The grounded/embodied cognition framework (Barsalou, 1999; Glenberg, 1997) has led to a renewed interest in affordances (Gibson, 1979) and more generally in whether action information is constitutive of manipulable objects’ conceptual representation. Although accumulating evidence from behavioural and neuroimaging studies supports the broader claim of an involvement of motor actions in object processing, their nature and role in various tasks remain controversial (Paisios, Labeye & Huet, 2019). Part of the reason for this comes from a lack of consistency in how ‘manipulability’ has been defined and assessed across studies. Particularly, objects that can be grasped (structural action) and those that afford an action related to their use (functional action) have been, to a large extent, confounded, thus limiting the conclusions that can be drawn about these two types of action knowledge (Buxbaum & Kalénine, 2010). Additionally, commonly used manipulability ratings (e.g. the ease to pantomime, Magnié, Besson & Dolisi, 2003) do not fully capture the extent to which objects are associated to functional actions. To address these limitations, we first provide a thorough review of the available semantic norms related to the manipulability of objects and discuss their relevance and shortcomings in light of current advancements on the role of actions in conceptual representations. We then present a hierarchical scheme for the collection of five object-related interaction ratings for French concrete object names, namely: the ease of physical interaction (body-object interaction – BOI, Pexman et al., 2018), the ease to grasp, the extent to which they are associated to an action, the ease to pantomime, and the number of actions they afford. Preliminary results of the BOI norms will also be presented and discussed.