ABSTRACT
Can alphanumeric characters be identified in parallel, or must they be identified sequentially? Previous experimental work on this issue has typically employed tasks in which subjects attempt to detect pre-specified target characters. Unfortunately, such tasks do not logically require either character identification or accurate conjoining of features. The present experiments used a task designed to force exhaustive identification of digits—naming the highest digit in an array of digits—and also a detection task involving stimuli selected to require accurate feature conjunction. Subjects’ accuracy in simultaneous versus rapidly successive presentations of briefly exposed characters was compared. The effects of predictability and order of presentation sequences were also examined. None of these variables produced the effects one would expect if character identification was operating sequentially. We argue that evidence for parallel identification, such as that presented here, does not have particularly direct implications for the issues of capacity limitations, voluntary attentional control, and early versus late selection, although these issues are commonly conflated. The independence of these issues is emphasised by contrasting the present results with a number of earlier findings; the contrast also leads us to a conception of how visual selective attention might operate that is rather different from the traditional early or late selection accounts.
