I accidentally started a Wikipedia hoax

[Hex]

EJ Dickson, on the Daily Dot: I accidentally started a Wikipedia hoax.

Last weekend, Grantland writer Molly Lambert tweeted an Iggy Azalea pun about Amelia Bedelia, the protagonist of the eponymous children’s book series about a “literal-minded housekeeper” who misunderstands her employer’s orders. Jay Caspian Kang, an editor at NewYorker.com, then tweeted a screengrab of the Wikipedia entry for Amelia Bedelia, with a chunk of text highlighted:

jay caspian kang
@jaycaspiankang

@mollylambert @rachsyme wait what is this. pic.twitter.com/LxHf3gYZZ5

11:37 PM - 27 Jul 2014

I saw the tweet last Sunday, while I was scrolling through my timeline right after I got out of the shower. When I saw Kang’s tweet, I immediately screamed “WHAT” and sat on the bed in a towel for a good hour, furiously tweeting and texting my friends until my boyfriend came in and yelled at me to put some clothes on, goddammit, we have somewhere to be in 20 minutes.

I wrote that Amelia Bedelia edit in 2009, with my best friend Evan during our sophomore year of college.

The article features a chat "with Milowent, a Wikipedia editor who's considered an expert on Internet hoaxes".

[Kumioko]

Thats pretty funny

[The Devil's Advocate]

The article features a chat "with Milowent, a Wikipedia editor who's considered an expert on Internet hoaxes".

Miss Dickson is still spreading lies I see. Also, lol.

[SB_Johnny]

This sort of thing almost makes me wish I were sub-pop-sub-culture-fluent enough to plant some fun hoaxes.

[HRIP7]

As chance would have it, we just had a blog post on this sort of thing the other day: How pranks, hoaxes and manipulation undermine the reliability of Wikipedia.

Googling around this today, I came across some apt comments from the user page of Flutedude (T-C-L):

I've been away from Wikipedia for three years. I thought it was just a fad that would go away. Now it looks like it's here to stay, and it will continue to be a harmful influence on everyone. I'm just one man, I can't stop this disaster.

Kozierok's first law
"The apparent accuracy of a Wikipedia article is inversely proportional to the depth of the reader's knowledge of the topic." — Kozierok's First Law

For example, I know a lot about the clarinet (T-H-L). The Wikipedia article on the clarinet seems to me to be a mixture of bald facts, half-truths, misunderstandings and outright hoaxes.

I know next to nothing about Hopf algebra (T-H-L). The Wikipedia article about that looks very accurate to me.

Give that man a coconut.

[HRIP7]

Submitted to Slashdot.

[Poetlister]

This shows the importance of removing all unsourced material from BLPs (indeed from all articles). I suppose though that by now there are several "reliable sources" with that information.

[HRIP7]

John E. McIntyre, The Baltimore Sun: Truth has not got its boots on

Writing at the Daily Dot, EJ Dickson confesses to perpetrating a fraud: planting a false entry in Wikipedia about the Amelia Bedelia children's books saying that the character is based on a maid in Cameroon.

She was, she concedes, stoned at the time and now is horrified to discover that her little lark has become so widespread that it is believed by the nephew of the late author of the Amelia Bedelia books.

Twitter is agog with the story. Please share widely.

[HRIP7]

Now on the Slashdot front page.

107 comments to date.

[Johnny Au]

Wow. Wikipedia is getting more bad rap among the mainstream.

That Amelia Bedelia hoax is getting noticed around the world.

[Captain Occam]

I wonder how many times things like this will have to happen before people stop repeating the "it's not a problem because the hoax was eventually fixed" argument?

[HRIP7]

Caitlin Dewey, The Washington Post, August 1, 2014: What was fake on the Internet this week: High-five emoji, killer whales and festivals for ‘straight white guys’

3. Amelia Bedelia’s character was not based on “a maid in Cameroon.” That ahistorical tidbit circulated the Twitters last weekend after a New Yorker editor spotted it on the Wikipedia page for the classic children’s books. EJ Dickson, a writer for the Daily Dot, later confessed to making the (totally fake) edit in 2009, when she was a sophomore in college. “It was the kind of ridiculous, vaguely humorous prank stoned college students pull, without any expectation that anyone would ever take it seriously,” Dickson wrote. But apparently, many people did take it seriously — including prominent academics, bloggers and the book author’s own nephew, who cited the non-fact in an interview. (For what it’s worth, Dickson’s essay on what that says about truth, fiction and Wikipedia is really worth reading in full.)

