| Summary: | Lukatela and Turvey (1994) showed that at prime presentation duration of 57 ms naming of a visually presented target word (frog) is primed not only by an associate word (toad), but also by a homophone of the associate (towed) and a pseudohomophone of the associate (tode). At prime presentation of 250 ms, priming with the homophone was no longer observed. Lukatela and Turvey's interpreted these findings as evidence for a strong phonological activation-verification model of visual word recognition, which entails that lexical representations are activated on the basis of a phonological code and subsequently disambiguated by a lexically mediated spelling check if more than one spelling corresponds to the phonological code. Four experiments are reported that further addressed the issue of phonologically mediated associative priming in visual word recognition. In Experiment 1, we replicated Lukatela and Turvey's findings in the Dutch language. Next, we demonstrate that the effect is not confined to the naming task but is also obtained in a lexical decision task with non-homophonic non-words (Experiments 2 and 3). Finally (Experiment 4), we show that when the lexical decision involves a word/pseudohomophone decision, phonologically mediated associated priming is still observed at 57 ms when the prime is a pseudohomophone of the associate (tode-frog) but not when the prime is a homophone of the associate (towed- frog). The results are interpreted within Lukatela and Turvey's activation verification model and we present evidence why we believe that a prime presentation duration of 57 ms yields a better estimate of the time course of the spelling check than prime presentation duration of 250 ms previously reported.
|
|---|
