Embodied approaches to the mental representation of concepts, called grounded cognition, often argue that brain regions which support action and perception are also those which are active during interaction with concepts, such that understanding the word kick draws on the same neural resources as the action itself. Yet, such research has made little progress in arriving at a theoretical description of mental representations as a result of an impasse in theory. The two prominent theories, perceptual symbol systems (Barsalou, 1999) and conceptual metaphor theory (Lakoff & Johnson, 1987), have gone largely unchanged since their inception. We argue that they were prematurely advanced, making too many assumptions, which generated a series of self-reinforcing issues. Theoretical work struggled because the high specificity of the original theories generated a fragmented field (Bringmann et al., 2022). This led, in turn, to agnostic empirical work which typically avoided focus on any one specific theory, preferring to demonstrate ‘embodied effects’ more generally (Anderson, 2008). Yet, this lack of theoretical allegiance prevents cumulative progress; findings which cannot be organized by the structure of a theory fail to build on one another. Our proposal is a critical methodological advancement to future work on embodiment because it removes this impasse, allowing even formerly-agnostic empirical work to cumulatively progress because it can explicitly position itself in the overarching theoretical framework we call the minimalist account. From past work, we have extracted those mechanisms which have exceedingly strong support, creating a useful, simple, and evidence-based framework that advances grounded cognition research.