The acquisition of speech rhythm by bilingual Spanish- and English-speaking four- and five-year-old children

This study investigates the acquisition of speech rhythm by bilingual Spanish- and English-speaking children between four and five years of age, and it compares their performance to that of their monolingual age-matched peers in both languages, as well as to the performance of monolingual and bilingual adults. Participants included younger children (between 3;9 and 4;5,15), older children (between 4;6,18 and 5;2), and adults (over 18 years). All age groups had data collected from bilingual speakers of English and Spanish as well as monolingual age-matched peers in both languages. A sentence elicitation task was used to obtain the data recorded directly onto a notebook computer using an external sound card and a wireless lavalier microphone system. The child participants' hearing and language were also evaluated to ensure typical and age-appropriate language skills and normal hearing. Sixty-four samples were collected and later analyzed via durational variability calculations of vocalic and intervocalic intervals based on E. Grabe and E. Low's (2002) normalized Pairwise Variability Index (PVI) measures so as to investigate the speech rhythm production of the participants.; The results revealed that younger bilingual children displayed distinct speech rhythm patterns for their target languages, but they also tended to deviate from their monolingual English-speaking peers, showing language interaction and somewhat of an equal speech timing bias. Older bilingual children also produced significant separation of speech rhythm, and differences between older bilingual children and their monolingual peers speaking English were also found. Comparisons between younger and older bilingual children indicated an effect for the vocalic PVI measure, but not for the intervocalic PVI, providing partial support for maturational differences for bilingual children. Contrasts between bilingual adults and bilingual children illustrated the existence of age effects. Furthermore, bilingual adults also show separation of their languages on the whole and do not differ on measures of speech rhythm from their monolingual peers in either language. Further analyses suggested that the normalized vocalic PVI measure was more stable and had more power to differentiate languages than the intervocalic PVI. Post-hoc analyses also found the normalized vocalic PVI to be unaffected by including or excluding the final vowel.