The name of the charity and its registration number if known

Wikimedia UK, registered Charity no. 1144513

Nature of complaint

Serious misrepresentation in its application to UKCC (letters 26 July and 26 September 2011).

Concern

Serious non-compliance in a charity which, left unchecked, could damage public trust and confidence in the Charity Commission as an effective regulator

Who is involved.

Wikimedia UK Board

Action taken by the charity because of what has happened

Complaint and emails ignored.  I was banned from their site after persistent questions.

Publicity

Not so far.

Documentary evidence in support of your concerns

See document attached.

Details of attempts you have made to get the charity to address your concerns

Posts on the charity website, ignored, followed by a 'block'; posts to their mailing list, ignored.

Your name, postal and email addresses and telephone number.

<redacted>

Connection (if any) to the charity

None.

Previous correspondence with UKCC

Freedom of Information access request

 


 

 

 

 

 

 

WIKIMEDIA UK'S MISREPRESENTATION TO THE UK CHARITY

 

COMMISSION

 

 

 

EDWARD BUCKNER

 

 

 

JANUARY 11 2012

 

 

 

 

 


 

Summary

 

In this document, I question the accuracy of a submission recently made by Wikimedia UK to the UK Charity Commission. In the document, it was claimed that Wikipedia's system of controls have 'very considerably' improved the accuracy of information; that users now cannot add unsourced defamatory or malicious material to biographies of living persons (BLPs) without being rapidly removed; that there are effective controls over notability and conflict of interest.  A number of case studies were cited, apparently suggesting that the controls were effective, and that it was difficult to add poor quality content to Wikipedia.

 

·        In the section " Neutrality, Verifiability and "No original research", I argue that the core policies dating almost from the beginning of Wikipedia have been entirely ineffective in ensuring a neutral, encyclopedic point of view.  I cite two important examples of control failure.

·        In the section "Controls over malicious biographies" I question WMUK's suggestion that defamatory or malicious material cannot now be added to the biography of a living person (BLP) without either very reliable sourcing or being rapidly removed.  I give a number of examples of recent BLP abuse, and I discuss its root cause, that only changes to an article are examined, and cursorily at that, rather than the whole article being regularly reviewed for error. 

·        In Conflict of interest, I comment on the Bell Pottinger affair, where a journalist discovered a PR agency adding biased and misleading information to Wikipedia.  This problem was not picked up by Wikipedia itself.

·        In Notability policy, I discuss how Wikipedia gives extensive coverage of the non-notable ('My Little Pony'), at the expense of the encyclopedic and educational.  I suggest that the underlying problem is the demographic of Wikipedia (young males with expertise in computers).

·        In Recent changes patrol I further discuss, with examples, the problem that Wikipedia reviewers give cursory attention only to changes to articles, rather than the whole article.

·        In Studies on quality I show how the evidence of Wikipedia quality studies is biased and misleading.  Biased, because it does not mention any of the negative studies, and misleading, because one of the studies demonstrated a serious problem with Wikipedia quality, which the submission did not even mention.

·        In Checkuser I question the effectiveness of Wikipedia's system of detecting 'sockpuppets'.

·        In Other, I question the figures about the number of users, and the number of administrators, mentioned in the submission.  These figures are blatantly incorrect.

 

The misrepresentation is serious.  If failure to comply is left unaddressed, the issue could damage public trust and confidence in the Charity Commission as an effective regulator.

 

 


Background

As an educator[1] I have been concerned for some time about the quality and accuracy of the internet encyclopedia Wikipedia.  Nearly every Google search on a subject returns the Wikipedia article in or near top place, and it is often the first place that the general public go to for information.  It is therefore in the public interest that there are strong controls over the quality and accuracy of Wikipedia articles.  However, Wikipedia's 'crowdsourcing' model, which allows anyone to edit, means that such controls are inherently difficult to apply. 

 

Wikimedia UK recently made a submission to the UK Charity Commission[2], making a number of statements about Wikipedia's system of controls, claiming that these have improved the accuracy of information.  The claims include:

 

·        that the accuracy of information has improved very considerably in the 10 years of Wikipedia’s existence (13.1.1, 13.1.3)

·        that a range of measures introduced since around 2005 have made it easier to remove vandalism, and the continual process of improvement of articles has removed or improved much inaccurate material (13.1.1)[3].

·        that defamatory or malicious material added to biographies of living persons (BLPs) cannot now be added to the biography of a living person without either very reliable sourcing or being rapidly removed (13.1.14).

 

The submission also includes (section 13.3) a list of 'comprehensive controls' over accuracy and quality that supposedly underpin these improvements.

 

These claims are grossly misleading and overstated, and are almost entirely unsupported by the evidence provided in the paper, as follows.

 


Neutrality, Verifiability and "No original research"

 

The Submission begins (section 13.3.1) by mentioning the core policies of Neutrality, Verifiability and No Original Research. 

