Chapter 6. Spanish rhotic variation and development in uninstructed immersion

Abstract

Perhaps due to the salient differences between English and Spanish rhotics, there has been a robust discussion regarding the acquisition of Spanish taps and trills by first language (L1) English speakers. Previous studies that have explored rhotic development have suggested that while there is a significant increase in accuracy of producing taps and trills as Spanish proficiency level increases (Face 2006; Olsen 2012), even advanced learners or learners with intense exposure to the language still face difficulty in producing the trill (Major 1986; Reeder 1998). Despite the contributions of previous studies, it remains unclear how learners develop the tap and trill in real time (as opposed to apparent time), to what extent the two rhotic sounds develop differently, and the role of additional linguistic and extralinguistic factors apart from phonological context and exposure. This study contributes to the discussion of Spanish rhotic development by tracking uninstructed L1 English learners in an Ecuadorian immersion setting for one year, providing an analysis of development in real time. By applying variationist methodology using the mixed-effects model Rbrul (Johnson 2009), we explore the role of phonological context, lexical stress, word class, cognate status, target word syllables, proficiency, exposure, speech style, and individual differences in the development of native-like taps and trills. We find that phonological context, word class, exposure, and speech style significantly condition tap production, while there are no significant predictors found for trill production. Furthermore, we see that target tap production accuracy grows throughout the year-long stay abroad, but target trill production shows no improvement.