From algorithmic value to journalistic practice: How Swedish public radio made algorithmic distinctiveness socially actionable

Abstract

While recent work has charted how journalistic values are inscribed into recommender systems, less is known about how such systems reshape news work after their implementation. Through an ethnography conducted at the Swedish public service broadcaster Sveriges Radio (SR) between 2018 and 2023, this paper examines how distinctiveness was operationalized through an algorithmic system for news distribution. The study shows how SR made their public service obligation to remain distinct actionable through a classificatory schema for metadata annotation that prescribed the production of radio news intended to evoke a sense of radiophonic authenticity – sonic expressions conveying place-specificity and presence. Post-deployment, reporters oriented their work toward producing such authenticity through less deskbound reporting and creative sonic practices. The paper argues that the algorithmic system’s primary organizational effects did not stem from computational sorting, but from how its classifications valorized certain practices and embedded journalistic work in a feedback loop of data that facilitated continuous auditing and managerial interventions. These findings highlight the plasticity of public values in algorithmic environments and show how their production is accomplished through algorithms and people acting together, mediated by classificatory schemas that can substantially reorganize journalistic workflows.