Foreign or Native-like? The Attitudes of Czech EFL Learners Towards Accents of English and Their Use as Pronunciation Models

Foreign or Native-like? The Attitudes of Czech EFL Learners Towards Accents of English and Their Use as Pronunciation Models Kateřina Brabcová, Radek Skarnitzl ABSTRACT: Attitudes of language users to English in the international context certainly do not rank among newly studied subjects. One of the frequent caveats of the ongoing research, however, is that it mostly targets university students of English, which may provide avery skewed perspective. This study focuses on young Czech speakers of English who have studied or are studying other disciplines and uses an online survey to examine their attitudes to English pronunciation in general and to their own pronunciation, to various accents of English and also to exposure to model accents. Analyses of 145 respondents show that 70% of them would like to acquire anative-like accent (most frequently General British), even though most of them use English with other foreigners (and not native speakers). They prefer to be exposed to many different accents of English, native and non-native, and believe that learners should themselves choose which accent they want to regard as amodel. One of the most important findings concerns the participants’ belief that pronunciation is teachable and that it is worth working on it; this should provide encouragement to teachers of English. In general, our results suggest that the ELF approach and the associated Lingua Franca Core concept do not seem to be relevant for young Czech users of English. KEY WORDS: English, English as aLingua Franca, foreign accent, language attitudes, pronunciation 1. INTRODUCTION There are approximately 1.5 billion speakers of English in the world, out of which only 400 million can claim it to be their mother tongue. The remaining 1.1 billion use English either as their second language or as alanguage that does not hold an official status in their country and is thus considered foreign (Crystal, 2002, p.10). Kachru (1982) uses amodel of three concentric circles to illustrate this distinction. The socalled inner circle contains all speakers who use English as their L1, the outer circle refers to users for whom it is their second language (ESL, alanguage that has been institutionalised in the speakers’ country of origin), and the expanding circle is reserved for learners of English as aforeign language (EFL). The rather obvious outcome is that about 80% of conversations happening in English worldwide (Timmis, 2002, p.240) are carried out not by its native speakers, but by speakers who do not share acommon L1 and use English mainly as acontact language. English has thus become an international language (EIL), and some scientists even go as far as saying that we can no longer speak of one English only, but should instead use the term World Englishes (Kuo, 2006, p.213). While EIL refers to communication between both native and nonOPEN ACCESS KATEřINA BRABCOVá, RADEK SKARNITZL 39 native speakers, there is another term— English as alingua franca (ELF)— which is used, in its narrower sense, to represent exchanges happening only between nonnatives (Seidlhofer, 2005). The fact that English is now used by speakers of various origins and of various mother tongues also means that we can encounter many different non-native, or foreign, accents. To define aforeign accent more precisely, it can be understood as any structured deviation of an L2 speaker’s pronunciation from astandard variety of the spoken language (in our context either General British, GB— previously referred to as Received Pronunciation, RP— or General American, GA; Cruttenden, 2014) (Derwing, 2013, p.2). Since the worldwide status of English in the new century has greatly changed and the dominant concept has become intelligibility in international communication, many researchers have called for new models of and approaches to pronunciation teaching (see Jenkins, 1998, 2000, 2006, 2009; Seidlhofer, 2005). There have been several suggestions how to change the model of pronunciation used at schools and to allow the speakers to be understood and to still be able to express their own identity while speaking English. With the thought in mind that intelligibility should be the main aim, Gimson (1978) came up with the Rudimentary International Pronunciation, which reduced the inventory of English sounds from 24 consonants and 20 vowels to 14 and 15 respectively (Gimson, 1978 cited in Jenkins, 1998). Its problem was that it disregarded the voicing distinction of consonants and was, as aresult, not intelligible after all. Another proposal, known as An International Approach, came from Pennington (1996). It gave the learner complete freedom to follow amodel for pronunciation, native or non-native, of their own choice (Pennington, 1996 cited in Jenkins 1998). This approach, however, could not guarantee mutual intelligibility either. Other suggestions followed such as Cruttenden’s model of Amalgam