In this commentary, we argue that the literate glasses bias highlighted by Kolinsky and Morais reflects a more general bias in language and cognitive sciences, which has to do with the way language is typically studied. Indeed, researchers tend to study language in a modular way, independent of other cognitive functions and its interactions with body and environment. Here, we focus on the limitations of this approach with respect to the literacy bias and we show how theories of embodied cognition and neural re-use shed new light on these debates.