1The role of pronunciation instruction in language learning has had a tumultuous history, with alternating bouts of extremism and indifference. For example, under the Audiolingual Method, pronunciation instruction had an ultimate goal of developing language learners’ speech to match that of native speech (Celce-Murcia et al., 2010). After this, pronunciation instruction tumbled from its peak of importance during the ascent of Communicative Language Teaching in the 1970s and 1980s, with researchers and teachers claiming that pronunciation instruction was unnecessary for communicative ability (Levis & Sonsaat, 2017). Consequently, pronunciation also fell into neglect in teacher education, and many teachers completed their teacher preparation programs without adequate training in this area (Murphy, 1997). Today, due to increasing research demonstrating that pronunciation instruction almost always leads to improvement (e.g., Lee, Jang & Plonsky, 2015), and due to a focus on the intelligibility of speech rather than native-like achievement, we are seeing a revival of pronunciation instruction in language learning (Derwing & Munro: 2015; Levis, 2005). Even though pronunciation is more visible today than perhaps ever before (Derwing, 2019; Levis, 2019), we continue to find studies that report a lack of teacher preparedness and confidence in providing pronunciation instruction (e.g., Macdonald, 2002; Foote, Holtby & Derwing, 2011; Murphy, 2014).
2Without formal training, teachers often report that they teach pronunciation either as prescribed in the textbook, as they learned it when they were language learners, or they forgo instruction altogether (Derwing, 2019). This has led some researchers to consider the effects of formal training on a teacher’s preparedness for providing effective pronunciation instruction (e.g., Burri, 2016; Buss, 2017; Kochem, 2022). The results from these studies show a positive correlation between training and a teacher’s confidence in, and valuing of, pronunciation instruction. However, limited research has been done to show what effects training has on a teacher’s knowledge of pronunciation. As Burri and Baker (2020: 3) comment, “[t]o what extent L2 teachers apply—in their classrooms—knowledge and skills they acquired in a pronunciation teacher preparation setting, and how their cognition and practices develop after completing a course on pronunciation pedagogy remains largely unknown.” More specifically, it is unclear how well teacher trainees understand (1) the phonetic and phonological processes of spoken language, (2) learner errors and their detrimental effects, and (3) the development of instructional materials and activities to target specific pronunciation features. Because of this, the current study explores how practical training influences the evolving practices of teacher trainees.
3For adult second language learners, pronunciation learning is essential for increased intelligibility and communicative success (Derwing & Munro, 2015). This need matches well with findings that pronunciation can and does improve with both face-to-face and technology-based instruction (Lee et al., 2015). It is also clear that such improvements in pronunciation can lead to increased comprehensibility, especially when pronunciation instruction includes a focus on suprasegmentals and global speech features (Derwing, Munro & Wiebe, 1998; Zhang & Yuan, 2020). However, pronunciation instruction is often neglected in language classrooms for a variety of reasons. It may not be taught consistently because:
1. it may be considered peripheral to a communicative approach to language teaching (Levis & Sonsaat, 2017),
2. it is not adequately assessed in formal examinations required by the educational system (Isaacs & Trofimovich, 2016),
3. it is not a prominent part of published materials (Derwing, Diepenbroek, & Foote, 2012),
4. teachers do not feel prepared to teach it due to lack of training or lack of confidence (Murphy 2014, 2017), or,
5. they do not believe that their own pronunciation is an appropriate model for learners (Phuong, 2021).
However, when teachers have the opportunity to teach in their pronunciation training, their teaching practices often develop in surprising ways as they respond to the needs of the learners.
4In 1969, L.G. Kelly stated that pronunciation had finally emerged from its long role of being a Cinderella in language teaching, at that time being considered of equal importance with grammar, morphology, and other central aspects of linguistic description and grammatical (or linguistic) competence. However, the advent of the communicative era changed the rules of the language teaching world, and it became expected that L2 pronunciation would largely take care of itself if teachers and learners focused on learning to communicate (Levis & Sonsaat, 2017). As a result, pronunciation fell into a new neglect in the language teaching world. The new focus on communicative competence (Hymes, 1972; Canale & Swain, 1981), of which linguistic competence was only a part, filled the language teaching world with needed new goals such as teaching appropriate language (sociolinguistic competence), learning authentic language that is cohesive and coherent (discourse competence), and learning how to communicate when linguistic resources were inadequate (strategic competence).
