Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

There is little doubt that the beliefs that learners hold about the process of foreign language learning and teaching are of paramount importance for the simple reason that, if we agree with Allwright and Bailey (1991) that language lessons or even entire courses are to a large extent co-constructed by teachers and their students, they are bound to affect what happens in the language classroom and also impinge on the effectiveness of the whole process of language instruction. In other words, somewhat contrary to widespread assumptions, it is not only the preferences manifested by teachers and the decisions that are informed by such preferences with respect to, for example, the content and sequencing of the syllabus, the choice of pedagogic activities or the overall approach to classroom proceedings, but also students’ experiences and expectations which shape their beliefs concerning different aspects of foreign language pedagogy that determine the nature of classroom interaction and the outcomes of what transpires in language lessons. Therefore, it should come as no surprise that learners’ beliefs are considered to be a crucial individual difference variable (e.g., Dörnyei, 2005; Ellis, 2008; Kalaja & Barcelos, 2003, 2012) and they have become the focus of empirical investigation. The studies conducted thus far have demonstrated, among other things, that beliefs can help us explain and predict learners’ behaviors in the classroom (Grotjahn, 1991), they are related to the use of strategies, motivation, the level of proficiency, foreign language anxiety and the adoption of an autonomous approach to language learning (cf. Kalaja & Barcelos, 2003; Mori, 1999; Yang, 1999), they can differ depending on the language studied, and can be both beneficial and detrimental (Ariogul et al., 2009), they can exert an influence on teachers’ classroom practices (e.g., Borg, 2003, 2008; Burgess & Etherington, 2002), they can get in the way of the process of learning if they are unrealistic (Sawir, 2002), and they can be modified to some extent through training (Mantley-Bromley, 1995). While many of these claims have been made about learners’ beliefs in general, with the relevant data being collected with the help of Horwitz’s (1987) Beliefs About Language Learning Inventory (BAALI), they apply in equal measure to the beliefs that learners display concerning different skills and language subsystems, and pronunciation is clearly no exception. In line with this assumption, the present paper reports the findings of a study which aimed to determine English majors’ beliefs about various facets of pronunciation instruction and to examine the relationship between such perceptions and success in mastering this foreign language subsystem, operationalized in terms of the scores on the oral part of the end-of-the-year examination in practical English and the pronunciation component of this score. In the first part, an attempt will be made to offer a brief overview of the available research dealing with learners’ and teachers’ beliefs in this area, which will be followed by the description of the research questions, the procedures for data collection and analysis, as well as the presentation and discussion of the findings. The paper will close with a handful of recommendations concerning the directions and methodology of future research on beliefs about pronunciation learning and teaching.

2 Previous Research on Beliefs About Pronunciation Learning and Teaching

As has been shown for the Polish context by, for example, Szpyra-Kozłowska (2008) and Pawlak (2010), there have been many foci of research into the teaching and learning of foreign language pronunciation, the most important of which include the status of pronunciation teaching at different levels of instruction (e.g., Baran-Łucarz, 2006), instructional targets, both when it comes to choosing the model to be taught (e.g., Sobkowiak, 2005) and the pronunciation features to be prioritized (e.g., Waniek-Klimczak, 2002), the main pronunciation-related difficulties experienced by different groups of learners (e.g., Nowacka, 2006), the effectiveness of different instructional techniques (e.g., Szpyra-Kozłowska & Stasiak, 2006), the utility of various teaching resources (e.g., Wrembel, 2005), the role of corrective feedback (e.g., Pawlak, 2004, 2013a), the use of evaluation procedures (e.g., Szpyra-Kozłowska et al., 2004), individual learner differences with respect to pronunciation instruction (Baran-Łucarz, 2012; Szyszka, 2011), learner autonomy in learning pronunciation features (e.g., Pawlak, 2006; Szyszka, 2006), or the use of pronunciation learning strategies (e.g., Pawlak, 2008). However, there has been surprisingly few empirical investigations that would have attempted to tap learners’ or teachers’ beliefs and preferences concerning pronunciation teaching, which must come as a surprise, given the importance of this field spelled out in the introduction to the present paper. In particular, somewhat in contrast to grammar teaching (e.g., Burgess & Etherington, 2002; Loewen et al., 2009; Pawlak, 2013b; Spada et al., 2009), very little emphasis has been placed, for example, on specific aspects of pronunciation instruction, such as the choice of the syllabus, the design of pronunciation-based lessons, the contribution of different categories of instructional techniques and procedures, or the role of various types of corrective feedback.

