|
Drawing on the example of Marie-Luise Vollbrecht, our BMBF-funded project KoKoKom investigates whether—and how—participants in polarized discourses like the one on the (non-)binarity of gender still establish a common ground. To this end, we have compiled a corpus (992,414 tokens) of social media posts and comments (YouTube, Reddit, Instagram, X, Facebook), transcripts of relevant YouTube videos and television programs, and texts from the German reference corpus DeReKo (Kupietz et al. 2018) across the period 07/2022–10/2022 containing the word Vollbrecht. One methodological focus involves contrastive analysis of word embedding models, comparing a model trained on our project corpus with one trained on DeReKo —the largest German reference corpus (see Fankhauser & Kupietz 2017; 2019 for details). This comparison reveals, for example, that the nearest neighbors of the word Biologin [biologist] have shifted substantially: instead of other subdisciplinary titles like Ökologin [ecologist], they now refer to Vollbrecht’s qualification as Wissenschaftlerin [scientist], reflecting a rhetorical strategy in our corpus, where Biologin is used to underline Vollbrecht’s authority. In our poster presentation, we will introduce the KoKoKom project, describe the corpus and methodological background, and put our initial analyses and findings up for discussion.
|