5A further consequence of these changes was that pronunciation, with its focus on accuracy, was no longer consistently part of many teacher education programs (Murphy, 1997), and many teachers did not receive adequate training on why and how to teach pronunciation. As a result, many language teachers did not teach pronunciation proactively because pronunciation was not part of their language teaching toolbox. If pronunciation errors became too prominent to ignore, teachers could always address them in an ad hoc fashion, correcting errors that either affected clarity or that were otherwise irritating. This ad hoc approach did not require systematic attention to pronunciation teaching and learning.
6By the early 1980s, evidence was mounting that pronunciation did not take care of itself in communicative language teaching, and that it was often necessary to systematically address pronunciation for intelligibility to improve. Hinofotis and Bailey (1981), studying international teaching assistants (ITAs), proposed that there existed a threshold for intelligibility, and that unless L2 speakers met the threshold, they could not be understood no matter how advanced they were in other aspects of communicative competence. This groundbreaking research on ITAs was followed up by other research demonstrating that ITA communicative challenges included both the construction of discourse through grammatical choices (Tyler, 1992; Williams, 1992) and pronunciation difficulties (Anderson-Hsieh, 1992; Gallego, 1990). Additionally, although pronunciation was often marginalized during this time, communicative language teaching contributed to other changes in pronunciation instruction by the mid-1980s, when there was a growing consensus on the importance of suprasegmentals for pronunciation learning, leading to innovative exercises that had a better fit with the principles of CLT (Kenworthy, 1987; McNerney & Mendelsohn, 1987; Wong, 1987).
7The importance of pronunciation instruction has continued to grow among teachers, and with the growth of empirical research on pronunciation’s role in evaluations of speech related to intelligibility, comprehensibility and accentedness (Jenkins, 2000; Munro & Derwing, 1995) and studies of teacher beliefs about pronunciation (Breitkreutz, Derwing & Rossiter, 2001; Burgess & Spencer, 2000; Foote et al., 2011), pronunciation instruction once again has become a Belle of the Ball in the language teaching world (Derwing, 2019). However, despite growth in research about teacher beliefs, we still know very little about what teachers actually do when teaching pronunciation.
8Studies in language teacher cognitions (e.g., attitudes, beliefs, and knowledge) continue to show that teachers lack confidence in their ability to teach pronunciation (Baker, 2014; Couper, 2017; Henderson et al., 2015), so the idea of incorporating technology into pronunciation instruction seems far-fetched for many language teachers. However, a myriad of studies have shown that incorporating technology into pronunciation instruction improves learning (e.g., Al-Qudah, 2012; McCrocklin, 2014; Neri, Cucchiarini, & Strik, 2003), which leaves us with the question of how to prepare teachers to use technology when teaching pronunciation. As researchers have stated for decades (e.g., Reinders, 2009; Torsani, 2016; Warschauer, 1997), the ability to use educational technology does not directly translate into an ability to teach with technology. The connection between pedagogy and technology led to the rise of techno-pedagogy (Guichon & Hauck, 2011), which explores the dynamic nature in which the two fields (i.e., pedagogy and technology) interact. Even with the understanding that direct instruction is needed for teachers to be able to incorporate technology effectively, professional development opportunities to develop the knowledge and skills needed are still somewhat limited (Kessler & Hubbard, 2017).
9As the lowest benchmark, training teachers to use technology for teaching pronunciation should include: (1) a dissemination of information, both empirical research and pedagogical practices, (2) methods for keeping track of all the new technologies, and (3) the importance of usable and accessible technology (Kochem et al., 2020). However, there seems to be a dearth of research exploring teachers’ knowledge and practices of incorporating technology in the teaching of pronunciation, although some scholars have suggested best practices and useful tools (e.g., Pennington & Rogerson-Revell, 2019; Yoshida, 2018). Therefore, understanding how training in pronunciation pedagogy and, more importantly, how training in technology-enhanced pronunciation pedagogy leads to teacher practices is needed.
