DH2014 day three notes

DH2104 had free portraits!

This was a great last day to what has been a rich week of learning and discovery. It was certainly worth the long flight to attend DH2014. The organizers did a great job with the logistics and communication. The professional portraits they offered for free were a great idea; I’ve never had such a nice photo.

Session 6, Friday

The Dog That Didn’t Bark: A Longitudinal Study of Reading in Physical and Digital Environments
Claire Warwick, UCL

Set out to study reading and to address some of the dire statements that are out there around reading on the Web, such as ‘reading is dead’ and so on. In particular, they want to research how behaviour changes in digital environments. It was a five-year study (2009-2013) that repeats year after year, which is unprecedented in terms of studying online reading.

They studied masters students in UCL Department of Information Studies programs. They used a diary study method, where students kept track of everything they read, where they were, the medium, how long, time of day, and made comments about their reading. In all, there were 1261 episodes of reading (628 digital and 633 analogue), with 5.13/student on average (5.11 digital/5.15 analogue). Of the 216 total students, 171 were female, with both sexes having the same response rate (56%). Women had more reading episodes on average, however. Their results also showed that reading habits don’t necessarily follow age assumptions, in other words, there was no trend that showed that older students read more than younger students.

She stressed at several points that their data show that, in general, students are reading more in total, both print and digital, than they were in 2009. Also made interesting and amusing mention of what the numbers show about reading in bed. For one, it’s not a major place of reading, but the trend line for print/digital is all over the place.

Supporting “Distant Reading” for Web Archives
Kari Kraus, University of Maryland

Showed examples of countries that keep Web archives, but while they have many TBs of material, they don’t see much use (e.g.- Denmark 280TB but only 20 users between 2005 and 2013). As she noted by quoting Rick Prelinger at the outset, people find archives comforting, but rarely actually use them.

Their project attempts to rethink how Web archiving is done and relies upon tools developed by Google: MapReduce, BigTable, and the Google File System (GFS). These now have open source equivalents such as Hadoop and HBase. Their project (WarcBase) is built on Hadoop MapReduce and HBase, and sets out to counter the limitations of Wayback. Uses commodity hardware and open tools and is highly fault tolerant, which is necessary for such a large mass of data and processing.

Their system assumes that the harvesting actually occurs outside the system, i.e.- that one already has the mass of Web materials in hand, so to speak. Showed an example based on a Web archive of the 108th US Congress that had been compiled by the Library of Congress that they pulled into WarcBase. Was able to show both the Web pages as well as some nifty visualizations.

Canon, Value, and Artistic Culture: Critical Inquiry About the New Processes of Assigning Value in the Digital Realm
Nunia Rodríguez-Ortega, University of Málaga

The Web enables the rise of new actors that ‘compete’ with established entities. In her field (art history), the established entities are art museums, galleries, etc. She based her comments around the notion of canon and what she termed hypercanonization, with the differences being based around differences in scale and scope that the Web introduces. On sites such as Europeana Exhibitions, one sees a great deal of material, but it is hard/impossible to discern who created the exhibitions and what their critical standpoint might be. Google Art Project also plays a role here, and she showed how museum Websites refer back to it, which establishes some sort of authority for Google (this is a poor summary of her comments).

The other phenomenon she described is decanonization, which is bottom up and driven by users. She showed a Tate Gallery page with the ‘canonical’ presentation of a particular Picasso work, but then a wide range of examples that riff on the original work and create variant versions, etc. These create different contexts and meanings than that set out by the Tate.

She also suggested there is intercanonization, where institutions interact with users and adopt social media practices to enhance their sites and collections.

She closed by posing four questions:

• To what extent are these measures marketing strategies?
• To what extent do they produce real and accepted hybrid knowledge?
• Are these institutionalizing processes?
• Are art history institutions willing to share their position of authority?

Session 7, Friday

Scholarly Primitives Revisited: Towards a Practical Taxonomoy of Digital Humanities Research Activities and Objects
Luise Borek, Quinn Dombrowski, Christof Schöch

TaDiRAH – Taxonomy of Digital Research Activities in the Humanities – combined from DiRT (Digital Research Tools) and DARIAH (Digital Research Infrastructure for the Arts and Humanities). Also not much found on the Web!

