Edited by: Aaron Williamon, Royal College of Music, United Kingdom
Reviewed by: Diana Mary Blom, Western Sydney University, Australia; Tsutomu Fujinami, Japan Advanced Institute of Science and Technology, Japan
This article was submitted to Performance Science, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Established pedagogical theories for classical piano usually do not consider the essential relationship between the musical structure, whole body movements, and expression. Research focusing on musicians' expression has shown that body movements reflect the performer's understanding of the musical structure. However, most studies to date focus on the performance of a single piece at a time, leaving unanswered the question on how structural parameters of pieces with varied technical difficulties influence pianists' movements. In this study, 10 pianists performed three contrasting Romantic excerpts in terms of technical level and character, while motion data was collected with a passive infrared motion capture system. We observed how pianists modulate their performances for each of the three pieces and measured the absolute difference in percentage of duration and quantity of motion (QoM) between four expressive conditions (normal, deadpan, exaggerated, immobile). We analyzed common patterns within the time-series of position data to investigate whether pianists embody musical structure in similar ways. A survey was filled in by pianists to understand how they conceive the relationship between body movements and musical structure. Results show that the variation in duration between the exaggerated and deadpan conditions was significant in one measure for one of the excerpts, and that tempo was less affected by the QoM used than by the level of expression. By applying PCA on the pianists' position data, we found that the head QoM is an important parameter for communicating different expressions and structural features. Significant variations in head QoM were found in the immobile and deadpan conditions if compared to the normal condition, only in specific regions of the score. Recurrent head movements occurred along with certain structural parameters for two of the excerpts only. Altogether, these results indicate that the analysis of pianists' body movements and expressive intentions should be carried out in relation to the specific musical context, being dependent on the technical level of the pieces and the repertoire. These results, combined with piano teaching methods, may lead to the development of new approaches in instrumental lessons to help students make independent choices regarding body movements and expression.
While it is common knowledge that musicians' body movements contribute to the audience's understanding of the musical score and the performer's expressive interpretation of music (Vines et al.,
In the literature on music performance, two types of gestures have received more attention: effective or instrumental gestures, and sound-accompanying or ancillary gestures (Delalande,
In piano pedagogy, arms and hands are often at the heart of learning the instrumental technique. This approach, although motivated by virtuosity achievement, does not integrate other types of body movements, which coexist with the gestures involved in the production of the sound. To investigate how body movements are connected to musical expression and structural parameters, previous studies used different experimental conditions with gradual levels of expression. Davidson (
The following studies suggest that musicians' movements are often related to the rhythmic and phrasing structure of an excerpt, as well as to its technical difficulty and character. In order to identify the relationships between the rhythmic structure and similarity in upper body movements, pianists played two Chopin Preludes, similar in character, but different in terms of the phrasing structure (MacRitchie et al.,
Other studies investigated the impact of movements on auditors' judgment of musical performances and assessed which parts of the body better convey the expressive intention or emotion of the performance. In piano performance, head and upper torso movements provided meaningful information to auditors, who were asked to discriminate between performance conditions, while the hand movements did not (Davidson,
Expressive manipulations and musical individuality of music performances have been linked mainly to temporal variations (Palmer,
Although previous research has focused on the expressive intentions a performer conveys to an audience, it is not clear yet how the structural parameters of musical excerpts with various technical difficulties are embodied in pianists' physical gestures. The study of different Romantic excerpts with various levels of complexity performed by a group of pianists may yield different results that may eventually clarify how auditors perceive and react to musical gestures and expression. This study seeks to understand better how experienced pianists use body movements and timing in relation to structural parameters of pieces with varied difficulties and contexts. First, we evaluate how pianists modulate their performances in terms of duration and quantity of motion (QoM) when asked to play excerpts from the Romantic period in different performance conditions. Second, we investigate how both the structural characteristics of the pieces and the conditions impact the pianists' body movements. Third, we analyze the recurrent patterns of head movement among all pianists when performing in a normal condition. The aim is to visualize where in the score do pianists tend to move in a similar way to understand whether certain movements are dependent on the musical parameters or the physical constraints brought by the instrument. Finally, we assess whether pianists are aware of the way they use body movements in relation to the musical structure and the various expressive conditions. The goal of this research is not to assess whether pianists express their ideas intentionally or not, but to observe the trends and differences among a group of pianists and how various musical excerpts influence body movements and expression. The survey provided us with additional information as regard pianists' expressive decisions and intentions. We hypothesize that the movements from the extremities of the body, such as the ones from the hand or head, will be more accentuated when exaggerating or limiting the expression and that they will vary according to the excerpt performed. We propose that changes in amplitude of movements will be restrained in more demanding passages, such as chromatic passages, and that tempo will be more affected in the deadpan and exaggerated conditions than in the immobile one.
