A Journal of Teaching English Language and Literature

ISSN Print : 2229-6557, Online: 2394-9244

Exploring Self-Evaluation of Teacher Talk (SETT) in the Indian ESL Context With In-Service Teachers

Monishita H. Pande


Recognizing the complex relationship that exists between language use, classroom interaction and learning can empower teachers to make more conscious decisions in class. Not only is this in line with the changing pedagogic goals of the moment, but it is also guided by the interactional choices available to the teacher. In this study, I will critically examine the tool proposed by Walsh (2006)—Self-Evaluation of Teacher Talk (SETT)—to see how it encourages teachers to reflect on their classroom discourse. I will also present a modified version of the tool after closely looking at teachers’ discourse in the Indian context. For the purposes of the study, I will critically analyse the classroom discourse of six in-service teachers based on data collected over a span of 5 weeks across 12 recorded lessons. I will attempt to unravel the dominant characteristics of teacher talk and identify new modes, pedagogic goals, interactional features and roles operating in the language classrooms. I would like to recommend this revised tool to teachers so that they can engage in self- and collaborative reflection.

Key words: SETT, classroom discourse, teacher talk, classroom interaction


In order to create an acquisition rich environment in the classroom, it is important for teachers to understand their classroom discourse in light of SLA research and findings. To make this connection, the teacher needs to make use of a tool which is accessible, comprehensible and promotes reflection. Walsh (2006) argues that in the area of classroom interaction, there is an absence of common metalanguage which can be used by researchers and practitioners alike. The Self-Evaluation of Teacher Talk (SETT) tool proposed by Walsh (2006) enables teachers to describe the classroom interaction in their lessons, reflect on their interactional patterns and foster an understanding of the same. The position adopted by him is that the “single, L2 classroom context does not exist; contexts are locally constructed by participants through and in their interaction in the light of overall institutional goals and immediate pedagogic objectives” (p. 62). In this study, I will report the challenges of using the SETT tool in the Indian context. I will also present a revised SETT tool, which may be used by Indian teachers to gain insights into their classroom discourse.

Context of the Study

This study was conducted in two Central government schools, S1 and S2 located in North Delhi. In both schools, teachers faced challenges such as learner drop out, bilingual education, mixed learner backgrounds in terms of socio-economic status, and limited resources and access to teacher development activities. The study involved six in-service English language teachers, who were teaching the CBSE curriculum, which is followed by a majority of schools in the country. In the initial phase of the study, the classroom discourses were recorded and analysed using the original SETT tool proposed by Walsh (2006). However, Walsh’s SETT tool was based on typical ESL classrooms in western countries, where the focus is on language skill development through communicative tasks. Therefore, during the course of the study, the SETT tool had to undergo several revisions based on the inputs from the teachers as well as the dominant features of classroom interaction found in the Indian ESL classrooms.

Interactions with the teachers revealed that the main focus of the English classes was on the literature reader. Books which helped to develop language proficiency were not used by the teachers. The reasons for this were: inability to understand how language tasks need to be conducted, and lack of teacher training and expertise in developing language skills.


The current study is based on a number of assumptions, which helped in formulating the research questions for the study. These assumptions are:

1.      There is an intrinsic relationship between language use and pedagogic goals, which requires a deeper understanding.

2.      The quality of teacher talk has a direct effect on the quality of classroom interaction and hence language acquisition.

3.      The teacher controls the patterns of classroom interaction and hence it is the teacher’s responsibility to promote meaningful interaction in the classroom.

4.      Self-Evaluation of Teacher Talk (SETT) can act as a tool for reflection to help teachers describe the classroom interaction of their lessons as well as to develop an understanding of their interactional processes.

5.      The SETT framework proposed by Walsh (2006) when applied in the Indian context, will need to be modified to suit the requirements of the teachers.

Research Questions

The study was guided by the following research questions:

1.      Does the teacher try to maximize the learning opportunities in the classroom by using effective strategies to clarify, check or confirm meanings? (Use of confirmation checks, comprehension checks, repetition, clarification requests, reformulation, turn completion, backtracking)

2.      What are the dominant interactional patterns and modes in the classroom? Are there modes and interactional features in the classrooms other than the ones identified by Walsh?

3.      What are the challenges involved in implementing the SETT framework? What modifications need to be made to make it effective in the Indian context?

