Abstract
Our eyes are constantly moving, jumping from one location to another, and yet we perceive our world as stable. Immediately after a saccade, attention has to remap from the previous “retinotopic” eye-centered location to the current “spatiotopic” world-centered location. Previous work has shown that when attention remaps after a saccade, it is briefly split between the previous retinotopic location and the current spatiotopic location. Researchers investigating this phenomenon have found that during this remapping period, perception of objects such as colored squares or colored, tilted bars can be subject to feature-binding errors such color mixing, illusory-conjunctions, and unbound-errors (Dowd & Golomb, 2019; Golomb, L’Heureux, & Kanwisher 2014). However, it is unknown what the consequences of feature degradation during remapping may be for perception of real-world objects. To investigate this question, we conducted an experiment in which participants viewed four colored objects presented simultaneously after either a short (50ms) or long (500ms) delay following completion of a saccade. Before making the saccade, participants attended a cued spatiotopic location, and then reported the color and identity of the object (e.g., a red car) presented at that cued location. Participants were overall less accurate for both color and object reports in the short-delay condition compared to trials with a long delay, in which there was sufficient time for remapping. Moreover, when participants made errors in the short-delay condition, they were more likely to make certain types of binding errors, such as illusory-conjunctions between the spatiotopic color and a non-target object, or reporting the correct spatiotopic target color with an object not originally present in the display. Participants also made unbound errors such as reporting the identity of the retinotopic object, but guessing on color. Overall, our results imply that saccadic remapping may have implications for the perception of real-world objects.