A key question in visual neuroscience is how our subjective experience of the visual world remains largely uninterrupted by the many eye movements we make, each of which abruptly displaces the retinal image. It has long been hypothesized that extra-retinal signals, generated by voluntary movements of the eye, alter our vision in such a way that facilitates visual perception across eye movements. However, it is unclear how these extra-retinal signals influence the integration of form information, giving rise to our ability to identify visual objects in spite of the fact that eye movements shift those objects on the retina. Thus, I used eye tracking and psychophysics to test the influence of eye movements on object perception in peripheral vision. In this thesis I present studies in which I show for the first time substantial changes in object identification that result from eye movement preparation and execution. The results from these studies show that our visual experience is not just constructed from the different images that hit the retina, but is influenced by eye movement signals that facilitate the perception of objects from one glance to the next.
The studies presented in Chapters 2 and 3 examined the ways in which eye movements alter visual crowding. Visual crowding refers to the reduced ability to recognize an object in peripheral vision when other objects surround it. A widely held view posits that crowding is caused by compulsory interactions of visual signals throughout early visual brain areas. In Chapter 2, I compared the spatial extent of crowding when no eye movements were made with the spatial extent of crowding during saccade preparation. I found that, just prior to a saccadic eye movement, the spatial extent of crowding is reduced for the target of the saccade despite no change in the retinal position of stimuli. In contrast, results presented in Chapter 3 show that during smooth pursuit eye movements, the spatial extent of crowding increases for objects contraversive to the direction of the eye movement. These results suggest that crowding is not caused solely by the interaction of visual signals in early brain areas, but is also modulated by neural processes involved in the generation of eye movements. Furthermore, these findings suggest that crowding is modified according to an object’s relevance to a current action. In the case of saccades, crowding is reduced such that discriminability of the saccade target itself is enhanced. In the case of smooth pursuit, on the other hand, discriminability is degraded for objects that are no longer in the path of the eye movement.
In Chapters 4 and 5, I examined “predictive remapping”. Predictive remapping refers to pre-saccadic changes in visual processing that are thought to anticipate the retinal position of objects following a saccade, and has been linked to phenomenological visual stability across eye movements. In Chapter 4, I used a task that combined a spatial cueing task with a saccadic eye movement paradigm. I found that visual processing is prioritized not only at the location of a visual cue, but also in the direction of an impending saccade. Interestingly, the direction of this shift in visual attention is opposite to what would be expected were the shift to compensate for the retinal displacement of visual objects across saccades. I therefore argue that this finding results from a general interaction between saccade programming and exogenous visual attention. In Chapter 5, I exploited visual crowding to test whether predictive remapping preserves an object’s visual features. When presented just prior to a saccade, I found that a probe in one visual field is “crowded” by distractors that surround the probe’s predicted, post-saccadic location in the opposite visual field. Importantly, the degree of such “remapped crowding” depends on how similar the distractors’ features are to the probe. These findings reveal that the predictive remapping signal conveys featural information such that pre-saccadic visual processing is prioritized according to the expected position and identity of an object that will shift on the retina across saccades.
Overall, the results presented in this thesis suggest that eye movements reliably alter peripheral visual processing. Importantly, my studies show that changes in visual perception act to increase or decrease the discriminability of objects in peripheral vision according to the oculomotor demands of a given task. Moreover, my findings show that these changes in perception are independent of changes in the image on the retina, and must therefore result from extra-retinal signals that are generated during the preparation and execution of saccadic and pursuit eye movements. The modification of vision by these extra-retinal signals likely facilitates the subjective impression of visual stability when eye movements continuously alter the image that falls on the retina.