Setting the record straight about Wikipedia
Sunday, August 17th by Ben YatesThree things that you really should know
Okay. So Wikipedia has a problem: most of the public still does not understand how it works. Even half the people who think they understand Wikipedia do not.
This is partly because the way Wikipedia works is just incredibly weird.
And it’s partly because once people start learning more about wikipedia, they’re inculturated into the wikipedian social sphere, and they start hanging out more with people who already understand wikipedia, and less with people who don’t.
(This is the story of the entire internet, by the way.)
Alright, class.
1. Wikipedia is not an encyclopedia
The first thing people think when they see Wikipedia is OMFG, these guys have everything!
People don’t discover Wikipedia by typing “wikipedia.org”. They discover it by typing “south america” or “techno” or “pope ratzinger” into Google, and then by staring at the screen for an hour discovering that every tiny facet of their favorite thing is covered in exhaustive detail.
They don’t read the “about” pages. They don’t even know those pages exist. They read the articles. And they see the logo in the upper-left corner, which lets them know they’re reading an encyclopedia.
But Wikipedia is actually not an encyclopedia — not in any normal sense of the word. It’s a million miles away from any other encyclopedia; the name obscures more than it reveals.
- Every other encyclopedia is written by an identifiable group of people — usually credentialed experts who were recruited and paid by the encyclopedia company.
- Every other encyclopedia is difficult and expensive to produce. Each edition of Britannica takes decades to complete. When Microsoft created Encarta, they didn’t write it themselves because they knew that would be fucking impossible; they bought and renamed an existing encyclopedia.
- Every other encyclopedia has gone through a traditional editing process wherein one person creates a document and a few other people fix its spelling mistakes and structural problems.
- Every other encyclopedia is released in canonical editions — the culmination of a long process that produces final, finished articles. The reader only sees the final version.
Wikipedia is none of these things.
- Unlike the staff of an encyclopedia, Wikipedia’s “staff” is permeable and informal. New people join and leave all the time, without stepping through a defined process — sometimes without identifying themselves or even using a pseudonym. It’s more like a subculture than a company.
- Wikipedia is officially nonprofit (not just unofficially, like Britannica :P ). It’s run like a charity, but it grows like a Kudzu. It’s already much larger than Britannica, and it’s only a few years old.
- Every Wikipedia article is written collaboratively, like a whiteboard. The most-edited wikipedia articles are effectively authorless! You can dimly see the imprint of individual authors if you read closely, and you can even view the changes each author made behind the history tab, but the cumulative effect is of reading a document created by a hive of bees. (Wikipedia’s distinct house style is a side effect of its fuzzy editing process.)
- In general, there’s no such thing as a finished Wikipedia article. See deletion by the numbers.
Wikipedia may be an encyclopedia in the abstract, in the same way that the State of Michigan is a nation. But it is not like any other encyclopedia; it is the black swan.
2. Anyone cannot edit Wikipedia
The second thing people realize about Wikipedia is that anyone can edit it. (This realization comes hard because universal editorship is such a strange idea. Wikipedia used to get panicked emails every day from people who thought they’d been granted improper access to the editing screen.)
And their next thought is holy shit, anyone could have edited this article? Anyone could have added misinformation or errors?
But anyone cannot really edit wikipedia — or rather, the question is philosophical: if you edit an article but your change is rolled back after 30 seconds, did you really edit it?
Basically, Wikipedia is very good at reversing people’s edits. There are two ways to look at how Wikipedia does this.
A. The immune system model.
Okay, so the first time your body encounters a harmful invader, it’s like “oh shit” and it scrambles to wipe out the bug after it’s already started breeding. But the second time it encounters the invader, it’s learned to recognize it. It wipes the bug out in no time flat.
Wikipedia has an immune system. Part of it is the recent changes page, which lists every new edit, anywhere on Wikipedia. Part of it is the hundreds of people who dedicate their time to checking facts and fighting vandals.
