Abstract
Attentional capture serves an important behavioural function—to ensure that salient stimuli in our environment are selected for detailed processing. But what determines salience? This question has primarily been studied by asking what types of low-level stimulus features capture our attention, such as colour or luminance, and how such capture by low-level features is regulated by our top-down goals (i.e., contingent on our attentional control settings). Building off of recent demonstrations that saliency can also be determined by higher-level representations such as conceptual knowledge (Wyble, Folk, & Potter, 2013, JEP:HPP), learned value (Anderson, Laurent, & Yantis, 2011, PNAS), and statistical regularity (Zhao, Al-Aidroos, & Turk-Browne, 2013, Psych Sci), we asked whether stimulus salience can be established based on long-term episodic memory representations. When searching our environment for one of a number of items stored in long-term memory, will stimuli resembling those memories automatically capture attention? To evaluate this question we asked subjects to memorize a set of 30 visual objects and then complete a Posner cueing task, for which the targets were any one of the studied visual objects. Cues were studied or novel objects, either matching or not matching the subjects' attentional goals, respectively. We observed that matching cues produced a cueing effect, but not non-matching cues, suggesting that attentional control settings can be specified based on long-term memory representations, potentially episodic memories. These results add to the growing evidence that salience reflects a complex interaction between internal representations and external features. More broadly, the results also suggest a role for visuospatial attentional capture during the retrieval of long-term memory representations, which may contribute to the rapid recollection observed in recent long-term memory studies (Aly & Yonelinas, 2012, PloS; Guild et al., 2013, PB&R).
Meeting abstract presented at VSS 2014