But that did not stop sanctimonious debate in the mainstream media about the unreliability of Wikipedia.
This is par for the course. The idea that a collaborative writing project which allows virtually anyone to edit entries could be regarded as an authoritative source is seen in some quarters as the epistemological equivalent of a sin against nature. So errors and lacunae in the encyclopedia are invariably seized upon with glee, and endlessly discussed in op-ed and other columns.
Wikipedia is a remarkable phenomenon. As of last Friday evening, it contained more than 2 million articles in English, all of them written and edited by volunteers. It has five employees in addition to Jimmy Wales, its co-founder, and carries no advertising. It meets most of its $750,000 (about £375,000) annual budget from donations, most of them of $20 or less. Anyone with an internet address can create or edit a Wikipedia entry. The site currently exists in dozens of languages and has 'hundreds of thousands' of contributors across the world. And for some time it has remained one of the top 10 sites on the web in terms of the number of daily visitors (it's currently number nine) - which puts it up there with Google, Yahoo, YouTube, MSN, MySpace and Facebook.
Not surprisingly, it's uneven. Articles on abstruse subjects sit side by side with pages on topics of the utmost triviality. Entries are sometimes vandalised by jokers (the page on George W Bush, for instance), or modified by contributors with more sinister motives. There have been some widely publicised cases of entries that have contained malicious or defamatory content.
And there is a persistent background hum of criticism about the accuracy or reliability of the enterprise, emanating either from aggrieved rivals (like Encyclopaedia Britannica) or more detached sceptics.
Truth and accuracy are contentious concepts, and objective measures of the reliability of encyclopedias are surprisingly difficult to come by. In 2005, the journal Nature published a survey comparing 42 entries on scientific topics in Wikipedia with their counterparts in Encyclopaedia Britannica, and concluded that Wikipedia had four errors for every three of Britannica's. (Cheekily, Wikipedia has an entry for 'Errors in the Encyclopaedia Britannica that have been corrected in Wikipedia'.)
Scepticism about Wikipedia's quality control has led Larry Sanger, who with Jimmy Wales was the co-founder of Wikipedia, to spin off his own version of an online collaborative encyclopedia. He calls it Citizendium and aims to improve on the Wikipedia model by requiring all contributors to contribute under their real names, by moderating the project for unprofessional behaviour, and by providing what Sanger calls 'gentle expert oversight' of everyday contributors.
A main feature of Citizendium is its 'approved articles', which have undergone a form of peer-review by accredited topic-experts and are closed to real-time editing. The only problem is that this all takes time and effort, which is why Citizendium is currently relatively tiny compared to Wikipedia.
The arguments about Wikipedia will continue to provide innocent amusement for decades to come. In the meantime, shouldn't someone be asking why newspaper obituarists choose to rely on a single source for a factual claim?