Research On Content-based Working Memory-driven Capture Of Visual Attention And Its Automaticity

Since there are usually much more stimuli in the scene than limited capacity visual system can process at any moment, visual attention selects only a minority of them for further processing. This selection process can be controlled either by bottom-up factors (e.g., salient stimuli in the visual scene) or by top-down factors, such as subjective intention or task target. The biased competition model suggests that working memory may be crucial to resolve the competition for selection amongst different stimuli in the visual field. Top-down feedback from the target template held in working memory will bias attention in favor of the matching objects. The present research focused on this content-based memory-driven attention mechanism and specifically investigated whether distractors that match the current contents of working memory can capture attention. A total of ten experiments were reported. Experient 1-3 explored attentional capture by visual working memory, and the results showed the reliable effect of working memory on selection which could be established on both feature-based and dimension-based matching between memory items and attention displays. Moreover, this effect could also occur when memory items were verbally encoded (Experiment 4-5). However, the top-down guidance of attention did not strictly satisfy the intentionality criterion for strong automaticity, since participants' state of attentional readiness could modulate attentional capture by working memory contents (Experiment 6). The present research makes strong contributions for understanding the automatic guidance of attention from working memory. First, since participants were provided with no explicit incentive for attending to memory matches in the present task paradigm, the current findings provide the strongest evidence for content-based working memory-driven capture of attention. Second, the present research extends previous studies by further demonstating that top-down guidance of attention can be based on different types of matching between working memory contents and attention objects, from feature-based matching to dimension-based matching. Finally, the present research is, to my best knowledge, the first attempt to directly test the intentionality criterion for the automaticity of attentional capture by working memory contents. A novel notion that content-based working memory-driven capture of visual attention is conditionally automatic is proposed here, which provides an important perspective for resolving the controversy as to whether working memory can involuntarily guide attention.