ISSN : 1226-9654
The hypothesis that activation of phonological representation is obligatory in visual access of word meaning was tested. In a series of five experiments which differed only in stimulus onset asynchronies (SOA; 51ms, 85ms, 119ms, 153ms, and 255ms, respectively) and subjects, a target word was named following an semantic associate of the target or a word homophonic with the associate. Semantic associates produced reliable priming regardless of the length of SOA, whereas homophones produced marginally significant priming at SOA 153 ms only. The results suggest that visual access of word meaning is primarily accomplished via activation of orthogrphic codes, and that the phonological constraint hypothesis is not tenable.