On the inhibitory consequences of visuospatial orienting: Inhibition of return?

Abstract

Responding is typically slowest toward (an “output” bias) or about (an “input” bias) targets presented at the location indicated by a spatially uninformative visual transient (a “cue”) when the onset asynchrony between the cue and target exceeds ~300 ms. The expression of this “inhibitory” (Posner, & Cohen, 1984) aftereffect depends crucially on the response characteristics of the task. If the task at any time requires reflexively-generated saccadic responses, manual and saccadic responses are slowest toward the cued location. If the task expressly forbids reflexively-generated saccadic responses and instead requires exclusively manual responses, responding is slowest about targets presented at the cued location (Taylor, & Klein, 2000). By mapping the time course of the “inhibitory” aftereffects of cueing, exploring their reference frames and manipulating response properties of the target, we identify (at least) three independent “inhibitory” cueing effects (ICEs). One form of ICE is “inhibition of return” (IOR; Posner, Rafal, Choate, & Vaughan, 1985). IOR is generated by cue-elicited activation in pathways responsible for reflexively-generated saccades (the superior colliculus) when the task involves reflexive, saccadic responses to transient onsets. Another form of ICE is “sensory adaptation” (Fecteau, & Munoz, 2006). Sensory adaptation is short-lasting and occurs most robustly when the oculomotor system responsible for reflexive saccades is in an active state and when two visual transients appear in close spatiotemporal proximity (as when the cue and target occupy the same location in space and when their onset asynchrony < 500 ms). A final form of long-lasting ICE is a bias against inputs at a cued object location when the oculomotor circuitry responsible for reflexively-generated saccades is tonically inhibited. In most cases this effect is a likely side-effect of cue-elicited response activation in the manual effector system (de Jong et al., 1994) and may be re-mapped into object-based coordinates by posterior parietal networks unique to the manual effector system. Unlike IOR, with which it is frequently confused, this form of ICE is expressed as an effect nearer the input end of the processing continuum.