Unveiling Hidden Agendas in Pakistani Media: A Corpus-Assisted Critical Discourse Analysis of Crime (Mis) Reporting

Abstract

This research aims to unveil hidden agendas in Pakistani media by analysing crime news reports using Corpus-Assisted Critical Discourse Analysis (CADS). Based on a cleaned corpus of 20 news reports published in Dawn and The News throughout the calendar year 2025, the study has applied Van Dijk's (1993, 1998) socio-cognitive model of Critical Discourse Analysis in three analytical dimensions: the micro-level analysis of discourse structure, the meso-level analysis of social cognition, and the macro-level analysis of societal power and ideology. AntConc (Anthony, 2022) was used to generate frequency lists, bigram clusters (N-grams), collocational patterns, and concordance lines from corpus data, which were then further analysed qualitatively. Results depict a systematic trend of institutional over-lexicalisation with the attribute of police (f=103) and crime-related words prevailing in the corpus. In contrast, victim-related lexis is significantly marginalised (victim, f=3). Police-generated statistical statements are epistemically neutralised by the amelioration narrative, which is realised through bigram reductions in, declines in, and drops in. Constructing perpetrators follows demonising collectivist vocabulary, and authorities are heroised with heavy clusters of speech-attribution. Collectively, these discursive practices are interested in building a collective mental model that both normalises punitive state power and silences structural analyses, in addition to murdering victims' agency. The research offers empirically sound and theoretically grounded perspectives on media, crime, and the development of power structures in South Asian societies.