Abstract
Recalling a past visual experience activates cortical structures that were engaged during the initial encoding of that experience. This phenomenon of neural reactivation is thought to underlie our subjective experience of remembering. However, when recalling past experiences, we are often trying to retrieve specific event features, such as information about object categories or their spatial locations. Does recalling one feature of a visual experience elicit incidental neural reactivation of other event features? If so, does such incidental reactivation of non-target event features strengthen memory for these features? We addressed these questions in a human fMRI study (n = 18) using pattern classification analyses to measure neural reactivation of memory features. During fMRI scanning, participants engaged in alternating study-test rounds. In study rounds, words were paired with images of faces or scenes that appeared on the left- or right- hand side of the screen. During test rounds, words were presented and participants were either instructed to recall the location (left/right; 1/3 of study items) or category (face/scene; 1/3 of study items) of the corresponding image; for the remaining 1/3 of study items, neither feature was tested. After scanning, participants completed a post-test that assessed their memory for both features (location and category) for each studied word. Pattern classification analyses revealed robust neural reactivation of feature information during the test rounds, including reactivation of non-target features (e.g., reactivation of face/scene information during location retrieval). Moreover, non-target features were much better remembered at post-test than features of items that were not tested at all, and this memory benefit for non-target features correlated with their incidental reactivation during test rounds. Thus, retrieving one feature of a visual event from memory can elicit incidental neural reactivation of other event features, thereby facilitating future remembering.
Meeting abstract presented at VSS 2012