The effect of attentional load on modal and amodal completion

Perceptual completion or filling-in is a remarkable ability of visual system to interpolate missing information from retinal input. It remains unclear whether modal and amodal completion or filling-in involve different or common neural mechanisms. In this study, we measured tilt after-effect (TAE) after prolonged adaptation to modal and amodal illusory gratings and no-filling-in control stimuli under different attentional load. Four moving gratings were presented in the apertures of a dark occluder. Modal or amodal completion was achieved by manipulating the relative depth between the inducers and the occluder. The phase and speed of the moving gratings were scrambled to generate the no-filling-in control stimuli. In the passive (or no-load) condition, subjects maintained fixation and judged the orientation of probes. In the low and high attentional load conditions, subjects performed either single- or conjunction-feature letter detection tasks in a rapid sequence visual presentation (RSVP) of colored letters. In the passive condition, both modal and amodal filling-in produced stronger TAE than their control counterparts, at similar magnitude. Diverting attention away from the stimuli almost eliminated TAE in both conditions, except for a marginal effect in modal filling-in under low attentional load. Our findings suggest that modal and amodal completion generate similar orientation representations at the early stage of visual processing, and both require attention. Whether lateral or feedback mechanisms are differentially involved in these two filling-in phenomena requires further investigation with neuroimaging techniques.