CWRC-shop at DHSI 2014

logo-dhsiWe just held our first CWRCshop at DHSI 2014 at University of Victoria where the campus was absolutely gorgeous with spring flowers and marauding deer (no bunnies, sadly).

DHSI was bigger than ever this year–28 courses!–including ours in Collaborative Online Scholarship. I was very pleased to have co-teaching with me Karyn Huenemann, manager of the Canada’s Early Women Writers Project at SFU, and Michael Brundin, CWRC metadata co-ordinator, and Mihaela Ilovan, CWRC project manager, both from U of Alberta. In the background, in Edmonton, was Jeffery Antoniuk, valiantly working away to help ensure that things were running. We had a big development push leading up to DHSI and were still testing thing and working to get some things working better while the course was proceeding: not the ideal scenario as these things go, but class members were extremely patient and forgiving of the glitches that we ran across, and the workarounds that we had at times to make, in the course of the week.

Our course description (submitted far in advance of the course) read as follows:

Online Collaborative Scholarship: Principles and Practices (A CWRCshop)

Susan Brown, with Mihaela Ilovan, Karyn Huenemann, and Michael Brundin

Online resources play an important role in contemporary scholarship, both at the point of production, where online resources and tools are playing an increasing role, and the point of dissemination. The web offers many opportunities for online publication, whether through social media such as Twitter, Facebook, or blogs, and through the dissemination of scholarship via more formal scholarly channels in online journals. In between these two stages, however, most scholars still collect materials for study, do much of their analysis, and prepare scholarship for publication on personal computers. This means that work in progress and working archives are stored in silos on individual hard drives, making it harder to share information and ideas or to benefit from those of other scholars while the work is still in formative stages. There is an unrealized potential for sharing, reducing the time spent on lower-level activities, and collaborating either passively or actively.

This course takes up the movement of scholarship to online environments, exploring possibilities for collaboration throughout the entire scholarly workflow. It is suitable for those wishing a general introduction to digital humanities as well as for those wishing to initiate a longer-term project, providing a general introduction to key principles associated with undertaking DH scholarship, ranging from platform-independent data formats and metadata standards to text markup, preservation challenges, and semantic web principles; it will consider practical, institutional and cultural challenges associated with collaboration, as well as strategies for deciding what types and levels of collaboration are right for particular individuals or projects. Through readings, discussion, and hands-on sessions, participants will engage with the topic of collaboration in broad terms, while also being introduced to a number of fundamental principles related to the sorts of choices facing scholars engaging with digital research environments. Hands-on experience will be provided in CWRC’s collaborative online environment as well as other tools; participants will be able to export the digital objects they create, so no long-term commitment to working in any specific environment is required.

This course is sponsored by the Canadian Writing Research Collaboratory / Le Collaboratoire scientifique des écrits du Canada.

To explore online scholarship we will use the Canadian Writing Research Collaboratory, a platform designed to support collaboration in a range of forms, from interlinking the research of individual scholars to larger group projects. CWRC supports a range of activities, including creating collections containing one or more of the following kinds of objects: bibliographical records, page and other images, audio records, video records, and born-digital texts. CWRC can be used to create focused collections of related research materials in a combination of forms (for instance, you could can have a combination of image files, text files, and sound files), biocritical scholarship such as that of the Orlando Project, timelines, critical editions, or side-by-side editions of texts and page images. Users can perform OCR (Optical Character Recognition) on page images to extract text, and can create or edit digital texts, annotate page images, add semantic encoding, or tag people, places, texts and organizations so they link to other mentions of those entities across the collections. Participants don’t need any background in text encoding or the digital humanities. Those who have some experience will be able to explore CWRC’s more advanced features and functionality.

Portions of the CWRC platform are still in development, so participants in the course will also have the opportunity to consider prototypes for possible inclusion and provide feedback on the platform’s interface and functionality, providing participants with insight into another aspect of working in the digital humanities.

That last sentence became more true than we would have wished, given some delays in development that meant the CWRC system wasn’t as far along as we would have liked. The course was really two courses in one: the first on the principles and practices of collaborative scholarship in a digital environment, and the second offering a kind of “meta-course” as we called it, that gave participants a glimpse of what a large-scale DH infrastructure project in progress looks like from the inside, as we worked as a team, with Jeff plugged in from Alberta, to get various things in place, and to troubleshoot problems that arose, over the course of the week. We’re grateful too to our partners at Open Sky Solutions and Discovery Garden, who worked hard with us in the period leading up to the CWRCshop to push things along, and did some mad debugging in the period immediately before.

