SDH »

SEMI 2009

Society of Digital Humanities/SEMI 2009 Conference Report

Note: this is being written during the conference. There will be gaps as my battery runs out and I have to present.

These are my notes on the 2009 Society for Digital Humanities conference.

Michael Best: The Shoulder of Giants

Michael Best was awarded the 2009 SDH/SEMI Award for Outstanding Achievement and gave the opening keynote that wove his current interest editing King John and reflections on computing and the humanities.

King John, the character, is reactive and rigid while the Bastard is flexible and gains from his illegitimacy. The challenge of a editorial representation of both play and character can be seen in both the strengths and weaknesses of digital techniques. One strength of the digital medium is that the ISE can publish both historical texts like The Troublesome Reign of King John and related performance texts linking them.

- Historical texts - Relevant historical documents - Performance texts - Multiple-media resources

Shakespeare's sources show contradictory positions about King John's reign.

The history of the play and its performance is likewise one of very different interpretations. King John is hard to like and hard for an actor to project. Now it is less known play, but this wasn't always the case. Major 19th actor managers liked King John and performed it regularly. King John could be seen as an allegory for England. Often

Often productions emphasized the pomp and majesty with hundreds of extras. The birth of the cinema superseded the spectacle of theatre and Best showed a clip from one of the first cinematic interpretations of Shakespeare. The multimedia capacity of the digital edition allows us to situate artifacts in their historical context.

A modern illuminated text. Best is experimenting with design, usability and visual language. The ISE is experimenting with various ways to visualize parallel texts using colour coding and animation (where variants are flipped in place.) The ISE is exploring a glossary function that links from terms in the text to maps and glosses.

The first cinematic representations placed the camera where a theatre audience would be. Likewise in digital editions we may doing the same thing. Can we change the view. How do we deal with the situation today where every browser can Google things and come in from searches. Users don't necessarily differenciate between the location field and search engine space. What are the effects of Google with sites ranked by popularity.

Best talked about a "sociable text" that we have theorized about, but have yet to exploit appropriately. The viral character of the social tends to focus on superstars rather than less well known works. Would King John be of interest to a crowd. "The Performance Chronicle" is the ISE's experiment that invites reviews of contemporary Shakespeare theatre. It has different levels of review of the reviews to avoid unsuitable materials.

The digital medium allows materials to be maintained as new research comes out. This is a challenge as it takes long term attention and it challenges our desire for stability. Best argued that the same energy should be devoted to maintenance as to software upgrading, but we have to avoid the danger of digital incunabula stranded in incompatible technologies. Standards are one way to try deal with change, but standards like the TEI may not scale for the complexity of intransigent documents.

The Bastard was willing to learn and adapt and that was his legitimacy. Best closed by commenting on how the variety of papers at our conference, just like a big red India rubber ball, are a sign of the dynamic and ludic flexibility of the digital humanities. The electronic media has a ludic element. The ISE has played with games like a big red India rubber ball.

In an academic game the goal will be to become an emeritus professor tending a virtual garden - a second Illyria.

Kirsten Uszkalo: What was she thinking?

Uszkalo presented a paper on her new research on cognition that came out her research on witchcraft (now devils and sex) and through the [http;//www.monkproject.org/ | Monk project].

"If we account for the use of crystallized and fluid intelligence in interpreting visualization and using tool sets we would be better able to support flexible and fluid intelligence"

Raymond Tallis argues we should avoid applying with a broad brush the premature conclusions of cognitive science to literary studies.

How do we develop and imagine tools?

• We pose a research question.
• Or we imagine a neat interface and then work back
• We imagine results or management

Kirsten argues that we should think about how tools help us think. User experience isn't sufficiently addressed in development. We don't ask about affect - how do people feel about what they are using. Social networking tools may enable us to process faster than we can empathize.

How do tools help us think?

• They can help us access analytics and understand them
• They can give us aesthetically pleasing visualizations that we explore
• They help us find patterns across evidence
• They can help us with thin-slicing functions - rapid cognition

But meaning is problematic for datamining because of the absence of explicit mention, because of allusion and other literary techniques, because of slippage, and because it isn't sensitive to the placement of ideas.

