There has been a flurry of attacks on one of the most useful sites on the Internet: the Wikipedia. It’s the source that I recommend most often for math and computer science. But students tell me that it’s disparaged by their other teachers, who consider it unreliable at best and scurrilous at worst. A recent USA Today article has exacerbated these opinions. My first reaction was to smile at the idea that USA Today was calling some other source unreliable. Then I read the somewhat sensationalistic article, after which I discovered that it was talking about a single unfortunate case in which one guy from Nashville had posted a fake biography as a misguided joke on a co-worker.
The Wikipedia is one of the wonders of the internet… for most of the time it worked remarkably well, reflecting the essential goodness of human nature in a supposedly cynical world and fulfiling a latent desire for people all over the world to cooperate with each other without payment. The wikipedia is now a standard source of reference for millions of people.
Back in the spring and summer, I decided to test out the accuracy of the Wikipedia. So I checked more than 200 entries on a variety of topics from mathematics, computer science, and linguistics — hundreds of entries, containing thousands of facts. Of course I didn’t know everything discussed in these entries, but I purposely picked topics on which I had a reasonable amount of expertise. Out of these thousands of facts, I found exactly two — count them, two — that I knew to be incorrect. This is an astonishing rate of accuracy. I’m sure I missed some errors, but still…
Would the New York Times be that accurate? I doubt it. It publishes several corrections a day, and it’s famous for a series of deliberately fraudulent articles. We trust the New York Times, the paper of record, but we’re apparently not supposed to trust Wikipedia because it’s written by ordinary people?
Or perhaps we trust the Encyclopedia Britannica. The Nature article referred to above (from the AP by way of MSNBC) found exactly the same number of serious errors in 42 articles on Wikipedia and a corresponding 42 articles in the Britannica. We need to take Wikipedia’s articles with several pinches of salt, but that’s true for all sources. We teach our students to be skeptical, not credulous, regardless of the source.
I do agree that citing the Wikipedia can be irksome to teachers, since the cited webpage may be different next year or even next week. But that’s a problem for all Internet-based sources, not just Wikipedia.
Finally, I have a conjecture about the possibility that Wikipedia might be much more reliable in certain disciplines than in others. People are not going to take the time to write and post an intricate article about the Mandelbrot Set or ergative languages unless they know what they’re doing. But I suppose a random person with an axe to grind or even just a strong opinion might well post inaccurate information about political or social topics, either intentionally or unintentionally. So maybe that’s why Wikipedia seems so reliable in math, linguistics, and computer science, whereas it might not be so reliable in English and social studies. In any case, I urge you to check out the entries for some topics in which you have some expertise of your own. I can’t judge the accuracy of humanities articles, but I have a suspicion that even they are more accurate than most people think. Let me know your experiences, especially in comparison to other sources.