Introduction
This research explores two types of assessments that teachers make in instrumental music lessons. Using a broad conversation analytic approach, it highlights how these assessments produce different actions which are responded to differently by the student and thus have implications for the opening and closing up of the learning dialogue. The value in focusing on verbal interaction in music lessons, whether one-to-one, group or class based, is built on the notion that talk achieves things, that it is action orientated (Hutchby & Wooffitt, Reference HUTCHBY and WOOFFITT1998). The focus is on the functions that an individual's talk has at that moment in the interaction, and it is through social interaction that shared meaning and mutual understanding are accomplished (Goodwin & Heritage, Reference GOODWIN and HERITAGE1990). In conversation analysis, key features that help construct the interaction are observed; these may be characterized by turns taken in the talk, how sequences in talk are organized, repair in talk (e.g., corrections in speech or overlap), and pauses and silences in speech, for example. It is only within this level of detail that we can see, for example, how assessments are embedded in the structures of pedagogical interaction, where even small nuances have potentially significant implications for learning (Carlgren, Reference CARLGREN2009).
One-to-one instrumental lessons (see, for instance, Burwell, Reference BURWELL2006; Gaunt, Reference GAUNT2008, Reference GAUNT2010; Presland, Reference PRESLAND2005) offer a type of learning environment different from more typical classroom contexts which have previously been studied (e.g., Cain, Reference CAIN2013; Freer, Reference FREER2009) because the close relationship between the teacher and student is known to be absolutely key to successful learning (Gaunt, Reference GAUNT2008). This is not to say that other teacher-learner relationships and contexts are not of equal value to explore interactionally. Indeed, research has also shown the importance of instructional and pedagogical interactions in classrooms and rehearsals as a means for highlighting learning practices (e.g., Cain, Reference CAIN2013; Merlino, Reference MERLINO2014; Reed, Reference REED2015; Weeks, Reference WEEKS1996).
This previous research illustrates the varied nature of the delivery of music tuition, ranging from one-to-one contexts, small group ensembles and classrooms. Each of these environments is rich in instructional interaction and conversation analysis enables the researcher to examine, microscopically, the complexities of these – socially produced – learning practices. In addition to the talk, as playing and demonstrating using the voice or instrument is a fundamental part of the music lesson (see, for instance, Ivaldi, Reference IVALDI2016; Nishizaka, Reference NISHIZAKA2006; Tolins, Reference TOLINS2013) exploring how this is woven into the sequence of the interaction is essential in our attempt to understand the learning process.
The nature of the interaction between teacher and student is integral to any examination of formative assessment, and this interaction may be a mutually shared (dialogical) process, or one where the student is more a recipient of the teacher's instructional discourse (Black & Wiliam, Reference BLACK, WILIAM and Gardner2012) – the latter being evident in previous research. Studies that have been conducted into the more detailed elements of the teacher-student interaction have found that more spoken interaction takes place within the music lesson than actual playing, and that the teacher is responsible for most of the talking. For example, Karlsson and Juslin (Reference KARLSSON and JUSLIN2008) found, in their analysis of music lesson interactions, that 61% of the lesson was taken up with spoken discourse rather than playing, and that teachers accounted for 73% of the words uttered. They coded the verbal interaction as having five educational functions: testing (e.g., questions such as ‘should I continue?’); instructional (e.g., instructions, evaluations, such as excellent); analytical (e.g., in order to play this phrase, you must use this type of fingering); accompanying (guiding the interaction e.g., ‘yes, that's right’); and expressive (e.g., ‘more expression’). They also looked at the feedback that was given and the frequency, which was coded in four further ways: verbal instruction, modelling (i.e., the teacher demonstrates by playing), outcome feedback (i.e., whether playing was good/bad etc.), and metaphors (as used when referring to the expressive features of the playing). Karlsson and Juslin (Reference KARLSSON and JUSLIN2008) found that outcome feedback and verbal instruction were the most frequently used forms of providing feedback by teachers. These same features of the music lesson are evident in the extracts from lessons included within this paper. However, the micro-analytical approach offered through conversation analysis explores these features in a different way, that is, by exploring the function they are serving within the interaction itself.
