Journal of Linguistics and Language Teaching (JLLT)

Journal of Linguistics and Language Teaching

Volume 12 (291) Issue 2, pp. 129-146

Accent Choices of English Majors in the Netherlands: Patterns and Motivations

Gezina Christien Dorothé Huttenga (Groningen, Netherlands) & Dick Smakman (Leiden, Netherlands)

Abstract

This paper investigated the accent choices and realisations of first-year students of English in the Netherlands and linked these choices to the social identity, or persona, they wished to express in English. Further, it was investigated how important students felt identity expression was relative to other criteria in their pursuit of a certain accent. This group gives us an impression of the norms that future teachers may apply in their classes. Most (70) of the survey participants preferred to follow the Standard British English (Received Pronunciation) articulation model, while a reasonable group (29) were more inclined towards more Americanised and international accents. Interviews demonstrated that students held traditional connotations toward accents of English: Received Pronunciation indexed tradition and formality and related characteristics while general American indexed dynamism, internationalism, and related characteristics. Identity expression was relatively low on the agenda of these students. The sound file analysis revealed students were not consistent in applying their stated RP preference when they actually speak. They mixed their accent not only with their L1 but also with American English, possibly below the threshold of awareness. The results show that this particular group distinguished clearly between school norms and real-life norms. Our suggestion is that this distinction deserves more attention in teaching so that students become better prepared for real-life communication and become mainly concerned about international and intercultural intelligibility. L1 influences may be embraced as being natural and not in need of suppressing if they do not hinder intelligibility.

Keywords: L2 identity, Received Pronunciation, American English, intelligibility, ELT

1 Introduction

1.1 Multilingual Identities

Speaking a certain language does not merely involve literal communication of a message. Language use choices are by the speaker and listener aligned with a real or imagined culture and group. Language choice may impact identity because identity is a construct strongly embedded in social contexts. A person defines him or herself according to his or her similarity and difference with others (Tajfel 1974). Burck (2011: 363) even found that switching languages can alter speakers’ perceptions of themselves.

Tajfel’s (1974) Social Identity Theory suggests that identity is primarily based on group membership and the value individuals place on belonging to those groups. Tajfel furthermore stated that a personal, more individual dimension of identity exists. More recent postmodern perspectives emphasise the unstable nature of identity. Individuals belong to different groups, and may portray different versions of themselves in each group they are a part of (Bucholtz & Hall 2010, Joseph 2010, Norton & Toohey 2011, Pennycook 2009, 2018).

Indexicality is a framework that captures this phenomenon of applying language to point to social meaning, such as group membership, personality traits, and ideologies (Bucholtz & Hall 2005, Eckert 2008). Different social meanings may not just be indicated by using different languages, but also different varieties of the same language. Speakers may put different pronunciation variables together in a bricolage, and create their own idiolect (Eckert 2008), which uniquely expresses their persona, or personae, and in discourse, this idiolect may vary. People can thus express their identity by using certain so-called indexes, or pronunciation variables to show who they are.

1.2 Accents and Identity

Of all the available language features to play with, accent is a particularly important indicator of such efforts to associate with identity because speakers can manipulate it, and it is easily perceived (Bucholtz & Hall 2010, Ortega 2009, Van der Haagen 1998). Different accents and contrasting aspects of these accents have divergent associations. Accent choices may signal status, friendliness, or other social aspects. Research has demonstrated that people associate speakers with certain accents with differing levels of competence, intelligence, and sociability (Coupland & Bishop 2007, Dent 2004). Moreover, specific accents may be suitable in different situations. Certain accents can signal a higher status and make listeners more likely to take the speaker seriously (Eckert 2000). Aware of this, people may seek to alter their accents to climb socially through association (Mugglestone 1995) or index identity. Geeslin & Long (2014) demonstrated that L2 students may index their personality, social identity, and other information through their subtle pronunciation choices, just like L1 speakers do.