 

These were introduced early in Wikipedia's history.  The first was 'Neutral Point of View' (NPOV), completed in December 2001 by one of the two co-founders of Wikipedia, Dr Larry Sanger, based on an earlier version written in November 2000 for the Wikipedia's predecessor Nupedia.  Dr Sanger famously left Wikipedia in 2002 after disagreements over experts not being involved in the project, and is still one of Wikipedia's severest critics.  He has published many papers about the problems inherent in the crowdsourcing model, including "The fate of expertise after Wikipedia", published in the philosophy journal Episteme in 2009.

 

The NPOV policy failed because of the difficulty of defining what a 'viewpoint' means.  That the earth is round is as much of a viewpoint as the view that it is flat.  Consequently, a further policy 'Verifiability' was drafted in August 2003 by Martin Harper.  This essentially dictated that Wikipedia accept the 'viewpoint' of the academic and scientific establishment, requiring that all claims must be supported by so-called 'reliable sources', and that all potentially dubious claims must be cited or 'supported' by such sources.  A third policy 'No Original Research', which is essentially an extension of verifiability, was added shortly afterwards in December 2003.  This policy distinguished between primary, secondary and tertiary sources, and discouraged the use of primary sources.

 

Secondary or tertiary sources are needed to establish the topic's notability and to avoid novel interpretations of primary sources, though primary sources are permitted if used carefully.

 

These policies were widely seen as a failure.  This was due to misunderstanding or ignorance of the policies, and more generally due to the lack of will to challenge often powerful crank groups, cults, political activists etc.  For example, Plane (esotericism), an article about the occult, was begun in October 2004 and reached approximately its present form in September 2005. It is in blatant violation of the neutrality policies, and was flagged in August 2009 as requiring citation and as containing original research.  More than two years later the flags remain. 

 

This is just one example - nearly 8,000 pages are flagged with the 'original research' template.  This also illustrates a common problem with Wikipedia, that even when errors are identified, it is easier to spot a problem than to fix it. Wikipedia editors rarely possess the skill or knowledge to do this. When there is a really serious dispute, it will often be resolved by a “consensus” of editors, namely a clear majority of those who happen to turn up and contribute.  These editors may be ill-informed or biased, and may include sockpuppets or friends solicited by e-mail, so the “consensus” is worthless.

 

I will give two further examples of how the neutrality policies have failed. 

 

Case study 1: Jagged 85

The first is the case of an editor called 'Jagged 85'. Jagged 85 began editing Wikipedia on 1 October 2005. His (or her) user page is crammed with 'barnstars', awards given by other Wikipedia editors for 'good contributions' to the Wikipedia project.  There is a barnstar for "outstanding contributions to Indian mathematics and science articles". There is another "for your outstanding work on medieval warfare in the Middle East", given by Kirill Lokshin, a member of Wikipedia's important governing committee, the 'Arbcom'.  There is even a barnstar for Jagged 85's work on the Islamic astronomy article which led to it being made a Wikipedia 'Good Article'. 

 

However, it was discovered much later that Jagged 85 had been systematically falsifying material in Wikipedia ever since he joined. His edits had a clear and consistent anti-Western agenda, systematically distorting source material in a way that untruthfully promoted Islamic (and also other non-Western) intellectual achievements, usually by claiming that a scientific development or invention or discovery was made or anticipated by some non-Western philosopher or scientist.  This was noted in a Spectator article in October 2009 by science historian James Hannan, who argued that a campaign was being waged in Wikipedia to persuade us that Islam’s role in the rise of modern science was a more significant than previously realised by orthodox historians.

 

The issue was taken in April 2010 to a Wikipedia request for comment page, when the scale of the falsification became evident. The editor had contributed to 8,115 pages, making 63,298 edits, nearly all of which were biased or falsified.  Jagged 85 had easily circumvented the requirement for citations.  Investigation revealed either blatant misrepresentation of the source, or a selective interpretation that went far beyond the meaning of the source. He also made frequent use of scholarly works not available on the internet, probably relying on the fact that most of Wikipedia's editors are amateurs who have no access to a university library.

 

Even the Good Article "Astronomy in medieval Islam" mentioned above was later found to have misused sources, and most of it pared down to a tiny stub on 28 April 2011 ("This article has been shortened from a longer article which misused sources. Details of the earlier versions may be found in the page's history. Please help us to rebuild the article properly.").  This clearly suggests that ordinary Wikipedians, who awarded 'barnstars' to the author of the original article, were unable to detect the failure of the Wikipedia policies on neutrality and verifiability and reliable sourcing.

 

Typical of Jagged 85's technique is this edit where he claims that "Avicenna developed an early theory of impetus, which he referred to as being proportional to weight times velocity, which was similar to the modern theory of momentum" citing Aydin Sayili (1987), "Ibn Sīnā and Buridan on the Motion of the Projectile", Annals of the New York Academy of Sciences 500. Yet the source attributes the theory to the fourteenth-century French philosopher Buridan, not Avicenna.

 

The problem still affects Wikipedia, as the cleanup page shows.  Investigators have checked and, where necessary, removed edits from the pages where he was most active, but many affected pages remain.  Yet Jagged 85 continues to edit Wikipedia, although he avoids medieval history and Islamic history articles.  Ironically, he was even asked to help with the 2011 Wikipedia fund-raising campaign as a 'story teller', recounting his work on, and contributions to Wikipedia.

