May 18

Aaron Swartz’s Essays about Wikipedia

Looking through my Zotero library, I found a series of essays by Aaron Swartz, written back in 2006. I added these essays years ago, before I know who Swartz was. I heard about him (mostly on Twitter) only in the last year couple years, first regarding the JSTOR lawsuit and then about his tragic suicide. If you don’t know who Swartz is, I highly recommend Larissa MacFarquhar’s piece in the New Yorker, “The Tragedy of Aaron Swartz.”

I originally saved the essays because they’re about Wikipedia, and I returned to them because I’ve been very interested in Wikipedia lately (as I’ve written about on this blog, I had the students in my Romanticism course add to a Wikipedia page as part of an assignment about literary vocabulary). Though seven years have passed, Swartz’s Wikipedia essays are still worth reading.

In “Wikimedia at the Crossroads” Swartz talks about his attendance at Wikimania and his campaign for the Wikimedia Foundation’s Board of Directors (he was not elected). The second essay is the one most worth reading. “Who Writes Wikipedia?” challenges Jimmy Wales’s claim that Wikipedia is in fact mostly written by a small group of people (you can watch a 2005 webcast of Wales). Swartz concludes that although “insiders account for the vast majority of the edits,” they’re mostly just formatting: “it’s the outsiders who provide nearly all of the content.”

The third essay, “Who Runs Wikipedia?“, begins with the question, why did Wikipedia work? He argues that Wikipedia’s strength is in its users, not the governing organization. In “Making More Wikipedians,” he argues that to promote Wikipedia, Wikipedians need to recruit, “giving talks and tutorials to groups that you know about, explaining the core ideas behind Wikipedia, and giving demonstrations of how to get involved in it.” Apparently, in 2006 only 10% of people knew about Wikipedia. Now (according to Alexa) it’s the 6th most popular site in the world.

The fifth of Swartz’s essays, “Making More Wikipedias,” argues that “Wikipedia’s real innovation was the idea of radical collaboration.” The software matters less than the ethos and the culture. This isn’t to say that software doesn’t matter at all, though, as the final essay, “Code, and Other Laws of Wikipedia,” makes clear. For example, the interface has to be user-friendly, otherwise non-experts are barred from entry. Yet Swartz maintains that these are really political choices, not technical ones, and the Wikipedia community (not a for-profit company, even if run by Wales) should make the decisions.

I’m in the process of compiling a list of articles about Wikipedia, so if you know of any good ones, please let me know!

Apr 18

Diary of a Visiting Assistant Professor, week 33: Wikipedia Pitfalls, Anticipated and Not

Last week I posted about an assignment in my Romanticism class, part of which requires students to add to Wikipedia’s glossary of literary terms, a potentially useful but mostly empty list. Thanks to my students, the list is now significantly less empty than it was a few months ago.

I’m generally pleased with the assignment. At minimum students have learned how Wikipedia works, and I’m convinced they now have a better understanding of the vocabulary of literary study — which was, ultimately, the goal. Nevertheless, I learned some things from this assignment which I’ll definitely apply if and when I use it again (and since I’m considering teaching a course on Wikipedia in the fall, it’s very likely).

First, the pitfall I had anticipated. While students use Wikipedia just about every day, very few had ever contributed. As every faculty member who has ever used any kind of technology in a class knows, just because students interact daily with a tool or website doesn’t mean they’ve ever thought critically about it. So I spent some time on basic editing: in class I showed them how to use the “edit” tab, and when we added citations (see below) I formatted the references myself and emailed them to students to copy and paste. A few still had questions, but the actual act of editing a Wikipedia page didn’t prove much of an obstacle.

What I wasn’t prepared for was that a user — “Green Cardamom” — had taken ownership of the glossary, policing it to ensure entries meet all the guidelines. I knew that Wikipedia pages require citations, but partly because I didn’t think glossary definitions really needed to be cited (all the terms already have their own pages, which don’t necessarily cite a source for the definition) and partly because it seemed an unnecessary step for students, I had told them not to stress about including citations. This was bad advice.

