On 5th March 2019, a full-day workshop entitled Twin Talks: Understanding Collaboration in DH took place during the 4th Digital Humanities in the Nordic Countries Conference. The event, which was organised by Steven Krauwer and Darja Fišer, brought together 28 researchers involved in interdisciplinary projects where humanities scholars and technical experts closely collaborate in answering humanities research questions in various disciplines. Hence, all eight papers presented at the event were co-authored by one or more humanities researchers and one or more digital experts who work side by side on a research problem. The aim of the workshop was to get a better understanding of the dynamics on the digital humanities work floor. The approach chosen was based on presentations of real-life examples of such collaborations.
Why research questions should come first
The programme started with an invited talk by Mikko Tolonen, who is the principal investigator of the Helsinki Computational History Group – a team of historians, computer scientists, and digital humanities scholars. The group studies European public discourse between the seventeenth and nineteenth centuries by combining historical and computational approaches. In his talk “Why humanities research questions should come first? Reflections on different kinds of collaboration in digital history”, Mikko explained that in the evolution of digital history research, one can distinguish stages of increasing complexity and depth, ranging from “checking historical data” to “finding out about historical data”, organising and classifying historical data”, “modelling historical data” and “understanding historical data”. He emphasised the importance of metadata by pointing out that, e.g. when studying intellectual history, it might be good not to start from trying to understand meanings and concepts and how they change, but instead to start by investigating the channels along which they travel: studying first the movements of vehicles, not the cargo.
The talks that followed all contained the following three components: (i) a presentation of the humanities research problem and its solution, (ii) a presentation of the technical aspects of the research, and (iii) a report on the collaboration experience itself, including the obstacles encountered and recommendations on how better training and education could help to make the collaboration more fruitful. For instance, Martijn Kleppe, Thomas Smits and Willem Jan Faber discussed the use of computer vision techniques to automatically classify historical newspaper images and presented their individual experiences in the collaboration on the basis of their research backgrounds as a humanities researcher (Smits), a research software engineer (Faber), and a digital scholarship advisor (Kleppe), concluding that the teamwork was directly beneficial for their main research goals. In a rather novel field of application, Konstantin Freybe, Florian Rämisch, and Tracy Hoffman discussed the development of a method and tool negotiation workflow used in their research of Japanese video game franchises (e.g., the Metal Gear Solid series) from the interdisciplinary perspectives of information science, cultural studies, Japanese studies, and information science. These and the rest of the papers presented at the workshop can be read in the proceedings.
Bridging the divide
The workshop ended with a round table discussion with all the participants in order to summarise the lessons learned from the presentations. These pertained primarily to the fact that the workshop served as a platform for the presentation of illustrative collaborations that clearly show the ways in which qualitative research takes advantage of new technological developments. This does not mean that there was no room for critical reflection. One of the observations made was that humanities research projects need to learn to articulate better what computer scientists could get out of such a collaboration in order to stimulate their involvement beyond mere technical implementation. As an example, one of the computer scientists in the room explained that the analytical challenge of working with humanities data is very high, as the information is often very pluriform. Apart from the message, another challenge will be to find the right channels to establish a stronger connection. Closer collaboration with industry would be one of the logical steps to bridge this divide.
If the TwinTalks demonstrated anything, it is that this will by no means be the end of the conversation on successfully bringing together humanities research and computational capabilities. The presentations have clearly demonstrated that technology is quickly becoming indispensable when tackling data-intensive humanities research questions, and that facilitating smooth collaboration between people with humanities and technical skills is a key factor. All this should be taken into account by those who develop training and education programmes.