Ten pianists (average of 29.6 years old,
In a pilot study, which sought to evaluate pianists' body movements when performing different excerpts in terms of their structural features and technical levels, eleven pianists performed different Romantic excerpts three times in the following order: normal, deadpan, exaggerated and immobile conditions. Similarly to Davidson (
The three 30-s Romantic excerpts chosen for the current study are listed below:
Medtner Sonata Reminiscenza Op.38 (mes. 253–274) Chopin 4th Ballade (mes. 152–160) Chopin Impromptu (mes. 43–51)
Table
Analysis performed by the authors of each excerpt's structural characteristics and summary of results from previous measurements.
-Very dynamic and changing character |
-Impetuous and constant character |
-Peaceful and gentle character |
-All conditions performed faster than normal |
-All conditions performed slower than normal |
-Exaggerated and immobile performance performed faster than normal and deadpan conditions |
For the rest of the article, each excerpt will be referred to as the “Sonata,” the “Ballade” and the “Impromptu.” Each excerpt was performed in the same four expressive conditions as used in the pilot study (normal, deadpan, exaggerated and immobile conditions). The pianists played each excerpt once in each expressive condition (for total of 12 performances per pianist). Participants could choose the tempo they found appropriate to convey the expressive conditions. The order of excerpts was randomized for each participant.
At the beginning of the experiment, pianists filled in a demographic questionnaire and, at the end of the measurement session, pianists completed a survey to assess how they experienced body movements. Participants were asked questions on their understanding of the structure of the excerpts and how it influenced their musical interpretation. Performances were video recorded with a Sony Wide Angle video camera and audio recorded with a Sennheiser MKH microphone. Motion data were collected, at a rate of 240 frames per second, with a 17-camera Qualisys motion capture system, using 49 passive reflective markers put on the pianists' hands, elbows, shoulders, torso, head, and pelvis. The placement of markers on pianists' upper body and head is shown in Figure
As discussed earlier, previous studies have shown that acoustical and kinematic parameters are important indicators of expression in piano performance. The term
To measure the duration of each excerpt in each condition for every pianist, a filter was applied to the absolute value of the audio signal, using the Matlab function movmean, which calculates the moving average across a sliding window. The length of the window used was 200 frames for every participant. Then a sound intensity threshold of 0.001 dB was applied to the signal to mark the beginning and end of each performance. Since pianists could choose the tempo in which to perform each excerpt, the signals also needed to be temporally aligned to the musical structure. Therefore, the exact time of each important gestural event (i.e., notes or beats) was identified and annotated with the audio editor Audacity. The time coordinates of the position data were aligned to their corresponding musical events using a time-warping algorithm (Verron,
First, we used principal component analysis (PCA) to determine which body parts vary the most across the performance conditions for each individual pianist. We calculated the cumulative QoM for all the body parts (i.e., head, torso, shoulders, elbows, hands and pelvis) using the MATLAB Motion Capture (Mocap) Toolbox (Burger and Toiviainen,
After identifying these body parts, we measured the absolute difference in the total QoM between each condition and the normal condition. We calculated the cumulative distance traveled by the markers to analyze the differences between the expressive conditions. All pianists' cumulative QoM values were averaged together. For each excerpt, the QoM of the normal performances was taken as a reference point (0%) to compare against the values obtained in the other conditions. Then, a series of one-way ANOVAs was conducted for each measure to identify whether there were significant differences between the conditions.
