Attentional Allocation within the Syllabic Structure of Spoken Words

Abstract

The role of the syllable during on-line speech perception was explored using a variant of the phoneme detection task developed by Pitt and Samuel (1990). In their task, listeners′ attention to phonemes in different serial positions inside word or nonword stimuli was manipulated by varying the probability that a target phoneme occurred in the various positions. In our experiments, French and Spanish subjects had to detect targets that appeared either in the coda of the first syllable or in the onset of the second syllable of carrier words. Subjects′ expectations about the structural position of the target were manipulated. In a series of five experiments (two using a decision paradigm and three using a detection paradigm), these expectations were shown to influence response latencies: that is, subjects who attended to the coda of the first syllable were faster when the target appeared in this position rather than in the onset of the second syllable; the reverse pattern was observed when subjects attended to the onset of the second syllable. This result held regardless of the serial position of the target. These results were equally valid for French and Spanish. Moreover, syllabification was present when Spanish pseudowords were used as carriers. The fact that subjects could focus their attention on a syllabically defined position, even when processing nonwords, suggests that syllabic information is specified at a prelexical level of representation. The phoneme detection task in which attention is manipulated provides us with an interesting new technique for exploring prelexical representations.