Two series of experiments explored the influence of experimental task, level of processing, and time course in speech production via the priming paradigm. The first series of experiments involved a reading task in which subjects were primed to produce slip-of-the-tongue errors. The targets were all word pairs that slipped to other real word pairs when the first phoneme of each word was reversed (e.g., the target pair rage wait slips to wage rate). Subjects were asked to read aloud the visually-presented target word pairs following visual presentation of priming word pairs that were related to the potential slip error phonologically (e.g., wake rain) or semantically (e.g., salary scale), or were unrelated to the slip error (e.g., table cap). Priming word pairs were presented for either 900 ms or 1500 ms. The second series of experiments involved a task in which subjects were asked to generate target words by producing aloud the opposite of a visually-presented cue word (e.g., on seeing the cue hot, produce cold). Cue words were preceded by primes that were related to the target words phonologically (e.g., cope) or semantically (e.g., ice), or that were unrelated (e.g., sheep). The time between onset to produce the prime and onset of the cue word was varied. Results for the slip-of-the-tongue task indicated that there was a tendency for subjects to show longer production durations for target pairs in the phonological condition than either the semantic or unrelated conditions at the 900 ms prime duration. A significant effect of phonological relatedness on the onset to produce targets was demonstrated for the generate task. Subjects showed longer onset times for the phonological condition than for the semantic or unrelated conditions. These data provide evidence for inhibitory effects in phonological priming. The findings were discussed in terms of their implications for the spreading-activation model of speech production presented by Dell (1986).
展开▼