In order to identify the sections of the score in which pianists perform with similar movements and to find the common patterns within the time-series of the position data, we used the instantaneous correlation algorithm developed by Barbosa et al. (
This section reports the results on a) the overall duration of the performance, b) the quantity of motion, and c) the recurrence of movements.
To evaluate how pianists vary the tempi in relation to the levels of expression and the different excerpts, we calculated the duration of every performance (total of 12 per pianist). The lengths of the performances are indicated in Table
Timing of performances of each condition for all pianists.
P1 | Normal | 43.32 | 25.89 | 31.97 | |||
Deadpan | 39.21 | –9.98 | 25.08 | –3.20 | 28.57 | ||
Exaggerated | 48.50 | 28.98 | 34.87 | +8.68 | |||
Immobile | 41.29 | –4.80 | 23.59 | –9.30 | 30.85 | –3.57 | |
P2 | Normal | 47.88 | 29.25 | 31.21 | |||
Deadpan | 55.83 | 33.49 | 39.64 | ||||
Exaggerated | 44.71 | –6.86 | 29.57 | +1.09 | 32.75 | +4.81 | |
Immobile | 50.2 | +4.73 | 30.79 | +5.13 | 28.28 | –9.88 | |
P3 | Normal | 40.45 | 26.98 | 32.29 | |||
Deadpan | 45.74 | 27.10 | +0.45 | 30.77 | –4.82 | ||
Exaggerated | 40.53 | +0.19 | 23.95 | 30.01 | |||
Immobile | 42.44 | +4.81 | 25.86 | –4.22 | 31.33 | –3.04 | |
P4 | Normal | 39.12 | 24.12 | 32.60 | |||
Deadpan | 36.48 | 21.28 | 27.74 | ||||
Exaggerated | 41.44 | +5.76 | 26.40 | +9.03 | 34.73 | +6.31 | |
Immobile | 39.01 | –0.28 | 24.25 | +0.54 | 33.35 | +2.28 | |
P5 | Normal | 35.59 | 25.70 | 39.93 | |||
Deadpan | 38.56 | +8.00 | 25.55 | –0.60 | 38.26 | –4.27 | |
Exaggerated | 42.59 | 27.05 | 42.44 | ||||
Immobile | 36.93 | +3.69 | 26.56 | +3.28 | 38.32 | –4.12 | |
P6 | Normal | 37.40 | 25.85 | 35.16 | |||
Deadpan | 34.85 | 24.27 | 31.95 | ||||
Exaggerated | 39.55 | +5.57 | 26.38 | +2.02 | 35.46 | +0.84 | |
Immobile | 39.25 | +4.82 | 25.80 | –0.19 | 35.95 | +2.21 | |
P7 | Normal | 43.11 | 39.91 | 39.27 | |||
Deadpan | 41.06 | 41.51 | +3.91 | 38.95 | –0.82 | ||
Exaggerated | 42.79 | –0.75 | 45.86 | 40.53 | |||
Immobile | 43.67 | +1.30 | 42.01 | +5.13 | 39.92 | +1.63 | |
P8 | Normal | 41.20 | 35.74 | 36.39 | |||
Deadpan | 43.11 | +4.53 | 37.05 | +3.58 | 36.15 | –0.67 | |
Exaggerated | 45.08 | +8.98 | 35.78 | +0.10 | 35.87 | ||
Immobile | 45.34 | 39.79 | 36.25 | –0.38 | |||
P9 | Normal | 47.36 | 34.74 | 39.71 | |||
Deadpan | 46.37 | 34.36 | –1.08 | 33.76 | |||
Exaggerated | 47.38 | +0.04 | 33.15 | 38.20 | –3.86 | ||
Immobile | 46.65 | –1.51 | 33.90 | –2.45 | 37.49 | –5.75 | |
P10 | Normal | 42.64 | 41.31 | 35.47 | |||
Deadpan | 46.07 | +7.73 | 44.12 | +6.59 | 34.56 | –2.61 | |
Exaggerated | 48.03 | 47.02 | 38.47 | ||||
Immobile | 43.63 | +2.30 | 43.28 | +4.67 | 33.70 | –5.12 |
Figure
Mean duration of performances for each condition and excerpt. The purple squares show the mean duration and the yellow bars the standard deviation between participants. The blue stars represent the longest performances, while the pink diamonds show the shortest ones. (N, Normal; D, Deadpan; E, Exaggerated; I, Immobile).
