Abstract
Our aim is to describe the phonetic variants of /r/ in South Tyrol Italian (STI) and to link this variation to both linguistic (phonetic-phonological) and extra-linguistic (typologies of bilingual speaker) factors (Flege, 2007). South Tyrol is a bilingual alpine region situated in Italy (since 1919) that lies on the Germanic – (Italo) Romance dialectological border. Although this area has been described as a societal bilingualism with two quite separate linguistic communities, (Bavarian) German and Italian (Mioni, 1990), nowadays the actual degree of overlapping between the two speech communities as well as the number of bilingual speakers seems to be increasing In this sociolinguistic context, South Tyrol Italian shows a higher degree of /r/ variation if compared to the common standard-dialect Italo-Romance linguistic repertoire (a preliminary instrumental analysis of /r/- sounds in Italy is offered in Romano, in prep.). Standard Italian is usually reported to have two or three (not so neatly contextual) realizations of /r/: The alveolar trill [r], the geminate [rr], and the alveolar tap [], e.g. raro [rao] /raro/ ‘rare’, carro [karo] /karro/ ‘cart, wagon’ (Canepari, 1999). Although in South Tyrol there is not a lively presence of Italo-Romance dialects, the neighbouring Trentino-Veneto dialects exerted a strong influence on STI (due to strong immigration fluxes in the Twenties). Thus, typical Venetian /r/ variants, such as the post-alveolar (retroflex) flap [] and approximant [], are to be found even in STI. To complete the complex picture of /r/ phonetic sub-system, Bavarian German dialects back /r/-sounds (trill [], fricative [], and approximant []) are well attested in STI. (a characterisation of /r/-sounds in Germany is offered in Wiese, 2001). Our data are drawn from 10 speakers (3 males and 7 females) living and/or working in Bolzano (the political and administrative centre of South Tyrol). The sample is stratified according to three broad categories of bilingual speaker: “early bilingual” (roughly a speaker with two native languages acquired during primary socialization), “German native speaker” (a late bilingual that acquired Italian as a language of secondary socialization, e.g. at school or from the peer group) and “Italian native speaker” (a late bilingual that acquired German as a language of secondary socialization). The elicitation techniques included reading tasks (e.g. word and sentence lists), visual stimuli and map tasks, using a modified version of the CLIPS (Corpora and Lexicon of Spoken Italian) protocol. However the present analysis is based only on reading tasks. The elicitation list contains sets of words controlled for frequency and phonological relevant variables. In order to assess if the high degree of /r/ variation in STI is context-independent or due to allophonic variation (as reported -at least to some extent- for standard Italian), special attention was paid to the selection of words displaying different phonetic environments (especially consonant clusters), syllabic positions (e.g. in the onset: respingere ‘to push back’ or in the coda: forte ‘strong’) and stress patterns. The main goal of our analysis is to provide an instrumental aided description of the different spectral- acoustic properties of /r/-sounds, with special attention to the back variants (uvular trill, fricative and approximant). Then, we seek for contextual distribution of the variants (against a free variation hypothesis). On an extra-linguistic level we look for relationships between sets of variants (front vs. back) and the three categories of speakers, early bilinguals and late bilinguals (German and Italian native speakers). Our main goal is to look for a relation between the occurrence of back variants and late bilingualism (of German native speakers). Back variants are thus considered as markers of German nativeness.