Digital humanities research has long been strongly textually oriented (Arnold and Tilton 2019). Increasingly, however, methods are being developed to incorporate the visual into DH analyses. This workshop will introduce participants to the basics of “distant viewing”: cutting-edge computer vision techniques in humanities research.
Like distant reading, these methods have proven useful to perform (historical) cultural analyses at a macro-scale. They can be used, for instance, to analyze the relationship between text and image in the nineteenth-century transnational press, to map the circulation of images in internet culture, to do visual stylometry (authorship attribution), or to study pictorial traditions, genres and motives in thousands of paintings. However, technological gains in computer vision go beyond merely increasing the scale at which we can research cultural phenomena. They also have the potential to change how we understand the cultural work of the visual vs. the textual, as they challenge traditional views of how images are consumed, cognitively processed, and assigned meaning (Moretti and Impett 2017; Arnold and Tilton 2019).
The day is intended for early-stage researchers who would like to learn about the principles, possibilities and pitfalls of research methods based on computer vision. Learning more about this may complement what you already know about digital humanities methods of ‘distant reading’, or help you think about how your current research questions could be operationalized at the larger scale. Or it may inspire you to formulate new project ideas. In any case, by the end of the day, you will have a sense of 1) what sorts of new research questions you can formulate with these methods, 2) what the workflow of this research looks like and 3) where to start: what are some collections, at the KB and beyond, which you can begin to explore using these techniques.
The day will start with a keynote by Leo Impett, whose work applies computer vision to analyze Aby Warburg’s Bilderatlas. After this, the trainers will give brief presentations on their own research, which are intended to inspire you to look at the possibilities of these methods for your own research interests. The afternoon consists of a hands-on workshop for max. 20 participants, in which we will go on a guided computational exploration of a dataset using the programming language Python. The day will also offer ample opportunity to discuss research ideas with trainers, peers and members of the KB team.
More information soon