Sage Journals: Discover world-class research

Abstract

Bodily gestures are essential in piano performance. They allow sound production and, at the same time, facilitate the communication of the expressive content of music. From pianists’ perspective, music expression-related parameters include not only single performance parameters (timing, sound intensity, articulation, etc.), but also more complex parameters (named hereafter abstract parameters), such as music structure features (e.g., phrasing) and extra-musical ideas (e.g., emotions, narratives, etc.). This systematic review aimed to investigate the impact of both performance and abstract parameters related to music expression on kinematics and muscle activity of expert pianists. As complementary objectives, we documented ontological and methodological differences between the studies included, and we addressed how music expression-related parameters affect pianists’ exposure to risk factors of injuries. The search strategy consisted of using concepts and keywords in Medline, Embase, SPORTDiscus, and Web of Science databases, and we followed the PRISMA guidelines. Sixteen studies were included. Eleven studies focused on performance parameters, four studies focused on abstract parameters, and one study addressed both performance and abstract parameters. Performance and abstract music expression-related parameters impacted pianists’ kinematics and muscle activity in a variety of ways. The specific effects were dependent on the type of task and the gestural variable investigated by studies. Important differences in ontological (performance or abstract parameters studied, gestural variable investigated) and methodological choices (experimental task and instrument used, data acquisition and processing procedures) prevent the establishment of a thorough dialogue between music research studies and biomechanics and motor control studies. A set of performance parameters (playing loud, playing fast, staccato articulation, large handspan chords) were identified as potential risk factors of injuries. Further interdisciplinary research mixing methods from empirical music research and biomechanics would help enhance knowledge on the impact of music expression on pianists’ gestures for both performance and injury prevention purposes.

Keywords

Piano performance expressive intentions biomechanics EMG PRMDs

Introduction

In music research, musicians’ body movements are usually characterized as gestures. The notion of gesture encompasses both the physical movement and its mental or cognitive aspects (e.g., expression of an idea or meaning; Jensenius et al., 2010). Analyzing pianists’ gestures is complex, as a great variety of multi-joint kinematic strategies and muscular activities can be used to produce a single piano tone (Furuya & Altenmüller, 2013). Moreover, pianists’ gestures are expertise-dependent, as novice and expert pianists show different movement strategies and muscle recruitment while playing similar tasks (Furuya & Kinoshita, 2007; Furuya et al., 2011). Musicians’ gestures during performance have been classified according to their function. Four functional categories of musical gestures have been reported in the literature: sound-producing gestures, sound-facilitating gestures, communicative gestures, and sound-accompanying gestures (Dahl et al., 2010; Jensenius et al., 2010). Sound-producing gestures are responsible for the effective production of sound (e.g., the striking of a piano key). Sound-facilitating gestures usually refer to bodily movements used to support sound-producing gestures for different music expression needs (e.g., the coordinated movements of musicians’ arms and trunk to shape the musical phrasing of the performance). These gestures are also called ancillary gestures in the relevant literature (Wanderley & Depalle, 2004). Communicative gestures are intended for communication with another performer and/or with the audience. Finally, sound-accompanying gestures are made in response to the music. Sound-producing and sound-facilitating gestures have been the primary object of study in the experimental research focusing on musicians’ gestures. These two gestural functions have been typically associated with different music expression-related parameters.

Music expression is a complex concept encompassing different phenomena and can be studied from a variety of disciplines (music theory, musicology, semiotic, semantics, psychology, neurosciences, performance science, biomechanics, among others). From pianists’ perspective, music expression-related parameters can be grouped in at least two main categories. First, performance parameters such as timing (related to management or adjustment of rhythm and tempo at both micro and macro levels of a musical piece; Repp, 1998), sound intensity (Drake & Palmer, 1993), and articulation (Repp, 1995). Parameters of this type are often called performance parameters because they are sound features effectively manipulated by pianists during practice and performance (i.e., piano tones can be louder/softer, longer/shorter, and time between tones can be longer/shorter). While performance parameters are defined (to a certain extent) by the composer in the musical score, pianists shape their performance of musical pieces by modulating these parameters according to their personal interpretation of the score (see, e.g., Gabrielsson, 1999). Pianists’ sound-producing gestures are generally associated with the effective control of performance parameters and have been investigated by research focusing on music biomechanics and motor control (for a review, see Furuya & Altenmüller, 2013; Goebl, 2017). One of the main goals of biomechanics and motor control studies focusing on music performance has been to assess if the manipulation of performance parameters may have an impact on exposure to risk factors of playing-related musculoskeletal disorders (PRMDs) (e.g., Degrave et al. (2020) focused on articulation and touch; Furuya et al. (2012) addressed loudness and tempo). This is an important topic for musicians, as lifetime prevalence of PRMDs ranges between 62% and 93% among professional instrumentalists (Kok et al., 2016). However, to the best of our knowledge, no systematic review has yet synthesized the current findings on how performance parameters affect pianists’ kinematics and muscle activity.

Music performance implies not only the effective control of performance parameters, but also the production and communication of more complex artistic content (either musical or extra-musical). Therefore, a second category of music expression-related parameters is needed to account for this essential aspect of the creative work of music performers. This category encompasses both music structure elements and concepts (e.g., phrasing, melodic and harmonic tension; Bigand & Parncutt, 1999) and extra-musical or semantic content (extra-musical ideas, such as a specific narrative, a picture, an emotion, a physical movement metaphor, and so on; Héroux & Fortier, 2015; Juslin & Västfjäll, 2008). As these music expression-related parameters usually refer to complex musical and extra-musical ideas rather than to specific parameters, we name them abstract parameters in this review for writing and reading simplification purposes. These abstract parameters are usually associated with sound-facilitating gestures in music research literature addressing musicians’ gestures, which have been studied in relation to structural music elements (e.g., phrasing and music tension; Vines et al., 2006) and music expression playing conditions (Davidson, 2007; Massie-Laberge et al., 2019; Thompson & Luck, 2012). Studies investigating sound-facilitating gestures have used various data collection tools similar to the ones used in the field of biomechanics, particularly 3D motion capture systems, while focusing often on markers’ linear kinematics rather than on more advanced methods intended to analyze human movement. These studies have shown that changes in music expression conditions (e.g., normal, exaggerated, and deadpan playing conditions) impact both movement of markers placed on performers’ body and the overall duration of the musical excerpt played (Massie-Laberge et al., 2019; Thompson & Luck, 2012).

Performance and abstract parameters are closely interrelated: abstract parameters can impact or inform performers’ choices in relation to performance parameters, and changes in performance parameters might result in changes in abstract parameters. Despite this interrelated nature, sound-producing gestures have been mainly addressed by studies focusing on the biomechanics of music performance, while sound-facilitating gestures have been addressed by empirical music research studies focusing on cognitive, musical, and learning aspects of music performance, usually from an embodied cognition theoretical perspective. As a result, there appears to be a limited dialogue between the studies examining the impact of performance and abstract parameters related to music expression on pianists’ kinematics and muscle activity. Ding (2024) reported a similar observation in violin performance research related to music expression, where biomechanical elements (e.g., muscle activity, joint kinematics) and musical elements (e.g., acoustical features) have been generally addressed in isolation.

