Abstract
Repetition blindness (RB, Kanwisher, 1987) is the failure to perceive the second occurrence of a repeated item in rapid serial visual presentation (RSVP). We have shown in previous studies (Yeh & Chen, 2002; 2003, Yeh & Li, in press) that RB for components in Chinese characters is a robust effect (e.g., ?-? was reported as ?-? by omitting the semantic component ?). However, all these findings of component RB were obtained when the occurrence frequency of the two critical characters (C1, C2) were held constant. The goal of this study was to examine whether RB for components could still be observed when the relative frequency of C1 and C2 was manipulated, and whether the size of RB would be affected by the semantic context. In each RSVP sequence, 3 unrelated characters and 4 symbols or 4-character word and 3 symbols were shown, and the participants were asked to write down all the characters. The relationship of the two critical characters was manipulated as follows: (1) the frequency of C1 was higher or lower than that of C2, and (2) the repeated components of C1 and C2 were semantic components or they were phonetic components, and (3) C1 and C2 shared a repeated component or they did not share any component. Results showed that RB for components was still reliably obtained even when the frequency range of C1 and C2 was not the same. In the unrelated condition, RB was higher for high-frequency C1 than for low-frequency C1. In the four-character word condition, only RB for phonetic components was found, but not RB for semantic components. These results suggest that component RB in Chinese characters is modulated by familiarity and semantic context, reflecting the similarity and the difference between English and Chinese.
Supported by the National Science Council of Taiwan, NSC92-2413-H-002-026