In fact, the few available studies in this area have pursued very disparate goals and focused, in the main, on teachers’ awareness of the importance of pronunciation and the ways in which it can successfully be taught, learners’ preferences concerning the pronunciation model to be taught and, to a much lesser degree, the use of particular instructional practices. Breitkreutz et al. (2001), for example, found that teachers of English in Canada were aware of the principles of pronunciation teaching, which found reflection in their classroom practices, and they believed that the role of suprasegmetal features for successful communication was greater than that of segmental features, but concluded that more training was needed in this area. A follow-up study conducted in the same educational context by Foote et al. (2011) produced similar findings and the researchers hypothesized that limited teacher awareness was the corollary of the fact that courses dealing with pronunciation instruction were not easily accessible during university education. In a recent development of this line of inquiry, Thompson (2012) investigated the beliefs and practices of 58 teachers of English from Canada and the United States, looking at such issues as the nature of a foreign accent (e.g., the impact of first language transfer), general instructional strategies (e.g., recording one’s voice to allow monitoring how sounds are produced), descriptions of English sounds (i.e., awareness of differences between particular segments) and specific teaching techniques (e.g., the role of reading aloud). The main finding was that the participants lack a coherent understanding of the issues included in the survey and many of them do not have the ability to critically evaluate their beliefs and practices, although, truth be told, the tool itself was flawed as the distinction between general instructional strategies and specific techniques does not seem to have been properly operationalized. An interesting study was also undertaken by Baker (2011), who explored the relationship between the cognitions (i.e., knowledge and beliefs) of five experienced teachers of English as a second language and their classroom practices with respect to pronunciation instruction, offering evidence that their knowledge, pedagogical choices and confidence were affected by the amount of training they had received, teaching experience and collaborative work with their colleagues. Finally, Wahid and Sulong (2013) found that the way in which teachers at a tertiary level approached pronunciation in their classroom was often at odds with current research findings, a result that was ascribed to lacking technical knowledge in pronunciation content and ignorance of the outcomes of empirical investigations in this area.

When it comes to learners’ beliefs about teaching and learning pronunciation, several studies of this kind have been conducted in the Polish context. Krzyżyński (1988), for example, reported that English majors were of the opinion that pronunciation was less important for attainment of proficiency in a foreign language than grammar or lexis, thus being critical of the importance accorded to accurate pronunciation in Polish schools. Very similar conclusions were reached by Sobkowiak (2002), since the university-level participants of his study also felt that pronunciation is less important than the other foreign language subsystems, many of them admitting, at the same time, that they practiced pronunciation regularly and wishing that they were given more opportunities for such practice by the institution they attended, views that are perhaps reflective of the fact that they were majoring in English as well as the requirements for successful completion of the program. Positive attitudes towards pronunciation were also reported by Waniek-Klimczak (1997) in a different group of English majors, but she also provided evidence for the tension between a desire to sound like a native speaker and the need to be fluent and confident when speaking the target language. Janicka et al. (2005), in turn, demonstrated that English majors display a marked preference for native models of English, British and American English being the varieties held in highest esteem. The same conclusion was also reached by Waniek-Klimczak, Rojczyk and Porzuczek (this volume) in their investigation of BA and MA students’ attitudes towards Polish-accented speech, or Polglish, to use the term proposed by Sobkowiak (1996). In line with the results of studies conducted elsewhere in Europe (e.g., Smojver and Stanojewic, 2013), they found that the participants opted for the native-like model and held negative views about the presence of the Polish accent, with the caveat that such opinions were a function of the educational level (BA or MA) and gender. As for studies conducted outside Poland, worth mentioning at this point is the research project undertaken by Simon and Taverniers (2011), who examined advanced Dutch learners’ beliefs about pronunciation teaching in comparison to grammar and vocabulary, demonstrating, among other things, that pronunciation and grammar were different from vocabulary in terms of bringing about communications breakdowns and involving the use of learning strategies, but similar when it comes to learners’ confidence in achieving native-like level proficiency and the utility of the instructional activities used in class.

3 The Study

3.1 Aims and Research Questions

In light of the fact that the available research on beliefs concerning pronunciation learning and teaching is rather scarce and limited in scope, the questionnaire study reported in the present paper aimed to contribute to this line of inquiry by investigating English majors’ perceptions of different aspects of pronunciation instruction and tapping the relationship between these perceptions and attainment, both in the case of speaking performance and with reference to this target language subsystem. More specifically, it sought to address the following research questions:

  1. 1.

