Thomas Sanocki¹ (), Teresa P. Pham²; ¹University of South Florida, ²University of Minnesota

Cognitive theories assume that perceiving and acting within a familiar scene is supported by top-down influences driven by internal knowledge about the scene. The knowledge may include typical objects and actions — information about object function. Here, we test whether viewing a scene can prime knowledge about object function. The primes were an image made from related, same-scene pieces, or from unrelated pieces of different scenes. That is, same-scene primes consisted of 3 different pieces of the same room (3 phrases of a living room, in terms of scene-grammar theory); different-scene primes consisted of 3 pieces that came from different scenes. A prime was presented for one second, followed by a copy of the image with one piece and one object within the piece highlighted. Observers indicated whether the object in this region (piece) was positioned appropriately to function. For example, a lamp functions if upright on a table top (YES response), but not if oriented sideways (NO response). (The stimuli were validated in a pilot study, to encourage rapid and accurate function-judgments. All objects were appropriate for the scene-piece they were depicted in, so that scene-belonging was not relevant to the response. Orientation was changed for some NO objects, and position within their scene was varied for others. 31 students participated.) Function judgments were overall accurate (97%) and rapid (Mean RT = 1171 ms), with RT improving across the session. The critical comparison was the mean RT after same-scene and different-scene primes. RT was not different between the two prime conditions, indicating a null effect of scene-organization. This similarity in RT held across the session. The results suggest that function perception can’t be primed by scene meaning; function perception is independent of larger scene-organization.