Abstract
This chapter investigates changes in medical topics, style and language across 500 years, from 1500 to 2018. To do so, we employ data-driven methods of Computational Linguistics and Digital Humanities: document classification, topic modelling, and automatically constructed conceptual maps. We trace changes from traditional thinking in the scholastic period to empirical methods, professionalised medicine, and finally the increasing importance of data, statis- tics and clinical studies, away from symptom-centred medicine. We conclude that medical discourse has undergone radical changes and that data-driven methods reflect these changes and offer an unprecedented overview. We also critically discuss shortcomings of our data and methods.