The Importance of Pronunciation Instruction in the English as a Foreign Language Classroom

(1)

Linköping University | Department of Culture & Society Linköpings universitet | Institutionen för kultur och samhälle Thesis 1, 15 credits | Secondary School Teachers’ Programme (upper secondary) - English Konsumtionsuppsats, 15 hp | Ämneslärarprogrammet (gymnasieskolan) - Engelska Spring Term 2020 Vårterminen 2020

The Importance of Pronunciation

Instruction in the English as a

Foreign Language Classroom

Vikten av uttalsundervisning i engelska som främmandespråk

Axel Tegnered Jonas Rentner

Supervisor/Handledare: Nigel Musk Examiner/Examinator: Michael Smith

Linköping University Linköpings universitet SE-581 83 Linköping, Sweden 013-28 10 00, www.liu.se

(2)

English

Institutionen för kultur och samhälle

Department of Culture and Society

581 83 LINKÖPING

Seminariedatum

Seminar date

Ämne Subject Språk Language Rapporttyp Type of Report

Engelska English Engelska English Examensarbete 1 (konsumtion) Thesis 1

Title (in English)

The Importance of Pronunciation Instruction in the English as a Foreign Language Classroom

Titel (svensk översättning)

Vikten av uttalsundervisning i engelska som främmandespråk

Författare Authors

Axel Tegnered och Jonas Rentner

Sammanfattning Summary (in English)

The field of pronunciation instruction has long been a neglected area. In recent years, however, researchers have found a renewed interest in the subject. For example, research has been conducted to investigate the views on pronunciation instruction held by teachers and learners. In light of this, the present study seeks to explore the field of research in pronunciation instruction to answer the question of whether pronunciation instruction is important and beneficial for learners of English as a foreign language. The results of the present study show that teachers and learners alike consider pronunciation instruction an integral part of language teaching. Furthermore, the results show that pronunciation affects comprehension and that pronunciation teaching has positive effects on pronunciation development. In the present study, these results are connected to the Swedish school context, where we draw the conclusion that pronunciation should be included in the English language classroom to a greater extent than is the case at the time of writing this thesis.

Nyckelord Keywords

Pronunciation, explicit pronunciation instruction, teacher views pronunciation, learner views pronunciation, comprehensibility

(3)

1. Introduction

When studying to become an English teacher, or any teacher for that matter, one naturally starts to reflect upon one’s own experiences as a learner. It seems natural to take inspiration from the experiences from one’s own education and apply them to the future teaching profession. Contemplating our own experiences as learners, we have noticed that very little time has been spent exploring the different aspects of pronunciation. Even though the core of the English classroom in Sweden is communicative, it seems that pronunciation, which is an important element in getting your message across, has rarely been explicitly taught. In fact, recent research shows that pronunciation has long been neglected in English teaching. Also, the Swedish syllabus offers no further insight into how pronunciation teaching should be considered. Since this is the case, it is important to consider why this might be and what pronunciation research can tell us about the matter. Therefore, this thesis aims to determine what teachers and learners think about pronunciation and what research has shown when it comes to the importance of pronunciation instruction.

1.1 Aim and Research Questions

The aim of the present study is to investigate to what extent pronunciation instruction is important and beneficial for EFL (English as a Foreign Language) learners. To answer this question, we investigate different aspects of pronunciation teaching, such as teachers’ and learners’ views on the subject along with the documented effects of pronunciation instruction on pronunciation development. Additionally, we describe what studies have found as regards the effect of pronunciation on comprehension. Four research questions have been constructed to guide our investigation:

1. What views are held by teachers on pronunciation and pronunciation instruction?

2. What views are held by learners on pronunciation and pronunciation instruction?

3. How does pronunciation affect comprehension?

(6)

2

1.2 Outline of the Review

The present study is divided into five chapters. Following this introduction, we present a theoretical and historical background of pronunciation instruction and research, including a review of the syllabus and its formulations on pronunciation teaching. Next, we account for our method in selecting and analyzing studies, and we provide an overview of the chosen data. In the fourth chapter, the research review, we present our analysis of the 25 included studies. Finally, we provide a discussion chapter, where we discuss our findings based on the research questions. Furthermore, the discussion provides an analysis of the implications that our findings have for the Swedish context. To conclude the discussion, we suggest some areas in need of further research.

2. Background

This chapter provides a historical and theoretical background to the field of pronunciation instruction. The first section provides a brief background to the history of pronunciation teaching, covering what the views on instruction have been and where they stand today, along with a presentation of the implications that different pedagogical approaches and perspectives have had on pronunciation instruction. The following three sections provide explanations and definitions of concepts central to this thesis. The final section of the background chapter accounts for the requirements and guidelines presented in the Swedish upper-secondary syllabus for English and its accompanying commentary material.

2.1 Historical Background to Pronunciation Instruction

Even though the interest in pronunciation instruction seems to be increasing (Brinton 2016: 258), historically it has long been a neglected area of English teaching (Tergujeff 2013: 10). During the 1800s and early 1900s, however, pronunciation was held in high regard as a means to learn a new language (Ketabi & Saeb 2015: 182). The Direct Method, a method essentially based around imitating “the rhythm and sound system of the target language” (ibid.), was a central approach in language teaching during the late 1800s. In the 1890’s, the Reform movement gained momentum and the phonetic system (henceforth referred to as IPA) was established (ibid.). Even during the era of Audiolingualism, based on the Contrastive Analysis Hypothesis, pronunciation was a very central aspect of acquiring a new language (Brinton 2016: 257). Using IPA, pronunciation drills were utilized to get learners to acquire sounds and features of the target language (ibid.).

(7)

3 However, from the 1960s onward, the advent of the cognitive approach had a huge impact on the view on pronunciation (Ketabi & Saeb 2015: 183). The idea that language was simply a set of habits that a learner could get accustomed to was banished, and Chomsky’s theory of Generative grammar took hold. This meant that language learning started to be viewed as a rule-governed system, and grammar became an increasingly important part of language instruction. Thus, pronunciation instruction was more or less abandoned (ibid.).

During the 1970’s, pronunciation continued to have a marginalized position in language acquisition. Different approaches, such as the Natural Approach and the Total Physical Response further deemphasized the role of pronunciation, since both these approaches consider receptive comprehension to be the key element of language acquisition. Accordingly, teachers and learners did not need to focus on pronunciation, as it was not considered a key skill for developing comprehension. Similarly, Communicative Language Teaching (CLT), gaining popularity during the 1970’s, was yet an approach that largely neglected the area of pronunciation (ibid.), as it focused on moving students’ attention to meaningful communication. Moreover, proponents of CLT argued that focusing on

pronunciation would interfere with communication and could possibly be harmful to students’ self-confidence (Ketabi & Saeb 2015: 183).

