Academic journal article Attention, Perception and Psychophysics

Spatial Interference between Attended Items Engenders Serial Visual Processing

Academic journal article Attention, Perception and Psychophysics

Spatial Interference between Attended Items Engenders Serial Visual Processing

Article excerpt

Published online: 10 November 2012

© Psychonomic Society, Inc. 2012

Abstract A pair of experiments investigated the architecture of visual processing, parallel versus serial, across high and low levels of spatial interference in a divided attention task. Subjects made speeded judgments that required them to attend to a pair of color-cued objects among gray filler items, with the spatial proximity between the attended items varied to manipulate the strength of interference between attended items. Systems factorial analysis (Townsend & Nozawa, Journal of Mathematical Psychology 39:321-359, 1995) was used to identify processing architecture. Experiment 1, using moderately dense displays, found evidence of parallel processing whether attended objects were in low or high proximity to one another. Experiment 2, using higher-density displays, found evidence of parallel selection when attended stimuli were widely separated but serial processing when they were in high proximity. Divided visual attention can operate in parallel under conditions of low or moderate spatial interference between selected items, but strong interference engenders serial processing.

Keywords Visual selective attention * Cognitive architecture * Systems factorial technology

(ProQuest: ... denotes formulae omitted.)

The typical visual environment contains too much information for the capacity-limited primate visual system to process at once, a constraint that appears to arise from the neurophysiology of the extrastriate visual cortex (Kastner & Ungerleider, 2000). Although receptive fields (RFs) in V1 are generally 2° or less in diameter (Hubel & Wiesel, 1968), extrastriate RFs can span many degrees of the visual field (Desimone & Gross, 1979; Kastner et al., 2001; Smith, Singh, Williams, & Greenlee, 2001). Because individual RFs are large enough to encompass two or more items simultaneously, a single neural response train can conflate the properties of multiple objects, producing an ambiguous or otherwise degraded representation (Kastner, De Weerd, Desimone, & Ungerleider, 1998; Luck, Girelli, McDermott, & Ford, 1997; Reynolds, Chelazzi, & Desimone, 1999). This effect, a manifestation of the binding problem (Luck & Beach, 1998; Reynolds & Desimone, 1999; Treisman, 1996), limits the perceptual quality with which the objects in a crowded scene can be represented at one time (Tsotsos et al., 1995) and demands attentional mechanisms to isolate and prioritize behaviorally relevant items (Broadbent, 1958; Bundesen, 1990; Desimone & Duncan, 1995; Sperling, 1960; Treisman & Gelade, 1980).

From Broadbent's (1958) filter theory onward, models of attention have often assumed that the capacity limitations of the visual system are circumvented by resort to serial processing. Posner's (Posner, Davidson, & Snyder, 1980) and Eriksen's (Eriksen & Hoffman, 1973; Eriksen & St. James, 1986) spatial selection models, for example, propose that regions of the visual field are attended in sequence by movements of a mental spotlight or lens. Both feature integration theory (Treisman & Gelade, 1980) and the early versions of the guided search model (Wolfe, 1994; Wolfe, Cave, & Franzel, 1989) posit an initial stage of parallel feature processing but hold that because the outputs of these front-end processes are impoverished (e.g., Treisman & Schmidt, 1982; Wolfe & Bennett, 1997), the accurate perception of multifeature objects requires serial selection via shifts of attentional focus. More biologically inspired accounts such as Desimone and Duncan's (1995) biased competition model (Desimone & Duncan, 1995) and Tsotsos's (Tsotsos et al., 1995) selective tuning model propose an early wave of parallel activation, after which objects compete through reciprocal suppression for control of extrastriate neurons. Selection occurs when competition resolves, in a winner-take-all manner, in favor of a single object (e. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.