Parents wishing the best for their offspring were once able to buy the hefty Encyclopaedia Britannica in WH Smith stores. It wasn’t cheap for the 28-volume set. With the best binding option the complete set was priced – for Christmas 1983 – at £3,000. That equates to more than £9,400 today. It only sells EB‘s yearbook now and is currently 35% off list price, at £47. The discount is applied, I guess, as it’s last year’s edition, which actually records events of 2013. In the age of 24/7 news, two-year-old recaps are ancient history.

EB and WH Smith are historic entities bestowed with superlatives. One is the oldest English-language dictionary of arts and sciences; the other is the world’s oldest chain store. Both developed revolutionary indexing systems: EB for categorising entries and Smith’s for creating the ISBN catalogue. I don’t know how factually accurate these points are, but Wikipedia says so.

It’s fair to attribute EB’s demise to Wikipedia. In its 1990 heyday EB generated annual revenue of $650 million, much of it by peddling knowledge door-to-door across America with a vaunted sales force. However the death warrant was really initiated by the CD-ROM. The publisher declined a digital deal with Microsoft in 1989, 12 years before Wikipedia was created. In 1995 EB was offered on disc, priced at $1,000. Microsoft’s Encarta, initially based on the rival Funk and Wagnalls encyclopaedia, was less than half that price. Encarta‘s price soon dropped to $99 and before long given away for free. The growing inconvenience of CD-ROMs, order of magnitude increases in download speeds and improved browser interfaces all helped accelerate online knowledge sourcing.

Under ownership changes and cost-cutting EB saw in the new millennium. Wikipedia didn’t appear until 2001. The viability of traditional encyclopaedia was then put under further strain. Still, it would be another decade before the online villain finally delivered the coup de grâce to EB’s print edition. It ceased print publication in 2012 when book revenue dwindled to just 1% of group income. EB still lives on, albeit solely in the digital dimension. It’s also been funked-up through rebranding, employing people who don’t wear tweed and even willing to poke fun about its fusty image.

In online access terms, Wikipedia is a wildly successful internet venture. It generates more than 430 million unique visitors monthly and credited as being the fifth most popular worldwide domain. Operated by San Francisco-based Wikimedia Foundation, the charity generated revenue of more than $50 million in its 2013-14 financial year. The site hosts more than 32 million Wiki articles in nearly 300 languages, far beyond EB‘s wildest goals.

More locally, Wikipedia‘s UK charitable arm raised almost £690,000 in its 2013-14 financial year. But Wikipedia is plagued by problems. Its heyday for contributing editors was almost a decade ago. In January 2005 it had around 12,500 active editors globally. That peaked in April 2007 with 90,000, and has declined broadly ever since, with some 73,000 active last March. This despite the honourable goal “of making the sum of human knowledge available to all humanity.” Part of its 2015 regeneration plan is to focus in future on the “Global South” for growth.

In Britain it’s also busy trying to improve the quality of its knowledge listings through partnerships. “Many of the UK’s cultural institutions have compatible aims with ours and a remit to make their collections available to all. In recent years the Wikimedia movement has been shifting priority from increasing quantity to increasing quality, and our work with GLAMs – [galleries, libraries, archives and museums] – is an important part of this,” its the charity said last year.

Universities are absent from its GLAM outreach. Instead, it’s developing a separate integrative education strategy. “In partnership with universities and support bodies, we are gradually establishing Wikimedia UK as part of the landscape of the British higher and further education sectors,” it added. Furthering that aim, the Wiki concept has gained ground within academic circles for its potential benefits in teaching, especially for distance learning.

However, there are inherent legacy issues it must deal with. Wikipedia’s easily editable entries – along with no need to register – has attracted an image of being the cyber Wild West. In October 2005 it was posting more than 1,500 new English-language articles a day, creating immense verification difficulties. The digital domain also creates demands of immediacy for editors. EB discovered this too after it dropped annual print edits for daily digital updates. But unlike EB, Wikipedia‘s can suffer from entry vandalism. Nevertheless a Nature study said Wikipedia‘s science topic accuracy almost matched that of EB. The claim was contested by EB, however it also conceded that the encyclopaedia was not without errors.

