In academia, different disciplines tend to represent themselves, their writers, and their readers in unique ways. To be able to probe into such cross-disciplinary discursive variations, the present cross-disciplinary study used Martin and White’s appraisal framework incorporating lexico-grammar and language evaluation strategies to explore the linguistic construal of evaluative stance in the conclusion section of academic research articles. To this end, a corpus of 160 research articles randomly selected from English international journals of psychology as a representative of soft sciences and industrial engineering as a representative of hard scienceswere examined in the light of Martin and White’s scheme, with a focus on graduation. To code the data, UAM Corpus Tool was used for the ease of descriptive analyses. In general, the findings indicated that the number of graduation words and graduation types were significantly higher in the hard science corpus compared to the soft science one, suggesting that the authors of hard science articles mainly attempted to express their appreciation and judgment and report the fluctuation of a specific feature in their findings.
