Keywords

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

1 Introduction

1.1 English Teachers in Sweden

It has been suggested that English in Sweden has some of the characteristics of a second language rather than a foreign language (SOU, 2002, p. 27; Hyltenstam, 2004, p. 52). Children in Sweden begin learning English in the first years of primary school and continue until they are in upper secondary school. English language proficiency (evidenced by passing grades in upper secondary courses in English) is part of the school leaving qualification and the requirement for university entrance. To this it could be added that English-speaking cultures are highly regarded by Swedish young people (Sundqvist, 2009; Sundqvist & Olin-Scheller, 2013), and that Swedish is a small language spoken almost exclusively in Sweden and parts of Finland, so a language of wider communication is seen as an essential part of the education of young Swedes. Consequently, Swedes are generally held to have a high level of proficiency. Hyltenstam (2004, pp. 53–54) attributes this to four factors: Swedish people frequently travelling abroad; the frequent use of English in Swedish media and the availability of media in English; Swedish people being interested in learning English and the fact that Swedish and English are related languages, making the learning of English fairly easy for Swedish speakers.

Primary school teachers in Sweden are expected to teach their classes all parts of the curriculum, including English. Nonetheless, until recently, primary teacher education in Sweden did not require students to study English. The latest primary teacher education programme requires students preparing to teach 6–10 year olds to have 15 ECTS credits of English at university, equivalent to 10 weeks fulltime study. This means that there are a large number of primary school teachers in Swedish schools who, while they are well able to understand written and spoken English, and to speak fairly fluently at the drop of a hat, have strong Swedish accents and are unable to write grammatically.

1.2 English Pronunciation in Swedish Schools

The Swedish curriculum for the school subject of English is, from the earliest stages to upper secondary, entirely based on the learners developing communicative competence in English. Students can be given top grades based on their ability to use English, with barely a nod to the grammar and norms of standard written English or the phonology of a native variety. In Swedish schools, there is no wish for a return to the prescriptive classrooms of yesteryear—communication is king! Communication is, in accordance with the intentions expressed in the Common European Framework of reference for Languages (CEFR: Council of Europe, 2011), the only reasonable priority for the teaching of English in Swedish schools. Those who continue studying English at university learn about and are expected to aspire to grammatical accuracy and the norms of native-like pronunciations. Teachers, as the main model for learners in the classroom, need to speak and write accurate English (even though, as Sundqvist (2009) established, Swedish young people hear a lot of English outside the classroom). Nonetheless, primary teachers, who may have high grades in English from upper secondary courses but have not studied English at university, will not necessarily be aware that their English is other than “awesome.”

University lecturers who teach the first year English courses often bemoan this; educational politics mean that the threshold for entrance to university study of English is low. Many students do not make it through the first semester of study and this has led to attempts to assess the needs of the new entrants and help them to improve their grammatical and pronunciation accuracy. However, neither a strong Swedish accent nor inaccurate grammar will necessarily lead to failure in university English, provided the student is able to communicate well. Once they have managed to pass the first semester courses, very little emphasis is placed on their English language proficiency. As proficiency is not the focus of the courses, by regulation, it is not allowed to be the focus of the learning outcomes or the grading criteria. This leads to some students coming to language teacher education after several years of university English, with a view to becoming teachers of English at upper secondary school, yet quite unable to speak without grammatical error and a strong Swedish accent. The same mechanisms apply to learning outcomes and grading criteria in language education as in courses in English linguistics and English literature. Even if the course is taught in English, and students use English in their assignments and in class, their proficiency, and still less, their accuracy (aka native-likeness) is not a learning outcome and may not be the basis for assessment.

There is a degree of doublethink in operation here. On the one hand, liberal forces argue that there is nothing wrong with a Swedish accent in English and that transfer from Swedish grammar to English is a normal language contact phenomenon, and that the resulting forms are characteristics of a Swedish variety of English. This might work for the curriculum for schools and the requirements for teenagers to pass English at school, but, on the other hand, few parents and principals want teachers to teach Swedish English.

In the 1980s, Swedish universities accepted only RP as a target for English pronunciation. This had some absurd consequences such as native speakers of American English failing exams on the basis of their failure to pronounce words and read texts in an adequate approximation to RP. The universities now accept other native accents as targets for English pronunciation, but a heavy Swedish accent is still frowned upon in university departments of English, in much the same way as failure to master subject-verb agreement is.

