Abstract
The visual system uses high-level knowledge to predict and constrain incoming sensory signals to increase the speed and efficiency of object recognition and action. Thus, we hypothesized that graspable objects would bias lower-level visual processing toward action-relevant object features (e.g., orientation) while non-graspable objects would bias toward less action-relevant features (e.g., saturation). Participants viewed images of objects that did or did not afford grasping and were located close to or far from the participant. Participants indicated, via a right-handed key press, whether the object decreased, didn’t change, or increased, in orientation or saturation. Participants were faster and more accurate at detecting changes in orientation versus saturation, graspable versus non-graspable objects, and far versus close objects, but no interaction effects were found. Upon eliminating objects that were easily parsed into their constituent elements, and thus encouraged bottom-up attentional strategies that circumvent top-down processes, participants were more accurate at detecting orientation changes in graspable objects and faster at detecting saturation changes in non-graspable objects. The results provide partial support for the hypothesis that object affordances modulate low-level visual processes in an action-relevant manner but highlight the importance of accounting for individual differences in attentional strategies when interpreting experimental results from visual cognition tasks.