Connected Speech Processes in Spontaneous Discourselinking /r/, T-tapping, and Yod Coalescence in Non-Rhotic English

Laburpena

The present dissertation aims to describe the correlation between the three connected speech phenomena mentioned earlier and several factors: phonetic (e.g., prosodic environment), linguistic (e.g., grammatical status and word-class collocations), sociolinguistic (e.g., gender), and usage-based factors (e.g., lexical frequency and collocational frequency) in spontaneous speech in the non-rhotic variety of English known as Received Pronunciation (RP). The research conducted in the current study addresses the following general research questions (RQ) to explore the status of linking /r/, t-tapping, and yod coalescence in contemporary RP and what factors motivate their production in connected speech: RQ1. How frequent is the use of linking /r/, t-tapping, and yod coalescence in conversational language use?; RQ2. What factors condition variation in the production of linking /r/, t-tapping, and yod coalescence? The study focuses on the speech of six UK political figures, all of them RP speakers, interviewed at The Andrew Marr Show, a talk show broadcast on BBC One from the mid-2000s to the early 2020s. Therefore, the intended contribution of this study is, empirical, involving a central methodological corpus-based procedure. In addition to using a corpus-based method, a usage-based approach makes it possible to model a correlation between the linguistic phenomena under study as well as the variation across the distinct factors (Bybee, 1999, 2001). In doing so, the potential contexts of linking /r/, t-tapping, and yod coalescence have been identified and analysed for the presence or absence of rhoticity, tap production, and coalescence assimilation in a small-scale corpus of interview speech. The results show that the three phenomena exhibit significant variability. The key findings indicate that linking /r/ production is notably frequent, although not categorical, which is consistent with its hiatus-breaking function alongside the use of laryngeal strategies to avoid hiatus when linking /r/ is not applied, as pointed out by previous studies (e.g., Cox, Palethorpe, Buckley & Bentink, 2014; Mompeán, 2022). Even though the findings suggest that t-tapping and yod coalescence are not very frequent connected speech phenomena in spontaneous speech in RP, the present study reveals that the processes can be recent developments in RP and that they are not exclusive of other varieties of English or non-standard variants (e.g., Wells, 1982). One interesting conclusion from the present thesis is that from the analysis of the factors that condition linking /r/, t-tapping, and yod coalescence usage is that multiple variables contribute to their variation, yet the weight of each of the factors is different across the phenomena. A remarkable finding that makes the current study different from previous ones (mostly corpus-based) is the fact that the results represent relevant and novel insights concerning empirical research and contribute to the necessity for an updated description of RP and the operationalisation of connected speech phenomena in this accentual variety (e.g., Cruttenden, 2014).