CORPUS APPROACHES TO ANALYZING DISCOURSE IN MEDIA ENGLISH

Abstract

This paper examines how corpus approaches can be used to analyze discourse in Media English, with particular attention to how news and other media genres construct social actors, evaluate events, and circulate ideological meanings through recurring lexical, grammatical, and phraseological patterns. Building on developments in corpus linguistics and corpus-assisted discourse studies, the study treats large-scale textual evidence as a means of grounding discourse interpretation in reproducible patterns while retaining sensitivity to context and communicative purpose. The proposed analytical logic integrates keyword analysis, collocation and colligation profiling, concordance-based pattern description, and dispersion-aware comparisons across outlets, genres, and time periods. Special emphasis is placed on methodological choices that matter for media discourse research: corpus design and representativeness, reference corpus selection, normalization of frequencies, handling of multiword expressions, and the triangulation of quantitative outputs with close reading of concordance lines. The paper also addresses recurrent challenges in media corpora, such as topic confounds, quotation practices, and editorial house styles that may distort inferences about stance or ideology if not controlled. Using illustrative analyses from contemporary English-language media texts, the study demonstrates how corpus techniques can identify systematic evaluative framing through adjectival and nominal patterns, reveal the discursive construction of legitimacy via attribution verbs and evidential markers, and map the reproduction of stereotypes through stable collocational networks around social groups. The findings show that corpus approaches are especially effective for detecting subtle but cumulative discourse features that may be invisible in small samples, while qualitative interpretation remains essential for explaining how patterns function rhetorically and intertextually. The paper concludes that corpus-informed analysis offers a robust framework for philological and applied linguistic research on Media English, supporting empirically anchored claims about discourse, ideology, and change over time, and providing transferable procedures suitable for university-level training in discourse analysis.