Abstract
<p>This paper presents a mixed-methods discourse analysis of image schemas—recurring spatial–temporal patterns that structure understanding—across six scientific and medical domains: chemistry, spectroscopy, biology, medicine, physics, and mathematics. Drawing on cognitive linguistic frameworks established by Lakoff (1987) and Johnson (1987), we examine how embodied schemas, such as CONTAINER, SOURCE-PATH-GOAL, UP-DOWN, PART-WHOLE, BALANCE, COMPULSION, BLOCKAGE, LINK, and CYCLE, operate within disciplinary textbooks. Unlike prior qualitative-only studies, we employ systematic quantitative coding of linguistic markers and diagrammatic features across an expanded corpus of 12 canonical textbooks (two per discipline). The analysis reveals discipline-specific schema profiles with high inter-rater reliability (κ = 0.84). Spectroscopy privileges SOURCE-PATH-GOAL and BALANCE; chemistry emphasizes CONTAINER and LINK; biology centers on CONTAINER, CYCLE, and BLOCKAGE; medicine prioritizes CONTAINER, BALANCE, and BLOCKAGE; physics is dominated by COMPULSION and UP-DOWN; while mathematics abstracts schemas into logical relations. We systematically test the claim that ENABLEMENT is entirely absent—and find it present at very low frequency (mean 0.03 instances per page) only in medicine and biology, supporting a nuanced version of the absence hypothesis. We further provide empirical validation of pedagogical transfer using a small reader study (n=48). Appendices provide full coding schemes, marker inventories, corpus details, and supplementary data. The study demonstrates how systematic schema analysis can inform science education and conceptual change.Keywords: image schemas, cognitive linguistics, science textbooks, discourse analysis, mixed methods, embodied cognition, ENABLEMENT</p>