wolf.dewitte@e-campus.uab.cat

Year of Study in Programme: First

Supervisor: Dr Juli Cebrian & Dr Núria Gavaldà (Universidad Internacional de la Rioja/UNIR)

Area/Topic of Research: Forensic Phonetics, L2 Speech

Title/Provisional Title of Thesis: Speaker Comparisons in Multilingual Contexts: Individual Variation in L1 and L2 Speech for Forensic Purposes

Short Abstract: The number of people who speak one or multiple languages in addition to their mother tongue is increasing significantly and so is, in turn, the number of forensic phonetic cases involving multilingual situations (Künzel, 2013; Rose, 2002). While the preferred method of analysis in these multilingual contexts appears to be that of automatic speaker recognition (Campbell et al., 2004; Kinnunen et al., 2006; Künzel, 2013; Luengo et al., 2008; van Leeuwen & Bouten, 2004), surveys have shown that the most commonly used method by forensic phoneticians in monolingual situations is the auditory-acoustic approach (Gold & French, 2011, 2019; Morrison et al., 2016). As such, the main aim of this thesis is to carry out an analysis of native (L1) Catalan speech and second language (L2) English speech in order to determine which features allow for the identification of speakers across languages and would, therefore, be suitable to be used for such ends in forensic phonetic contexts involving an auditory-acoustic approach. More specifically, this study wishes to explore how certain features which are highly similar across Catalan and English—such as nasal consonants—perform when analysed in multilingual contexts; that is, whether the analysis of these features can aid the identification of speakers when the speech samples under investigation consist of an L1 Catalan and a highly proficient L2 English sample produced by the same individual. It is hypothesised that a comparison of the features mentioned across languages and within speakers will result in a low amount of variability being observed, while a similar comparison across speakers will result in the detection of a high degree of variation.