Authors:

Duncan, J.

Reference:

Year of publication:

1985

CBU number:

1799

Abstract:

Research on visual attention deals with the limit on our ability to identify several stimuli at once. In visual search, this leads to concern with both the number of targets and the number of nontargets in a display. Although research usually concentrates on the number of nontargets, I suggest that it is useful to consider targets and nontargets together. Effects of multiple targets motivate the distinction between a first, parallel stage of perceptual processing, and a second, limited-capacity stage (Broadbent, 1958). They show that even when a target attribute can be used at the first stage to guide access to the second, the attribute cannot be reported until such access has occurred. This holds even for simple attributes such as tilt and closure. When an experiment manipulating the number of nontargets suggests parallel search, this does not show that the task can be performed without involving the limited-capacity stage. Rather, it shows the efficiency of selective access to that stage, and as such cannot be clearly interpreted without knowing the effect of multiple targets. One way to put the conclusions might be that many stimulus attributes are “identified” at the first stage but cannot yet be reported. I suggest, however, that it is misleading to talk of a stage at which information is “identified” or “analysed”. The enterprise instead is to understand the different processes of information use.