10The purpose of this multiple case study was to explore the evolving teaching practices of teacher trainees enrolled in an L2 pronunciation pedagogy course. Special attention was given to three elements of the course: (1) the trainees’ ability to analyze and categorize learner errors in a diagnostic assignment; (2) the translation of coursework into actual teaching practices during a tutoring project; and (3) the incorporation of technology (or lack thereof) into the trainees’ tutoring sessions. This study is of relevance to both teacher educators and student teachers of L2 pronunciation pedagogy. First, it identifies how well student teachers are able to identify and address learner errors after taking an explicit course in pronunciation pedagogy. Second, it identifies what student teachers found to be most difficult when teaching pronunciation in a one-on-one setting. Finally, it reveals how technology was incorporated into the tutoring sessions. The findings affirm the need for more specialized training in L2 pronunciation pedagogy in teacher preparation programs (Murphy, 2014). In particular, this study addresses the following questions:
1. What role does learning to do diagnostic pronunciation assessments play in the development of pronunciation teaching skills?
2. How do teachers’ cognitions and practices evolve over a four-week tutoring project in English pronunciation?
3. In what ways is technology incorporated into the teaching of English pronunciation?
11Fifteen teacher trainees in a graduate-level L2 pronunciation pedagogy course at a large university in the Midwest US participated in the study. Table 1 provides demographic information related to their speaker status, teaching experience (including languages and levels taught), and degree currently sought during the course. The participants had varying backgrounds and teaching experience, which was viewed as advantageous for the current exploratory study. Pseudonyms were randomly selected for each participant.
Table 1: Demographic Information
- 1 NS = native speaker; NNS = non-native speaker
Participant
|
Speaker Status1
|
Years Teaching Experience
|
Language(s) Taught
|
Level(s) Taught
|
Degree Sought
|
Amy
|
NNS
|
0
|
n/a
|
n/a
|
MA
|
Andrew
|
NNS
|
6
|
English
|
Beginner and Intermediate
|
PhD
|
Autumn
|
NS
|
2
|
English; Arabic
|
Beginner and Intermediate
|
PhD
|
Ben
|
NS
|
3
|
Spanish
|
Beginner
|
MA
|
Chloe
|
NS
|
0
|
n/a
|
n/a
|
BA
|
Chris
|
NNS
|
8
|
English; Chinese
|
Intermediate and Advanced
|
MA
|
Eve
|
NS
|
1.5
|
English
|
n/a (tutoring)
|
MA
|
Faith
|
NNS
|
8
|
English
|
Beginner and Intermediate
|
PhD
|
Jacob
|
NS
|
0
|
n/a
|
n/a
|
MA
|
Mason
|
NNS
|
0
|
n/a
|
n/a
|
PhD
|
Mia
|
NS
|
0
|
n/a
|
n/a
|
BA
|
Sarah
|
NS
|
10+
|
English; Turkish
|
Beginner, Intermediate, and Advanced
|
PhD
|
Scarlett
|
NS
|
0
|
n/a
|
n/a
|
BA
|
Sofia
|
NNS
|
8
|
English
|
Beginner and Intermediate
|
MA
|
Zelda
|
NS
|
1
|
English
|
Intermediate
|
PhD
|
12The graduate course focused on the teaching and researching of L2 pronunciation with a particular focus on English pronunciation. The course took place within a program that emphasizes the inclusion of technology in applied linguistics and language teaching though there was no requirement in this class to include technology in teaching. Some topics covered in the course included segmentals, suprasegmentals, materials analysis and development, and computer-assisted pronunciation training (CAPT). Table 2 provides an overview of the course, including major topics and assignments. The textbook for the course was Celce-Murcia et al.’s (2010) Teaching Pronunciation: A Course Book and Reference Guide, which introduced the participants to the communicative framework (p. 44-48), which outlines five types of exercises or activities that language teachers can use when teaching pronunciation: description and analysis, listening discrimination, controlled activities, guided activities, and communicative activities.
Table 2: Structure of the Graduate Course
Week
|
Topic(s)
|
Assignment(s)
|
1
|
Researching and Teaching Pronunciation
|
|
2
|
Consonants
|
|
3
|
Vowels
|
Materials Analysis 1
|
4
|
Word Stress
|
Materials Analysis 2
|
5
|
Rhythm
|
Materials Analysis 3
|
6
|
Connected Speech
|
Materials Analysis 4
|
7
|
Prominence / Intonation (1)
|
|
8
|
Intonation (2)
|
|
9
|
Assessment (1)/Materials Development (1)
|
Materials Analysis 5 / Diagnostic
|
10
|
Assessment (2)/Materials Development (2)
|
Lesson Plans Week 1-3 for Tutoring
|
11
|
Technology
|
Week 1 Tutoring
|
12
|
Morphology / Orthography
|
Week 2 Tutoring
|
13
|
Speaking / Fluency
|
Week 3 Tutoring
|
14
|
Listening / Perception
|
Week 4 Tutoring
|
15
|
Variation and Social Aspects/Ethics of Instruction
|
Research Paper
|
13As the capstone project for the course, each participant was given the task of providing explicit instruction to a learner of English over four weeks (i.e., a tutoring project). The learners were current students, both undergraduate and graduate, at the same university and from a variety of disciplines and majors. The purpose of the project was to allow the participants to put into practice the knowledge they had gained throughout the course. By the time the tutoring project started, the participants had received training in the major features of English pronunciation (see Table 2 for a complete list). The tutoring sessions took place in a space determined by the participant and their learner, so long as the space was quiet (e.g., empty classroom, office space, conference room).
14Before tutoring, each participant received audio files of their learner’s speech, which they then had to analyze for pronunciation errors. This diagnostic assignment was meant to give the participants a chance to notice and categorize learner errors. At the end of the diagnostic, each participant was to choose three pronunciation features to work on with their learner. For these features, tutors were encouraged to address the pronunciation errors that were most detrimental to the learner’s comprehensibility and intelligibility. From there, the participant developed a three-week lesson plan based on a template, in which the three features selected should be targeted in each lesson. While the participants were not required to follow their lesson plan exactly, they were required to address each of the three features each week. This flexibility allowed participants to spend more time on a particular feature if they thought it was necessary. For the lesson plan for the fourth week, tutors were asked to create their own plan that did not have to conform to the template plan of the first three weeks.
15Four instruments were used to collect data in this study: diagnostic assessments, oral reports, written reflections, and tutoring observations. The use of multiple methods aided with methodological (between-methods) triangulation (Denzin, 1989), as each instrument was used to provide another angle to analyze the participants’ experiences and their evolving teacher practices. In addition to the four instruments, there were a number of other data sources that were used to further contextualize the findings, including unstructured interviews with the participants and the course instructor, video recordings of the course, and the lesson plans that were developed by the participants.
16Diagnostic Assessment
17Teachers in the study started the tutoring project by being assigned a student for the tutoring project. Before the first tutoring session, the tutee recorded their speech in both read-aloud and spontaneous tasks. Tutees were recruited through university ESL classes or when they approached the course instructors for extra help. Tutees were assigned to tutors by the instructors of the class based on what we knew of the tutee’s needs and the tutors’ backgrounds. Our main intended restriction in matching pairs was that we tried to avoid tutors and tutees sharing the same language background to encourage the use of English throughout the tutoring.
18The diagnostic assessment included a variety of tasks, each of which targeted different pronunciation features. This was done to acquaint tutors with different ways to elicit pronunciation targets, to make it easier to choose targets for tutoring, and to make the diagnostic grading more transparent. Tutors were given a week to complete the diagnosis before receiving feedback on their accuracy from the instructors. The tasks and targets for diagnosis are listed in Table 3.
Table 3: Structure of the diagnostic assessment
Task
|
Target(s) and Number of items
|
Comments
|
Reading tasks
|
Passage reading
|
Consonants (not counted)
-ed & -s endings (17)
Word stress (28)
Intonation (16)
|
Passage reading is a common way of diagnosing pronunciation. The passage is from Celce-Murcia et al. (2010). Consonant sounds were not counted because of their quantity
|
Minimal pair reading
|
Vowels (72)
|
Potentially challenging vowel minimal pairs were presented together.
|
Short phrase reading
|
Linking (11)
|
Connected speech features such as linking are advocated for pronunciation teaching. These phases highlight three types of linking: CV, VV, and CC
|
Sentence reading
|
Distinguishing full vowels (16) from reduced vowels (12)
|
The production of schwa is critical to English rhythm. These Sentences include both full and reduced vowels diagnosed during different listening tasks.
|
Dialogue reading
|
Prominence (28)
|
Prominence (or sentence stress) is important for marking contrasts and new information. This task targets final and non-final placement of prominence.
|
Free speech tasks
|
Describe favorite holiday
|
All targets
|
Reading aloud is more efficient at eliciting target features, but less authentic as a task. This task asks tutees to speak freely on a familiar topic.
|
Video prompt retelling
|
All targets
|
Picture and video retelling elicit speech that is less monitored for pronunciation features by requiring retelling. This video narration uses a video without speech as a prompt.
|
Besides the immediate purpose of the diagnostic, that is, as a needs analysis for the tutoring, the diagnostic also was intended to help tutors learn to listen to pronunciation as part of controlled and free speech, to identify individual errors and error patterns, and to make decisions about priorities. Tutors were required to choose three errors to teach, at least one of which had to be a suprasegmental (word stress, intonation, rhythm, or prominence). When the diagnostic was graded, they were given feedback both on the accuracy of their listening and the appropriateness of their choices.
19Written Reflections
20Written reflections were collected from the participants during weeks 1 and 4 of the tutoring project (weeks 11 and 14 of the course, respectively). The participants were required to write two to three pages for each reflection, as they answered the following questions:
1. Describe how your tutoring session went. Try to provide as much detail as possible.
2. Reflect and evaluate your tutoring session. What was successful and what was not successful? Why?
3. What changes, if any, do you think you would make if you had the chance to do the tutoring session over?
4. How do you think your tutee responded to your instruction? Why?
21Oral Reports
22Oral reports were conducted during weeks 2 and 3 of the tutoring project (weeks 12 and 13 of the course, respectively). Each oral report lasted roughly 60 minutes and was broken into two constituents: a semi-structured interview (SSI) and a stimulated recall interview (SRI). The SSIs were used to gather general information about the participants’ experiences with the tutoring project, while the SRIs were used to have the participants interact with video footage of their actual tutoring sessions.
23Tutoring Observations
24Video-recorded tutoring observations were conducted with each participant in weeks 2 and 3 (approximately 110 minutes in total for each student teacher). These weeks were decided in tandem with the course instructor as the ideal weeks to observe the participants. Each observation was recorded and used as part of the SRI and to provide further contextualization for any findings.
25First, data collected from the diagnostic assessment were analyzed for frequency of errors. The frequencies were then entered into an Excel spreadsheet for further comparison of the pronunciation errors that the participants made.
26Next, the oral reports were transcribed verbatim. These data, along with the video-recorded tutoring observations, were entered into NVivo 12 (released in March 2020) for coding purposes. The tutoring videos were segmented into individual activities, which allowed for the coding of the activity types, and the oral reports were segmented into their two constituents (i.e., SSI and SRI). This step was done as soon as the authors realized that the data types provided separate views about the participants’ teaching practices. Finally, the written reflections were added.
27A top-down approach to coding was adopted to analyze the SSIs, SRIs, and written reflections using Shulman’s (1986, 1987) theoretical model of teachers’ seven categories of knowledge. For the current study, the authors emphasized three of these categories: content knowledge, pedagogical content knowledge, and knowledge of the learner. These three categories were chosen to highlight the student teachers’ knowledge of pronunciation and phonological features (i.e., content knowledge), their ability to translate content knowledge into practical teaching applications (i.e., pedagogical content knowledge), and their ability to analyze and categorize learner errors (i.e., knowledge of the learner). This coding scheme was further adapted using the work of Baker (2014), Brown (2007), Burri (2016), Celce-Murcia et al. (2010), and Couper (2017) to allow for a more in-depth understanding of pronunciation pedagogy and practices.
28Diagnostic
29The pronunciation diagnostic was evaluated in two ways: for accuracy by the second author before the initial tutoring sessions, and retrospectively by the teachers in their reflective comments. The second author, who has several decades of experience in supervising tutoring experiences and diagnostic assessment, checked his impressions with the first author, who had previously taken the class before serving as the teaching assistant and carrying out the primary research reported in this paper.
30The tutors had a week to complete the diagnostic assessment and to use their findings to recommend targets for tutoring. They were asked to recommend three targets and two alternative targets (in case some of the initial choices were misguided or based on inaccurate listening). Tutors were generally accurate in their choices, but in a minority of cases, tutors were asked to change their targets because they had chosen a feature that they had not heard accurately. The diagnostic assessment assignment, which was able to be listened to in about 25 minutes from beginning to end, took almost every tutor the full week to complete, and most tutors turned it in but expressed a lack of confidence that they had heard everything accurately. However, the tutors became very familiar with their tutees’ speech patterns before their first meeting, and the diagnostic served to make them more confident about meeting their tutees and getting started right away with focused practice.
31Their expressed lack of confidence was in line with their overall accuracy on the assessment and often showed that they did not hear pronunciation errors consistently well (see Table 4 for error frequencies). This was true both for features that they were already sensitive to (e.g., vowel contrasts) and those that they were not used to identifying (e.g., differences between full and reduced vowels, prominence). Sometimes their mishearings came from the influence of their L1. For some vowels, for example, L2 English-speaking tutors admitted they struggled to hear certain contrasts themselves and they were uncertain they were classifying sounds correctly. For other features, such as prominence, tutors would hear some aspects of prominent syllables well (pitch movement) but not others (deaccenting syllables after the prominence). Often, the result of their own struggles was that the tutors wanted to teach these features. If L2 English-speaking teachers struggled to hear vowel contrasts (such as the difference between bad and bed), they found models to use on the internet and took the role of a co-learner. This strategy was almost always successful in connecting to the tutee’s difficulties and in stepping back from the role of an infallible expert.
Table 4: Error Frequencies of Student Teachers on Diagnostic Assessment
Student Teacher
|
-ed/-s Endings (17)
|
Word Stress (28)
|
Intonation (16)
|
Vowels (72)
|
Linking (11)
|
Reduced Vowels (12)
|
Full Vowels (16)
|
Prominence (28)
|
Amy
|
0
|
0
|
0
|
18
|
0
|
0
|
7
|
6
|
Andrew
|
0
|
0
|
0
|
12
|
1
|
2
|
8
|
1
|
Autumn
|
0
|
2
|
0
|
2
|
2
|
1
|
4
|
3
|
Ben
|
0
|
2
|
0
|
8
|
3
|
0
|
1
|
6
|
Chloe
|
1
|
0
|
0
|
4
|
1
|
1
|
6
|
4
|
Chris
|
0
|
3
|
1
|
13
|
1
|
3
|
5
|
3
|
Eve
|
0
|
0
|
0
|
7
|
0
|
1
|
3
|
3
|
Faith
|
0
|
5
|
1
|
15
|
4
|
3
|
3
|
5
|
Jacob
|
2
|
5
|
1
|
8
|
2
|
2
|
5
|
2
|
Mason
|
0
|
8
|
0
|
6
|
1
|
0
|
0
|
4
|
Mia
|
0
|
0
|
0
|
7
|
1
|
0
|
4
|
5
|
Sarah
|
0
|
0
|
3
|
2
|
3
|
5
|
4
|
3
|
Scarlett
|
0
|
2
|
0
|
5
|
0
|
0
|
1
|
3
|
Sofia
|
0
|
6
|
0
|
10
|
1
|
4
|
4
|
3
|
Zelda
|
0
|
3
|
1
|
16
|
0
|
6
|
4
|
5
|
Mean
|
0.19
|
2.38
|
0.44
|
8.94
|
1.38
|
1.81
|
3.94
|
3.75
|
32Tutoring
33During the first week of tutoring, the student teachers reported that the sessions went well overall, though a number mentioned pitfalls or hurdles that they needed to overcome in future sessions. Two of these challenges, which provide a good snapshot of the student teachers’ evolution as pronunciation instructors, included (1) piloting, ordering, and executing activities and (2) teaching and assessing suprasegmental features.
34Piloting, ordering, and executing activities. Thirteen of the 15 student teachers mentioned they would change their choice or execution of activities, from simply having more activities on-hand in case they finished early to piloting the activities beforehand. The first week’s reflections from the student teachers revealed that time management was a factor that many misjudged. For example, seven student teachers mentioned that listening activities were either significantly longer or shorter than they had expected, which resulted in the timing of their session to be off. As Ben noted after the first week:
I now have a better idea about which aspects of each feature we can focus on [in each session] and how long various activity types take […]. Therefore, I will simply try to be more realistic next week with how long each activity might take and perhaps have extra activities in the case that we progress more quickly than expected. (Reflection #1)
Mia shared a similar hurdle with the timing of activities, noting that she “need[s] to work on timing things out a bit better. The first two topics [received] sufficient time, but the last [topic] was shortened considerably” (Reflection #1). This hurdle—that each pronunciation feature did not receive equal attention—was noted by over half of the student teachers.
35To resolve this issue with time management, some student teachers mentioned that piloting their activities beforehand could be a possible solution. In addition to time management, piloting the activities would also ensure that the activity would do what the student teacher wanted it to do, as Ben noted:
[…] I plan to do a mock run-through of my materials prior to the [second] session in hopes of identifying any possible bugs in the materials[…]. I realized that one of the /ʌ/-/ɑ/ activities didn’t actually work as I had planned (I had initially thought that I could restructure it on the fly to work differently than the activity was intended for), and so I had to scrap the activity just minutes before starting. (Reflection #1)
36Teaching and assessing suprasegmental features. Studies have found that teachers prefer to teach segmentals over suprasegmentals, as they are generally perceived to be easier to teach (Foote et al., 2011; Foote, Trofimovich, Collins, & Urzúa, 2016; Huensch, 2019). Even though some participants noted an increase in confidence—“After this course, I feel more knowledgeable and confident with suprasegmentals” (Mia, Reflection #2)—eight participants painted a different picture about suprasegmental instruction. Frequent unwanted pauses and missed errors often were reported by participants while watching themselves teach suprasegmental features, particularly intonation and prominence. The unwanted pauses were typically the result of providing descriptions and examples, as Faith remarked while reflecting on teaching intonation: “I’m just not confident about that. I don’t know if I have the right intonation or not. Not describing, just [producing intonation]” (SRI #1). Scarlett presented a similar reflection when talking about prominence instruction as a NS:
I think prominence is harder than I thought it would be because prominence is something that is—like production of a consonant or a vowel, you know, you can think about where it is in your mouth, you explain how to produce it—but prominence is something that is just so intuitive. And […] I can be taught […] prominence is what this means, and this is how we use it, and when I’m trying to explain it to somebody, it’s a lot harder because it’s like explaining how to walk. (SSI #1)
The ‘intuitive’ nature of prominence, as Scarlett refers to it, is likely the result of acquiring this feature as a NS rather than learning it as an NNS.
37While NS and NNS participants mentioned challenges with teaching suprasegmental features, only NNS participants unequivocally mentioned having a lack of confidence in their examples, with their L2 accent typically being the underlying cause. Mason more explicitly stated this lack of confidence as an obstacle for NNS teachers: “I don’t know but when it comes to actually explaining what the pronunciation feature is, especially suprasegmentals, and I’m not saying that [NS speakers] are better, I’m just saying that nonnative trainees wouldn’t be as confident” (SSI #2). A lack of confidence by the teacher could lead towards neglecting instruction in the future, as some suggest (e.g., Derwing, 2019), which is concerning because some writers have said that pronunciation instruction “should focus first and foremost on suprasegmentals as they have the greatest impact on the comprehensibility of the learner’s English” (McNerney & Mendelsohn, 1992: 132).
38Use of Technology
39While not explicitly asked to use technology, six student teachers decided to incorporate technology into their tutoring sessions. Among the six student teachers, five used technology to show a video from YouTube, and one used technological resources aside from YouTube. Autumn, Faith, Mia, Sarah, and Scarlett used a laptop to show videos from YouTube. Faith incorporated two types of videos, both an instructional video and short clips from the sitcom Friends. Her motivation to use the instructional video was to provide a description of the pronunciation feature that also gave a sagittal view of the mouth and tongue, which she thought would be “more helpful” for her learner than her description alone (SSI #1). After the instructional video, Faith provided explicit activities to focus on her pronunciation features, such as intonation and rhythm. Additionally, she introduced methods of improving pronunciation by watching American sitcoms, which was her attempt to get her learner motivated, stating, “It was not easy to keep her motivated for the whole class time. But watching a video seemed to amuse her and it was also a new strategy for her to learn” (Reflection #4).