Classroom Interaction and SLA

Research studies support the view that there exists a strong relationship between classroom interaction and SLA. According to Gass (1997), conversation is considered to be an important devise that sets the stage for language acquisition to take place. An argument which one finds repeatedly in literature is that meaningful interaction increases the possibility of greater amount of input becoming available to the learner. This results in an increase in learning opportunities, which is required to initiate certain basic processes that are crucial for L2 development.

Interaction is considered to be at the heart of language acquisition, particularly the interaction that takes place between teachers and learners (Ellis, 1998). Allwright (1984) states that effective teaching involves “successful management of the interaction” (p. 154). Further, the quality of interaction is largely determined by teachers during their face-to-face communication with the learners. As the quality of interaction has an impact on learning, it is important to optimize it rather than to maximize it. Therefore, to improve language learning in the classroom, appropriate interaction matching the desired learning outcomes has to be encouraged.

Walsh (2006) argues that an awareness of interactional processes is important for teachers and learners to understand how language is learnt in formal settings. Johnson (1995) writes, “the teacher plays a crucial role in understanding, establishing and maintaining patterns of communication that will foster, to the greatest extent, both classroom learning and second language acquisition” (p. 90).

According to van Lier (1996), “interaction is the most important element in the curriculum”. Ellis (2000) states that “learning arises not through interaction but in interaction”. Broadly speaking, interaction needs to be understood to promote learning. Besides, the lack of empirical evidence for the contribution of learner-learner interaction to SLA (Foster, 1998), there is an increasing awareness that the teacher has an important role to play in shaping learner contributions (Jarvis & Robinson, 1997).

Long’s Interaction Hypothesis (1996) and Swain’s Output Hypothesis (1995, 2005) have been modified in recent years to acknowledge and redefine the role of the teacher in constructing knowledge and understanding. Long’s Interaction Hypothesis (1996) has been adjusted to take into account the importance of negotiation for meaning in the feedback that the learners receive from the teacher. Swain, in her latest version of Output Hypothesis, takes up a socio-cultural dimension in order to highlight the importance of teacher-learner dialogues in promoting acquisition. Smith (2004), while looking at the primary L1 context, emphasizes the importance of “interactive whole-class teaching”.

Ellis (1998) discusses that quality interaction which is “acquisition rich” has to be initiated, managed and sustained by teachers through careful and informed management of turn-taking sequences that occur in face to face communication.

Based on the studies which highlight the importance of interactive processes in acquisition, one may claim that teachers as well as learners need to gain a comprehensive understanding of the interactive processes that facilitate learning. Walsh (2006) states that effective interaction either takes place on its own, or it is a function of the teaching methodology. In fact, interaction in an acquisition rich classroom is initiated and sustained by the teacher.  He further argues that interaction facilitates SLA. Therefore understanding how classroom talk is carried out is crucial to understanding of the role of interaction in SLA. According to Long (1996), SLA is promoted when, through communication breakdown, learners have to negotiate for meaning. He further argues that acquisition of language occurs when learners ask for clarification and confirm comprehension (key features of Long’s interaction hypothesis).

It is clear that routine questions and teacher talk play an important role in enhancing learning opportunities in the classroom. Therefore, it is important to arrive at a context sensitive tool, which will allow teachers to engage with questions around negotiated interactions.

The Self Evaluation of Teacher Talk (SETT) Framework

Walsh (2006) designed the SETT framework based on his analysis of classroom interaction in which he adopted a variable approach to analysing classroom interaction. He identified four micro-contexts, characterized by specific patterns of turn-taking called modes. These were: managerial mode, classroom context mode, skills and systems mode, materials mode. Each mode is made up of specific interactional features related to instructional goals. He added that the modes that he has identified are not comprehensive and other modes may be incorporated depending on the micro-contexts of a classroom.

Walsh indicated that certain interactional features facilitate learning opportunities, while others appear to hinder opportunities for learning. That is, depending on a teacher’s pedagogic goal, the choice of language could either construct or obstruct learning opportunities.

Critical Evaluation of the Existing SETT Framework

Highlighting the challenges involved in using his framework, Walsh (2011) acknowledged the fact that there could be more modes in different L2 contexts. He also mentioned some delineated cases in the data, which do not fit into any of the four modes. Also, there are points where it becomes impossible to identify the exact interactional features.

In the Indian ESL context, while applying the SETT framework I found that English lessons in the two schools selected for the study revolved around the literature reader. Teaching English as a skill was completely ignored; instead it was treated as a content subject.