This systematic review aimed to establish a dialogue between experimental research on expert pianists’ sound-producing and sound-facilitating gestural functions by investigating how both performance and abstract parameters related to music expression impact the kinematics and muscle activity of expert pianists. In addition, we addressed the following two complementary research questions: i) what are the ontological (i.e., focus of studies, gestural variable investigated) and methodological (i.e., data collection tools, experimental tasks, musical instrument used) differences between the available studies on music expression and pianists’ gestures, and ii) how do music expression-related parameters affect pianists’ exposure to risk factors of PRMDs.

Methods

The present systematic review was reported in accordance with Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) guidelines (Page et al., 2021).

Search Strategy

A professional university librarian assisted with the creation and execution of the search strategy. Keywords within each concept were combined with the OR Boolean operator, and three concepts were combined with the AND Boolean operator (Table 1). The electronic databases OVID (Medline, Embase), EBSCO (SPORTDiscus with Full Text), and CLARIVATE (Web of Science) were systematically searched. The last search was performed on December 14, 2022. Following the search, all identified studies were collated and uploaded into EndNote (X9, Clarivate Analytics, USA) and duplicates were removed. Then, references were uploaded to the Web-based system review software Covidence for the study selection process.

Table 1.

Concept and keywords used to identify relevant articles.

Concept	Related keywords
Participant	Piano OR pianist
Music expression-related parameters	Performance OR expressive OR expression OR “musical structure” OR “structural parameters of pieces” OR intention OR communication OR patterns OR loudness OR louder OR tempo OR tempi OR duration OR velocity OR velocities OR “sound intensity” OR timing OR rhythm OR articulation OR timbre OR “musical tension”
Kinematics and muscle activity	Movement OR motion OR gesture OR kinematic OR biomechanics OR electromyography OR electromyogram OR electromyograph OR muscle OR muscular OR position OR posture

Eligibility Criteria

Publications in English in peer-reviewed journals were included. Specifically, the focus centered on experimental studies on piano involving expert or professional adult pianists. The task was required to be a musical excerpt or a series of notes, encompassing either isolated keystrokes (such as repeated notes or octaves) or technical melodic exercises (such as scale or arpeggio tasks). Consistency across participants was a key requirement: all participants were to undertake the same task. Moreover, the designated task had to incorporate at least one change in expression-related parameters: performance parameters (e.g., sound intensity, articulation, tempo), abstract parameters (e.g., use of experimental conditions such as deadpan, immobile, normal, exaggerated), or type of playing context used (e.g., tone sequences, chord sequences, etc.). The dependent variable for analysis was required to be based on the measurement of pianist muscle activity and/or kinematics during the piano task. Lastly, studies comparing novices and experts were included.

Conversely, studies with only novices were excluded, as novice and expert pianists show different movement strategies and muscle recruitment while playing (Furuya et al., 2011). Additionally, studies focusing on the analysis of performance parameters without a formal analysis of gestural features (muscle activity or kinematics) were excluded. Finally, studies involving duo performance were excluded (as it involves additional elements, such as synchronization and communication with the other performer, that are not present in a solo piano performance and may therefore introduce risk of interference).

Study Selection

A first screening was performed using only titles followed by a second screening of abstracts performed by two independent reviewers (authors RM and CT). This process determined whether a study was to be included based on the predetermined eligibility criteria, while minimizing reviewer bias. Subsequently, a final sorting was performed by two authors (RM and CT) using the full texts of the remaining studies. The list of selected articles was discussed between authors until consensus was achieved. All authors’ conflicts were discussed internally and resolved by a third author (FV).

Quality Assessment

In line with the PRISMA guidelines, the methodological quality of the studies included in this review was evaluated by two independent reviewers (RM, FV) using a modified version of the Downs and Black checklist (Downs and Black, 1998). Out of 27 items, 11 items were identified as relevant by the authors, which allows the evaluation of overall reporting bias (items 1, 2, 3, 4, 6, 7, 10) and internal validity bias (items 16, 18, 20, 23) in the included studies. For this review, two items were replaced to ensure relevance for the investigated literature: item 4 (“Are the interventions of interest clearly described?”) was replaced by “Are the experimental conditions clearly described?”, and item 23 (“Were study subjects randomized to intervention groups?”) was replaced by “Were experimental conditions randomized for participants?”. The items were scored as 1 (“yes”), 0 (“no”), or “UD” (“Unable to Determine”). The maximum total consists of 11 points per study. Each study was assigned a score of “high” (≥75%), “moderate” (60–74%), or “low” (≤60%) (Desmyttere et al., 2018). When the quality scores differed among the reviewers, consensus was finally reached through discussion.

Data Extraction

Main Results

The details of each study were extracted by the author RM and were verified by two authors (CT or FV) in a table containing i) general information (author names and year of publication, study design); ii) methodological information (participant characteristics, study type, body segment studied, device used, experimental task, independent variables, and dependent variables); and iii) main findings.

Risk Factors of PRMDs

PRMDs can be the result of a vast variety of biomechanical, psychosocial, and environmental risk factors (Adams & Turk, 2018). Biomechanical risk factors of PRMDs relate to overuse and misuse factors, including repetitive movements, increased muscle load, muscle fatigue, and poor/static postures (Rousseau et al., 2021). Therefore, to determine how expression-related parameters affect pianists’ exposure to risk factors of PRMDs, RM and FV assessed the main findings of each study included in the present review in relation to the above-listed overuse and misuse risk factors of PRMDs.

Inter-Rater Agreement

Cohen's kappa was calculated to analyze inter-rater agreement for the overall study selection process. Kappa values 0.00–0.20 indicate poor, 0.21–0.40 fair, 0–41–0.60 moderate, 0.61–0.80 substantial, and greater than 0.81 almost perfect agreement.

Results

Search Results

The search yielded a total of 657 results. Following screening, 38 full-text articles were assessed for eligibility, of which 22 were excluded. A total of 16 studies (including a total of 167 participants) were eligible for this review (Figure 1). Inter-rater agreement for the overall study selection process yielded a Cohen's kappa of 0.78, suggesting a substantial agreement between the two authors (RM and CT).

Figure 1.

PRISMA diagram of the study screening process and article selection.

Quality Assessment

The median quality score of the included studies was 82% (range from 55% to 100%), indicating a moderate to high quality (Table 2). Eleven studies were of high quality (Dalla Bella & Palmer, 2011; Degrave et al., 2020; Furuya et al., 2010, 2011, 2012; Goebl & Palmer, 2013; Goubault et al., 2021; Massie-Laberge et al., 2019; Thio-Pera et al., 2022; Verdugo et al., 2020a, 2020b; Wong et al., 2022), four studies were of moderate quality (Sforza et al., 2003; Shoda & Adachi, 2012; Thompson & Luck, 2012; Turner et al., 2022), and one study was of low methodological quality (Castellano et al., 2008).

