Abstract
Do object memories sharpen the perception of familiar objects' borders? Memories represent the norm of previously seen objects; for familiar objects, these are likely to have sharp, focused borders because they have often been fixated and attended. Since perception arises from the integration of the current stimulus with memories, blurry borders might be perceived as sharper in familiar than novel objects. Observers judged whether the borders of two black silhouettes presented on a gray background were the same or different levels of blur. One object was familiar (lamp); the other was a matched novel object created by spatially rearranging the parts of the familiar object. Across trials one of the objects varied in blur (Test) from low to high through 9 levels; the other was held at a constant medium blur (Standard). Both objects served as Standard and Test equally often; their left/right location varied across trials. If object memories influence the perceived sharpness of blurry borders, then the peak of the "same" distribution should be at a higher level of blur when the Test object is familiar rather than novel. We observed this effect in a previous experiment, p < 10-6. There, the pair of stimuli was preceded by masked word primes; on ~17% of trials the word named the familiar object. No effect of word prime was observed. In Experiment 1, we replicated the effects without a word prime, p < 10-7, supporting the hypothesis that memory representations increase the perceived sharpness of the borders of familiar compared to novel objects. Experiment 2 tests the generalizability of these results by using silhouettes of two other familiar objects, each presented alongside their part-rearranged, novel matches. If we replicate the results of Experiment 1, that will support the hypothesis that object memories can alter the appearance of an object.
Meeting abstract presented at VSS 2018