Abstract
Binocular rivalry (BR) is observed when the two eyes receive conflicting information, leading to perceptual switches between the eyes’ images. Evidence indicates that rivalry ceases, and the two images are processed equally, when attention is fully withdrawn. Consistent with this, computational models of BR typically have a region in the parameter space where both competing representations reach an equilibrium, and recent model work suggests that strong attention withdrawal drives the system into such a region (Li et al., 2017). But there is an unresolved issue: in existing models those parameter changes that move the system toward, but not into, the equilibrium region (e.g. modest attention withdrawal) increase the perceptual switch frequency, yet experiments (Paffen et al., 2006; Alais et al., 2010) indicate that modest attention withdrawal has the opposite effect of reducing switch frequency. Due to the key position of this latter result as an obstacle to reconciling empirical results and models, we aimed to verify the result. Moreover, the original work raises a potential methodological concern because observers reported BR dominance while also performing a secondary attention task, which may have invited a strategy of shifting attention back and forth between tasks. To avoid this potential concern, we used optokinetic nystagmus to track perceptual switches of a task-irrelevant rivalry stimulus (foveally presented dot fields moving in opposite directions in the two eyes), while observers performed an auditory attention task. Our results show that switch frequency decreases as the auditory attention task becomes more challenging, thus confirming the existing finding with our new method that does not share the original method’s potential shortcomings. Because computational model predictions are inconsistent with this finding, our work indicates that modifications to existing models are needed.