 

Of course, this case was identified, and may eventually be resolved.  But the point is that this was not down to the neutrality policies themselves, as they are so easily circumvented.  In order for these policies to work, editors with the ability to make judgments about the quality of sources are needed, and they are in short supply.  One of them (user Syncategoremata) left in July 2010, shortly after the investigation began.  More of them are leaving, particularly in the wake of the bitter internal battle between the so-called 'content contributors', who are not officially recognised as a body within Wikipedia, and the 'administrators' who are the police of Wikipedia, and who are generally unskilled and lack the ability to make expert judgments.  The dispute was even escalated to the Arbitration Committee, but even this body is not qualified to resolve technical issues of content and sourcing.

 

Case study 2: Haiduc

Although the Jagged 85 problem is likely to be resolved, there are others which have not, including the contributions of user Haiduc.  Haiduc is a Romanian word for 'rebel'.  Haiduc's real identity is unknown: like many editors, particularly those who promote idiosyncratic or eccentric 'points of view', he has remained anonymous.  His editing was exclusively focused on what he called 'Pederasty', meaning a sexual relationship between an older man and an adolescent boy, usually in puberty.  He claimed to be writing a book on pederasty in which he promoted the view that heterosexuality was regarded as 'normal' only in relatively modern times, resulting from Christian propaganda and moralising, that pederasty was the form that homosexuality took in pre-Christian culture, and was widely regarded as 'normal' in pre-Christian times.  The view is derivative of a theory promoted by gay activist and NAMBLA apologist William Percy III that has limited acceptance within the academic community, and is frequently used as an apologetic for pedophilia or 'man-boy love'.

 

Haiduc promoted his view across many articles in Wikipedia, most of which remain in their original form.  See, for example, Pederasty in ancient Greece, Athenian pederasty, Cretan pederasty etc.   His views aroused support from the many pedophile activists who were then editing Wikipedia, and provoked strong reactions from others in the community.  One pedophile activist[4] remarked that "those who compose articles for the resource are a natural target for the sex-abuse kooks who see pedophiles under every rock and bush".

 

The issue was eventually resolved by automatically banning anyone who argued for or advocated pedophilia on Wikipedia.  Most of the activists were banned in 2007 or early 2008, following an initiative by the Arbitration Committee.  Haiduc, who argued he was not advocating pedophilia, but 'pederasty', was not banned until February 2010.  Little or no effort was expended in clearing up or correcting the many articles he wrote.  Nor was the ban the result of the enforcement of the neutrality policies, on account of the expert judgment required to enforce them.  To an inexpert observer Haiduc's citations always seemed to support his claims, although they were of questionable relevance, and Haiduc himself had a level of education and ability that was difficult for the average Wikipedia editor to challenge.

 

Conclusion

I have given two examples of significant known problems with the neutrality.  There are others we know about which we must omit here for sake of brevity, and there are almost certainly many others which we do not know about at all.

 

Controls over malicious biographies

 

In section 13.1.2 it says “Accuracy is especially vital in the biographies of living persons, which Wikipedia could be said to have been slow to recognise before the John Seigenthaler incident in [May] 2005”, which strongly implies that Wikipedia recognised that accuracy is important after 2005. Yet in section 13.1.2 it mentions Taner Akcam, a Turkish historian who was wrongfully detained in February 2007 after Canadian authorities referred to an inaccurate version of his biography on Wikipedia that described him a terrorist.  Clearly the problems had not been addressed by 2007, nearly two years after Seigenthaler.  It misleadingly adds (13.1.4) that Akcam’s biography is now indefinitely "semi-protected" so that only established registered users can edit it.  However, this is true of only a small number of BLPs, with the majority remaining unprotected and open to abusive and malicious editing from ordinary IP addresses, and in any case malicious editing can still occur from registered accounts. Note that anyone can create an account within seconds on Wikipedia, and there are few checks on the use of multiple accounts, i.e. the same user registering dozens, possibly even hundreds of ‘sockpuppet’ accounts.  (For controls over such 'sockpuppet' accounts, see the checkuser section below.)

 

Finally, it states that WMUK believe that such [defamatory or malicious] material cannot now be added to the biography of a living person without either very reliable sourcing or being rapidly removed (13.1.14).  This is grossly inaccurate. Malicious editing to BLPs occurs all the time, and there are many known instances of recent abuse.  For example,

 

The article MILF[5] had a picture of Pamela Anderson added in December 2010, with the caption "Pamela Anderson in 2003, sometimes labelled as MILF by gossip websites", which was not removed until six months later despite at least two of the site’s administrators and a member of their arbitration committee editing the page, during that period. Note that one of them was adding categories in order to embarrass an American politician (Rick Santorum) by increasing the Google rating of a supposed neologism.  On 9 December 2011 an allegation about Lord Bell that had been there since November 2005 was removed.  This was prompted only because of Wikipedia's discovery that Bell's PR agency had been editing Wikipedia for years (see the Conflict of Interest section about that).

