Abstract
One of the challenges posed by a world in which visible patterns or textures may be generated from surface reflectance or from externally produced patterns of light is deciphering whether a pattern viewed on an object's surface is a property of the object. If the pattern is a property of the object, then it is advantageous for pattern information to be included in the object representation to facilitate accurate individuation or identification of the object upon later viewings. In contrast, if the pattern is not a property of the object, and is instead a shadow cast on the objects' surface, representing the pattern as an object property can lead to errors in object individuation and identification. In the current study, we tested 7-month-old infants' (N = 40) ability to individuate objects based on pattern information when the pattern was projected and thus, not a true property of the object. In a between-subjects design using the narrow-screen task (Wilcox & Baillargeon, 1998a, b), we assessed object individuation under conditions in which spatiotemporal and motion cues could be used to discern the origins of the pattern. Results indicated that when a pattern was projected onto the surface of an object but was no longer visible when the object was occluded (i.e., appearance of the pattern and object were temporally linked), infants appeared to regard the pattern as a property of the object even though the pattern remained stationary as the object moved. In contrast, when the pattern remained visible during object occlusion, infants regarded the pattern as distinct from the object. Together these finding suggest that when forming object representations based on patterns, 7-month-olds rely more heavily on temporal cues relative to motion cues in order to distinguish patterns inherent to the object from other types of patterns.
Meeting abstract presented at VSS 2018