[HRIP7]

Coverage in the Wikipedia Signpost.

[Poetlister]

But apparently, many people did take it seriously — including prominent academics, bloggers and the book author’s own nephew, who cited the non-fact in an interview.

There we are, so now it's in Reliable Sources so it's OK to have it on Wikipedia. After all, verifiability and not truth.

[Mancunium]

The Lesson from a Hoax
by Annoyed Librarian
Library Journal, 4 August 2014 link

Something shocking has happened in the world of information. It turns out that there was an error in the Wikipedia, a deliberate error introduced into an article for fun by a couple of stoned college students. This is as shocking as the time that other factual error was found on the Internet. [...] As for the hoax, nobody saw a lie become a fact. A lie was quoted as a fact. People believed said lie. Apparently a number of gullible people repeated that lie in writing as a fact, but it was still a lie. [...] So is this an example of the way a lie becomes a fact? Or maybe an example of how people should check their sources, as good librarians should tell them to? [...]

Instead, I think it provides just another example of how a lot of people are gullible and lazy. Librarians might talk about information literacy and the importance of finding reliable information, but people only want reliable information when something really matters, unlike their political opinions or who a fictional character in some children’s books is based on. When people are going to pay money for something, for example, they want reliable information about it. Oh, wait, except for all the people investing in subprime mortgages or buying lottery tickets. Okay, I take it back. People don’t want reliable information. They want comforting information. That’s it. People are gullible and lazy and prefer comforting, easily found, and easily assimilated information to reliable information. That’s why librarians are always fighting a losing battle. [...]

Cranky Annoyed Librarian evidently doesn't know the difference between Truth (T-H-L) and Truthiness (T-H-L).

[HRIP7]

In the discussion at the Signpost article, Llywrch (T-C-L) points out that the alphabetical list provided by Special:UnwatchedPages (T-H-L) (admins only) breaks off after 2,000 articles, with the 2,000th article being 1975–76 Israeli League Cup (T-H-L).

To see what proportion of the total those 2,000 articles might represent, I checked another large category, Category:All articles needing cleanup (T-H-L). That particular maintenance category contains 23,725 articles in total, and 1975–76 Israeli League Cup (T-H-L)‏‎ would be the 58th article in that alphabetical list. Now, knowing that 1975–76 Israeli League Cup (T-H-L) is the 2,000th (not the 58th) article in an alphabetical listing of the English Wikipedia's unwatched articles, we can guesstimate that the unwatched articles category is 2,000/58 times larger than Category:All articles needing cleanup (T-H-L) with 23,725 articles. So, based on that guesstimate, the English Wikipedia would contain about 820,000 unwatched articles that are wide open to hoaxes: anything that gets past ClueBot and Recent Changes is unlikely to be checked by anyone else.

The English Wikipedia contains 4.6 million articles, so we are talking about more than 1 in 6 articles being unwatched.

If that's true, that's a staggering number of articles.

[Johnny Au]

The Cranky Annoyed Librarian should know that one is never fully attainable, except from mathematics and computer programming on the most part, while the other is coined by a clown of French ancestry.

[Poetlister]

In the discussion at the Signpost article, Llywrch (T-C-L) points out that the alphabetical list provided by Special:UnwatchedPages (T-H-L) (admins only) breaks off after 2,000 articles, with the 2,000th article being 1975–76 Israeli League Cup (T-H-L).

To see what proportion of the total those 2,000 articles might represent, I checked another large category, Category:All articles needing cleanup (T-H-L). That particular maintenance category contains 23,725 articles in total, and 1975–76 Israeli League Cup (T-H-L)‏‎ would be the 58th article in that alphabetical list. Now, knowing that 1975–76 Israeli League Cup (T-H-L) is the 2,000th (not the 58th) article in an alphabetical listing of the English Wikipedia's unwatched articles, we can guesstimate that the unwatched articles category is 2,000/58 times larger than Category:All articles needing cleanup (T-H-L) with 23,725 articles. So, based on that guesstimate, the English Wikipedia would contain about 820,000 unwatched articles that are wide open to hoaxes: anything that gets past ClueBot and Recent Changes is unlikely to be checked by anyone else.