I conducted workshops with the teachers to familiarize them with the SETT tool, and help them engage with the audio recordings of their lessons using the SETT framework. I also got their feedback on the usability of the tool. During the discussions, I tried to create an awareness among the teachers to view language as a skill rather than as a subject. I felt that I had to address the dominant practice of using literature readers solely to teach English. The teachers on their part expressed an interest in exploring the different ways in which literature texts could facilitate language teaching and increase the interactional space in their classrooms. Therefore, the existing SETT framework was expanded to include new pedagogic goals and interactional features to capture the events of such lessons.

Theorists and researchers claim that literature has the potential to generate a kind of quality talk that is characterized by reflection and exploring intertexual links (Almasi, 1996; Langer, 1995; McGinley et al., 1997; Wells, 1990). Boyd & Maloof (2000) suggest:

By introducing substantive issues through literature, and encouraging students to relate them to their experiences and perspectives, teachers can facilitate opportunities for students to compare and contrast, to make connections with other literary texts and other experiences of their lives. (p. 166)

The six teachers selected for the study were found to have stark differences in terms of their educational background, and exposure to English and ELT training. Although these teachers were teaching in the same institution, their engagement with the SETT tool was quite different in terms of their ability to use it for reflection; in fact, one can argue that each teacher-participant explored the tool at her own pace. The following challenges were identified with regard to the existing SETT framework:

1.      The ESL situation under study was clearly a bilingual situation (if not multilingual). Thus, the use of another language (Hindi in this case) by the teachers needs to be considered.

2.      Two new modes were identified in the data and thus the revised SETT framework had to make space for these modes. These were called bilingual mode and read aloud/lecture mode. Also, a number of changes had to be made to address the use of literature texts within material mode.

3.      In some of the existing modes, a few pedagogic goals and interactional features were added in view of the classroom data collected.

4.      The skills and systems mode had to be revisited as “skills” were missing in all lessons. Since the system mode was operational quite often, it could be seen as a separate mode.

5.      In the lessons, the teachers tended to ignore the materials centred around language skills and planned all their lessons around the literature reader (as discussed earlier). Thus, new pedagogic goals were identified within materials mode. Ways to enable the teacher to use literature texts for effective classroom discourse had to be addressed.

6.      Intertextual links had to be strengthened for effective use of literary texts (literature reader) to encourage meaningful exchanges in the classroom; the revised SETT framework was supposed to address these links. 

7.      The SETT framework does not include teacher and learner roles. Each of the modes triggered certain teacher and learner roles in the classroom in line with the interactional features and pedagogic goals. The argument is that teachers’ self-evaluation of classroom discourse also needs to analyse learner contributions and roles. The original framework did not have space for this. Therefore, teacher and learner roles had to be introduced within the framework. Assigning roles in every mode also served another purpose—it became easier for the teacher-participants to understand them.

Walsh (2006) mentioned the following procedure for the application of his tool:

  • Do an audio recording of 10-15 minutes of the lesson
  • In the first listening of the recording identify the mode
  • In the second listening, use the SETT instrument to keep tally of the different features of teacher talk. Give examples of each feature from the data (take help of the SETT key if required)
  • Evaluate teacher talk in light of the overall aim and modes used
  • Give feedback in the interview

This procedure was found to be problematic, especially when it came to mode identification and the different features of teacher talk. Segregating the two meant that the teacher was analysing them as two separate components. However, modes and the interactional features are intrinsically linked. Therefore, it was important to link mode identification with feature description. Keeping a tally of the two was also not so useful as the quantitative count of how many times the interactional features were used failed to feed into the overall analysis.

The application of the framework revealed that it needs to be implemented in different stages. This is because teachers initially felt overwhelmed by the number of interactional features and modes involved in the tool. 

Analysis of ESL Classroom Interaction and Arriving at the Revised SETT Tool

Classroom data was collected through audio recordings of 12 lessons (Classes VIII to XII) and classroom observations over a span of five weeks.

