I discussed this elsewhere, and someone else brought the subject up, but it got lost in the noise.

One of the fundamental problems of the Wikipedian editing model about people, though this extends further, is that the majority of editors are locked into using the Internet as its main resource.

This leads to a few biases:

* News driven agenda for articles. People are defined by their newsworthy events, rather than a rounded view of their life.

* News articles are of the time and there may not be a newsworthy need to re-visit the subject after the event (this is happening, we think because of... vs that happened and we now know that...).

* Important topics are discussed in terms of the newsworthy elements which tend to be the controversy of the subject. Again, a well rounded article, though acknowledging the elements of public interest, needs to take a rounded view of the subject matter.

* Qualified editors who have access to specialised sources may have their sourcing excluded as it is not generally available, therefore not verifiable.

* The Internet represents a dubious subset of all possible sources: potentially missing sound, copyrighted printed sources of high value; sources written to push an agenda; poor evaluation of the quality of sources: if a quality name can be introduced into the sourcing, little account is given to the type of source (e.g. WP:Verification looks to mainstream newspapers, without explaining the myriad of qualifications that need to be considered).

What this type of problem tells me is that the current Wikipedia can never expect to produce a reliable article because there is no process to address this, it is simply illogical to claim that a system not designed to produce the ideal article can magically be expected so to do by the simple application of many pairs of hands of people who have no specific knowledge of the subject, and who will only gain a consensus of a subject based on reviewing the common flawed sources of the Internet?

The main process that a real encyclopedia would use to solve these issues is the skill of the authors (a word that comes to mind is TRUSTWORTHY), but the Wikipedian model, especially through policy of NOR, denies that it is appropriate for editors to apply reasoning in their editing process - this being identified as original research or synthesis (albeit that the original reasoning for NOR was sound, the impact of extending this concept into an all-embracing solution to improper editing is flawed).

All of the points you raise are good, and I'm sorry they got lost in the noise. I'll comment on one: yes, I ran into precisely the problem that a reference I made to a book printed in 1905, was not acceptable in WP. No one could find the book in a ‘library’. What’s that?

There is a related point, that perhaps deserves a separate thread, namely that the Wikipedia model is somehow about ‘IT’ or ‘technology’. People are in awe of it because of that. But it’s nothing really to do with IT, which is just a delivery mechanism.

The key factors in building an encyclopedia are no different than they were before the ‘information revolution’. You need people who can write English well, and who can grasp the difficult skill of key point summarisation and thread. You need (though this is missing from WP) a process of selection – both on the basis of the skills just mentioned and on personal relationships. In real companies, this is achieved by exams or tests (for the skills) and by interviewing people (to ensure they will fit in with the corporate culture). You need other things such as a referencing system, house style, policies and so on (some of which WP has). But none of this is anything to do with ‘technology’.

A further irony. Most of the articles in WP that are any good in my subject area are those cut and pasted from the Catholic encyclopedia, Britannica 1911 or other outdated sources. Thus, the white-hot technological revolution of the 21st century largely consists of delivering what people thought about in the early 20th century.

It goes to the problem of the "No Original Research" issue.

Apparently, if the information is culled from a reference book and not from someone's MySpace page, then it's "Original Research" and is disallowed. If the book happens to be out of print, then people cannot quote it because the teenage admins don't know what "book", "library" or "research" mean - such terms having been deleted from the 11th Edition of the Newspeak dictionary and replaced by one word: Google

I ran into this problem when I sought to improve the sourcing of content appearing in the article Icons of Evolution. The same problem arose in the BLP of David Berlinski.

Actually, there's a further problem, and that is that the higher the difficulty of finding sources physically--- how many stacks you have to search for misplaced tomes and microfiched newspaper cards that you have to blow the dust off, after driving a hundred miles to the library that promises them--- the more difficult it becomes to defend Wikipedia's BLP policy as "merely indexing knowledge about persons which is available to anybody by other means, anyway."

The time that it takes to do actual research, as opposed to manic quick-draw googling, pretty much defeats the purposes of the Wikipedia MMORPG, and would limit it to being merely an online encyclopedia.

I'd like to add another feature to DB's catalogue of systemic bias, and that is the tendency to treat US and Commonwealth press sources as if they were the Revealed Truth, while treating Russian, Asian or Third World press sources with derision. There is a chauvinist/racist aroma coming off this.

Russian? I think that anything that's not in English is considered suspect as a source by the majority of WP Cabal members. Since the majority of the sources are from English-language sources, the bias becomes not only evident, but glaring.

Professor Sharman Lichtenstein of Australia's Deakin University, who was recently in the news, has published a 7-page conference paper, The Wikipedia: Experts, Expertise and Ethical Challenges which she presented last February at the Fifth Conference of the Australian Institute for Computer Ethics (AICE-2008).

The paper is not yet posted online, but it may be of interest to at least some readers of this thread.


There are 32 references in the paper, six of which are available online.

Ethics are only a concern for people with morals. Wikipedia ain't got none.

In fact, its a severe handicap to edit Wikipedia if you have integrity, because you'll sooner or later fall foul of the many who don't - then the fun begins.

It took only 9 days for Filll to start the fun for me.

He must have been on vacation.

He had to round up a posse big enough to tackle me.

I've found that the bias extends to English-language publications that come from the wrong parts of the world.

Its not so much chauvinism as unfamiliarity with dealing with multiple sources for the same subject - each with a different view and different strengths and weaknesses.

In other words, Wikipedia is dominated by people who can't research - they instead rely on Google rankings and familiarity with a limited number of Western periodicals which they've heard of because their parents used to read them at breakfast.

The strictures expressed in WP:NOR (including WP:SYNTH) tends to weed out of WP any would-be encyclopedists who are experienced in researching primary and secondary sources and synthesizing a coherent treatment of a subject, suitable for an encyclopedia entry.