Phonetic imitation in L2 speech: Immediate imitation of English consonant glottalization by speakers of Polish

ABSTRACTPhonetic imitation is a ubiquitous process in speech production. Speakers have a strong tendency to imitate their interlocutors both in a native and a non-native language. It is especially important in acquiring non-native speech, because it allows forming new sound categories. In the current study we investigated whether and to what extent Polish learners of English are able to imitate t-glottalization observed especially in British English. A total of 25 Polish learners of English imitated English models’ productions with t-glottalization that were subsequently compared to their default productions (pre-test) and post-exposure production (post-test). The results showed that the participants successfully imitated t-glottalization after the exposure to the model talker. The generalization effect was limited in its magnitude in that only some of the non-imitated words had traces of glottalization. The results are discussed in terms of the differences in the implementation of glottalization in Polish and English and of how phonetic imitation informs second-language speech acquisition. Disclosure statementNo potential conflict of interest was reported by the authors.Data availability statementThe authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials https://osf.io/p3zes/.Notes1 See Docherty & Foulkes (Citation1999) and Ashby & Przedlacka (Citation2014) for illustrations of glottal events in non-SSBE varieties of English. For glottalization in American English, see, for instance, Seyfarth & Garellek (Citation2020) or Kaźmierski (Citation2020).2 Interestingly, when we fit the data with a numeric dependent variable (0 or 1) using a standard lmer model with the same effect structure as before, the predicted means align with the observed means almost perfectly. P-values are generally higher now, and the significance of pairwise comparisons changes somewhat. There is now no significant difference between positions in T1 (p = 0.081), and the effects of task are altered for the comparison T1/T3 (in intervocalic position, the difference is no longer significant with p = 0.128, while in non-prevocalic position, there is now a significant difference with p = 0.006).Additional informationFundingResearch supported by the National Science Centre Poland grant Phonetic imitation in a native and non-native language (UMO-2019/35/B/HS2/02767) to the first author and by the funds granted under Research Excellence Initiative of the University of Silesia in Katowice.