Abstract
Recent studies have revealed that the deployment of visual attention can be biased by the content of visual working memory (VWM). However, stored visual memories do not always interact with attention. This has led to a model which proposes a functional division within VWM, between a single active template that interacts with perception and multiple accessory memory representations that do not. The present study investigated whether memory-based attentional guidance is indeed limited to a single representation or whether multiple items in memory are able to bias attention. Participants performed a visual search task while maintaining a variable number of colors in VWM. We observed increased attentional capture by memory-related distractors when VWM was filled with a single item. However, memory-related capture disappeared completely for memory loads beyond a single item. The absence of memory-related capture at higher VWM loads was not dependent on individual VWM capacity, nor was it attributable to weaker encoding, forgetting, or reduced precision of memory representations. When analyses were limited to those trials on which participants had a relatively precise memory, there was still no sign of attentional guidance at higher loads. However, when observers were required to remember multiple items, but were then cued towards a specific memory item after encoding, interference with search returned for the cued item. Uncued items did not interfere with search even though the task still required them to be remembered. These results are consistent with a multilayer model of VWM, which makes a distinction between representations that interact with perception and those that do not. Only a single item can have access to the visual input, and cueing an item, or simply being the single item in VWM, automatically causes it to acquire this status of attentional template.
Meeting abstract presented at VSS 2014