Still, this ‘playing with language’ has its limitations. Firstly, L2 speakers are limited by their linguistic capabilities (Ortega 2009) and cannot model their L2 accent perfectly and consistently. A second limitation is their incomplete knowledge of the cultural implications of their language use. Learners may rely on very general stereotypical associations, when in fact, many different associations to a pronunciation variable exist. A strongly rhotacised rhyme, for instance, does not directly link one’s pronunciation with United States English and its typical association with dynamism or enthusiasm. Within and outside this region, there is great variation in the occurrence of this sound, and rhoticity is also used in Scotland and various regions in England, for example. Non-rhotic accents exist in the United States as well. As (Eckert 2008: 467) noted, different hearers may select disparate meanings from the larger indexical field of associations and pronunciation feature indexes. Therefore, a person’s accent choice may not always be interpreted in the way they intend.

1.3 Accent choices: RP vs GA

Dutch learners are exposed to many English accents through the mass media such as the internet and television. The most internationally dominant of the accents they encounter are Received Pronunciation (RP) and General American (GA) (Ladegaard & Sachdev 2006). Learners may adopt RP for reasons of status. This particular accent is still favoured in Europe, both by official educational bodies and informally by ordinary language users (Ladegaard & Sachdev 2006, Van der Haagen 1998, Edwards 2016, Broersma 2015), though this may be gradually changing. Hughes, Trudgill & Watt (2005: 5) noted that this accent can index “affectation, social snobbery, arrogance, and aloofness” in Great Britain. GA is not considered as strongly class-based (Mair 2006), and it is less localisable.

Pronunciation differences between Received Pronunciation and General American are quite salient (Wells 1982, Van der Haagen 1998). Most notably, GA is rhotic, while RP is not, and this feature, in combination with phonemic differences as well as speaking style (speech-tract position), makes these two accents highly different (Wells 1982, Van der Haagen 1998). Many other such differences exist, and in formation, they make these two model accents highly mutually different; not only linguistically / phonetically, but also their indexical meanings.

Van der Haagen’s (1998) study of the attitudes of Dutch secondary students found that these students generally considered RP normative and correct, associating it with more status. However, these same students preferred GA themselves. They associated US accents with more dynamism and considered the US accents as more “witty, dynamic, spontaneous”. Moreover, they also felt such accents held more affect, considering them “honest, friendly,” and “companionable” (Van der Haagen, 1998: 55, 57-60). As Van der Haagen noted, RP then appears to have more overt prestige, and GA more covert prestige. This means that while RP is considered to have more official status, GA has the “hidden values associated with non-standard speech” (Trudgill 1972: 183-184).

1.4 Previous Research on Attitudes towards Accent in the Netherlands

In the Netherlands, Edwards (2016) and Van der Haagen (1998) found a general preference for RP as an accent model. Yet, these same investigations revealed much tolerance towards American and more international accent models. They also found that GA was gaining popularity. Younger people, the leaders of language change, were somewhat more likely to choose an American accent, while older people more often chose a British accent (Edwards 2016). It appears, then, in line with Van der Haagen’s (1998) study that the normative status of RP may be shifting.

2 Methodology

2.1 Research questions

As Geeslin & Long (2014) noted, social factors, including identity formation, are underrepresented in second language acquisition research. Accent attitudes in the Netherlands are changing and dynamic, and it is therefore interesting to examine what factors influence and motivate Dutch future English-language experts to choose a specific accent. The present study relates English majors’ accent choices explicitly to their identity and does so through the following research questions:

1. What accents do these students aspire to speak?

2. What are the indexicalities these students attach to various accents of English?

3. What are the accent features of these students’ English?

2.2 Participants

Ninety-nine Dutch students of English from Leiden University (Netherlands) participated in our questionnaire. Seventeen of these participants had also produced sound files in which they spoke English. Furthermore, four participants were interviewed. Students were all in their first year of study and were generally 18 to 20 years old. They had not had much explicit pronunciation training at university level and only general instructions at secondary education. Table 1 shows the overall attributes of the participants in the tests:

Gender	Questionnaire (N=99)	Sound File Analysis (N=17)	Interviews (N=4)
Women	77	9	2
Men	22	8	2

Table 1: Participants

The students’ level of English before their studies was generally intermediate to high, as is common in Northern European countries. The students had a choice between two pronunciation seminars: ‘International English’ (which often translates to GA in practice) or ‘Received Pronunciation’ (standard British English as typically spoken by BBC newsreaders). All first-year students, irrespective of accent or other preferences, were asked to participate in the questionnaire. For practical reasons, the audience for the sound file analysis and interviews was only with students aspiring toward a British accent. The ‘International English’ group was not large enough. Also, their pronunciation criteria were less strict, which means that this group was more diverse than the ‘British’ group.