    What are the participants’ beliefs concerning the overall value of pronunciation instruction, the choice of the syllabus, the design of classes devoted to pronunciation, the introduction of pronunciation features, the ways of practicing these features, and the role of error correction in this area?

  2. 2.

    What is the relationship between beliefs held in these areas and attainment on the final examination, both with respect to speaking in general and pronunciation in particular?

  3. 3.

    What are the reasons why the participants like or dislike learning pronunciation?

  4. 4.

    What instructional techniques are the most and the least favored by the participants?

3.2 Participants

The participants were 110 advanced Polish learners, 86 females and 24 males, in an institution of higher education, who were majoring in English and were enrolled in a 3-year BA program. Based on the information provided in the demographic section of the questionnaire (see Sect. 3.3 below), their average experience in learning English amounted to about 11 years, with the minimum of 5 and the maximum of 15 years. As is the norm in most programs of this kind, the participants had the opportunity to attend numerous English classes, with separate courses dealing with grammar, pronunciation, speaking, writing and integrated skills. When requested to self-evaluate their ability in the target language on a scale of 1 (lowest) to 6 (highest), they provided the following ratings: overall mastery—4.04, grammar—3.75, vocabulary—4.17, pronunciation—4.09, reading—4.49, writing—3.98, listening—4.08, and speaking—3.92, which shows that they were rather confident of their ability to use various segmental and suprasegmental features. The students were convinced of the importance of correct pronunciation in learning English, as indicated by the rating of 4.08 on a scale of 1 (lowest) to 5 (highest). As regards attainment, the average grade in the pronunciation course was 3.80 on a scale of 1 to 5, while the score on the end-of-the-year exam in practical English was 64.18 % for the oral interview and 61.10 % for the pronunciation component of this interview.

3.3 Data Collection and Analysis

The data were collected by means of a specifically designed survey, which was fashioned on a tool for investigating beliefs about form-focused instruction (i.e., grammar teaching and error correction), designed by one of the present authors (Pawlak, 2012, 2013a, b), in accordance with the assumption that different aspects of pronunciation can be viewed as target language forms and thus the same categories can be applied. Therefore, in contrast to the instruments used in many of the studies overviewed above, the tool was much more specific in the sense that the statements it contained were intended to offer insights into various aspects of pronunciation teaching. On the one hand, these statements were reflective of the latest developments in theory and research into teaching language forms, mainly grammar (e.g., Larsen-Freeman, 2010; Loewen, 2011; Nassaji & Fotos, 2011), and modeled on questionanires dealing with learners’ beliefs in this area (e.g., Loewen et al., 2009; Schulz, 2001; Spada et al., 2009). On the other, the necessary modifications were introduced so that the items included were reflective of pronunciation instruction, and the literature on pronunciation teaching and learning was consulted as well (e.g., Kelly, 2000). The questionnaire was worded in English and constructed in such a way that it provided factual (e.g., experience in learning English, self-assessment of overall proficiency as well as specific skills and subsystems, access to the target language outside school, final grade in the pronunciation course) and attitudinal (i.e., beliefs about overall importance of pronunciation in language learning and specific facets of instruction in this area) information. The core of the survey was constituted by 30 Likert-scale items, where the respondents were requested to indicate the extent of their agreement on a five-point scale (1—strongly disagree, 5—strongly agree) with respect to: (1) overall importance of pronunciation instruction (also in connection with different skills), (2) syllabus design (i.e., structural, where a list of features is determined in advance vs. task-based, where pedagogic intervention is mainly determined by learner need), (3) planning classes devoted to pronunciation teaching (i.e., isolated, where pronunciation features are practiced in separation from communication, vs. integrated, where they are targeted in the course of conveying messages), (4) introduction of phonological forms (i.e., deduction vs. induction, the use of the mother tongue, metalanguage and demonstration), (5) ways of practicing pronunciation features (controlled vs. communicative practice, as well as such based on reception and production), and (6) the provision of corrective feedback on pronunciation errors (i.e., focus, timing, source, corrective technique). This was complemented by four open-ended questions which were more general and focused on the reasons why the respondents liked or disliked learning pronunciation as well as preferred and dispreferred ways of being taught this language subsystem. The tool was piloted with a comparable group of respondents, which allowed introducing changes to some of the items, and internal consistency reliability was determined by computing Cronbach’s alpha, which stood at 0.77, a value that was satisfactory.Footnote 1

