Abstract
When targets are presented in the visual periphery with nearby flankers they may be crowded to the extent that they are unidentifiable. To what extent is the crowded target processed in the brain? A central question in crowding research is the nature of the processing for crowded information as the answer to this question will reveal where and how this visual limit is imposed. Here we examined whether semantic information can be extracted in the crowded situation using a primed lexical decision task with Chinese characters as stimuli. The prime and the target were semantically related or unrelated, and the prime was presented in isolation or with flankers. Manipulation checks on the effectiveness of crowding were conducted to allow the adoption of stringent criteria on a subject-by-subject basis for analyzing the semantic priming effect: Only the trials in which the prime was identifiable when presented alone and not identifiable when crowded were included. Results showed clearly that despite being unrecognizable and undistinguishable from matched non-characters, crowded Chinese characters still generated robust semantic priming in a subsequent lexical decision task. Indeed, the semantic priming effect on the lexical decision task was similar whether or not the primes were crowded and unrecognizable or not crowded and recognizable. These findings show that crowding does not prevent semantic activation suggesting that the locus of crowding lie beyond the level at which semantic information is extracted, and further that semantic activation may pool information over a smaller spatial extent than that sampled for identification.
This study is supported by grants from Taiwan's National Science Council, NSC96-2413-H-002-009-MY3 and NSC98-2410-H-002-023-MY3.