From the speed with which she responded, Green Cardamom must subscribe to the page’s rss feed. I created a Juxta a collation of the edits made on January 28th, the date on which the most students contributed. The histogram view clearly shows what’s happening (the blue bars represent differences, and the edits are in chronological order): histrogramAs you can see, the progress isn’t linear — edits are being made and deleted (you can see the specifics if you view the collation in juxtacommons). To my students’ chagrin, Green Cardamom was deleting their definitions because they lacked citation. After a brief exchange on the Talk page, I asked my students if they’d be willing to go back and add citations, which they already had (to create their definitions I asked them to consult at least two places.) This class had good rapport with me and with each other, and they were invested in the assignment, so they were happy to do so. It was a lesson for me, though: it had never occurred to me that a person would take responsibility for a certain page, especially a page that remained largely empty. From what I’ve read since, this actually isn’t uncommon, so maybe the real lesson is that I should have done more homework.

Have you encountered a similar roadblock in a digital assignment? How did you handle it?

Feb 26

Wikipedia, the Oscars, and Juxta

Like millions of other people, Kate and I watched the Oscars this week. As Daniel Day-Lewis walked towards the stage to accept his Academy Award for Best Actor, the voiceover informed us that he was the first male actor to win the award three times. Wondering what the first two awards were for, we turned where everyone in the twenty-first century turns: Wikipedia. Kate had the page open before Day-Lewis was more than a few sentences into his speech, and of course we learned the answer to our question (My Left Foot [1989] and There Will Be Blood [2007], if you’re wondering). But Wikipedia also told us what we’d just learned from the announcer: that Day-Lewis was the first male actor to win the award three times. That’s right: Day-Lewis’s Wikipedia page was updated almost before he began his speech.

I’ve shared this anecdote with a few people, and many others have had similar Wikipedia experiences (often with celebrity deaths). Information moves quickly. But since I’m working on an article about Wikipedia — specifically, about an assignment in my Romanticism class — I wanted to use this opportunity to test juxtacommons, a web-based collation tool, currently in beta. Developed by NINES, Juxta allows you to compare multiple versions (“witnesses”) of the same text, and visualize the differences between them (you can do this on Wikipedia too, but only two at a time). About a year ago, as a NINES fellow, I wrote about using Juxta to compare different versions of the Wikipedia entry for Digital Humanities. At that time Juxta was desktop-only and I entered the data by hand, but the new online version has the capability built into it: you just select the versions of the Wikipedia page you want to compare. It’s very user-friendly.

What’s more, Juxta Commons also lets you embed the collation directly into WordPress. The heat map below compares the first 15 edits following the Academy Award announcement (15 is the max Juxta can handle at one time). All occurred within 8 minutes. Note the absence of changes in the citations (which Wikipedia supposedly requires).

If you’ve never used Juxta before, the blue highlighting represents the difference from the “base text,” in this case the most recent version. The darker the highlighting, the more this version differs from the others. We can also view the comparison from the other direction, setting the base text to be the version of the page right before the announcement:

If you view the comparison in Juxta Commons (here), you can get into the data even more: you can compare versions side by side, choose different base texts, or see it in a histogram.

Juxta is a powerful tool, and while it was created for literary uses (early examples compared versions of poems by Rossetti and Tennyson), it’s since found new audiences: another NINES fellow, Emma Schlosser, looks at the Wikipedia page for the Benghazi attacks; Dana Wheeles, project manager for NINES, has compared versions of itunes and Instagram terms of service; and Juxta has even been used on Broadway, to compare versions of Tennessee Williams’ “Masks Outrageous and Austere.”

So, do you have different versions of a written document you’d like to compare? Or are you interested in tracking versions of Wikipedia pages? Juxta Commons is for you. And check back here for more about using Wikipedia as an assignment.