Abstract
People are able to hold more items in visual working memory (VWM) when asked to remember meaningful stimuli (e.g., umbrella) than abstract stimuli (e.g., blue square) (e.g., Brady et al. 2016). However, in previous work, real-world objects and abstract stimuli were not controlled for perceptual equivalency. It is therefore possible that differences in VWM capacity were driven by perceptual properties (e.g., visual information load) rather than representational meaning. Here we address this concern by using perceptually-matched stimuli and manipulating only their meaningfulness. We used two-tone images (Mooney faces) that can be perceived as meaningful faces when upright, but meaningless blobs when inverted or shuffled. In particular, we measured VWM capacity for faces vs. non-face stimuli and recognized vs. unrecognized faces while simultaneously measuring the contralateral delay activity (CDA). The CDA is believed to be a neural marker sensitive to the number of items being actively held in mind (e.g., its amplitude increases as the number of items held in VWM increases). By combining behavior and CDA we can assess the effect of meaningfulness on VWM capacity. In Experiment 1, we found that participants had higher VWM capacity for trials with more faces present compared to perceptually-matched non-faces (t(11)=3.23, p=0.008), and on trials where participants recognized more of the faces compared to trials with the same stimuli where they recognized fewer faces (t(11)=3.38, p=0.006). In Experiment 2, we found that in addition, CDA amplitudes were larger when the memory sets consisted of more faces than when they consisted of fewer faces (t(12)=2.98, p=0.01). Together these results suggest that meaningfulness plays an important role in enabling more items to be held in VWM, independent of perceptual properties. Broadly, this suggests that VWM capacity is not fixed but critically depends on what type of information is being remembered.
Meeting abstract presented at VSS 2018