Absolute difference in percentage of duration for each measure per condition. The red line represents the reference point, the normal condition, against which the other conditions are compared.
As demonstrated in Figure
As shown in Figure
Contrary to the other two excerpts, pianists tend to perform the immobile condition faster (
PCA was used in order to verify which body parts were the most altered when pianists perform in various expressive conditions. Table
First PC's component feature and level of variance (in %) across all expressive conditions and excerpts for all pianists.
P1 | Head x-axis | 95.89 | Head x-axis | 95.69 | Head x-axis | 94.82 |
P2 | Rhand z-axis | 94.62 | Head y-axis | 84.45 | Head y-axis | 89.41 |
P3 | Head y-axis | 95.95 | Head y-axis | 96.12 | Head x-axis | 94.56 |
P4 | Head y-axis | 96.79 | Head x-axis | 93.63 | Head y-axis | 91.66 |
P5 | Head y-axis | 95.94 | Head y-axis | 94.78 | Head y-axis | 98.01 |
P6 | Head y-axis | 98.88 | Head y-axis | 98.44 | Head y-axis | 98.48 |
P7 | Head x-axis | 91.18 | Head x-axis | 96.66 | Head y-axis | 98.66 |
P8 | Head y-axis | 89.88 | Head x-axis | 93.03 | Head x-axis | 96.95 |
P9 | Head y-axis | 97.53 | Head y-axis | 98.17 | Head y-axis | 99.00 |
P10 | Head x-axis | 85.18 | Head x-axis | 94.72 | Lelbow x-axis | 96.50 |
Figure
Mean cumulative head QoM for each condition and excerpt. The purple squares show the mean QoM and the yellow bars the standard deviation between participants. The blue stars represent the largest values, while the pink diamonds show the smallest ones. (N, Normal; D, Deadpan; E, Exaggerated; I, Immobile).
Absolute difference in percentage of head QoM for each measure per condition. The red line represents the reference point, the normal condition, against which the other conditions are compared.
Medtner Sonata Reminiscenza - Results from the one-way ANOVA performed on the cumulative distance traveled by the head marker for the regions presenting significant differences between the normal condition and the other expressive conditions.
Region A | Bar 1 | 11.6 | 0.009 | Normal-Deadpan |
0.01 | Normal-Immobile | |||
Bar 2 | 11.9 | 0.05 | Normal-Deadpan | |
0.04 | Normal-Immobile | |||
Bar 3 | 10.5 | 0.007 | Normal-Deadpan | |
0.03 | Normal-Immobile | |||
Bar 4 | 12.3 | 0.003 | Normal-Deadpan | |
0.01 | Normal-Immobile | |||
Bar 5 | 11.1 | 0.01 | Normal-Deadpan | |
Bar 6 | 8.9 | 0.01 | Normal-Deadpan | |
Bar 7 | 12.8 | 0.05 | Normal-Deadpan | |
Region B | Bar 13 | 11.9 | 0.004 | Normal-Deadpan |
0.01 | Normal-Immobile | |||
Region C | Bar 16 | 6.2 | 0.03 | Normal-Deadpan |
Bar 17 | 9.2 | 0.005 | Normal-Deadpan | |
Region D | Bar 19 | 9.4 | 0.02 | Normal-Deadpan |
0.02 | Normal-Immobile | |||
Bar 20 | 14.7 | <0.001 | Normal-Deadpan | |
0.001 | Normal-Immobile | |||
Bar 21 | 13.2 | 0.002 | Normal-Deadpan | |
0.007 | Normal-Immobile |
Musical examples for Tables
During performances of the Sonata, pianists used on average 20.61% more QoM in the exaggerated condition than in the normal one, significantly higher than for the other two excerpts. Figure
As revealed in Figure
Chopin 4th Ballade-Results from the one-way ANOVA performed on the cumulative distance traveled by the head marker for the regions presenting significant differences between the normal condition and the other expressive conditions.
