Vanessa Michel¹, Mark Eckert², Alex White¹; ¹Barnard College, ²Columbia University Irving Medical Center

Background: Oculomotor freezing (OF) is the rapid inhibition of eye movements caused by sensory stimulation. While well-documented in the visual domain, cross-modal freezing and its modulation by cognitive state remain poorly understood. This study answers two questions: Do visual and auditory stimuli elicit comparable OF, and do concurrent task demands modulate the magnitude of this inhibition? Methods: Across four experiments, participants performed two tasks that elicited frequent saccades but differed in top-down control. In the reading task, participants read passages and answered comprehension questions. In the dot-following task, they tracked a randomly moving dot and counted its color changes. During both of these tasks, visual and auditory distractor stimuli were presented at unpredictable intervals. The visual stimulus was a briefly flashed frame around the edge of the display. The auditory stimuli were pure-tone beeps and white noise bursts. Results: Visual flashes elicited strong OF during both tasks – the saccade rate dropped to a minimum within ~ 110 ms. In contrast, auditory stimuli showed a task-dependent effect. They elicited minimal freezing during reading but significant freezing during the dot-following task. Notably, when auditory OF did occur, it was roughly 30 ms faster than visual OF. Furthermore, white noise bursts elicited stronger freezing than pure tones, potentially reflecting stimulation of a broader population of frequency-tuned neurons. Conclusions: Saccade planning is inevitably disrupted by salient visual onsets, but it can be “shielded” from auditory stimuli during tasks that require strong top-down control, such as reading. In contrast, saccades in the dot-following task are more stimulus-driven and thus more susceptible to inhibition by salient irrelevant sounds. These findings position OF as a sensitive, objective tool for measuring how the brain prioritizes sensory inputs, which may differ across age and in clinical populations.

Acknowledgements: Special thanks to Martin Rolfs.