2.3 Questionnaire

The group of 99 students was asked to fill in an online questionnaire that included nine questions. These questions related their accent choice in English to the persona they wished to portray in this language. The first question was an open one: ‘Do you have a specific accent or model you want to emulate (e.g., Scottish English, British English, American English)’. Students could specify whether they had a target accent, and if so, which one. The remaining questions were scaled questions, which are shown in Table 2. The responses were measured on a six-point scale. The questionnaire took five to ten minutes to complete:

Questionnaire Statement	Identity Label
I feel Dutch.	Dutchness
I feel close to the culture of my model accent.	Cultural alignment
I want to sound smart when I speak English.	Intelligence
I want to sound cool when I speak English.	Coolness
I use my English accent to show who I am.	Self-expression
I want to sound international when I speak English.	International outlook
I want to sound like an English native speaker.	Nativeness

Table 2: Questionnaire Items

Students could mark to what degree they agreed with these statements using a graded scale, called the empirically grounded, well-anchored (EGWA) scale, which is used in Psychology (Pelham & Blanton 2007) and which can be viewed in Table 3. Students were also given the option of ticking ‘not applicable’. This scale enables participants to choose values between ‘not at all’ and ‘slightly’, between ‘slightly’ and ‘quite a bit’, and between ‘quite a bit’ and ‘extremely’:

0	1	2	3	4	5	6
not at all		slightly		quite a bit		extremely

Table 3: The EGWA Scale Model

2.4 Recordings Analysis

To gain insight into the realities of students’ accents early on in their pronunciation course, recordings of the texts read out by 17 of the students were described phonetically. All the selected students aspired toward a British accent. The text can be found in Smakman (2015: 15-17))

The MP3 recordings were transformed to WAV (Waveform Audio File Format) and subsequently processed through the speech-analysis programme Praat (meaning ‘talk’ in Dutch). This tool was designed by Boersma & Weenink (2016) and is regularly updated. Through this tool, voice, voice-onset, silence, aspiration, and vowel development can be determined. Besides this tool, auditory analysis, i.e. listening carefully, was carried out by the first author of the present article. This technique helped us to determine the presence or nature of consonants. To increase the reliability of the auditory analysis, a so-called semi-consensus transcription was performed; the two authors had a joint listening session to fine-tune their perception as much as possible, and after that the first author continued by herself. Different occurrences of /r/, /t/, and of the vowels were selected in specific phonetic / phonological contexts, because these features are considered to distinguish different types of accents, especially British and American English (Van der Haagen 1998). The second author, an experienced English pronunciation teacher (British and International English) and native speaker of Dutch, assessed the degree to which each student’s pronunciation tended towards Dutch, British English, or American English. He gave them a percentage score on how Dutch, British, and American they sounded, with the percentages adding up to a 100%. The categorisation was based on knowledge of all three accents. Inevitably, the phonetic features this article focuses on were important determiners of this categorisation.

2.4.1 Consonant /t/

For /t/, seven tokens were collected and two variants were distinguished. These tokens and the variants that were distinguished are shown in Tables 4 and 5:

Context	Tokens
(-VCtV-)	created, cheetah, meters, motivated, water
(-VCt)#(V-)	but I’ll, habit is

Table 4: Distinguished Contexts of /t/¹

The first author of the present listened and then wrote down the variants she encountered. These variants are in Table 5:

Variants
No /t/	[ʔ]	glottal stop
	[∅]	empty realisation
Full /t/	[t]	neutral
	[th]	aspirated
	[ˀt]	post-glottal
alveolar	[d]	neutral ‘d’
	[d̥]	voiceless
	[ɾ]	voiced

Table 5: Distinguished Phonetic Variants for /t/

2.4.2 Consonant /r/

For /r/, ten tokens were collected and two variants were distinguished. These tokens and the variants that were distinguished are presented in Tables 6 and 7:

Context	Tokens	Non-linking or Linking /r/
(-VCrC)	world, first	non-linking
(VCr)(C-)	furthermore, furthermore	non-linking
(VCr)#(C-)	for more, there must	non-linking
(-VCr)#	lover, furthermore	non-linking
(-VCr)(V-)	under a, answer is	Linking