40Mia, Scarlett, and Autumn also used YouTube videos to show clips of American movies or television shows. Mia and Autumn, while overall pleased with how the activity progressed, mentioned they would have gotten input from their learner before selecting the clips, as Mia expanded, “Yeah, I don’t think she hated [the Disney clips], but she didn’t enjoy them as much as I did” (SSI #2). Scarlett also used clips from YouTube, but she mentioned difficulty in preparing for the activity during the tutoring session:
I had thought I had watched it over and over and had marked on my paper where the [features] I wanted to talk about were and what I want to talk about with him. But I didn’t really write out a whole explanation for everything, and so I’m sitting there and I’m like, “Oh, I know I wanted to bring this up there” but that was kind of hard because I didn’t know how. (SRI #2)
One trainee, Eve, used technology beyond showing YouTube videos. She incorporated listening exercises from EnglishAccentCoach into each of her tutoring sessions to work on the /f-θ/ distinction. While this practice became successful by the end of the tutoring project, Eve reported in the first week that “one aspect of the session that was [least] successful was [her] use of EnglishAccentCoach” (Reflection #1). Her learner found this tool to be frustrating because it focused solely on sounds, while the learner was trying to understand them as words. Eve found a quick and effective solution for this hurdle in the next session: “I began with other listening activities first in which I provided her with real words with the target sounds, and then I gave her the EnglishAccentCoach.com activity” (SSI #1). In this order, Eve found that her student had been “primed” for hearing the specific sound after hearing it in context first.
41This article used three separate lenses to look at how student teachers evolved in their knowledge and practice of teaching pronunciation in a one-on-one setting: (1) diagnosing and evaluating learner errors; (2) teaching pronunciation in a one-on-one setting; and (3) incorporating technological tools to enhance pronunciation instruction. The diagnostic was an essential first step, both to focus the tutors’ listening and to prioritize their teaching targets as they decided what was most important. Second, although working with one learner is different from working with a full class, it focused the tutors’ attention in a way that a full class would not. Tutors were faced with success or lack of success because they had to concentrate on one person who was constantly providing feedback on how well they were doing. Tutors could not put learners into pairs or groups, and thus tutors had to be constantly engaged. Tutors also found that they needed to improvise and adjust as they saw the success (or failure) of their teaching plans and activities. Finally, although tutors were not required to use technology, they often did. Experienced pronunciation tutors often act as communicators, models, and feedback providers at the same time. These multiple roles can be exhausting for teachers, and the use of technology, especially in providing a reliable model, allowed tutors to concentrate on communicating and making decisions about feedback. Because these tutors were also digital natives, they often used technology in ways that the teacher of the class had not thought of, indicating that they were connecting their teaching goals to other sources of input that they knew of before learning to teach pronunciation. For example, Eve used her computer’s webcam as a mirror. In this way, the tutee could watch Eve’s mouth and then mimic her own facial movements when trying to distinguish two sounds. This was further utilized when Eve made recordings of herself making the sounds, which the tutee could use later to continue practicing by having the recording on one side of the computer screen and the webcam video on the other side.
42Even though pronunciation is becoming more visible than ever in language teaching and in applied research contexts (Levis, 2015), teacher training in pronunciation pedagogy remains neglected in most language teacher preparation programs. It is highly desirable to have classes devoted to the teaching of pronunciation, and content knowledge about pronunciation teaching and learning must be the backbone of such a class. Teachers’ growth, however, is not likely to develop deeply without putting content knowledge into practice. The tutoring project reported on in this paper shows how different practical tasks provide the context for tutors to put their knowledge to work with a learner with specific challenges. There are other ways of putting theory into practice, but the tutoring project reported here is both flexible and effective in helping students experience how content knowledge works in practice.