Conversational analysis has been used in a range of educational settings such as academic supervision, secondary school maths lessons, and adult second language classes, and has explored the ways in which teachers give feedback (e.g., Vehviläinen, Reference VEHVILÄINEN2009); how students demonstrate knowing and understanding (e.g., Koole, Reference KOOLE2010); and the construction of evaluations of teacher-student responses (e.g., Zhang Waring, Reference ZHANG WARING2008). This latter study has indicated that when teachers give positive assessments (e.g., good, very good, excellent) students are less likely to explore alternative answers with their teacher, or express other concerns. For example, it has been shown that positive evaluations are often delivered slowly with falling pitch, delays in the teacher's assessment can often be heard by the student as potential negative evaluation, and repeated questions are heard as marks of a failed answer (Hellermann, Reference HELLERMAN2003). By looking at interactions at this level of detail, Lee (Reference LEE2007) posits that we can uncover important pedagogical practices in teaching.
Conversation analysis in music pedagogic interactions
Although still in its infancy, there is an increase in the research employing conversation analysis in music domains that offers an alternative micro-analytic approach to exploring music verbal interactions. For example, in formal learning contexts, Szczepek–Reed, Reed and Haddon (Reference SZCZEPEK–REED, REED and HADDON2013), and Reed (Reference REED2015) have looked at masterclasses where they focused on immediate versus delayed (‘now’ or ‘not now’) directives, and the teacher's (master) relinquishing moves when interacting with the student respectively. Tolins (Reference TOLINS2013) looked at how teachers made assessments of students’ playing through mirroring back – or quoting – specific elements through vocalizations, and Ivaldi (Reference IVALDI2016) looked at how students and teachers orient to learning and performance in their playing during a music lesson.
In a rehearsal context, Weeks (Reference WEEKS1996) was interested in how talk related to the non-verbal expressions within a group practice session, and the frequency with which sung instructions were used in place of talk. Weeks distinguished between verbal expressions and illustrative expressions (such as singing or counting) techniques that were used by the conductor to correct musicians’ playing. He found that illustrative expressions, as a form of correction, were rarely used in isolation, in contrast to verbal expressions, which were more commonly used on their own. Weeks argued that, in order for the illustrative expressions to be most effective for the players, they need to be clarified verbally as well. In terms of how corrections are made in the sequence of the rehearsal, Weeks found that when the conductor presented verbal and illustrative expressions together the verbal expression usually came first in order to explain how the illustrative expression should be interpreted. In addition, in cases where more than one illustrative expression was given, the first one tended to be an account of what the musicians had just played, as heard by the conductor (often delivered in an exaggerated or negative way), and the second one tended to be the conductor's account of what it should sound like (delivered clearly). Overall, evaluations of playing – although made verbally – were often achieved by using illustrative expressions as well.
These research examples of instructional contexts in music illustrate that assessments of playing are continually made by the teacher or conductor as a resource for learning that leads to performance. Within the context of this study the term ‘assessment’ is used rather than ‘feedback’ due to the interactional practices taking place within the lessons. Firstly, Pomerantz (Reference POMERANTZ, Atkinson and Heritage1984) states that ‘assessments are produced as products of participation; with an assessment, a speaker claims knowledge of that which he or she is assessing’ (p.57). The teacher, being in the category of expert, therefore speaks from this position of knowledge. Secondly, assessments are receipted by the student (mainly as an acknowledgement utterance, but also as an agreement or disagreement) therefore indicating that the student is part of the production of the assessment. Thirdly, an assessment might also come from the student themselves, in response to their own playing. It could be argued that this assessment may also be delivered from a position of self-knowledge; as the performer, they themselves hold the knowledge of what they were aiming to achieve, or how they were intending to negotiate certain difficulties with the instrument or score, for example. Assessments, therefore, are jointly produced, which may be viewed in contrast to that of feedback which is argued here to be more focused on just the teacher's response to the students’ displays of knowledge (Zhang Waring, Reference ZHANG WARING2008), as shown here through the actual performance given by the student.
The teacher's treatment of the assessments as feedback may be understood in context with previous feedback literature. For example, Hattie and Timperley’s (Reference HATTIE and TIMPERLEY2007) model proposes four levels of feedback that is centred around (1) the task: how well it is being performed, or how knowledge is being acquired; (2) process: how the task is being created or completed – the strategies that might be in place; (3) students’ self-regulation: the student's assessment, monitoring or regulation of their own learning; and (4) self feedback: feedback that is orientated to the person themselves. Whilst the findings of the current study are contextualised within this model of feedback, it is important to note, however, that a conversation analytical approach does not allow for an interpretation of the impact of these feedback levels to be made on the interaction per se, and therefore subsequent learning; rather, conversation analysis allows us to explore how the student is ‘doing’ showing knowledge, or ‘doing’ self-regulation.