Table 6: Distinguished Contexts of /r/

In all these contexts, /r/ occurred postvocalically, while in two of them – namely word-finally before a vowel in the next word and word-finally before a pause – it was subject to possible linking to the subsequent syllable (linking /r/). The two variants were distinguished as given in Table 7:

Variants

/r/ pronounced; rhotic

/r/ not pronounced; non-rhotic

Table 7: Distinguished Phonetic Variants for /r/

2.4.3 Vowels

Four vowels for analysis were selected. These four and the tokens (4 per vowel) in which they occurred in the texts are in Table 8:

‘bath’ vowel	‘lot’ vowel	‘trap’ vowel	‘dress’ vowel
giraffe	spotted	cats	then
asked	dog	land	head
answer	logically	habit	get
after	not	Jack	kept

Table 8: Distinguished Contexts of /r/

All tokens were in word-stressed syllables. Following suggestions in Deterding (1997), Smakman (2006), and Botma, Sebregts & Smakman (2012), these vowels were purposefully selected so as not to follow or precede /ɹ/, /w/, /j/, and /l/, as these approximants can affect vowel quality. At times, participants’ vowels in these contexts were of insufficient quality as they were too short, unstressed, or demonstrated substantial creakiness and irregular periodicity. In this case, the same vowels were selected, but in different contexts. Three vowel occurrences, namely in ‘habit’, ‘logically’, and ‘spotted’, were in open syllables. The assumption was that this did not affect the vowel quality, only length. Vowel length was not taken into consideration. For the ‘lot’, ‘trap’, and ‘dress’ vowel, Praat was used; for the ‘bath’ vowel, auditory analysis was deemed sufficient, as only two variants needed to be distinguished. These two variants are classified as fronted ‘bath’ vowel [æ, æː] and back ‘bath’ vowel: [ɑ, ɑː]. In cases where the vowels did not quite fit the phonetic categories of [æ, æː] and [ɑ, ɑː], they were classified into the category they were judged to sound closer to. Table 9 summarises the vowel token selection:

Variable		Variant
‘bath’		fronted [æ, æː]
		back [ɑ, ɑː]
‘lot’		formant values
‘trap’
‘dress’

Table 9. Distinguished Variants for the Vowels

For the acoustic analysis, boundaries were placed near a well-formed period at the beginning and the end of each vowel in PRAAT. Segments were continuously voiced, and a minimum of 30 milliseconds long. Through a modified Praat script, automatically extracted averages for F1 and F2 of the selected vowels were generated. This average was computed over the most prominent area of the selected vowel space. The beginning and end-points of the vowels were not included, because their quality is often affected by the consonants surrounding them. Midpoint values, a customary procedure in formant analysis (Colantoni, Steele & Escudero 2015) were avoided, as this approach seemed less effective; some midpoints did not contain the most central or prominent quality of the vowel in question.

2.5 Interviews

To establish the identity associations attached to the various accents of English, informal interviews were held amongst four of the students who had already participated in the questionnaire.

3 Results

3.1 Students' Accent Preferences

Table 10 shows the accent preferences of the 99 students who filled in the survey:

Model Accent	Number of students
British accent model	70
US or North-American model	10
None	10
Mixed	6
Scottish model	2

Table 10: Model Accents of the 99 Students Surveyed

As shown below, a large majority of the students strove for a British accent model. A minority aspired for a U.S. or North-American accent. Those with ‘mixed’ models had multiple models, or changed accents depending on the setting.

The results in Table 11 show the average score on the graded scale from 0-6 of how strongly students agreed with the statements below, which have been arranged in order of preference (by the participants):

Statement	Label	Average Score
I want to sound like a native speaker.	Nativeness	5.1
I feel Dutch. I want to sound smart when I speak English.	Dutchness Intelligence	4.1 4.0
I feel close to the culture of my model accent.	Cultural alignment	3.6
I want to sound international when I speak English.	International outlook	3.1
I use my English accent to show who I am.	Self-expression	2.6
I want to sound cool when I speak English.	Coolness	2.5

Table 11: Participants' Pronunciation Goals (N=99)

Most of these criteria are identity-related, though nativeness and internationalness may be less explicitly so. Emulating the language of native speakers is ranked most highly. Feeling Dutch and wanting to be perceived as intelligent are also moderately important to students. Self-expression, a more personal aspect of identity expression, receives a lower score and seems relatively unimportant to students. Coolness scores lowest.