Table 2.

Methodological quality assessment scores of included studies using the modified version of Downs and Black checklist.

Music-expression-related parameters	Author & date	Reporting							Internal validity (bias)				Score (%)	Quality
Music-expression-related parameters	Author & date	1	2	3	4	6	7	10	16	18	20	23	Score (%)	Quality
Performance parameters	Dalla Bella and Palmer (2011)	1	1	1	1	1	1	0	1	1	1	UD	82	High
	Degrave et al. (2020)	1	1	1	1	1	1	1	1	1	1	UD	91	High
	Furuya et al. (2010)	1	1	1	1	1	1	0	1	1	1	UD	82	High
	Furuya et al. (2011)	1	1	1	1	1	1	1	1	1	1	UD	91	High
	Furuya et al. (2012)	1	1	1	1	1	1	0	1	1	1	1	91	High
	Goebl and Palmer (2013)	1	1	1	1	1	1	0	1	1	1	UD	82	High
	Goubault et al. (2021)	1	1	1	1	1	1	1	1	1	1	1	100	High
	Sforza et al. (2003)	0	1	1	1	1	1	1	1	1	0	UD	73	Moderate
	Thio-Pera et al. (2022)	1	1	1	1	1	1	0	1	1	1	UD	82	High
	Turner et al. (2022)	1	1	1	1	1	1	0	1	0	1	UD	73	Moderate
	Verdugo et al. (2020b)	1	1	1	1	1	1	1	1	1	1	1	100	High
Abstract parameters	Castellano et al. (2008)	1	1	0	0	1	1	0	1	1	0	UD	55	Low
	Shoda and Adachi (2012)	1	1	1	1	1	1	0	1	1	0	0	73	Moderate
	Thompson and Luck (2012)	1	1	1	1	1	1	0	1	1	0	UD	73	Moderate
	Massie-Laberge et al. (2019)	1	1	1	1	1	1	1	1	1	0	1	91	High
Performance and abstract parameters	Wong et al. (2022)	1	1	1	1	1	1	1	1	1	1	1	100	High

1 = Yes; 0 = No; UD = Unable to Determine. Quality score: “High” (≥75%), “Moderate” (60–74%), “Low” (≤60%). Studies have been classified by the performance/abstract parameter manipulated. Q1: clear aim, Q2: clarity of reporting outcomes, Q3: clarity of participants’ characteristics, Q4: clarity of experimental conditions, Q6: description of main findings, Q7: estimation and report of random variability, Q10: reporting actual probability values, Q16: clarity of probable data dredging, Q18: appropriate statistical tests, Q20: accuracy of outcome measures, Q23: randomization of experimental conditions.

Studies’ Characteristics

Out of the 16 studies included, 13 were cross-sectional studies (Dalla Bella & Palmer, 2011; Degrave et al., 2020; Furuya et al., 2010, 2011, 2012; Goebl & Palmer, 2013; Goubault et al., 2021; Massie-Laberge et al., 2019; Sforza et al., 2003; Thio-Pera et al., 2022; Thompson & Luck, 2012; Verdugo et al., 2020b; Wong et al., 2022), and three were cross-sectional case studies (Castellano et al., 2008; Shoda & Adachi, 2012; Turner et al., 2022). Eleven studies focused on the modification of performance parameters (Dalla Bella & Palmer, 2011; Degrave et al., 2020; Furuya et al., 2010, 2011, 2012; Goebl & Palmer, 2013; Goubault et al., 2021; Sforza et al., 2003; Thio-Pera et al., 2022; Turner et al., 2022; Verdugo et al., 2020a, 2020b), four studies focused on the modification of abstract parameters (Castellano et al., 2008; Massie-Laberge et al., 2019; Shoda & Adachi, 2012; Thompson & Luck, 2012), and one study addressed both performance and abstract parameters (Wong et al., 2022). To enhance the readability and clarity of this systematic review, Table 3, which summarizes the findings in each study, has been subdivided in three sections: Table 3.A reports on studies investigating changes in performance parameters, Table 3.B reports on studies investigating changes in abstract parameters, and Table 3.C reports on the only study investigating changes in both performance and abstract parameters. The studies in Table 3.A are classified by the performance parameter manipulated, and the studies in Table 3.B are presented in chronological order.

Table 3.A.

Summary of studies included in the review that investigated the impact of changes in performance parameters on kinematics and/or muscle activity of pianists (the column “Main outcome” summarizes one or maximum two main outcomes relevant for the present literature review).

