Send message to Author

Short abstract:

The article explores how algorithmic imaginaries transform the production of science communication content distributed on social media. Drawing on the analysis of algorithms in practice, I argue that it is the sense-making around algorithms that transforms content creation in science communication.

Long abstract:

This article explores how algorithmic imaginaries as ‘‘ways of thinking about what algorithms are, what they should be and how they function’’ (Bucher, 2017, p. 30) transform the production of science communication content distributed on social media. Drawing on ethnographic fieldwork on the development and production of a social science format on YouTube – a collaborative project between a German public broadcast service and social science academics – I explore how “digital experts” who act as strategic consultants in the project, interpret recommendation algorithms. More particular, I examine how they translate algorithmic imaginaries into requirements that impact content creation in social science communication. Drawing on the analysis of “actual practices surrounding algorithmic technologies” (Christin, 2017, p. 1), I argue that it is precisely the “sense-making” surrounding recommendation algorithms that transforms content creation in science communication. In more detail, I examine the consulting practices of "digital experts" and delineate three requirements that they derived from algorithmic imaginaries: (1) Include diverse audiences, exclude complexity (2) Address everyday needs of users, (3) Persuade first, explain later. While acknowledging the significant role of science within the YouTube platform, they advised to restructure the content according to their understanding of the recommendation algorithms (e.g., prioritising persuasion practices over scientific rigour) making themselves complicit with the platform’s interests. By focusing on algorithmic imaginaries, the paper proposes that it is not the algorithms alone that transform science communication content on social media but rather the context-specific interpretations of algorithms and the adjustment of practices towards this interpretation.