#HitTheSlopes : En kritisk diskursanalys av algospeak i drogrelaterat TikTokinnehåll

Sammanfattning

This study critically examines how drug-related discourses are formed on the social media platform TikTok, within a moderated and algorithm-driven environment. Using a qualitative approach with a focus on critical discourse analysis, the study analyses content in the form of videos and comments on TikTok published within the last six months. The aim of the study is to critically investigate how drugs and drug use are represented on the platform, what discursive strategies are used to communicate about drugs, and how TikTok's algorithms and moderation can be understood as factors shaping these discourses. In doing so, the study aims to contribute to existing research on how discourses are constructed in relation to platforms' technical and communicative conditions in the digital public sphere. The study's point of departure is that social media does not function merely as a neutral communication channel, rather, platforms actively contribute to shaping which expressions are given visibility. The study draws on a theoretical framework combining critical discourse theory, critical linguistics, social constructivism and the concept of affordance. The analysis is conducted with inspiration from Norman Fairclough's three-dimensional model for critical discourse analysis, which enables an examination of the relationship between text, discursive practice and social practice. The findings show that drugs and drug use on TikTok are frequently represented through playful, aestheticising and indirect expressions rather than explicit designations. Coded expressions such as algospeak, metaphors, emojis and slang are recurrently used to communicate about drugs without employing words that users assume the platform will moderate. For instance, cocaine is represented through expressions such as 'slopes' and other snow-related metaphors and emojis associated with winter. These linguistic choices alter the connotations of the content by connecting drugs to humour, lifestyle, community and aesthetics rather than counter discourses such as criminality, dependency or risk. The study further demonstrates that comment sections function as a central arena for collective meaning making, where users reproduce and reinforce the discourses expressed in the videos. The study also shows that the use of algospeak can be understood as a discursive strategy shaped in relation to TikTok's moderation and algorithmic logic. Content creators adapt their language to reduce the risk of content being restricted, removed or losing visibility on the platform, something that previous research has also identified as a consequence of platforms' automated moderation systems (Steen et al., 2023; Gomes & Sultan, 2024; Tillewein et al., 2024). The findings indicate that TikTok's algorithms and moderation function as communicative conditions that shape how drugs can be expressed and understood. The platform's lack of transparency regarding how the algorithm operates simultaneously contributes to users developing their own assumptions about how the system functions, which in turn influences their content and language use. Drawing on previous research on algorithms, moderation and digital platforms, the study argues that TikTok's technical structures both restrict communication about drugs and contribute to shaping it. As explicit formulations risk being limited while coded expressions are granted visibility, discourses about drugs risk appearing less serious and more socially acceptable. The study thereby illuminates how the interplay between users, algorithms and moderation influences which representations of drugs are granted visibility and legitimacy in today's digital public sphere. At the same time, the study highlights methodological challenges associated with analysing moderated content, where deleted or restricted material may have influenced which discourses are visible in the data. That the few critical comments present in the material are not reproduced or built upon by other users nonetheless suggests that the normalising discourse is dominant within the context analysed. Keywords: TikTok, Algospeak, Critical discourse analysis (CDA), drugs and drug use