Author and date	Participants	Study design	Body segment studied	Measurements	Experimental task	Independent variables	Dependent variables	Main outcome
Furuya et al. (2011)	N = 5 (2 males, 3 females) Professional pianists with more than 20 years of formal classical piano training	Cross-sectional study	Right upper limb (finger and elbow joints) 2 right upper-limb muscles: flexor digitorum superficialis (FDS), and extensor digitorum communis (EDC)	3D motion capture system (120 Hz, 23 markers) 2 EMG sensors (960 Hz) Digital piano	Alternate keystrokes	Five tempi: 70, 90, 110, 130, and 260 bpm	Peak angular velocity (degree/s) of the elbow, little finger, and thumb Mean joint angle (degree) at the metacarpo-phalangeal (MCP) and proximal phalangeal (PIP) joints for the index, middle and ring fingers Mean muscle activation (%MVC)	Expert pianists used hand pronation/supination (a degree of freedom at the elbow) to reduce metacarpophalangeal muscle load at the five tested tempi*
Sforza et al. (2003)	N = 5 (3 males, 2 females) Professional pianists	Cross-sectional study	Right hand	3D motion capture system (100 Hz, 5 markers) Digital piano	A scale	Three tempi: 80, 112, and 160 bpm	Overlapping coefficients between trajectories for all fingers	Repeatability of finger movements was lower in concert pianists than in teachers and learners
Dalla Bella and Palmer (2011)	N = 4 (1 male, 3 females) 16.3 years of piano performing experience	Cross-sectional study	Right hand	3D motion capture system (120 Hz, 5 markers) Digital piano	Two melodies	Five tempi: 60, 180, 210, 240, 250 bpm	Mean movement amplitude (mm) and mean anticipation time (ms) of peak height for fingers Mean finger velocity (m/s) and mean acceleration (m/s²) MIDI data	Peak finger heights preceding keystrokes and key velocity increased as tempo increased
Goebl and Palmer (2013)	N = 12 At least 10 years of piano instruction	Cross-sectional study	Right fingers, hand, and wrist	3D motion capture system (250 Hz, 25 markers) Digital piano	A melody	Ten tempi: 7.0, 8.4, 9.6, 10.7, 11.7, 12.3, 13.3, 14.1, 15.0, and 16.0 tones per second	Joint angle trajectories for all adjacent finger phalanges, the hand, and the wrist	Each finger joint did not change its relative contributions to the fingertip movements across tempi
Turner et al. (2022)	N = 3 (2 professionals and 1 intermediate-level with 11 years of piano study)	Case Study	Trunk and right hand	3D motion capture system (200 Hz, 6 markers) Grand piano	An excerpt	Three tempi: 6, 8, and 10 notes/s (N/s)	Starting position Initiation intervals Trunk range of motion (ROM) Right hand velocity (m/s) Right limb coordination	As tempi increased, trunk and right-hand medio-lateral shifts were more synchronized
Furuya et al. (2012)	N = 18 (5 males, 13 females) More than 15 years of classical music	Cross-sectional study	Right upper limb (finger, wrist, elbow, and shoulder joints) Six right upper-limb muscles: anterior and posterior deltoid (AD and PD), triceps brachii, biceps brachii, flexor digitorum superficialis (FDS), and extensor digitorum communis (EDC)	2D motion capture system (300 Hz, sagittal plane) 6 EMG sensors Upright piano	Repetitive chord keystrokes	Four loudness levels: piano (p), mezzo-piano (mp), mezzo-forte (mf), and forte (f) Four tempi: 180, 240, 300, and 360 bpm	Peak angular velocity (rad/s) of the shoulder, elbow, wrist, and finger Mean muscle activation (%MVC)	Interaction effect of loudness and tempo on peak angular velocity for all joints except for the elbow, and on muscular activity for all muscles
Furuya et al. (2010)	N = 7 (3 males, 4 females) More than 15 years of classical piano training	Cross-sectional study	Right upper limb (upper arm, forearm, hand, and finger)	2D motion capture system (150 Hz, sagittal plane) Upright piano	Repetitive isolated keystrokes	Two loudness levels: piano (p) and forte (f) Pressed and struck touches	Mean of joint angles (rad), fingertip and key’s (mm) vertical position and velocities (rad/s, mm/s) Mean peak angular velocities (rad/s) Net joint acceleration (rad/s²)	The pressed and struck touches effectively took advantage of the distal-to-proximal and proximal-to-distal inter-segmental dynamics, respectively
Verdugo et al. (2020b)	N = 9 (7 males, 2 females) Holding or currently pursuing a doctoral degree in piano performance	Cross-sectional study	Pelvis, thorax, right upper limb, and left lower limb	3D motion capture system (150 Hz, 68 markers) Grand piano	Repetitive isolated keystrokes	Eight combinations of these three variables: 1) Body implication (use of trunk and upper-limb motion) or use of only upper-limb motion 2) Touch (pressed or struck) 3) Articulation (staccato or tenuto)	Upper-limb linear velocities (m/s) Joint angular contribution (m/s)	All upper-limb segments presented forward velocities during the key descent regardless of touch and articulation. Pelvic anterior rotation effectively contributed to creating forward linear velocities at the upper limb
Degrave et al. (2020)	N = 12 (10 males, 2 females) Professional pianists	Cross-sectional study	Nine right upper-limb muscles (flexor digitorum superficialis, extensor digitorum communis, biceps brachii, triceps brachii, anterior deltoid, middle deltoid, great pectoral, upper trapezius, serratus anterior)	9 EMG sensors (2100 Hz) Grand piano	Repetitive isolated keystrokes	Four possible combinations of: two types of touch (pressed or struck) and articulation (staccato or tenuto)	Time histories of mean muscle activation (%MVC)	Compared to tenuto articulation, staccato articulation induced a higher muscle activity on shoulder muscles
Goubault et al. (2021)	N = 49 (30 males, 19 females) All participants had at least a university (or equivalent) degree or were enrolled in undergraduate or graduate studies in piano performance	Cross-sectional study	Forearm muscles (finger and wrist flexor and extensor muscles)	49 monopolar EMG electrodes (2048 Hz) Grand piano	Digital and chord excerpts	Digital and chord excerpt Fatigue	Rate of perceived exertion EMG median frequency	Finger/wrist extensor muscles showed greater signs of fatigue than finger/wrist flexor muscles Pianists showed extremely different levels of endurance. Muscle fatigue negatively affected key velocity and note-accuracy
Thio-Pera et al. (2022)	N = 8 (6 males, 2 females) Professional piano players	Cross-sectional study	Forearm muscles (hand and wrist muscles)	32 monopolar EMG electrodes (2048 Hz) Upright piano	Octaves and three excerpts	Octaves were played in four conditions and two tempi: Four conditions: spezzate, loading the forearm, the wrist, and the fingers segments, each at a time Two tempi: self-paced speed and as fast as possible	EMG average map	Depending on playing contexts, professional pianists consistently load specific finger/wrist muscles, whether performing octaves (extensor muscles) or classical excerpts (flexor muscles)

*As this study compared novices and experts, the main outcome has been reported according to experts only.

Table 3.B.

Summary of studies included in the review that investigated the impact of changes in abstract parameters on pianists’ kinematics.

Author & date	Participants	Study Design	Body segment studied	Measurements	Experimental task	Independent variables	Dependent variables	Main outcome
Castellano et al. (2008)	N = 1 female Professional concert pianist	Case Study	Overall body movement	Video cameras (25 Hz) Grand piano	An excerpt	Five different expressive modes: personal, sad, allegro, serene, and over-expressive	Quantity of motion of the upper body and the velocity of head movements	Velocity of head movements was influenced by expressive modes
Shoda and Adachi (2012)	N = 1 male Professional pianist (who studied the piano for 20 years)	Case Study	Overall body movement	Video cameras (29.97 Hz) Grand piano	Two excerpts	Three levels of expression: deadpan, artistic, and exaggerated	Mean value of the movement amplitude (rad)	Pianist's range of movement in the artistic condition differed from the other two for a fast, energetic piece, whereas it only differed from the deadpan for a slow, romantic piece
Thompson and Luck (2012)	N = 8 (3 males, 5 females) Between 10 and 20 years of piano-playing experience	Cross-sectional study	Upper body (hip, torso, neck, head, shoulders, elbows, wrists, middle fingers)	3D motion capture system (120 Hz, 26 markers) Digital piano	An excerpt	Four expressive intentions: normal, deadpan, exaggerated, immobile	Duration of the performance (sec) Cumulative distance traveled by markers (mm)	Markers at the head and at the shoulder exhibited more movement per measure, compared to the fingers, wrists, and lower back, for the normal and the exaggerated conditions
Massie-Laberge et al. (2019)	N = 10 (4 males, 6 females) The participants were all graduate or post-graduate students	Cross-sectional study	Hands, elbows, shoulders, torso, head, and pelvis	3D motion capture system (240 Hz, 49 markers) Digital piano	Three excerpts	Four expressive conditions: normal, deadpan, exaggerated, and immobile	Overall duration of the performances Cumulative distance traveled by markers (mm) Head movement recurrence	Head movements are important for communicating different expressive playing conditions and structural features (ascending movements, crescendo dynamics, and at the beginning of the melodic theme and its repetition)

Table 3.C.