Fluid intelligence is different from domain knowledge. It is the intelligence that deals with novel problems - it is linked to multitasking, pattern finding, and thinking independent of a defined content or task. It is the ability to manipulate abstract symbols. We should think about how to support this type of thinking in addition to the thinking of an expert in an area. Fluid intelligence is what we use negotiating our roles and trends in a field.

There is a link between "open people" to fluid intelligence. If we could develop tools for open people we could better reach out. Uszkalo finished by arguing that developing for fluid intelligence would make us smarter.

Susan Brown: All of Orlando's Interfaces

Susan Brown talked about work they are doing on the interfaces to Orlando - orlando.cambridge.org . They ran an study on their existing interface. They did log analysis, semi-structured interviews, study user study, scholarly reviews, and web-based questionnaires.

From the logs it looks like people are search most of all for "people" and then for "chronology", though they also surprisingly use the "tag" searching. The linking is not being used much and neither is the text search.

From the interviews the "rabbit holes" that one can go down and explore are one of the advantages, though for others the possibilities for distraction are a problem.

Students respondents thought the site looked boring and academic, though this could be a good thing. Students were turned off by the rich results of searches. They seem to want Google.

The reviews are long and from domain experts who took the time to learn the system. They are positive about the interface, but had suggestions like more links outside Orlando. They also talk about Web 2.0 features that could enhance the interface.

Brown concluded with thoughts about the difference between novice and expert uses and the challenge of supporting both.

Alejandro Giacometti: Tell Me What Those Blogs Are About

Alejandro talked about a study of reliability of tagging. User tags are cheap and can be useful when accumulated.

Tags, however, can be inconsistent, they lack a hierarchy, and associations may be obscure depending on motivations of the tagger.

They looked at different algorithms for automatic tagging and had an experts tag a collection. Then they had others rank the different types of tagging (user tags, algorithmic, and expert.) The consistency was low. There was no significant measure of inter-rater reliability between the three raters.

After his talk we had a spririted talk about tagging and how unreliable it is. We all seemed to want user tagging to be reliable and talked about ways to make it so.

Carolyn Guertin: Rethinking Authorship and Text

Does it make sense to talk about writing, reading and distribution with new media? Authorship is about property - digital culture occupies a space of free culture. Authorship is a collaborative act. It is only recently (18th century) that solitary authorship has been romanticized and legalized. The printed book and copyright fenced of the commons. The author is now originator and owner even though copyright was invented to protect publishers. Authors tend to then present themselves as

Digital narratives are deterritorialized spaces. See Access by Marie Sester. The third space of authorship is Peer2Peer Networks. The first space is the oral storyteller. The second is the romantic solitary poet. The third is a new space of hybridity. She talked about Twistori that harvests tweets that include phrases like "i feel". This is authorship as territory or landscape. They make social relations art. There is no art object, what is privileged is others text mashed up. They turn the commons into art.

James Boyle talks about a drive to privatize the public resources as a second enclosures movement. IP law is shrinking what we can draw from which is part of authorship. The open movements is trying to fight this. Stallman and others are trying to make sure that IP remains free to be reused and played with for creation. Lessig's Creative Commons movement is similarly trying to preserve some freedom. We need new collaborative models for the digital world. See Perry Bard's 2008 Man With a Movie Camera: The Global Remake: "Vertov in the age of YouTube". An experiment in database cinema.

We should worry about the spread of DRM and copyright restrictions - they impinge on fair use for research. Digital art works and open movements that resist should be important to us.

Martina King: Visual Communication: A Study of Library Virtual Reference Icons Online

Martina talked about the design of icons. "The value of icons lies in their ability to communicate with visitors or users in a more intuitive way." Icons are important especially for the small screen of smart phones.

Martina then talked about virtual reference services and how icons can be designed to communicate. One problem is the variety of iconic representations for the same service (virtual reference chat). Can the iconography be standardized? How do standards emerge?

Martina took a three pronged approach. First she assessed different library icons. Then she interviewed library staff about icons and finally she interviewed students about the icons. She then compared the student perceptions to the library staff views. Libraries are creating icons in-house which is why they are so different.

