Embodied Object Concepts: The Contribution of Structural and Functional Manipulability Depends on Available Visual Information

Object identification is driven, in part, by the extent to which we have sensorimotor experience with the object. Importantly, the activation of embodied object representations depends on contextual information. In the present study, we use a visual masking paradigm to investigate how the availability of visual information modulates the role of manipulability in the representation of object concepts. Using both an object naming task (i.e., linguistic response) and a picture-word matching task (i.e., manual response), we provide evidence that structural manipulability (the ability to pick up an object with one hand) and functional manipulability (the action information that pertains to the ultimate use of the object) have dissociable effects on object identification. In both tasks, the effects of structural manipulability were greater when structural information was available in the image (i.e., when the objects were unmasked); in contrast, the effects of functional manipulability were greater when the objects were masked. Importantly, these effects were not due to object familiarity or the age at which the name of the objects was acquired. Our results are consistent with the activation of the two pathways within the dorsal visual stream that are part of a distributed neural network that represents embodied action information. We extend previous research by showing that visual information determines which type of embodied information drives object identification. L'identification d'un objet s'effectue, en partie, par le degré d'expérience sensorimotrice que nous avons avec l'objet en question. Fait important, l'activation de représentations d'objet incarnées dépend de l'information contextuelle. Dans la présente étude, nous utilisons un modèle de masquage visuel pour examiner comment la disponibilité de l'information visuelle module le rôle de la manipulabilité dans la représentation de concepts d'objets. L'utilisation combinée d'une tâche de dénomination (c.-à-d. une réponse linguistique) et d'une tâche d'association mot-image (c.-à-d. une réponse manuelle) nous permet de fournir la preuve que la manipulabilité structurelle (la capacité à saisir un objet avec une main) et la manipulabilité fonctionnelle (l'information de l'action ayant trait à l'utilisation finale de l'objet) ont des effets indissociables sur l'identification d'objets. Dans les deux tâches, les effets de la manipulabilité structurelle étaient supérieurs lorsque l'information structurelle é