The questionnaire was sent out to the participants electronically together with instructions that both Polish or English could be used in the case of the open-ended questions, with the completed copies being returned to the researchers by e-mail. A combination of quantitative and qualitative analytical procedures was employed to analyze the collected data. The former were mainly applied in the case of Likert-scale items and involved tabulating the averages and frequencies of the responses (i.e., 1, 2, 3, 4 and 5), collapsing them into three categories (i.e. strongly agree/agree, undecided, strongly disagree/disagree) and computing their percentages. In addition, in order to establish the relationships between beliefs and attainment, Pearson product-moment correlations were computed. When it comes to the latter, it was employed with responses to the open-ended questions and involved identification of recurring themes, although the frequency of occurrence of the most frequent of those themes was also tabulated.

3.4 Research Findings

Before taking a closer look at the results for the specific Likert-scale statements and the categories that they comprise, it should be noted that the participants were largely convinced of the importance of correct pronunciation in learning English, as evidenced by the rating of 4.08 on a scale of 1–5 in the demographic section of the questionnaire. When it comes to the students’ perceptions of various aspects of pronunciation instruction, they are presented in Tables 1, 2, 3, 4, 5 and 6, each of which is related to one of the areas investigated in the present study (i.e., overall importance of pronunciation instruction, choice of the syllabus, format of pronunciation-based classes, introduction of pronunciation features, instructional options used to practice pronunciation features, and different ways of correcting pronunciation errors). In each case, the percentages of responses to the Likert-scale items in the agree (A), undecided (U) and disagree (D) categories are supplied, together with the mean (M) and the value of standard deviation (SD) for each statement.

Table 1 Beliefs manifested by the participants about overall importance of pronunciation instruction
Full size table
Table 2 Beliefs displayed by the participants about syllabus type
Full size table
Table 3 Beliefs manifested by the participants about the design of classes devoted to pronunciation
Full size table
Table 4 Beliefs manifested by the participants about introducing pronunciation features
Full size table
Table 5 Beliefs manifested by the participants about practicing pronunciation features
Full size table
Table 6 Beliefs manifested by the participants about corrective feedback on pronunciation errors
Full size table

As can be seen from Table 1, containing statements reflective of the perceptions of the role of pronunciation in the process of foreign language learning, the majority of the participants were convinced that the knowledge of segmental and suprasegmental pronunciation features and the ability to use them correctly is highly beneficial. This is evident from the fact that for as many as 5 out of 7 items in this category there was over 80 % agreement as to the important role of pronunciation and the mean values exceeded 4, sometimes considerably. To be more precise, the students were convinced of the positive role of correction (statement 5–90 %, M = 4.50), they approved of such teacher intervention (statement 2–86.4 %, M = 4.39), they were of the opinion that the knowledge of pronunciation aids communication (statement 10–86.4 %, M = 4.19), they enjoyed learning English pronunciation (statement 6–82.7 %, M = 4.19), and they believed that good pronunciation enhances their listening comprehension skills (statement 1–80.9 %, M = 4.11). Although the results for the remaining two statements were lower, they were also indicative of favorable attitudes towards learning and teaching pronunciation since 67.3 % of the participants agreed that good language learners are cognizant of pronunciation issues (statement 9, M = 3.76) and 58.2 % expressed the opinion that knowing and practicing pronunciation will lead to the improvement of their ability in English (statement 3, M = 3.67). It should also be pointed out that the percentage of the students questioning the facilitative role of pronunciation instruction never exceeded 10 % (it was the highest for statements 3 and 10), in some cases a considerable number of students were undecided, in particular in the case of statement 3 linking overall improvement to pronunciation (32.7 %) and statement 9 concerning good language learners (26.4 %). As to the values of standard deviation, they oscillated between 0.7 and 0.9 and although they are not extremely high, they do indicate that there was considerable individual variation when it comes to responses to the statements in this category.

As regards the participants’ beliefs concerning syllabus design, the results included in Table 2 testify to their preference for a structural rather than a task-based model of setting the agenda for pronunciation instruction. This is because while 62.7 % of the students declared that they would like to be given a list of pronunciation features to be covered in a course (statement 25, M = 3.72) and only 9.1 % were of the opposite opinion, just 27.2 % would like to focus only on the forms that are a source of problems in communication (statement 26, M = 2.96) and 31.9 % did not find this pedagogic option appealing. What is striking in this case are quite high percentages of responses in the undecided category, particularly for the statement reflective of a preference for a task-based syllabus (40.9 %), which might perhaps indicate that the students are not familiar with this pedagogic option with respect to pronunciation teaching. Also interesting is the fact that the values of standard deviation for both statements are over 0.90, a result which points to somewhat greater dispersion of responses than in the case of overall value of pronunciation instruction.

