Wednesday, August 31, 2011

Power To The Wiki People

While reading Robert Kurzban's (mostly good) book Why Everyone (Else) Is a Hypocrite, I came across a fragment of a sentence which annoyed me.

Kurzban is in the midst of explaining a computer science topic when he writes, "according to Wikipedia, which I am usually hesitant to use, but will for this purpose," and then block quotes, starting with "creating a domain-specific language...".

As it turns out, Wikipedia is written by people. Rp, a Dutch software developer, added this sentence on April 24, 2008. The only difference between Rp's change and the current version is a superfluous "of course" which was removed on May 12, 2008. This is not hard to decipher based on a gander at the page's revision history. It took me about four minutes.

Listen, I recognize that Wikipedia is low status, which is why Kurzban had to express his hesitation to use it. But we all know how Wikipedia works now. People edit it. You aren't constrained to merely cite "Wikipedia", you can look back and see exactly which user wrote that passage first, and cite that user.

Some of my clever readers are probably already mentally defending the status quo by saying that some sentences or portions of articles have been edited so many times that it would be difficult to say who wrote them.

Yes, in very rare cases like The Iraq War this is the case, but it happens much less often than you'd expect. The vast majority of articles are edited in chunks of sentences, paragraphs, or sections, and these chunks are eminently traceable.

Perhaps the best thing about such a shift in citation norms is that it would help incentivize people to edit Wikipedia. If you think this is not an issue, you are sorely mistaken.

Consider the page on epigenetics. Inspired in part by Razib's manifesto about the growing importance of this topic, I have subscribed via RSS to the changes made to the page since January. What I expected were the vitriolic edit wars deserving of such an unfolding, important topic.

But I've seen nothing of the sort. In fact the page hasn't changed significantly since Team Cytokine Storm made some edits last November. In the meantime, how many words have been typed about epigenetics for publication elsewhere?

For example, in the past two months, there have been at least four academic reviews on topics in epigenetics (see here, here, here, and here). See for yourself--here is a pubmed search for "epigenetics review."

How many people will read these reviews? Do you think that more will read those reviews than will read the Wikipedia page? Is this a healthy division of labor?

nobody does homework on saturdays
Bottom Line: When quoting or referencing an article hosted on Wikipedia, cite the major user(s) that contributed, instead of just "Wikipedia."