There is a long tradition in the linguistic sciences of considering some phenomena as peripheral to the core of language. These include exceptions to the Saussurean idea of the arbitrariness of the sign such as sound symbolic words, and linguistic expressions that deploy extra-systemic structural features such as interjections, particles, and conversational routines. Moreover, other linguistic signs which make use of modalities other than the auditory, such as manual gestures, are also considered peripheral. More generally, such phenomena are ignored because they cannot be adequately accounted for in current linguistic theories. However, these marginalia are tools for social interaction and are indispensable in the language games that people play cross-linguistically, as recent lessons from documentary linguistics, pragmatics and interactional linguistics show. They also hold the key to addressing various theoretical questions like given their multi-modal nature, what is the nature of their meaning, and how are their meanings to be represented and compared? How should their imagistic significance be represented? How are multi-modal utterances to be accounted for, and how do we approach their extra-systemic and structural nature? In this lecture, I reflect on how linguistic theory and practice would look like if we turn the tables and make marginalia the core business of the language sciences. This requires broadening the “structural grammar” approach (Boasian Trilogy) to a “grammar of use” and “ethnography of communication” (David Wilkins’ Communicative Practice Pentagon) perspective.