 

On April 2 2011 a Wikipedia editor added libellous information to the article on Vijay Bahadur Singh, in order to test the effectiveness of Wikipedia controls on biographies. On April 4 he added a further libel. For both edits he added a reference, to circumvent controls on unreferenced information. A quick check would have shown that the references were bogus, but this did not happen, even though "(Tag: possible BLP issue or vandalism)" appeared as an edit summary after the first edit (as we note above, it is rare for citations to be checked by patrollers). The editor then created another account to remove the libel, but when he effected his change to repair the encyclopedia, alarmingly his action was reverted by another Wikipedian as vandalism.  That is, Wikipedia now favored the malicious edit and didn't want it removed!

 

After removing the content a couple more times, (once with the edit summary "remove lies") the account was blocked for edit warring. At no time did the blocking admin perform a basic check to see if the information was indeed false, and so the admin restored the libel. The problem was only spotted when the admin read the thread about the test on the Wikipedia watchdog site Wikipedia Review. This suggests a systemic failure of control over unsourced and libellous information added to biographies. The editor claimed he added libellous content to a further 50 BLPs, all of which remained unreverted and showed up in Google searches.

 

A ‘biography’ of teacher Brian Read was created in December 2010, possibly by his students. It contained completely false and libellous information. It was brought to light in a Wikipedia Review thread on Fri 18th March 2011, 8:00am, and deleted by administrator and Arbitration Committee member Ira Matetsky 12:15 on the same day, noting “It is dismaying that this article was overlooked in our internal patrolling and had to be brought to our attention offsite by a bitter critic of the project”[6].

 

Glenn Beck is an American conservative radio host and political commentator whose followers love to fill the article of anyone mentioned by Beck with out-of-context quotes and Beck's accusations. Single-purpose accounts work in teams together with established Beck-promoting accounts, and various one-off accounts and IP addresses.

 

The article on Tahir Abbas (deleted as recently as 25th December 2011) was a BLP that contained personally damaging information, based on a single secondary source from two years ago that the publisher recently removed from their website following lengthy legal action by the subject. Registered editors at the article insisted (November 2011) that without a formally published retraction, the original article, although no longer available online, is still a reliable source, and inclusion of the material is fully in line with BLP policies. The BLP subject was marginally notable, and has asked, pleaded even, to have his biography deleted from Wikipedia, because for the past two years it has been used to hound him, and has literally ruined his life.

 

For those two years, the article has been the top search result for anyone Googling his name, with the smear representing more than 25% of the entire article content. The article was not deleted without bitter internal wrangling at Wikipedia.  It is suspected that the Wikipedia editors concerned, who were registered under anonymous user names, were academics who had sided with the Head of Department and against Abbas, and were bringing a bitter academic dispute to Wikipedia, trying to end Abbas’s career.

 

In 2009, an anonymous account edited the Wikipedia entry of the art dealer Philip Mould, questioning the importance of “discoveries” and suggesting other dealers had made far more important finds, and claiming that Mould was cheating on his wife. Although unsourced, the information was not removed by other editors editing the relevant section. Then, in October 2009, the same person sent a “press release” to national newspapers, falsely claiming Mould was having an affair with Charlotte Barton, a 42-year-old artist.

 

Although these were complete fabrications Mould could not remove the entries, because the slanderous allegations were now in the tabloid press, and Wikipedia could now substantiate the same unsourced allegations with 'reliable sources'  (all national newspapers, even tabloids, being reliable sources for every article in them).  Mould was forced into an 'edit war' with Wikipedian administrators who were determined to defend the source.  At one point, Mould was disallowed from removing the allegations because he had a ‘conflict of interest’, being the subject of the article.  This was even when the article read 'The couple separated in May 2009, after Mould started an affair with notorious marriage wrecker Charlotte "Charlie" Barton, although an administrator did allow a change from 'notorious marriage wrecker' to 'artist'.  The allegation stayed there for a year and a half, until May 2011 when a senior administrator removed it, commenting "Section removed. There is evidence of a deliberate plot to defame the subject of this article. For those investigating this misuse of Wikipedia, the content formerly here can be found in the page history".". (The poison-pen attack, and Wikipedia’s role in it, was eventually reported in the Daily Telegraph.)

 

In December 2011, Rita M Gross, a noted feminist scholar, noticed that someone had added incorrect and improperly sourced information about her traffic citations to her biography some time prior (it turned out that the anonymous edits had been made by someone using her own university’s computer system). When she deleted the libellous information, in line with Wikipedia policy, the libel was several times restored by experienced Wikipedians, and she was warned on her user talk page for editing with a conflict of interest.

 

Malicious edits to Clare Brown’s article (sister of Gordon Brown), by PR agency Bell Pottinger, are covered in the section ‘Conflict of Interest’.

 

Breaking news: as this document was being finalised (January 2012), the Wikipedia administration, including Jimmy Wales himself, have been removing unsourced allegations about the academic Fabio Barraclough.  This included a claim, added nearly three years ago, that he was a spy and money launderer.  The claim was even featured on Wikipedia's Main Page in the "Did you know?" column on 26 January 2009, viewed approximately 4114 times. Once  again, the allegations were not spotted by the 'change patrollers', but rather by the victim contacting Wales himself.