English or his International English (Cruttenden, 1998 cited in Szpyra-Kozłowska, 2015) or Collins and Mees’s (2003) method of Error Ranking which hierarchizes the importance of phonological features, giving them preference based on how much they contribute to intelligibility breakdown (Collins & Mees, 2003 cited in Szpyra-Kozłowska, 2015). But the one proposal that has become especially popular is the concept of ELF and, specifically, Jenkins’s Lingua Franca Core (LFC) (Jenkins, 1998, 2000). Jenkins (1998, 2000) suggests that we should distinguish core and non-core phonological features. Core aspects of pronunciation are those that are likely to cause acommunication breakdown and should thus be taught to all learners according to the standard pronunciation of English. On the other hand, non-core features should not lead to any such problems, and learners should therefore be allowed to use their own non-native variants. As for segmental features, Jenkins says that most of the consonants should be taught, as well as the distinction between long and short vowels. She also believes that initial consonant clusters should be respected and that the rhotic variety of the language is better understood in international communication. On the other hand, vowel quality would be considered anon-core feature together with the th sounds and the use of the dark [ɫ]. Other core features include nuclear stress (sometimes also called sentence stress) and articulatory setting that would allow the learner to produce core sounds correctly. What should not, however, be OPEN ACCESS 40 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2018 taken into account is word stress, features of connected speech (elision, linking, weak forms), and rhythm. At first glance, Jenkins’s approach sounds reasonable. Its goal is to make learners’ speech intelligible regardless of where their conversational partner happens to be from. However, her LFC has been criticized from various perspectives. First, the classification of core features does not seem to be based on empirical findings: for instance, the idea that lexical stress is unimportant for English seems difficult to accept, even in international communication (see Field, 2005 or McCrocklin, 2012). Christiansen’s (2014) study even illustrates that the alleged “core” features correlated with intelligibility the least. Second, LFC completely overlooks other important features of interpersonal communication, such as style, aesthetic concerns, and social functions (Kuo, 2006, p.215). It is the social aspect of spoken communication which we regard as crucial: it has been repeatedly shown that people judge others based on the way they speak, whether in their native or in aforeign language. Foreign-accented speech is consistently associated with negative stereotypes and biases, sometimes even with discrimination (see Dovidio et al., 2010 for areview). To mention but few recent studies, Timming (2017) reports discrimination of applicants during job interviews in the USA due to their non-native accent (cf. the study of Munro, 2003 for the Canadian context). Espinosa (2017) studied the Spanish-accented English of Ana Botella during her speech to the International Olympic Committee, which, although intelligible and not diverging much from the LFC, received severe criticism in Spain and was considered an example of bad English. Dragojevic et al. (2017) further show that stereotyping is not the only factor that leads to negative evaluations of non-native speakers. In their study, they proved that the participants listening to recordings of heavy and mild foreign-accented speakers judged the former group more negatively not because of prototypicality effects, but because their speech was more difficult to process. LevAri and Keysar (2010) investigated how native listeners evaluate statements produced by native and non-native speakers in terms of credibility. Their results show that the foreign-accented speech was perceived as less truthful. Their experiment was replicated by Hanzlíková and Skarnitzl (2017) with non-native listeners and again they demonstrated anegative bias against non-native speakers due to their accent, even in what approximates the ELF context. Foreign accent is, therefore, clearly astrong factor which influences the impression aspeaker makes on listeners. Jenkins (2000) herself is well aware of this fact and therefore argues that we need to present learners of English with many different varieties of accented English and to teach them some tolerance. She claims that, as aconsequence, they would not choose the native-speaker model of pronunciation anymore because they would realise that there are other, more achievable options. However, this still remains only ahypothetical situation because the socio-psychological character of spoken communication, including stereotyping and biases, cannot be easily turned off or abstracted from. Alarge amount of research is needed to find whether learners’ attitudes to English and its model accents are likely