Representing and Attending to the Temporal Structure of Visual Events in Working Memory

Visual events can be characterized by their spatiotemporal structure, for instance by the points in space and time at which they occur, by their spatial distribution and temporal duration. The representation of temporal structure is, however, not nearly as well understood as that of spatial characteristics. This project examines how the temporal structure of a visual event is represented once the event itself has passed, that is, in working memory, and how this structure can be used to optimize the capacity-limited visual working memory system by orienting attention to representations related to relevant points in time. Different variants of change detection tasks have been designed that allow answering specific questions on the characteristics and use of temporal structures. The first part of the project serves to establish how the temporal structure of visual events is represented. First, we will examine whether temporal information is incidentally encoded and used along with spatial information to provide a frame of reference for the storage of visual objects. We will also clarify whether encoding in such a spatiotemporal frame of reference is obligatory or flexibly adjustable to task demands, that is, to a predictable retrieval context that emphasizes either space, time or both. Second, we will investigate the metrical properties of the represented spatial and temporal structures. Specifically, we will determine whether the storage of visual objects is sensitive to global, relational or ordinal transformations of their spatial and/or temporal coordinates. The second part of the project will examine how the representation of temporal structure is used to optimize visual working memory processing by prioritising relevant information based on temporal coordinates. One series of experiments will establish whether attention can be retrospectively oriented towards representations based on their temporal locations, when information about the relevance of points in time in the past becomes available during maintenance. A second series will investigate whether temporal expectations induced with rhythms can tune attention to the temporal structure of perceived and remembered visual events.