Positive vs. Negative Collocates of Indian-Origin Words in Western Media: A Corpus-Based Critical Discourse Analysis

This study explores how Indian-origin lexical items—such as karma, chai, Bollywood, and yoga—are framed evaluatively within Western English-language media. Drawing on the News on the Web (NOW) corpus (2015–2023) and guided by the principles of Corpus-Assisted Discourse Studies (CADS), the research investigates whether these borrowed cultural terms are systematically associated with positive, negative, or neutral sentiments in journalistic discourse. Fifteen Indian-origin words were selected based on frequency, cultural relevance, and dictionary inclusion. Using ±5 word collocation windows, the most frequent co-occurring words were extracted via AntConc and Sketch Engine. Collocates were manually annotated for sentiment polarity—positive, neutral, or negative—supported by using lexicons (SentiWordNet, NRC), and their accuracy was verified through inter-coder reliability. Sentiment scores were statistically analysed using Chi-square, ANOVA, and Kruskal-Wallis tests. Results revealed that terms like yoga and ayurveda were predominantly associated with positive collocates, whereas curry and Bollywood attracted significantly more negative sentiment. The Chi-square test confirmed significant variation in sentiment across keywords (p = .035), although ANOVA and Kruskal-Wallis tests were not statistically significant. Visualizations such as heatmaps and stacked bar charts highlighted distinct sentiment patterns, reflecting broader ideological framing. The study concludes that Indian-origin words are not just linguistically assimilated but ideologically recontextualised in Western discourse, often affirming wellness culture while trivialising or exoticising ethnic identities. These findings offer critical insights into semantic prosody, media representation, and postcolonial power dynamics in global Englishes.