to change. In the present study, we have thus decided to find what the learners’ current attitudes towards foreign-accented English are, what pronunciation goals they set for OPEN ACCESS KATEřINA BRABCOVá, RADEK SKARNITZL 41 themselves, and what they believe pronunciation teaching should look like. This is certainly not anew research topic, and anumber of studies have addressed such questions already. Ladegaard and Sachdev (2006), for example, investigated the preferences of Danish EFL learners to a variety of native English accents. Although the learners indicated that they preferred American culture, they strongly favoured the British RP accent as amodel for pronunciation. Szpyra-Kozłowska (2004, cited in Szpyra-Kozłowska, 2015, pp.49–55) wanted to know whether, if given the choice, learners would opt for astandard English accent or for that which is promoted by the ELF movement. Out of the 134 Polish pupils aged 16–17 who participated in her questionnaire survey, 40.2% opted for RP, 32.8% for GA and only 13.4% for ELF despite the fact that prior to the questionnaire, there was a15-minute talk on English accents explaining the concept of ELF. Among the reasons why they chose RP, the students stated that it was the variety used by many native and non-native speakers, it ensured international intelligibility, and they wanted to be perceived as educated and cultured. Nowacka (2012) conducted asimilar survey asking international university students from Italy, Spanish and Poland and found out that 98% believed good English pronunciation was important and that 89% of the students stated that people should aim for anative-like accent. Similarly, the learners expressed their belief that such avariety signalled they were competent users of the language. In Sa’d and Modirkhamene’s (2015) study, 213 Iranian male EFL learners were asked whether they would choose the accent of the target language or the foreign-accented variant and again, 81.3% selected the first option and only 18.7% the second. They said that they perceived the native-like accent as beautiful and effective and that its acquisition would prevent possible mockery. Waniek-Klimczak, Porzuczek and Rojczyk (2015) surveyed over 500 university students of English in Poland and found that there was astrong preference for anative model. Lastly, Timmis (2002) carried out aquestionnaire study addressing both students and teachers from 14 and 45 different countries, respectively. While 67% students preferred the native-like accent as their target for pronunciation, the teachers expressed adifferent opinion. 27% favoured the pronunciation of anative speaker, 39% would encourage their students to preserve their own accent as long as their speech can be understood, and the remaining 34% indicated no preference. The teachers who selected the third option did so because they believed the decision was context-dependent and also because they wanted to let their students choose for themselves. Many of the studies mentioned above thus support anative model of pronunciation, which was reflected in the compromise position of NELF, Native English as aLingua Franca. As Szpyra-Kozłowska (2015) explains, NELF is intended for learners who want to learn English to communicate with both native and non-native speakers of this language, and amodel of anative speaker should be used because “native English serves auseful communicative function” (p.24). The aim of our research is to follow up on the previous studies and to determine the attitude of Czech learners of EFL towards their own accent and to model accents. There is an already existing survey conducted by Jakšič and Šturm (2017) which also investigates the attitude of Czech students towards accents in English. Unlike our study, Jakšič and Šturm analysed high-school students’ perception of and opinions on the RP versus the GA accent. Our study, however, targets not only other native OPEN ACCESS 42 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2018 accents, but also non-native accents. It is thus similar to the research conducted by Güttnerová (2016) as part of her diploma thesis. The respondents’ answers in her questionnaire suggest that their point of reference was the pronunciation of anative speaker and that they believed they were taken more seriously when speaking with anative-like accent. One of the drawbacks to many of the cited studies is their partial or exclusive reliance on university students of English; such students are likely to be highly motivated when it comes to learning English pronunciation, and their views may reflect that. In the present study, our aim was to examine attitudes of exactly those respondents who do not, did not, or were not planning to study English as auniversity major. The research questions we would like to answer are: 1. What are the attitudes of Czech learners of English towards their own foreign accent? 2. Which accent do they believe learners of English should acquire in general? 