Much less straightforward are responses to statements reflecting beliefs about whether the teaching of pronunciation features should be separate from communicative activities or integrated with such activities. As illustrated in Table 3, on the one hand, the participants seemed to manifest a predilection for being given the opportunity to focus their attention on a specific pronunciation target, reflect on an explanation and then engage in practice activities dealing with this feature. This is evidenced by the fact that 70 % of them agreed with statement 27 (M = 3.84) concerning their awareness of what pronunciation feature is the focus of a particular class, 63.7 % stated that they liked to be provided with an explanation and practice opportunities, and 64.5 % preferred to practice individual sounds before they used them in speech (M = 3.67), with the percentages of those disagreeing standing at 8.2 %, 11.8 % and 16.3 %, respectively. On the other hand, however, even more participants were of the opinion that pronunciation practice should be incorporated into communicative activities, as demonstrated by the fact that 86.4 % agreed with statement 28 (M = 4.25) and not a single person disagreed with it, and 71.9 % concurred with statement 30 (M = 3.89), with just 5.4 % of disagreement. The students were somewhat less convinced of the superiority of suprasegmental features, such as rhythm and intonation, over individual sounds, although 48.7 %, a clear majority, agreed with statement 18 (M = 3.43) and only 12.7 % disagreed. On reflection, though, these findings should not be regarded as overly surprising or contradictory because they might indicate the participants’ conviction that the two options should be combined, with pronunciation features first being introduced and practiced, and later employed in meaning and message conveyance. It should be noted that, with the exception of statement 28 (13.6 % of undecided students and the value of standard deviation standing at 0.67), quite a few respondents were in two minds about the value of isolation and integration, as indicated by the fact that the percentages of responses in the undecided category approached or exceeded 20 %, and they also varied in their opinions to a considerable extent, as shown by the SD values ranging from 0.88 to 0.97.

The situation is no less complex when it comes to the introduction of new pronunciation features, with the key distinction lying in the choice between deduction (i.e., rule provision) and induction (i.e., rule discovery), as well as specific techniques within each of these two options. As shown in Table 4, the participants manifested a strong preference for being provided with an explanation of the pertinent pronunciation rules, since as many as 80 % agreed with statement 11 (M = 4.01) and just 4.5 % disagreed with it. These results do not mean in the least that the students downright rejected induction as a viable instructional option, because as many as 81.8 % stated that would rather be exposed to new pronunciation features in spoken texts than be provided with rules (statement 8, M = 4.04), 56.4 % were of the opinion that it is best to discover pronunciation rules in cooperation with others (statement 7, M = 3.61), and 56.4 % stated that they liked to discover such rules on their own (statement 4, M = 3.59), with the percentages of disagreement for these items standing at 1.8 %, 14.5 % and 15.4 %, respectively. As regards the specific ways in which the introduction of pronunciation features should proceed, the students were by and large in favor of the use of demonstration, as indicated by 60 % agreement and 17.2 % disagreement with statement 17 (M = 3.60), and they were cognizant of the facilitative role of the mother tongue in teaching pronunciation, as evident in the fact that 50.9 % agreed with statement 12 (M = 3.42) and 20.9 % were of the opposite opinion. By contrast, the students were much more skeptical about the need for reliance on metalanguage when teaching pronunciation features, since only 33.7 % agreed with statement 14 (M = 3.03) while 27.2 % disagreed with it. What is striking for this category are quite high values of standard deviation, exceeding 1.0 for items 4, 7, 12, 14 and 17, which indicates that there is considerable individual variation among students when it comes to their preferences concerning the introduction of phonological forms.

