Prosody and syntactic structures in continuous speech in French Free

Infant-directed speech contains dominantly multi-word utterances. Segmenting speech into linguistic units is crucial for language acquisition. This study inquires if prosodic cues exist in speech and mark syntactic units. Participants were Quebec-French speakers. In Experiment 1 participants read determiner+noun and pronoun+verb utterances. Nouns and verbs were pseudo-words (e.g., mige, krale) counterbalanced in their occurrences in the utterances. Their prosodic properties (duration, pitch, intensity) were measured. Results showed that the two categories did not differ in prosody: noun versus verb productions of these pseudo-words were equivalent. Experiment 2 tested whether larger utterances were produced with prosodic cues supporting grammatical categories and phrasal groupings. The same pseudo-words were the final words (counterbalanced) in 1) [determiner+adjective+noun] and 2) [[determiner+noun]+verb] structures. Results showed that the last word as nouns versus verbs differed significantly in duration, pitch and intensity. Moreover, the initial consonant of verb productions was longer, with a distinct preceding pause. The word preceding the verb (2) exhibited categorical and boundary cues, differing from the word preceding the noun in (1) in duration, pitch and intensity. We suggest that the acoustic cues may help infants first parse larger utterances and then acquire the syntactic properties of phrases and words based on their distribution.