Summary of the study included in the review that investigated the impact of changes in both performance and abstract parameters on pianists’ kinematics.

Author and date	Participants	Study design	Body segment studied	Measurements	Experimental task	Independent variables	Dependent variables	Main outcome
Wong et al. (2022)	N = 15 (3 males, 12 females) Completed Level 9 music training or majored in piano at university at the time or before the data collection	Cross-sectional study	Upper body (head, neck, trunk, sacrum, great trochanter, forearm)	3D motion capture system (100 Hz, 14 markers) Digital piano	A scale and an excerpt	Three expressive conditions: deadpan (no variation in dynamics or tempo), projected (played as they normally would in a performance), exaggerated (exaggerate all expressive features: tempo, dynamics)	Spine angles for each segment (deg)	Spine joint angles showed an average posture closer to neutral 1) in the deadpan playing compared to projected and exaggerated conditions, and 2) in playing an excerpt, compared to playing a scale

Impact of Performance Parameters on Pianists’ Kinematics

Technical Melodic Exercises

As tempo increased in technical melodic exercises, one study showed that mean movement vertical amplitude averaged across all fingers increased (from ∼17 mm with a tempo of 60 beats per minute (bpm) to ∼27 mm with a tempo of 250 bpm) and key velocity increased (from ∼43 MIDI units with a tempo of 60 bpm to ∼68 MIDI units with a tempo of 250 bpm; Dalla Bella & Palmer, 2011). Another study showed that finger joints did not change their relative contributions to the vertical fingertip movements across tempi; only the wrist vertical movement contributed slightly more to the fingertip motion at fast tempi than at slow tempi (from a wrist vertical efficiency score of ∼0/1 with a tempo of 7 tones per second to a wrist vertical efficiency score of ∼0.3/1 with a tempo of 15 tones per second; Goebl & Palmer, 2013).

Isolated Keystrokes

With an increase in tempo (from 180 bpm to 360 bpm and with a loudness of forte) during isolated keystrokes, Furuya et al. (2012) showed that peak angular velocities increased at the shoulder (from ∼0.3 to ∼0.37 rad/s) and the wrist (from ∼−1.8 to ∼−2.1 rad/s), but decreased at the elbow (from ∼−2.2 to ∼−1.3 rad/s). During alternate keystrokes, elbow velocity (pronation/supination) increased with an increase in tempo (from 70 bpm to 260 bpm; Furuya et al., 2011). As loudness increased (from piano to forte and with a tempo of 180 bpm), peak angular velocities increased at all joints (shoulder: from ∼0.14 to ∼0.3 rad/s, elbow: from ∼−0.7 to ∼−2.2 rad/s, wrist: from ∼−1.2 to ∼−1.8 rad/s, and finger joints: from ∼−1.8 to ∼−2.5 rad/s; Furuya et al., 2012). Verdugo et al. (2020b) found that shoulder-girdle joints contribution to finger upward velocity was greater during staccato articulation compared to tenuto articulation (absolute difference = 0.207 m/s, percentage difference = 206%). These authors also showed that pianists produced systematically forward upper-limb velocities during isolated keystroke attack and key holding/release phases, regardless of the choice of articulation and touch.

Repertoire Excerpts and Mixed Tasks

Wong et al. (2022) found that spine joint angles showed an average posture closer to neutral while playing an excerpt (head tilt of 3.2 ± 8.3° in projected playing), compared to playing a scale (head tilt of −4 ± 8.9°). One study showed that trunk and right-hand movement were more synchronized at faster tempi when playing an excerpt (Turner et al., 2022). Moreover, when averaging between the three musical sections, the shortest pianist (1.65 m) had the greatest trunk range of motion (276 mm), and the tallest pianist (1.90 m) had the smallest trunk range of motion (101 mm).

Impact of Abstract Parameters on Pianists’ Kinematics

Three out of five studies found that playing conditions with a higher level of expressiveness (e.g., deadpan compared to normal, normal compared to exaggerated) resulted in more head and proximal movements compared to deadpan condition (Castellano et al., 2008; Massie-Laberge et al., 2019; Thompson & Luck, 2012). For example, Thompson and Luck (2012) found that the distance traveled by the right and the left shoulder was between ∼200 and ∼300 mm per measure for the exaggerated conditions, and between ∼10 and ∼50 mm per measure in the deadpan condition. Similarly, Shoda and Adachi (2012) found that a pianist increased upper body movements in the artistic and exaggerated conditions compared to the deadpan condition. In addition, one study found that spine joint angles showed an average posture closer to neutral in the deadpan playing (craniovertebral angle of 43.5 ± 7.6°), compared to the other two conditions (craniovertebral angle of 38.3 ± 7.6° and 37.9 ± 9° for the projected and exaggerated conditions, respectively; Wong et al., 2022).

Impact of Performance Parameters on Pianists’ Muscle Activity

Studies investigating muscle activity focused on isolated keystrokes and musical excerpts.

Isolated Keystrokes

Both the activation level of six muscles (anterior and posterior deltoids, biceps brachii, triceps brachii, flexor digitorum superficialis, and extensor digitorum communis) and the co-activation index between the anterior-posterior deltoid, biceps-triceps brachii, and flexor digitorum superficialis-extensor digitorum communis muscle pairs increased at a tempo of 5 keystrokes per second or higher (Furuya et al., 2012). As loudness increased (from p to mp, mp to mf, and mf to f), the activation level of the above-mentioned muscles and their co-activation index increased (Furuya et al., 2012). The activation level particularly increased for distal muscles (the flexor digitorum superficialis, and the extensor digitorum communis increased their muscle activity from ∼4–5% maximum voluntary contraction (MVC) at slow tempi, to ∼10% MVC at faster tempi). During and after key descent and release, staccato articulation showed a higher activity in the shoulder muscles, compared to tenuto articulation (upper trapezius +2.1% MVC, anterior deltoid +2.5% MVC, great pectoralis +3.8% MVC; Degrave et al., 2020). One study showed that professional pianists activated more finger/wrist extensor muscles than finger/wrist flexor muscles when performing octaves (Thio-Pera et al., 2022).

Musical Excerpts

Thio-Pera et al. (2022) showed that professional pianists activated more finger/wrist flexor muscles than finger/wrist extensor muscles when performing repertoire excerpts compared to octaves. Constant repetition of a digital exercise and a chord musical excerpt, both performed loud and fast, showed higher levels of muscle fatigue at finger/wrist extensor muscles (the EMG median frequency decreased between 10 and 20 Hz at task termination) compared to the respective flexors (the EMG median frequency decreased between 2 and 10 Hz), and pianists showed different levels of endurance in their time-to-task termination (from around 2 min to 12 min, which was the maximum time allowed for the task; Goubault et al., 2021).

