Abstract
Feature Integration Theory (FIT, e.g., Treisman & Gelade, 1980) suggests that the presence of a feature singleton can be detected preattentively based on activity in a feature map. We first replicated a result of Theeuwes, Van der Burg, and Belopolsky (2008) which seems to refute this prediction of FIT. Subjects responded to the presence or absence of a single red letter in a circle of gray letters surrounding fixation. Subjects were faster to indicate the presence of a red letter when that letter's identity was repeated from the previous trial, even when spare attentional capacity was occupied by a demanding second task. There was no repetition priming when a gray letter was repeated. Assuming that the letter form of the color singleton could only be picked up by focal attention, these results suggest that focal attention was directed to the letter, even in a simple feature detection task. However, the use of letters may have biased the outcome. It has been suggested that in some circumstances alphanumeric characters may elicit “compulsive encoding” so it is possible that participants were involuntarily directing their attention to the red letter in order to read it even though the task simply required detection of a color (Teichner & Krebs, 1974; see also Stroop, 1935). In a subsequent experiment, we replaced the English characters with unfamiliar symbols (Chinese letters). This eliminated intertrial priming effects in both single and dual task conditions. Detection of the red letter remained highly accurate, as with the English characters. A control experiment confirmed that intertrial priming effects are possible in the same task with Chinese characters when attention was directed to them with a spatial precue. The results suggest that the detection of a feature singleton does not require the application of focal attention.