Training of binocular rivalry suppression suggests stimulus-specific plasticity in monocular and binocular visual areas

Primitive visual channels have a causal role in cognitive transfer

Scientific investigations have long emphasized the cortex's role in cognitive transfer and arithmetic abilities. To date, however, this assumption has not been thoroughly empirically investigated. Here we demonstrated that primitive mechanisms-lower visual channels-have a causal role in cognitive transfer of complex skills such as symbolic arithmetic. We found that exposing only one monocular channel to a visuospatial training resulted in a larger transfer effect in the trained monocular channel compared to the untrained monocular channel. Such cognitive transfer was found for both novel figural-spatial problems (near transfer) and novel subtraction problems (far transfer). Importantly, the benefits of the trained eye were not observed in old problems and in other tasks that did not involve visuospatial abilities (the Stroop task, a multiplication task). These results challenge the exclusive role of the cortex in cognitive transfer and complex arithmetic. In addition, the results suggest a new mechanism for the emergence of cognitive skills, that could be shared across different species.

Attention to Monocular Images Bias Binocular Rivalry

When monocular images cannot be fused, perception alternates between the two (or more) possible images. This phenomenon, binocular rivalry (BR), is driven by the physical properties of the stimuli (size, contrast, spatial frequency, etc.) but it can also be modulated by attention to features of one of the rival stimuli (Chong et al., 2005; Dieter et al., 2016) and by attentional demands independent of the BR assessment (Paffen et al., 2008). Instead of the perceptually demanding tasks previously used to bias BR, we designed a simple counting task. We monocularly presented a number of trials (around 10 min) with a set of symbols and asked participants to count them. We found that after this task, dominance durations decreased for the unattended channel, and did not change for the attended channel. The results parallel those of Paffen et al. (2008) and square nicely with Levelt’s second proposition, suggesting that the counting task effectively increased the sensibility of one channel which led to increased strength of the images presented to that channel. Alternatively, the results could be explained assuming that the non-attended channel was inhibited during the counting task, and the inhibition was carried over to the BR task.