Using a Corpus-Assisted Discourse Studies Approach to Analyse Gender: A Case Study of German Radiology Reports

Abstract

This paper explores how a Corpus-Assisted Discourse Studies (CADS) approach can be utilised to investigate representations of gender as well as potential gender bias in radiology reporting, which constitutes a form of professional, medical discourse. The database collected for this purpose consists of three specialised German sub-corpora (332,901 cranial, thoracic, and whole-body computed tomographies, with more than 61 million tokens), which were extracted from a larger medical corpus called MedCorpInn that was built as part of an interdisciplinary project conducted jointly by the University of Innsbruck and Innsbruck Medical University. As a basic premise, CTs are assumed discursive, linguistic events, which are influenced by social and institutional factors. They represent an essential everyday communicative practice among radiologists and referring doctors and they function both as documentation and as a legal record of imaging procedures. To investigate whether there are differences and/or subtle similarities (Taylor 2018; Brezina 2018) in the largely standardised reports on female vs on male patients, a CADS-approach focusing on gender is applied. Keywords, collocation, and concordance techniques will be introduced and used to explore how male and female patients are discussed in the medical discourse studied here. Research into internal clinical communicative practices could also be of interest from the perspective of gender medicine.