Region A | Bar 1 | 7.1 | 0.02 | Normal-Deadpan |
0.01 | Normal-Immobile | |||
Region B | Bar 3 | 6.2 | 0.04 | Normal-Deadpan |
Bar 4 | 16.3 | 0.001 | Normal-Deadpan | |
<0.001 | Normal-Immobile | |||
Bar 5 | 9.8 | 0.008 | Normal-Deadpan | |
0.01 | Normal-Immobile | |||
Bar 6 | 6.2 | 0.05 | Normal-Deadpan | |
0.03 | Normal-Immobile | |||
Region C | Bar 8 | 26.9 | <0.001 | Normal-Deadpan |
<0.001 | Normal-Immobile |
As shown in Figure
Chopin Impromptu-Results from the one-way ANOVA performed on the cumulative distance traveled by the head marker for the regions presenting significant differences between the normal condition and the other expressive conditions.
Region A | Bar 1 | 17.1 | 0.002 | Normal-Deadpan |
0.003 | Normal-Immobile | |||
Bar 2 | 22.6 | <0.001 | Normal-Deadpan | |
<0.001 | Normal-Immobile | |||
Bar 3 | 18.8 | <0.001 | Normal-Deadpan | |
<0.001 | Normal-Immobile | |||
Region B | Bar 4 | 14.6 | 0.004 | Normal-Deadpan |
0.005 | Normal-Immobile | |||
Bar 5 | 9.1 | 0.04 | Normal-Deadpan | |
0.02 | Normal-Immobile | |||
Bar 6 | 12.1 | 0.001 | Normal-Deadpan | |
0.001 | Normal-Immobile | |||
Bar 7 | 8.9 | 0.004 | Normal-Deadpan | |
0.002 | Normal-Immobile | |||
Bar 8 | 10.5 | 0.01 | Normal-Deadpan | |
0.002 | Normal-Immobile | |||
Region C | Bar 9 | 17.0 | <0.001 | Normal-Deadpan |
0.001 | Normal-Immobile | |||
Bar 10 | 11.8 | <0.001 | Normal-Deadpan | |
0.003 | Normal-Immobile |
To assess whether several pianists embody the musical structure in a similar way, the head position data and the motion recurrence map analysis were used jointly. In the top graphs of Figures
Medtner Sonata Reminiscenza-Top plot: average time-warped amplitude of the head movement in the four expressive conditions. The arrows delimitate the regions of interest. Bottom plot: motion recurrence map indicating the regions with high recurrence (red regions).
Figure
The beginning of the Ballade is marked with several regions of recurrent movement patterns, as shown in Figure
Chopin 4th Ballade-Top plot: average time-warped amplitude of the head movement in the four expressive conditions. The arrows delimitate the regions of interest. Bottom plot: motion recurrence map indicating the regions with high recurrence (red regions).
As Figure
Chopin Impromptu-Top plot: average time-warped amplitude of the head movement in the four expressive conditions. The arrows delimitate the regions of interest. Bottom plot: motion recurrence map indicating the regions with high recurrence (red regions).
Pianists filled in a survey about their perception of how they move in relation to the musical score. The survey includes open-ended questions related to the strategies pianists employed to convey the different expressive conditions, as well as to the types of movements they used to communicate the musical structure. Pianists' answers to the survey were then used to compare the movement data with pianists' personal assessment of their movements.
