Elsevier

Linguistics and Education

Volume 61, February 2021, 100886
Linguistics and Education

How teachers use prosody to guide students towards an adequate answer

https://doi.org/10.1016/j.linged.2020.100886Get rights and content

Highlights

  • How prosodic features display evaluative stance in teaching and exams.

  • Acknowledgments with wide pitch span treat a student answer as less adequate.

  • Acknowledgments with narrow pitch span treat a student answer as more adequate.

  • How teachers encourage students to continue answering without intervening.

Abstract

This paper focuses on the role prosodic features play in displaying evaluative stance in desk talks and oral exams in Norwegian secondary schools. We explore the extent to which teachers make available, to students, their treatment of student answers as more, or less, adequate with the acknowledgment token “ja” (yeah/yes). We found that, within extended question-answer sequences, acknowledgments with wider pitch spans treated a student answer as less adequate compared to acknowledgments with narrow pitch span. We show how the prosodic design of third turn acknowledgments is consequential to how both teachers and students proceed pursuing an adequate answer, and to how teachers subsequently attempt to remedy any negative interpretation of their evaluation. The paper demonstrates how prosodic features can direct the further trajectory of question-answer sequences, without, or prior to, making any distinct evaluation explicit. We discuss the relevance and limits of prosodic features to teachers’ assessment practices.

Introduction

In educational contexts, not all answers are treated as equally adequate as an answer to a question, and educators routinely steer students’ problem solving through various interactional means, towards an adequate answer e.g., (Lee, 2008). Research has shown how various linguistic and multimodal resources support this steering process (Lee, 2007; Margutti & Drew, 2014), but there remains further scope for identifying how different types of evaluation are associated with specific prosodic features such as pitch, loudness and voice quality. This paper explores how teachers and students use prosodic features systematically to differentiate more, or less, adequate answers in evaluation sequences. This research has implications for understanding in detail the range of subtle, implicit resources educators employ in their everyday practice to guide and support students.

Our analysis explores teacher – student interactions in “desk talk”1 and oral exams2 in Norwegian secondary schools. The target of the analysis is teachers’ design of “acknowledgment tokens”3 (Jefferson, 1984; 1993) produced in third position - that is, the turn following a teacher/examiner question and a student answer. Different forms of acknowledgment tokens, such as “mhm”, “yeah”, “okay”, are well documented across English languages (Gardner, 2001; Jefferson, 1984) and in other languages including Mandarin (Xu (2016). In Norwegian, like in English and other languages, acknowledgment tokens can be produced vocally both with lips closed (“mhm” or “mm” in English; “mm” in Norwegian), with opens lips (“uh-huh” in English, “øøh” in Norwegian), and with lexically more explicit agreement forms (e.g. “yes” or “yeah” in English; “ja” in Norwegian). Research on the relationship between lexical and prosodic forms sits at the interface between linguistics and conversation (see Couper–Kuhlen & Selting, 1996), and also between linguistics and educational practice. In educational contexts, acknowledgment tokens are one of many resources available for teachers to show their evaluative stance toward a student answer (Lee, 2007). In sequential terms, such evaluative use of acknowledgment tokens takes place in third position, or the Evaluation (or E) part of the well-known three-part IRE sequence (e.g., Macbeth, 2004; Mehan, 1979). Previous research has shown how teachers use acknowledgment tokens in the third position to provide an evaluative stance and also to move the question-answer sequences forwards (Lee, 2008). Acknowledgment tokens are thereby central to the kind of steering work that teachers so routinely do (Lee, 2007). But while a range of research has documented the systematic use of the third turn in educational contexts, we know less about the prosodic nuances in the way teachers display evaluative stance towards a student answer, and the extent to which students also recognise the teacher's evaluation based on prosody. There are also few studies in general that explore the prosodic variability of acknowledgment tokens.

In this paper we explore how teachers, from one third position acknowledgment token to the next, within extended IRE sequences, may modify the prosodic form of their acknowledgments according to how they treat the student answer. With this approach, we address how teachers use prosodic features to contextualise an answer in progress as “closer to”, or “further from” an adequate answer. Generally, acknowledgment tokens can represent less explicit (and more subtle) evaluations than third position assessments such as “exactly”, “correct”, and herein lies also their interactional potential: acknowledgment tokens may further steer the answerability of a question without treating the answer so far as (“on the record”) incorrect. We address how such subtle evaluation work might be relevant to evaluations both in the classroom and during oral exams. We provide further background on evaluative stance across institutional contexts (Section 1.1), the interactional work done with acknowledgment tokens (Section 1.2) and how these approaches apply to studies on teacher-student evaluations in oral exams and in the classroom (Section 1.3).

Studies on evaluations in institutional contexts, such as interviews and courtroom proceedings, demonstrate how the presence or absence of assessment forms (e.g., “good”) serve a specific interactional purpose. For example, interviewers actively avoid assessing a participant's answer to not influence its validity, restricting themselves to acknowledgment tokens including “yes” and “okay”, or a repetition of the answer (Houtkoop-Steenstra, 1996). However, interviewers may also produce more explicit assessments, as one way of “doing being personal” with the interviewee (Houtkoop-Steenstra, 1997). Furthermore, assessments may help shaping the interactional environment for getting institutional tasks done. (Antaki, 2002) found that interviewers, following an answer to a question, and an initial acknowledgment token (“right”, “okay”) would produce a high-grade assessment (e.g. “brilliant”) ahead of moving into a next item on the interview agenda, thereby marking the completion of an institutional objective as successful. In courtroom proceedings, on the contrary, lawyers regularly omit third position assessments of witness responses to show that they already know the answer to the question (Drew & Heritage, 1992). Thus, there is a reflexive relationship between third position assessments (their absence/presence and form) and the institutional context in which they take place: institutional roles and contexts both shape, and are shaped by, the types of third position assessments used.

Third position assessments can take explicit forms, such as “wonderful”, “well done”; however, they are not limited to lexical or syntactic content as they may be produced nonvocally or prosodically through phonetic variation of tokens such as “Ah:::” (Goodwin & Goodwin, 1987, p. 11), and also through the phonetic design of lexico-syntactic structures (Ogden, 2006). In the current paper we primarily focus on prosodic variability within uses of the lexical token “ja” in Norwegian, that is, the focus is on prosodic form rather than lexico-syntactic form of third position evaluations.

A range of conversation analytic research has demonstrated how conversationalists use acknowledgment tokens routinely and systematically to operate on ongoing talk. With the potential to be “noncommittal” to any form of evaluation or assessment – to align (Steensig and Sørensen, 2019) but also to affiliate Stivers (2008) with ongoing talk - the variability of interactional uses of acknowledgment tokens are wide-ranging, and they remain far from being adequately understood (Gardner, 2001). To identify their differential roles and treatments in talk requires careful sequential analysis.

Gail Jefferson was amongst the first researchers of conversation to study the range of interactional functions represented by acknowledgment tokens in conversations. Her research was predominantly based on data in English (American and British). Through detailed sequential analyses she identified what she called the “distinctive work of [acknowledgment] tokens” (1984, p. 7), ranging from simply acknowledging, to projecting some level of affiliation and agreement on the one hand, and, on the other hand, subverting, through passive recipiency, the ongoing trajectory of talk. Jefferson (1984; 1993) hypothesised and tested the extent to which interactional functions of acknowledgment tokens were associated with their form. She showed how listeners regularly move from “mhm” to “yeah” when they intend to speak next (Jefferson, 1993), suggesting that the form of acknowledgment tokens is associated with the potential to project transitions in sequence and speakership.

Most research on acknowledgment tokens has relied on lexical categories, investigating the extent to which various forms of acknowledgment tokens (e.g. “mhm”, “mm”, “uh-huh”, “yeah”) behave differently in interaction, without paying much attention to how prosodic features (forms) may contribute to functional distinctions. Jefferson (1984) hinted at the possibility of prosodic features doing distinctive work, observing that “although the token type is constant, the token shape changes” (p. 8). In one single case, Jefferson suggested a “yeah” with wider pitch contours projects engagement with the on-going topic, as opposed to flatly intoned “yah” [sic] which disengages with the topic (1993, p. 5), however, she never provided any systematic phonetic analysis of acknowledgment tokens in her research.

Gardner (2001), as part of a larger study on acknowledgment tokens, provided some prosodic analysis of acknowledgment tokens. For example, he described “mm” tokens generally as weaker forms of acknowledgment than “yeah”; however, depending on their pitch contour, the “mm” may display a continuer-type response (rising/flat intonation), as opposed to an affirming acknowledgment (rising-falling contour) and a somewhat disengaging acknowledgment (falling contour).

In Norwegian and Scandinavian languages, Svennevig (2001) and Lindström (1999) have researched “ja” as answers to questions/propositions in ongoing talk. In linguistic terms, “ja”, like “yes/yeah” in English, is typically used to affirm a positive proposition (e.g., proposition: “the capitol of France is Paris” – affirmation: “ja”). In contrast, “nei” (“no” in English) affirms a negative or positive proposition (e.g., proposition: “the capitol of France is (not) London” – negation: “nei”), and “jo” (no direct translation in English) is used to negate a negative proposition (e.g., proposition: “the capitol of France isn't Paris” – negation: “jo”) (see Svennevig, 2001). There are also hybrid versions of “ja” and “nei” in use, produced with a nasal onset of “ja” – “nja”. Similar uses are reported in English Jefferson (1978), as a way of ambiguating or weakening an affirmation. Svennevig (2001) shows how “ja” not only occurs as stand-alone utterances but to initiate a more elaborate response to wh-questions. His research shows how “ja” contributes to regulate the interactions as they project a delayed answer, a multi-unit response or a need for more processing. Lindström (1999) shows how elongated “ja” (and its equivalents “a” and what she labels curled “ja” pronounced with slight rise in pitch toward the end of the syllable) projects a non-aligning response. While Lindström (1999) investigated how prosody/phonetics of a specific type of “ja” contributes to the action it performs, none of these studies have looked at the acknowledgement tokens in third position in educational contexts – to which we turn next.

The third turn in classroom interaction is generally conceptualised as the Evaluation part in the three-turn sequence, Initiation – Response – Evaluation (IRE) (Mehan, 1979). However, as research has shown, the third turn offers more local and immediate contingencies than the generic notions of evaluation and feedback imply (e.g. Macbeth 2003; Lee, 2007). For example, teachers’ third turns respond in different ways to the correctness, adequacy and relevance of a student's response/second turn. The teachers’ rich range of responses in third position does important pedagogical and interactional work as they show the students what the teacher was setting out to achieve, thus building interpretive resources for the students which makes the students able to participate effectively in classroom interactions (Lee, 2008; Macbeth, 2011; Margutti & Drew 2014).

Extract 1 is an illustrative example of teachers provide such “steering work” (Lee, 2007), operating on the student's answer to their question. “TEA” stands for “teacher”, “S1” refers to “student 1” and “S2” to “student 2”.

Both acknowledgment tokens, “mm” (line 5) and “↑JA” (line 10) are produced in third position, i.e. as successive E turns in an extended IRE sequence. However, the difference lies in how they are used and treated in the ensuing interaction. As the teacher provides an acknowledgment “mm,” in line 5 (a “continuer”; cf. Schegloff, 1982), S1 re-initiates his answer in line 6. In comparison, following the teacher's “↑JA.” in line 10, the teacher proceeds with an explicit affirmation in line 12 and an account in lines 14-15. Here, both teacher and students treat the teacher's acknowledgement “JA” as an opportunity to make the positive evaluation more explicit: the teacher by providing the explicit evaluation, and the students by providing no further attempt to progress their answer. The question remains whether and how lexical and prosodic design features of the “mm” versus “↑JA.” may be systematically tied to such a differential conversational outcome.

Other conversation analytic studies have shown how teachers’ evaluation turns are used to indicate whether the student's answer to teachers’ question is (for all practical purposes) correct (e.g., Hellerman 2003; Margutti & Drew 2014). Hellerman (2003) has shown how absence of a teachers’ third position response is treated as negative assessment, making students offer an alternative answer. A teacher's repetition of students’ answers, however, is sometimes treated as a positive confirmation of the students’ answer: Hellerman (2003) suggests that, depending on prosodic form, the third position response may index either a positive evaluation and the end of a sequence (with features including matching pitch onset/peak with student and falling pitch contour), or an evaluation that indicates an incomplete answer and thereby leads to an extension of the sequence (with features including non-matching pitch onset/peak and final rising pitch contour). Based on this, and other research regarding the relevance of prosodic features to differentiate types of sequence expansion (e.g., Walker & Benjamin, 2017), there is already some evidence of systematic ties between evaluative moves and prosodic form. Margutti and Drew (2014), also focusing on positive evaluations in the classroom, show how teachers through five different formats of their third position response provide the students with various “interpretive resources for understanding more broadly the type of activity being enacted and teachers’ pedagogic goals” (p. 447). Margutti and Drew (2014) found some version of lexical repetition to be the most frequently occurring in classroom teaching, but also noted that third position particles such as “mm”, with no precise lexical meaning, nonetheless endorse prior answers, and that prosodic features help to disambiguate such an endorsement. That is, when treating an answer as “correct”, “mm”- type particles were associated with marked high onset pitch and rising-falling intonation, and louder than normal for the speaker. These are features similar to what we observed for the “JA” in Extract 1, and also to what Gardner describes as “heightened involvement” associated with acknowledgment tokens (Gardner, 1997, p. 132).

One may use a range of different terms to account for the target that the teacher is steering towards, including “correctness”, “completeness” and “adequacy”. In this paper we use the term “adequacy” to describe the range of cases we have to hand, specifically in terms of how prosody systematically foreshadows the evaluative stance of the next turn, as “more”, or “less”, adequate. This is in line with Macbeth (2003), who suggests that a third position evaluation does not in general remark on the correctness per se of the student's answer, “but rather on its adequacy as a reply, for the practical, instructional, discursive purposes at hand” (2003, p. 263). As Macbeth (2003) points out, “students can hear the adequacy of replies in the production of teachers’ third turns, and thus can know an adequate or failed reply without themselves knowing what a correct reply would be” (Macbeth 2003, p. 260). In this view, the evaluation process is less about distinguishing correct/incorrect answers than giving students some way of accessing the adequacy of the answer so far. One might also conceptualize this evaluation in terms of “completeness”; however, our view is that a measure of “completeness” assumes the student is on target for a complete answer, while the concept of adequacy also opens for the answer so far not being on target. Although the students can hear that something is not quite right in their answer (based on the teacher's acknowledgment token), it may be unclear whether they are indeed on the right path.

Section snippets

Data and Methods

The dataset builds on a collection of video-recorded teacher student interactions in six Norwegian secondary schools. It contains ten hours of “desk talk”, that is, interactions in which the teacher facilitates discussion with individual or groups of students as they work on tasks at their desks, and twenty hours of oral examinations. The collection of desk talk is extracted from lessons involving teacher-fronted instruction; however, our focus is on the desk talk interactions excluding

Analysis

Our analysis shows, first, how teachers treat answers unequally with regards to how adequately the answer fits the question (Section 3.1). In Section 3.1 we focus primarily on the sequential features, based on a core set of examples; in Section 3.2 we use the same examples to show how prosodic features tie with the sequential distinctions. In Section 3.3 we demonstrate how the identified ties between prosodic features and evaluation are treated as consequential by students. Finally, in

Discussion

Prosodic features are available to perform a wide range of actions. This paper has shown how teachers employ prosodic features, specifically width of pitch span, to carefully steer students towards an acceptable answer without explicitly rejecting or accepting students’ answer so far. The teacher's acknowledgment “ja” (and neighbouring forms such as “jo” and “(n)ja”) does not explicitly confirm nor disconfirm the adequacy of the student's response, but nevertheless makes an interpretation, and

Acknowledgments

This work was supported by The Research Council of Norway under grant 273417.

References (38)

  • C. Goodwin et al.

    Concurrent operations on talk: Notes on the interactive organization of assessments

    IPrA papers in pragmatics

    (1987)
  • M. Hayashi et al.

    Negotiating boundaries in talk

  • J. Hellermann

    The interactive work of prosody in the IRF exchange: Teacher repetition in feedback moves

    Language in Society

    (2003)
  • H. Houtkoop-Steenstra

    Probing behaviour of interviewers in the standardised semi-open research interview

    Quality and Quantity

    (1996)
  • I. Hutchby et al.

    Conversation analysis

    (2008)
  • G. Jefferson

    What's in a ‘Nyem'?

    Sociology

    (1978)
  • G. Jefferson

    Notes on a systematic deployment of the acknowledgement tokens “yeah”; and “mm hm”

    Papers in Linguistics

    (1984)
  • G. Jefferson

    Caveat speaker: Preliminary notes on recipient topic-shift implicature

    Research on Language and social interaction

    (1993)
  • G. Jefferson

    Glossary of transcript symbols with an introduction

  • Cited by (11)

    • Going beyond the post-observation's interactional agenda: The observers’ references to their practices and pedagogical understandings

      2022, Linguistics and Education
      Citation Excerpt :

      The findings of our study maintain that this experience is not only related to one's own practices but also related to the observation of other practitioners. Such findings may highlight the existing Language Teacher Education frameworks that make use of video-recordings of other practitioners to stimulate reflective accounts (see Sert, 2015; Sikveland, Solem, & Skovholt, 2021; Waring, 2021) and inform the ones based on teacher's own recordings (such as Hale, Nanni, & Hooper, 2018; Walsh, 2006) to expand and include the use of others’ practices to foster reflection. Having a peer-to-peer interaction in advice-giving creates a space that the asymmetries that are taken for granted in pre-service's mentor-mentee interaction are induced because of the peers’ equal positioning in the institutional sense.

    • Asking more than one question in one turn in oral examinations and its impact on examination quality

      2021, Journal of Pragmatics
      Citation Excerpt :

      During this same process, the examiner also uses a manual gesture, initiated during the first interrogative (lines 09–11) and maintained during the increment (line 11) and the second interrogative (lines 10–14). Thus, though producing syntactically complete TCUs, the examiner also holds her turn by means of gestural as well as vocal prosody (c.f. Sikveland et al., 2021), showing that the second interrogative does not orient to an answer as absent, as with vertical MUQs (Kasper and Ross, 2007), but pro-actively (re-) developing her question to set appropriate conditions for a relevant answer. After a micropause, the examiner formulates a second interrogative (“Do you remember that there is a shift when she takes the step out in reality?”)

    • Enhancing learning expériences through artificial intelligence: Classroom 5.0

      2023, Fostering Pedagogy Through Micro and Adaptive Learning in Higher Education: Trends, Tools, and Applications
    View all citing articles on Scopus
    View full text