 

Conclusion

Problems with living person biographies are persistent and have not, as WMUK suggest, been cleared up in the wake of the Seigenthaler incident of 2005.  The problem is not just crude vandalism, but rather malicious and agenda-driven editing from within Wikipedia, and a failure to recognise such editing.

 

Conflict of interest

 

The extract from the solicitor's email dated 26 July 2011 says (4.7.4) "Conflicts of Interests behavioural guidelines" which set out generally accepted standards that editors should follow, when addressing an incompatibility between the aim of the Wikimedia projects and the aims of an individual editor.

 

These guidelines are widely ignored.  There is a natural conflict of interest in Wikipedia, given that no qualifications or credentials or personal identification are required to edit (or even become an administrator), and given that Google gives priority to any subject mentioned in Wikipedia.  Few people edit Wikipedia for the altruistic dissemination of knowledge, and their reason is more likely to be one of the following reasons: (1) promoting positive information, and suppressing negative information about companies or individuals, often by paid editors or employees of the subject; (2) legitimising or normalising deviancy such as pedophilia or bestiality; (3) publishing information about celebrities, games or shows of which they are fans; (4) promoting quack cures, bogus medicine or psychotherapy; (5) legitimising and promoting the activities of cults; (6) supporting and promoting political interest groups, often on the fringe; (7) publishing theories or systems that have been rejected by the academic establishment; (8) writing about themselves, friends or family in order to see their names come first on a Google search.

 

Thus any restaurant, high school band, obscure author, or crackpot theory automatically gets visibility and publicity by being on Wikipedia[7].  

 

This is too detailed and wide a subject to be treated in any comprehensive way.  Here are some examples. Many members of the Wikipedia administration have articles about themselves, their friends or family. Bogus and 'new age' cures are widely promoted on the encyclopedia, see for example the article on Ken Wilber, a new age 'philosopher' who gets comprehensive coverage on Wikipedia, despite there being no academic recognition of his work.  There is a whole category devoted to his supporters.

 

Nor are there are any effective controls on paid editing.  The most well-known incident is the Bell Pottinger affair, uncovered as recently as December 2011.  Bell Pottinger is one of Britain's largest lobbying agencies, who used a sockpuppet farm to "sort out" Wikipedia coverage of their clients by adding positive information, and removing negative information. The farm was not uncovered by Wikipedia’s systems and controls, but rather during an undercover investigation by the Bureau of Investigative Journalism who were probing PR abuse generally. Reporters from the Bureau posed as agents for various clients, including the government of Uzbekistan, which has a poor human rights record, and its cotton industry, which uses child labour, to investigate the ethical standards of the agency. Ten firms were contacted, of which five were prepared to work for the 'Uzbekistan' agents. The discussions were secretly recorded.

 

The agency claimed that it has all sorts of 'dark arts', which could not be included in a presentation because it would have been 'embarrassing' if revealed. This included creating fake blogs which appeared to be independent, but which were in reality biased, and would rank high in Google searches. Another 'dark art' was the team of Wikipedia sockpuppets, who removed damaging allegations about Bell Pottinger clients. For example, edits made by the account Biggleswiki included removal of a reference to the university drugs conviction of a businessman who was a client of Bell Pottinger, editing of material relating to the arrest of a man accused of commercial bribery, editing of the entries for prostate cancer expert Professor Roger Kirby and his firm, The Prostate Centre, both of whom are clients of Bell Pottinger, and editing of the articles of both Chime Communications - parent company of Bell Pottinger – and Naked Eye Research after the former company bought 55% of the latter.

 

One of the targets of 'Biggleswiki' was Clare Brown, Gordon Brown's sister. Brown is a campaigner against the destruction of rainforests in the Malaysian state of Sarawak. In this edit, for example, a 'criticism' section is added to Brown's Wikipedia biography, saying that her radio station 'Radio Free Sarawak' was operating illegally, in breach of the Communications and Multimedia Act 1998, and that it was inciting racial hatred in Sarawak.

 

The failure of Wikipedia’s system to pick up this abuse is all the more surprising given the unsophisticated approach used by the agency.  When I consulted other paid editors, they told me that "Bell Pottinger don't know the rules of the game".  First, they edited through their usual office system. These tend to have fixed IP addresses, so that several accounts have the same or similar addresses.  Thus, alternative accounts can be easily spotted by the Wikipedia checkuser system. Also, once the IP address is known, this can be used to identify the firm itself. Second, the socks consistently worked on client articles, so that the conflict of interest is more likely to be spotted. "Don't spend all your time on this sort of work. If half your time is writing apparently good stuff about Nepal or the First Duke of Marlborough (or, better, both) you'll be less noticeable. I see accounts that spend all their time laundering, and they're fairly obvious". Third, the sock team failed to diversify onto other WMF sites, making them look obsessive.  Yet even an easy target such as this was missed.

