This is the second time in as many months that Wnt (T-C-L-K-R-D) has said something useful, among all the inane waffle he usually spouts. The man is possessed of such eloquence and intelligence ... if only he knew how to use them.

Anyway, Wnt's story concerns the medical condition known as glucojasinogen, sourced to two scholarly papers by non-overlapping sets of authors, both of which appear to have copied a hoax sentence inserted by an IP in 2007 ...

Wikipedia, serving the advancement permanent destruction of knowledge.

Nice, very nice.

Excellent example of what can go wrong. "Breathing its own exhaust" is right.

The authors of that first paper have a bit of a plagiarism problem on their hands too.

Thanks, I didn't catch this one. Wnt presented it as a "joke", typical of Wiki fanboys.

I'll write it up.

(Wikipedia: other people make misteaks for us! But we never do!)

This reminds me of the whole "oh, well, Wikipedia isn't perfect, don't always trust it for information" stuff some Wiki apologists put out when cases like this arise. If your website is the #1 Google search result for just about every subject, when it claims to be an "encyclopedia," and when it indirectly discourages actual encyclopedias and their utilization by the average man, I think it's a high-priority issue to have content that is assured to be consistently, if not always correct, at least not patently false. Since it is not a high priority for Wikipedia the end result is vandalism of the more malicious kind (parading as true information especially via face-value 'legitimate' citations) actually begins making its way outside the internet.

The real point of all this is simple: they have no idea how much subtle vandalism they have, right now.
The WMF has not even shown interest in making a study, using a sample of articles, to make some
kind of vague inference of how big the problem is. They don't care. Meanwhile, an unknown quantity
of similar cases waits for them....

A good example of how untrustworthy Wikipedia is can be seen in the fact that most articles written by people who actually know about what they're writing on are filled with thousands of footnotes, since without them the case for reverting a false or incorrect statement is made much harder. Not to mention that it ironically opens the article up to being called "unreliable" or "unverifiable" to begin with. But of course Wikipedia allows for you to forge or make dishonest "sources" quite easily, thus undermining the supposed point of citations as a quality control thing.

I understand why extensive footnoting is done, it just demonstrates a fundamental flaw with Wikipedia. Real encyclopedia articles shouldn't really need footnotes, since you're only writing about established facts that can be checked via a bibliography. But Wikipedia isn't a real encyclopedia, it's an "encyclopedia anyone can edit," thus it is forever untrustworthy in the minds of all critically-thinking persons.

This this this. They literally have no idea who is writing their articles most of the time.... And again the same issue when it is corporations and political organisations working underhandly, like the Larouche party's deliberate corruption of anything they can lay their hands on...

I think the big problem with the LaRouche page is that it's just too damn big, and they even have this big page separated into other, also large pages ("Views of Lyndon LaRouche and the LaRouche movement," "LaRouche movement," "LaRouche criminal trials," and even "Lyndon LaRouche U.S. Presidential campaigns," although that's more an issue of "why does this warrant a separate page" than size) which creates problems since:

A. LaRouche is not a figure who has amassed many scholarly or otherwise notable analyses of his life, his work, or his political following;
B. ... that's basically it.

So really, the article should be at least half its original size. The longer it gets the more needless debates over him grow because the sheer length of an article alone assumes that it's really significant, whereas in fact the end result is two sides throwing random random not-very-encyclopedic sources at each other, e.g. newspapers on one side and LaRouche literature on the other, because there's really not much scholarly material to use to justify such a big page, let alone its almost equally large branched pages.

It is quite evident that there are many people slanting articles to their benefit or the benefit of their clients. We all know that here and to some degree accept it. I do not understand why we need to pick on the Larouche party. They may well be doing it, and are as at liberty to do so as anyone, no more but on the other hand no less. I do not believe for one minute that they are particularly big players in such a game. If anyone has evidence to the contrary, I would love to see it. Indeed, the Republican and Democrat parties are very probably doing it far more.

I am still blissfully unaware of who or what this LaRouche is, and intend to remain so, regardless of the number of pages dedicated to the phenomenon.

I think people focus on LaRouchists because it demonstrates that a tiny group, widely seen as a cult by those who know of them, can easily slant pages in their favor.

So can Democrats and Republicans, of course, but those are two major US political parties so it's expected and just another flaw Wikipedia has. The fact that tiny fringe groups can do it with such ease is a much bigger flaw. There's also a bit of perceived insidiousness involved, since, again, LaRouche is seen as a cult leader. It's seen as creepier than usual and an attempt to rope people into a cult.

I think they care, because being able to publicly say "we've done such and such to keep vandalism levels down" is a potentially big selling point for the WMF. But they can't say that so they want to shove it under the table.

The problem is, 1) a study would require numerous experts and hundreds of hours of fact-checking, especially if some really long articles were randomly selected, and 2) knowing how much vandalism there is does not help the WMF because they just want a quick and dirty way to get rid of it all. A study would only hurt them because it would almost certainly reveal that there is much more vandalism, and if not that then unencyclopedic fancrufty crap, then the average reader expects. And that will hurt donations.

