The Language of Hate Speech on Nigerian Twitter: A Discourse Analysis

Hate speech on social media platforms has emerged as a pressing challenge for democratic societies, and Nigeria's ethnically and religiously diverse public discourse on Twitter (now X) provides a particularly rich context for examining how linguistic features construct and propagate harmful group-directed language. This study analyses the linguistic features of hate speech directed at ethnic and religious groups in Nigerian Twitter discourse, examining a corpus of five hundred tweets collected during two peak periods of inter-communal tension in 2023 and 2024. Tweets were identified using a keyword-based collection protocol targeting known hate speech markers, with a subsequent manual verification step to exclude satirical and counter-speech uses of targeted language. Discourse analysis draws on critical discourse analysis and recent digital hate speech frameworks. Feature analysis identifies dehumanization metaphors, tribal slur deployment, in-group/out-group pronoun polarization, and imperative constructions directing action against targeted groups as the primary linguistic resources of hate speech construction in the corpus. Religious hate speech shows more complex syntactic constructions than ethnic hate speech, frequently embedding derogatory propositions within apparently neutral religious commentary. The study identifies a pattern of constructed ambiguity in which speakers strategically maintain deniability through indirect reference while communicating unambiguous hostile meaning to in-group audiences. Implications for platform content moderation and Nigerian hate speech legislation are discussed. Keywords: hate speech, Twitter, Nigerian social media, discourse analysis, online language.