Seeing without a fovea? : Eye movements in reading and visual search with an artificial central scotoma

The effects of a central scotoma on eye movement control were investigated in the present study. A gaze-contingent display technique was developed to study this issue in normal-sighted subjects. Subjects developed a stable pseudofovea within five hours of training. A further aim was to investigate the effects of pseudofovea location on reading performance. Reading performance was superior with a pseudofovea to the right of fixation, intermediate with a pseudofovea to the left and worst with a pseudofovea below fixation. Eye movement patterns indicated that the congruence between pseudofovea location and the direction of eye movements required by the task can explain this finding: if the pseudofovea is to the right of fixation, both the identification of details presented at the pseudofovea location and the requirement to execute eye movements in the direction of the text require covert attention shifts to the right. This assumption was further investigated in a visual search task that requires eye movements either on the horizontal (from left to right/ right to left) or the vertical meridian (i.e., from top to bottom/ bottom to top). Performance was best whenever the pseudofovea was congruent with the direction required by the task (e.g., a pseudofovea above fixation when the task requires eye movements from bottom to top). There was no overall superiority for directions on the horizontal as compared to the vertical meridian. This finding rules out hard-wired properties of the visual system as the underlying reason for the observed eye movement behaviour. Instead, I suggest that the pseudofovea location serves as a potential saccade target, competing with task-relevant targets (e.g., the upcoming word or the next visual search target).