Most of the pianists mentioned that the arms are important for a better control of the fingers and the keys, and to play in a more natural and fluid manner. Using arm weight helps staying connected with the rest of the body and the instrument. The torso and head are generally used to communicate creativity and emotional investment. The hips, although less often mentioned than arms, help project the sound and are used for openness.
Most of the pianists stated that during the deadpan performance they decreased the QoM by restricting mainly the motion from the head and arms. The exaggerated condition required them to move with more amplitude, more arm motion and weight, and more hip movement. One pianist perceived that playing in the exaggerated condition created useless tension and imprecision in movements for all the excerpts, but particularly during the Ballade. For the same excerpt, two pianists reported that the immobile condition was easier to perform than the exaggerated condition because for that excerpt, playing with less movement is closer to a natural performance than playing with exaggerated ones. However, for the two other excerpts, pianists found that the immobile manner felt generally unnatural and prevented them from playing fluidly. To perform the immobile condition, they tried to limit the head and torso movements. However, playing with a restricted amount of movements while trying to be natural in the expression helped one pianist identify the regions in the score where excessive efforts were normally made. That pianist mentioned that, while restricting the movements, the focus was put on listening to the performance.
Pianists said that they used specific movement strategies to convey the respective structural parameters of each excerpt. Overall, pianists mentioned that the movements are mainly connected to the phrase structure, the dynamic shape and the melodic and rhythmic form, and that these parameters influence the amplitude of motion.
According to the pianists, the Sonata was performed with more hip and torso movements in passages that required playing a series of chords. For them, larger movements from the forearms and elbows were needed for crescendos in this excerpt, while the hips were more implicated before accentuated chords or notes and for attacks.
Three pianists specified that it was difficult to exaggerate the expression in very energetic passages, since these moments already required an investment from the whole body. For instance, for the Ballade, pianists found that the polyrhythm between the hands and the fast displacements of the left hand made it difficult to exaggerate the performance. Many variations in tempo make it difficult to keep a stable rhythmical precision. One pianist mentioned that because of the figurations (i.e., short succession of notes) contained in the excerpt and the many repetitive patterns, special attention on the finger and hand movements was necessary.
Because of its rhythmic simplicity and uniform writing, most of the pianists found that the Impromptu was the easiest excerpt to perform in different expressive intentions. They also claimed that the expressive variations were mainly done in very melodic parts, which naturally induce larger amplitude of motion in an exaggerated performance. Three pianists specified that fluid and larger arm movements are often used in rubato sections. The moderate tempo of this excerpt therefore gives more flexibility in the movements.
Pianists revealed that playing in a deadpan manner affected their sense of phrasing and several other expressive parameters, such as tempo and dynamics. Five mentioned that they noticed that their tempo was faster and more stable. They reduced the rubatos, the variations in nuances, as well as the contrasts naturally present between the hands. These same parameters were accentuated in the exaggerated conditions. Four pianists noted that certain regions might have been emphasized, while other passages might have been disrupted by an exaggerated expression because this condition made it difficult to control the sound. Again, most pianists found it difficult to play in the immobile condition, saying that it prevented them from rendering the appropriate expressive result. They mentioned feeling rigid and tense, and as a consequence, they did not perform the dynamic contrasts as well as they would have wanted. On the other hand, one pianist noted that she had the impression that she could play more efficiently while still achieving similar or better sound results.
This paper focused on the kinematic analysis of pianists' body movements in order to understand better how experienced pianists use body movements when performing different Romantic excerpts and when asked to play different performance conditions. We measured the duration and QoM of each performance and identified the regions in the score where pianists use common patterns of head movement.
