Abstract
The sensory recruitment model of visual working memory (VWM) suggests that VWM representations depend on the early visual cortex (EVC). However, the model has been challenged by the logic that ongoing perception should interfere with EVC VWM representations. Moreover, previous research showed that the contents of VWM can be decoded from the parietal cortex regardless of distractor presence but EVC decoding was reduced or absent if visual distractors were presented during the memory delay (Bettencourt & Xu, 2016, doi:10.1038/nn.4174; Rademaker et al., 2019, doi:10.1038/s41593-019-0428-x). Here, we adopted two manipulations to evaluate VWM reliance on EVC vs. parietal cortex. First, we applied anodal (or sham) transcranial direct current stimulation (tDCS) to either occipital cortex (Oz) or right posterior parietal cortex (P4) prior to a Gabor patch orientation VWM task. We predicted distinct effects of EVC and parietal tDCS on overall VWM performance. Second, we displayed distractors during the delay period in half of the testing blocks. We reasoned that the oblique effect, which is thought to rely in part on EVC (Berkley et al., 1975, doi:10.1016/0042-6989(75)90213-8), would be reduced by distractor presence and the accompanying shift to parietal mechanisms. Contrary to our predictions, neither tDCS manipulation was effective. The oblique effect was present under all conditions, and surprisingly, it was stronger under visual distraction. To the extent that the oblique effect indeed has its basis in EVC, these results suggest paradoxically increased dependence on EVC when memories are maintained through incoming visual stimuli. Alternatively, such an effect would be consistent with the loss of VWM detail during distraction, or these results could suggest that the oblique effect is less dependent on EVC than previously believed.