Specific to assessments in music lessons is the ‘embodied production of assessments’ (Mondada, Reference MONDADA2009, p. 330). That is, where the teacher's non-verbal talk and gestures help to make up the assessment of the students’ playing. In conversation analysis research in general there has been a move from just looking at verbal assessments to embodied assessments; a notion that has particular relevance to music contexts. For example, Fasulo and Monzoni (Reference FASULO and MONZONI2009) found in talk between a designer and two tailors finalising an item of clothing that negative evaluations could be more sensitively handled when directed to a mutable object – the evidence – rather than making a verbal critique that is more directly related to the individual. In the data presented here, we see how the critique of the students’ playing is managed and negotiated through embodied assessments made by the teacher.
Drawing on conversation analysis research that has looked at assessments, and that which has focused on interactions in a musical domain, the aim of this research is to explore the types of assessments made by teachers of their students’ playing, and to examine the direct response to those assessments by the student. Black and Wiliam (Reference BLACK, WILIAM and Gardner2012) argue that in dialogical interaction, the teacher's interpretation of the students’ learning is crucial as this then leads to their next action and subsequent response. It is proposed in this study that this is not enough to fully understand how assessment and feedback is being interactionally produced. How the student interprets and responds to the teacher's interaction, whether it is dialogically produced, or merely receipted, becomes of equal importance. In other words, it is the intersubjectivity (the speakers’ shared understanding of each other's actions) of assessment and feedback talk that is the analytical basis to this study. The research questions are: (1) how do teachers make assessments of students’ playing and how are they responded to by the student? Thus illustrating how an assessment is socially produced; and (2) what function do the assessments serve in opening up the learning dialogue?
Method
Participants
Seven teachers working with 18 students from a UK music conservatoire took part in the study; all but two of the lessons involved different pairs of teachers and students. Instruments represented were violin (five lessons), piano (five lessons), clarinet (three lessons), cello (two lessons), and singer, viola, and oboe (one lesson of each). Four of the teachers took part in more than one lesson recording with different students.
Procedure
Video recordings of one-to-one instrumental lessons were made over a three-week period towards the end of Semester 2 (May–June) where a number of undergraduate and postgraduate students were close to taking their final recital assessments; thus much of the lesson time was spent working on the final touches of their performance. For other students in a different year group, they were in the earlier stages of learning new pieces. The structure followed that of a typical lesson in that the students played pieces of music upon which the teacher made comments. Similar to the research cited previously, the lessons were taken up with more spoken interaction than playing. 18 lessons were recorded, generating more than 22 hours of data; lessons ranged from 50 minutes to two hours. At the start of each lesson the researcher set up the recording equipment and then left for the lesson to continue as normal. The research was conducted in accordance with the British Psychological Society's ethical code, and had been approved by the institution's own ethics committee. Both teacher and student (and accompanist, if relevant) consented to the recording being made at the start of the lesson, and again at the end to ensure they were happy for that specific lesson to be used. Participants were made aware that they could stop the recording at any time, or ask for it not to be used. All extracts included here have been anonymised.
Data analysis
In the first instance, each recording was transcribed verbatim and transcripts of the lessons were read and re-read in order to discover recurring phenomena and patterns. Any instances where assessments were being made – either by the teacher or student – were highlighted, for example, assessments of the student's playing (e.g., good, bad) or instances where the student's playing was deemed incorrect or inadequate. The selected extracts were then subjected to basic Jefferson (Reference JEFFERSON and Lerner2004) transcription – the most frequently used transcription system in conversation analysis – where everything from pauses, sighs, to intonation etc. are documented (see Appendix for transcription symbols). In addition to spoken interaction, references to playing and other key non-verbal gestures (e.g., pointing to the score) were also transcribed. Given the range in length of lessons, particularly with those running to two hours, it is important to note that the timing in which events occurred within the lessons is only interactionally important, and not procedurally important. In other words, should the teacher decide to hear a new piece of music just before the end of a lesson, how the teacher shifts their talk to a change in topic and action is of significance, and not their reasons for doing so. In a similar way, conversation analysis does not allow for the impact of the length of the lesson, on the perceived quality or variety of talk, to be measured as its sole focus is intersubjective and in how one speaker interprets another person's talk.
