Aims and objectives/purpose/research questions: The study examines bilingual children’s prosodic competence, specifically the ability to correctly assign word stress in both languages, and contrasts it with participants’ segmental competence. To this end, we estimated and compared the magnitude of prosodic and segmental transfer in L1 and L2 speech of typically developing Polish–English migrant bilingual children raised in the UK. We also explored the influence of cumulative language exposure on both types of transfer. Design/methodology/approach: A non-word repetition task was used in both languages to assess children’s faithfulness in repeating segmentals and stress patterns in two to five syllable-long items. Also, a parental questionnaire on cumulative language exposure was conducted to estimate the quality of input in both languages. All children ( N = 59, M = 5;8) were early bilinguals raised in the UK. Data and analysis: A 2x2 ANOVA and a correlation analysis were conducted to compare the magnitude of prosodic and segmental transfer, within and across languages. Also, multiple regression analysis was performed to establish the predictors of transfer in L1 and L2. Findings/conclusions: The bilingual children repeated stress patterns in both languages correctly, showing resistance to transfer in word stress, even though bidirectional transfer was observed in segmentals. The magnitude of segmental transfer in Polish and English was predicted by cumulative exposure to English, while prosodic transfer in Polish was predicted by cumulative exposure to Polish. Originality and significance/implications: The study provides evidence on word stress placement in the age group of bilingual children, 4–7 years of age, in comparison with segmental data. It uses a novel methodology in comparing the magnitude of transfer between prosody and segmentals, within and across languages. Moreover, the study highlights the importance of language exposure for reducing the amount of segmental transfer in this group of children.

No abstract

Polish is considered a classic example of a bidirectional stress system with internal lapses. Such systems are typologically rare and were recently hypothesized to be non-existent [1]. Latest studies [2, 3], based on comparisons of paired five-and sixsyllable words, revealed that secondary stress in Polish is iterative and is expressed in terms of onset consonant duration. No acoustic study of Polish words having more than six syllables has been conducted thus far. However, heptasyllabic words, unlike six-syllable words, can be revealing not only about the presence of secondary stress iteration but also about the direction of stress assignment ([(σσ)(σσ)σ(σσ)] vs. *[σ(σσ)(σσ)(σσ)]). The present paper reports on an acoustic study of heptasyllabic words compared to segmentally matched five-and six-syllable words (eight triplets, e.g. acetylenowego acetylenuacetylenowy, collected from eight native speakers of Polish). Four parameters are investigated: onset consonant duration, vowel duration, intensity and fundamental frequency. The results point to the presence of stress on the third syllable of heptasyllabic words, manifested in terms of longer onset consonant duration. This supports traditional descriptions of Polish as a bidirectional stress system.