Abstract: Working memory (WM) and attention are regarded as two of the most important cognitive faculties. WM includes many core cognitive processes such as information processing, temporary dynamic storing, and executive controlling. Attention acts as a “gatekeeper” to the information processing system for accessing external information. The interaction between WM and attention forms the central mechanism of the information processing system. Several previous studies have found that when visual working memory (VWM) shares some features with visual search items, the content of VWM guides attention orienting in a flexible manner at the attention stage. However, the processing of visual attention includes two stages: preattention stage that has a limitless capacity and witnesses parallel processing and attention stage that has a limited capacity and witnesses serial processing (see Wolfe, 1998). For many decades, several researchers have considered that attention orienting at the preattention stage is automatic and involuntary and that it is not controlled by consciousness. It is unclear whether the content of WM can completely guide attention orienting at the preattention stage. Further, the manner in which WM guides attention orienting and the difference of guidance at this stage from that at the attention stage constitute extremely important questions. However, few studies have explored these issues. The current research combines WM tasks and feature-based visual search tasks and explores how the contents of VWM affect attention orienting at the preattention stage.
Thirteen undergraduates, including 6 men, participated in Experiment 1 that investigated whether or not the orientation-feature-based visual search task was parallel. This visual search task was also used in Experiments 2 and 3. Experiment 2 comprised of Experiment 2a, involving 16 subjects (8 men), and Experiment 2a, involving 17 subjects (8 men), and aimed at ascertaining the effect of attention orienting by using VWM when the content of VWM could be the target of the next parallel visual search (target-relevant condition). Experiment 3 that involved 17 subjects, including 8 men, ascertained how VWM affects attention orienting when the content of VWM is not the target of the next visual task (target-irrelevant condition). All the experimental programs were compiled with E-Prime 1.1 and run on an IBM Pentium IV computer with a 17-inch (1024 × 768) LCD monitor.
The results revealed the following: (1) in Experiment 1, the reaction time (RT) and error rate were almost equal in every condition, which suggested that the visual search task herein was a feature-based visual search task; (2) the results of Experiments 2a and 2b indicated that the participants searched for the target more slowly in the distractor-matched condition than they did in the target-matched and control conditions; and (3) there were no differences between the distractor-matched and control conditions in Experiment 3.
The results suggested that even at the preattention stage, attention orienting was top-down modulated by VWM. However, at the attention stage, when the items related to the following visual search target were in VWM, this item guided attention to its location, slowed the processing of the target when employed as a distractor in the visual array, and exhibited no effects when employed as the target in the visual array. However, when the content of VWM was irrelevant to the next search target, attention was not given to the distractor that matched the content of VWM and the target was efficiently processed

Key words: visual working memory, attentional orienting, target-relevant, target-irrelevant