3. How important do they think pronunciation teaching is? 4. Which varieties of English should be, according to the respondents, presented to students in lessons? 2. METHOD 2.1 DATA COLLECTION In order to address the research questions, aquestionnaire was created with the use of Google Forms (https://www.google.com/forms). It was divided into two parts; the first asked the respondents to provide some basic demographic information about themselves (namely their nationality, mother tongue, sex, age, subject of study, proficiency in English, how long they have stayed/lived abroad, and what speakers they commonly talk to in English). The second part of the survey consisted of 11 items, as shown below. In item 1, respondents had to self-identify the degree of foreign accent in their English. Item 4 was atwo-part question regarding their wish to acquire aspecific accent of English. The remaining 9 items were formulated as either personal or general statements and the participants’ task was to express the degree of their agreement with them on a5-point Likert scale (1 = strongly disagree, 5 = strongly agree). The questions were given in Czech, so as to allow even less advanced users of English to participate, and they appeared in the following order: 1. My accent in English is: 1 = strongly non-native-like, 5 = native-like 2. Iam satisfied with my accent in English. 3. Iwould like to acquire anative-like accent in English. 4. Is there aspecific accent in English you would like to acquire? Which one and why? 5. It is important for me to speak with aforeign accent to indicate Iam non-native. OPEN ACCESS KATEřINA BRABCOVá, RADEK SKARNITZL 43 6. Students of English should be exposed to only one variety of native English in lessons. 7. Ithink that students themselves should choose which accent they want to use in English. 8. Ithink it is generally important for the students to acquire anative-like accent in English. 9. Students of English should be mainly exposed to varieties of non-native English in lessons. 10. Ithink that trying to learn the pronunciation of English is awaste of time and energy. 11. Ithink that teachers of English should present both the accent of native and of non-native speakers in lessons. The order of the questions was designed to make the respondents take different perspectives and to prevent them from answering aquestion immediately following asimilar one with opposite polarity. The snowball technique was used for the distribution of the questionnaire. Friends and acquaintances of the researchers received alink to the questionnaire via e-mail. They were asked to send it further to their own friends or colleagues. The link was not freely accessible on social media or any other websites. 2.2 PARTICIPANTS In total, 168 people responded and filled out the questionnaire. Twenty-three completed questionnaires had to be discarded as the respondents were either not of Czech origin or answered that their main subject of study was English. We have thus analysed the responses of the remaining 145 participants. The majority (141) of the participants was aged 15–30, only 4 were between 30 and 40 years of age. Eighty-eight respondents were female, fifty-seven male. One hundred and eight of them (had) studied non-linguistics subjects (or indicated giving preference to these at ahighschool level already), the remaining 37 focused on the study of languages other than English. As for their self-reported linguistic proficiency, most of the participants were at the B1, B2, or C1 level (18, 73 and 41, respectively), only one was at the A1 level, four at A2 and eight at C2 (CEFR, 2001). The responses were extracted into atable, and subsequent analyses were performed and the results visualised using R (R Core Team, 2017) and the R package ggplot2 (Wickham, 2009). 3. RESULTS AND DISCUSSION First, we were interested in finding who our respondents’ communication partners were when they spoke English. The respondents could tick any of the three groups of interlocutors shown in Figure 1 (other Czechs, speakers of other L1s, native speakers of English). As shown in the figure, our participants are most likely to use their OPEN ACCESS 44 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2018 English when communicating with other non-native speakers. This result is important, as it confirms the relevance of our respondent sample when addressing English as aLingua Franca. Only relatively few of the participants communicate with native speakers of English. Of the 145 respondents, 94 reported communicating only with one type of interlocutor, 40 with two groups. Seven respondents reported communicating with Czechs, other non-natives, as well as native speakers of English, while four respondents do not use English at all. Figure 1:Responses indicating the participants’ most frequent communication partners in English, broken down according to their field of study. Next, we wanted to see how responses to questions 1–3 (see section 2.1) correlated with each other. Since the data consists of discrete rather than continuous values (scores between 