3.2 Indexicality

In informal interviews, four participants aiming toward a British accent were asked about associations connected to GA and RP. The reason for asking only these students was to keep the group of interviewees homogenous. From the interviews, a very traditional view of British and American English arises. British English signals prestige, intelligence, and beauty. Qualifications that were mentioned included: ‘snobbish’, ‘more refined’, ‘less neutral’, ‘very small group’, ‘elite’, ‘unemotional, cold’, ‘more intellectual or smart’, ‘thought through’, ‘fancy’, ‘elegant’, ‘beautiful’, ‘cultural’, ‘proper’, and ‘polite’. American English, on the other hand, seems to be considered less intelligent, less beautiful, yet a more neutral, common, and unmarked variety. Qualifications that were attached to GA were: ‘more neutral’, ‘accentless’, ‘more expressive’, ‘more international’, ‘easier’. These qualifications are, to a large degree, in line with Van der Haagen (1998). Another interesting finding was that the participants distinguished between classroom English and outside-classroom English. Adhering to pronunciation norms was perceived as less important outside the classroom, and three students reported that their accents changed depending on who they spoke with.

3.3 Students’ Pronunciation

We also looked at the realities of students' pronunciation, irrespective of their self-reported preferences. As mentioned in Section 3.2, the pronunciation of seventeen students’ was analysed more specifically, and these students all aimed toward a British accent. The results for the /t/ realisations are in Table 12:

fully pronounced [t, th,ˀt]	58
alveolar [ɾ, d̥, d]	26
no /t/ [ʔ, ∅]	6

Table 12: Medial /t/ Realisations (N=90)

The most recurrent realisation is the non-reduced one (a clearly audible /t/), which could denote a British accent or be a leftover from a Dutch accent. In Dutch, too, medial /t/ is pronounced fully and usually not glottalised or tapped (alveolarly). The tapped realisation could denote either a British or American accent, but it is mostly associated with an American accent (Van der Haagen 1998: 16). The occurrences of glottalised /t/ are likely to be imitations of a British English accent, because they are less likely to occur in Dutch or GA. In fact, they are perceptually salient features of a certain style of British English.

The realisations of postvocalic /r/, not including those /r/s in linking contexts, are presented in Table 13. The realisations of postvocalic /r/ which occurred before a word that started with a vowel sound, and thus could be pronounced as linking /r/, are found in Table 14:

Postvocalic /r/
non-rhotic	97
rhotic	39

Table 13: Realisations of Postvocalic /r/ (N=136)

Linking /r/
non-rhotic	26
rhotic	8

Table 14: Realisations of Linking /r/ (N=34)

This particular group of students showed a relatively strong preference towards not realising the postvocalic /r, even in linking position, which suggests a tendency towards RP. Linking /r/s are generally pronounced in RP, so the participants were possibly overgeneralising a rule. While their pronunciation appears to be predominantly British, some influence that could be due to the GA norm could be traced in their pronunciations of /r/ as well, as not all postvocalic /r/s were left unpronounced. It should be noted that in very fast speech, postvocalic /r/ might not be realised in word-stressed words, even in rhotic accents.

Table 15 contains the data for the ‘bath’ vowel:

‘Bath’ Vowels
back [ɑ, ɑː]	50
fronted [æ, æː]	18

Table 15: The ‘bath’ vowels (N=68)

Most participants pronounced the ‘bath’ vowels with a more fronted realisation. Individual participants often showed variation in their pronunciation, switching from one pronunciation to the other.

