Mining Emotions in Academic Writing: A Subdisciplinary Probe into Medical Research Articles

Songyun Chen
DOI: https://doi.org/10.1016/j.esp.2024.01.002
IF: 2.417
2024-01-01
English for Specific Purposes
Abstract:Sentiment analysis has garnered significant attention in the academic field lately, which gives academic scholars a full comprehension of how emotion states are communicated in texts and how it works as a powerful persuasion strategy for academic writers. Based on a large dataset of over five million tokens, this study examined specific emotions in research articles across 18 subdisciplines of medicine, aiming to reveal the emotions presented by medical writers in academic writing. The results showed the following: (1) four emotions with positive valence-trust, anticipation, joy, and surprise-were observed; (2) the tendency of emotion represented by trust > anticipation > joy > surprise was identified, with the first (trust) being overwhelmingly dominant and the rest indicating minor presence; and (3) subdisciplinary emotion variation was found and the contributing variables were identified as research methods and paradigms, research objects, interdisciplinary features, and disciplinary status in knowledge accumulation. The findings contribute to our understanding of emotion strategy applied in academic writing and genre characteristics of medical science. This article concludes with pedagogical implications for EAP teachers and suggestions for possible future research. (c) 2024 Elsevier Ltd. All rights reserved.
What problem does this paper attempt to address?