Academic journal article Canadian Journal of Experimental Psychology

Unattended Words Need to Be Primed to Be Recognized

Academic journal article Canadian Journal of Experimental Psychology

Unattended Words Need to Be Primed to Be Recognized

Article excerpt

Unattended words need to be primed to be recognized

Abstract The categorical relation between a target word and a flanking, to-be-ig nored, nontarget word can influence target response. Although usually taken as e vidence of a full and automatic analysis of stimuli whether or not they have bee n attended, this flanker effect may only point to the failure of focused attenti on when nontarget stimuli have been primed and made task-relevant. The present s tudy examined the role of priming in the flanker task. In one condition, schemat ic and semantic priming of nontargets was potentiated by having subjects categor ize the target as an instance of a living or nonliving thing. In a second condit ion, priming was minimized by requiring only a shallow analysis of the target fo r a response; subjects searched the target for the presence of the letter R. A f lanker effect was found only in the categorization condition, and then only when the target was the name of an animal. There was no evidence that unattended non targets had been fully and automatically encoded to a semantic level.

Evidence of capacity limitations in human information processing has led neurosc ientists, computationalists, and psychologists to distinguish between two stages or levels of processing. At a preattentive stage, processing is automatic (inde pendent of strategic control), unlimited in capacity, and spatially parallel. Ou tput from the preattentive stage passes to an attentive stage where processing i s controlled, limited in capacity, and serial (e.g., Enns & Rensink, 1991; Posne r & Presti, 1987; Shiffrin & Schneider, 1977; Treisman, 1986).

There is a consensus that simple object properties and features such as orientat ion, size, and location are analyzed preattentively. But there is an ongoing deb ate over whether the semantic attributes of familiar objects are also preattenti vely available (Folk & Egeth, 1989; Yantis & Johnston, 1990). In reaching for a cup, I successfully ignore pens, papers, and other spatially-adjacent objects. B ut what happens to my sensory representations of these unattended objects? Are t hey processed only to the level of shape and location information? Or are they f ully processed to the level of meaning? At what point in stimulus processing doe s selective attention occur?

According to one of two main theoretical positions on the locus of selection que stion, objects are selected for attention from the detection of task-relevant fe atures. It is only when these features are integrated through the operation of a n internal process of focused attention that objects recognition occurs and that the meaning of an object becomes available to conscious awareness. In this earl y-selection view, selection is by default. At the time my attention is focused o n the cup, other unattended objects do not gain internal representation in terms of their meanings; only their low-order properties are encoded (Broadbent, 1982 ; Kahneman & Treisman, 1984; Treisman, 1986).

Another theoretical position holds that all objects within the limits of sensory resolution, attended and unattended, are automatically analyzed to a semantic l evel of representation. In this late-selection view, the mechanism of selection is presumed to operate either (a) on the relevant and attended representations ( for their transfer to, or maintenance in, a limited-capacity store) while ignore d representations decay (e.g., van der Heijden, 1981), or (b) on the irrelevant and ignored representations, which are actively inhibited (e.g., Tipper & Driver , 1988).

The Flanker Effect

The flanker effect provides potentially strong evidence for the late-selection p osition. In the simplest task, two items are shown at tachistoscopic speeds, one positioned where the subject is fixating and the other positioned at a to-be-ig nored spatial location. The semantic or associative relations between the focall y-attended target item and the unattended nontarget item influence target respon se time (RT), showing that the meaning of the nontarget was access even though i t was not focally attended. …

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.