A renewed interest in pronunciation could be observed from the late 1970s into the 1980s (Ketabi & Kateb 2015: 184). Among other ideas, learner involvement and self-monitoring were theories that brought this view forward (ibid.). In the 1990s, the concept of English as a Lingua Franca (ELF) began to gain momentum, mainly under the influence of researcher Jennifer Jenkins (Deterding 2013). This concept further enhanced the interest in pronunciation as a crucial part of foreign language instruction (Ketabi & Saeb 2015: 184), arguing that a core set of pronunciation aspects should be taught. Even though ELF has had a wide spread over the last decades, it continues to be controversial (Deterding 2013: 4), as some say it is simply an “impoverished code that has limited domains of use” (ibid.: 5) and “merely a [...] last-resort means of communication” (ibid.).

2.2 The Nativeness Principle vs the Intelligibility Principle

There are two major principles affecting views and practice in pronunciation instruction: the Nativeness Principle and the Intelligibility Principle. In this section, we begin by providing

(8)

4 brief definitions of the two principles, after which we describe their characteristics. To

conclude, we report that recent research seems to advocate the Intelligibility Principle, and we highlight some arguments for this from articles significant for the field.

The stance of the Nativeness Principle is that achieving native-like pronunciation is the ultimate goal of pronunciation instruction (Levis 2005: 370). According to Levis, the Nativeness Principle “drives the accent reduction industry” (ibid.), which holds that

instruction in conjunction with learners’ motivation can “eliminate a foreign accent” (ibid.). Central to the discussion of the opposing principles, then, is not only that the Nativeness Principle holds that native-like pronunciation is the aim, but also that it conveys the message that such an aspiration is indeed achievable (ibid.). In contrast, the Intelligibility Principle emphasizes that the main aim of pronunciation instruction should be for learners to be understandable (ibid.), and not to attain a native-like accent.

An analysis of research from the last decades seems to indicate that the Intelligibility Principle has gained momentum. For example, Jenkins (1998) writes:

Many of us working in the field of pronunciation […] no longer regard […] the target of pronunciation teaching as a native-like accent, with the eradication of all traces of a 'foreign' accent, however unrealistic that target always was. (119)

Jenkins’ argument is supported by Thomson and Derwing (2015), who conducted a meta study investigating the results of 75 pronunciation studies in a second-language context. They argue that “native-like pronunciation is an unrealistic goal, but that improved intelligibility and comprehensibility are achievable” (ibid.: 335). Furthermore, Thomson and Derwing’s findings suggest that an improved accent does not necessarily generate increased

comprehensibility and, vice versa, that increased comprehensibility can be achieved without improved accent (ibid.: 337). Moreover, Derwing & Munro’s (2005) findings suggest that there is no research indicating “a link between pronunciation instruction and the elimination of a foreign accent” (384).

However, despite the emphasis on the Intelligibility Principle, Jenkins (1998) and Levis (2005) point out that the Nativeness Principle continues to affect teachers and curricula (Jenkins 1998: 119; Levis 2005: 371). Furthermore, Jenkins (1998) highlights the importance

(9)

5 of avoiding “political correctness”, meaning that if learners wish to aim for native-like

pronunciation, they should be allowed to do so (125).

2.3 Segmentals and Suprasegmentals

Segmentals are the individual sound units in pronunciation, in other contexts often referred to as phonemes (Field 2005: 402). A phoneme, in turn, is described by Encyclopædia Britannica as the “smallest unit of speech distinguishing one word (or word element) from another” (2020).

The term suprasegmentals, on the other hand, refers to the aspects of word stress, rhythm, and intonation, often collectively called prosody (Jenkins 1998: 121; Field 2005: 402; Brinton 2016: 258). Suprasegmentals are aspects of pronunciation that “are not limited to single sounds but often extend over syllables, words, or phrases” (Encyclopædia Britannica 2020). Suprasegmental features are sometimes also referred to as global aspects of pronunciation, especially in less contemporary research (see for instance Derwing, Munro & Wiebe 1998).

2.4 Intelligibility, Comprehensibility and Accentedness

For the evaluation of pronunciation production, three major aspects are identified and

consistently used in the literature covered: intelligibility, comprehensibility, and accentedness. The terms are defined by Derwing (2010: 29) and Derwing & Munro (2005: 385) as follows: intelligibility refers to how well a speaker’s utterance is understood by a listener;

comprehensibility is the listener’s subjective judgement of how difficult the speaker’s utterance is to understand; accentedness refers to the degree to which the speaker’s accent differs from the variety that acts as a norm in a given context.

2.5 The Swedish School Syllabus

In this section, we present the parts from the Swedish upper-secondary syllabus for English (Kursplan Engelska gymnasiet) that concern pronunciation instruction. Additionally, we present aspects of the commentary material accompanying the syllabus and discuss what the Swedish National Agency of Education (Skolverket) says about the subject of pronunciation, also commenting on what is not said.

(10)

6 Indeed, the subject of pronunciation instruction is sparse in the English syllabus for the upper-secondary school. The passages that relate to the subject describe what should be achieved by students in quite general terms; for instance, one aim of the subject is to “develop [...]

confidence to use English in different situations and for different purposes” (Skolverket 2019: 1). Furthermore, it emphasizes that students should be given the opportunity to “develop all-round communicative skills” (ibid.) and “develop correctness in their use of language in speech and writing” (ibid.). The additional commentary material states that inherent in “develop[ing] correctness” in students’ use of language are skills such as “vocabulary, [...] pronunciation [and] prosody” (Skolverket n.d.: 2, authors’ translation).

There are three English courses of upper-secondary education in Sweden: English 5, English 6 and English 7 - each building on preexisting knowledge from the former course. The core content of each course is divided into three major sections: “Content of communication”, “Reception” and “Production and interaction” (Skolverket 2019: 3). The commentary material notes that “Content of communication” contains the subjects which the other two should contain (Skolverket n.d.: 3). In the “Production and interaction” section of English 5, the syllabus states that the course is to cover “oral [...] production and interaction of various kinds, also in more formal settings” (Skolverket 2019: 4). A similar passage is found in the English 6 and 7 courses, which further add “in different situations and for different purposes” (ibid.: 8, 12). The commentary material gives no further explanation to these passages, other than explaining the terms used to differentiate to what extent this skill should be

comprehended in each corresponding grade (Skolverket n.d.: 10).

As is evident in the paragraphs above, the English syllabus for upper-secondary education in Sweden gives no clear indications or instructions as to how pronunciation should be

incorporated into the English-language classroom. In light of the methods and approaches presented above, it seems that the Swedish syllabus is constructed mainly with the CLT approach in mind. One of the indications that supports this can be found in the main titles of the core content, labeled “Content of communication”, “Reception” and “Production and Interaction”, emphasizing the focus on developing communicative skills. Thus, the syllabus leaves room for interpretingthe teaching of skills concerning language form, while also lacking a clear pronunciation model indicating what the aim of pronunciation teaching should be.