Their work is pragmatic, not theoretical, and is intended to facilitate discovery. They also seek to balance “correctness” with common practice. They adopted work done by DiRT and DARIAH and began to think about how to adopt and expand upon what those two taxonomies had started. Also borrowed from the arts-humanities.net taxonomy, which they found a bit too narrow and technical.

What emerged was a basic structure that breaks into four areas: goals (short, closed list), methods (longer, closed list), research objects (open list), and techniques (separate open list). They developed their model through feedback via workshops, Skype calls, Google Doc comments, etc. Released v0.5 via GitHub.

Where can it be applied? They use it in DiRT, Doing Digital Humanities (a DARIAH bibliography), the DARIAH Teaching Resources Registry, the DHCommons project directory, and also on the DARIAH-DE Website. They’ve had some further requests from other projects to use the taxonomy.

In the future, they want to encourage usage, localize it (translate), apply revisions, and cooperate with more entities to drive credibility and visibility.

Introducing Digital Humanities through the Analysis of Cultural Productions
Everardo Reyes, Paris 13

Trying to introduce undergraduate students to using toolkits to describe cultural productions and lead them to the point of getting involved in designing tools to do this work. They select a cultural production, gather media and metadata, do some information processing, create their own analytical maps, and then analyze this with social and human sciences methods.

Showed an example, which was a student’s large Photoshopped .jpg that brought together the materials she found. He tells them not to use PowerPoint for their research and pushes them to use new tools. They encourage the use of a broad toolkit that consists of free and widely available tools.

What do they choose? Mostly albums and films, with the top artists being Michael Jackson, Amy Winehouse, The Velvet Underground, Lou Reed, Rihanna, and Beyoncé.

Beyond the Tool: A Reflexive Analysis on Building Things in Digital Humanities
Stéfan Sinclair, McGill

Read the talk, which was mainly written by Stéphane Couture, a graduate student at McGill.

Made the case for using IRC logs as a source material for studying how tools are developed. They’ve got an IRC log tool (IRCMine) that is in its extended alpha phase, in their words. Showed a brief demo, but it was hard to see due to the font size and the large room. At any rate, showed how the tool allows a researcher to drill down into a log file and isolate conversations.

Argued that building a prototype is a critical practice. For example, one element is that one must reflect on the material that is studied. What, precisely, is an IRC log and how is it constructed, how do people communicate, etc.? Another issue concerns the ethical aspects of interface design. For example, digging into logs like this raises issues around privacy and confidentiality, not least since their tool is intended to be Web-based. Design and coding practices must also be considered, as they influence how tools develop.

Let DH Be Sociological!
Andrew Goldstone, Rutgers

Using topic modeling he wants to argue that if sociological methods are being used in DH, then it should be more sociological. Much of this is driven by the need for DH researchers to work at scale and to cope with scale.

Outlined a practical project where he applied topic modeling to a corpus of literary journals to analyze how scholarship has evolved over time. Showed us how he narrowed the results to a usable scale where he could draw a sensible conclusion about how the scholarly discourse has shifted over time.

He then shifted to what he termed the polemic part of his short talk. He mocked the notion of “return,” i.e.- the notion that we can return to close reading from distant reading, or go back to the text. He argues we should not return from topic modeling. He argues that in DH we have some quantitative methods that tend to merge with the social sciences. Beyond that, the interest is shifting from the “single, sacred text” more toward the socio-cultural context. If that’s happening, he feels we should be more explicit about this convergence.

Literary Canon and Digital Bibliographies: The Case of the United States
Carolina Ferrer, Université du Québec à Montréal

Showed a graph of MLA-indexed publications that mention the notion of canon. While it stretches back to 1887 for the first mention, but in more recent decades there is an explosion of interest as documented by the number of articles that discuss the notion of canon.