1 and 5), we used the non-parametric Spearman’s rank correlation coefficient rho (ρ). The results in Table 1 indicate amedium-strength negative correlation (ρ = –0.43) between the respondents’ self-reported general level in English (CEFR level) and the self-reported strength of their accent in English: in other words, the higher their general English level, the more native-like their (self-assessed) pronunciation. Perhaps not surprisingly, the strongest discovered correlation (ρ = –0.52) concerns the relationship between accent strength and the degree of respondents’ satisfaction with their accent, meaning the more non-native-like the participants judged their accent, the less satisfied they were with it. Participants with astronger accent also tended to express awish to acquire anative-like accent in English more than those who reported having aless accented English (ρ = 0.42). The other relationships manifested only very weak and insignificant correlations. Our data indicate no link between, for instance, the degree to which respondents are satisfied with their accent on the one hand and the strength of their wish to acquire anative accent on the other; such awish seems to be amatter of personal preference and experience rather than dissatisfaction with one’s pronunciation. linguistic other linguistic other linguistic other 0 25 50 75 100 responses (%) Respondents frequently speak English with ... other Czechs non-native speakers native speakers yes yes yes yes yes yes no no no no no no OPEN ACCESS KATEřINA BRABCOVá, RADEK SKARNITZL 45 CEFR level accent strength accent satisfaction accent strength –0.43 *** accent satisfaction 0.14 * –0.52 *** wish for native accent 0.12 0.42 *** 0.11 Table 1.Correlation matrix with Spearman correlation coefficients ρ pertaining to selected comparisons. *** p < 0.0001; * p < 0.1 Let us examine the participants’ wish to acquire anative accent in more detail. Figure 2a shows that over 70% of the respondents would like to acquire anative accent (question 3 in the questionnaire), and only 15% explicitly state that they do not want to speak in anative accent. It is interesting to point out that one half of those who would like to have anative accent did not identify any specific accent, while the other half did have one in mind. The preferences for this latter group are shown in Figure 2b; since this was an open-ended question (question 4), the responses were grouped (for instance, “British” and “RP” were taken to mean the same thing). We can see that there is astrong preference for British English (GB), especially among female respondents. It is also noteworthy that it was mostly males who expressed awish to learn one of the Celtic accents (Scottish or Irish English) or the Australian accent. Figure 2.Responses to question 3, broken down according to the participants’ sex and field of study: a. agreement with the statement expressing awish to acquire anative accent; b. specific accents listed by participants who wished to acquire anative accent (see text). The following analyses are related to model accents, and the role of pronunciation and exposure to accents in the curriculum. Figure 3 provides results for questions 10, 5, 7, and 8 (the order in which they appear in the figure). In answer to question 10, 90% of the respondents expressed disagreement (i.e., responses 1 or 2) with the statement reading Ithink that trying to learn the pronunciation of English is awaste of time FM 0 25 50 75 100 responses (%) a. FM b. GB linguistic other FMFM linguistic other Aus GA Celtic strongly disagree strongly agree OPEN ACCESS 46 STUDIE ZAPLIKOVANÉ LINGVISTIKY 1/2018 and energy; we may regard this as afavourable and motivating message for teachers of English in general and specifically those teaching English pronunciation. The exact wording of the question is important here: in other words, the respondents— both linguistically oriented and not— believe that it is worth trying to work on one’s pronunciation in English. The answers to question 5 (It is important for me to speak with aforeign accent to indicate Iam non-native.) show that it does not seem to be important for most of our participants to manifest their identity through their foreign accent in English: overall, 73% of the participants disagreed with the statement. The responses to the latter two statements are similar. The majority of the participants agree that speakers should choose which accent they want to regard as amodel one. At the same time, more of them believe that it is generally important to acquire anative- -like accent in English. Figure 3.Responses to questions 10, 5, 7, and 8 (the shortened version of the questions is given above each chart), broken down according to the participants’ field of study (see text). 0 25 50 75 100 responses (%) pronunciation is a waste of time strongly disagree strongly agree important to speak in my own accent linguistic other linguistic other 0 25 50 75 100 students choose whichever accent important to acquire a native accent OPEN ACCESS