 

As implied by the remarks above, more sophisticated paid editors are operating undiscovered on the project. One even blogs about his work.  Their techniques include the ones mentioned above, the use of ‘hot spots’ (wifi networks) to conceal IP addresses, or the use of IP proxies.  Other ways of evading detection are the use of ‘dynamic IPs', as provided by ISPs like btinternet. Dynamic IPs change daily or hourly, and make abusive editing almost impossible to detect.  Wikipedia uses programs to detect the identity of the internet browser, but these are easily bypassed.

 

A basic control on abusive accounts would be a requirement of personal identification to the Wikimedia Foundation through a separate and independent office or service. However, this has always been unacceptable to the Wikipedian community, who prefer – for whatever reason – the freedom to use anonymous and multiple accounts.

 

Breaking news:  Wales is now putting together schedule of visits and calls to major PR firms and to people who he can identify "as paying for inappropriate behavior at Wikipedia".  This is proof in itself that the controls by Wikipedia itself are not, and never have been, efficient.   Wikipedia was founded on the assumption that the good will drive out the bad, and that conflict of interest will never arise because the community of editors will generally act in good faith. This is clearly unrealistic.

 

Conclusion

Conflict of interest policies on Wikipedia are widely ignored and abused. Conflict of interest is inherent in the way Wikipedia is set up, namely as a volunteer organisation which anyone can join under an anonymous and virtually undetectable account.  A recent and well-known abuse by a PR firm (Bell Pottinger) was never picked up by Wikipedia's controls, despite the firm's crude and unsophisticated technique.

 

Notability policy

 

Section 13.3.3 of the submission says that "Another important method of controlling content is the Notability policy, with its many sub-policies for particular areas.  This defines which subjects are appropriate for their own article and which are not, with sub-pages giving more detailed guidelines for areas such as academics, politicians, athletes, companies, films, books and even numbers".

 

This is the policy, but like the other policies on content, it is almost completely ineffective.   Most of the content on Wikipedia is biased towards banal and trivial subjects, without any encyclopedic or educational value.  The ephemeral List of My Little Pony characters is as large as 180k.  By contrast, the article on William of Ockham, the fourteenth century logician after whom 'Ockham's Razor' is named, and who is one of the five greatest philosophers England has ever produced, is only 24k.  It is full of errors, including inconsistent dates of birth[8].

 

Similarly the writer Ayn Rand, who received no formal training in philosophy and received only one (disparaging) mention in The Oxford Companion to Philosophy, received extensive coverage in Wikipedia, whereas the philosopher and expert on Bertrand Russell, Mark Sainsbury, who is one of England's greatest living philosophers, receives a poor one-paragraph article that has been flagged as having 'BLP concerns' for two years.  Likewise, the proportion of British university professors who are included is quite low, whereas the great majority of professional footballers, even in the lower divisions, are included.

 

This is probably explained by the demography of Wikipedia. Wikipedia and its administration is dominated by young computer-literate males, whose interests reflect the interests of that group as a whole, namely computer games, video games, Japanese animation, usually pornography. There are serious doubts about the ability of a group with this profile to produce genuinely neutral, balanced encyclopedic coverage appropriate to a project that aims to bring 'the sum of all human knowledge to every person on the planet'.

 

Conclusion

Wikipedia notability policies have failed to correct the bias towards the trivial and non-encyclopedic.  This is inherent in the demographic of Wikipedia.

 

Recent changes patrol

 

Section 13.1.10 of the submission alludes to the problem of 'recent changes'.

 

It seems Wikipedia corrects the vast majority of errors within minutes, but if they're not spotted within the first day the chances of them being corrected dwindle, as you're then relying on someone to spot the mistake while reading the article rather than reviewing the edits.”

 

This is certainly true, but the submission does not dwell on the serious consequences of this. In May 2005, someone claimed that Belgian businessman Georges Jacobs (born in the late 1940s) was a commander of the Waffen SS's 28th SS Volunteer Grenadier Division Wallonien (presumably disbanded in 1945). The claim was removed only recently (August 2010), and so had been there over 5 years without any patroller noticing.

 

On 26 April 2009, an account called ‘Ivanelo’ added some material to the Wikipedia article on ‘Climate of Antarctica’, including the claim that “Since mid 1960s, all major studies detect cooling over the most of Antarctica”. On 28 November 2011, climate expert William Connolley reverted the edit, with the comment “rm twaddle (how did that stand for so long)”.  His question was rhetorical.  As he commented here, " I presume you know: most articles are watched for changes, not regularly re-read from scratch. Once something sneaks in and is layered over, it can stay for a long time."  Connolley himself had worked on the article a few times without spotting the error.

 

The submission to UKCC completely ignores, and thus misrepresents, this control failure.  The only effective control would be a regular review of all articles (of which there are 3 million) to ensure accuracy and objectivity.  However, this would be impossible owing to the severe shortage of expert editors who would be able to make such judgments.

 

Note also the comments above on recent BLP abuse.

 

Conclusion

The 'recent change' method of patrolling Wikipedia is fundamentally flawed, as it is unable to detect the 'layering' and 'laundering' of biased or incorrect information.

