Abstract
Discussion sections are central to the interpretation, justification, and disciplinary positioning of research findings. This study investigated argumentative strategies in 61 English-language EFL and applied-linguistics research articles published in three SINTA 2 journals affiliated with Muhammadiyah and Aisyiyah universities during 2024–2025. The corpus comprised 19 JFLTL, 26 JEES, and 16 ELTEJ/UAD articles. Only discussion discourse was analyzed: in articles with combined Results and Discussion sections, descriptive result statements, tables, statistics, and uninterpreted participant quotations were excluded. A qualitative corpus-assisted content analysis integrated Swalesian rhetorical moves, Toulmin’s argument components, and Hyland’s stance and engagement framework. Evaluating significance and presenting implications/recommendations each occurred in 58 articles (95.1%), while comparison with previous literature occurred in 55 (90.2%). Claims, grounds, warrants, and qualifiers were widespread, but explicit limitations appeared in only nine articles (14.8%). Concordance-audited stance counts showed 1,151 hedges and 390 boosters (2.95:1), confirming a strong preference for cautious qualification. Supportive citation functions were substantially more common than explicit contrast, and only 26 articles contained a clearly identifiable counterargument or alternative explanation. The findings show that these Discussion sections were generally interpretive and evaluative rather than merely descriptive; their principal weakness was uneven critical reflexivity, especially in acknowledging limitations, testing preferred claims against alternatives, and engaging discrepant evidence. The study offers an empirically grounded basis for academic-writing instruction and editorial guidance in Indonesian EFL publishing.