We first looked at the variations in duration between the conditions for each excerpt. Although no distinct pattern was found among pianists regarding the overall duration of the performances of each expressive condition and excerpt, we found that the deadpan performances were generally played faster and the exaggerated performances slower as compared to the normal condition. Similarly to the results found in Thompson and Luck (
The similarity between the results of the present study, in which pianists performed three Romantic excerpts with contrasting difficulties and those found in Thompson and Luck (
Another purpose of the current study was to examine the effect of different pieces with various technical levels on pianist' head QoM and expression. By applying PCA on the pianists' position data, we showed that pianists' head QoM is an important parameter for communicating different expressions and the structural features of various excerpts from the Romantic period, which corroborates results from other studies (i.e., Davidson,
Davidson (
Similarly to Camurri et al. (
Pianists' answers to the questionnaire gave us important insights regarding the physical and acoustic strategies they can use to convey different levels of expression potentially associated with the musical structure. For most of the pianists, the arms movement and weight are considered as important motion cues to communicate their expressive ideas in a normal performance. Most of them found it difficult to exaggerate the performance in the Ballade, and found that performing in an immobile manner while trying to produce a normal expression was difficult for the Sonata and the Impromptu. For them, it was almost impossible to produce an accurate performance by restricting their movements the way they did.
This research provided new knowledge regarding the types of strategies pianists used to convey expressive intentions and structural parameters through body movements. Although pianists used varied strategies in terms of tempo and QoM to communicate different expressions, we identified similar trends in specific areas of the score. Our results indicated that when ten pianists performed three excerpts from the Romantic repertoire in difference expressive conditions (normal, deadpan, exaggerated and immobile): (a) the duration of performances was less affected by the QoM used than the level of expression regardless of the technical level of the excerpt, (b) the head QoM communicated well different expressions and structural features, and was only significantly different in the immobile and deadpan conditions when compared to the normal condition for all the excerpts, but mainly during the Impromptu, (c) the Sonata allowed more variations in amplitude of the head movements in the exaggerated condition than the two other excerpts due to the variety of elements in the writing, whereas the complex polyrhythm and melody in the Ballade prevented pianists from performing with exaggeration in the movements, and (d) recurrent head movements were found in specific regions of the score for the Sonata and the Ballade only. The results of this kinematic analysis, combined with common piano teaching methods, can benefit the field of piano pedagogy by helping teachers implement and integrate a more systematic approach in instrumental studio lessons in terms of accurate feedback related to movements and musical expression. Learners would be able to compare their movements to those of experienced pianists and become aware of the effect of that movement on the communication of expressive and structural parameters. Providing more systematic feedback in instrumental lessons can help students transfer teachers' explanations to various musical contexts so they may make independent creative choices, and aim to increase their musical communicative abilities.
Further studies investigating the ability of auditors to discriminate between a normal and immobile conditions could help evaluate whether reducing the movements in a performance affects auditors' perception of musical expression. The authors of this article have shown that even a slight modification in movements, such as the amplitude or acceleration of head motion can influence the sound parameters in a way that is noticeable for auditors (Massie-Laberge et al., unpublished manuscript). Additional work is also needed to identify whether there are distinct groups of pianists who tend to perform with similar body movements and whether these groups differ in terms of individual musical formation, influences and pianistic styles. Extensions of this work could also consider the impact of pieces from various musical periods on pianists' movements. Finally, expressive parameters, such as loud sound dynamics, accents, fast rhythms and rich texture, can be heavily dependent on the motion coming from the hip region. Complementary studies may examine the co-variations between the force applied on the piano stool and body movements to understand further the mechanisms involve in the movements, as well as weight compensation strategies used by pianists.
The raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher.
This study was carried out in accordance with the recommendations of the McGill University Policy on the Ethical Conduct of Research Involving Human Participants and the Tri-Council Policy Statement: Ethical Conduct For Research Involving Humans, McGill University Research Ethics Board II (REB-II). All participants gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the McGill University Research Ethics Board II, a unit within the Office of the Vice-Principal (Research&Innovation). REB File number: 101-0815.
CM-L, IC, and MW: design of the experiment; CM-L accomplishment of the experiment, data processing and analysis, and writing of report; IC and MW: research supervision and review of the report.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors would like to thank the performers for their participation and collaboration.