Explained how she devised her data mining techniques and how she applied them to the MLA Bibliography to document the appearance of United States literature as a subject of scholarly attention. Essentially uses data mining to demonstrate the existence of a canon and the nature of this canon. By applying scientometric methods, she moves past subjective notions of canon. Moreover, the methods can be repeated and revised.

Session 8, Friday

Making Digital Humanities Work
Trevor Muñoz, Jennifer Guiliano

Talk arises from their work with the DH Incubator, which is a program to engage librarians in DH work. They see four major needs, one of which is to reflect on the nature of labour in DH. He covered the other three very quickly; perhaps the best summary is to say that he made an argument for critical analysis of our practices with regard to DH, rather than relying on or privileging anecdote.

The Digital Humanities Incubator ran during the 2013-2014 academic year and included librarians of all ranks. 42 participated in at least one workshop or exercise. The workshops included developing DH project ideas, finding data, evaluating tools, and developing projects.

She spoke mainly to the issue of faculty fellowships and what she termed a “conceptual flaw” in how these are realized. MITH adopted the model that ITAH at the University of Virginia had established. Their model provided a graduate assistant (10 hrs/wk), a $10K project budget (1992), office space/equipment, and consulting and sustainability for the life of the project. This is expensive, to say the least. Unsworth (then director of IATH) estimates that this costs hundreds of thousands per fellow.

These faculty fellowships tend to go to tenured or tenure-track faculty. In IATH’s case, these went entirely to tenured faculty (had some “star allure”). In short, this model reinforces traditional hierarchies of power and status and the quantification of output.

They had to cut some content, but closed by focusing on the where of digital humanities work. They really needed more time to do justice to what they were presenting (they did say it’s a gloss of what they expect to be a 25-30 page paper). I wanted to capture much more of what they were laying out, but they spoke at lightning speed to get in what they were able to cover in 20 minutes.

The Q&A after their talk was rich and active, but I was trying hard to listen and follow rather than take notes. Also, I was a bit preoccupied with my own concerns around the role of research in libraries and specifically in the work of librarians. I tossed out a feeble question to probe this, and their response made a lot of sense to me, but I need to have a much richer conversation around this. My main concern is that I see us (research libraries collectively and globally, not any particular library) failing to address some of the tasks and mandates that are right out there in front of us, and it often comes down to a question of resources. We’re never going to get a lot more money, so we have to point our resources in the direction of the critical work. Trevor responded to my question (think it was Trevor who said this) by saying (or this is what I heard) that it need not be a binary situation, where we either do a or b, but not both. In particular, encouraging librarians to do research and to engage in thinking as researchers can drive our practice and increase our ability to do the meaningful work we need to do. That makes sense to me in the abstract, but in my concrete experience making this happen is somewhat challenging.

Developing a Physical Interactive Space for Innovative Digital Humanities Exhibition
Jyi-Shane Liu, National Chengchi University

Gave a brief into to visualization and its history as a DH tool, noting that visualization has been established as a research process as it generates new questions, etc. Their work is to take visualization beyond the personal display and do it at a larger scale.

Their goal was to create a space in an old research library and establish its role in promoting and driving DH at the institution. They took a space that had been cleared of its previous function and reconceptualized it to serve multiple issues. They created a mixed auditorium and gallery; included a 12m projection wall and other features. Beyond hardware, they also wanted to embed visualization technology into the interior space design. He showed a video where that became apparent; the screens provide colour and richness to what could otherwise be a fairly arid space.

Project has been a major success and has become a popular space. For the library, it has allowed them to become part of an interdisciplinary group working around DH. He noted that they need to develop in-house content production capacity; there are many free tools and frameworks, but they take knowledge and experience to use well.

Digital Pedagogy is About Breaking Stuff: Toward a Critical Digital Humanities Pedagogy
Jesse Stommel, University of Wisconsin

Asserts that pedagogy should be at the core of what DH is as an academic discipline. As his slide put it, it’s not just a delivery device. Builds on the notion that teaching and research feed each other, as opposed to cannibalizing each other as many people feel.

Nice line: handle our technologies roughly. Don’t let technology drive what we do, but the other way around.