Facebook has faced growing criticism regarding its handling of hateful user-generated content (UGC) with research revealing how the platform can foster both covert and overt racism. This research has tended to focus on racist content while relying on abstract references to the general logics of social media platforms. In this article we consider how Facebook shapes the production of racist discourse in more concrete ways by integrating a concern for the platform’s architectures and affordances within a broader analysis of the immigration-related discussions of a large Swedish Facebook group. We combine a quantitative topic modeling of a large data set of the group’s UGC with a qualitative critical discourse analysis (CDA) of a sample of that data set. Our findings show how Facebook enables and influences various discursive strategies of identification and persuasion—within which covert and overt racist discourses are embedded—through processes of cybertyping, role-playing, crowdsourcing and (counter-)reaction.

中文翻译：

---

为瑞典站起来？一个反移民Facebook团体的种族主义话语，建筑和嗜好

Facebook对其处理可恨的用户生成内容（UGC）的批评日益增加，研究表明该平台如何促进秘密和公开种族主义。这项研究倾向于关注种族主义内容，同时依赖于对社交媒体平台一般逻辑的抽象引用。在本文中，我们考虑了Facebook如何通过更广泛地分析瑞典大型Facebook集团与移民相关的讨论中整合对平台架构和服务的关注，从而以更具体的方式塑造种族主义话语的产生。我们将小组的UGC大数据集的定量主题建模与该数据集样本的定性关键话语分析（CDA）相结合。