The students’ preferences concerning practicing pronunciation features were considered with respect to the key distinctions between reliance on controlled activities and communicative tasks, on the one hand, and the use of production-oriented and reception-based techniques, on the other. As can be seen from Table 5, also in this case, the respondents seem to be rather eclectic in their approach and recognize a beneficial role of what specialists sometimes view as opposite, perhaps even mutually exclusive, instructional options. This is because, although 62.8 % of the participants emphasized the importance of the use of pronunciation features in communicative activities (statement 15, M = 3.73), 54.5 % acknowledged the value of performing traditional exercises, such as minimal pair practice (statement 19, M = 3.58). What is noteworthy in this case are very low percentages of disagreement (7.2 % for statement 15 and 8.2 % for statement 19) and large numbers of undecided students (30.0 % for statement 15 and 37.3 % for statement 19). The responses to both of these statements testify to the conviction on the part of the participants that pronunciation practice has to involve the production of spoken output, be it more or less spontaneous. On the other hand, however, as many as 81.8 % of the students were of the opinion that learning pronunciation can be enhanced by listening to authentic communication (statement 20, M = 4.23), a mere 1.8 % rejected this assumption, and 16.4 % were undecided. In addition, 55.5 % expressed the view that they would prefer to know how a specific pronunciation feature is produced, before they are requested to use it themselves (statement 13, M = 3.73), while 13.6 stated the opposite. Also here the number of responses in the undecided category was substantial, amounting to 30.9 %, and the value of standard deviation was the highest, equaling 1.08.

As illustrated in Table 6, the participants were rather conservative in their beliefs concerning corrective feedback, which should not come as a surprise as this outcome is consistent with the results of studies on learners’ preferences concerning form-focused instruction in general (e.g., Pawlak, 2012, 2013a, b). In the first place, an overwhelming majority of the students were in favor of teacher correction, which is evidenced by the fact that only 6.4 % expressed a preference for peer-correction (statement 23, M = 1.92) and as many as 74.5 % rejected it as a viable instructional option. They were also against confining the provision of corrective feedback to pronunciation errors which get in the way of communication, as shown by 20.1 % agreement and 43.5 % of disagreement with statement 24 (M = 2.75). This could be interpreted as meaning that the focus of this type of intervention should be much broader and it perhaps should even be applied to the majority of pronunciation errors, although it has to be kept in mind that as many as 36.4 % of the students were undecided. The situation is more complex when it comes to the timing of corrective feedback because while 63.6 % of the participants expressed a preference for immediate correction (statement 21, M = 3.73), 50 % were of the opinion that delayed correction is a better option (statement 22, M = 3.38). Still, it should be noted that many more students were against delayed correction (33.6 %) than against immediate correction (17.3 %), which could mean that the latter is perceived as more advantageous. What is noteworthy in this category are very high values of standard deviation for all the items, ranging from 0.94 to 1.37, which implies that the provision of corrective feedback is an area where consensus regarding specific instructional options is hard to find.

An attempt was also made to identify relationships between the participants’ beliefs regarding pronunciation instruction with respect to the categories discussed above and attainment, operationalized as the overall score on an oral interview, which is part of an end-of-the-year examination in practical English, and the pronunciation component of that score. It turned out, however, that all of these correlations were very weak and not statistically significant, with the highest value of the correlation coefficient (r = −0.17, p = 0.076) being identified in the case of the relationship between a favorable attitude towards pronunciation learning and teaching (see Table 1 above) and the pronunciation score. The fact that this relationship is negative could perhaps be cause for concern were it not for the fact that it is almost negligible and the participants’ beliefs can at best be said to account for about 3 % of the variance in achievement. Interestingly, there was a very strong positive correlation between the pronunciation score and the overall result of the oral interview, with r = 0.784 (p < 0.0001), explaining over 60 % of the variance in the exam scores.

The more general preferences of the students’ concerning learning and teaching pronunciation were tapped by means of four open-ended questions included at the end of the survey. When asked about the reasons why they liked learning pronunciation, the participants most often (54) pointed to the fact that it helps the communicate more effectively, as illustrated by such statements as: “Because the better my pronunciation, the better people will understand me and the easier I will find it to understand them. Knowing vocabulary, grammar and so on is kind of useless if you are unable to pronounce those structures correctly”, or “It improves my English and I feel more comfortable in communication when I know more about pronunciation of specific words”. Apart from this, the students also stated that it helps them develop other language skills (19), makes it possible for them to become more native-like (14), boosts their confidence (9), helps them understand others (5), and they like the sound of English (4), with single respondents also commenting that learning pronunciation is fun, it is interesting, important, entertaining, enjoyable or it is simply something that a student of English has to do. As regards their justifications for disliking learning pronunciation, the students most often wrote that it is difficult (34) or boring (24), and some of them also commented that it is less important than grammar or vocabulary, it can be acquired naturally through listening, it is time-consuming or it is demotivating because the native-speaker level is beyond their reach. It should be noted, however, that as many as 30 participants simply reiterated that they liked working on this subsystem and thus provided no response to this question. With respect to the most preferred instructional techniques, a vast majority of the students (51) mentioned different types of listening (e.g., to native speakers, the teacher, other students, radio, songs, CDs, podcasts), a representative example being the following comment: “Listening to recorded conversations among natives, listening to the teacher and noticing how he pronounces things, speaking with other students in class”. Apart from this, the students also mentioned repetition (20), making dialogs (14), the use of the pronunciation features taught in communication (13), doing typical exercises and practicing (12), reading and listening at the same time (11), being corrected (9), being provided with explicit information about articulation (3), or reading and repeating words (3), with single respondents also pointing to paying attention to sounds while doing other things, working with others, analyzing phonetic script, or simply stating that all the ways of learning pronunciation are beneficial. By contrast, the most dispreferred techniques included different forms of repetition (i.e., of single words or words that nobody uses, after the teacher or CD, in a chorus, all of these without being corrected) and techniques based on theoretical considerations (e.g., the use of complex terminology, provision of rules, the use of charts). As one of the respondents put it, “I do not like learning exactly the description of the way of articulation of new sounds. In my opinion, the practice of pronunciation of a new sound is more important and helpful in successful communication with others, not theory”. Other participants were not very happy about using phonetic transcription, being corrected by the teacher or other students without an explanation, reading and recording or even incompetent teachers.