In Table 16, the vowel formants of the participants are compared to available RP and GA values from previous research, namely Deterding (1997) for RP and Hillenbrand et al. (1995) for GA. The Deterding RP values in the table can be classed as more conservative RP, as they are older values that proceed from a BBC-based corpus. As discussed in Adank (2003: 12), higher F1 values are associated with more closed vowels, and lower F1 vowels are associated with more open vowels. Conversely, higher F2 values are associated with more fronted vowels and lower F2 values are associated with vowels further back in the mouth:

		Men		Women
		F1	F2	F1	F2
‘dress’	RP, 1980s	494	1650	719	2063
	GA, 1990s	580	1799	731	2058
	Our study	532	1717	684	1940
‘trap’	RP, 1980s	690	1550	1018	1799
	GA, 1990s	588	1952	669	2349
	Our study	629	1579	836	1817
‘lot’	RP, 1980s	558	1047	751	1215
	GA, 1990s	768	1333	936	1551
	Our study	530	1092	616	1230

Table 16: Vowel formants, compared to RP/GA

In 10 of the 12 formant values (three vowels, two formants, women and men separately), the students’ formant values were closer to the RP model formants. In the remaining two formant values, they tended more towards GA than RP, namely the F1 for the ‘trap’ vowel for the men and the F2 for the ‘dress’ vowel for the women.

In general, it therefore appears that students sounded more British than American. This agrees with the students’ stated British accent preferences. However, variability is found among the recordings, where students do not always sound British, but also have traces of Dutch and what could be labelled as ‘American’ sounds in their accents. This is likely due to a lack of skill or awareness.

3.4 Expert ratings

The second author rated these seventeen students on how British, American, and Dutch they sounded, on the basis of native (for Dutch) and professional (for the two Englishes) knowledge of these varieties. The average results for the rated students are presented in Table 17:

	RP	GA	Dutch
Average	52%	8%	40%
St. Dev.	0.169	0.155	0.144

Table 17. Accent ratings of participant recordings (N=17)

These results corroborate that students sounded largely British. Students sounded British more than half of the time (52%). However, as can be seen in the table, there were large standard deviations among students. On average, there was a high amount of Dutch influence (40%), and students did not often sound American overall according to the rater (8%).

Such results, like other studies, show that participants succeed variably in accurately emulating their accent model. Besides looking at the degree of success of imitating a strict model, it seems that they were also avoiding sounding like Dutchmen or like native speakers of American English. The latter may show their allegiance to British English, as one might expect a move towards more international or American pronunciation choices for those students who were less successful in approximating British English.

4 Conclusions

4.1 Answers to the Research Questions

With respect to Research Question 1, i.e. the accents students aspire to speak, Table 10 shows that a majority of the students aspired to speak with a British accent. Some students mostly preferred a more Americanised or international accent (Table 10). Students aspired very strongly to nativeness (Table 11).

With reference to Research Question 2, i.e. the lexicalities attached to various accents of English, our interviews demonstrated that students associated RP strongly with more intelligence and status, and viewed GA as a more neutral and common variety.

In response to Research Question 3, referring to the accent features of students’ English, the seventeen participants who attempted to approximate the RP pronunciation model on average showed a dominance of this model in their accents. However, considerable variation existed amongst participants, with participants mixing accents and including accent choices not typically RP, such as flapping their /t/s and including postvocalic /r/s.

4.2 The Role of Identity

As Tajfel (1974) noted, social identity is linked to group membership and the value individuals place on belonging to those groups. These participants’ role as future professionals is based on identity or belonging in specific ways. There are three groups these participants associate with.

First, it appears that they value being part of an academic group and signal their membership to that group through an accent that agrees with the Dutch norm and which signals intelligence and professionalism. As found in the interviews, their accent may vary outside the classroom. Therefore, it appears that they associate strong language norms with formal and professional contexts. This emphasises the association with professionalism.

Secondly, they signal their belonging to the group that can be defined on the basis of specific expertise, namely (future) English experts and Anglophiles. As future professionals in the field of English, they desire to have an accent similar to that of a native speaker. Moreover, these are participants who have a certain love or appreciation for the English language and may therefore choose an accent model that they consider more ‘beautiful’. Three of the four participants interviewed associated the British English accent with esthetic criteria, calling it ‘beautiful’, ‘a little more beautiful’, or ‘elegant and correct’.

While signalling membership with the aforementioned groups, it can be argued that these participants do not emphasise their personal identity. Though they stress personal characteristics, such as intelligence and competence, these can be argued to be shared with other individuals, and these are thus not emphasising uniqueness. They seem, in fact, to be hiding much of their own personality in their accent choices. Emulating pronunciation norms is high on these students’ agenda and in that respect for their desired pronunciation behaviour may be perceived as a type of cosplay: when they speak English in professional contexts, they take on the accent of native speakers, but only in that context.