Ontological and Methodological Choices

Ontological choices (focus of studies, gestural variable investigated) and methodological choices (kinematic and EMG data collection tools, experimental tasks, musical instrument used) are reported in Figure 2.

Figure 2.

Overview of the included studies. Distribution of A) Focus of studies, B) Gestural variable investigated, C) Kinematic and EMG data collection tools, D) Experimental tasks, and E) Musical instrument used. Definition of experimental tasks: isolated keystrokes (isolated notes, octaves, or alternating keystrokes), technical melodic exercise (scale or arpeggio type of tasks), repertoire excerpts (actual excerpts from the repertoire), mixed tasks (studies using both repertoire and other types of tasks). Other: analysis of head movements and other body segments using 2D videos of performances.

Regarding kinematic analysis, the gestural variables investigated were different between studies focusing on the modification of performance parameters and studies focusing on abstract parameters. Studies focusing on the modification of performance parameters assessed either i) finger and/or wrist linear velocities (Dalla Bella & Palmer, 2011; Turner et al., 2022), joint angles (Goebl & Palmer, 2013), and movement repeatability (Sforza et al., 2003); or ii) right upper limb (Furuya et al., 2010, 2011, 2012) or upper body (Verdugo et al., 2020b) linear and joint kinematics (e.g., joint angular velocities, segmental linear velocities, etc.). Studies focusing on abstract parameters measured either i) the quantity of motion, (i.e., an approximation of the amount of detected movement, based on Silhouette Motion Images, which represent all variations of a simplified white-body silhouette, obtained using background subtraction; Castellano et al., 2008), and the cumulative distance traveled by markers (Massie-Laberge et al., 2019; Thompson & Luck, 2012); or ii) postural angles of the spine (Shoda & Adachi, 2012; Wong et al., 2022). Regarding EMG analysis, three studies calculated mean muscle activation over the entire trial (Furuya et al., 2011, 2012; Thio-Pera et al., 2022), one study calculated time series muscle activation (Degrave et al., 2020), and one study calculated the EMG median frequency to assess the myoelectric manifestation of muscle fatigue (Goubault et al., 2021).

Risk Factors of Playing-Related Musculoskeletal Disorders

A complementary objective was to address how music expression affects pianists’ exposure to risk factors of PRMDs. Risk factors of PRMDs associated with music expression could only be extracted from the studies focusing on performance parameters. The main results of four of these studies were directly or indirectly linked to an increase in muscle load (Dalla Bella & Palmer 2011; Degrave et al., 2020; Furuya et al., 2012; Thio-Pera et al., 2022), while one study showed an increase in muscle fatigue (Goubault et al., 2021). Among these five studies, the following four playing factors were identified as potential risk factors of PRMDs: playing loud, playing fast, staccato articulation, and large handspan chords (Table 4). The type of task, playing factor, biomechanical impact, and PRMDs risk factor addressed by these studies are summarized in Table 4.

Table 4.

Type of task, playing factor, biomechanical impact, and PRMDs risk factor of included studies.

Author and date	Type of task	Playing factor	Biomechanical impact	PRMDs risk factor
Dalla Bella and Palmer (2011)	Melodic exercise	Playing fast	Increase in fingertip height*	Increase in muscle load
Degrave et al. (2020)	Isolated keystrokes	Staccato articulation	Increase in shoulder muscle activation	Increase in muscle load
Furuya et al. (2012)	Isolated keystrokes	Playing loud and fast	Increase in muscle activation of six upper limb muscles (anterior and posterior deltoids, biceps brachii, triceps brachii, flexor digitorum superficialis, and extensor digitorum communis)	Increase in muscle load
Goubault et al. (2021)	Repetitive melodic and chord tasks	Playing loud and fast	Increase in finger/wrist extrinsic muscle fatigue	Muscle fatigue
Thio-Pera et al. (2022)	Consecutive octaves	Large handspan chords	Increase in finger/wrist extensor muscle activity	Increase in muscle load

*At the hand level, an increase in pianists’ fingertip height during melodic excerpts is primarily driven by the metacarpophalangeal joint (extension movement) (Goebl & Palmer, 2013). Therefore, the increase in fingertip height with faster tempi reported by Dalla Bella and Palmer (2011) points to an increase in both metacarpophalangeal joint extension and fingers’ extrinsic extensor muscles activity, a muscle group more prone to developing fatigue (Goubault et al., 2021).

Discussion

This systematic review investigated how both performance and abstract parameters of music expression impact pianists’ gestures. It also addressed ontological and methodological differences of the included studies and the impact of music expression-related parameters on exposure to risk factors of PRMDs. Sixteen studies were included. Eleven studies focused on the modification of performance parameters (i.e., sound intensity, tempo, articulation), four studies focused on the modification of abstract parameters (structural and/or semantic), and one study focused on both. Performance and abstract parameters impacted pianists’ kinematics and muscle activity, the specific effects being dependent on the type of task (i.e., isolated tones, digital task, chord task) and the gestural variable investigated by studies. Important differences in ontological (performance or abstract parameters studied, gestural variable investigated) and methodological choices (experimental task and instrument used, data acquisition and processing procedures) prevent the establishment of a thorough dialogue between studies focusing on performance and abstract parameters. Risk factors of PRMDs associated with music expression parameters included playing loud, playing fast, staccato articulation, and playing large handspan chords.

Impact of Music Expression on Pianists’ Gestures

Loudness and Articulation

In isolated keystrokes, loudness and articulation had clear and consistent effects across studies. An increase in loudness led to greater angular velocities and muscle activity at both distal and proximal joints/segments of the right upper limb (Furuya et al., 2012). These results are consistent with piano sound-production mechanics, as loudness is closely related to the key attack velocity (e.g., Dannenberg, 2006). To increase the targeted key velocity of louder tones, pianists increased velocity and muscle activity at upper-limb joints (shoulder, elbow, wrist). Moreover, this pattern appears consistent across skill levels, as novices and amateurs showed greater levels of muscle activation during repetitive octaves at wrist extrinsic extensor and flexor muscles as loudness increased (from pp to mf, and mf to ff; Oikawa et al., 2011). Articulation had a similar effect, but in relation to the release motion. Staccato articulation (in opposition to tenuto articulation) increased upper-limb upward/forward velocities (Verdugo et al., 2022a, 2022b) and shoulder muscle activity (Degrave et al., 2020) during and after the key descent in the context of isolated keystrokes, as the shoulder-girdle joints were the primary mover of the rapid lifting motion of the arm and hand after the attack (Verdugo et al., 2022a, 2022b). In brief, the production of faster key attack (louder tones) and key release (staccato tones) movements demands an increased velocity and muscle activity at the joints responsible for those faster endpoint movements. These studies have been conducted on simple performance tasks (repetitive isolated keystrokes). However, in more complex musical contexts, it seems possible to hypothesize that the same relation may prevail (i.e., increase of joint velocity and muscle activity allowing faster key attack and release movements).

