1. Introduction
Training in pronunciation pedagogy helps pre-service and in-service teachers develop cognitions (i.e., beliefs, thoughts, and knowledge about teaching; see Borg, Reference Borg2003) to implement systematic pronunciation teaching practices in class (e.g. Burri & Baker, Reference Burri and Baker2021). Additionally, many teachers develop a knowledge base for pronunciation teaching based on their previous teaching and learning experiences (Gordon, Reference Gordon2019). Such a knowledge base is necessary to implement explicit instruction that could help learners develop intelligible second language (L2) speech (Lee et al., Reference Lee, Jang and Plonsky2015). One key component in instruction is corrective feedback (CF), which is effective in pronunciation development when combined with explicit instruction (Saito, Reference Saito2013; Saito & Lyster, Reference Saito and Lyster2012). CF is beneficial not only when implemented by the teacher but also during peer interaction (Martin & Sippel, Reference Martin and Sippel2021). However, very little has been investigated about the cognitions that direct teachers to implement CF in L2 pronunciation (but see Baker & Burri, Reference Baker and Burri2016; Couper, Reference Couper2019). One particular group of teachers who has not been investigated in this regard is nonnative-speaking (NNS) teachers of English. While previous research has demonstrated the effectiveness of NNS teachers in L2 pronunciation instruction (Levis et al., Reference Levis, Sonsaat, Link and Barriuso2016), and the knowledge base that allows them to systematically implement pronunciation teaching (Gordon, Reference Gordon2019), this is still an area that has not been investigated in depth. The qualitative case study presented here investigated the way an experienced NNS teacher implemented CF in a stand-alone English as a Foreign Language (EFL) pronunciation class at a university in Costa Rica, and the rationale behind the implementation of CF techniques. The study was guided by the following questions:
1. How does an experienced NNS teacher implement CF in an EFL classroom?
2. What cognitions allow this experienced NNS teacher to implement CF in pronunciation instruction in a classroom?
2. Data collection and analyses
In this qualitative case study (Merriam & Tisdell, Reference Merriam and Tisdell2016), I interviewed Carlos (a pseudonym) through a semi-structured interview to learn about his teaching-learning background, in addition to his ideas about pronunciation teaching, intelligibility, and being a NNS teacher. I chose Carlos for this study because of his experience teaching pronunciation and working with a variety of learners of different ages and proficiency levels. For instance, he had undergraduate and graduate degrees in Teaching English to Speakers of Other Languages (TESOL), and he had almost ten years of teaching experience at the moment of the study. It is important to mention that as part of his teacher training, he did not take a pronunciation pedagogy course. However, he took an introduction to linguistics class and pronunciation courses like the one of this study (as an L2 learner) where he learned basic aspects of English pronunciation and phonetics and phonology in general. After the initial semi-structured interview, I observed Carlos for a total of eight hours of pronunciation instruction as a non-participating observer. Finally, I carried out a stimulated-recall interview in which Carlos observed video clips from his class and provided a rationale for using different CF techniques. I coded the CF techniques he implemented based on Lyster and Ranta's (Reference Lyster and Ranta1997) taxonomy of CF (i.e., explicit correction, recasts, clarification requests, metalinguistic feedback, elicitation, and repetition). Then, based on the explanations Carlos provided through the semi-structured and stimulated-recall interviews, I coded his CF techniques and their rationale based on three categories from Shulman's (Reference Shulman1987) knowledge base of teaching framework: content knowledge (e.g., basic knowledge of phonetics and phonology), pedagogical content knowledge (e.g., practical ability of a teacher to make phonetic/phonological content accessible to learners), and knowledge of learners and their characteristics (e.g., knowledge of learners’ first language (L1) and their L2 pronunciation needs). Because these categories overlapped, I also coded some of this overlapping information into a fourth category, mixed category, as it has been done in previous studies (e.g. Johnston & Goettsch, Reference Johnston and Goettsch2000). The results are presented as follows.
3. Knowledge base of pronunciation teaching and corrective feedback techniques
3.1 Content knowledge
Although recasts can be a simple reformulation of learners’ production, and some of them are more salient than others, the rationale behind their use by Carlos evidenced his basic knowledge of phonetics and phonology – which he learned in a basic linguistics class as an undergraduate student and pronunciation courses similar to this one that he took as a student. For example, he mentioned that using recasts allowed him to indicate key linguistic differences in sounds, such as voicing in pairs of consonants like /s/-/z/ or /p/-/b/, or differences in vowel sounds like /i/-/ɪ/ or /ɛ/-/æ/-/a/. In addition to recasts, he also used phonetic transcription on the board – sometimes highlighted with color markers to stress pronunciation differences. However, in many cases these recasts were not provided in isolated form but were accompanied by linguistic explanations.
