Multimodal performance of individuals on the web
How do people proposition themselves on the web, how are they presented by others? The study entitled 'Multimodal performance of individuals on the web' analyses this from the cultural studies angle using a Russian person in the Russian-speaking internet.
Along with texts, images play an integral role in numerous cultural contexts, also in the internet: images and videos are copied, edited, shared and then turn into what we call memes. This produces a huge number of images and videos, which elude any traditional 'close reading' on account of their sheer bulk. At the same time, the images and videos are available in digital format, and some of them have been annotated with metadata by the previously used web platform (e.g. YouTube).
On this account, image memes and viral videos lend themselves admirably for use in quantitative methods. In a sample study, a new distant watching method is being developed using knowledge discovery. This new method places the focus on the image and its content in a quantitative study for the first time.
A state-of-the-art regional convolutional neural network (RCNN) is being trained using specific pre-selected symbols in videos so that it can recognise them automatically in a large video corpus. This will allow the image content of videos to be automatically captured and quantitatively measured for the first time. Depending on the (co-)presence or absence of symbols, varying conclusions can then be drawn about the content of a given video.
This study examines how the Ukrainian nationalist Stepan Bandera is received. This reception encapsulates the ambivalence of Ukrainian memory culture and repeatedly sparks controversy in the current Ukraine conflict: for the pro-Russian camp he is a fascist and mass murderer; his followers are branded Banderites and equated with fascists. For the Ukrainian nationalist faction, Bandera figures as an idealised hero who fought relentlessly for the country's independence. New media are being extensively used to advance that view of Bandera that each side prefers.
A first survey revealed that this instrumentalisation pervaded all major digital media and had been deeply ingrained even before 2014. The project builds on this groundwork; we compare the YouTube corpus before war broke out in Ukraine with a corpus of today in order to show whether and, if yes, how the Ukraine conflict has changed the previously existing modes of instrumentalisation.
Alongside instrumentalisation for propaganda purposes, the 'post-memory' level (Hirsch 2012) also merits consideration. Marianne Hirsch uses this concept to describe an encounter with a traumatic past that was not endured in person. Visual media play a decisive role in this because they are emotionally more immersive than texts. What needs to be clarified, both qualitatively and quantitatively, is how these emotional components can be taken into account in distant watching.
Team: Bernhard Bermeitinger, Gernot Howanitz, Erik Radisch