4 Discussion

The picture that emerges from the findings reported above is exceedingly complex and in some cases quite difficult to interpret. As regards the first research question concerning the participants’ beliefs about various aspects of pronunciation instruction, it is clear that they manifest very favorable attitudes towards learning this target language subsystem and they are convinced that good pronunciation can be beneficial, both with respect to effective communication and the development of specific language skills. Such results should not be surprising, given the fact that they were English majors who are required to attend theoretical and practical courses dealing with phonetics, they are expected to achieve high levels of mastery in this area, and pronunciation is included among the criteria taken into account during final examinations. The students also expressed a strong preference for the structural syllabus, where pronunciation features are preselected, ordered and taught one by one, which, again, can be accounted for in terms of the instruction they receive since pronunciation courses are structured in exactly such a way, with segments typically being covered in year 1 and aspects of suprasegmental phonetics in year 2. Since there is usually little integration between this course and other courses in practical English, such as those devoted to speaking or integrated skills, where problematic pronunciation features could be targeted, these sentiments are perhaps understandable. On the other hand, the participants’ beliefs concerning the design of pronunciation-based lesson, the introduction of phonological forms or the techniques which can be used to practice these forms demonstrate that they were in favor of combining quite disparate approaches, as long as it serves the purpose of improving their pronunciation. Thus, even though they are clearly in favor of deduction, they also see the value of a more inductive approach, they recognize the contribution of controlled and communicative practice, as well as production and reception of the elements of pronunciation. Such a stance seems to be fully warranted in view of the fact that different stages in the process of pronunciation instruction may require the application of different techniques and procedures (e.g., reception and controlled practice quite naturally precede production and the performance of communicative tasks), and it seems to be reflective of the students’ considerable experience in learning this subsystem. The situation is more straightforward in the case of the provision of corrective feedback, because the participants were overwhelmingly in favor of teacher correction, preferably such that would target a number of pronunciation errors, not only those that impede communication. On the other hand, they were less decisive with respect to the timing of correction, since they approved of both immediate and delayed intervention, in all likelihood depending on the type of activity being performed. A comment is also in order on the high percentages of neutral responses and high values of standard deviation in some cases. Yet again, this trend is to some extent predictable in light of the fact that the students could have been aware that different instructional options could be beneficial for different tasks and stages of learning, and that, quite advanced as they were, they could have opted for ways of learning that might be regarded as disparate but were effective for them.

Much more difficult to address is the second research question dealing with the relationship between the participants’ beliefs about different aspects of pronunciation instruction and attainment on the oral interview, for the simple reason that the observed correlations were negligible and not statistically significant. At first blush, these results might be viewed as inexplicable since, also in line with the results of previous research on beliefs (see e.g., studies mentioned in the introduction to this paper), it could be assumed that learners’ preferences should have a bearing on their performance and ultimate level of achievement. Such an assumption, however, may be overly simplistic for a few reasons. First, while the conviction about the overall value of pronunciation instruction could indeed be expected to be related to attainment, such a relationship may not be the case for the various facets of instruction investigated in the present study, since, as was demonstrated above, different instructional options may be perceived as equally useful in different situations by the very same students, which makes it difficult to pinpoint a direct link with attainment. Second, the culprit for this lack of relationship could be a considerable degree of individual variation, which is predictable in the case of learners at this level, and their cognizance of the need to adjust instructional practices to different targets, tasks, goals and current priorities, which, in line with the tenets of complex system theories (Larsen-Freeman & Cameron, 2008), indicates that the interaction between beliefs and learning outcomes is intricate, dynamic, and affected by other variables. Third, it should be remembered that the limited attention and working memory capacities may preclude many students, even those at seemingly high proficiency levels, from effectively monitoring their pronunciation in real-time communication, where so many other things are at a premium (e.g., the choice of lexis, grammatical accuracy, pragmatic issues, keeping track of what is being said, planning what to say next), and lack of automaticity is bound to expose all the deficiencies. Such problems are inevitably exacerbated in an exam situation, not least because of the negative impact of affective factors, such as anxiety, which might indicate that the measures of attainment used in the present study might have been problematic.