There were suspenseful moments when we waited to see whether a particular functionality would work, moments of gratification when they did, frustration when they didn’t, and some brilliant fixing and work-arounds by the CWRC team. I learned a huge amount from the course. While “Principles and Practices” initially felt like a handy rubric to indicate the blend of discussion and hands-on activities in the class, a sense not only of the connection and also a tension between them emerged very strongly, since when it comes down to it there is seldom world enough and time to live up to all principles when it comes to actually putting things into practice. Getting a sense of where various people are with their thinking about working collaboratively, with specific projects, and of the wide range in the level of technical knowledge and extent of experience of working digitally, either individually or with others, was hugely useful to me. There’s no doubt that the diversity of aims, methods, content, and technical ability in the humanities makes for a big challenge when it comes to thinking about collaborative online environments, but it is equally certain that the rewards of collaboration and getting our materials to talk to each other would be huge.

One exciting moment was getting to test a much more functional (though still “very alpha”) version of the HuViz knowledge visualization tool prototype (for which we need a better name) than we had seen before, thanks to some very hard work by Shawn Murphy of Semandra. HuViz developed by the Text Mining and Visualization for Literary History project, building on the OVis prototype developed through a grant from Sharcnet (we were very glad to have a couple of folks from Compute Canada‘s new “Team DH” join us towards the end of the course), is a browser-based knowledge visualization tool that graphs relationships embedded in humanities data. In this case, what is being visualized is quads (RDF triples with contextual information about provenance) extrapolated by John Simpson (you can see the paper about it here) from the relations implied in the Orlando Project‘s XML data. And a great recent win is the ability to see the portion of the Orlando textbase on which that relation draws, and to dive from there into the entire entry. Another aspect I love is the ability to remove nodes from the graph if they aren’t relevant to what you are exploring. The interactivity of the graph over the web, the ability to refer back to the source material from which the graph is abstracted, and the wide range of ways in which the graphs can be shaped, pruned, and engaged with: these are all affordances essential for a broadly useful graph visualization tool for humanities research.

Brontes genres and topics

The Bronte sisters’ connections to genres and topics, selective nodes labelled

There were lots of points of larger collaboration: we had a number of participants in the course from existing CWRC projects such as the Canada and the Spanish Civil War project and the Cabaret Commons; Dean Irvine came to share the advances on critical edition creation and collation made by the Editing Modernism in Canada Project; and Zailig and Josh Pollock of the Digital Page project, and Alan Stanley from Discovery Garden, and others came into the class (including the George Whalley project’s  Mike DiSanto and Robin Lsard via Skype from Algoma University) for a collaborative colloquy on collaboration that went surprisingly well given the number of people in the room and the range of topics we wanted to (and managed to) cover.

We were one of the bigger classes, at 23 participants, with folks ranging from grad students in English to professional programmer and IT folks to senior professors. Quite a range of expertise and interests, but everyone got on very well and helped each other out when needed. Some fun collaborations developed within the course, beyond those who had come with the intention to collaborate. Indeed, the suggestion was made that next time we might just assign the entire class a collaborative project to give people a hands-on taste of working together.

Collaborative timeline produced by class at DHSI 2014

Collaborative timeline produced by class at DHSI 2014

We did, however, collaborate on one project, which was the production of a shared timeline that reflected the eclecticism and broad expertise in our group. Using the CWRC timeline and mapping tool (still looking for a better name!), which is itself based on the Simile timeline out of MIT, a shared set of events we created in the form of structured data was taken by Michael as the basis for a custom timeline to show in the poster session at the end of DHSI.  The data transformation process was abetted by John Simpson’s “Fundamentals of Programming/Coding for Human(s|ists)” class: collaboration all the way down! You can explore our timeline here. I certainly learned a lot: who knew that Don Cherry only played one NHL game, for instance? Maybe every other Canadian, but not me.

Going forward, we have a number of things to consider based on our experience this spring of this and of the mini-CWRCshop at Congress at Brock University the week before: whether to offer the course again at DHSI next year, whether to try to demonstrate so much of the CWRC functionality in a single course, what the right balance is between discussion and practical work, the ideal length of CWRCshops, and how best to offer ongoing training and support for researchers both new and continuing.

What is certain is that this was a hugely valuable learning experience for all of us at CWRC, and we are so grateful to those who chose to hang out with us for a whole week in sunny Victoria!