Abstract

Color repetition blindness (CRB) refers to a poorer identification of a repeated color than that of a nonrepeated color under perceptually-limited viewing conditions. Four experiments tested representational as well as attentional assumptions of token-individuation failure hypothesis (Kanwisher, 1991). Experiment 1 varied the brightness of target colors and Experiment 2 changed the contour brightness, both to explore the nature of color tokens. CRB disappeared when the subjects could not use brightness tokens. Experiment 3 explored the effects of attention allocation to input colors in the pre-and postcue conditions. Experiment 4 used both moving and static color inputs. These two experiments showed that attention allocation has nothing to do with CRB. The results of four experiments were not consistent with predictions derived from the token-individuation failure hypothesis. Instead, many of our results were consistent with predictions based on a token-based inference hypothesis regarding CRB.