 

Studies on quality

 

The submission to cites a number of studies that are supposed to confirm Wikipedia's quality controls.  The studies are either misrepresented, or flawed, or of restricted scope. Moreover, the citations are selective and biased in that they omit (or misrepresent) studies that highlight serious problems with quality.  It even omits one of the most important recent studies, which has provoked a storm of controversy.  In a series of papers, Harvard research follow Andreea Gorbatai has been exploring whether the collective methods of production used in Wikipedia are effective at creating social utility. In a paper first presented in 2010[9], she demonstrated that such collective production fails to respond to social need despite being designated as high value, and despite high 'demand'[10] for such products.

 

A follow-up to her study by the Wikipedia community concluded that the reward system on Wikipedia was perverse, with greater recognition being given to editors producing large numbers of low importance articles than to editors producing small numbers of high importance articles[11].  The follow-up study claimed there was a fundamental problem with the way that Wikipedia 'rewards' its contributors. 

 

The results are evident in the prevalence of [mushrooms and hurricanes].

 

The Nature study (cited in 13.1.8) has often been represented as a scientific study, though it was not peer-reviewed.  There is no mention in the submission of the well-known flaws in the study, such as the small sample size, the poor way the comparison articles were chosen and constructed, and the failure to quantify the degree of errors or the quality of writing. The most significant problem, as Wikipedia co-founder Larry Sanger[12] points out, is "that the comparison articles were all chosen from scientific topics. Wikipedia can be expected to excel in scientific and technical topics, simply because there is relatively little disagreement about the facts in these disciplines … One has to wonder what the results would have been if Nature had chosen 1,000 Britannica articles randomly, and then matched Wikipedia articles up with those."  Generally, non-scientific subjects are handled poorly in Wikipedia[13].  See e.g. the excruciatingly written article Civilisation, which still contains the sentence "The Inca civilization of the Andes did not use writing at all, yet it still functioned jolly well as a society", or practically any article on Philosophy[14], including the awful flagship article itself, which has been the subject of constant edit warring and fluctuation for the last 10 years, and which has barely improved since the original version in 2001.

 

The study by Magnus (cited in 13.1.9) is misleadingly represented.  Magnus is an associate professor of philosophy who inserted humorous, absurd and blatantly incorrect 'fibs' into Wikipedia articles towards the end of 2007. 18 of these fibs, including the claim that G.E. Moore's influence outside philosophy includes a reference to him in the signature line of the musical Oliver, and that after retiring, Gilbert Ryle bought a small farm and tinkered with automated processes to care for livestock, "although they never proved to be commercially viable" were never reverted[15]. A smaller number were reverted, largely due to the facts that (a) Magnus was editing from an 'IP' address, which are closely scrutinised (b) he made different edits from the same address, making it easy for vandal fighters to 'follow the trail' and revert all the edits, (c) he did not provide fake citations to support the claim – as we observed above, citations are rarely checked to ensure that they support sources, thus allowing Jagged 85 to edit for four years with nearly 64,000 edits before being found out.

 

In a number of cases, erroneous information on Wikipedia leaks out into the web or even published sources, and turns into citable information that can then be recycled as verifiable back to Wikipedia. In February 2011 a Wikipedia editor spotted the following strange edit made in November 2002, referring to a scientist’s claim that the Van Allen belt is the result of volcanic activity. This is nonsense. The editor Googled for this and got plenty of hits, and it nearly went unchallenged. However, suspecting that the age of the Wikipedia entry had caused this claim to become accepted fact, he emailed the scientist, who confirmed it was nonsense, and the edit was finally reverted, nearly 10 years later.

 

In another example of recycling misinformation, when an editor corrected the absurd claim that Sharaf al-Din al-Tusi was the first to discover the derivative of cubic polynomials, he was immediately challenged by another Wikipedia editor, who says that the claim is supported by the Encyclopedia of Ancient Egypt. However, the Encyclopedia of Ancient Egypt had itself mistakenly used Wikipedia as a source.

 

Yet another example was the case of art dealer Philip Mould (see above), where a malicious editor contacted the tabloid press with a fake story, then used the story as a 'reliable source' to insert malicious falsehoods about Mould.

 

Conclusion

The submission is biased in that it does not mention negative studies of Wikipedia quality, and it blatantly misrepresents one of the positive studies.

 

Checkuser

 

Section 13.3.8 of the submission says ""There is also “CheckUser” software that enables a small number of selected and vetted volunteers to establish, in many cases, whether two editors are from the same ISP [sic[16]], and often where that ISP is located.  This information is mostly used to detect blocked editors who try to return under a different account name. "

 

Checkuser is almost entirely ineffective for any editor who is determined to circumvent it.  It uses internet 'IP' addresses to identify users, however many users have 'dynamic' or 'floating' IP addresses (including most of btinternet customers).

 

There are many, many devices that allow for easy IP switching, such as Linux based routers, tethered smart phones, 3G to wifi local hotspots. There are many classes of device that allow you to change your IP immediately. By combining these and using different browsers or use a plug-in that allows for browser string spoofing, there is literally no way for the checkuser tools to find you[17].