It is beyond the remit of this paper to outline the many analytical tools that conversation analysts use in studying talk, and interested readers are encouraged to consult the many excellent books on the topic (e.g., Hutchby & Wooffitt, Reference HUTCHBY and WOOFFITT1998; Schegloff, Reference SCHEGLOFF2007; Sidnell, Reference SIDNELL2010; ten Have, Reference TEN HAVE2007). However, one key speech device that is evident throughout the extracts presented here is the acknowledgement token, that is, speech that acknowledges the preceding talk (Putcha & Potter, Reference PUCHTA and POTTER2004). Acknowledgment tokens have particular functions in talk: ‘Mm’ ‘mhm’ or ‘uh huh’ are techniques used to continue conversation and can be used in a number of ways. For example, they can be used to acknowledge what was just said by making a ‘receipt’ of the speaker's turn, encouraging the speaker to continue talking, or highlighting a potential problem or ambiguity in the talk (see, for instance, Puchta & Potter, Reference PUCHTA and POTTER2004). ‘OK’ or ‘right’ can be used by speakers to acknowledge that they have understood, while the use of these words also serves the function of moving the conversation on (see, further, Beach, Reference BEACH1993). ‘Oh’ can be used to indicate that what the speaker has just heard is new information. Acknowledgement tokens, therefore, have particular analytical interest in this study as they not only help to illustrate when students are ‘doing’ knowing and understanding as they respond to the teachers’ talk, but also are used in the way that speakers respond to each other when assessments are made.
In the study of academic discourse, Guthrie (Reference GUTHRIE1997) found that the most frequent occurrence of acknowledgement tokens was in informing and/or advising, that is, working through the problem. This was attributed to the fact that offering advice was the biggest focus of the academic session. In data that came from university sessions, 77% of the tokens were produced by the students, thus illustrating the teacher-student role within the discourse where, according to Guthrie, the teacher generates many more extended turns at talk than students, creating opportunities throughout for students to respond with acknowledgement tokens. In the extracts below, I will highlight how the delivery of different acknowledgement tokens provides different functions in the sequence of the talk.
In the readings of the transcripts two types of assessments were identified by the author as: (1) Explicit, definite assessments of the students’ playing; and (2) Performative, instructive assessments. Each is discussed in detail below.
Credibility of analysis
The interpretations made within conversation analysis are played out within the sequence of talk between the speakers, rather than being identified by the researcher; the researcher's role is to illustrate how the speakers are making the interpretations. As Speer (Reference SPEER2002) states, the conversation analyst ‘generally avoids imposing their own categories and agenda on their data, and instead considers what is going on from a member's perspective’ (p. 785). Furthermore, the inclusion of the extracts allows the reader to validate the analyst's illustrations of the speakers’ interpretations in the talk, or come up with alternative readings, if appropriate. Thus, readers can draw their own conclusions as to how, and the extent in which, assessments are being constructed.
Data Analysis
Exploration of the dataset highlights two types of assessments that the teacher makes of the students’ playing: (1) overall explicit assessments (such as good, excellent etc.), frequently made by the teacher as a way of providing immediate feedback for the student and moving on to a correction or closing of the sequence; and (2) performative, instructive assessments made by the teacher as a way of offering suggestions for improvement, or alternative ways of approaching the music. The types of assessments made generated different actions whereby the explicit assessments closed down the sequence, providing little opportunity for the student to enter into the dialogue, whilst the performative assessments highlighted that more work or interaction needed to take place as the teacher and student worked through the problem.
Explicit, definite assessments of the student's playing
The explicit assessments that were made either lead to a clear action from the teacher or closed the sequence (either to move onto a new piece, or to end the lesson), thus having a transitioning function. These assessments are referred to by the author as explicit in that they are clear and specific statements about the student's playing, rather than implied, assumed, or embedded. (The use of the term ‘explicit’ has also been documented in a similar way in Zhang Waring, Reference ZHANG WARING2008.) Collectively, the assessments made in this category are most in line with Hattie and Timperley's (Reference HATTIE and TIMPERLEY2007) first level of feedback that focuses on the task itself. This level is characterised by informing the student as to how correct or incorrect their knowledge or performance is, and tends to be the most common source of feedback provided to students in learning contexts.
