Abstract
Both phonological and phonetic priming studies reveal inhibitory effects that have been interpreted as resulting from lexical competition between the prime and the target. We present a series of phonetic priming experiments that contrasted this lexical locus explanation with that of a prelexical locus by manipulating the lexical status of the prime and the target and the task used. In the related condition of all experiments, spoken targets were preceded by spoken primes that were phonetically similar but shared no phonemes with the target (/bak/–/dεt/). In Experiments 1 and 2, word and nonword primes produced an inhibitory effect of equal size in shadowing and same–different tasks respectively. Experiments 3 and 4 showed robust inhibitory phonetic priming on both word and nonword targets in the shadowing task, but no effect at all in a lexical decision task. Together, these findings show that the inhibitory phonetic priming effect occurs independently of the lexical status of both the prime and the target, and only in tasks that do not necessarily require the activation of lexical representations. Our study thus argues in favour of a prelexical locus for this effect.
Notes
1The nonword primes have fewer lexical neighbours than the word primes. Hence, similar inhibitory priming effects with word and nonword primes cannot be attributed to strong lexical activation of the neighbours by the nonword primes.
2Experiment 1 does not actually constitute a strict replication in French, given that similarity between primes and targets in the English studies was based on confusions and subjective similarities rather than on distinctive features.
3The three words excluded from the analyses in Experiment 3 were also excluded from the set of data of Experiment 1B. As a result, the comparison between Experiment 1B and Experiment 3 was performed on exactly the same number of items.