‘Dark metrics’ and the mainstreaming of political extremism on Dutch-speaking Telegram: A comparative reading of platform affordances

Telegram is a social media and messaging platform known for its hands-off approach to content moderation. This makes it both a popular ‘mainstream’ platform and an attractive medium for discourse less welcome on other platforms, such as those of politically extreme actors. Additionally, owing to its anonymity and prominence in Eastern Europe, it has been linked to campaigns of Foreign Information Manipulation and Interference (FIMI), particularly by Russian state actors. The objective of the present paper is to investigate how Telegram, through both user-curated and algorithmic recommendations, facilitates connections between channels pertaining to mainstream electoral politics and those propagating fringe political discourses. To this end, we analyse the Telegram environment surrounding the channels of two far-right political parties from the Netherlands and Belgium, respectively: Forum voor Democratie (‘FVDNL’) and Vlaams Belang (‘vlbelang’). We specifically propose a comparative, ‘centrifugal’ reading of these channels, in which we trace outgoing connections to map the broader issue space to which they connect. On a methodological level, we argue that an analysis of the dataset based on outgoing links represented by forwarded messages – a popular method in Telegram research – is insufficient for a comprehensive understanding of this issue space and propose an approach centred around Telegram's new ‘similar channels’ feature. This under-researched algorithmic mode of amplification is to some extent based on shared audiences between channels, but the actual inner workings of this recommender engine remain largely opaque. Tracing these connections, we observe that both political seed channels exist in a context in which far-right, identitarian and conspiratorial discourse predominates. Moreover, channels concerned with geopolitical current events that propagate a pro-Russian narrative are present. We thereby conclude that Telegram's ‘similar channels’ feature affords new methodological opportunities that can enrich Telegram case studies by providing access to the platform's ‘ambient’ modes of amplification.