Collating the “sum of human knowledge” is a burdensome task. Wikimedia UK has admitted that apart from trustee turnover, key risks are “reversing editor decline and managing our reputation“.

That reputation was sullied significantly before the 2015 UK general election over claims of underhand enhancement and depreciation of political personalities. In particular, media attention focused on an account allegedly linked to Conservative Party chairman Grant Shapps, then a minister in the coalition government. This followed similar media claims several years ago of Wikipedia edits undertaken to boost Mr Shapps. New accusations were raised in this year’s electoral campaign, of rivals being undermined through another Wikipedia account. A Wikimedia administrator suspended the account, Contribsx, just weeks before the May vote.

A Fleet Street storm grew over the incident. Wikimedia employee Richard Symonds admitted past political affiliations but denied these motivated him to suspend the account supposedly linked to Mr Shapps. “I’ve never been a political activist. I have been an armchair member of the Lib Dems for maybe two years of the past five, but I don’t think it’s right to call me an activist,” Mr Symonds told The Guardian.

Nevertheless, some of his colleagues complained that although special software was used to check editing history of Contribsx, Mr Symonds had acted too hastily. “I stand by the decision I made, but some other volunteers are concerned that I could have been more clear in my reasoning and could have run it past other people first,” he added. “They may be wrong or right, but that doesn’t change the outcome or the facts of the case – I felt I had to do something to stop Wikipedia from being whitewashed.”

In June, Wikipedia’s audit subcommittee censured Mr Symonds as “no evidence” was discovered to link Contribsx to Mr Shapps. It also said Mr Symonds had an apparent conflict of interest. Much rested on Mr Symond’s usage of Wikipedia’s CheckUser software verification tool. The committee said: “There does not appear to be a major breach of policy. However, it could be argued that that tool was used to ‘exert political or social control’, which would be a violation of the [CheckUser] policy.”

In response to the findings, Mr Shapps said: “Wikipedia’s investigation has resulted in the strong disciplinary action now being taken … However, the failure of various media outlets to check even basic facts meant that this false and damaging story ran for an entire day during the general election campaign … My hope is that this case serves as a reminder that both the source, as well as the content of a story, should be carefully checked before it is broadcast in future.”

Politics can be a highly subjective area, particularly so when so-called collective intelligence drives knowledge collation. Counter-views can be rife. Last year, an assistant professor at Harvard Business School co-authored a comparative study of 3,918 articles about US politics published by both Wikipedia and EB. The study analysed bias and slant between “crowd” contributions versus the expert/editor process. It concluded that Wikipedia articles were more slanted towards Democrats than Republicans. They also held greater bias. Contentious topics, such as war, peace and foreign policy, also showed a much stronger bias. And yet through time as entries are repeatedly edited – up to 1,900 times for some – the Wikipedia bias wanes. Eventually, little separates the two sources.

The Shapps incident highlights a wider philosophical issue in the age of Big Data, when collectively produced intelligence can dominate over expert-based knowledge. Vertical disaggregation of central control merely shifts decision-making to an opaque communal command. Wikipedia says it is trying to be as transparent as possible. And the Harvard study raises further concerns of political affiliations being a driving force for where Web users go to seek knowledge. The issue of ultimate responsibility was a political puzzle for Jeremy Bentham, in the twilight of the age of Enlightenment. It was central to his conceptualisation of surveillance inside the Panopticon. Quis custodiet ipsos custodes?, he asked in 1787. Who watches the watchers?

Bentham may well have asked the same of EB, which was commencing initial publication of the third edition of the encyclopaedia in 1787. Scottish Episcopalian clergyman George Gleig completed the task after the death of its incumbent editor. By then the encyclopaedia was already considered a venerable source of Anglo-Saxon knowledge. And yet, under Gleig, the edition dismissed Newton’s concept of motion, instead preferring to assert the Classical belief that gravity is caused by elemental fire. Wikipedia told me that. Should I give trust to that claim? A footnote is cited, but nowhere in the reference publication is the Newtonian omission spelled out.

Pete Norman is Ben Pimlott Writer in Residence in Birkbeck’s Department of Politics.