In schools, however, the Swedish curriculum for English has stepped away from any kind of pronunciation teaching. Pronunciation and intonation are actually specified content in the syllabus from year 4–6 onwards, but the grading criteria make no mention of anything more demanding than clarity in oral production, even for the highest grades in year 6. By year 9 the highest grade also requires fluency of speech.

Naturally, this situation means that pronunciation is not prioritized in Swedish schools. Teachers have no reason or remit to guide students towards native-like pronunciation as long as their speech is clear and reasonably fluent. This is, in fact, entirely in line with the personal targets of many young Swedes. As English approaches second language status in Sweden (Hyltenstam, 2004), it is entirely reasonable that Swedish-accented speech be afforded the respect due to any regional accent of English. In other words, if it is acceptable for a teacher to speak English with a Northern Irish, Scottish or New Zealand accent, none of which can lay any great claim to being clear or intelligible, why should it not be acceptable for a Swedish speaker to speak English with a Swedish accent rather than aspire to fool the listener into believing that they grew up in Oxford or Ohio? Furthermore, some learners take pronunciation teaching as much of a personal affront as a Northern Irish person might if urged to work a bit harder at upholding the GOOSE-FOOT distinction. If the native speaker of such a variety can argue that this distinction happens not to be part of their phonological system, and that speaking with an accent is an expression of identity, how can anyone insist that Swedish students work at distinguishing ice and eyes?

This increasing acceptance of non-native pronunciation is by no means confined to Sweden. Certainly, Tergujeff (2013, p. 84) found a young Finn who reported not wanting to speak English without an accent, saying (in her translation) No, it wouldn’t be nice. I want to emphasise that I’m not British but a Finn. Buckingham (2014) found that Omani learners responded favourably to both British accents and Arabic accents in teachers. Nonetheless, this position is not universally accepted by all stakeholders. Prescriptive forces, including some students and their parents, as well as many teachers, insist that learners are given the opportunity to aspire to native-speaking targets, usually British or American unless there is a compelling reason, such as the student having spent time in another English-speaking environment. This means that there is a case for encouraging teachers to work on the most salient Swedish-accented features of their English pronunciation. Cunningham (2009), Henderson et al. (2012), D. Murphy (2011), J. M. Murphy (2014), Smith (2011) and Van den Doel (2008) also discuss the question of whether a particular variety of native or non-native English is a good model, and the answer has to be that it depends entirely on the beliefs and targets of the learners, and that learners should probably be allowed to choose their targets as well as their models.

1.3 Swedish Accent of English

Even a strong Swedish accent does not often lead to any real lack of clarity. Swedish has an inverse temporal relationship between vowel and consonant length, such that short vowels are followed by long consonants e.g. vitt, and long vowels are followed by short consonants, e.g. vit. This means that Swedish speakers often transfer this relationship to English. Also, Swedish has dentals where English has alveolars (/t, d, n, s/) and no /z/. Swedish has no affricates and its voiceless fricatives, while several in number, do not include a palatoalveolar that occurs initially like the English shoe. Swedish does not use pre-fortis clipping as a cue to post-vocalic voicing as English does, so the Swedish accent will not make the vowel in bid longer than the vowel in bit. In addition, vowels are often confused or distinguished using cues that are not salient to non-Swedish listeners.

1.4 Teaching Pronunciation Online

Online language teaching is not new, although universities have not, generally been among the first to move in this area. Ubiquitous connectivity has led to a range of more or less genuine operations offering to connect teachers and learners for private tuition. Tertiary distance learning has a long tradition of text-based courses, which clearly do not lend themselves to modern communicative spoken language learning. Videoconferencing would have offered a reasonable step up if the technical requirements were not so expensive and complex. The realization that the development of synchronous tools such as chat and voice chat such as Skype could be applied to learning met some resistance by those who had not tried or who had faced technical challenges. Desktop video conferencing through e.g. Adobe Connect is a huge step forward and, even though there is a learning curve for teachers and students alike, there are advantages in the multimodal affordances of these systems (cf. Cunningham et al., 2010; Cunningham, 2011). I would argue that effective language learning and teaching requires synchronous communication. Others go a step further into virtual space, to create liminal experiences such as those in an integrative environment such as Traveler or more recently, Second Life (Sobkowiak, 2012). One advantage of this is that an environment is created where tools, including games, for teaching can be collected and used.