 

Other claims

 

Other claims in the submission are 13.3.6, that "each registered editor (currently about 144,000 of these are active each month) has a “watchlist” of articles that they have previously edited or elected to watch (and not removed from their list)", and 13.3.7 that "There are currently over 1,500 administrators on the English Wikipedia.". 

 

Both these statistics are misleading, if not completely inaccurate.  The page for active editors for November 2011 shows only 34,516 such editors for en.wikipedia (where an 'active editor' is defined as one who has contributed more than 5 times in one month).  The number of very active Wikipedians (defined as Wikipedians who contributed 100 times or more in a month) is only about 3,400.  This seems insufficient to keep an eye on the estimated 600,000 biographies of living persons, even if there were a way to co-ordinate them to do this[18].  It is unlikely that a significant number of these will keep BLPs on their watchlist.  The only effective control on BLPs, as we have mentioned above, would be compulsory registration  for that category, or editing under real names, or any other control to ensure that editors of malicious or defamatory BLPs were directly accountable for their work. In 2010, people were told that they were going to introduce pre-screening ('flagging') of edits, but that proposal was rejected.

 

The figure of 1,500 administrators is also grossly misleading.  The page here, which is a list of all active[19] administrators, shows that there are fewer than 700 active administrators.  Nor does the submission mention the well-known (to Wikipedians) problem that many administrators are leaving the project, without being replaced.

 

 



[1] I used to teach philosophy at Bristol University, and have published papers on the Philosophy of language and the philosophy of mind. I am now retired from teaching, but continue to work on philosophy, and have recently completed a translation of and commentary on an early work by the 13th century philosopher-theologian Duns Scotus.

[2] Hereafter 'the submission'.  The submission was sent via WMUK solicitor’s e-mail of 26 September 2011.  WMUK have so far refused to provide the whole submission to me. However, UKCC supplied parts of it, namely sections 2 and 13, in a 'Freedom of Information' request made by Ian Spencer on 24 November 2011, supplied by Mark Edwards on 20 December, reference ME/1144513/C-340380/RTN CC:01721035 .

[3] Vandalism is defined as "any addition, removal, or change of content in a deliberate attempt to compromise the integrity of Wikipedia".

[4] Corax, since banned.

[5] See the article itself for an explanation of this offensive and abusive term.

[6] Wikipedia Review is a watchdog of Wikipedia which has publicised many quality problems with Wikipedia. It is regularly read (and acted upon) by the senior members of the Wikipedia administration, but has no official status on Wikipedia. On the contrary many members of the review have been punished for posting there, and the Review is generally hated by the administration of Wikipedia.

[7] Even the co-founder of Wikipedia, Jimmy Wales, is not above this – as is well-known in Wikipedia circles, he created an article Mzoli's, an unnotable shop in Gugulethu Cape Town.

[8] The introduction gives d.o.b. as 1288, but he is categorised under 1285 births.

[9] "Failures of Social Production: Evidence from Wikipedia." Paper presented at the Sunbelt Social Networks Conference, Sunbelt Series, Riva di Garda, Italy, June 2010.

[10] As measured by Wikipedia page views.

[11] "Part of the problem is over concentration on certain topics. 10% of the GA+ articles [i.e. articles equal or better to a Wikipedia 'Good Article'] come from four peculiar categories: tropical cyclones (hurricanes), mushrooms, trains, and US roads. What would an outside scholar think of this? Those four categories together account for only 3 of the 1000 VAs [Valued Articles] (0.3%)." He wondered whether these topics are a focus "because they are easier to mechanically write award winners on?".

[12] Who Says we Know, On the New Politics of Knowledge by Larry Sanger, The Edge 2007.

[13] There is too much to mention here, but see e.g. Intellectual history (“The concept of the intellectual is relatively recent”) almost entirely written by one person. Or how about the really awful History of Europe, an important subject and deserving of thoughtful treatment, but contains such gems as “During this time many Lords and Nobles ruled the church. The Monks of Cluny worked hard to establish a church where there were no Lords or Nobles ruling it. They succeeded.”.  (Version of 17 November 2011, since cleaned up as a result of my comments on a blog post).

[14] Version of November 2011. There are serious problems with the ‘most important philosophy articles’ in Wikipedia.  Nine of these articles are of ‘C’ class. The article on Metaphysics is a mess, containing gems like “Nihilism represents and [sic] extremely negative view of being”. The article ‘History of Philosophy’, which is not even ‘C’ class, is just plain weird (“an historical perspective … emphasizes the existence of a long period of transition between the teologically [sic] driven centuries (running up the XIII or XIV Centuries) and the rationalists-empiricists debates”; “early and late Renaissance philosophers were a more heterogeneous population, including rhetors, magicians and astrologues, early empirical scientist, poets, philologists”).

[15] Except by the hoaxster himself.

[16] 'IP' is clearly meant.

[17] Information was provided by an experienced sockpuppeteer, 31 Dec 2011, by email.  He said "In summation, checkuser is feeble and the people who use it and expect results are even worse. They only catch the most lazy and incompetent sockpuppeteers. Or those who don't give a sh-t."

[18] Which there isn’t, of course.

[19] 'Active' means having made 30 or more edits during the last two months.