Assessing cross-linguistic perception of L2 Portuguese vowels

Description

Two of the key questions that guide L2 speech learning research are the perceptual relationship between the L1 and the L2 sound inventories (‘the mapping issue’ [1]) and how this relationship may impact learning when acquiring a non-native phonological system (‘the learning difficulty/ease issue’ [1]). Theoretical models of L2 speech acquisition [e.g. 2-5] hypothesize that the degree of perceived phonetic similarity between the L1 and L2 phonological systems predicts the relative difficulty/ease in L2 speech learning. In this study, we address these issues by examining the perceived cross-linguistic (CL) similarity between L2 European Portuguese and L1 Canadian English (CE)/Southern British English (SBE) sounds to predict the learning of the Portuguese oral vowels /i/, /e/, /ɛ/, /ɐ/, /a/, /ɔ/, /o/, /u/ at the onset of L2 acquisition. 26 L1 CE-speaking participants (13F; Mn age: 32.2 (SD=6.4)), and 29 L1 SBE speakers (15F, 2Nonbinary; Mn age: 32.2 (SD=7.2)), who were naïve to European Portuguese (i.e., who have never been exposed to or had any experience with the target language), were asked to complete a crosslanguage perceptual assimilation task (PAT) and a language background questionnaire (adapted from LSBQ [6]). Two tasks were designed, one for the Canadian English group and another for the British English cohort. In each
of the PATs, participants listened to pseudowords ‘CVC([ɨ]) (C=consonant, V=vowel) with both the target L2 European Portuguese vowels (for example: fite, site) and either the 11 SBE vowels /i/, /ɪ/, /ɛ/, /æ/, /ʌ/, /ɜ/, /ɑ/, /ɒ/, /ɔ/, /ʊ/, /u/, or the 12 CE vowels /i/, /ɪ/, /eɪ/, /ɛ/, /æ/, /ʌ/, /ɜ/, /oʊ/, /ʊ/, /u/, /ɑ/, /ɒ/ (for example, sit, fit). Participants were first asked to map both EP and CE or SBE phones to English vowels using a set of 11 or 12 response labels, respectively, that consisted of keywords with the orthographic form <hVd>,
representing each of the English vowels. Immediately after, participants listened to the same pseudowords again and were asked to rate category goodness-of-fit (GoF) of each vowel token in a 7-point Likert scale ranging from 1 (poor example) to 7 (very good example). The experiment took approximately 30 minutes. We used the Gorilla Experiment Builder (www.gorilla.sc) to create and host our study [7]. All data was collected online, and participants were recruited through Prolific (www.prolific.com). For our preliminary analysis, we
calculated the mean percentage of assimilation of both L2 and L1 phones to L1 vowels. We considered phones to be categorized when consistently mapped to a specific L1 category above a 50% threshold, otherwise they were deemed uncategorized (e.g., [8], Table 1).
These results will be discussed within the PAM framework [1-2], including the predictions for L2 vowel ease/difficulty of learning for each L1 English variety, as well as the intergroup vowel assimilation and ranking differences.

Period	23 Oct 2024
Event title	Encontro Nacional da Associação Portuguesa de Linguística
Event type	Conference
Conference number	40
Location	Ponta DelgadaShow on map
Degree of Recognition	National