Teaching and learning expressive music performance 1 Methods for teaching expressive performance

Several studies have demonstrated how verbal teaching using metaphors, verbal teaching explaining musical properties and aural modelling (Woody, 2000; Woody, 2006) can be effective for improving expressiveness in tertiary students and adult musicians. Verbal teaching using metaphor employs imagery to illustrate what the music should portray or sound like (e.g. Barten, 1992). Instruction using metaphors for enhancing expressiveness seems to be used extensively by conservatoire teachers (Lindström et al., 2003) and in various musical styles (Schippers, 2006). Metaphors are culturally defined (Schippers, 2006) and can be used in a variety of ways; for example to illustrate the musical character, for explaining technical instruction (Woody, 2006b) or to induce feelings in the performer (Lindström et al., 2003; Woody, 2000; 2006b; Schippers, 2006). It seems that some musicians translate metaphors into expressive musical devices via cognitive processes, while others personalize the imagery used to enhance their performance (Woody, 2006a). Although metaphors can be useful for explaining concepts that are difficult to describe verbally (e.g. Schippers, 2006) they can also be problematic when the imagery used is vague, complex or obscure (Persson, 1996; Schippers, 2006).

Verbal teaching explaining concrete musical properties describes how expressive tools such as variations in articulation, dynamics, and tempo can be employed to produce certain effects. Woody (2006b) found that instruction based on this type of teaching is inefficient as it requires significantly more practice time than instruction using metaphors or aural modelling.

Aural modelling is widely used (Lindström et al., 2003; Woody, 2000) and can be described as listening to expert performances from professional musicians, life or

via recordings, in order to build up internal aural representations of the music (Hallam, 1998) and to develop an expressive style (Meissner, 2017; Sloboda, 2005). Although some researchers proposed that aural modelling might be less effective than verbal teaching because there is a considerable amount of detailed information contained in an aural model (Sloboda & Davidson, 1996; Woody, 2000) research has shown that this is not the case. Woody found that all three instructional strategies, verbal teaching using metaphors, verbal teaching explaining musical properties and aural modelling, were effective for improving expressivity (Woody, 2006b).

Contrastingly, Rosenthal (1984) found that aural modelling was more effective than concrete verbal instruction, or modelling combined with verbal teaching. The inconsistency between these findings might be the differing levels of complexity of the test extracts and the content of the verbal instructions. Woody used short, simple extracts and the concrete verbal instructions as well as the metaphors were clear and focused on a few performance aspects. Conversely, in Rosenthal’s study a relatively difficult test piece was used with many detailed instructions. It seems likely that participants in Rosenthal’s study had to concentrate on too many performance aspects which might have hindered their expressiveness.

In a recent study, Van Zijl and colleagues explored the effect of musicians’ experienced emotions on performance characteristics (Van Zijl, Toiviainen, Lartillot, & Luck, 2014). Eight violinists were asked to play a musical phrase that was expressive of sadness in response to three different instructions: (1) playing while focusing on technique; (2) giving an expressive performance; and (3) playing while focusing on experienced emotions after a mood induction procedure. Experienced emotions were defined as ‘music-related felt or induced emotions of the performer, as opposed to practice- or performance related emotions, or perceived emotions’ (Van Zijl et al., 2014, p. 35, referring to Van Zijl & Sloboda, 2011, and Gabrielsson, 2001-2002). The violinists performed the melody three times in each performance condition. The authors analysed the articulation, dynamics, tempo, timbre, and vibrato of the performances in addition to interview data. Findings revealed that a focus on technique produced a technically controlled performance; ‘it seemed that the performers were just playing the notes as written, without adding much personal

expression to the music’ (Van Zijl et al., 2014, p. 44). Concentrating on expressiveness resulted in an extraverted and ‘projected’, audience-directed performance, while a focus on experienced emotions created ‘more introverted and personal performances’ (Van Zijl et al., 2014, p. 33). Six violinists thought that this ‘emotional’ performance was their best. As mentioned by the authors, these findings concern the performance of music expressive of sadness, and it is possible that these treatment conditions have different effects for extracts containing other musical characters. As all participants participated in all three treatment conditions in the same order, it is conceivable that the order and practice of the melody influenced the research outcome too (Van Zijl et al., 2014). Additionally, it would be interesting to know how listeners would evaluate the expressiveness of performances generated after these different conditions; which performance would an audience perceive as the most convincing or moving.