(11)

7

3. Method

This chapter accounts for the data used for this study. In the following sections, four different aspects of this procedure are presented. First, search terms and digital resources are provided, followed up by a description of how our data was sorted, organized, and selected. After this, we present a review of the central aspects of our chosen literature, such as the quantity of qualitative and quantitative articles, their data set and from which countries they originate. Finally, we describe the problems that we encountered in this process.

3.1 Selection

We began our search for material relevant to this study by using digital search engines available online. The main database used was ERIC, but we also used Uni Search, our university’s main search engine. Upon examining what results the different search engines provided when entering specific search terms, we came to the conclusion that the results were similar, indicating that either of the search engines was appropriate for our purposes. Starting out broadly, we used the search term “pronunciation teaching”, which resulted in 2058 hits in ERIC. To narrow down the result and increase the reliability of articles found, we applied filters such as “Peer reviewed” and “Academic journals”. Upon getting a substantial amount of hits from countries such as Turkey, Iran, and China, we also filtered hits geographically, using the function which displays hits from specific countries and regions, selecting among others the UK, California, Austria, Canada, and Germany. We did this in order to incorporate a wider spread of countries and to include a western perspective.

We continued our search by applying search terms connected to “speaking” or “oral skills”, with various additional terms such as “instruction” and “teacher attitudes”. We did this, since at this stage, we aimed toward broader research questions concerning speaking in general, before narrowing down our focus to only the pronunciation aspect. Having decided on the final research questions, we then altered our search terms to incorporate “pronunciation”, along with other terms such as “norms”, “importance” and “effects”. We also included the search terms “ESL” and “EFL” in order to find articles that concerned English specifically. These search terms were all narrowed down using the filter “peer reviewed”, and often the filter “full text”, so that articles we found would be readily available in the search engine.

(12)

8 Apart from using search terms to find articles, we used some reference lists of relevant

articles to further deepen our findings. Furthermore, weactively searched for and found some articles that were frequently mentioned and cited in the research we found through ERIC. These articles, in turn, were used both for the literature review and the background chapter. In addition to the abovementioned literature, we used other sources that were recommended to us, such as a chapter on pronunciation from a research handbook (Brinton 2016), a doctoral dissertation and language conference proceedings.

3.2 Reading and Selecting Articles

After finding a substantial number of articles, the time came to sort them and select the ones appropriate for our study. From the outset, choosing from the hits matching our search terms, we looked at the titles of articles shown in the search engines, and made our first rough selection based on the titles that seemed to concern our main research question or one or more of our sub questions.

In the second sorting of articles, we read the abstracts to see if they seemed relevant to our study. After reading numerous articles, we engaged in discussion and sorted out the ones that did not match our research questions. At this stage, we had roughly 50 articles that either related to the sub question of teachers’ and learners’ attitudes toward pronunciation teaching, or the sub questions regarding the effects of pronunciation on comprehensibility and

pronunciation teaching. From here, we quickly read the articles through again, mainly to see if the result and discussion sections said something relevant to the present study. Another process of exclusion took place, and the articles that were kept were read thoroughly with the help of an overview template. In this process, three additional articles were excluded: one because of ethically inappropriate methods, one because of a lack of pre-test information, and one because of irrelevant contents. From here, after supervision, we opted to exclude any articles that did not concern the English language; that is, studies that examined the effects of pronunciation teaching in other languages were excluded. This process gave us the articles at hand, and these are elaborated on in the next section.

3.3 The Nature of Sources

For the literature review, we used a total of 25 studies, all of which are empirical studies published in academic journals. In the selection and analysis process, these 25 studies were

(13)

9 divided in accordance with the sub questions for our thesis. The description of the chosen studies is divided into two parts, one for each sub question.

For the question concerning teachers’ and students’ views on pronunciation, 14 empirical studies from academic journals were read and analyzed. These studies examine the views held by teachers and students in questionnaires and interviews, thus eliciting both qualitative and quantitative data. Six of the articles are both qualitative and quantitative in the sense that they make use of both questionnaires and interviews. Five of the articles elicit qualitative data by using interviews and conversation analysis, while the remaining three articles are purely quantitative, being based on questionnaires only. The 14 studies in this section examine views and attitudes held by students and teachers in 13 countries, spanning five continents: Turkey, the USA, Hong Kong, Iran, China, New Zealand, South Africa, Pakistan, Japan, South Korea, Australia, Poland and Finland. It is worth mentioning that for teacher views, we decided to incorporate studies on both in-service and pre-service teachers, that is, students in teacher programs, since these students are the teachers of tomorrow.

For the questions about the effects of pronunciation on comprehension and the effects of pronunciation instruction, we used a total of 11 studies, 10 of which elicit quantitative data. One study is purely qualitative, as it makes use of conversation analysis in order to analyze what role pronunciation plays in communication breakdown. The fact that there is a strong emphasis on quantitative data in this section is no coincidence, as research focusing on pronunciation development often makes use of rating scales, thus eliciting quantitative data (Derwing, Munro & Wiebe 1998: 296). Furthermore, many of the studies in this section are experimental, meaning that they utilize a pre-test/post-test design examining the longitudinal effects of pronunciation instruction, which also provides quantitative results. Finally, the 11 studies in this section originate in nine different countries, spanning four continents: Brunei, Turkey, New Zealand, Canada, Costa Rica, Iran, the USA and the UK.

For the background chapter, we used the syllabus for English as a foreign language at the upper-secondary level provided by the Swedish National Agency of Education (Skolverket), including the accompanying commentary material. Additionally, we used multiple sources covering a wide range of pronunciation research both in a historical and present-day

context. The syllabus material was used in order to examine what the Swedish school requires when it comes to pronunciation instruction. These texts are also used when discussing the

(14)

10 implications that our findings have for teaching English in a Swedish context. Regarding the historical and theoretical background, some of the sources used were found through our searches for the literature review, and while we determined them irrelevant for the review itself, they provided relevant information for the background as regards the historical and current contexts of pronunciation research. Additionally, some sources were found via lists of references in other works, most notably articles by Jenkins (1998) and Levis (2005) which are mentioned in a plethora of studies on pronunciation teaching. In addition, we used Brinton’s (2016) chapter on pronunciation from a handbook on language acquisition research. These sources all contribute to establishing a theoretical and historical background for our work and for the field of pronunciation research.

3.4 Limitations and Problems

The first problem we encountered while searching for relevant literature was the number of articles that for one reason or another were unavailable to us. The main reason for this was that some hits encountered in the search engines led to articles that were either blocked behind paywalls or unavailable as a student of Linköping University. This was frustrating as some of the titles of the unavailable articles seemed to match our purposes very well. Another problem of finding and selecting articles was that the authors of this study are fluent only in Swedish and English, and thus only articles available in those languages were available to us. However, this turned out to be less of a problem, since our English search terms were

sufficient for finding the research we needed.

