On analog movements of visual attention

On analog movements of visual attentionSTEVEN YANTISJohns Hopkins University, Baltimore, MarylandA great deal of evidence has accumulated over the last20 years or so suggesting that decisions concerning thecontents of a visual .scene can be enhanced with the as-sistance of advance information about the spatial locationor locations in the scene that are likely to contain task-relevant information. For example, Eriksen and his col-leagues have demonstrated that reaction time to identifya target letter in visual search is improved when a spatialcue is provided about 200 msec before the onset of thesearch display (Colegate, Hoffman, & Eriksen, 1973;Eriksen & Hoffman, 1972, 1973, 1974). Eriksen andHoffman (1972) suggested that the spatial extent of theenhancement is limited to about 1 o of visual angle (seealso Eriksen & St..lames, 1986). Posner and his col-leagues have also provided evidence that attention can bespatially directed in response to an appropriately timedcue. Reaction time in a dot-detection task is enhancedwhen the dot appears in an attended location and is slowedwhen it appears in an unattended location (e.g., Posner,1980; Posner, Snyder, & Davidson, 1980). Jonides (1981)demonstrated that the effectiveness of spatial cues in focus-ing attention varies ’with their location: Peripheral cuesapparently capture attention automatically, whereas cen-tral cues require effortful allocation procedures. LaBerge(1983) and Downing and Pinker (1985) showed that whensubjects focus attention on a restricted spatial location,reaction time to objects at unattended locations increaseswith their distance from the focus. There is also evidencethat attention is more efficiently applied to objects thatare physically proximal to one another (e.g., Hoffman N Kahneman & Henik, 1981; Podgorny S Posner et al., 1980).A quite natural heuristic for this body of results is thatthe application of spatial attention is analogous to shin-ing a spotlight on the visual field. Objects that fall underthe "spotlight of attention" can be identified or detectedmore rapidly than other objects. This sort of analogy hasbeen advocated by many writers (e.g., Broadbent, 1982,p. 271; Downing & Pinker, 1985; Eriksen & Hoffman,1973; Eriksen & St. James, 1986; Eriksen & Yeh, 1985;LaBerge, 1983; Posner, 1980; Posner, Cohen, & Rafal,1982, p. 188; Posner et al., 1980, p. 172; Shulman,Remington, & McLean, 1979; Tsal, 1983).1 Only a fewtheorists (e.g., Duncan, 1981) question the viability ofthe spotlight metaphor. Certainly the spotlight metaphoris held with varying degrees of concreteness and differ-