• Research data

The internet has brought on a variety of new forms of literary expression: from self-published novels, over complex hyperfiction forms or automatically generated texts, to large forums of variegated fanfiction. A removal of authorities (”Wegfall der Kontrollinstanzen” (Simanowski 2005)) allows writers to release texts which due to their style or content most probably would not work in a traditional book setting. This can reach from curiosities like the mid-2020 release of the Kindle eBook Kissing the Coronavirus, which became an internet sensation and has been deemed as a ”masterpiece of bad erotica” by a Reddit user (see ParadoxicallyItWas 2020), to the complex system of fanfiction writing. Systemic approaches to digital literature, especially in German studies, are rare (see Ries 2021), though studying the commonalities of literature on the internet may give valuable insights on readers and writers interests. ”Fanfiction R-rated - A Quantitative Study of Sensitive Topics in German Fanfiction”, deals with fanfiction by setting it into the context of literary writing on the internet and describing the effects of the previously mentioned low boundaries for publication.

Fanfictions are literary texts of appropriative-derivative or transformative nature that are produced in a fandom by fans and are based on e.g. novels, films or TV-Series (see Stemberger 2021, p. 10). Due to their fannish nature, fanfiction writers operate from a standpoint of adoration of and frustration with a specific source material (see Jenkins 1992). This results in a variety of stories that try to expand the source material and/or seek to rework it to reflect authors’ needs. Apparently, artistic freedom is used to full capacity, particularly, as Catherine Tosenberger describes it, ”with regard to non-normative and taboo forms and representations of sexuality” (Tosenberger 2014, p. 17). The research objective of this thesis is thus to study the effects of the especially low boundaries for publication in fanfiction and what role more sensitive topics play in contemporary fanfiction. The study at hand will look at how to examine these phenomena on a corpus of German fanfiction written between the beginning of 2020 and the end of 2021, which has been collected by the LitLab of the TU Darmstadt. To manage the size of this corpus, examine trends, and integrate the rich metadata the data set offers, quantitative approaches are most suitable. The project therefore will include metadata analysis as well as quantitative textual analyses.

Theoretical background will be laid out in chapters 2 and 3, setting this study in the context of wider studies of literature on the internet and describing the mechanisms underlying the process of fanfiction writing. Following this, the two key hypotheses will be explained in chapter 4. Thereby the first hypothesis says that in the fanfiction corpus at hand trends reflect the use of sensitive topics and the second hypothesis says that word meaning is altered in a way that blurs the difference between sensitive and non-sensitive topics. As a first analytical method, the metadata will be evaluated statistically in order to create a useful cut of the corpus and to describe how age restriction and genre tags hint at the discussion of sensitive topics (see chapter 6). Then, chapter 7 will describe how word embedding models can be created and used to examine the use of certain keywords in a corpus. An exemplary analysis will show how certain keywords express sensitive and non-sensitive topics and how the difference between these areas can become blurry. The results will be evaluated (chapter 8) and concluded with an outlook (chapter 9).