Technology has swept our world making it easier to produce and disseminate verbal, visual and aural content on various platforms. This makes communication in the twenty-first century increasingly multimodal. Such a tendency necessitates an inferential approach, rather than an approach based on the code model as in traditional multimodal analysis, wherein “codes and the tacit rules and constraints that underlie the production and interpretation of meaning within each code” are identified (Chandler 2017, 185; also see Kress 2010, 34–36). In pragmatics, text comprehension is often conceptualized as “a context building process” (Maillat 2013), through which contextual assumptions are made salient to help the addressees draw inferences about the text producer’s intended meanings. In multimodal texts, the context building process is rather more complex, as it requires inferring contextual assumptions based on the contents communicated across the different modes. Although such an inferential process is a pragmatic process par excellence, multimodality has tended to fall outside the interest of mainstream pragmatic scholarship. This is because pragmatics, broadly defined as the study of content-sensitive meaning, has traditionally been concerned with the study of verbal communication (Dicerto 2018, 37). Fortunately, the field has recently witnessed an increase in the number of studies applying pragmatic theories to account for the meaning communicated multimodally (e.g. Forceville 2020; Forceville and Clark 2014).