Abstract:

PURPOSE: The present study aimed to examine whether two visual working memory (VWM) representations can simultaneously guide attention and automatically interfere with concurrent visual search when participants are required to memorize two items with feature conjunctions. In addition, we tested whether collective representations of multiple features or representations of integrated objects with feature conjunctions guide attention.
METHODS: The present study adopted the typical dual-task paradigm, participants were required to memorize two objects before starting a visual search task which was completely irrelevant to the memory task.
RESULTS: The present study showed that two VWM representations can involuntarily capture attention and interfere with concurrent visual search. In addition, each of two VWM representations can interfere with concurrent visual search as much as can a single cued VWM representation, irrespective of whether the cue validity is 75% (Experiment 3), or 100% (Experiment 4). Moreover, each of two VWM representations can also produce a large memory-driven capture e?ect, whether or not there is a short (800msec) or long (3000msec) SOA between memory display and visual search.
CONCLUSIONS: Thus, two VWM representations can simultaneously control attention. Furthermore, when participants are required to memorize two objects de?ned by a conjunction of color and texture, a distractor that matches the color of one memorized item and the texture of the other memorized item (such that it is not fully matching either memorized item) can also capture attention. This might suggest that features of the memorized object are free-?oating in VWM.

Key words: working memory, attentional capture, visual representation