Moving on to research questions three and four, the results can be regarded as promising because most of the students seemed to be aware that pronunciation is not studied and practiced for its own sake, and to realize that its mastery is important because it has the potential to contribute to successful communication. It is also comforting that the students reported attaching so much importance to different forms of listening, also to authentic input, which is undoubtedly indispensable for the transition from the use of pronunciation features in controlled activities to their accurate employment in spontaneous speech. On the other hand, the quite widespread opinion that pronunciation is difficult to learn and the process of learning is boring are insightful in the sense that they should alert teachers to the need to raise students’ awareness in this respect and define the learning challenge in more accessible ways. By the same token, the visible disgruntlement with mindless repetition, often overused in pronunciation classes or the lack of appreciation for theoretical explanations should sensitize teachers to the fact that the palette of instructional options should be extended and efforts should be made to explain to students how metalinguistic information can contribute to greater mastery of pronunciation features.

It is also necessary to acknowledge the limitations of the study, which might be partly responsible for the fact that the results were in some cases inconclusive and the expected relationships failed to be identified. In the first place, the questionnaire used to collect the data is clearly in need of further development and refinement, both in terms of the choice of the main categories, the statements included in these categories, and the wording of the items themselves, a task that may necessitate performing factor analysis. This is the corollary of the fact that its construction represented the first attempt to apply the findings of more general research in form-focused instruction to learning pronunciation and it was clear from the very outset that the instrument was far from perfect on a number of counts. Another problem is connected with the fact that attainment was operationalized in terms of performance on a high-stakes examination, a situation which is perhaps not very conducive to objectively gauging students’ proficiency in any area, for the reasons explicated above. Although examination scores can be regarded as objective, not least because they are arrived at by more than one person on the basis of preset criteria, and there is no reason why they should not be considered in future studies, what could also have been taken into account were the final grades in the pronunciation course of the participants self-assessment with respect to pronunciation. Finally, it should be borne in mind that, being English majors expected to achieve high levels of proficiency and having the benefit of extensive theoretical and practical courses in phonetics, the participants represented a very distinctive group of foreign language learners and their beliefs are very unlikely to be shared by other groups of students for whom good pronunciation is hardly a priority.

5 Conclusion

Although there have been numerous studies of pronunciation learning and teaching, few of them have addressed learners’ or teachers’ beliefs in this area, and even fewer have done so in a principled manner, targeting different aspects of pronunciation instruction. The empirical investigation reported in the present paper has sought to remedy the situation by approaching pronunciation teaching in the same way as any other type of form-focused instruction and adopting the categories derived from theory and research in this area. Obviously the study represents merely the first step in this direction, mainly because the instrument still suffers from a number of weaknesses and it is clearly in need of further validation and modification for the purposes of future research. It is also evident that future studies should target other age groups, proficiency levels and educational contexts, and an attempt should be made to determine links between such beliefs and attainment, take into account variables that may impact differences in beliefs about pronunciation instruction (e.g., learning styles and strategies, gender, goals, previous experiences), explore the relationship between learners’ and teachers’ beliefs in this area, investigate the connection between such cognitions and actual classroom practices, as well as examining more temporal and situation-specific nature of beliefs. While the use of well-designed surveys distributed among large populations is one way of investigating such issues, a situated, context-sensitive approach is also necessary to obtain a more multi-faceted picture thereof, which would necessitate the use of a variety of data collection tools and the application of more longitudinal research designs. Such methodological considerations aside, the study of beliefs concerning pronunciation instruction is without doubt a worthwhile undertaking, as it can help us better understand whether, why and how learners like to be taught aspects of this target language subsystem, which can contribute to greater effectiveness of instructional practices.