Wednesday, December 03, 2008

Don't Cite Wikipedia

[See also this more detailed post.]

There are all kinds of uses for Wikipedia. Because Wikipedia is open to input from anyone who is interested in a subject, you sometimes find little nuggets of specialized information that you would not get (as easily) elsewhere. Its "neutral point of view" policy also often ensures that arguments on both sides of controversies are presented (sometimes in a somewhat artificially "balanced" manner). A good article really can teach you something about the subject matter.

Wikipedia is also a reasonably dependable source of reliable sources of information. That is, it often cites perfectly good sources. But it is not itself a reliable source of information.

Much of its unique value comes from its open and dynamic platform, which is also what makes it a poor source of authoritative knowledge. Wikipedia is "the encyclopedia that anyone can edit" and many of its articles change several times a day. So it is of little use to the reader to know that Wikipedia makes a particular claim about an event, or that Wikipedia defines a certain concept in a particular way.

It does not help to specifiy the date you read the article, or to cite the particular version of the article you're quoting from. Either way, you would have to have some special reason to believe that this version of the article is right. And if you have such a reason then you also have a better source for the point in question. Cite that source, not Wikipedia.


Tim said...

Disagree. Most articles in Wikipedia are reliable and valid.

Thomas Basbøll said...

That's not true statistically. (Wikipedia has millions of articles, very few of them have passed even Wikipedia's internal review.)

Like I say, many of the articles are informative and even valid. But the nature of the project precludes using it as an authority. The fact that Wikipedia says something is the case gives your reader no particular reason to think so. But if you really have found a valid article, then that article will give you all the information you need to provide a solid reference.

CLULEY said...

Surely the problem with citing Wikipedia, as Thomas points out, is that it changes so often. Even if the reference is a valid piece of information, it's impossible for anyone to return to the original source of the reference. The url might be the same but, statistically, the content will have changed.

Thomas Basbøll said...

Well, you can actually avoid the instability problem. E.g.:

Finance is, among other things, "the study of money and other assets" (Wikipedia 2007).


Wikipedia, s.v. "Finance," (version as of 17:55, 31 May 2007, accessed December 3, 2008).

But this obviously doesn't solve any problems.

Jonathan said...

I cited Wikipedia once in my Lorca book. I learned of the existence of Leonard Cohen's song "Take this Waltz" and Joan Baez's reading of some Lorca poems on her album "Baptism" from the Wikipedia article on Lorca. I thought it was the correct thing to do to note the place where I first came upon this information. I could have just not acknowledged this source. Nobody but me would have been the wiser. It's not like I cite Wikipedia often, or believe it to be reliable, but that in this particular case this was, in fact, a source of some "leads."

Thomas Basbøll said...

Say it isn't so, Jonathan!

I can imagine crediting Wikipedia in a footnote, or in the acknowledgements. You can even work some kind words elegantly into the main text. One can always find a way of weaving a research process narrative into the presentation of the results.

But say you didn't just take Wikipedia's word (and expected your reader to do the same) for what Leonard has sung or Joan has read?

I mean, these are things that can be verified independently. Those albums do actually exist. Their relation to Lorca can be established by applying your own expertise. It is that application we read your book for.

In fact, if there was no reliable source in Wikipedia when you read about this, after your book is published a source can be added: your book.

Think about this: where did Wikipedia get the info? What if it was taken from some uncredited source?

Jonathan said...

That's exactly what I did: credit Wiki in a footnote. Of course I verified the info myself.

Thomas Basbøll said...

That's what I figured. Whew.

I think I'm going to write a post on the difference between crediting someone (or thing) with a discovery, and citing a source of evidence.

When your book arives I'll be sure to add the reference to the Wikipedia article. First thing.

Anonymous said...

I will be sharing that comparison, source of "Discovery v.s. Evidence", with my students. Thanks!