And part of it is the reflexes that these vandal-fighters have developed after months or years in the trenches, and the attitudes they pass on to each other. They’re like policemen on a dark city street — Does that guy look suspicious? He must be up to something, look at the way he’s standing around.
If you submit an article that’s not written in the house style, if you make an edit without registering an account first, if you look at someone a little funny, your edit might get reversed — you’ll look like a hoax-writer or a vandal or a PR guy, even if you’re not.
Okay, make that the immune system/police force model.
B. The subculture model.
People clump. They clump a lot in high school, but also out in the real world, and especially in situations where there’s no formal hierarchy.
Different subcultures have different ways of thinking, different ways of doing things. In theory, you could go up to a bunch of london punks and just sit down and chill with them. But in practice, you can’t — not unless they think you view the world the same way they do.
Anyone can be a wikipedian, but not anyone can be a wikipedian. This is a bit of a problem, but it does work pretty well at keeping articles free from vandalism.
3. Wikipedia tends to be pretty accurate (and when it’s not, it’s easy to tell)
Really, accuracy is generally not a problem. The level of misinformation in Wikipedia is low. For example, the journal Nature compared wikipedia favorably to britannica back in 2005. (Nature’s study did have some flaws.)
But you can rigorously gauge Wikipedia’s accuracy by reading about topics you already know a lot about.
And you can even estimate the accuracy of articles about unfamiliar topics — for example, by checking the article’s revision history and contacting the authors of individual edits, or by going back to the primary sources listed at the bottom of the article.
If you read wikipedia enough, you’ll start to develop an intuition for its accuracy. This sounds mystical, but it’s the same way you determine which newspapers and TV shows to trust.
In the meantime, I recorded a few screencasts to demonstrate clues you can pick up from the article’s context (these are all from a book I’m co-authoring, How Wikipedia Works).
First, check for outgoing Wikilinks. (If you can’t see the embeds, the videos are all here.)
Second, check for incoming Wikilinks.
Third, check for a stub notice.
Fourth, see if the article’s topic is particularly controversial.
Jimmy Wales talked about this a little:
Neutrality Point of View (NPOV) is absolute and non-negotiable in Wikipedia. The problems come up in obscure topics, such as Japanese anime. For common topics people come together and make a decent statement on what it is. It turns out what is really important is that participants have a shared vision of what they want to accomplish.
Mutually-assured destruction is inherent in Wikipedia. People who want to push an agenda end up having to write “for the enemy” rather than to those who share the same bias. Most people are pretty reasonable, but you don’t get that sense from TV where they put up two people on opposite sides. Most people are in the middle and aware of pros and cons of issues.
Fifth, check for categories
Okay, class dismissed.
But if you have any other techniques — or questions, or annoyances — there’s a discussion at the cracked forums.
August 18th, 2008 at 9:32 pm
I’m going to copy a few of those Cracked posts here.
August 18th, 2008 at 9:33 pm
Disco Stu:
Onebrain:
Johnny Roastbeef:
DevHyfes:
kelvinc:
The Furlinator:
The Evil Sloth:
Kathana:
August 21st, 2008 at 12:44 am
Wikipedia is not an encyclopedia by definition… At least according to the definition from the Wikipedia in Spanish:
“Una enciclopedia es un compendio exhaustivo del conocimiento humano. La enciclopedia reúne y divulga datos especializados o dispersos que no podrían ser hallados con facilidad y que presentan un importante servicio a la cultura del hombre moderno.”
Therefore there are more deletionists in the Wikipedia in Spanish.
The lack of a firm definition of the objectives of the project derived in the sterile fight between inclusionists and deletionists
Wikipedia is successful because is not an encyclopedia (Is bigger and more useful) and because promises the impossible ideal of the NPOV. NPOV could be subjective like Justice and Beauty, but we want it anyway.
IMHO Wikipedia should fork in a “Wikibase” (a site for inclusionists, and a source for the other proyects), a “general Wikipedia” (A selection of “the most encyclopedic content” for a printed version and the new main battlefield for POV pushers), and multiple Thematic Encyclopedias (Where super-specific articles will not be erased).
We just need more Transwikists xD