Here, the teacher assigns value to the playing in terms of whether it is excellent, good, poor etc. Overall, these assessments tended to be short in delivery, repetitive, and were used to engineer a more detailed evaluation where further assessments regarding how the student could improve their playing were to follow (as discussed in Section 2 below). They are also characterized by a lack of dialogue between the teacher and student. For example, in the following extract the teacher (T - male) makes a clear assessment of the student's (S - female) playing, whilst swiftly moving on to provide an explanation. Two pianos are positioned side by side, one for the teacher and one for the student:
Extract 1:
In this extract the words laden with value are repeated twice (line 2). Often, repeats in interaction are a form of repair in troubled talk (see, for instance, Puchta & Potter, Reference PUCHTA and POTTER2004). However, here, they serve a different function, particularly as in this sequence there is no evidence to suggest trouble. Puchta and Potter state that repeats may highlight the ‘worthiness of an answer’ (p. 101) and this notion could be applied to assess the worth of a student's playing. In addition, the repetition of the same assessment indicates to the student that their level of playing has remained constant, and this gets upgraded in line 4 to excellent. The student's playing is ended by the teacher changing from ‘excellent’ to ‘OK’. As stated earlier (see Beach, Reference BEACH1993), ‘OK’ can be used as a form of closure and to move the conversation on. Indeed, this is heard by the student who stops playing (line 5) and, through her response ‘yes’, anticipates further commentary from the teacher. Here, the teacher summarizes her overall playing as ‘super’ but moves straight into a critique. This involves the teacher illustrating aspects that need improvement, on the piano, simultaneously supporting the illustrations with talk. Supporting Weeks's (Reference WEEKS1996) finding, the teacher evaluates the student's playing using a verbal illustration and then provides an illustrative expression via demonstration on the piano. A similar example of an explicit assessment followed by an evaluation is given in Extract 2. Both the female teacher (T) and male student (S) play on their own cellos:
Extract 2:
In this example it might be suggested that the repeated words, delivered quickly (line 3), serves the purpose of moving on; an evaluation is given, and the teacher is wanting to move on to give a demonstration. Here, however, this illustrative demonstration occurs before the verbal expression (line 4) and this could be because there is nothing in the teacher's talk to suggest a correction is required. The teacher makes an embodied assessment by playing the cello to quote what has been played and to contextualize the next assessment (‘so it's a lull’). The pause in line 6 instigates a new turn in talk, and the teacher continues to offer further guidance. The talk is more in keeping with correction talk in that the teacher is offering both verbal and illustrative expressions as she suggests that more care needs to be taken at the end of the phrase. In lines 7–8 the teacher continues to make embodied assessments by playing the relevant aspects of the score in line with their talk, ending with a use of a non-verbal gesture of turning the page in the score (line 10) to move the lesson on.
Further functions of explicit assessments were used to close the sequence, either to change the direction of the lesson or to end it altogether. This is shown in Extract 3 below which is taken from a piano lesson with a male teacher (T) who is talking to a female student:
Extract 3
The teacher achieves this transition between the two pieces by firstly ending the instruction on the current piece of music by stating that there is nothing else to say, corroborated by the fact that it is not followed up with further guidance, correction or playing, as seen in Extracts 1 and 2. Secondly, the lengthy pauses (lines 4 and 6 when the teacher is checking the score) indicate a potential shift in topic, and this is confirmed by the teacher who makes reference to a different piece of music, thus signifying to the student that the teacher has firmly moved on from the piece just played and that a new phase of the lesson has begun. These assessments serve a different purpose in that they are overall summaries of the teachers’ views of their students’ playing, and are also framed in such a way that they achieve closure of the sequence and, at times, actual lesson. Closing conversation is a highly negotiated activity and further research (e.g., Schegloff & Sacks, Reference SCHEGLOFF and SACKS1973) has looked at this in more detail.
In summary, the explicit assessments shown here provide a brief and immediate assessment of the students’ playing, and there is very little interaction from the student. However, unless they are used to negotiate the end of a section, these evaluations are always followed by ‘interpretative’ assessments, as explored in the next extracts, whereby the teacher offers suggestions for ways in helping the student to improve their playing, with the link between them often being an ‘OK’ or a pause. What is already evident in these extracts is that the assessments always follow a playing episode as they function as an effective turn from the students’ playing to the teacher.
Performative, instructive assessments
This second type of assessment explores a more complex process of evaluating the students’ work through the construction of performative assessments relating to the performance. Collectively, these assessments may be understood further in conjunction with levels two and three of Hattie and Timperley's (Reference HATTIE and TIMPERLEY2007) feedback model, that of process and self-regulation.