The international scope of the research led to discussions of applicability to the Swedish school context. While the studies included in this thesis cover a total of 17 countries, only five studies originate in a strictly European setting: two from Poland, one from the UK, and two from Finland. The other 20 studies originate in other parts of the world. One example where this was an issue is one study (Saito and Lyster 2012) which examined the possibility of explicitly teaching the voiced alveolar approximant /ɹ/ to Japanese learners of English, which is a known difficulty for these learners. While the results are interesting and possibly

applicable in a general Swedish context, it is harder to apply that specific result to the Swedish school context, since the above-mentioned pronunciation problem is uncommon among Swedish learners. Another example is a Turkish study carried out by Karakaş (2017), examining the underlying ideologies that shape the view of Turkish teacher students’

(15)

11 Swedish teacher students, it is hard to discern how much the Turkish cultural context is an influence on the views observed, and what results would be observed had that study been carried out in Sweden.

Furthermore, some studies originate in countries where English is taught as a second language rather than a foreign language, for example the USA and Canada. This differs from Sweden, where English is taught as a foreign language. While the learners in the mentioned studies are foreign, they learn English in an English-speaking community, thus having some learning opportunities that Swedish learners do not. This, in turn, could mean that their pronunciation development differs from that of Swedish students, which is another example of issues of applicability.

While the problems covered above arguably pose questions as to the studies’ applicability in the Swedish context, we still believe that the sheer number and wide range of origins of the studies compensate for the low number of studies within a European setting. Additionally, one could argue that the studies covering the Finnish context are especially applicable to the Swedish context due to cultural similarities, giving some balance to the cultural differences of our chosen research.

4. Research Review

This chapter provides an account of the research review. It is divided into four major sections: two for views on pronunciation, one for the effects of pronunciation on comprehensibility and intelligibility, and one for the effects of pronunciation instruction.

4.1 Teachers’ Views on Pronunciation

In this section, in-service and pre-service teachers’ (henceforth called PSTs) views on pronunciation are divided according to the major themes found in the gathered research. Initially, the section accounts for general views held by teachers. Then the preferred model for pronunciation instruction is reviewed, that is, whether the teachers lean towards a native-speaker model or an intelligibility model. After this, the question of teaching pronunciation explicitly or implicitly is brought up and followed by a presentation of some teachers’ reluctance to teach pronunciation. Finally, a section devoted to teachers’ views on their teacher training concludes this section.

(16)

12 4.1.1 General Views on Pronunciation

To account for teachers’ general views on pronunciation, we used four studies (Bai & Yuan 2019; Tergujeff 2012; Üstünbaş 2018; Uzun & Ay 2018) generally indicating that teachers consider pronunciation to be an important aspect of English teaching.

In Bai and Yuan’s study, 16 teachers took an open-ended survey and expressed their views on pronunciation teaching, and four of those teachers were chosen for further interviews. The teachers interviewed generally have a positive view towards teaching pronunciation (2019: 137), the reason being that they see benefits such as enhancing students’ communication, increasing self-confidence and motivation, as well as benefiting other elements of language acquisition (ibid.).A comparable tendency is shown in Tergujeff’s study from 2012. The study is a part of her dissertation from 2013 and is based on a survey answered by 103 teachers active in different parts of the Finnish school system. Tergujeff highlights that the teachers view pronunciation teaching as an important aspect of language acquisition when compared to other language skills (2012: 34). However, she also notes that teachers report not devoting as much time to pronunciation teaching as they would like to (ibid.). Üstünbaş (2018) gives a similar view in his study, investigating among other things how teachers consider

pronunciation in language learning. Analyzing the 5-point Likert scale used in the study, he comes to the conclusion that the teachers think that pronunciation is an important part of language teaching (ibid.: 81). Finally, the study by Uzun and Ay (2018), investigating the views of 147 PSTs at a university in Turkey, concludes that more than 90% of them consider pronunciation an integral part of learning English (ibid.: 123-124).

4.1.2 Teaching Model

Going into teachers’ preferred pronunciation model, findings generally suggest that teachers seem to submit to the Nativeness Principle, indicating native-like pronunciation as the main aim of instruction. Five studies shed light on the matter (Moradkhani & Asakareh 2018; Tergujeff 2012; Çeçen & Serdar Tülüce 2019; Karakas 2017; Uzun & Ay 2018).

Moradkhani & Asakareh’s (2018) study investigates teachers’ perceptions onthe teaching model in private and public schools in Turkey. 106 teachers participated in a questionnaire, which was followed by interviewing six teachers for further analysis. Generally, the teachers seem to consider the accents spoken in native-speaking countries as a pedagogical model for

(17)

13 pronunciation (ibid.: 9). A similar tendency is shown by Tergujeff (2012). In her quantitative study analyzing teachers’ preferences of pronunciation model, she states that the

“pronunciation model is traditional” (ibid.: 41) in that teachers use RP and GA to a greater extent than “a type of international English” (ibid.: 40), even though 42.1 % of the teachers reported also using an international English to some extent (ibid.). The term “international English” is, however, not further specified by the author.

Çeçen and Serdar Tülüce’s (2019) study investigates pre-service EFL teachers’ attitudes toward accent. The researchers analyzed dyadic conversations between 12 Turkish pre-service EFL teachers in order to elicit data on what their attitudes are toward native and non-native speakers, and what the reasons are for these views. The results of the study indicate that the participants tend to use the native accents as a means of comparison; that is, their attitudes toward speakers with accented English are shaped in light of the native accents (ibid.: 130). The PSTs thus attribute lower status to speakers with accents and higher status to native speakers of English (ibid.).

A similar study was carried out by Karakaş (2017), who investigated PSTs view on “good English” by analysing an open-ended questionnaire filled out by 42 participants (487). An analysis of the participants' answers conclude that the vast majority of the participants fall into the category “normative perceptions of good English” (ibid.: 494), meaning that they associate good English with “correctness” and “nativeness” (ibid.). On the same topic, Uzun and Ay (2018) show that the PSTs in their study tend to think that a native-like accent is the ultimate goal (124), while only 38.1% of the participants feel that intelligible pronunciation would suffice for their learners (125).

4.1.3 Explicit or Implicit Pronunciation Teaching

The following section addresses teachers’ views on teaching pronunciation implicitly or explicitly, that is, whether they devote time especially to pronunciation instruction by for instance teaching the IPA or if they prefer to incorporate pronunciation instruction into other elements of English teaching. The studies included in this section reveal varying views among teachers on the topic.

Üstünbaş (2018) finds that the teachers in his study prefer to teach pronunciation integrated into other skills (82), even when it comes to assessment (79). However, Uzun and Ay (2018)

(18)

14 present a different view. Their results indicate that the PSTs are unsure of whether

pronunciation should be taught explicitly or implicitly (ibid.: 124), correlating with the findings which indicate that the PSTs’ generally require more instruction on pronunciation teaching (ibid.).