The English Wikipedia contains 4.6 million articles, so we are talking about more than 1 in 6 articles being unwatched.

If that's true, that's a staggering number of articles.

The number of unwatched articles is quite meaningless. Firstly, it may be that the "article watchers" for any given article are all banned, or haven't logged in for years, so the article isn't really being watched. Conversely, there are huge numbers of recent change patrollers, and while they are of highly variable quality, it is quite possible that vandalism to an apparently unwatched article will be picked up and reverted quickly.

[HRIP7]

In the discussion at the Signpost article, Llywrch (T-C-L) points out that the alphabetical list provided by Special:UnwatchedPages (T-H-L) (admins only) breaks off after 2,000 articles, with the 2,000th article being 1975–76 Israeli League Cup (T-H-L).

To see what proportion of the total those 2,000 articles might represent, I checked another large category, Category:All articles needing cleanup (T-H-L). That particular maintenance category contains 23,725 articles in total, and 1975–76 Israeli League Cup (T-H-L)‏‎ would be the 58th article in that alphabetical list. Now, knowing that 1975–76 Israeli League Cup (T-H-L) is the 2,000th (not the 58th) article in an alphabetical listing of the English Wikipedia's unwatched articles, we can guesstimate that the unwatched articles category is 2,000/58 times larger than Category:All articles needing cleanup (T-H-L) with 23,725 articles. So, based on that guesstimate, the English Wikipedia would contain about 820,000 unwatched articles that are wide open to hoaxes: anything that gets past ClueBot and Recent Changes is unlikely to be checked by anyone else.

The English Wikipedia contains 4.6 million articles, so we are talking about more than 1 in 6 articles being unwatched.

If that's true, that's a staggering number of articles.

The number of unwatched articles is quite meaningless. Firstly, it may be that the "article watchers" for any given article are all banned, or haven't logged in for years, so the article isn't really being watched. Conversely, there are huge numbers of recent change patrollers, and while they are of highly variable quality, it is quite possible that vandalism to an apparently unwatched article will be picked up and reverted quickly.

I wouldn't say it is meaningless.

First, I agree with you that the mere fact that an article is on someone's watchlist does not mean that edits to such an article will be scrutinised appropriately. Many people have thousands of articles on their watchlists, but only take a practical interest in a small subsection of them. Others, as you say, may not have logged in in years, may have greatly reduced their involvement, or indeed may be banned, yet the article will still fail to show up as unwatched, because it is still on such an editor's watchlist. However, what this means is that the number of articles that are unwatched for all practical intents and purposes is even greater than the list of unwatched articles would seem to indicate.

Secondly, I have less confidence than you that Recent Changes patrollers will pick up false information. In my experience, Recent Changes patrollers are rushed, and focus on the obvious stuff – deletions of sourced material, nonsense edits, etc. If an edit on a topic they know nothing about looks plausible, they will not give it a second look, especially if it appears to cite a reference.

[Poetlister]

Conversely, there are huge numbers of recent change patrollers, and while they are of highly variable quality, it is quite possible that vandalism to an apparently unwatched article will be picked up and reverted quickly.

I have less confidence than you that Recent Changes patrollers will pick up false information. In my experience, Recent Changes patrollers are rushed, and focus on the obvious stuff – deletions of sourced material, nonsense edits, etc. If an edit on a topic they know nothing about looks plausible, they will not give it a second look, especially if it appears to cite a reference.

I don't think our positions are very different. I just said that it was quite possible that they'd pick up vandalism.

[Mancunium]

The Downside Of An Elite Education, A Wikipedia Prank & Best New Reads for September
New Hampshire Public Radio, 2 September 2014 linkhttp://nhpr.org/post/9214-downside-elit ... -september[/link]

E.J. Dickson shares her story about the time she made up some facts about Amelia Bedelia that stayed on Wikipedia for five years.

Audio embedded.

[EricBarbour]

Chart34.png