Furthermore, constructive and effective feedback is useful for improving expressivity (Woody, 2003). Tutors can give constructive verbal feedback, addressing technical and musical issues (Hallam, 1998). Additionally, audio or video recordings of students’ performances can be used for obtaining feedback (e.g. Juslin, Friberg, Schoonderwaldt, & Karlsson, 2004; Woody, 2000; 2001), or musicians can receive feedback from technology (Timmers & Sadakata, 2014; Juslin, Friberg, Schoonderwaldt, & Karlsson, 2004). Although research has demonstrated that it is possible to improve performers’ emotional expression by using computer software (Juslin, Karlsson, Lindström, Friberg, & Schoonderwaldt, 2006), musicians seem reluctant to adopt this technology (Karlsson et al., 2009) as they prefer actual tutors explaining strategies for enhancing expressiveness (Karlsson, Liljeström, & Juslin, 2009; Timmers & Sadakata, 2014). Additionally, the use of computer-assisted feedback can be seen as cumbersome or time consuming (Timmers & Sadakata, 2014).

The aforementioned studies investigating strategies for teaching expressivity were conducted with students in higher education or adults. Results obtained from these studies should not automatically be extrapolated to children as it might well be that the acquisition of expressive performance skills by young musicians is accomplished by other means, depending on their age, development and ability (Meissner 2017, p. 121; Woody, 2006b).

2.5.2. Methods for teaching young musicians expressive performance

As mentioned in the introduction, little is known about effective approaches for facilitating young musicians’ performance expression (Lisboa, 2008; McPhee, 2011). Several studies suggest that teaching in children’s instrumental lessons tends to concentrate on technique and sight reading skills (Karlsson & Juslin, 2008; West & Rostvall, 2003, cf. McPherson et al., 2012; Hallam, 2010). Ward (2004) found that aspects of teaching performance expression such as ‘encouraging spontaneity’, ‘individuality’ and ‘personal choice’ were not seen as important elements of instruction by music educators. Similarly, prospective music tutors in Fredrickson’s (2007) study intended to work on notational accuracy first, before working on expression.

Some small-scale studies have observed instrumental lessons to investigate methods for teaching children expressiveness. Brenner and Strand (2013) observed lessons of four, and interviewed five tutors, and found that aural modelling, verbal instructions and task repetitions were used for improving children’s (aged 7-15) performance expression. Brenner and Strand observed that tutors adjusted methods to children by exaggerating dynamics and melodic contour of verbal instructions, enlarging gestures accompanying verbal explanations, and overdoing aural modelling. Although tutors in this case study thought that emotion is important for musical expression they varied ‘on ideas about how and why to focus upon emotions and physical sensations during lessons as well as how to go about drawing these out from the students’ (Brenner & Strand, 2013, p. 92).

Griffiths (2017) conducted year-long case studies with four of his own pupils exploring how expressive gestures might be used for developing expressive performance skills of teenage pianists. He concluded that teaching should be learner- centred because of pupils’ differing motor skills, kinaesthetic awareness, technique and practice habits, and ought to build up confidence first. Similarly, Broomhead and colleagues (Broomhead, Skidmore, Eggett, & Mills, 2012, 2018) found that boosting confidence during pre-performance routines facilitated teenagers’ expressiveness in singing. McPhee (2011) interviewed two tutors and analysed video-recordings of one- to-one brass and cello lessons with two teenagers and noted that teachers used

