ABSTRACT
Facilitative effects of bilingualism on general aspects of third language (L3) proficiency have been demonstrated in numerous studies conducted in bilingual communities and classrooms around the world. When it comes to L3 phonology, however, empirical evidence has been scarce and inconclusive in respect to the question of whether and/or how knowledge of two (or more) phonologies enhances the acquisition of an additional one. Adopting the Focus on Multilingualism [Cenoz, J. (2013). The influence of bilingualism on third language acquisition: Focus on multilingualism. Language Teaching, 46(1), 71–86] approach, the present study investigates the L3 phonological development in 19 German learners of Spanish with prior knowledge of English. Five of the young L3 learners were also speakers of Croatian, Israeli Hebrew, Italian and Polish, Russian, or Spanish as a heritage language. All the learners were tested on their developing ability to produce the Spanish /r/ and /ɾ/ segments in the course of three years of formal target language instruction. The results suggest a positive effect of specific, rather than broad-based, bilingual experience on L3 speech learning, in particular the production of universally difficult phonemes. Evidence for a degree of a ‘connected growth' of L2 and L3 phonology further illuminates the results.
Acknowledgements
The author wishes to express sincere gratitude to all the young multilinguals as well as their Spanish teacher for their unfailing and enthusiastic involvement in the longitudinal research study. Many thanks are also due to the editors and the three anonymous reviewers for their helpful comments and valuable suggestions.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1. It is acknowledged that additionally recording the active bilinguals' rhotics in their other home languages would have been desirable to verify the speakers' productions in those languages, and indeed to capture their whole linguistic repertoire. Due to practical considerations and constraints related to data collection in the school setting, this was not possible in the present study. Future studies in this area should nevertheless attempt at collecting phonological data in all the languages of their multilingual participants to be able to draw more valid conclusions about potential facilitation of prior phonological experience on L3 perception and production (see Cabrelli Amaro, Citation2013 for a detailed discussion on methodological challenges of conducting research in L3 phonology).
2. Analyses of five tokens of the English rhotic sound productions were performed due to the number of elicited target tokens during the English interview, in which all the participants produced at least five (and a maximum of eight) renditions of the rhotic segment in a prevocalic position.
3. The Spanish thrill is known to be very sensitive to alterations in its articulatory and aerodynamic configurations. As Martínez-Celdrán (Citation1997) points out, the trill depends on the Bernoulli effect to sustain vibratory movement, in contrast to the voluntary movement associated with the production of the tap. The production of the Spanish trill can thus be expected to display a degree of variability in both native and non-native speakers (see e.g. Johnson, Citation2008).