Whilst performative assessments were often generated from initial explicit assessments, they are given separate attention here as their delivery is often more intricate and open-ended (compared to the more explicit ‘excellent’ and ‘well done’ assessments seen above). As a result, these performative statements might be heard more as suggestions by the student rather than explicit instructions, and characterized by the following dimensions of playing (based on Chaffin et al., Reference CHAFFIN, LISBOA, LOGAN and BEGOSH2010): Basic techniques (e.g., ‘you need to bring the elbow forward a bit’); Interpretation – to include phrasing, quality of sound, intonation, and dynamics (e.g., ‘the intonation in the first phrase wasn't quite right’); and Expressive (e.g., ‘what you've got to get hold of is a kinda performing extinct’). In Extract 4 the male teacher (T) is evaluating the pianist's playing:
Extract 4:
This extract begins with the teacher articulating the aspect of performance on which Beth needs to work first: achieving a balance between the different parts (instrumental line) in the music. Beth's ‘Mhm’ might be regarded as an acknowledgement of the teacher's evaluation, but more specifically, is an acknowledgement that the speaker has not yet completed their turn (see, further, Guthrie, Reference GUTHRIE1997). This is further confirmed by the completion of the statement by the teacher in line 3 where he qualifies what he means by ‘top heavy’. Here, the teacher separates out the identities of the student (‘you’ i.e., player) and everyone else (‘we’ i.e., audience), and qualifies this further by locating the problem straight away – that they are not paying enough attention to the lower lines. The assessments made in this example have enabled the teacher to arrive at, and demonstrate, the solution to the problem. The emphasis on ‘you are not paying attention’ might be heard by the student as accusatory, putting the responsibility on the student to correct the problem. However, the teacher offers his own judgement (line 3) as to why the student's playing is too top heavy, and she does not question the teacher's assessment.
In addition, the teacher's use of ‘I don't think’ (line 9) enables him to manage, more sensitively, the aspects of the student's playing that require improvement. Wooffitt (Reference WOOFFITT2005) argues that in discursive psychology researchers should be careful to acknowledge that use of terminology such as ‘I think’ reflects the speaker attempting to negotiate their talk, rather than merely representing his own mental states.
In the following extract we see the (male) teacher trying to resolve one specific problem with the (female) student, and the student offering her assessments in an attempt to contribute to finding a solution; both teacher and student play on their own clarinet:
Extract 5:
In this extract the teacher is working in detail with the student on her playing of one note (‘slightly more on the first A natural’, line 5). We learn early on that the student has already made an improvement ‘that's closer’, and that when she plays it again she is ‘getting closer still’. However, the student has still not achieved this completely. Indeed, the student actually encourages an assessment from the teacher by asking ‘is that better’ (line 3). This is a rare feature in the extracts where the student explicitly requests that the teacher makes an assessment. The student's ‘mmm’ (line 8) acts as a continuer, anticipating a further turn at talk from the teacher, which he provides. The teacher then helps to resolve the problem by stating that they both need to experiment with it, with reference to experimenting suggesting a possible number of alternatives that need to be tried out before a conclusion is made. The teacher puts the responsibility on both of them to achieve this. In line 11 the student attempts to offer her own judgement as to what she thinks the problem is (although the pauses suggest that she is having problems articulating exactly what it is). Here the student also makes an embodied assessment as she points to the score in order to evidence her evaluation. She concludes that getting the same volume on the high, as well as the low notes is difficult to execute; the teacher confirms the student's judgement (line 13). Here, there is a greater sense of the student contributing to the learning process by offering her own interpretations through her assessments. Her reference to ‘be quite a nice slur’ (line 12) suggests a considered response that helps the student to demonstrate her understanding.