2 The Course

2.1 Course

The course that is the focus of this study was not in any kind of rich 3D environment, and there were compelling reasons for keeping the synchronous elements to a minimum. The stated aim of the course was that students would develop their awareness of common “problem areas” in English as well as their ability to work communicatively with grammar and pronunciation in their teaching at different school levels.

The course was structured on the University learning platform as a series of ten weekly packages of tasks including two written hand-ins and six oral hand-ins. Each week there were web-lectures or videos to watch, podcasts to listen to, texts to read, sound recordings to make, texts to write, forums to read and interact in. Each week’s package included tasks in each of grammar, pronunciation and language education. The course ended with the second written assignment, a lesson plan working communicatively with a formal aspect of the language and a real-time group seminar in Adobe Connect where five students discussed each other’s lesson plans according to a pre-arranged schedule.

2.2 Students

This was a summer course, offered to active teachers and teacher students who were so concerned about their professional development that they were prepared to give up some of their free time over the long vacation to work on their English and their teaching of English. Many of them took part from holiday homes or boats and they may not have had much access to fast broadband. For that reason, an effort was made to limit the need for synchronous communication. One hundred and fifty students were registered on the course, 111 handed in the first assignment and 91 of them completed the course, 73 with passing grades. This is a fairly good throughput for an online summer course. European students do not pay tuition fees in Sweden, and many sign up for courses like this that they never really start. There are no repercussions for students who do not start or who drop out of courses.

2.3 Course Materials and Technologies

As well as helping teachers to teach English using a communicative approach, the course had a second, somewhat covert aim: it was designed to raise the students’ own proficiency in English. As indicated above, university structures meant that this course aim could not be clearly expressed. A textbook (Cunningham, 2013) was written for the course, to give a brief explanation of twenty of the most noticeable features of Swedish speakers’ English with a view to helping users to work on these “easy targets” for improvement. Eleven of these twenty features were grammatical, and the remaining nine were:

  • Confusion of words like eyes and ice

  • Confusion of the vowels of man, men, main

  • Confusion of the vowels of hot, hut, heart

  • Confusion of the vowels in sheep and ship

  • Confusion of the initial sounds of sheep and cheap

  • Interdentals

  • /v/ and /w/

  • Confusion of the initial sounds of yes and Jess

  • Difficulty expressing emphasis and contrast prosodically.

While these features do not typically hinder intelligibility, they do create a strong impression of a Swedish accent. As well as the textbook, web-based lectures were prepared for the students as mp4-files using TechSmith’s screen capture and video editing software Camtasia and Snagit. These were about course administration, e.g. on how to activate the course books’ web material, or about aspects of language education, e.g. how to use rhymes and songs to teach pronunciation. There were also audio lectures which accompanied another textbook on the publisher’s website, audio lectures produced to accompany Cunningham (2013) and audio and other material from external sources such as the BBC Learning English website or the British Council website. The students were introduced to TechSmith’s free screen capture software, Jing, which they would need to use to record their oral hand-ins. This was chosen as it allowed them to have a text on the screen that they are talking about as they speak, and they and their own students could use it later at no cost.

The course was centred on the learning platform. There were ten folders in the course resources, one for each week of the course. These each contained a hypertext document (such as that shown in Fig. 1). This document linked the students to course materials and to other sites that they were asked to access. The course materials were made available to the students a week before the beginning of the course week, and remained available until the end of the course. The students were assigned to groups and each group had a forum on the learning platform. They were asked to upload links to their recorded films (hosted in TechSmith’s own Screencast servers) to the forum and to view and comment on each other’s work. Towards the end of the course, the students signed up in groups of five on a schedule for the real-time seminars that were part of the examination of the course. Students were told they needed a headset or earbuds to take part in the Adobe Connect sessions.

Fig. 1
figure 1

Weekly instructions for the course

Full size image

2.4 Strategies

The nine pronunciation chapters of Cunningham (2013) (the themes of these are listed above) were used as the basis for the pronunciation part of the course. The following strategies were used in the course.

2.4.1 Needs Analysis

The publisher’s website accompanying Cunningham (2013) includes a simple web-based, automatically corrected multiple choice grammar test, which served as a needs analysis for the grammar part of the course. In addition, students were asked to record themselves reading a short text aloud and introducing themselves and their reasons for taking the course. These sound recordings were analysed for the presence of any of the nine target pronunciations, and anything else particularly striking. This formed the basis of a pronunciation needs analysis.

2.4.2 Individual Feedback from the Needs Analysis

The students were given individual feedback on their pronunciation based on the analysis of their pronunciation in the first oral hand-in. They were told which of the pronunciation chapters in the textbook they would need to work most with. Other feedback was also given on anything that did not fit into any of the chapters of the textbook (Cunningham, 2013). See Results below for further detail of the feedback given.

2.4.3 Texts Explaining the Pronunciation of Specific Sounds

The chapters of the textbook contained detailed explanation of these common features of Swedish-accented English. These explanations often took similar Swedish sounds as the departure point for the description of an English sound, such that the English word far was described as being like the Swedish word far but without the strong lip-rounding of the Swedish vowel and with a shorter vowel and no pronunciation of the final 〈r〉. IPA transcription was not used.

2.4.4 Web-Based Audio Lectures on these Texts

The web version of each of the pronunciation chapters in the textbook includes a sound file with the examples given in the text. In addition, an audio lecture was prepared to talk the student through each chapter.

2.4.5 Web-Based Lectures on English Articulatory Phonetics and Reading the IPA

Three radio programmes about pronunciation were linked in as weekly activities in some weeks for the students. These are part of extensive pronunciation learning material from the BBC Learning English site. In addition, this site has videos about each phoneme in RP with example words, and materials for practicing the IPA.

2.4.6 Perception Practice Activities

The students were asked to practice listening to problematic sounds in exercises where students can practice their perceptual skills. Students are asked to listen to words being read and to pick out the one different word in triads of words, such as man, main, man.

2.4.7 Production Practice Activities

One of the oral assignments that the students do is to record themselves reading a text, an excerpt from either Winnie the Pooh or Harry Potter, as though to children or adolescents. This involves not only fluency and reasonable segmental pronunciation, but also engagement and good understanding of the text.

Another activity involved the participants producing a 60-s idea to change the world, based on the BBC radio programme of the same name.

2.4.8 Peer Feedback

The students gave each other feedback, not directly on pronunciation, but on their thoughts about reading aloud to learners as a meaningful language teaching activity. In fact, some of the students did spontaneously give each other positive feedback on pronunciation. The only times negative feedback was given between peers was when the students totally mispronounced a word.

2.4.9 Second Feedback Session

Near the end of the course, as the sixth oral hand-in, the students rerecorded themselves reading the text they read and received individual feedback on at the beginning of the course. They were invited to compare the recordings and to reflect on any improvement in pronunciation they have noticed, recording their reflections as well. Then they were given feedback again on these new recordings.

3 Results

3.1 Student Experience

In the anonymous web-based end-of-course evaluation, all 47 of the students who responded expressed finding the web-based lectures helpful or very helpful. Comments were that they appreciated the lectures being short and accessible. Similarly, the course book (Cunningham, 2013) and its associated online material were found useful and easy to understand.

The students expressed frustration at the limited amount of feedback they received from the teachers and did not generally appreciate the peer feedback, although some groups worked better than others. Some students felt that it was awkward to have both inexperienced teacher students and teachers with many years of classroom experience in the same groups. The students enjoyed the immediate response from self-correcting quizzes that were used to practice differentiating between similar sounds.

3.2 Outcome of Course

In the first oral hand-in, the students were asked to record themselves reading a short text (The North Wind and the Sun) and then to comment on their own pronunciation and how they felt about it. The sixth and final oral hand-in, about 8 weeks later, had them revisit the first recording and the feedback received then and record a new reading of the same text and reflections on any perceived progress made in pronunciation.

There were three kinds of pronunciation feedback given to the students on the marking template used for oral hand-in 1. First, there was a matrix, indicating the nine pronunciation chapters of the main course textbook (Cunningham, 2013), shown here as Fig. 2. Any feature the student had trouble with was highlighted.

Fig. 2
figure 2

Template used in feedback to indicate chapters in the course book dealing with pronunciation features that students needed to work with

Full size image

Second, there was a heading called Other issues intended for any comments that did not come under any of the chapters of Cunningham (2013). Third, there was a General comment heading for more general comments, e.g. about tempo, suprasegmental features of the student’s pronunciation or intelligibility or some matter of content mentioned by the student, e.g. about the classes they teach.

The twelve students who received the lowest passing grade, E, for the course were selected to represent the pronunciation difficulties and feedback experienced by the students on the course. These students participated fully in the course, and while there may have been different reasons for them not achieving a higher grade on the course, between them they received most of the pronunciation feedback that was given to the class under the heading Other issues. Their results are shown in Table 1.

Table 1 Uptake in oral hand-in 2 of feedback from oral hand-in 6 for students receiving grade E passes for the course
Full size table

Some of the most frequent comments were feedback on the pronunciation of specific words, like pronunciation or obliged. Other comments were on the pronunciation of specific vowels. Table 1 shows that 17 specific word comments were taken up by the students, while five were not taken up, meaning that the students continued to mispronounce these same words in the second recording of the text and/or their comments on their pronunciation. In this very limited data set, that is a 77 % uptake rate, which is quite satisfactory. Other kinds of feedback, regarding the pronunciation of sounds that occurred in several words, were less successful, except in the cases where the feedback was illustrated in particular words, such as the non-rounded final vowel in immediately and closely and English words like first where /s/ is preceded by /r/ being pronounced with a retroflex voiceless fricative. In these cases it is possible that the feedback is processed as word-specific feedback. General feedback on individual sounds, such as the interdental fricatives, or over-rounded vowels, was not usually taken up by the speakers.

The uptake of the marking of specific chapters for individual students was not analysed as all the students were in any case asked to work through all the chapters in the book. However, part of the final oral assignment was for the students to reflect on any development they were aware of in their own pronunciation. Again limiting the sample to the twelve students who received the lowest passing grade, a number of themes emerged:

First, students expressed gaining in confidence from being asked to prepare and record oral presentations of various kinds:

I feel like my English pronunciation has improved since I entered this course. (S2)

I feel a bit more confident and I feel I have developed. Before it was quite a while ago that I spoke English. So that was good with all the oral assignments that we have had to practice. (S10)

I feel much better confidence speaking English aloud. (S12)

Second, students felt they had learned about the pronunciation of specific words:

I feel like I can pronounce words like first and wind better than before. (S2)

I used to say pronunciation (S3)

Now I know better. (S7)

I did not really know that but now I know and will think about it. (S9)

I got more knowledge to why you are to pronounce something in a certain way. (S9)

Third, students felt they had learned strategies for continuing to develop their pronunciation:

Shaping my tongue is a problem, but I am practicing. Now I am really listening to pronunciation, and repeating. (S12)

Finally, fourth, they were motivated to continue working on pronunciation:

I will practice my pronunciation in the future and hope that in my future teaching that the students understand me and learn how to pronounce in the right way. (S2)

I will continue learning for ever until I sound like, as close as possible, to a native speaker. (S9)

I guess I just have to keep on practicing and reading English books. (S10)

My pronunciation still needs a lot of work. (S11)

4 Conclusions

It appears that the students gained in their explicit knowledge about English pronunciation through the course, but that their implicit knowledge, as evidenced in their own improved pronunciation in spontaneous speech was relatively unaffected by the course. This is supported by the fact that several students self-corrected for the pronunciation of some of the individual words they had been given information about, both in their second reading of the text, and in their spontaneous speech as they recorded their reflections. In addition, learning the pronunciation of individual words was mentioned by several of the students in their second recordings.

The relationship between explicit teaching and implicit knowledge is not well understood. In the kind of course that has been in focus in this study, the very act of having to speak English, even if it is without an interlocutor, is likely to improve the students’ proficiency. Individual feedback given some time after a recording is made may be useful only as information about the pronunciation of specific words, rather than being available to affect the phonological system of the learner’s interlanguage.

The advice given to students who needed to continue improving their pronunciation was to work holistically, listening extensively to English, e.g. in the form of podcasts on (non-language oriented) topics that are meaningful to them and sometimes stopping to pay attention to accents and to repeat aloud the pronunciation of words and phrases.