Abstract
AbstractThe world around us is inherently structured and often repetitive. Research has shown that we can implicitly learn to prioritize relevant objects and locations while filtering out distracting information, creating an integrated priority map for attention allocation. The current study examines whether these attentional biases are tied to environment-dependent (allocentric) or viewpoint-dependent (egocentric) coordinates. The search display consisted of six stimuli that were surrounded by a wheel and square frame. In two experiments, either a distractor or a target appeared more frequently in one location, leading to the suppression or enhancement of that location, respectively. Learning blocks were followed by test blocks, where the surrounding frame rotated, creating egocentric-matching and allocentric-matching locations. These experiments showed that both target and distractor learning relied on an egocentric reference frame only. In follow-up experiments, the likely target and distractor location rotated dynamically with the frame during learning. This revealed that participants can learn to enhance a likely target location in an object-centered, allocentric manner. We hypothesized that while space-based learning feeds into a priority map reliant on an egocentric reference frame, object-based learning allows for implicit prioritization of subparts of objects independent of their spatial orientation.Public significance statementAttention is shaped by past experiences, guiding us to suppress locations likely to contain distractors while enhancing locations likely to contain relevant information. This study explores how these attentional biases behave when the search environment is viewed from different perspectives. Do these biases persist relative to our viewpoint, or do they remain stable within the environment? The findings reveal that implicitly suppressed and enhanced locations are tied to viewpoint-dependent coordinates. However, attentional biases can also be formed in an object-centered manner, where a likely target location within an object is prioritized, irrespective of the orientation of that object.
Publisher
Cold Spring Harbor Laboratory