various strategies: Verbal metaphor; Discussion of dynamics and articulation; Consideration of phrasing; Expressive markings in the score; Own recordings; Giving pupils some choices regarding expressive characteristics of the music. McPhee observed that these strategies achieved the same outcome as long as pupils had a clear understanding of how their playing changed. Additionally, she noted that pupils’ final performances followed the teachers’ rather than the pupils’ interpretation. McPhee proposed that beginners are unlikely to possess the technical skills to learn about expressive performance but suggests that teachers could allow teenage students some responsibility for making interpretative decisions. Likewise Graham (1998) recommends that teachers and pupils discuss the interpretation of a musical work to encourage students’ creativity in performance. In research with singers, Broomhead (2006) noticed that teachers’ enquiry, providing problem-solving opportunities and encouraging students’ initiatives (Broomhead, 2005), can be used to facilitate teenagers’ expressivity in choir rehearsals. Broomhead proposed that conducting consisting of expressive gestures and modelling combined with verbal explanations makes students dependent on their teacher-conductor. He suggested that singers need opportunities to make expressive decisions in order to construct their understanding of performance expression (Broomhead, 2005). In a study with four high school choirs, Maas (2016) found that conductors used various approaches, including analysis, metaphors and dialogue for practising expressivity. Teenage students in her study appreciated learning environments that invite dialogue and collaborative practices (Maas, 2016).

Furthermore, Davidson, Pitts & Correia (2001) proposed that teachers should employ physical movements and gestures to help their pupils to experience and understand the direction and content of the music they are learning. Likewise, Nijs (2018) suggested integrating movement-based technologies in instrumental teaching to provide young musicians with visual feedback and an embodied experience of music making. Although the ideas summarized above might be useful, these studies did not explore the effectiveness of the proposed methods on young musicians’ expressiveness and learning, or studies were conducted with a small number of participants (except for Broomhead, Skidmore, Eggett, & Mills, 2012; 2018).

A few studies investigated the effectiveness of instructional approaches during short experimental sessions. Vandewalker (2014) examined the relative effectiveness of aural modelling, concrete verbal instruction (through marking in the score combined with verbal explanation), and verbal instruction using metaphors, on wind-band students' (aged 12-13) expressiveness during a short (6-8 minutes) experimental session. Vandewalker found that aural modelling and verbal teaching using metaphors were significantly more effective than concrete verbal instruction for changing students’ use of dynamics, tempo and note duration. Contrastingly, Chester (2008) found no significant differences between aural modelling, concrete verbal instruction, verbal explanation using metaphors, and no instruction, on middle school instrumentalists’ expressiveness. Instruction time had been short, and her participants consisted of a mixed ability group which might explain these divergent findings. There are some limitations to Chester (2008) and Vandewalker’s (2014) studies. Both studies were conducted within a very limited time frame and they assessed pupils’ expressivity by measuring their use of only a few dynamic cues, namely articulation, dynamics and ritardando in Chester’s, and dynamics, tempo and note duration in Vandewalker’s study. Neither overall expressiveness nor expression of character was assessed, nor did they ask their participants for their views on their learning or the instructional strategies used.

Lisboa (2000) compared the effect of three instructional conditions on the practice and performance of three of her cello pupils. She compared the following approaches: ‘no instruction’, ‘analytical’ instruction and a ‘multi-modal approach’ which included singing before playing on the instrument; colouring the score; watching a video of a professional cellist; and discussions of general musical issues, such as interpretation and pupils’ approaches to practice and performance. Lisboa found that the multi-modal approach led to the most expressive performances and that her pupils concentrated on notation and fingering when left to their own devices. However, it is not possible to determine whether the combination of complementary methods of the multi-modal approach had been especially effective, or whether one particular method was successful, as four new strategies were introduced at the same time.

