Institut Jean Nicod

Accueil > Séminaires/Colloques > Séminaires > PRISME > Seminaire PRISME - Barbu Revencu - Depictions in cognition and (...)



Seminaire PRISME - Barbu Revencu - Depictions in cognition and communication

Seminaire PRISME - Barbu Revencu - Depictions in cognition and communication
 
Vendredi 13 mars 2026, 11h, salle de réunion Jean Nicod (Rdc Pavillon jardin 29, rue d’Ulm 75005 Paris). Nous accueillerons Barbu Revencu (Collège de France et Neurospin) qui parlera des représentations dans la cognition et la communication.

From static diagrams to dynamic animations, humans routinely convey information through depictions—representational stimuli in which visual objects are arranged in spatiotemporal configurations for communicative ends. I will argue that depictions recruit two representational layers : one that tracks object symbols in space and time, and one that represents the entities under discussion through discourse referents. An assignment function links the two layers by providing pointers from the symbols to their discourse referents, enabling information about the referents to be communicated via the object symbols. I will then move on to present four sets of studies on semantic, pragmatic, and syntactic aspects of symbols in depictions. Study 1 offers an existence proof that 19-month-olds already interpret one class of stimuli—animations—as depictions. Study 2 goes one step further and shows that 15-month-olds readily accept arbitrary visual objects as ad-hoc symbols of familiar discourse referents. Study 3 provides evidence from adult participants that even realistic photos of objects are interpreted as symbols, in a way that is sensitive to pragmatic considerations. Finally, Study 4 indicates that human adults encode geometric shapes—visual objects often used as symbols in communication—in hierarchical, tree-structured representations. At the end, I will provide an overview of where this research project stands now and outline a big-picture open question on the relation between depictions and natural language.

CNRS EHESS ENS ENS