The feature [sonorant] in auditory word identification

The distinctive feature [sonorant] separates resonants and obstruents. Does this feature mark a difference in the perceptual processing of these consonants? Marks et al. [Marks et al., ‘‘Word reconstruction and consonant features in English and Spanish,’’ Linguistics, 40, 421–438 (2002)] showed such an influence in a word reconstruction task. Target words containing an obstruent segment were harder to reconstruct when the obstruent was replaced by a resonant (mismatch condition) than when it was replaced by another obstruent (match condition). Resonant target words showed no such effect. The present study extended Marks et al. (2002) using an auditory priming lexical decision task. Participants heard a nonword prime 50 msec prior to another word or nonword target. The prime differed from the target item in just one consonant which either matched the target segment on the feature [sonorant] or mismatched it. Listeners judged whether the target item was a real word or nonword. The obstruents were divided into fricatives and stops, which were compared with resonants in both cases. Latencies for the fricative target words were significantly slower in the mismatch condition than in the match condition but not for stops. The present study validates the effect found by Marks et al. (2002) with an online task but shows that it is limited to fricatives.