56 teenagers’ (aged 12-15) emotional expression in singing: (1) ‘Concrete verbal instruction’ explaining verbally to participants how to use their voices and performances to sing expressively; (2) ‘Aural modelling’ in which participants listened to a pre-recorded vocal track of a different melody with the intended emotional expression; (3) ‘Kinaesthetic exploration’ that encouraged participants to explore emotions in a ‘physical and active way’, by acting, drawing, walking or moving around; (4) ‘Audio-visual learning’ which consisted of viewing 20 pictures representative of a particular emotion while listening to a musical work that was expressive of the emotion that was conveyed in the pictures. Ebie found that the effectiveness of concrete verbal instruction was significantly lower than that of aural modelling and audio-visual learning, while the other three teaching conditions were similar in effectiveness for improving emotional expressiveness. Although Ebie’s aural modelling method seems less straight forward than conventional modelling where pupils hear a performance of the music they are learning, it had been effective for improving expressiveness. Perhaps the aural modelling, kinaesthetic exploration and audio-visual learning conditions in this study were similar in effectiveness because they all referred to emotion, whereas the concrete verbal instruction focused on technical aims. It might have been helpful for the participants to realise that musical emotion is important for expressive performance. However, the problem with Ebie’s (2004) study is that each participant performed the test melody 16 times; with four different expressions (happiness, sadness, anger and fear) after four different teaching strategies with one-week break between treatment conditions. Therefore, there could be an effect of learning, or possibly boredom, on pupils’ expressiveness.

As mentioned in the introduction, instrumental tutors in an exploratory action research project used various instructional strategies for teaching young musicians expressiveness, including enquiry and discussion, explanation of expressive devices, gestures and movements, metaphors and imagery, aural modelling, ‘projected performance’ and listening to own recordings. Tutors emphasized the importance of pupils thinking about the interpretation, and ‘owning the performance’. Although the teachers reported that all these methods had been effective in lessons, analysis of performance assessment scores did not show a significant improvement. Interestingly,

four out of five pupils who did improve their expressiveness scores were taught by tutors who used enquiry and discussion of musical character and the use of expressive devices (Meissner, 2017). This finding suggests that teachers’ enquiry and discussion is helpful for improving pupils’ expressive performance and it would be interesting to explore this further.

Several studies have suggested that young musicians tend to have ineffective practice strategies as they are inclined to play through pieces and focus on pitch and fingering (e.g. Lisboa, 2008; McPherson & Renwick, 2010; Pike, 2017; Pitts & Davidson, 2000; Pitts, Davidson, & McPherson, 2000). If pupils concentrate on reading from notation and technical issues during practice and performance, they might overlook issues of musical communication and expression. If this is the case, teachers’ questions and tutor-pupil dialogue regarding musical character and phrasing might facilitate young musicians’ learning of performance expression. Some might object that talking about the musical character or verbally labelling emotions is too complex or an extremely cognitive approach for teaching children. However, research has shown that most children can produce verbal labels to express happiness from the age of 2 and sadness, anger, fear surprise and disgust from the age of 3-4 years old (Wellman, Harris, Banerjee, & Sinclair, 1995 [sample 2-5 years olds]). Furthermore, a study by Brechet, Baldy, & Picard (2009 [sample 6-11 years old]) that explored children’s labelling and drawing of basic emotions, revealed that awareness and verbal labelling of happiness (90% correct responses at 6 years old) preceded that of sadness and fear (70% correct responses by 8 years old) followed by understanding of anger and disgust (40% correct responses at 11 years old). Interestingly, children in this study often labelled the person in the ‘neutral’ description as a happy person. Additionally, Bunte (2014) discovered that children in her study (aged 7-11) had musical concepts in mind which they could use for verbal descriptions of their musical preferences. Herbert and Dibben (2018) explored teenagers’ subjective experiences with music, particularly which meanings they attach to experimenter-selected musical excerpts. These authors found that their participants (10 - 18-year olds) frequently reported mental imagery, while especially the 10 – 12-year olds spontaneously described induced affects when asked for written impressions associated with the music. Based on these studies it

seems likely, that children and teenagers possess the mental concepts and verbal labels to talk about character or affect associated with their music by the time they start instrumental music learning.