Tempo

In the case of technical melodic exercises, an increase in tempo led to an increase in the height of the fingers before the keystroke (Dalla Bella & Palmer, 2011). The most important finger joint to produce the fingertip vertical movement during this type of melodic exercises was the metacarpophalangeal joint, and its contribution remained stable across different tempi (Goebl & Palmer, 2013). These results point to two ideas. First, faster tempi impose spatiotemporal constraints that led to an increased distance between the fingertip and the key before the attack to produce the targeted key attack velocity. Second, this increased height of the fingertip might increase the extension motion of the metacarpophalangeal joint. Furuya et al. (2011) showed that during fast alternate keystrokes (i.e., tremolo), expert pianists used hand pronation/supination (a degree of freedom at the elbow) to reduce metacarpophalangeal muscle load. As pianists can use pronation/supination not only during alternate keystrokes but also during a wide range of melodic passages (scales, arpeggios, etc.), the increased finger height at faster tempi reported by Dalla Bella and Palmer (2011) could be achieved by multi-joint hand/forearm movements rather than by isolated metacarpophalangeal joint movements. Dalla Bella and Palmer (2011) did not address interactions of loudness and tempo in their study. However, as louder sounds entail faster joint and key velocities, the required finger height needed to play at a certain tempo might also be affected by the targeted sound intensity. This must nevertheless be confirmed by future research.

During repetitive isolated keystrokes, varying the tempo produced distinct effects on peak angular velocities at different joints. Despite the interactions between loudness and tempo reported in Furuya et al. (2012), these authors observed that an increase in tempo resulted in faster peak flexion/extension velocities at the shoulder and wrist and slower peak velocities at the elbow. Similarly, two recent studies showed a reduction of elbow flexion/extension range of motion while playing repetitive chords at faster tempi (Turner et al., 2023; Wang et al., 2023). If elbow extension was the main contributor of the fingertip downward attack velocity during slow isolated keystrokes (Verdugo et al., 2020b), the results of the above-mentioned studies show that the leading role of the elbow to produce the attack downward velocity of the fingertip decreases while tempo increases. However, this tempo-dependent change of inter-joint coordination did not imply a reduction of elbow muscle activity, as faster tempi produced greater mean muscle activity at proximal and distal joints of the upper limb and higher co-contraction levels of elbow and finger muscles (Furuya et al., 2012).

Type of Task

Thio-Pera et al. (2022) found different task-dependent finger/wrist muscle loads regardless of tempo, where consecutive octaves induced greater activations at extrinsic extensors while other types of excerpts (melodic “finger” passage, slow-loud chord passage) induced greater activations at extrinsic flexors. To play consecutive octaves, pianists constantly hold the hand in a fixed position characterized by finger extension/abduction (which is coherent with the increased activity of extrinsic extensors reported by the authors). This is not the case in the other excerpts used in Thio-Pera et al. (2022), as distal joint posture and movements can be adapted at each keystroke or group of keystrokes. Chong et al. (2015) also found that muscle activation in hand extrinsic muscles was dependent on the configuration of the notes imposed in the score. However, Goubault et al. (2021) found higher levels of fatigue at finger/wrist extrinsic extensor muscles (compared to flexors) regardless of the type of task. Extensor muscles showed higher signs of fatigue during both a chord passage (involving octaves) and a melodic “finger” passage played repetitively in cycles. These results suggest that even though muscle load is dependent on note configuration (i.e., task-dependent muscle load), muscle fatigue might greatly affect specific muscles due to their intrinsic characteristics or the duration of the repetitive activations.

Abstract Parameters and Gestural Functions

Regarding abstract parameters, Thompson and Luck (2012) and Massie-Laberge et al. (2019) found that playing conditions with a higher level of expressiveness (e.g., deadpan compared to normal, normal compared to exaggerated) resulted in more movement at markers placed on the head and proximal segments. These findings underline the role of performers’ whole-body movements as a tool to encode or embody the expressive content of music while playing (Krumhansl, 2002; Leman & Maes, 2015). These gestures have been associated with sound-facilitating gestures (or ancillary gestures) by the cited studies and the related literature (Massie-Laberge et al., 2019; Thompson & Luck, 2012; Wanderley et al., 2005). However, delimitation of what body movements are labeled as sound-facilitating and as sound-producing is not clearly addressed. Typically, in this literature, sound-facilitating gestures involve the trunk and the head, while sound-producing gestures involve distal segments close to the performer-instrument interface (forearm, hand, fingers; Jensenius et al., 2010). Thus, these two types of gestures are usually understood as distinct gestures and have been studied separately in the literature. However, Verdugo et al. (2020b) showed that pianists’ pelvis and thorax movements can play a role in the control of performance parameters related to articulation and loudness. Moreover, other recent studies have also highlighted the role of trunk motion in pianists’ sound-production strategies (Turner et al., 2022, 2023; Verdugo et al., 2022a, 2022b). In the case of piano performance, it therefore seems clear that sound-producing and sound-facilitating gestures are not distinct gestures but are rather gestural functions embedded in the same gestural space incorporating the entire kinematic chain (pelvis, thorax, upper limbs, and potentially lower limbs). To enhance dialogue and coherence between the literature addressing the impact of performance and abstract parameters on pianists’ gestures, we recommend a more systematic use of the concepts of sound-producing and sound-facilitating gestural functions, rather than sound-producing and sound-facilitating gestures.

Ontological and Methodological Differences of the Targeted Literature

One of the main ontological differences of the studies in this review relates to the type of music expression-related parameters investigated. The 11 studies addressing performance parameters focused on “score-imposed” variations of performance parameters. For example, playing the same task at the piano with different imposed tempi (e.g., Sforza et al., 2003), different articulation (e.g., Degrave et al., 2020), and different loudness levels (e.g., Furuya et al., 2012). In addition, these studies, generally from the field of biomechanics and motor control, did not consider abstract parameters (music structure parameters and extra-musical or semantic ideas) from research questions. On the contrary, studies from the music research domain addressing abstract parameters investigated the effect of music expression on pianists’ gestures in relation to the performers’ interpretation of the score. By using notions such as expressive intentions and experimental conditions based on different levels of expression (deadpan, normal, exaggerated), these studies addressed music expression from the performers’ point of view. This different focus on expression (score-imposed features versus performers’ interpretation features) is a key difference that hampers the establishment of a connection between research on musicians’ gestures from biomechanics and motor control, on one hand, and from music research and empirical musicology, on the other hand. Biomechanical studies addressing the impact of pianists’ personal management of abstract parameters would be necessary to strengthen the link between literature from biomechanics and music research, and enable a deeper understanding of the impact of music expression (linked to the creative work not only of the composer but also of the performer) on pianists’ gestures.