In the final extract a number of assessments are made by both the (male) teacher and (female) student in a piano lesson. These assessments not only relate to the playing of the piece, but also include judgements made by the student of what the teacher is going to say:
Extract 6:
Here, the teacher makes reference to the student having misjudged an aspect of her playing but it is not clear what this is until he starts playing and then stops at the exact point and informs her she was too loud (line 5). The continuer ‘mhmm’ (line 3) is used here as the ambiguity in what the teacher is referring to suggests that there must be a continuation of his turn in order to demonstrate what he means. This assessment is not news to the student as she refers to hearing him say ‘shhh’ whilst playing (line 6). The student, however, makes an unsuccessful attempt to question this (line 10) as he talks over her by giving another example. She makes another attempt by explaining why she is doing it, corroborating her decision with the use of the score (line 13). However, despite this justification that she is offering, she makes her own assessment that she knows what the teacher is going to say, thus answering her own question as to what she should be doing. The teacher confirms the student's assessment through the repetitive use of ‘yeh’ and does not need to give an explanation as the student has done it for him. However, this sequence is not resolved; the teacher continues to play and she confirms her understanding – but not as a question, but as a statement (line 17). The remaining lines of the extract suggest that the teacher is still working it out, as shown by his continual playing, the elongated ‘s::o’ (line 18), and then a correction of his own playing (of a different bar), thus suggesting that he is either confirming as to whether his assessment is correct, or giving a second demonstration to the student. The teacher then settles on what it should be ‘it has to be that’ (line 22), and it is at this point that the student fully understands what the teacher is trying to demonstrate. This is later confirmed by her pronounced statement ‘oh:: I see’ (line 23). According to Koole (Reference KOOLE2010), the student's ‘oh:: I see’ is a claim at understanding, rather than a demonstration of their understanding which the teacher does not ask her to demonstrate. This sequence is finally resolved – and thus indicated to the student – by moving on to a different part of the piece.
Discussion
Two types of assessments made by the teacher were identified in the interactions of music lessons. However, whilst they were presented separately for the purposes of exploring the differences between them, the assessments frequently appear in combination with each other, suggesting that explicit assessments were often followed by performance assessments. There is a sense, therefore, that the teacher is continually helping to improve the student's playing or interpretation of the music, despite referring to the student's playing as ‘very very good’ or ‘excellent’. Indeed, there are data where the teacher has described their student's playing as ‘outstandingly beautiful . . . it was almost definitive . . . it had me in tears’ and, yet, one week before their recital, the student could still improve their performance (as indicated by the teacher's use of ‘almost’). The explicit assessments tended to generate little response from students, as often the teacher's assessments closed the conversations. Performative assessments, in contrast, generated more continuers (e.g., ‘mhmm’) or opportunities for the student to take a turn in the talk. It could be argued that, in these instances, learning is still very much taking place within the dialogue, and, therefore, these assessments appear to be the most useful in terms of extending pedagogical interaction.
Both the explicit and performative assessments may be contextualized within Hattie and Timperley's (Reference HATTIE and TIMPERLEY2007) feedback model. Whilst the complexity of the interaction means that the assessments given move backwards and forwards between the four levels of feedback, the first level of task feedback is illustrated most frequently in the explicit assessments, and the second and third levels of process and self-regulation tend to be most evident in the performative assessments. With the latter, in particular, we see evidence (Extracts 5 and 6) of the student precipitating an assessment response from the teacher, whilst also reflecting on their own understanding of the problem at that moment, thus playing a key role in the self-regulation of their learning. It should be noted that there is some evidence in the dataset to suggest that assessments are also explicitly aligned to level four of Hattie and Timperley's (Reference HATTIE and TIMPERLEY2007) model, that of the self (e.g., ‘you'll be fine’). These tended to occur at the end of the lesson as a means for closing down the interaction and as a way of encouraging the student before their upcoming recitals. Further examples of such instances would need to be collected, however, in order to make more concrete analytical claims of these data.
It is evident in the extracts that the teacher is continually making suggestions throughout the lesson on how to improve the students’ playing (as played out in the performative assessments), but what are the consequences of this for the student? In most of the extracts we do not sense – in the students’ responses to the teacher's evaluations – that they are having a (an immediate) negative effect on the students’ playing. Indeed, there is a shared understanding that the subjective nature of music means that a musician's performance is continually open to interpretation in that there is no such thing as a definitive performance, regardless of the teacher's own performative evaluations. The teacher's evaluations reflect this: their judgements are generally constructed positively, offering further suggestions and interpretations, rather than being laden with negative evaluations or corrections of the students’ playing.
Analytically, the findings have helped to unpack some of the findings of Karlsson and Juslin's (Reference KARLSSON and JUSLIN2008) study. The five educational functions that they coded using content analysis of their data – testing, instructional, analytical, accompanying, and expressive – were all evident in these data, but by looking at assessments alone, we can see the complexity of these functions being played out, more explicitly, in the interaction. For example, Karlsson and Juslin's ‘instructional’ function included both instructions and evaluations given by the teacher, but yet their discourse, such as ‘play it from the top’ or ‘that was excellent’ (p. 216), appears to ignore the fact that these two aspects perform very different functions within the interaction. These, in turn, have implications for how the student responds. Furthermore, this study's focus on the assessments given to students carries much more significance in terms of how students show understanding of the feedback than Karlsson and Juslin's analysis has allowed.