One aspect of teaching pronunciation is making use of the IPA, which we consider as explicit pronunciation teaching. The teachers in Tergujeff’s (2012) and Bai & Yuan’s (2019) studies are divided in this regard. In Bai and Yuan’s study, most teachers emphasize the use of this tool, stating that the IPA enables students to develop their pronunciation on their own (ibid.: 138). Similarly, Tergujeff (2012) shows that 72.8% of the teachers in her study report that they teach how to recognize phonetic symbols (Tergujeff 2012: 38). However, Bai and Yuan (2019) mention that some teachers are concerned that the IPA might be too difficult and thus demotivating (138), and Tergujeff (2012) shows that many teachers feel that teaching the IPA is unsuitable for young learners (39).

4.1.4 Teacher Training

Concerning teacher training, findings indicate that teachers feel that they have received insufficient training on pronunciation instruction, thus making them unsure of how to teach it to their students. One example of such a study was conducted by Uzun and Ay (2018), who show that 35.4% of the PSTs feel that they do not know how to teach pronunciation, and that 41.5% were “undecided” (124). This is in line with Tergujeff’s (2012) study, where the teachers indicate that they received proper instruction on pronunciation themselves, but an insufficient amount of training on how to teach it to their students (34). Additionally,

Macdonald (2002) shows that the teachers in his study receive inadequate training in how to incorporate pronunciation into their communicative approach, causing them to instead teach pronunciation ad hoc, whenever issues arise (10). Finally, comparable results can be found in a study by Bai and Yuan (2019), who argue that the teachers receive insufficient instruction on pronunciation teaching, which leads to “limited professional competence” and,

subsequently, low efficacy as regards teaching pronunciation (140).

4.1.5 Reluctance to Teach

One aspect that Bai and Yuan (2019) and Macdonald (2002) have examined is what the reasons might be for some teachers’ reluctance to teach pronunciation to their students. Bai and Yuan highlight that while the teachers in their study seem to hold pronunciation teaching

(19)

15 in high regard, many of them are reluctant to teach it because of their own accent (2019: 139). The non-native teachers report that they are unsure of their own pronunciation and therefore leave the pronunciation and speaking classes to teachers that are native speakers of English (ibid.). Furthermore, the responsibility for pronunciation teaching in Hong Kong is often placed upon the native-speaking teachers, while non-native-speaking teachers’ attempts to teach pronunciation are often suppressed (ibid.). According to Bai and Yuan, this reduces the legitimacy of the teachers, and affects their efficacy as regards pronunciation instruction (ibid.).

Other possible reasons for teachers’ reluctance to teach pronunciation are highlighted by Macdonald (2002). One example is that in the Australian context, there is no particular demand to teach pronunciation specifically from the national curricula, resulting in pronunciation teaching being “easy to overlook” (ibid.: 6). Additionally, the lack of a framework for the teaching of pronunciation affects the availability of teaching resources (ibid.: 7), which is another possible reason for teachers’ reluctance to teach pronunciation (ibid.: 11). Furthermore, Macdonald’s findings suggest that the teachers find it difficult to assess students’ pronunciation (ibid.: 7). Rather than having clear criteria for how to assess pronunciation, instruction is introduced in the classroom when the intelligibility of students is insufficient (ibid.). This lack of assessment guidance is yet another reason for the reluctance to teach pronunciation. To add to this, teachers in the study report that they are unsure of how to give feedback on pronunciation in the classroom, as they feel that it might cause

“embarrassment and discomfort” (ibid.: 9) and interrupt students’ “flow of speech” (ibid.).

4.2 Learners’ Views on Pronunciation

This section provides an account of learners’ views on pronunciation. Divided into three parts, the section presents results concerning general views on pronunciation, including pronunciation instruction, followed by findings on the preferred model of pronunciation. In the final section, attitudes towards accented speech are presented.

4.2.1 General Views on Pronunciation and Pronunciation Instruction

Four studies were found that investigate learners’ views on pronunciation and pronunciation instruction: Kang (2015), Jarosz (2019), Üstünbaş (2018) and Tergujeff (2013). The studies all look into learners’ views on pronunciation, but with different foci; the results from Kang

(20)

16 (2015) reveal general views held by learners on the importance of pronunciation for their language development. Jarosz (2019), Üstünbaş (2018) and Tergujeff (2013), on the other hand, investigate learners’ views on pronunciation instruction in particular, eliciting what role they believe instruction plays for the development and possible enhancement of

pronunciation.

The study by Kang (2015) investigates adult learners’ views on pronunciation in six different countries by collecting responses to a questionnaire. Although she investigates the differences in learners’ responses between the countries, the results covered in our research review applies to all the learners. Thus, a dissection of the results pertaining to the differences between contexts is left for further research. In terms of their views on pronunciation in general, learners from all six countries indicate that they consider pronunciation an important aspect in communication, with mean scores ranging from 4.59 to 5.40 on a 6-point scale (ibid.: 69). Similarly, the learners express concerns about their pronunciation (M = 3.93-5.14) and that they want to improve it (M = 4.61-5.31) (ibid.). Additionally, the learners recognize that improved pronunciation could lead to increased confidence in English, with mean scores ranging from 4.81 to 5.29 on the same scale (ibid.). It should be mentioned that learners from countries where English is the main language consistently supply the highest mean score values on these topics (ibid.).

Learners’ positive views on the role of pronunciation are also presented in Jarosz’s (2019) study, which covers pronunciation instruction specifically. She investigates 10 Polish

secondary-school learners’ views on pronunciation instruction through a set of questionnaires given both before and after a 30-week pronunciation course. The results from the post-course questionnaire indicate that the learners generally believe that teaching pronunciation is important for its development and that good pronunciation strengthens communication (ibid.: 301). In particular, the learners emphasize phonetic training and deem it relevant for their language learning (ibid.). In her discussion, Jarosz further emphasizes the learners’ acquired speaking efficacy through instruction, and argues that this, in turn, “strengthens the argument that pronunciation must be taught” (ibid.: 301). Similarly, Üstünbaş’ (2018) study

investigating learners’ views on pronunciation instruction concludes that they consider the teaching of pronunciation to be important. Furthermore, she investigates whether learners prefer an explicit or implicit approach to pronunciation instruction. The study reveals that learners mainly want to learn pronunciation implicitly, that is, they prefer to have

(21)

17 pronunciation instruction integrated into the teaching of other language skills (ibid.: 79).

A qualitative study on the same topic that provides slightly different results was written by Tergujeff (2013), as a part of her dissertation from the same year. This study investigates views on pronunciation teaching held by learners of English attending primary, lower secondary and upper secondary school in Finland. The study is based on interviews of 10 learners ranging from the age of 10 to 18 (ibid.: 81). Even though the learners in the study do not clearly express their wish for explicit pronunciation teaching, the learners attending lower and upper secondary levels express that they feel that the time devoted to pronunciation teaching is insufficient (ibid.: 89). However, the primary school learners give a different view, indicating that they are satisfied with the amount of pronunciation teaching provided (ibid.). Tergujeff explains this by stating that Finnish schools put more emphasis on oral skills in lower-level English classes (ibid.). It is worth mentioning that the learners of the study do not count the implicit pronunciation instruction integrated into communicative tasks as actual pronunciation teaching (ibid.: 86), which might indicate that what the lower and upper-secondary school learners are asking for is indeed explicit pronunciation instruction. In that case, Tergujeff’s study shows somewhat different results than those of Üstünbaş’ (2018).

Another aspect of explicit or implicit teaching is the use of the IPA in the classroom. In this regard, the views of learners in Tergujeff’s study (2013) are varied. Teaching pronunciation using the IPA occurs mostly in primary school (ibid.: 87) and when asked if learners thought it was useful to learn the phonetic symbols, some state that learning the IPA was useful, while others feel that it is unnecessary, stating that the system interfered with regular spelling (ibid.: 88).

4.2.2 Learners’ Pronunciation Model

The following section covers the learners’ pronunciation model, that is, what their goals of learning pronunciation are. Furthermore, findings are presented on what the possible reasons are for these goals. A majority of the learners observed in the studies seem to submit to a native model, while some submit to a non-native model, correlating with the nativeness and intelligibility principles respectively. Five studies were found that provide insight into this topic. Wach (2011), Kang (2015) and Jarosz (2019) all found evidence that point towards a nativeness principle. Similar findings were made by Kung and Wang (2019), although they found evidence of the intelligibility principle as well. Finally, Tergujeff (2013) found

(22)

18 evidence which indicates that the learners in her study did not seem to submit to a native model.

In her study from 2011, Wach investigates 234 Polish English majors’ views on

pronunciation, especially considering native and lingua franca norms. Generally, the findings reveal that the learners hold positive views on native-speaker norms (ibid.: 260). Some learners, however, express positive views on lingua franca norms, for instance by highlighting that they could be useful for communicating with other non-native speakers (ibid.: 256). Still, most learners are either unaware of the concept of lingua franca norms, or express negative views about them, for example that they are insufficient as a target for learners or that they are a threat for the value of native accents (ibid.). Finally, the respondents express that they believe that native-speaker pronunciation norms should be incorporated into instruction from the start, and that teachers should have native-like pronunciation (ibid.: 259). In line with these results, the learners in Kang’s (2015) study emphasize their will to sound like a native speaker of English, with mean scores from 4.42 to 5.20 on a 6-point scale (69). Moreover, the findings in Jarosz’s (2019) study are comparable to the above-mentioned results. In her study, the learners seem to emphasize native-like proficiency over intelligibility in the pre-course questionnaire; four learners indicate that they aim for native-like pronunciation, while only one learner emphasizes intelligibility as their aim (ibid.: 299). Additionally, the learners point out correct pronunciation as a pivotal aspect of speaking English, and that phonetics is

relevant in language learning (ibid.: 300-301).

Similar results are provided by Kung and Wang (2019). In their qualitative study, 34 undergraduate students studying at a university in southeast China are interviewed and respond to an open-ended questionnaire about their views on accents for effective ELF communication (ibid.: 397). Findings reveal that 76% (25 out of 33) show a preference for speaking with a native-like accent, while 24% do not (ibid.: 398). Going into the reasons for these views, the authors elicit three major themes why learners wish to sound native. One reason is cultural and material influences from native-speaking cultures (ibid.: 399). Other reasons are the influences of teachers and peers, further enhancing their tendency to strive for a native accent (ibid.), as well as the university learning context in which teachers tend to assess pronunciation to a greater extent than teachers in earlier school contexts did (ibid.: 400).

(23)

19 Going into reasons for not wanting to sound native, that is, striving for being intelligible rather than native-sounding, Kung and Wang (2019) report that the minority of students who did not pursue sounding native did so for two major reasons. One is that international

interlocutors communicating with the participants had stated that the participants were intelligible, removing the need for sounding native (ibid.:402). The other major reason was that the participants mainly communicated with other Chinese speakers of English, thus reducing the necessity to sound native, because the Chinese accent is mutually intelligible (ibid.: 402). The intelligibility view of this minority of students is in line with the findings of Tergujeff’s (2013) study. In her interview with Finnish learners of English, learners generally state that they do not have a goal to sound native-like (ibid.: 88), a statement that is in contrast to the findings of the studies mentioned above. Moreover, students report not aiming to sound native-like despite acknowledging that there is a clear native-speaking model in their schools (ibid.).

4.2.3 Negative Attitudes Toward Accented Speech

Relating to the findings on pronunciation models in the previous section, the following section reveals what learners’ attitudes are toward accented speech and possible reasons for these attitudes. Two studies are included in this section: Ito (2019) and Kung and Wang (2019). Both studies investigate EFL learners’ attitudes and highlight findings that reveal negative attitudes towards non-native speakers’ accents.

The study conducted by Ito (2019) examines learners’ attitudes toward accent, and

specifically how accent affects Japanese EFL learners’ perceptions of native and non-native speakers’ speech (65). Data was gathered by having 78 Japanese EFL learners listen to

controlled speech samples and judge them as grammatical or ungrammatical (ibid.: 65-66). Ito found evidence indicating that the learners tended to judge non-native speech samples as ungrammatical more often than they did native speech samples (ibid.: 68). Additionally, native speech samples were more often judged as grammatical than non-native speech samples (ibid.). Also, speech samples from non-native speakers that were deemed to have “mild” accents were more likely to be interpreted as grammatical than those from non-native speakers with “heavy” accents (ibid.: 69). According to Ito, these results imply that the accents of the speakers have some effects on the learners’ perceptions of them (ibid.: 76). In line with Ito’s findings, Kung and Wang (2019) find that there are several negative

(24)

20 reports that having an accent gives the impression of not “work[ing] hard” (ibid.: 401), and that it is “uncomfortable” speaking to that person (ibid.). To add to this, learners report that having a native-sounding accent gives the impression of sounding “professional” and that it will provide more “opportunities” (ibid.).

4.3 The Effects of Pronunciation on Comprehensibility and Intelligibility

In this section, we provide research concerning the effects that pronunciation has on comprehensibility and intelligibility, establishing why teaching pronunciation is important, and why investigating the effects of pronunciation instruction is relevant. Three studies were found that investigate how word stress affects comprehensibility and intelligibility (Lewis & Deterding 2018; Hahn 2004; Field 2005) while one study (Kagitci Yildiz 2017) investigates the effects of accented speech. Generally, findings suggest that both word stress and accent have an effect on comprehensibility and intelligibility.

In their qualitative study from 2018, Lewis and Deterding analyze two corpora of data which consist of interactions between non-native speakers of English. Their main aim was to discern what impact word stress has on intelligibility. However, several instances of

misunderstanding due to segmental aspects of English were found, such as the confusion of /l/ and /r/ (ibid.: 168). Despite that Lewis and Deterding found only six instances out of 183 tokens where word stress had an impact on intelligibility, they draw the conclusion that “unexpected or unclear stress can sometimes cause problems for intelligibility in ELF contexts” (ibid.). More commonly, the study found several instances of miscommunication due to the loss of the second consonant in consonant clusters such as /kl/ and /pr/ (ibid.), resulting in misinterpreting, for instance, the word “clouds” as “cow” (ibid.: 171).

Interestingly, the study also found one example of where the “American flap for [...] /t/ in

bottle” (ibid.) was used and communication was disrupted. Although there was only one

example of the error, this could possibly give some indications that attempting to sound native might be a contributing factor to communication breakdown.

Another study which investigates the effects of pronunciation on comprehension was

conducted by Hahn (2004). One aim of her study was to determine whether the misplacement or omission of primary stress affects native-speaker listeners’ comprehension of spoken English (ibid.: 205). 90 native-speaker students at a university in the USA were asked to listen to scripted recordings from the same non-native speaker, where primary stress was placed a)

(25)

21 correctly, b) incorrectly, or c) missing entirely (ibid.: 206-207). The results of the study

indicate that it is more difficult for native speakers to comprehend non-native speech where primary stress is misplaced or missing, as opposed to speech where primary stress is placed correctly (ibid.: 210). For example, the listeners were able to remember a greater number of ideas from the utterances when primary stress was placed correctly (ibid.). This also applies to what the author calls “main ideas”, which were also remembered to a greater extent in

utterances with correct stress placement (ibid.). Thus, the findings of Hahn’s study suggest that comprehension is enhanced by correct placement of primary stress (ibid.: 215). In her conclusion, Hahn argues that her research strengthens the argument that instruction on suprasegmentals is an important part of English teaching (ibid.: 218).

Similar findings are reported by Field (2005), where he investigates native and non-native listeners’ comprehension of utterances with non-standard stress patterns. The aim of the study was to examine whether incorrect stress placement affects comprehension, and if so, what kinds of errors affect comprehension the most. In line with Hahn (2004), Field (2005) finds that incorrect stress placement negatively affects listeners’ comprehension (410).

Furthermore, Field’s findings suggest that native and non-native listeners’ comprehension seems to be affected to the same extent, indicating that comprehension is affected more by the nature of the error rather than the listeners’ mother tongue (ibid.: 413).

On the same topic, Kagitci Yildiz (2017) conducted a study to investigate whether accent had an effect on students’ performance on listening comprehension tests, and how the students in the study perceive non-native accents. 107 students with an intermediate level of English were selected for the study, and they listened to recordings of different topics by six teachers of English, three of whom were non-native speakers (ibid.: 111-112). To elicit whether accent had an effect on comprehension, the students took a multiple-choice comprehension test. After the test, a questionnaire based on a 5-point Likert scale was handed out to collect students’ opinions of the different accents. The results show that accent played a significant role in the comprehension of some texts (ibid.: 116). However, the author notes that “it is difficult to say that there is a direct relationship between certain speaker accents and students’ listening comprehension” (ibid.: 118). Concerning the opinions of the students, 83.2%

(26)

22

4.4 The Effects of Pronunciation Instruction

Having investigated the importance of pronunciation with regard to comprehensibility and intelligibility, we now present findings from studies that examine whether or not instruction on pronunciation has an effect on its development. This section is divided into three parts, the first two investigating the effects of segmental instruction in controlled and spontaneous speech, while the third part investigates the effect of suprasegmental instruction in both contexts.

4.4.1 The Effects of Segmental Instruction on Controlled Speech

This section accounts for the effects of segmental instruction on controlled speech. Segmental instruction refers to instruction that focuses on specific phonetic features of speech, for instance epenthesis and absence (Couper 2006: 48) or pronouncing the alveolar approximant /ɹ/ (Saito & Lyster 2012). Controlled speech, in turn, refers to being able to pronounce the target sound or sounds when reading a set of pre-written phrases or sentences. The articles used in this section (Couper 2006; Kayaoğlu & Çaylak 2013; Saito & Lyster 2012; Derwing, Munro & Wiebe 1998; Saito 2011) generally suggest that segmental instruction has a positive effect on controlled speech.

Couper (2006) studies the effects of pronunciation instruction on 21 immigrant learners resident in New Zealand (49). The study examines the improvement of controlled speech by performing pre- and post-tests, as well as a delayed test to see differences over time. Couper’s study focused solely on segmentals, omitting suprasegmentals due to the difficulty of

quantifying those kinds of errors (ibid.: 50). Upon examining the error rates between the pre-test and the post-pre-test, all learners made “substantially fewer errors after instruction” (ibid.: 55), and even though the error rate slightly increased in the delayed test, the number of errors was still significantly lower than it was in the pre-test (ibid.).

Kayaoğlu and Çaylak (2013) show comparable results in their study on the effects of controlled speech, where they examine the “audio articulation method” to see if it increases the tendency to correctly pronounce the dental fricatives /θ/ and /ð/ (ibid.: 273). The “audio articulation method”, in essence, is constructed to remedy specific segmental pronunciation problems over the course of one single class (ibid.: 272). The study includes 36 elementary-level students aged 19-26, studying at a Turkish vocational school. It utilizes a pre- and post-test along with a delayed post-test to determine the participants’ overall pronunciation

(27)

23 development. Moreover, a control group was used for comparison to further enhance the reliability of the study. Even though only half of the participants managed to produce the target sounds in the post-test (ibid.: 275), this was still a significant increase in producing the target sounds when compared to the control group (ibid.: 274). To add to this, the participants showed similar progress in the delayed test, even though a slight decrease could be observed (ibid.: 297).

A comparable study, conducted by Saito & Lyster (2012), shows somewhat similar results investigating whether instruction and corrective feedback develop the pronunciation of the voiced alveolar approximant /ɹ/ by Japanese learners of English. Using a pre-test/post-test design, Saito and Lyster use three sets of tests to verify improvement, two of which measure controlled speech by looking at word-reading and sentence-reading (ibid.: 610). To elicit whether speakers improved or not, acoustic analyses informed by human rating were used. In line with the studies mentioned above, segmental instruction had a significant effect on improving the target phoneme. However, the study suggests that this instruction was only effective in conjunction with corrective feedback (ibid.: 622). Thus, segmental instruction without corrective feedback was insufficient in developing the pronunciation of the voiced alveolar approximant (ibid.). Derwing et al. (1998) investigate 48 intermediate students in an ESL program in Canada. The study differs from the above-mentioned studies in that it measures the extent to which the participants increased their comprehensibility and

accentedness in controlled speech (ibid.: 401). The participants in the study were divided into three different groups, one of which received instruction with a segmental focus (ibid.: 396). The study shows that receiving segmental instruction significantly improves both

comprehensibility and accentedness in controlled speech (ibid.: 402-403). However, Saito’s (2011) study, investigating the effect of segmental instruction on comprehensibility and accentedness, presents different results. His study partially looked at the controlled speech of 20 intermediate learners of English, comparing them to a control group of uninstructed students, by utilizing a sentence-reading task. A human-rating method was used to elicit the progress of the participants. In contrast to Derwing et. al. (1998), Saito’s results show that segmental training has no significant effect on accentedness (2011: 53). However, Saito’s findings are in line with Derwing et. al. (1998) in the sense that they show statistically significant evidence of improved comprehensibility (Saito 2011: 54).

(28)

24 4.4.2 The Effects of Segmental Instruction on Spontaneous Speech

Five studies were found that investigate the effects of segmental instruction on spontaneous, unscripted speech. The studies investigating these effects (Saito & Lyster 2012; Kayaoğlu & Çaylak 2013; Derwing, Munro & Wiebe 1998; Saito 2011; Gordon and Darcy 2019) show that segmental instruction does not have a significant effect on spontaneous speech, even though there are some contrasting results.

As mentioned in section 4.4.1, the learners in Kayaoğlu and Çaylak’s (2013) study showed improvement in controlled speech in the post-test and delayed test. However, when the researchers observed the learners in classroom activities during the period between the post-test and the delayed post-test, learners seemed to fall back to incorrect pronunciation of the target sounds (ibid.: 282). These results indicate that the learners are able to use the acquired sounds in controlled speech but have difficulties applying them to spontaneous speech. Similar findings are presented by Derwing et al. (1998), where the group that received segmental instruction did not improve their spontaneous speech in comprehensibility, accentedness, or fluency (405). According to the authors, the reason for this lack of development could be that spontaneous speech moves speakers’ attention to other linguistic aspects, thus inhibiting their capacity to “allocate enough resources to phonological concerns” (ibid.: 406).On the same topic, Saito (2011) shows that the pronunciation of the learners in his study did not improve as regards accentedness in spontaneous speech (53). This result is in line with the results on controlled speech (see section 4.4.1). However, the results on controlled and spontaneous speech differ regarding comprehensibility. While the comprehensibility of the learners improved in controlled speech, Saito found no statistical significance indicating that their spontaneous speech improved, even though some learners “showed remarkable improvement” (ibid.: 54). According to Saito, the lack of significance could be due to great variance among the participants (ibid.).

One study that contrasts somewhat with the results of the research covered above was

conducted by Saito & Lyster (2012). They present results that seem to indicate that segmental training with corrective feedback improves speakers’ production of the alveolar approximant /ɹ/ in spontaneous speech (ibid.: 617). However, it is important to note that the authors acknowledge that the task they used “might not be the most appropriate way to tap learners’ spontaneous speech” (ibid.: 623). The reason for their skepticism is that the participants were provided with written words including the target sound /ɹ/ that they could read when

(29)

25 producing their so-called spontaneous speech (ibid.). Thus, it could be argued that the

spontaneous speech samples include some aspects of controlled speech.

Gordon and Darcy’s (2019) experimental study presents results that stand out among the other studies’ results when it comes to spontaneous speech. The study investigates the

pronunciation development of EFL learners at a university in Costa Rica. The aim was to find out whether the EFL learners improved in comprehensibility after 10 weeks of instruction, and whether segmental instruction, suprasegmental instruction, or a mix of both was most effective. While mixed instruction yielded insignificant improvement in comprehensibility over time, segmental instruction turned out to decrease the comprehensibility of the learners, thus indicating that segmental instruction seems to harm rather than aid pronunciation development (ibid.: 121).

4.4.3 The Effects of Suprasegmental Instruction

Three studies were found that investigate the effects of suprasegmental instruction on

pronunciation development: Ghorbani (2019), Derwing, Munro and Wiebe (1998) and finally Gordon and Darcy (2019). While the low number of studies found for this section is not ideal, it is a result of the scarcity of research on the matter. In fact, only about 19% of the studies that Saito & Plonsky (2019) examined in their meta study on pronunciation research investigate suprasegmental features only (676). In contrast, more than 70% of the studies focus solely on segmentals (ibid.), thus indicating an emphasis on the latter in pronunciation research. As opposed to the previous sections, this section includes studies investigating both controlled and spontaneous speech mainly due to the low total number of studies.

One study that examines how suprasegmental instruction improves controlled speech was conducted by Ghorbani (2019). In his quasi-experimental study, 34 undergraduate students at an Iranian university participated in a course based on “The Flatmates series from the BBC’s learning English Website” (ibid.: 404). The participants were grouped into one control group and one experimental group, where the experimental group received explicit instruction on the phonological sound system of English. A pre-test was constructed, where the participants received 60 lexical items and were asked to mark where the primary stress should be (ibid.). After instruction, the participants received the test again as a post-test and were asked to mark the same lexical items and then pronounce them with the help of their written test. The results indicate that the explicit suprasegmental instruction had a significant impact on the learners’

(30)

26 word stressing (ibid.: 406). It should be mentioned that the study does not mention in what manner the participants were assessed. Also, the findings of the study are hard to generalize due to the small number of participants, according to the author (ibid.: 407).

Other aspects of the effects of suprasegmental instruction are presented in the study conducted by Derwing et al. (1998). The results of the study show that suprasegmental instruction improves controlled speech to some extent, for instance that the participants who received suprasegmental instruction significantly improved their comprehensibility when compared to the control group (ibid.: 402). However, when analyzing the results on

accentedness, Derwing et al. found that while the suprasegmental group improved, so did the control group (ibid.: 403). Thus, the results do not indicate that suprasegmental instruction has any significant effects on accentedness in controlled speech.

Derwing et al. (1998) also provide relevant findings on the effects on spontaneous speech. The participants who received suprasegmental instruction showed significant improvement in comprehensibility and fluency in spontaneous speech, as opposed to both the control group and the segmental group, who did not show any significant improvement (ibid.: 405).

Regarding accentedness, however, none of the groups showed any improvement (ibid.). It can thus be concluded that while suprasegmental instruction does not seem to have a significant effect on accentedness, it clearly improves comprehensibility, and more so than segmental instruction (ibid.). Another study that investigates the effects of suprasegmental instruction on spontaneous speech was conducted by Gordon and Darcy (2019). The authors investigate how comprehensibility is affected by instruction of three different kinds; segmental,

suprasegmental and a mix of the two. The findings show that suprasegmental instruction significantly improved the participants’ comprehensibility, while the mixed instruction did not yield any significant improvement (ibid.: 120-121). Given that the instruction took place over 10 weeks, Gordon and Darcy argue that the results of the study “suggest that a treatment based on suprasegmentals seems most effective in a short period of time” (ibid.: 122).

However, it should be mentioned that the authors acknowledge that the lack of significance for the mixed group could be due to the brevity of the instruction (ibid.). Gordon and Darcy argue that it is possible that 300 minutes of instruction over 10 weeks was not enough for the implementation of both segmental and suprasegmental aspects (ibid.).

The Importance of Pronunciation Instruction in the English as a Foreign Language Classroom