When ISO tests their symbols they get a ratio of acceptable to unacceptable messages communicated. They look for an acceptable ratio as no icon is unambiguous.

The majority of icons in her sample don't conform to ISO standards and guidelines. Staff are often aware of guidelines but ignore them - why is that?

We had an interesting talk about what made for good icons and how some of the most horrid tested well.

Keith Lawson: Using social networking sites in undergraduate teaching

Keith started with some principles:

• Connect but keep your distance - students don't really want to communicate with profs through social networking sites (SNSs). Students use SNSs for immediate negotiation of social relationship with peers. The don't want us in this space. SNSs are collaborative texts where some things are out of their control so they may not want us looking at their walls. We need to respect the ecology of the SNSs and keep our distance. How do we do this? See Courses plug in which lets us post stuff, schedule, create groups, carry on discussion and so on with people who are not our friends.
• Be a model of collaboration and encourage academic use - The reason for this is to increase students' collaboration and IT skills. It is increasingly important for students to learn how to collaborate using IT outside school.
• Teach students how to use (SNSs) - We can teach them to use SNSs in a different way than just for immediate social use. We can direct them to other tools where they can build on SNSs to do work.

Keith mentioned The ECAR Study of Undergraduate Students and Information Technology - the results seem to be fairly stable over time except that now more students are beginning to say that IT doesn't improve their learning. Why is this? Are students getting tired of all the IT and wanting education to be more humane? Or could they be feel that the IT we use in teaching (PowerPoint and Blackboard) are outdated?

We had an interesting discussion about how youth are using SNSs and perhaps need a space to "act out" and show themselves taking risks. Should we try to teach students to develop a "responsible" persona that won't harm them when they apply for jobs? Or, is it important to let them have their space.

Paul Dyck and Ryan Rempel: Expanding Textualities in the Undergraduate Classroom: A Course on the “Digital Word”

Paul and Ryan ran a course together on digital textuality starting with a book history approach. Ryan designed a series of markup exercises starting with an open ended question task about marking anything not visual. The idea was to get students to want XML rather than teaching it first. Interstingly they didn't teach XSLT but had the students work with a "programmer" (Ryan) to describe what XSLT transforms they wanted.

Ryan walked us through http://tei.pauldyck.com - a really extraordinary online tool that uses mostly JavaScript and XSLT on the client side. They have TEI texts woven with tools.

Stéphane Lévesque: "Why can't you just tell us?" Learning to think like historians with the Virtual Historian

This paper was a about a project with high school students using the Virtual Historian.

Technology calls us to rethink the point of teaching. It is hard to integrate IT into the classroom - itself a technology. In the old model learning history was "getting the story right". Now there is the potential for a more dialogical model where the student works with evidence and tools. Stéphane talked about two types of knowledge:

• Substantive knowledge - facts and dates
• Procedural knowledge - methods and tools

They ran an experimental study to see if IT (the Virtual Historian) would improve outcomes both in terms of substantive and procedural knowledge (and epistemic.) The Virtual Historian is based on cases where the student has a "mission" involving a contested landscape. The case they used was Dieppe - was it useless slaughter?

They ran the experiment on class groups with and without IT. Both classroom and VR groups both learned, but the students using VH did better with papers and had better epistemological improvements.

Textbooks push the idea of history as stories to master. The VH pushes them to think about history as the making of interpretations and issues of authority.

After the talk we discussed inquiry based learning and whether IT was essential.

Panel on the Future of Graduate Education in the Digital Humanities

A very large group of us were on a panel about graduate education. I talked to the question:

How do we make theory (in humanities computing)?

I argued that in Willard McCarty's "methodological commons" view the theory and the theoretical questions come from the neighboring disciplines and what we do to create models to help answer these. This means that there is no theory in humanities computing unless it is a theory of methods. I'm not sure this simple story is enough, but nor do I want there to be some theoretical canon forming process so we get our own theory. Instead I suggested two ways that we might think about:

• Theoretical Fabrications - where we think about theory through fabrication and modeling. We ask about theory by trying to make theoretical things.
• Theoretical Commons - where we think about theory formation as a social process that we can help support through social network technology.