In line with the previous research in music lessons, these data show that the teacher does most of the talking in the lessons, and that the types of assessments that are made are key to ensuring whether the interaction is opened up to the student or not. More often, students respond using acknowledgement tokens, showing little evidence of challenging or questioning the authoritative discourses. In his study of student responses to teachers’ explanations, Koole (Reference KOOLE2010) found that the teachers designed their talk around the student, and achieved this by responding to the information received by the student in terms of what they know and understand (or demonstrate in regards to music). Much of the way the teachers in this study responded to the information (the playing) given to them was through the use of assessments. The acknowledgement tokens ‘OK’ and ‘mmm’ often demonstrated the presence of assessments, with ‘OKs’ (from teachers) suggesting that an evaluation was about to follow, and ‘mmms’ (from students) demonstrating that they were in the middle or end of an assessment being given by the teacher.
In addition to the lack of turns made available to students, and their consequences for creating minimal opportunities to contribute to the learning dialogue, one further problem is that when teachers give positive evaluations, students are less likely to question them or seek alternative answers (Zhang Waring, Reference ZHANG WARING2008). This might be partly due to the possibility that, for students, receiving positive feedback is an acknowledgement that what they are doing is a valid and worthy account of the music, and thus have reached their end goal. More interactionally significant, however, is the way in which the teacher constructs their turn in talk. The extracts provided here show that the way teachers set up their turns when offering positive evaluations can limit the type of receipt available to the student. In other words, the teacher gives little invitation to the student to take up the next turn in the sequence in any way other than by using an acknowledgment token. Thus, while positive evaluations might serve, pedagogically, as sources of encouragement for the student, they do little to enhance the learning process.
So what do assessments mean for the conservatoire lesson? Here it is argued that assessments provide important functions within the instrumental lesson. Firstly, at the most basic level, verbal assessments allow the teacher to move from the student's playing to interaction, as an indication to the student to stop playing. Secondly, the nature of teachers’ assessments are different to those of the student in that the teachers’ evaluations are all forms of providing feedback, whether it is of how good/bad they are, their interpretations and what can be approved, and in how they are approaching the piece. Students’ assessments, which are rarely shown in the interaction, are used more as a way of checking with the teacher their understanding and evaluation of their playing. Thirdly, the assessments used by the teacher have helped to provide and demonstrate a solution to the problems that arise in the music lessons, acting as interactional diagnostic tools that enable the learning process to evolve. Fourthly, the notion that musicians can continually work on and improve their interpretation and performance of a piece means that the performative, instructive assessments, in particular, aid this process throughout. It is these assessments that could be considered the most valuable – albeit challenging – to students as they are encouraged to continually question and consider what it is they are trying to achieve in their performance. Finally, the way that the assessments were produced, using non-verbal features, provide support for the need to look at embodied assessments (see, further, Mondada, Reference MONDADA2009), and that certainly in this context, the multimodal features of the interaction cannot be ignored.
Further research in this area may also be developed. Whilst it is beyond the scope and epistemological foundations of conversation analysis to measure the impact that such interactions might have on students’ self-regulation within, and beyond, the lesson, the study does open up the possibility for researchers with a particular interest in music pedagogy to propose alternative research questions and methods in order to explore such a relationship. Developing the research using conversation analysis, future work might examine two further observations of assessments not documented in this study that suggest that (1) assessments are constructed in terms of the perceived progression of the student (for example, ‘that was much better than yesterday’) and in terms of themselves as a performer (for example, ‘you are not totally happy there’), and (2) teachers’ assessments that are challenged by the student. Furthermore, a natural way of extending this work would be to look at beginners’ lessons as it might be anticipated that there will be more overall explicit assessments than performance assessments in lessons in which students are novices, compared to the heavy use of performative assessments employed by teachers that are deemed necessary for developing expertise in musical performance at a higher level. It is possible that the assessments constructed here are unique to advanced music lessons and thus provide new insights into this specialized and complex learning environment.
Appendix
Transcription conventions (from Hutchby & Wooffitt, Reference HUTCHBY and WOOFFITT1998):
Additional conventions used in the transcription of playing and non-verbal gestures:
