What renders a saccade target relevant?

• Applicant: Dr. Christian Wolf
• Subject Area: General, Cognitive and Mathematical Psychology
• Term: since 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 427754309

Only the central region of the visual field enables high-acuity processing. Consequently, humans shift their gaze approximately three times a second, each time choosing a different part of the visual environment for high-acuity processing. These gaze shifts are achieved by saccadic eye movements. Given their ballistic nature and the close link between eye movements and visual perception, saccades have become a role model to study motor learning as well as visual decision making. Yet, the exact mechanisms how what we foveally see affects where we look next are not fully understood. The aim of the present grant proposal is to reveal how visual information affects where we move our eyes – both, in terms of the decision where to look next as well as the maintenance of eye movement accuracy. To this end, five experiments, grouped into four work packages, are described in which the content or quality of pre-saccadic or post-saccadic foveal information is manipulated and its influence on eye movement control is investigated. We recently showed that the foveal image content biases the decision between peripheral targets. The first work package aims to distinguish between two potential mechanisms how what we currently see in the fovea affects our decision where we look next. Two further work packages will measure the adaptation of saccades and will address whether and how post-saccadic vision and proprioception jointly contribute to oculomotor learning when the quality of visual information is poor (work package 2), and will reveal the mechanisms driving oculomotor learning in the presence of multiple post-saccadic targets (work package 3). The fourth work package bridges the gap between visual decision making and motor learning and will unveil whether reinforcement learning and/or error-based learning can help to select optimal eye movement endpoints. Taken together, these experiments would foster our understanding of how the brain utilizes visual information to learn from error, how it assigns priority to targets in the visual environment and how visual perception and eye movement control interact.

• DFG Programme: Research Grants
• International Connection: Netherlands
• Cooperation Partner: Privatdozent Dr. Artem Belopolsky