Phonetic convergence in the speech of Polish learners of English

Abstract

This dissertation examines variability in the phonetic performance of L2 users of English and concentrates on speech convergence as a result of exposure to native and non-native pronunciation. The term speech convergence refers to a process during which speakers adapt their linguistic behaviour according to who they are talking or listening to. Previous studies show that the phenomenon may take place both in a speaker’s L1 (e.g. Giles, 1973; Coupland, 1984; Gregory and Webster, 1996; Pardo, 2006; Babel; 2010) and L2 (e.g. Beebe, 1977; Berkowitz, 1986; Lewandowski, 2012; Rojczyk, 2013; Trofimovich and Kennedy, 2014). Speech convergence can be subdivided into three types of linguistic behaviour: convergence (the process of making one’s speech more similar to that of another person), divergence (the process of moving away from the speech of another person) and maintenance (the process of maintaining one’s default linguistic behaviour in spite of exposure to the speech of another person). The dissertation consists of four chapters; the first two provide theoretical background, the next two describe the study and its findings. Chapter One is concerned with previous research on speech convergence. The chapter reviews the methodology and approaches used in previous work and discusses the range of factors that may affect convergence strategies. Chapter Two provides an overview of relevant studies in the field of L2 phonetics. It describes the structure and formation of the L2 sound system and the numerous socialpsychological, linguistic and psycholinguistic variables that may influence L2 phonetic performance. Chapter Three describes the study on speech convergence in the pronunciation of Polish learners of English, i.e. the aims, hypotheses, methodology and results. In Chapter Four, the results of the study on phonetic convergence in the speech of Polish learners of English are analysed and discussed. The phenomenon of speech convergence has been explored under different names and with the use of various frameworks and methodological procedures. Some researchers refer to the process as accommodation and investigate it by analysing spontaneous conversational data (e.g. Giles, 1973; Bourhis and Giles, 1977; Coupland, 1984; Gregory and Webster, 1996). Other researches use the term imitation and examine the phenomenon in socially minimal, laboratory-based settings (e.g. Goldinger, 1998; Schokley et al., 2004; Delvaux and Soquet, 2007; Nielsen, 2011). Irrespective of terminological and methodological differences, the results of previous studies on phonetic convergence indicate that the process is conditioned by 171 a variety of linguistic (e.g. Mitterer and Ernestus, 2008; Babel, 2009; Brouwer et al., 2010; Nielsen, 2011) and social-psychological factors (Giles, 1973; Bilous i Krauss, 1988; Gregory and Webster, 1996; Pardo, 2006; Babel, 2009, Yu et al., 2013) Research on L2 acquisition and non-native pronunciation shows that the development of the L2 sound system is a complex and dynamic process. It has been argued that the productions of L2 users are generated by interlanguage (IL), an independent linguistic system that encompasses elements of the learner’s L1 and L2 but does not correspond exactly to either the NL or the TL (e.g. Selinker, 1972; 1992). Importantly, previous findings indicate that the phonetic performance of non-native speakers is influenced not only by their L1 and L2 sound systems but also by a range of various psycholinguistic (e.g. Flege, 1987; Flege et al., 2003) and social-psychological factors (e.g. Taylor et al., 1971; Zuengler, 1982; Gatbonton et al., 2011). The process of adapting one’s pronunciation as a result of exposure to another person’s speech has been detected in the productions of L2 users (e.g. Beebe, 1977; Berkowitz, 1986; Lewandowski, 2012; Rojczyk, 2013; Trofimovich and Kennedy, 2014). Similarly as in the case of L1 speech convergence, previous studies show that the magnitude of L2 speech convergence may depend upon a variety of social-psychological and linguistic variables. An interesting aspect of L2 phonetic convergence that has not yet been thoroughly explored is the comparison of pronunciation shifts upon exposure to the speech of native speakers of the TL as compared with pronunciation shifts upon exposure to the speech of other learners. The aim of the study was to address this issue by investigating and comparing L2 convergence strategies upon exposure to native and non-native pronunciation. The study concentrated on the phonetic performance of advanced Polish learners of English, who were exposed to two pronunciation varieties: Polish-accented English and native English. The participants were 38 native speakers of Polish, majoring in English Studies and recruited from the University of Lodz. The subjects listened to pre-recorded productions provided by two model talkers/interlocutors: a native speaker of Standard Southern British English and a native speaker of Polish (a qualified phonetician imitating a heavy Polish accent in English). The phonetic variables under investigation were the following: aspiration in word-initial /p t k/, pre-voicing in word-initial /b d g/, vowel duration as a cue for consonant voicing in English /æ e ɪ iː/. The experimental procedure consisted of several phases. First, the informants were instructed to identify the target words in an auditory naming task (baseline condition). Next, they were asked to listen to pre-recorded English words provided by the two 172 model talkers/interlocutors and to identify the words by saying them out loud (imitation condition). Finally, the subjects were required to read the target words for the two model talkers/interlocutors to listen to at a later time (accommodation condition). Following the production stage of the experiment, the participants completed a questionnaire whose purpose was to gauge attitudes towards native and foreign-accented English. Three hypotheses were formulated to be tested in the course of the study. Hypothesis 1 predicted that convergence strategies following exposure to native and non-native English will vary as a function of model talker/interlocutor. Hypothesis 2 predicted that convergence strategies following exposure to native and non-native English will be affected by the subjects’ attitudes towards native and Polish-accented English. Hypothesis 3 predicted that convergence strategies following exposure to native and non-native English will differ as a function of phonetic context (place of articulation and vowel category). Acoustic and statistical analysis of the data revealed that the subjects modified their linguistic behaviour following exposure to the speech of the model talkers/interlocutors, which corroborates the claim that L2 speech convergence phenomena are present in nonnative pronunciation. Hypothesis 1 was partially supported by the results of the study. It was found that speech behaviour following exposure to native and non-native English varied as a function of model talker/interlocutor in all but two instances (accommodation on pre-voicing and imitation of vowel duration). The results suggests that when using a second language, speakers may use different convergence strategies depending on the native/non-native status of the model talker or interlocutor. Hypothesis 2 was partially supported by the data. The results indicate that a strong preference for target-like pronunciation may prompt learners to converge towards native speech and diverge from foreign-accented speech. However, the factor does not seem to operate if a learner has not succeeded in mastering a given TL pronunciation feature, i.e. the impact of attitudinal factors on the magnitude of convergence in non-native pronunciation appears to be conditioned by the stage of acquisition of a given TL phonetic feature. Hypothesis 3 was not borne out the results obtained in the study. It was found that convergence strategies following exposure to native and non-native English did not vary depending on phonetic context. Overall, the findings of the study provide support for the claim that the process of speech convergence operates in L2 pronunciation and imply that certain social-psychological and psycholinguistic factors may have an impact on learners’ convergence strategies.