Linguistic Radicalisation of Right-Wing and Salafi Jihadist Groups in Social Media: a Corpus-Driven Lexicometric Analysis

Abstract Social media groups, for example on Facebook, WhatsApp or Telegram, allow for direct exchange, communication and interaction, as well as networking of different individuals worldwide. Such groups are also used to spread propaganda and thus allow for self-radicalisation or mutual radicalisation of their members. The article reports selected results from a research project analysing online communication processes of extremist groups. Based on data from group discussions in social media, corpus linguistic analyses were carried out, examining quantitative relationships between individual lexical elements and occurring regularities. To this end, four different corpora were built. These consist of data collected in right-wing and Salafi jihadist groups of a low or medium radicalisation level on Facebook and VKontakte via fake profiles, and of group communication in forums, messenger apps and social networks of highly radicalised persons, which were extracted from files of (e.g. terrorism) cases prosecuted in Germany. Quantitative linguistic analyses of social media data continue to be challenging due to the heterogeneity of the data as well as orthographic and grammatical errors. Nevertheless, it was possible to identify phenomenon specific sociolects that point to different levels of linguistic radicalisation. Based on the results of the analyses, the article discusses the prospects, problems and pitfalls of lexicometric analyses of online communication, especially as a tool for understanding radicalisation processes.