PROMOTING NOVELTY IN VISION: Inhibition of Return Modulates Perceptual-Level Processing

To facilitate visual search of complex scenes, information arising from recently attended locations is subject to a selective inhibition in processing known as inhibition of return (IOR). Although the mechanisms of IOR remain unresolved, both motor and perceptual influences have been proposed based on reaction time (RT) studies. Here we report the results of two reflexive cuing studies in which signal detection methodology was employed to directly examine the effects of IOR on perception. IOR was found to be associated with a significant reduction in the accuracy of target discriminations at recently attended locations. Further, these effects of IOR on response accuracy were independent of whether emphasis was placed on the speed of responding. These results provide the first direct evidence that IOR can affect the perceptual quality of visual processing. Humans are remarkably efficient at orienting to relevant events in the visual world. Effective orienting is subserved in part by a temporary inhibition in neural processing that is associated with recently attended locations in the visual field. This inhibitory phenomenon— known as inhibition of return (IOR)—lasts for several seconds, and follows a brief facilitation in processing attributable to reflexive attention (e.g., Posner & Cohen, 1984; for a review, see Taylor & Klein, in press). IOR can be characterized as a selective inhibition in processing because only information arising in recently attended locations is inhibited (see also Tipper, Weaver, Jerreat, & Burak, 1994). Although it is generally agreed that IOR promotes attending to novel events in the visual field by inhibiting recently attended information, there has been contentious debate over whether IOR can occur at perceptuallevel stages of processing. Initial efforts to examine perceptual effects in IOR involved the use of temporal-order judgments (TOJs), during which subjects must decide which of two presented stimuli occurred first. Both Maylor (1985) and Posner, Rafal, Choate, and Vaughn (1985) found that IOR did not affect TOJs, which was taken as evidence that IOR does not involve perceptual inhibition because the temporal characteristics of the visual input remained unperturbed. Although Gibson and Egeth (1994) cast doubt on these conclusions, showing that IOR can affect TOJs under certain experimental conditions, their positive results have now been questioned (Klein, Schmidt, & Müller, 1998). As a result, the data from studies of TOJs remain equivocal on whether IOR can affect perceptual processing (see also Schmidt, 1996). Perceptual effects in IOR have also been examined using reaction time (RT) paradigms. In similar studies, Abrams and Dobkin (1994) and Rafal, Egly, and Rhodes (1994) measured saccadic RTs as a function of whether or not the imperative stimulus for moving the eyes required the detection of a nonfoveal target. Although Abrams and Dobkin and Rafal et al. both concluded that IOR can inhibit stimulus detection, motor-level inhibition was also implicated in their results (see Klein & Taylor, 1994; Rafal & Henik, 1994), and both studies were faced with the problem of dissociating perceptual from motor effects. As a result, although Abrams and Dobkin and Rafal et al. agreed that perceptual-level IOR can occur, they ultimately disagreed over the specific conditions in which it is manifest, underscoring the inherent difficulty of ruling out motor-level explanations of IOR when using RT measures (see, e.g., Müller & von Mühlenen, 1996). The question thus remains: Can IOR affect perceptual-level processing? In the following study, we adopted a novel approach to this question based on signal detection measures of perceptual sensitivity (e.g., Macmillan & Creelman, 1991). Because these measures reflect the accuracy of target responses, they should remain relatively unaffected by any motor-level inhibition that may contaminate RT measures of IOR, thus providing a more direct test of whether IOR can modulate perceptual processing.