The following dominant features of L2 classroom interaction were found in the data collected:

1.      Excessive use of teacher echo and repetition

2.      Lack of interactional strategies such as confirmation requests, comprehension checks and clarification requests

3.      Strict control over topic and turn by teachers

4.      High use of display questions

5.      Pre-dominance of IRF pattern

6.      Teacher directed classrooms

7.      Grammatical errors in teacher’s speech, which were repeated by the learners without correction (fear of fossilisation of errors)

8.      Reduced or total absence of wait time on the part of the teacher

9.      Tendency to complete learner contribution without allowing enough time to think

10.    Lack of scaffolding

11.    Students expected to stand up every time they make a contribution (increase in affective filter)

12.    Lack of negotiated interaction in most cases

13.    Positive use of direct repair without wasting much time on repair

14.    Content feedback found in most of the classes

15.    More display questions than referential questions

16.    Presence of new modes (Bilingual mode and Read aloud mode)

A portion of the data collected is presented here to show how it exhibited some of the dominant features of classroom discourse:

Lack of Comprehension Checks, Clarification Requests and Confirmation Checks

In a number of lessons, teachers did not carry out any comprehension checks, clarification requests and confirmation checks. The following extract illustrates this point:


21. T: Special qualities, what special qualities would you like to contribute to Mr Lamb?

22. L5: Very kind hearted person, he used to do gardening, he used to grow apples and anybody could come to the park and eat apples, eat those apples but a…if…=

23. =its very good…very good…if you wish to summarize all the activities in a single phrase then he was a man with positive attitude, extremely positive attitude…

24. L5 optimistic…

25. T yes…very good word… optimistic. I do hope most of you know this word. And this story is a mingle of two opposite qualities. Yes till now we discussed the qualities of Mr. Lamb.

In the extract, in line 22 the learner makes an effort to express himself but before he could complete the contribution, the teacher cuts him short. Instead of clarifying what he was trying to say and helping him in doing so (scaffolding), the teacher gives him a positive feedback (very good…) and moves on to his next question. The opportunity to negotiate meaning was lost and therefore an opportunity for learning too was lost. Again, in line 25, the teacher assumes that the class knows the meaning of the word “optimistic”. He does not confirm it with the help of comprehension check. Referring back to the issue of language problem, in line 25 the teacher uses the word “mingle” in an odd manner.

Revised SETT

The revised SETT is presented as follows. The modifications in the framework have been marked in bold.


Pedagogic goals

Interactional features

Dominant teacher and learner roles


To transmit information

To organize the physical learning environment

To refer learners to materials

To introduce or conclude an activity

To change from one mode of learning to another

To get learner’s attention and help them focus on the material

Single, extended teacher turn which uses explanations and/or instructions

Use of transitional markers

Use of confirmation checks

Absence of learner contributions

Use of comprehension checks

Learner roles: Listener

Follower of instructions

Teacher roles:



Materials (mostly literature texts)







To provide input or language practice around a piece of material

To elicit responses in relation to the material

(Fact, quote or question about the literary work)

To check and display answers

To clarify when necessary

To evaluate contributions

To strengthen intertextual links

To discuss perceptions regarding author’s perspective or intent closely relating to the text

Predominance of IRF pattern

Extensive use of display questions

Form-focused feedback

Corrective repair

Use of scaffolding

Direct nomination

General solicit

Teacher echo



Learner roles:



Teacher roles:







(Skills Mode wherever applicable)

To enable learners to produce correct forms

To enable learners to manipulate the target language

To provide corrective feedback

To display correct answers

To provide metalinguistic knowledge

To provide practice in grammatical structures and vocabulary

To provide learners with practice in sub-skills

Use of direct repair

Use of scaffolding

Extended teacher turns

Display questions

Teacher echo

Clarification requests

Form-focused feedback

Learner roles: Responder




Teacher roles:

Explanation giver






To enable learners to express themselves clearly

To establish a context

To promote oral fluency

To get opinions about the literary work

To enable learners to relate the literary text to self-experience and self-identity

To enable learners to relate to universal qualities and concepts

Extended learner turns

Short teacher turns

Minimal repair

Content feedback

Referential questions


Clarification requests

Learner roles:


Sharer of experience



Teacher roles:

Opinion seeker/giver


Sharer of experience








To encourage learners to establish cultural links (relate to experiences in Indian culture)

To encourage learners to relate to native language (Hindi in this case)

To facilitate understanding of concepts and ideas, and enable learners to access their understanding of those concepts and ideas already available to them in their native/other language

To clarify whenever necessary

To check if learners have understood

Short or extended teacher/learner turn

Display and referential questions

Clarification requests

Confirmation checks


Comprehension check

Form and content focussed feedback

Learner role:




Sharer of experience



Teacher roles:




Sharer of experience



Read aloud Mode/Lecture Mode

Learners read the given text

To explain the content of the text

A learner reads aloud the text (or portions) loudly

The teacher reads aloud the text (or portions) loudly

Long teacher turns

Explaining the material

Paraphrasing and summarizing the material

Personal solicit


Absence of learner turn or participation

Learner role(s):




Teacher role(s): Summarizer

Explanation giver






The Revised SETT Framework


The revised SETT tool is based on data collected from typical Indian ESL classrooms and feedback collected from in-service teachers working in that context, having engaged with the original tool. It captures new modes, identifies new interactional features, includes more pedagogic goals and makes space for learner contributions. All of these reflect the features of a typical Indian ESL classroom, acknowledge the use of mother tongue and also address the use of literature texts to teach language.