Similar studies in the Netherlands (Edwards 2016, Broersma 2015, Van der Haagen 1998) showed that Leiden students were particularly conservative in their tendency towards RP. It is true that Leiden University is considered a slightly elitist university in the Netherlands. Important political figures have attended this institution, and the city boasts a strong sorority and fraternity culture.

4.3 Consequences for English Language Teaching

The question remains how useful it is that students of English at Leiden University seem to gravitate to RP English when the current Dutch environment seems in flux, and preferences are changing. Those students who will become teachers may find themselves teaching an audience who are aspiring to a different accent: teenagers who are strongly influenced by the media and would rather shirk old-fashioned norms. A disconnect may then occur between the teacher and their students. In this globalising world, where people are in increasing contact with those of other cultures, it may be inevitable that those who speak English as an L2 will switch accents on a regular basis. It may then be best to teach them traditional associations with these models, while simultaneously questioning these associations and instilling critical attitudes in students.

Finally, there is another question which also arises from this study. Many of the English students do not consistently apply one accent when speaking English. This inconsistency will also enter the classroom. The question arises whether a teacher had best appropriate one consistent accent when teaching, and whether switching would be detrimental for students’ learning progress. Or perhaps this switching by the teacher is reflective of realistic L2 communication. This mixing is not negative per se, if intelligibility is not hindered.

Mixing rather than committing to strict pronunciation models fits well in the era that is postmodernism. This period is often said to have started in the late 20^th century, and it is characterised by principles such as scepticism, individualism, inter-cultural awareness, and the right to self-realisation (Pennycook 2018). Pronunciation choices by today’s students are obviously affected by these principles. Conscious mixing or allowing mixing to take place is in line with postmodern principles. Committing to one accent means committing to the indexicalities of a certain culture, and this is likely to lead to increasing discomfort amongst today’s learners, especially now that informal exposure to a myriad of accents is increasing.

Teaching about variation and connotations of accents but also about individual sounds could be considered more important than teaching a one-accent-for-all approach and expecting this specific accent (and register) to work in all social contexts. First-language influence may be presented as not only inevitable but also a tool to sound natural and authentic. This mixing may be viewed as an attempt towards the construction of an authentic accent, rather than simply being a result of incompetence. This new approach will help to avoid the common ridicule of ‘foreign’ accents and improve the status of accented English. The language classroom may be seen as a place where sounds and connotations are offered as well as a basic pronunciation model that meets the norms of intelligibility. It is up to students to construct their own intelligible accent, which will meet their own, personalised demands in the specific contexts in which they speak English. Inclusivity is the end result of such an approach.

References

Adank, Patti. 2003. "Vowel Normalization. A perceptual-acoustic Study of Dutch Vowels." Ph.D., Department of Linguistics, Nijmegen University.

Praat: Doing phonetics by computer 6.0.14.

Botma, Bert, Koen Dirk Corné Jac Sebregts, and Dick Smakman. 2012. "The phonetics and phonology of Dutch mid-vowels before /l/." Laboratory Phonology 2:273-299.

Broersma, N.B. 2015. Accent goals, attitudes, and accent realization of English majors at the University of Groningen (BA thesis). Groningen: University of Groningen.

Bucholtz, Mary, and Kira Hall. 2005. "Identity and Interaction. A Sociocultural Linguistic Approach." Discourse Studies 7:585-614.

Bucholtz, Mary, and Kira Hall. 2010. "Locating identity in Language." In Language and Identities, edited by Carmen Llamas and Dominic Watt, 18-28. Edinburgh: Edinburgh University Press.

Burck, C. 2011. "Living in several languages: Language, gender and identities." European Journal of Women's Studies 18 (4):361-378.

Colantoni, L., J. Steele, and P. Escudero. 2015. Second Language Speech: Theory and Practice Cambridge: Cambridge University Press.

Coupland, Nikolas, and H. Bishop. 2007. "Ideologised values for British accents." Journal of Sociolinguistics 11 (1):74-93.