3.2 Pedagogical content knowledge
Carlos also provided recasts with metalinguistic information. For example, he recasted the right pronunciation of the word played and then reminded learners that all vowels are voiced – which is why the pronunciation of the past tense -ed morpheme is [d] in this specific case. Some of these metalinguistic comments were also reinforced with phonetic transcriptions on the board for learners to visually see differences between sounds. In some other instances, Carlos elicited the right pronunciation from learners. He explained that these CF techniques (metalinguistic feedback and elicitations) were aimed at making learners aware of common pronunciation problems that in some cases were due to the learners’ L1 (Spanish). Additionally, he also claimed that eliciting the right pronunciation from learners could enhance learning autonomy by making them aware of mispronunciations that could cause problems in communication. In many cases, such elicitations and metalinguistic comments were used to make learners aware of differences between their L1 and L2.
3.3 Knowledge of learners and their characteristics
Carlos did not correct his students all the time, and he explained that in some cases he needed to be selective as to what to correct. However, recasts, metalinguistic comments, and elicitations were still used to raise learners’ awareness of differences between Spanish and English. For instance, he elicited correct pronunciation not only from learners being corrected but also from the entire class to see if the students remembered specific rules discussed in class before (e.g., the pronunciation of cooked in which a student said /kukɛd/). These elicitations also prompted discussions in class about differences between English and Spanish pronunciation. Carlos used these discussions as opportunities to make learners aware of pronunciation differences between both languages and to indicate to learners the pronunciation features that other speakers could expect to hear from them to maintain intelligible communication. Carlos also explained that eliciting information from the rest of the class felt like a more natural form of interaction that resembled common conversations that happen in real life.
3.4 Mixed categories
It is also important to mention that the rationale behind the use of recasts, metalinguistic feedback, and elicitations overlapped on characteristics of the previous three categories. For example, recasting the right pronunciation of a word (e.g., circus) and explaining to the students the linguistic reasons for its right pronunciation contrasting English and Spanish vowels and the voicing sound /z/ at the end evidenced Carlos’ knowledge of the subject matter (phonetics and phonology). Additionally, it demonstrated his knowledge of ways to make such content accessible to learners (e.g., recasts with metalinguistic information and then using phonetic transcription on the board) based on his knowledge of the students’ L1 and their pronunciation needs. All of these aspects evidenced the complexity of a teacher's knowledge base to implement CF in pronunciation instruction regardless of his linguistic background.
4. Discussion and conclusion
The previous results demonstrated that an experienced NNS teacher uses a variety of different techniques to provide CF in pronunciation instruction. Carlos’ use of recasts, metalinguistic feedback, elicitations, or no feedback at all (using comparisons between the L1 and L2, or reinforcing his explanations with phonetic transcriptions on the board) demonstrated that NNS teachers like him use a variety of CF techniques based on their learners’ needs and what they know about the pronunciation system of the L2. While this is evidence of implementing instruction known to enhance pronunciation development (Saito, Reference Saito2013; Saito & Lyster, Reference Saito and Lyster2012), these findings align not only with previous research that demonstrated that NS and NNS teachers deliver explicit pronunciation instruction based on their knowledge base of teaching (Levis et al., Reference Levis, Sonsaat, Link and Barriuso2016), but also with results that demonstrated experienced NS teachers using a similar rationale for implementing specific techniques (e.g., recasts, elicitations) in English as a second language (ESL) contexts to make learners aware of linguistic features and enhance learner autonomy (Baker & Burri, Reference Baker and Burri2016; Couper, Reference Couper2019).
These results also demonstrate that the knowledge base that allows a NNS teacher to implement CF is a complex synthesis of interrelated categories of knowledge influenced by formal training and shaped by past teaching-learning experiences. The interrelationship of these categories of knowledge is reflected in the selection of different CF techniques. For example, Carlos relied on his knowledge of phonetics and phonology (content knowledge) to bring learners’ attention to phonetic-phonological characteristics of the L2 that could cause problems in communication. However, more than pointing out such linguistic features through explicit metalinguistic explanations, recasts, or elicitations – reinforced with phonetic transcriptions on the board (pedagogical content knowledge), his knowledge of differences between his students’ L1 and the L2 also influenced his pedagogical choices (knowledge of learners and their characteristics). This knowledge base that allowed Carlos to implement CF is similar to what has been found in pronunciation teachers implementing CF in ESL contexts (Baker & Burri, Reference Baker and Burri2016; Couper, Reference Couper2019), or in teachers of other skills such as grammar (e.g. Johnston & Goettsch, Reference Johnston and Goettsch2000).
These results also imply that training on pronunciation pedagogy should allow teachers to develop knowledge of different ways to implement CF. Due to its importance in L2 pronunciation development (Martin & Sippel, Reference Martin and Sippel2021; Saito & Lyster, Reference Saito and Lyster2012), training on pronunciation pedagogy should allow teachers to experiment with different CF techniques – which could be done using specific frameworks that could be adapted to specific teaching situations (see Couper, Reference Couper, Levis, Derwing and Sonsaat-Hegelheimer2022). For instance, training could include classroom activities such as microteaching simulations or tutoring projects where learners could put into practice different CF techniques that could give future teachers experience to start shaping their knowledge base of teaching and make sound pedagogical decisions based on specific teaching contexts.
Supplementary material
To view supplementary material for this article, please visit: https://doi.org/10.1017/S0261444823000265.
