Monday 7 November 2011

The Birth of TEI By Example

Sometimes academic projects come about due to a combination of necessity and wishful thinking. TEI by Example was a product of needing to fill a gap in funding, whilst wishing something into existence that I wanted someone else to do in the first place.

I've talked before about a pop-stack of ideas that exist for "need further work!". As well as that I have a pop-stack, or wishlist, of Things I Wish Existed That Would Make My Academic Life Easier. One of these was an online set of tutorials to teach TEI.

The Text Encoding Initiative guidelines, as we all know, are one of the bedrocks of Digital Humanities. If you want to be in DH, you need to know XML, and TEI, and markup. Which means it has to be taught. Even if, like me, you are not terribly interested in markup per se, and you never use it in your research, and you are not part of the TEI community, you will still need to cover it at some point in a class in Digital Humanities.

Now, there are some excellent people within the TEI doing some excellent teaching. There is some good teaching stuff available online. But what I needed, really, was some point and click tutorials that I could direct my masters students to after an introductory lecture on TEI. When learning code, it's the done thing that you look at and play with examples of code. Where oh where oh where was TEI by example? Where were examples of marked up texts people could see to learn from?

In late 2005 I was chatting to Edward Vanhoutte about his team at the Centrum voor Teksteditie en Bronnenstudie, Ghent, about how things were going up in the Centre. Unlike me, Edward and Ron Van den Branden markup texts all the time, and have considerable expertise in the markup of correspondence material. Edward was saying there was a gap in funds coming up for Ron, as they would be between projects. He needed to find a couple of months of work to pay him. Easy, I said. Apply to the Association for Literary and Linguistic Computing for a small grant, to build a set of tutorials for teaching TEI. Call it TEI by Example.
So we did. And we got the money, and we did.

It will take two months! we said. We started in 2006. Now, given I dont really do TEI, I couldnt really write the tutorials. Edward took that on, but he was in the phase of doing a lot of childcare with very young kids (and this was above and beyond our dayjobs). Ron built the infrastructure and did a fantastic job sorting out the quizzes, and set to building the online validator. I chivvied and tested and harangued. I was asked to give a plenary about the project at TEI@20: 20 Years of Supporting the Digital Humanities Conference, University of Maryland, in November 2007. It will be finished by then! we thought. Of course it wasnt. I gave the plenary when I was 8 weeks pregnant with the bump that turned out to be The Boy. (I cant recommend flying across the pond to give a plenary when the only thing you can stomach is salt and vinegar crisps. I was terrified that customs would take away my stash, and what would I eat when I was there? I also cant recommend giving a plenary when you think you are going barf over the front row's shoes. But I digress). We worked on the tutorials some more. 2008 came and went,with me mostly on Maternity leave. Then Ron also joined the parenting club! Hurrah! TEI by Example was finally born in July 2010, a mere 4 years late. Some gestation that was. Its like the closing credits of Toy Story 3, when they list all the kids born during the production phase.

Its worth also saying that we encountered a fair bit of resistance to TEI by Example from the TEI community. Folks were not interested, in general, in giving us access to fragments of their code. Promises were made and not kept, emails not replied to, snark was thrown on mailing lists. Who were we to build something beyond the TEI community! Who did we thinks we was!

Meanwhile, TEI by Example has been a quiet success. As Edward said in his plenary to this year's TEI meeting:
Fifteen months after the launch of the tutorials, the site has attracted close to 30,000 unique page views with 1,900 unique views for the modules on primary sources and critical editing together. The statistics and logs show that users are finding their way to the tutorials directly, via Digital Humanities courses or via the TEI website and we see that there is high activity from the US, Germany, the UK, France, and Canada: not surprisingly countries with a high digital humanities and digital editing profile. And we're particularly proud of our single visit from Vatican City. 18% of the visitors stay for more than 15 minutes on the site, which suggests that they really do some work. We also see a decent amount of returning visitors.
We hear its being translated into French. We know people within the TEI use it in their teaching. We're getting positive comments about it, and relationships with certain folks have improved dramatically. And I can sit my students in front of it for an hour, after an introductory lecture about TEI. Well I will do, when I return from my second maternity leave break to resume teaching next year.

As for the resulting paper? Here it is - my plenary which explains why we needed to go down this route.

Terras, M and Van den Branden, R and Vanhoutte, E (2009) Teaching TEI: The Need for TEI by Example. Literary and Linguistic Computing , 24 (3) 297 - 306. 10.1093/llc/fqp018. PDF

No comments: