Contents
0. Abstract
31. Introduction
41.1 Thesis outline 5
2.
Background
72.1 The development of English as a lingua franca 7
2.2 Nativeness, pronunciation norms and language ownership 8 2.3 Intelligibility vs. native-likeness in English pronunciation pedagogy. 10 2.4 English as a Lingua Franca and the and the Lingua Franca Core for 11
pronunciation pedagogy.
2.5 Judgements of Dutch-accented English. 13
2.6 Goals, motivation and attitudes in pronunciation training. 15
2.7 Statement of purpose 17
3. Method
183.1 Participants 18
3.2 Materials 19
3.3 Procedures 21
3.4 Design and analysis 21
4. Results
225. Discussion
305.1 Motivational profiles of the clusters 30
5.2 Motivation and its implications for English pronunciation pedagogy 32
6. Conclusion
35References 37
0. Abstract
In the Netherlands, current pronunciation teaching and methods still focus on the traditional English pronunciation norm of British English. However, as English has become the international lingua franca, language ownership and the authority of traditional pronunciation norms have become major topics of debate. Jenkins (2000) developed a pronunciation model, the Lingua Franca Core (LFC), which is based on a combination of English varieties and aims to promote intelligibility among non-native English speakers. However, the Socio-educational model (Gardner, 1985) and the L2 motivational self-system (Dörnyei, 2005) explain that motivation and goals of language learners influence the language learning process and its outcomes.
1. Introduction
The English language has rapidly become the lingua franca in international communication, increasing the demand for English speakers and ESL or EFL education. As a result, the number of L2 or L3 English speakers is larger than the amount of native English speakers and the language has a plethora of accent varieties that are non-native forms of English which thus diverge from pronunciation norms (Jenkins, 2000). However, this development has not affected the goal of English pronunciation training in educational environments, as teaching methodologies have not adapted to the increase in pronunciation varieties. In the Netherlands, the aim of pronunciation training traditionally entails acquiring a native-like English accent, preferably RP, but in some cases General American (GA) is also allowed.
However, the possibility for L2 English learners to attain a native-like English accent has long been a topic of debate among linguists. The landmark theory of the Critical Period Hypothesis (henceforth CPH) posited by Lenneberg (1967) described that the onset of L2 acquisition severely impacts the degree of L1 interference in L2 pronunciation (Lenneberg, 1967). Further studies proposed that when L2 acquisition starts after the age of 12, high standards can be achieved, yet native-like L2 accent attainment is highly unlikely (Long, 1990). However, the CPH has remained controversial, as other studies have indicated that with major L2 exposure, often due to immigration, native-like L2 pronunciation is possible post-puberty.
The difficulty in native-like L2 accent attainment can be explained by Equivalence Classification theory (Flege, 1987). This theory describes the entrenchment of articulatory and auditory settings in L1 acquisition which impedes the ability of older L2 learners to discern and produce L2 phonemes that are different, yet similar, to phonemes in their L1 phonemic inventory.
Although such a perspective on pronunciation training seems just and egalitarian, it is important to consider the attitudes native and non-native speakers have towards pronunciation norms. Attitudes to accents influence pronunciation training and might affect the reception of Jenkins’ (2000) model in the classroom environment. It is important to consider that accent stereotyping influences the general public’s view of correct pronunciation. Language learners might be concerned with native speakers’ attitudes towards foreign-accented English. More importantly, they might judge foreign-accented English differently and prefer to aim for native-likeness in pronunciation training. As it remains unclear how ESL learners view the roles of native-likeness and intelligibility in pronunciation training, this thesis aims to provide insights into the goals, motivation, efforts and performances of Dutch ESL learners in pronunciation training and investigates how the motivations and goals relate to current pronunciation pedagogy and Jenkins’ (2000) pronunciation model.
The following research questions were posed to achieve this goal:
1. Do Dutch high school pupils aim to acquire an L1 English variety in pronunciation? 2. What are the motivational profiles of Dutch high school pupils in pronunciation
training?
3. Would Dutch high school students’ motivation increase with implementation of the Lingua Franca Core?
1.1 Thesis outline
In the literature review, the transformation of English as a foreign language to an international language is briefly discussed in chapter 2.1. It will become clear how native speakers were granted authority over the English language and how native varieties were established as pronunciation norms. The need for a revision of the authority of native varieties and the prescription of prestige, yet minor varieties in pronunciation training will become apparent in chapter 2.2. In chapter 2.3, the role of intelligibility versus native-likeness in pronunciation pedagogy will be discussed and it will become clear that a change in pronunciation norms could benefit intelligibility.
and Csizér and Dörnyei (2005) are explained. These constructs form the basis of the questionnaire and the focus of this study, for if English teachers in the Netherlands were to adopt Jenkins’ (2000) LFC model in pronunciation teaching, the personal and individual goals of Dutch ESL learners need to be taken into consideration. In the method section, the motivational variables an the analyses that were chosen for this study will be presented.
2. Background
2.1 The development of English as a lingua franca.
In order to understand the demand for a revaluation of the current predominating English pronunciation norms, it is important to study the changing functions of English through time and the development of its status as a Lingua Franca which creates the demand for a revision in English pronunciation pedagogy. Prior to the expansion of the British Empire and the colonisation of the Americas, the authority of native speaker pronunciation norms seemed very logical. As the purpose of English language acquisition was to communicate with native English speakers, the pedagogy aimed at meeting the set standards of the native speakers and the attainment of a native-like accent (Jenkins, 2000). Since the goal of language learning was interaction with native speakers, the latter were granted ownership of the English language and the authority to prescribe pronunciation standards.
However, the economic and political developments on a global scale have led to a change in the functions of English (Jenkins, 2000). The imperialism of the British empire caused the English language to spread globally. However, the decolonisation of nations that were under British influence caused the function of English to change. Firstly, in countries like India, the process of decolonisation did not lead to reimplementation of L1s other than English in the public domain (Jenkins, 2000). Instead, local L1s influenced and nurtured novel English pronunciation varieties (Jenkins, 2000). Consequently, the debate about maintenance of intelligibility and standard pronunciation norms on the one hand and tolerance towards new pronunciation varieties arose within these nations (Jenkins, 2000).
Furthermore, the American domination of the economy in the 20th century in
addition to the prevalence of English in decolonised nations caused the occurrence and influence of the English language to increase (Jenkins, 2000). These developments not only caused debates about language norms within English-speaking nations, but also raised questions about ESL and EFL pedagogy.
Categorisation of the present functions of English might put the current situation in perspective. In order to provide insights into the different functions of English, Kachru (1985) developed a model consisting of three circles, each describing the role of English in a nation and each encircling countries that pertain to the role described in the circle. The inner circle contains nations where English is the first language for most people, like the British Isles, the USA, Australia and New Zealand. The outer circle encompasses nations where English is not necessarily the first language, but it is the language of communication in national education, politics, law and business. Finally, the expanding circle contains the remaining countries where English serves as an international language, necessary in international business, education, communication and politics conducted with foreigners (Melchers & Shaw, 2011).
nativeness, pronunciation norms, authority and language ownership will be discussed and it will become clear why a revision of pronunciation norms is in order.
2.2 Nativeness, pronunciation norms and language ownership.
The authority of standard varieties in prescribing pronunciation norms is often ascribed to nativeness. The increasing influence of English as an International language (henceforth EIL) has sparked debates about the terms native and non-native speaker, the former’s ownership of English and its pronunciation norms and the authority that is given to native English varieties. Which factors determine nativeness and why should certain ‘native’ English varieties be granted the authority to become English pronunciation norms? Inspection of these issues will reveal that a revision of terminology and their values is in order.
So-called native English varieties, originating from inner circle nations and spoken by native speakers, are considered to be adequate norms. The prescription of these norms and the distinction between native and non-native speakers has extensive consequences. Inherent to the NS/NNS division is a value judgement and an unequal distribution of authority and ownership (Jenkins, 2000). The NS/NNS discrepancy maintains the authority of ‘native’ English accents as it contributes to the notion that native speakers are fluent speakers, whereas the non-native speaker group consists of L2 English speakers with a range of fluency levels. Due to the NS/NNS distinction, non-native speakers can never reach native-speaker status, regardless of their proficiency level. As the native speaker group is considered to contain proficient speakers, primarily from inner circle nations, its accent is regarded as the norm and its standards are valued. Diverging from this standard and belonging to the NNS category is not valued. Thus, even highly proficient L2 English speakers remain non-native speakers and this status prevents them from being granted ownership of the language. As a result, non-native varieties are not granted the authority to serve as pronunciation norms. The terms NS and NNS are clearly of a static and categorical nature and do not represent the dynamic reality of English language use and the proficiency of its speakers.
are outer circle countries like Singapore where English is the current language in the public domain and at home. Furthermore, the model does not consider immigrant minorities like Spanish-speaking immigrants in the USA (Yano, 2009). The assignment of certain nations to the other circles is also disputable, since a country like the Netherlands would be assigned to the expanding circle even though it has outer circle qualities as well. English is becoming increasingly more important in the Netherlands, especially in the field of education; the number of bilingual and international secondary schools increases and Dutch universities offer fully English taught programmes and courses. These examples indicate that the categorisation of English by the degree of monolingualism in various domains of nations is inadequate as it disregards the possibility of bilingualism in individual speakers. Thus, although Kachru’s (1985) Circles Model provides insights into the different functions of English within and among nations, its categorical nature and monolingual orientation subscribe to the negative differentiation between native and non-native speakers. It is clear that linguistic nativeness of a speaker cannot be determined by nationality, nor can it be ascribed correctly to nations as it generalises the linguistic diversity of their populations.
By focusing on the linguistic background of individual speakers a more accurate manner of discriminating different English speakers can be employed. Jenkins (2000) proposes distinguishing individual English speakers by monolingualism and bilingualism. L1 speakers of English who do not speak other languages are termed Monolingual English speakers (henceforth MES), whereas L1 or L2 speakers of English who speak two or more languages fluently are considered Bilingual English speakers (henceforth BES). A final, third group of English speakers can be established comprising English speakers who have not acquired a high L2 proficiency level of English. Jenkins (2000) terms this group Non-bilingual English speakers (henceforth NBES).
However, categorisation of speakers according to this model is not without difficulty, since fluency and proficiency levels are not categorical. The proficiency levels of English speakers should be seen as part of a continuum, which complicates the distinction between bilingual and monolingual English speakers. There is no clear point in the continuum that indicates the transition from a monolingual speaker to a bilingual status. Jenkins (2000) explains that the term ‘bilingual’ should not be interpreted as denoting equal competence in two languages. Jenkins (2000) states, “the term should mean that the speaker has attained a specified degree of proficiency in both languages, although in practice probably going well beyond this level in one of them (p.10)”. However, it remains unclear what the specified degree of language proficiency entails.
2.3 Intelligibility versus native-likeness in pronunciation pedagogy.
The rejection of NS/NNS terminology and NS pronunciation norms presented by Jenkins (2000) seems ideal. However, the shift to more egalitarian views of English varieties and their authority affects English pronunciation pedagogy. Although the current pronunciation norms are outdated and unjust, they have provided a clear guideline for teachers and learners in English teaching. Consequently, the rejection of native speaker norms is simultaneously the cause for a demand of a new guideline or framework in English pronunciation pedagogy.
In creating new strategies for pronunciation pedagogy, it is important to emphasise that EIL pronunciation pedagogy should take models rather than norms as its foundation. This viewpoint draws on the notion that there is a discrepancy between the two terms. Dalton and Seidlhofer (1994) state that pronunciation teaching should focus on native speaker models; as a native speaker norm is inherently linked to an unrealistic aim of attaining a native-like accent, whereas the native speaker model is of a more lenient character, since a model functions as a point of reference and a guideline. A model is based on the idea of approximation and thus allows the variability in pronunciation that is apparent in language use, instead of demanding attainment of a standard that is considered to be the epitome of correct pronunciation (Dalton & Seidlhofer, 1994). It is clear that the adoption of native speaker models relates to a more realistic view of pronunciation variability and provides English learners with an achievable goal.
However, the emphasis on native speaker varieties in pronunciation models undermines the previously discussed equal distribution of language ownership and authority among native and non-native English varieties. In order to grant authority to all English varieties, they should all be considered as English pronunciation models. This approach would maintain accent equality as it provides different English speaker communities with the opportunity to set their own standards and norms and models.
pronunciation models are used in pronunciation teaching. In the next chapter, a pronunciation model will be discussed that might offer a solution to the problems of intelligibility and inequality in pronunciation norms.
2.4 English as a Lingua Franca (ELF) and the Lingua Franca Core for
pronunciation pedagogy.
Instead of maintaining a Circles model that generalises the function and users of English to SL and FL, Jenkins (2000) proposes to come to a new definition for occurrences of the language among non-native speakers: English as a Lingua Franca (ELF). This term diverges from the differentiation between native-speakers, SL speakers and FL-speakers that is based on expectations of monolingualism in people. Jenkins states the advantages of establishment of the new term for the use of English (Jenkins 2000, p.11)
ELF emphasises the role of English in communication between speakers from different L1s, i.e. the primary reason for learning English today; it suggests the idea of community as opposed to alienness; it emphasises that people have something in common rather than their differences; it implies that ‘mixing’ languages is acceptable and thus that there is nothing inherently wrong in retaining certain characteristics of the L1, such as accent; finally, the Latin name symbolically removes the ownership of English from the Anglos both to no one and, in effect, to everyone.
This perspective on English language use formed the foundation of Jenkins’ (2000) new pronunciation model. Jenkins (2000) explains that in order to establish a fully realistic approach to pronunciation variability and accommodate language learners with an achievable goal while maintaining phonetic intelligibility, the pronunciation model should not be based on one or two native speaker varieties, but all English varieties.
To safeguard intelligibility Jenkins (2000) developed a phonological core, the Lingua Franca Core (LFC) that forms a solution to the problem of intelligibility among NBESs. Over several years, Jenkins (2000) studied the role of Inter-speaker variation in NBES in order to create the LFC. The transfer of L1 segmental (i.e. phonemic and phonetic) features onto the production and reception of target L2s was found to be the greatest obstacle for intelligibility in NBES communication. This result was in contradiction with other findings that indicate that intelligibility is mainly impeded by suprasegmental (i.e. prosodic) L1 features (Jenkins, 2000).
perceives them as L1 phonemes that are familiar to the learner. A case in point is the common pronunciation error Jenkins (2000) found in a Japanese subject. The Japanese English speaker pronounces /r/ as /l/ since the two phonemes are allophones in the L1 phoneme inventory of the speaker. In order to promote intelligibility with the Lingua Franca Core, it is thus important to consider which phonetic features cause L1 interference and unintelligibility in NBES of different L1 backgrounds and to replace these features with phonemes that approximate the L2 (Jenkins, 2000).
Jenkins (2000) found that the most important features for intelligibility were most consonant sounds, appropriate consonant clusters simplification, vowel length distinctions and nuclear stress. In establishing the core features of the LFC, Jenkins (2000) found that some RP and GA features are crucial to the intelligibility among NBESs and therefore, these varieties form the basis of the LFC. However, other RP and GA phonological features were found to have no effect or a negative effect on intelligibility. In these cases, the LFC was adapted by means of L1 English varieties, thus the LF maintains an egalitarian approach to different varieties of English (Jenkins, 2000).
The inventory of features and the considerations Jenkins (2000) made are too extensive to be discussed in full detail in this study. However, I will discuss some important features for Dutch ESL learners. For example, the dental fricative pair /θ/ and /ð/ are difficult to acquire for NBES and often substituted with approximating phonemes, a process that also occurs in many L1 English varieties. As substitution often does not lead to loss of intelligibility, these phonemes are not included and replaced by /v/ and /f/ in the LFC. Another example of an excluded consonant is the dark [ɫ] which is problematic for the majority of English learners (Jenkins, 2000). It is often replaced by a clear /l/ or /ʊ/, a process that does not harm intelligibility (Jenkins, 2000). This substation of dark /l/ is also found in native speakers of RP and Welsh English speakers do not even have a dark [ɫ] in their phonemic inventory. Jenkins (2000) explains that the acquisition of syllabic /l/ and dark /l/ takes great effort, whereas the /ʊ/ is a familiar phoneme to many English language learners and if not, it is relatively easy to acquire. Therefore the dark, velarized [ɫ] is not included in the LFC.
Further considerations have led to include GA [ɻ] in the LFC rather than the RP realisation of /r/ , as the RP [ɹ] is not realized in post-vocalic positions, despite orthographic occurrence (Jenkins, 2000; Collins & Mees, 2003). However, RP is used as reference in the LFC for realisations of /t/ as RP has a more reliable relationship with orthography (Jenkins, 2000). In intervocalic positions, /t/ in GA becomes flapped, making the phonetic realisations closer to /d/. Since this could create a loss of contrastiveness in words like madder and
matter, this realisation risks intelligibility.
are accustomed to pronunciation of /r/ when it is orthographically present (Collins & Mees, 2003). Lastly, in Dutch /t/ is, similar to RP and the LFC, realised as /t/, although substitutions of /t/ with /d/ also occur to approximate the GA alveolar flap (Collins & Mees, 2003).
However, the previously discussed phonemes of the LFC do not guarantee an increase of intelligibility in all NBES speech. As discussed earlier, many Asian English speakers perceive /l/ and /r/ as allophones and have difficulty in perceiving a contrast between the two sounds (Deterding, 2010). Maintenance of these two phonemes in the LFC does not solve the problem of incorrect perception and realisation of these phonemes by Asian English speakers. Moreover, the issue becomes even more complicated for English speakers from southern China due to maintenance of /n/ in the LFC. English speakers from southern Chinese regions do not perceive a contrast between /l/ and /n/ in onset-initial position (Deterding, 2010). Consequently, they continue to distribute and realise /l/,/n/and /r/ incorrectly in English speech. For these English learners, the LFC does not simplify pronunciation training, but more importantly, it does not increase intelligibility if these speakers continue to struggle with these phonemes.
Furthermore, substitution of dental fricatives with /v/ and /f/ might prove to be difficult for many learners with different L1s, as they confuse /v/ with /w/ or /b/, devoice or even elide the consonant in word-final positions (Dauer, 2005). In addition, many NBES substitute the dental fricatives not only with /f/ and /v/, but also with /s/ and /z/. This does not only lead to a lack of contrastiveness in words like sink and think, but it also means that English learners need to learn to substitute the phonemes correctly.
Thus, the LFC might not be suitable for all NBES, although it theoretically has the potential to make English pronunciation easier for native Dutch speakers. However, pronunciation is not merely a theoretical phenomenon. NBES might have their own opinions about pronunciation norms and it is important to investigate sociolinguistic topics like attitudes to pronunciation and their effect on ESL learners ideas about pronunciation training.
2.5 Judgements of Dutch accented English.
As a social phenomenon, language and accents convey a message as they invoke emotions and judgements about someone’s background (van den Doel, 2006). Accents are part of speakers’ identities and the fact that dominant accents like RP and GA have maintained their status as pronunciation norms thus far suggests that, in general, these accents are favoured by non-native speakers and relate to an image of status and successful language acquisition.
By studying the judgements of accented English, van den Doel (2006) and Koet (2007) demonstrated how native English speakers and native Dutch speakers judge Dutch-accented English. Van den Doel (2006) explains how native speakers often stereotype foreign accents, a practice that becomes evident in the assignment of accents to characters in (animated) movies and television programmes. It can be expected that the Dutch are aware of foreign accent judgement and stereotyping by native speakers, as the influence of the English language is increasing in Dutch society and media. A survey conducted among 615 Dutch university students of English indicated that the majority of the subjects (56%) believed the British and Irish to be the strictest judges of foreign accented English and the Americans and Canadians to be the most lenient judges (7%) (van den Doel, 2006). The belief that the British and Irish are the strictest judges of foreign accented English might be caused by the fact that British accents remain the norm in English pronunciation training in Dutch secondary schools (van den Doel, 2006).
To investigate whether the Dutch students were right in their beliefs, van den Doel (2006) studied judgements of Dutch-accented English given by native speakers of different English varieties. The American judges of Dutch-accented English were found to be stricter judges than the British judges. This finding contradicts the beliefs that the participants had, as they believed the British and Irish to be the strictest judges. However, in Koet’s (2007) study, there was no significant difference in the strictness of British and American judges. The strictest judges of English speech were not even native English speakers, as Koet (2007) found that Dutch speakers of English were more strictly judged by fellow native Dutch speakers.
In order to determine why accented English is considered incorrect, it is important to consider which features are considered most disturbing. Van den Doel (2006) established a hierarchy of errors to determine which phonetic features were deemed most disturbing by the native speakers. The establishment of a hierarchy of errors based on the judgements of native speakers from different varieties is crucial to understanding the irrelevance of standard native speaker norms in pronunciation training.
judgements of British native English speakers and American native English speakers. The Americans scored RP features as erroneous, whereas the RP speakers scored American features as erroneous. Thus, no matter which variety is used as a pronunciation model, speakers of different varieties will find differences and might consider these disturbing, if not erroneous.
The results demonstrated that Jenkins’ (2000) idea of intelligibility being more important than pronunciation norms is not aligned with reality. Even though Jenkins’ Lingua Franca Core is designed for communication between non-native English speakers, it does not take into consideration that both native and non-native speakers still favour pronunciations norms and that the latter value the judgements of the former (van den Doel, 2006; Scheuer, 2005). The judgements of native and non-native English speakers maintain the idea of appropriateness of current goals in pronunciation teaching, but it is important to consider that they might also influence the goals and motivation of English pronunciation learners.
2.6 Goals, motivation and attitudes in pronunciation training.
The previous section explained how accent stereotyping and native and non-native speaker judgements of accented English subscribed to English pronunciation norms. However, it is important to consider that the impact of attitudes of native speakers and ESL learners exceeds the predominance of native speaker norms. The attitudes ESL learners have towards accents might also affect their goals and motivation in pronunciation training. The role of motivation is vital in L2 acquisition (Csizér & Dörnyei, 2005). However, the goals and motivation of English learners is overlooked in the development of new pronunciation models, although they might be of greater significance to pronunciation training than expected.
However, in order to determine the level of motivation a student has, it is important to consider the factors that influence it. Motivation in itself has been recognised as a dynamic construct consisting of several interacting factors. Gardner (1985) has been of great significance in the field of psycholinguistics, as he established a construct of motivation that formed the bedrock for present dynamic motivational constructs. However, Gardner’s (1985) motivational construct is limited to two types of motivation that do not fully represent the construct of motivation. The first type of motivation is instrumental motivation, which is based on the idea that L2 acquisition functions as a means to attain another goal, such as a desire to please a teacher, or the promise of reward (Gardner, 1985). The second type, Integrative motivation refers to the learner’s willingness to identify oneself with the L2 community and its speakers (Gardner, 1985).
The recognition of different types of motivation lead to the development of a dynamic, interactive motivational construct that can also be applied to pronunciation training. Csizér and Dörnyei (2005) explain that the construct of motivation is subject to individual differences in the affective factors of motivation. In addition to Gardner’s (1985) motivational factors of integrativeness and instrumentality, other factors that might affect the motivational construct are: attitudes toward L2 speakers, cultural interest, vitality of L2 community and linguistic confidence (Csizér & Dórnyei, 2005). Attitudes to L2 speakers concerns the attitudes toward having contact with L2 speakers and traveling to their country and cultural interest indicates the appreciation of L2 culture. These items can be considered to be different parts of integrative motivation. The vitality of L2 community describes the perceived importance and wealth of the L2 communities and lastly, linguistic confidence relates to the learner’s confidence and perception of his own L2 acquisition capacity (Csizér & Dörnyei, 2005). This factor is related to Locke and Latham’s (2006) condition of self-efficacy, as both refer to a person’s belief about their capacity. Thus, linguistic confidence is an important component of motivation and pronunciation goal attainment.
Furthermore, the composition of these motivational factors can be related to an L2 motivational self-system (Dörnyei, 2005). The traditional disposition of integrative motivation is associated with the ideal L2 self, whereas an ought-to self is related to instrumental motivation (Dörnyei, 2005). In addition to integrative and instrumental motivation, Dörnyei (2005) introduces a third dimension called “L2 learning experience”. This motivational factor concerns the executive motives related to the immediate learning environment and experience (Csizér & Dörnyei, 2005).
the L2 community and cultural interest need to be considered to come to a more dynamic and thorough understanding of the construct and the interactions within motivation (Dörnyei, 2005). By studying these motivational factors and the students’ goals in pronunciation training the relationship between pronunciation goals, motivation and accent performance can be explored.
2.7 Statement of purpose
It has become clear that current pronunciation norms are based on the notion that ESL learners need to be prepared for communication with native speakers. Since English has become the Lingua Franca, there are more non-native than native English speakers and pronunciation pedagogy should aim at preparing ESL learners for interaction with other non-native speakers.
Furthermore, the current international status of English and increasing NNS-NNS interaction demand revision of language ownership and the authority of native pronunciation norms. Since English is a lingua franca, ownership should not be limited to native speakers from inner circle nations. However, granting non-native varieties the authority to set their own pronunciation norms or models could impede intelligibility among non-native English speakers.
Jenkins’ (2000) Lingua Franca Core aims to improve intelligibility in NNS-NNS interaction and distribute language ownership and authority equally over native and non-native English varieties. However, van den Doel (2006) and Koet (2007) demonstrated that deviating from current pronunciation norms is considered disturbing by both native English and native Dutch speakers. These findings indicate that Dutch speakers of English have a negative attitude towards Dutch-accented English and might not favour a pronunciation model that allows foreign-accented speech.
It is important to consider the importance of ESL-learners’ attitudes and goals in pronunciation pedagogy. Locke & Latham (1990, 2002) explain how goals, motivation, performance and effort influence each other. Individuals who agree with a goal are more likely to be motivated, make an effort and thus perform better. In pronunciation pedagogy, pronunciation models and norms are imposed on students and disagreement with these goals will not benefit the process of pronunciation training. It is important to study the pronunciation goals and motivational profiles of ESL-learners, as they can provide insights into the suitability of current pronunciation norms and the Lingua Franca Core for pronunciation teaching.
profiles based upon Dörnyei’s (2005) framework, the current attitudes to pronunciation norms and pedagogy can be determined. From the perspective of Goal-Setting theory (Locke & Latham, 1990, 2002), the relation between the students’ motivation and current pronunciation goals is determined and the possible effect of adoption of LFC in pronunciation teaching is discussed. By studying the effect of goals and motivations can be decided if implementation of LFC in pronunciation pedagogy could improve English pronunciation teaching.
It is expected that the Dutch participants aim to acquire a native-like accent as studies by Koet (2007) and van den Doel (2006) have found that native Dutch speakers find Dutch-accented English annoying to listen to. Furthermore, it is expected that the students with the strongest preference for native-like accent acquisition are most motivated and fit the profile of Dörnyei’s (2005) ideal L2 self.
3.
Method
In order to investigate what the pronunciation goals and motivational profiles of Dutch learners of English are, a questionnaire was designed based upon former studies by Dörnyei (2005), Csizér and Dörnyei (2005) and Smit (2000). With the results of the questionnaire it can be determined whether or not the implementation of a Lingua France Core model in English pronunciation meets the goals and ambitions of Dutch learners of English, and what the views of Dutch upper secondary school pupils are on English pronunciation training and the role of the prevalent native speaker norms.
3.1 Participants
3.2 Materials
Information about the goals, motivations, efforts and performance of the participants were gathered by means of a questionnaire consisting of 43 statements to which the participants were required to grade their agreement on a 5-point Likert Scale (1 = strongly agree, 2 = agree, 3 = neutral, 4= disagree, 5 = strongly disagree). The questionnaire was designed in Dutch to ensure that the participants would fully understand the items of the questionnaire. The original Dutch questionnaire can be found in Appendix A. Additionally, the participants were asked to provide background information about their L1s, age, sex, education level and target English accent.
The first five questions of the questionnaire enquired about the preference of the participants for different pronunciation goals. The listed pronunciation goals were:
1. My goal is to speak intelligible English, without a clearly retraceable accent. 2. I would like to speak intelligible English and maintain my Dutch accent. 3. I want to acquire a standard British English accent.
4. I want to acquire a standard American accent.
5. My goal is to acquire an English accent, regardless which variety.
In case the subjects preferred the final item, they could list their preferred accent at the end of the questionnaire.
The other variables of the questionnaire was based on Dörnyei’s (2005) L2 Motivational Self System, consisting of the previously discussed three dimensions: the Ideal L2 Self, the Ought-to L2 Self and the L2 learning experience. The different selection of content areas that might influence the L2 Motivational Self System of the subjects were based upon a review of previous studies (Csizér & Dornyei, 2005; Smit, 2000). The factors that were selected expected to influence the levels of motivations, as described in the background section, formed the following 5 variables:
- Linguistic confidence: in this study, this variable concerns the confidence of participants in English communication with native and non-native speakers, confidence in pronunciation capacity and abilities, confidence about current English accent and fear of judgement.
- Integrativeness: in this study, concerns the interest in English accents and cultures, attitudes to native English speakers and the desire to be accepted by them.
- Instrumentality: this variable concerns the participants’ desire to achieve other (long-term) goals through accent acquisition, such as a prestigious image, pleasing the teacher and job opportunities.
- intended effort: this variable measures the effort the participants are willing to make in order to acquire an English accent.
These variables are tested with items that measure the attitudes to different factors that are inherent to the variables. In order to check whether the participants filled out the questionnaire seriously, certain statements were proposed twice, with slight alterations. Each control pair enquired the same subject and would thus evoke a similar response in the participants.
The following items were included in the questionnaire (control pairs in italics): Linguistic confidence
6. I have the capacity to acquire an excellent English accent.
7. I am afraid to speak English because of my accent. 8. I believe I have a good English accent.
9. It is difficult to acquire an English accent.
10. My Dutch accent increases the fear to speak English with native English speakers. 11. It is more important to be intelligible in English, than to have a perfect English accent. 12. I feel more comfortable speaking English with people who do not have an English accent. 13. When I am speaking with non-native English speakers it’s important to have an English accent.
14. It is impossible for me to fully acquire an English accent. Integrativeness
15. My Dutch accent in English is part of my Dutch identity.
16. I want to improve my accent, because I like a particular accent.
17. I want to improve my accent, because I appreciate speakers of a particular accent.
18. It is important that native English speakers cannot retrace a Dutch accent from my English speech.
19. Having a Dutch accent causes native English speakers to have a negative image of me. 20. I want to improve my accent, because I am interested in English-speaking cultures. 21. I want to acquire an English accent in order to be accepted by native English speakers. 22. Having a Dutch accent in English causes native Dutch speakers to have a negative image of me.
23. It is important to be seen as a native English speaker when I speak English.
Instrumentality
24. I want to acquire an English accent, because it is expected of me at school. 25. I have to acquire an English accent to get good grades at school.
26. Having an excellent English accent will increase my chances of getting a good job. 27. Having an English accent is NOT important for my future.
28. I would like to improve my English accent, because I want to enrol in an English-taught programme.
29. I am taken more seriously when I speak English with an English accent.
31. I would like to improve my English accent, because I want to live in an English speaking country.
Intended Effort
32. I would consider staying in an English-speaking country to improve my pronunciation. 33. In addition to my regular homework I practice my English pronunciation independently. 34. I would like to improve my English pronunciation through contact with native speakers. 35. If given the possibility, I would enrol in an extra course to improve my English pronunciation.
36. I practice my English pronunciation through communication with native English speakers. 37. I try to imitate native English speakers as much as possible.
38. I put my full effort into English pronunciation exercises in class.
Learning Experience
39. More attention should be paid in class to English accent acquisition.
40. Acquiring an English accent is the epitome of successful pronunciation training. 41. Correct pronunciation is an important part of English language acquisition.
42. There should be a smaller focus on English accent acquisition in class.
43. Pronunciation training exercises to improve my English are boring.
In case participants responded similarly on two control items, their results were found unreliable and excluded from data processing and analysis. The questionnaire was tested on reliability and found reliable with the Cronbach Alpha for the 43 items at .76.
3.3 Procedures
The questionnaire was filled out by the students in similar environments. All participants were asked to fill out the questionnaires in a classroom, at the beginning of their English lesson. Prior to handing out the questionnaire I emphasised the correct interpretation of the term “English accent” to ensure that the participants would not interpret this term as referring to British English only. The subjects were asked to fill out the questionnaire individually, without discussing their answers with their peers. The procedure cost 10-15 minutes per class.
3.4 Design and analysis
In Csizér and Dörnyei’s (2005) study, the mean scores of variables are given, yet the means of the individual items that test the variables are not presented. In this study, the clusters’ means on the variables as well as the mean scores of the individual items are calculated and presented. In this manner, any variation in item scores within the variables will be visible and thus it can be determined on which aspect of each variable the clusters differ. By calculating the mean scores of items and variables, the differences between the clusters become apparent. In order to test whether the differences between the clusters were significant, various statistical analyses were used. Since the data of the clusters was not normally distributed, the Kruskal-Wallis test was run to test for significant differences in item medians and variable medians among all clusters. Additionally, for post-hoc analysis Mann-Whitney tests were run to find significant differences in item medians between cluster pairs. It was decided that the differences reach significance at p<0.05 on the initial Kruskal-Wallis analyses. On the post-hoc analysis, the alpha level was adjusted with Bonferroni correction. As there are 4 clusters and thus 6 comparisons, the results would reach significance at p<0.0083 (0.05/6) on the post-hoc analyses. In the tables, the mean scores of variables and items are given instead of the medians, as clusters might have the same medians due to the Likert scale. By presenting the means rather than the medians, the differences between the clusters’ item and variable scores are more easily identified.
4. Results
One of the participants’ results were found to be unreliable, as the respondent had different responses to the control items. The results of the questionnaire were processed in Excel. The participants had 5 main options for pronunciation goals: speaking English without a clearly detectable L1 accent, speaking L1-accented English, speaking standard British English (RP), speaking General American English (GA) or acquiring any native English accent.
Mean Standard
deviation
1. My goal is to speak intelligible English, without a clearly retraceable accent. 1.92 0.87 2. I would like to speak intelligible English and maintain my Dutch accent. 3.36 1.08
3. I want to acquire a standard British English accent. 3.16 1.08
4. I want to acquire a standard American accent. 3.14 1.04
5. My goal is to acquire any English accent. 2.83 0.92
Table 1 Preference of Pronunciation Goals in percentages
However, it is clear that the goal to speak intelligible English without a distinct accent is preferred by the students and that acquisition of any native accent also has a slight preference. In order to determine which native English accent the participants prefer, the students were asked to list any English accents they would like to acquire if they were striving to acquire a different native accent than standard British English or standard American English. Of the 133 participants, only 14 listed an English accent they aim to learn, 2 of which listing 2 accents. The non-standard accent preferences of the 14 participants can be found in figure 1.
Figure 1 Non-standad accent preference of 14 participants.
The fact that only 14 participants (9,5%) listed a non-standard English accent as their pronunciation goal, indicates that not all respondents who indicated to desire acquisition of any English accent (item 5) have a preference for a particular accent. Among the participants that listed accents, Australian English and Irish English were the most popular non-standard varieties.
In order to identify the motivational profiles that belong to the different pronunciation goals of the respondents, a non-hierarchical cluster analysis was run to divide the participants into 4 groups according to their item scores. The non-hierarchical cluster analysing resulted in four clusters which contain participants with similar scores, as can be seen in table 2. The cluster sizes varied greatly, but since it was decided to use Mann-Whitney tests, this would not be problematic.
cluster 1 N=26 cluster 2 N= 50 Cluster 3 N=45 Cluster 4 N=12 19.5 37.6 33.8 9
Table 2 cluster membership of participants in percentages
The mean scores on the pronunciation goals, the variables’ and items were calculated per cluster. The results of the analysis of pronunciation goals are presented in Table 3. Initial inspection of the mean scores indicates that the clusters have different pronunciation goals. Of all clusters, the first cluster is most in favour of losing a Dutch accent or acquiring a native English accent. The first cluster scores high on Dutch accent maintenance, which demonstrates that the members do not agree with this pronunciation goal. Cluster 1 prefers native English accent acquisition, as the scores on items 1 and 3-5 are relatively low,
indicating agreement with the statements about English accent acquisition as pronunciation goals.
In stark contrast with cluster 1 is the fourth and smallest cluster, which is marked by a strong preference to either retain or minimise a Dutch accent in English. This cluster also scores highest on statements concerning native English accent acquisition, which indicates disagreement with these goals.
There seem to be no salient differences in the mean scores of clusters 2 and 3. Both clusters score higher than clusters 1 and 4 on the first item, indicating the strongest preference of all groups for pronunciation training that aims at minimising an L1 accent without acquiring a native-like accent. However, in order to determine whether these differences were significant, the clusters were tested by means of the Kruskal-Wallis test and subsequent Mann-Whitney U tests to find differences between cluster pairs.
Significant differences among the clusters and individual cluster pairs were found. As could be expected from the findings in Table 1, there were no significant differences between the clusters on the first item. Almost all of the participants, regardless of their cluster, agree with the goal of intelligible English pronunciation without a distinct and retraceable accent. However, the opinions varied greatly regarding maintenance of a Dutch accent in English (item 2). Significant differences were found among all cluster pairs, with clusters 1 and 3 indicating disagreement with Dutch accent maintenance, cluster 2 agreeing nor disagreeing and cluster 4 agreeing on Dutch accent maintenance.
Clusters 1 and 4 are significantly different in their preference for all pronunciation goals, except the first. Clusters 2 and 3 differ significantly from the first and last cluster, falling between the two extreme opposites, however, they do not differ significantly from each other. Inspection of the scores on motivational variables might indicate on which aspect of motivations clusters 3 and 4 differ
.
Pronunciation goals
item cluster 1 cluster 2 cluster 3 cluster 4 KW results
1. My goal is to learn to speak intelligible English, without having a clearly detectable and retraceable accent.
2.35 1.84 1.78 2.08 H= 5,545 p= .136 2. I want to learn to speak intelligible English while
maintaining my Dutch accent.
4.12 3.0 3.64 2.0 H= 38,460
P=<.001 3. I want to acquire a standard British English accent. 2.27 3.28 3.27 4.33 H= 27,714
P=<.001 4. I want to acquire a standard American English accent. 2.81 3.06 3.09 4.58 H=21,790
P=<.001 5. My goal is to acquire an English accent through
pronunciation training, regardless of the variety.
2.27 2.88 2.93 3.67 H=13,876 P= .003
Post-hoc comparison* Item 1 Item 2 Item 3 Item 4 Item5
* Numbers refer to clusters. Clusters presented on different rows indicate significant differences among clusters. Individual statistical results in appendix D.
In order to determine the motivational profiles of the clusters, the clusters’ scores on the motivational variables were compared. Since the variable scores of all clusters were not normally distributed, it was decided to run Mann-Whitney U tests to find significant differences between the variable data of the clusters. It is important to bear in mind that this non-parametric test does not compare means, but medians and sums of ranks.
Analyses of the overall scores of the clusters on the different variables indicates that there are significant differences between the clusters (Table 4). All clusters differed significantly from each other on integrativeness, instrumentality and intended effort. The results in Table 4 demonstrate that cluster 1, which contains students who aim to acquire a native-like accent, shows the strongest agreement with the items from integrativeness, instrumentality, intended effort and learning experience.
The majority of the items of integrativeness, instrumentality and intended effort tested positive attitudes of the participants, thus the lower the scores, the stronger the agreement with items that indicate high levels of the motivational variables. This means that the low scores of cluster 1 indicate high levels of integrativeness, instrumentality and intended effort. The first cluster is followed by clusters 2,3 and 4 which lean to disagreement or indicate strong disagreement with the items on the motivational variables. Thus, it can be concluded that cluster 1 has significantly higher levels of integrativeness, instrumentality and intended effort.
On linguistic confidence, cluster 1 differed significantly from cluster2, which also differed significantly from cluster 3. The results in table 4 demonstrate that cluster 1 disagreed most with the linguistic confidence items, whereas cluster 2 had the highest level of agreement. As the majority of the items of this motivational variable tested the level of anxiety and lack of confidence, the disagreement of cluster 1 should be interpreted as an indication of high linguistic confidence, and the slight agreement of cluster 1 indicates a lack of linguistic confidence.
The final motivational variable that was tested was learning experience, however, statistical analyses did not yield any significant results. The mean scores in table 4 indicate that the clusters agreed with the learning experience items.
Linguistic
confidence integrativeness instrumentality Intended effort Learning experience
Cluster 1 3.33 2.85 2.78 2.74 2.77 Cluster 2 2.96 3.25 3.07 3.26 2.78 Cluster 3 3.28 3.65 3.38 3.38 2.89 Cluster 4 3.16 3.81 3.72 4.40 2.89 * Numbers refer to clusters. Clusters presented on different rows indicate significant differences. Mann-Whitney results in Appendix C. 2 4 4 3 1 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
Table 4 cluster mean scores and test results per motivational variable
In order to determine on which aspects of the motivational variables the clusters differ, the clusters’ scores on the individual items of the variables were tested for significance. The most important results on the items of the motivational variables will be discussed. The results of all items can be found in Appendix C and the p-values of the Mann-Whitney U tests are presented in Appendix D.
The first motivational variable that was tested in the questionnaire, was linguistic confidence. The linguistic confidence items measured the level of confidence regarding accent performance, capability and fear of judgement (table 4). The mean scores of the items that showed the most significant difference among the clusters are presented in table 5. Clusters 1 and 3 scored low on first two items, concerning confidence about acquisition abilities and current accent performance. This indicates that they believe themselves to be able to acquire an English accent and are not unconfident about their current English accent. Unlike clusters 1 and 3, the other two clusters are unconfident about their accent and their capacity to acquire an English accent, as they scored high on these items. The post-hoc comparison of the clusters demonstrated that clusters 1 and 3 were significantly more confident than clusters 2 and 4.
Linguistic confidence
Item cluster 1 cluster
2
Cluster 3
Cluster 4 KW test results
6.I have the capacity to acquire an excellent English accent.
1.96 3.22 2.07 3.5 H = 52.118
P = <.001 8. I believe I have a good English accent. 2.96 3.82 3.07 3.75 H=37.86
P=<.001 14. It is impossible for me to fully acquire an English
accent.
4.27 2.84 3.8 2.33 H= 47.264
P= <.001
Post-hoc comparison* Item 6 Item 8 Item 14
* Numbers refer to clusters. Clusters presented on different rows indicate significant differences among clusters. Individual statistical results in appendix D.
1 3 2 4 1 3 4 2 4 2 3 1 Table 5 Mean scores and test results of linguistic confidence variable per cluster and item
be found in table 6. The scores demonstrate that cluster 1 scores low on integrativeness items, indicating a positive attitude towards L2 accents (item 16) and the L2 community and its speakers (items17 & 21) among students who wish to acquire a native-like accent. Cluster 1 was followed by cluster 2, cluster 3 and finally cluster 4 which have a low level of integrativeness.
The Mann Whitney U tests demonstrated that there were significant differences between some clusters. Clusters 1 and 4 were significantly different from each other and from clusters 2 and 3 on the first three items concerning cultural identity, attitudes to English accents and attitudes to native speakers of English accents. Cluster 4 is significantly different from all clusters in indicating that their native accent is part of their Dutch identity (item 16), a sentiment that is not shared by cluster 1 and felt significantly less by clusters 2 and 3. The 4th cluster also scored significantly lower on appreciation of English accents and its speakers as a motif for pronunciation training.
Similar to the results of preference in pronunciation goals, the difference between the two intermediate clusters are not significant in the majority of the items. However, on items 18, 20 and 21 clusters 2 and 3 differ significantly. From these findings can be concluded that the groups are different in the level of certain aspects of integrativeness, namely native speaker acceptance and judgement and cultural interest.
Integrativeness
Item cluster 1 cluster 2 Cluster
3
Cluster 4 KW test result
15. My Dutch accent in English is part of my Dutch identity. 3.85 2.92 3.13 2.08 H= 26.491 P= <.001 16. I want to improve my accent, because I like a particular accent. 1.92 2.72 3.07 3.92 H= 33.815
P=<.001 17. I want to improve my accent, because I appreciate speakers of a
particular accent.
2.38 3.22 3.38 4.25 H=27.427
P= <.001 18. It is important that native English speakers cannot retrace a
Dutch accent from my English speech.
2.42 3.30 3.76 3.58 H= 28.318
P= <.001 20. I want to improve my accent, because I am interested in
English-speaking cultures.
2.31 3.30 3.80 4.33 H= 40.567
P= <.001 21. I want to acquire an English accent in order to be accepted by
native English speakers.
3.19 3.40 3.93 4.0 H= 18.099
P= <.001
Post-hoc comparison* Item 15 Item 16 Item 17 Item 18 Item 20 Item 21
* Numbers refer to clusters. Clusters presented on different rows indicate significant differences among clusters. Individual statistical results in appendix D. 4 2 3 1 1 2 3 4 1 2 3 4 1 2 4 4 3 1 2 3 4 1 2 3 4
For the other motif of motivation described by Gardner (1985), instrumentality, results can be found in table 7. The participants from cluster 1, who demonstrated a high level of linguistic confidence and integrativeness, indicated a wish to enrol in English-taught education (item 28). On this item, they differed significantly from all other clusters. Furthermore, the prospects of moving abroad to an English-speaking nation and intelligence and prestige associated with an English accent are important reasons for desiring a native-like accent (items 29, 30 & 31). Clusters 2 and 3 also scored positive on some of the items, although their scores were significantly lower than those of cluster 1.
Similarly to the results on pronunciation goals and integrativeness, no significant differences between cluster 2 and 3 were found. However, the distinction between clusters 1 and 4 is once more apparent. The results show that the two clusters differ significantly in their agreement with statements about relocation to English-speaking countries and educational plans. This shows that the instrumental motivation is important for cluster 1, and to a lesser extent for clusters 2 and 3. To cluster 4, however, instrumentality is not a motivating factor. Instrumentality Item cluster 1 cluster 2 Cluster 3 Cluster 4 KW results
28. I would like to improve my English accent, because I want to enrol in an English-taught education.
2.5 3.7 3.51 3.92 H= 27.780 P= <.001 29. I am taken more seriously when I speak English with an English accent. 2.27 2.9 2.96 3.58 H= 17.588
P=.001 30. To be considered successful and intelligent, it is important to have a good
accent when I speak English.
2.46 2.74 3.11 4.17 H=20.876 P= <.001 31. I would like to improve my English accent, because I want to live in an
English speaking country.
1.69 2.54 2.87 4 H= 48.025 P= <.001
Post-hoc comparison* Item 28 Item 29 Item 30 Item 31
* Numbers refer to clusters. Clusters presented on different rows indicate significant differences among clusters. Individual statistical results in appendix D.
1 3 2 4 1 2 3 4 1 2 3 4 1 2 3 4
Table 7 Mean scores and results of Instrumentality variable per cluster and item
The other clusters show stronger, yet similar disinterest in additional training and exercises. Cluster 4 in particular scored very high on almost every item, thus demonstrating a very low intended effort for pronunciation training.
The post-hoc comparison which compared the medians of the cluster on each item demonstrated great contrasts among cluster pairs on intended effort. Cluster 1 scores significantly lower on intended effort than the other clusters, indicating greater intended effort. Clusters 2 and 3 only differ significantly on item 34, with cluster 2 willing to improve pronunciation through communication with native speakers and cluster 3 disliking the idea. The differences between clusters are once again most significant between clusters 1 and 4. The final cluster’s intended effort is significantly lower than that of all other clusters.
Intended effort
Item cluster 1 cluster 2 Cluster 3 Cluster 4 KW results
32. I would consider staying in an English-speaking country to improve my pronunciation.
1.69 2.54 2.87 4 H= 31.747
P= <.001 33. In addition to my regular homework I practice my
English pronunciation independently.
3 4.02 4 5 H= 26.999
P= <.001 34. I would like to improve my English pronunciation
through contact with native speakers.
2.31 2.82 3.51 4.58 H= 49.737
P= <.001 35. If given the possibility, I would enrol in an extra
course to improve my English pronunciation.
3.31 3.36 3.78 4.58 H= 14.788
P= .002 36. I practice my English pronunciation through
communication with native English speakers.
3.35 3.76 3.62 5 H= 23.884
P= <.001 37. I try to imitate native English speakers as much as
possible.
2.5 3.46 3.64 4.83 H= 40.937
P= <.001
Post-hoc comparison* Item 32 Item 33 Item 34 Item 35 Item 36 Item 37
* Numbers refer to clusters. Clusters presented on different rows indicate significant differences among clusters. Individual statistical results in appendix D.
1 2 3 4 1 3 2 4 1 2 3 4 1 2 3 4 1 3 2 4 1 2 3 4
Table 8 Mean scores and results of intended effort variable per cluster and item
The final variable that was tested was learning experience. It is important to consider the influence of these scores on the attitudes of the students toward pronunciation training and exercises. The ideas students have about pronunciation teaching and their experience with it can affect their motivation to train their pronunciation (Locke & Latham, 2006).
In table 9, the mean scores on items 39 and 42 illustrate that cluster 1 and 4 have opposing opinions about the role of pronunciation teaching in English class. Cluster 1 is in favour of a greater emphasis on pronunciation training, whereas cluster 4 is not. This was to be expected as cluster 1 wishes to acquire a native-like accent, whereas cluster 4 only aims to speak intelligible English and maintain a Dutch accent in pronunciation.
Learning experience
Item cluster 1 cluster 2 Cluster 3 Cluster 4 KW results
39. More attention should be paid in class to English accent acquisition.
2.31 2.94 3.22 4.08 H= 28.733
P= <.001 42. There should be a smaller focus on English accent acquisition
in class.
3.73 3.20 3.33 2.08 H= 19.953
P= <.001
Post-hoc comparison* Item 39 Item 42
* Numbers refer to clusters. Clusters presented on different rows indicate significant differences among clusters. Individual statistical results in appendix D.
1 2 3 4 4 2 3 1
Table 9 Mean scores and results of Learning experience variable per cluster and item
5. Discussion
5.1 Pronunciation goals and motivational profiles of the clusters
At the onset of this study it was stated that it was expected that Dutch upper-secondary school students would prefer native-like accent acquisition as van den Doel (2006) and Koet’s (2007) studies indicated an awareness of current pronunciation norms and possibly negative native speaker judgements among native Dutch speakers.
The results indicated that a majority of the students does not favour native-like accent acquisition. All groups of participants showed preference for the goal of learning to speak intelligible English, without having a clearly detectable and retraceable accent. Only one small group indicated the desire to maintain a Dutch accent. This means that only a minority of the participants favours Dutch-accented English and that the majority still finds Dutch-accented English undesirable, even when it does not impede intelligibility. It is evident that Dutch-accented English is not appreciated, which is in line with Koet (2007) and van den Doel’s (2006) findings on judgements of accented English.
The participants who aim at native-like accent acquisition also form a minority. The results demonstrated that participants from the first group had the highest native-like accent preference, favouring a standard British English accent, unspecific English accents and to a lesser extent, a standard American accent. The participants’ preference was found to differ significantly from the other groups of students on standard British English and unspecified English accent preference. However, the second and third group also indicated to aim for unspecified English accents, albeit to a significantly lesser extent than group 1 and a significantly higher extent than the students who aim for Dutch accent maintenance. Thus, although preference for intelligible English without a distinct accent is higher, the majority of the students aim to acquire a native-like accent, but not standard British English or standard American English.
non-standard accent, it remains unclear which non-non-standard varieties are favoured by the students. It is possible that the majority of the students who indicated preference for any native accent did not have a particular non-standard accent in mind. The students might feel like they should acquire an English accent, but they do not like the standard American and British accents that are the current norm. From the non-standard accents that were listed, Australian English and Irish English were most popular, as both were selected by 25% of the participants. The preference for Australian English and Irish English came from 2 classes. The desire of the participants to acquire an Irish English accent might be explained by the fact that all participants who indicated this preference had an Irish teacher. The students’ familiarity with and exposure to the accent and a perhaps a positive disposition to the Irish teacher might have influenced the participants’ attitudes toward the accent. The preference for Australian English found in participants from another class might be caused by the increase in Australian television programmes on Dutch television channels. Exposure to these non-standard accents might have caused the students to favour them and aspire to acquire them.
By studying linguistic confidence, integrativeness, instrumentality, intended effort and learning experience of the students, their motivational profiles can be established. It was expected that the students who aim to acquire a native-like accent would have high levels of integrativeness and intended effort. The results on the motivational variables demonstrated that all groups of students differed significantly from each other on integrativeness, instrumentality and intended effort.
Furthermore, from the results can be concluded that the lower the preference for native-like accent acquisition, the lower the motivation of the students seems to be. The group of students who expressed the desire to acquire British, American or another native English accent had, as expected, a significantly higher level of integrativeness, instrumentality and intended effort than all other groups of students. This minority was the only group that demonstrated positive levels of integrativeness, instrumentality and intended effort, despite two other groups indicating a preference for accent acquisition of any English variety. Furthermore, the minority of students that aims to speak either intelligible English without a clear accent or maintain a Dutch accent in English scored significantly lower than all other groups on integrativeness, instrumentality and intended effort.
It is clear that the students who scored high on all native-accent pronunciation goals are the most motivated students. This group of students is followed by two large groups of less motivated students who wish to speak intelligible English, but also have a slight preference for acquisition of any native English accent. Finally, the least motivated students are the students who desire to speak intelligible English without a distinct accent or Dutch-accented English. These students scored consistently low on the variables, making them the least motivated group of students. Cluster 4 scores lowest on intended effort, integrativeness and instrumentality on almost every item of the questionnaire
