The map above shows the geographic distribution of Wikipedia articles. Darker blues mean more geotagged articles per country. According to researcher Mark Graham, who created the map, Wikipedia has some pretty prominent blind spots:
Almost all of Africa is poorly represented in Wikipedia. Remarkably there are more Wikipedia articles written about Antarctica than all but one of the fifty-three countries in Africa (or perhaps even more amazingly, there are more Wikipedia articles written about the fictional places of Middle Earth and Discworld than about many countries in Africa, the Americas and Asia).
This morning, Nora talked to Mark about representation on Wikipedia, and how sometimes, our virtual worlds don’t match up to the real world. A shorter version of this interview will air on Spark 93, but you can hear the full, uncut interview below, or download the MP3. [runs 15:56]
Play audio:
If you like hearing these extended interviews, why not subscribe to Spark Plus? You’ll get regular weekly episodes, plus additional blog-only content like this. [Subscribe via RSS] or [Subscribe with iTunes]
Great interview, and glad to see you covering Wikipedia.
However, I'm not sure it's fair to only use *geotagged* articles as the metric for how well a country is covered. If only 500K articles are tagged, that's less 1/6th of the total article count. And there is all kind country-specific articles that can't be tagged to a particular location: Politics of Canada, History of Ghana, etc. If you compare "Category:Burkino Faso" to the root categories of other African countries, it's not any more well-covered; it's just that someone spent the time to geotag the articles.
That being said, glad to hear someone eviscerate the theory that Wikipedia is running out of potential material!
Correction: my fraction of 1/6th isn't accurate, because I was basing it on only the English Wikipedia. The map is for all languages, so the fraction should be more like 1/28th.
Hi Padraic. Thanks for the comment. You make some very good points. It's true that only a relatively small number of Wikipedia articles are geotagged. The main reason for this is that a lot of information simply isn't geotaggable: It wouldn't make sense to assign coordinates to the vast majority of articles on topics like apples, Marxism, post-rock etc. But, as you point out, some explicitly spatial articles do remain untagged. You're probably right that the reason that Burkina Faso has more geotagged articles (1071) than South Africa (945), Kenya (217) and the rest of Africa is diligent editing rather than more actual content in Burkina Faso. However, in all cases, these numbers pale in comparison to the huge number of articles in places like the US (89,549) and Germany (54,634). So, I'd argue that: (1) the geographic biases in tagged versus untagged articles are relatively small; and (2) because those biases exist we should pay more attention to the general patterns of geographic inequalities in content (i.e. the fact that there is much more content in the Global North than the Global South) than to the relatively minor differences between places.
- Mark Graham
I agree with your points. Thanks for drawing attention to this massive disparity and keep up the great work.
i was pleased to hear the interview on spark, it shone a light on some of the concerns that have been swirling around in my own head for some time now. i think the idea of wikipedia is fascinating, as is the ideal of 'democratization' that the web is supposed to offer generally, but we need to understand whose democracy it is. wikipedia becomes an easy default source for information, as more and more people turn to it for a quick answer to questions on all sorts of things. however i really feel that as contributions of new articles appear to have plateaued, the absences become more important than the content. it can't be considered 'authoritative' on anything (other than perhaps popular culture for geeks) if we are not as aware of what is not there as what is.
OMG. This was probably the most over-wrought and sophomoric interview with a pseudo-intellectual that I've heard in a long time (well at least this week).
Believe it or not, before the Internet, there was bias whether it was on purpose or not. I've been doing research for 20 years and the first thing that I learned was that as soon as you write something about a subject is that you have to decide what to leave out which would automatically create bias.
The constant refrain that I heard by both Mark Graham and Nora Young was an ominous refrain of, " you really have to know where you are getting your data from" as though this just happened when the internet came about. THIS HAS BEEN HAPPENING FOR CENTURIES! THIS IS HOW WE EXTEND OUR KNOWLEDGE!
Are we supposed to think that every year, Encyclopedia Britannica produced the most perfect and balanced articles about every location in the world without bias?
Personally I would rather have a detailed, obsessive entry about some village in England that I don't care about than worry that it isn't balanced by some slap-dash article about some equally obscure village in Ghana.
Of course, I would like to see more geographical balance in wikipedia, how could you not. I would also like ice cream to cure cancer. But to throw out the wikipedia experiment just because it isn't perfectly balanced geographically, I think is, well , idiotic.
I can here you now, saying, "but that isn't what I'm saying". Listen to the interview again and it is exactly what you are saying.
I'm sorry for the spelling mistakes and bad grammar but I'm so mad that these so-called profound "insights" are getting air time that I'm just spitting this out and I haven't re-read this. But these "ideas" are just quite frankly, crap wrapped up in faux concern that I just can't stand it.
Hi William. Thanks for your comment. I fully agree with you that this has been happening for centuries. Representations are always biased and always highlight some aspects of place at the expense of others. However, with Wikipedia, I often hear the argument that we are running out of topics to write about. Thus implying that the 3 million or so articles in English cover almost any topic/place/event/thing we can imagine.
In no way am I advocating throwing out the Wikipedia experiment, and I would never deny that Britannica and others probably have equally problematic biases. My point with these maps is to highlight some of the gaps in knowledge that we can: (1) work on filling in; and (2) be aware of when using Wikipedia as a resource.