The revised SETT tool can act as a powerful framework for teachers to engage in self and collaborative reflection with regard to their classroom discourse in order to identify the dominant interactional patterns in their classroom talk. Also, raising awareness among teachers about the importance of managing classroom interaction in a way that maximises interactional space in the classroom can help in increasing learning opportunities and hence contribute to creating an acquisition rich environment in the ESL classroom.


Allwright, L. (1984). Why don’t learners learn what teachers teach?: The interaction hypothesis. In D. Singleton & D. Little (Eds.), Language learning in formal and informal contexts (pp. 3-18). Dublin: IRAAL.

Almasi, J. (1996). A new view of discussion. In L. Gambell & J. Almassi (Eds.), Lively discussions: Fostering engaged reading (pp. 2-24). Newark, DE: International Reading Association.

Boyd, M., & Maloof, V. M. (2000). How teachers can build on student-proposed intertexual links to facilitate student talk in the ESL classroom. In J. K. Hall & L. S. Verplaetse  (Eds.), Second and foreign language learning through classroom interaction (pp.163-180). London: Lawrence Erlbaum Associates.

Ellis, R. (1998). Discourse control and the acquisition rich classroom. In W. A. Renandya & M. Jacobs (Eds.), Learners and language learning. Anthology Series 39. Singapore: SEAMO Regional Language Centre.

Ellis, R. (2000). Task-based research and language pedagogy. Language Teaching Research, 49, 193-220.

Foster, P. (1998). A classroom perspective on the negotiation of meaning. Applied Linguistics, 19, 1-23.

Gass, S. M. (1997). Input, interaction, and the second language learner. Mahwah, NJ: Lawrence Erlbaum Associates.

Jarvis, J., & Robinson, M. (1997). Analysing educational discourse: An exploratory study of teacher response and support to pupils’ learning. Applied Linguistics, 18(2), 212-228.

Langer, J. (1995). Envisioning literature. New York, NY: Teachers College Press.

Long, M. H. (1996). The role of the linguistic environment in second language acquisition. In W. Ritchie & T. Bhatia (Eds.), Handbook of second language acquisition (pp. 413-468). New York, NY: Academic press.

McGinley, W., Kamberelis, G., Mahoney, T., Madigan, D., Rybicki, V., & Oliver, J. (1997). Re-visioning reading and teaching literature through the lens of narrative theory. In T. Rogers & A. Soter (Eds.), Reading across cultures: Teaching literature in a diverse society (pp. 42-68). New York, NY: Teachers College Press.

Swain, M. (1995). Three functions of output in second language learning. In G. Cook & B. Seidlhofer (Eds.), Principle and practice in applied linguistics (pp.125-144). Oxford: Oxford University Press.

Swain, M. (2005). The output hypothesis: Theory and research. In E. Hinkel (Ed.), Handbook on research in second language teaching and learning. Mahwah, NJ: Lawrence Erlbaum.

Smith, F., Hardman, F., Wall, K. & Mroz, M. (2004). Interactive whole class teaching in the national literacy and numeracy strategies. British Educational Research Journal, 30 (3), 395-411.

Van Lier, L.(1996). Interaction in the language curriculum: Awareness, autonomy and authenticity. New York: Longman.

Walsh, S. (2006). Investigating classroom discourse. London: Routledge.

Walsh, S. (2011). Exploring classroom discourse. London: Routledge.

Wells, G. (1990). Talk about text: Where literacy is learned and taught. Curriculum Inquiry, 20, 369-405.


Monishita H. Pande is Assistant Professor, at the Centre for English Language Education (CELE) in Ambedkar University Delhi (AUD). She teaches proficiency courses and language theory courses. She is also involved in teacher training projects. Her research interests include English language teacher education, cross-linguistic transfer of skills, multilingual education and learner autonomy.