This project focuses on the perception of movie sequences. These are of special interest because the stimulus is dynamic, but the observer has no role to play in causing this movement. In fact, films and videos routinely include changes that do not occur in normal perception (for example, abrupt switches to completely new scenes or changes in the camera's "point of view"). Nonetheless, the viewer appears able to glue these disparate images into a single coherent mental representation that includes not only the identity of objects depicted, but also where they are located in the space defined by the film. Visual transitions obviously play a role in the narrative flow of a movie. However, little is known about their consequences and virtually nothing about how apparently effortless perceptual coherence is achieved.

Previous work from our lab has explored eye movements and memory for objects across a single transition (or “cut”) - from a dynamic panned sequence to a static view of the same scene from a different viewpoint. We have shown that information encoded during a pan survives a cut to another point of view and even a cut to a completely different scene. Viewers retain information about the identity and colour of objects, but the way position information is coded differs both qualitatively and quantitatively. Essentially, people seem bad at encoding and remembering positions in movies. However, the reasons for this difficulty in movies are uncertain. This project aims to establish the psychological mechanisms underlying the mental representation of space in movies.