The Temporal Dynamics of Willed Attention in Vision

Abstract

Most models of attention distinguish between voluntary and involuntary attention, the latter being driven in a bottom-up fashion by salient sensory signals. Studies of voluntary visual-spatial attention have used informational or instructional cues, such as arrows, to induce or instruct observers to direct selective attention to relevant locations in visual space in order to detect or discriminate subsequent target stimuli. In everyday vision, however, voluntary attention is influenced by a host of factors, most of which are quite different from the laboratory paradigms that utilize attention-directing cues. These factors include priming, experience, reward, meaning, motivations, and high-level behavioral goals. Attention that is endogenously directed in the absence of external cues has been referred to as self-initiated attention, or in our prior work as "willed attention". Such studies typically replace attention-directing cues with a "prompt" that signals the subject when to choose where they will attend in preparation for the upcoming target stimulus. We used a novel paradigm that was designed to minimize external influences (i.e., cues or prompts) as to where, as well as when, spatial attention would be shifted and focused. Participants were asked to view bilateral dynamic dot motion displays, and to shift their covert spatial attention to either the left or right visual field patch at a time of their own choosing, thus allowing the participants to control both when and where they attended on each trial. The task was to discriminate and respond to a pattern in the attended dot motion patch. Our goal was to identify patterns of neural activity in the scalp-recorded EEG that revealed when and where attention was focused. Using machine learning methods to decode attention-related EEG alpha band activity, we were able to identify the onset of voluntary (willed) shifts of visual-spatial attention, and to determine where attention was focused. This work contributes to our understanding of the neural antecedents of voluntary attention, opening the door for improved models of attentional control, and providing steps toward development of brain-computer interfaces using non-invasive electrical recordings of brain activity.