There are at least a dozen such studies. And they've generally been getting better over time. I'm not sure anyone should trust the Foundation to do such work though, as the potential for conflicts of interest, intentional and otherwise, is too great. (Similarly to how they wouldn't look at admin attrition after saying they would for years, or after it was shown that it leads to editor attrition, and then getting really upset when someone finally does. -- Note that admin departures evolve power to Foundation admins.)

But just hitting "Random article" ten or twelve times and comparing the state of the articles to their late 2007 versions before editor recruitment and retention started falling apart is a fair look. Every time I do that I'm invariably astonished in a good way. I must have done that fifty times in the past year and I can't ever remember seeing something that gave rise to concern. I wonder if WRers have a different experience?

It's easy to find censorship and deliberate bias in controversies, of course, but random controversial articles (they have a category) only disappoint me maybe 5% of the time when compared to their late 2007 versions.

James, those studies are all very limited. And most of them, especially the Nature one, have been
heavily criticized for being badly designed or badly implemented. And they are all aimed at general
article accuracy, not at vandalism specifically. Most of them did not count the factual errors, just
the correct items. And the latest one was in 2008--four years ago.

You ought to talk to Robert McHenry, he will have some things to say about these "studies".

If you want to do your own study, feel free. My issue, and the focus of the book, will not be on
article "accuracy". We are fully aware that there are hundreds of thousands of good, usable
articles on English Wikipedia. We are focusing on the inherent biases and the weirdness that
isn't obvious or simple. Like, for instance, the massive focus on Doctor Who. At present,
4,025 articles about bloody Doctor Who. Other TV shows, that have won awards and are considered
"classics", tend to be ignored or glossed over by Wikipedia. But Doctor Who, that's important, baby.

Why? What fantasy universe do Wikipedians live in? And why hasn't any of this been studied by
researchers before?

Ahem. Most of those studies were commissioned by PC magazines, not renowned for their grasp of the humanities. Of those which weren't, the one by George Bragues, on 'top Western philosophers' shows that Wikipedia's performance is 'decidely mixed', and 'fails to emerge as clearly superior to the traditional alternative of relying on individual expertise for information'. I also question his claim that there are no actual errors in the articles. On Rosenzweig's study, the Wikipedia article grossly misrepresents what he actually says. On what he actually says, I have a short piece here http://ocham.blogspot.com/2010/10/universi...-future_21.html .

Please make up your mind. If random articles invariably astonish you in a good way, how can random controversial articles disappoint you 5% of the time? And do you really have the competence to assess dozens of random articles on presumably a very wide range of topics?

Wow. Good stuff here. Definitely gonna look into this a bit deeper for the film: encyclopediagame.net

I have noticed that, for some reason, journalists and academics are really drawn to assertions on WP that turn out to be subtle acts of vandalism. Those assertions are just interesting enough, but just believable enough, (and conveniently not already "owned" by anyone) to make them prime targets for plagiarism. Then they get published, then cited, and we have the emergence of a new wiki-fact.

I anticipate one of the hardest parts of shooting "The Encyclopedia Game" will be confronting academics and journalists who have plagiarized WP, only to have their publications cited on WP. I am not eager to ruin any careers or anything. Especially not over plagiarism. Truth is truth, after all. Who cares who takes the credit? But journalists and academics have got to know better than to take what they read on WP for granted. Especially those really enticing bits, which turn out to be much more likely to be subtle acts of vandalism.

However, perhaps the ruin of a couple of careers will present a lesson for posterity that might salvage hundreds of careers that could have been lost to Wikipedia-related trust.

Copying word for word from Wikipedia as these guys did should be a world of career-hurt.

People mistrust science enough as it is. The whole system falls apart if you tolerate plagiarism.

I am guessing that academics and journalists who find themselves plagiarizing wikipedia probably aren't very happy with their careers anyway.

Perhaps not, although that would be difficult to prove one way or the other.

Keep in mind: plagiarists have always been with us and will always be with us. The main, and extremely
disturbing, point about Wikipedia is that plagiarists love it--because it makes plagiarism so easy.
And because contentious articles are constantly being changed, they can copypaste chunks of it, and
months or years later, the article will look completely different, so anyone looking for their theft might
not notice the changes. Or copy a chunk out of an old revision of the article. All hidden, automatically.

This is where the "article gnomes" come in. A couple thousand OCD sufferers, using custom bots to
constantly reformat sentences, change page formatting, move articles, change categories, etc. etc.
They are usually considered "important" by the Wiki-nabobs, even though their constant grinding does
NOT improve articles. They add no information, they just move it around. An improvement? Hah hah.

You should talk to the folks at Turnitin. They are the largest service for the discovery of school-paper plagiarism.
They have custom software that evidently has to scrape older revisions of WP articles, making it
slow and time-consuming, therefore costly. Meaning, in this era of bankrupt school districts,
middle-school and high-school students can usually get away with WP theft, because Turnitin is too
expensive to use on all the class assignments.

Wikipedia is Turnitin's biggest headache. Even theft from printed books isn't such a big deal.
Most modern school-paper plagiarism is from digital and online sources, which Turnitin can scan easily.
Those sources don't change very much--when Turnitin's server samples a Britannica article or an academic
paper online, it probably won't change, so they can keep using it for years. But Wikipedia is a constantly
moving target, so they have to sample it constantly....