Dent, Stacey R. 2004. Attitudes of native and nonnative speakers of English toward various regional and social U.S. English accents (MA Thesis). Iowa: Iowa State University.

Deterding, D. 1997. "The Formants of Monophthong Vowels in Standard Southern British English Pronunciation " Journal of the International Phonetic Association 27. (1-2):47.

Eckert, Penelope. 2000. Linguistic Variation as Social Practice: The Linguistic Construction of Belten High. Malden: Blackwell.

Eckert, Penelope. 2008. "Variation and the indexical field." Journal of Sociolinguistics 12 (4):453-476.

Edwards, A. 2016. English in the Netherlands : Functions, Forms and Attitudes. Amsterdam: John Benjamins Publishing Company.

Geeslin, Kimberly L., and Avizia Yim Long. 2014. Sociolinguistics and Second Language Acquisition: Learning to Use Language in Context New York: Routledge.

Hillenbrand, J., L.A. Getty, M.J. Clark, and K. Wheeler. 1995. "Acoustic characteristics of American English vowels." The Journal of the Acoustical Society of America 97 (5):3099-3111.

Hitlin, S. 2003. " Values as the Core of Personal Identity: Drawing Links between Two Theories of Self." Social Psychology Quarterly 66 (2):118-137.

Hughes, Arthur, Peter Trudgill, and Dominic Watt. 2005. English accents and dialects: An introduction to social and regional varieties of English in the British Isles. London: Hodder Arnold.

Joseph, John E. 2010. "Identity." In Language and Identities, edited by Carmen Llamas and Dominic Watt, 9-17. Edinburgh: Edinburgh University Press.

Ladegaard, Hans J., and Itesh Sachdev. 2006. "‘I Like the Americans... but I certainly don't aim for an American accent’. Language attitudes, vitality and foreign language learning in Denmark." Journal of Multilingual and Multicultural Development 27 (2):91-108.

Mair, C. 2006. Twentieth-Century English. New York: Cambridge University Press

Mugglestone, L. 1995. 'Talking Proper': The Rise of Accent as a Social Symbol Oxford:: Clarendon Press.

Norton, Bonny, and K. Toohey. 2011. "Identity, language learning, and social change " Language Teaching 44 (4):412–446.

Ortega, L. 2009. Understanding Second Language Acquisition London/New York: Routledge.

Pelham, B. W., and H. Blanton. 2007. Conducting research in psychology. Measuring the weight of smoke Belmont: Thomson Wadsworth.

Pennycook, Alastair. 2009. "Plurilithic Englishes: Towards a 3D model." In Global Englishes in Asian contexts, edited by Kumiko Murata and Jennifer Jenkins, 194-207. Basingstoke: Palgrave MacMillan.

Pennycook, Alastair. 2018. Posthumanist Applied Linguistics. Abingdon/New York: Routledge.

Smakman, Dick. 2006. "Standard Dutch in the Netherlands. A Sociolinguistic and Phonetic Description." PhD, Landelijke Onderzoekschool Taalwetenschap (LOT).

Smakman, Dick. 2015. Accent Building. A British English pronunciation course for speakers of Dutch. Leiden: Leiden University Press.

Tajfel, H. 1974. "Social identity and intergroup behaviour." Social Science Information 13:65-93.

Van der Haagen, Monique. 1998. Caught Between Norms. The English Pronunciation of Dutch Learners. Utrecht: Landelijke Onderzoekschool Taalwetenschap.

Wells, John C. 1982. Accents of English 3. Beyond the British Isles. Cambridge: Cambridge University Press.

Authors:

Gezina Christien Dorothé Huttenga

Hanze University of Applied Sciences

Zernikeplein 7

9747 AS

Groningen

Netherlands

Email: geeskehutt@protonmail.com

Dr. Dick Smakman

Leiden University Centre for Linguistics (LUCL)

Arsenaalstraat 1

2311 CT

Leiden

Netherlands

Email d.smakman@hum.leidenuniv.nl

____________________

1 In the tables below, round brackets denote syllable boundaries, subscript /t/ or /r/ denotes the position of /t/ or /r/ in the syllable, ‘V’ denotes ‘vowel’, ‘C’ denotes ‘consonant’, ‘-’ denotes ‘any phoneme’, and ‘#’ denotes word boundary.