Increasingly, studies are taking account of multimodality when analyzing language textbooks. Due to the diversity of multimodal frameworks used in analyses, and the interdisciplinary nature of language textbook studies, conceptual differences arise that are important to discuss -- which is the purpose of this paper. Specifically, I argue that multimodal analyses of language textbooks can be divided into two groups based on how they conceptualise meaning. One examines how textbooks' textual-visual content encodes and communicates "ideas about the world," treating meaning as "representation." A second set of research studies is focused on how multimodal elements in textbooks foster interpersonal relations between text producers and readers, thus viewing the meaning textbooks communicate as "interaction." While each approach is valid, I argue that neither places sufficient emphasis on the fact that textbooks are a didactic genre where learners' engagement with any meaning is heavily guided. As such, critical textbook analysis should attempt to demonstrate the ideological nature of meaning-making by examining the interplay of multimodal representations, the interactive meaning of textbooks' multimodal material as well as the pedagogic-didactic frame within which learners encounter them. The article utilises an example from a popular English as a foreign language textbook to illustrate these points.