Others talked about how difficult it is to wear two hats (traditional disciplinary and digital humanities.) There was discussion about the place of programming and technical skills. A student nicely pointed out that what they learned at U of Alberta in the MA was to manage and be part of projects. They see themself at the interface of different traditions and skills - able to talk across the boundaries.

Implementing New Knowledge Environments

Ray Siemens, Claire Warwick, Teresa Dobson, Alan Galey, Richard Cunningham, and Stan Ruecker

Wednesday morning started with a panel on INKE (Implementing New Knowledge Environments) - a new SSHRC MCRI. Ray began by giving us a context for the project.

We have an exciting future for e-books and e-textuality, but an inconvenient present. We have a problem "finding" the humanities. We have an increasing variety of devices for reading and increasing variety of content providers. Readers are sophisticated and computationally empowered. The humanities are not always involved, but we have the capacity to imagine new knowledge machines reflecting back on the history of the book and reading.

Some of the things we have to offer:

• Evolution of reading and writing
• Mechanics and pragmatics
• Strategies of reading and organization
• Computational possibilities

The INKE interdisciplinary areas are:

• Textual studies
• User experience
• Interface design
• Information management

Alan Galey talked about textual studies. We need to ask again what a book is. Is it a machine? Is it similar to the computer as a machine. For a textual scholar a book is a synthesis of the material, expectations, traditions of interpretation and distribution.

• History - text as agency and power
• Literature - text as tradition of interpretation
• Bibliography - text as artifact

Textual studies studies process at different moments in the history of the book rather than just the product. We focus on particular cases but also look for patterns across time.

Claire Warwick talked about user experience (UX). Information seeking systems are often designed for scientists and not for humanists who have very different seeking behaviours. In the 90s there was a rhetoric of "beyond the book" as if books and e-textuality were in competition. Now we know that the book has not been killed by the hypertext. The User Experience group will be look at reading (and writing) in real situations. UX wants to study the

I couldn't help think that the first way we encountered electronic text is through the billboard. What was the first billboard? Did people think of it as electronic text or as a variant on neon? What was the first pixel board.

Stan Ruecker talked about the Interface Design group. (I am part of this group.) We will be drawing heavily from the other groups and trying to generate interesting prototypes.

Ray Siemens talked about the Information Management group. This group will bring Synergies and CRKN together and TAPoR too. They are trying to develop the best of analysis to the large body of knowledge. We have lots of data and trying to figure out what to do with it.

Lynne Siemens talked about the challenges of the MCRI program objectives. It is hard to be a very collaborative project. INKE has been setting up an administrative governance document and asking people to commit to deliverables.

Ray finished by giving us an image of what we are talking about - the small devices from the iPod to the OLPC. He talked about impact. Can the humanities find this problem worth engaging? Can we find interdiscplinarity engaging? We wonder if our results will be social applicable or embraceable?

The History and Future of Reading

Christian Vandendorpe, Richard Cunningham, Alan Galey, and Jon Saklosfske talked about modes of reading.

Christian talked about a new ergotive reading that includes the potential of writing (wreading) in a way that few used to have access to. It used to be that a few wrote and many read. The authority of the book began to fail in the 60s with new theories of reading. This transition was accelerated by the digital. The reader can build their own context out of atoms in an ocean of the web.

Richard talked about ArchBook a wiki about the architecture of the book which is envisioned as being an online encyclopedia about the features of the book - features like openinings, citatations, substrates, and inscriptions.

Jon Saklosfske talked about the William Blake Archive and new ways of presenting it. He showed a neat radial visual exploration tool. He uses thumbanils of the manuscript pages. He has ways of resorting the pages into rings. You can drag them out, zoom in and out, and create visual links. I thought it was quite elegant. A mix of the Light Box and the Mandala project. I want it for JiTR.

There were interesting questions about the social elements. Why would people need permission to contribute to the ArchBook? I think there is a lot of anxiety about opening things to just anyone. In my experience the problem is not amateur contributors but the spambots.