Abstract
Spatial location is believed to have a special role in binding other visual features both in perception and working memory. Consistent with this view, a recent study found specific deficits in feature binding in working memory when memoranda were presented sequentially at the same location, compared to a control condition in which items were presented sequentially at different locations (Pertzov & Husain 2014). In the present study, we investigated whether these apparent working memory deficits may in fact be due to perceptual interference at the stage of encoding, as suggested by findings of temporal crowding (Yeshurun, Rashal & Tkacz-Domb, 2015). Subjects viewed four oriented bars in discriminable colors, and, following a brief delay, were cued with a color to reproduce the orientation of one bar from memory. Bars were presented sequentially either at the same or different display locations. Importantly, we varied the inter-stimulus interval (ISI) either to match the duration from the previous study (300 ms), or to be twice as long so as to remove any possible perceptual interference between stimuli. In the conditions with a short ISI, we reproduced the findings of the original study: there was a higher proportion of swap errors (i.e. incorrectly reporting the orientation of a non-cued stimulus) when memoranda were presented at the same location than when memoranda were presented at different locations. However, this effect of stimulus location disappeared in the long ISI conditions, without any decrease in overall performance. Our results suggest that the increase in swap errors is due to interference at the level of perception rather than due to a binding deficit specific to working memory. We argue that temporal order can replace the special role of location in feature binding when items are presented sequentially.
Meeting abstract presented at VSS 2018