A key methodological difference in studies focusing on pianists’ kinematics relates to the choice of the kinematic variable investigated. It is noteworthy that most studies (9 out of 13 focusing on pianist's kinematics) used 3D motion capture systems (Figure 2). Despite one exception (Wong et al., 2022), studies on pianists’ sound-facilitating gestures analyzed segmental kinematics based on data of markers placed on the body (e.g., quantity of motion, distance traveled by markers), with little or no attention to joint kinematics or more thorough methods for the computation of segmental kinematics (using for example segment endpoint or center of mass). Therefore, studies in music research do not usually take advantage of methods from biomechanics to analyze and address the interdependent nature of movements of multi-body systems such as the human body. An interdisciplinary approach mixing methods from empirical musicology and biomechanics would, first, facilitate a better understanding of the interrelated nature of sound-producing and sound-facilitating functions in the context of multi-joint movements of pianists. Second, as abstract parameters might not only affect pianists’ kinematics (Turner et al., 2024) but also muscle activity (Verdugo et al., 2020a, 2022a, 2022b), this interdisciplinary approach would allow the assessment of the impact of pianists’ expressive intentions on their muscle load, with implications for both embodied cognition and injury prevention research. As an example, in a recent case study on two participants, Mailly et al. (2024) showed that pianists can embody their expressive intentions in different musical contexts through upper-body muscle activity, including proximal (upper trapezius, external oblique) and distal muscles, such as flexor digitorum superficialis and extensor digitorum communis.

Another important methodological difference was associated with the instrument used. Six studies used a grand piano, seven studies used a digital piano, and three studies used an upright piano (Figure 2). Acquiring pianists’ movements using optoelectronic cameras and passive markers remains a challenge due to marker occlusions caused by the piano itself. Despite this obstacle, several studies have been conducted with grand piano using these motion capture systems (Turner et al., 2021, 2022, 2023; Verdugo et al., 2020b, 2022a, 2022b). The grand piano is the actual instrument where pianists usually perform and practice, and its specific key action mechanism influences pianists’ touch and sound control (Traube et al., 2017). Therefore, the use of digital instruments to facilitate motion capture procedures may change how pianists manipulate both performance and abstract music expression-related parameters, and consequently, influence research results. This methodological limitation was highlighted in a previous literature review on piano touch (MacRitchie, 2015). However, it remains relevant for the current state of the literature.

Risk Factors of PRMDs and Considerations for Injury Prevention

The following four playing factors were associated with increased muscle load or muscle fatigue: playing loud, playing fast, staccato articulation, and playing large handspan chords. These results might be applicable to other instruments, as they support the results of a violin study that showed greater muscle activation in the forearm muscles caused by playing loud and playing fast (Mann et al., 2023). The included studies in this review did not address the relationship between the reported increase of muscle activation or fatigue and PRMDs history in the participants recruited. Nevertheless, increases in muscle load and muscle fatigue are typically considered prominent risk factors of musculoskeletal disorders in music (Ling et al., 2018), sports, and daily life activities (Côté, 2014). In addition, questionnaire-based studies have shown that large handspan chords (such as octaves; Allsop, 2007; Sakai, 2002; Shields & Dockrell, 2000), and playing chords loudly (Furuya et al., 2006) were associated with the development of PRMDs. In this direction, future studies should investigate gestural strategies that can reduce exposure to the four risk factors listed above. Moreover, considering that these playing factors are extremely common in the piano repertoire, quantification of practice load based on these playing factors might also help reduce occurrence of pianists’ PRMDs.

Recent studies in sport science literature focusing on injury prevention (e.g., running, tennis) have centered on the concept of “training load” (Coutts et al., 2011; Drew & Finch, 2016; Gabbett, 2016). Training load is quantified by multiplying the duration of each training session by its intensity, usually assessed using the perception of effort on a scale of 1 to 10 (Foster et al., 1996). Subsequently, the evolution of the training load is assessed on a weekly basis. Literature has shown that when training load increases by more than 15% above the previous week training load, prevalence of injury increases by 21% to 49% (Gabbett, 2016). Therefore, it has been concluded that to minimize the risk of injuries in athletes, it is advisable to maintain training load weekly increases below 10%. While the concept of playing load has recently been studied in relation to music performance (McCrary et al., 2022), monitoring playing load based on the playing factors identified in this review as a prevention strategy of PRMDs has yet to be studied. Therefore, future studies could develop playing load models to both quantify playing load and its fluctuation and predict occurrence of PRMDs in pianists, especially during the preparation process of upcoming performances or music competitions.

Limitations

Limitations of this systematic review include search strategy and methodological differences between studies. First, limitations in database coverage may have resulted in some relevant studies being overlooked. However, the search strategy was performed in four databases, which improved coverage, and decreased the risk of making inappropriate conclusions (Ewald et al., 2022). Second, the included studies differed in their design (cross-sectional studies or case studies). The studies also varied in their focus, addressing different parameters (performance or abstract parameters, or both), types of tasks (i.e., isolated tones, digital task, chord task, musical excerpts), and gestural variables investigated (linear segmental kinematics, joint angles, muscle activity), which limited the comparison between studies. However, most of the included studies showed a moderate to high methodological quality assessment score, suggesting that the overall quality of the results was not compromised.

Conclusion

This systematic review addressed how both performance (timing, sound intensity, articulation) and abstract parameters (related to music structure and extra-musical or semantic content) impacted pianists’ kinematics and muscle activity, showing that the specific effects were dependent on the type of task and the gestural variable investigated by studies. Therefore, findings on the relationship between musical expression and pianists’ gestures are specific to the studied contexts and cannot be generalized to other musical settings or gestural variables. The present review also highlighted important methodological and ontological differences in the included studies. These differences prevent the establishment of both a more unified view on the impact of music expression parameters on pianists’ gestures and a thorough dialogue between studies focusing on performance and abstract parameters of music expression. Future research on music expression and pianists’ gestures should develop a deeper interdisciplinary approach to bridge the gap between studies from empirical music research and biomechanics. In this direction, empirical music research could make use of joint kinematics and muscle activity analysis used in biomechanics. Similarly, biomechanics and motor control studies could also integrate abstract parameters of musical expression in research questions and protocols. Finally, this systematic review also identified four playing factors associated with performance parameters of music expression as potential risk factors of PRMDs (playing loud, playing fast, staccato articulation, and large handspan chords). The proposed interdisciplinary approach for future research could help enhance knowledge on the impact of abstract parameters of music expression on pianists’ exposure to risk factors of PRMDs.

Footnotes

Acknowledgments

We would like to thank Denis Arvisais (librarian at University of Montreal) for his help in the creation and execution of the search strategy.

Action Editor

Eckart Altenmüller, Hochschule für Musik, Theater und Medien Hannover, Institut für Musikphysiologie und Musikermedizin

Peer Review

Two anonymous reviewers

Author Contributions

RM: Conceptualization, methodology, writing – original draft, writing – review and editing. CT: Methodology, writing – review and editing. EG: Methodology, writing – review and editing. FDM: Methodology, writing – review and editing. FV: Conceptualization, methodology, supervision, writing – original draft, writing – review and editing, funding acquisition.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical Approval

This research did not require ethics committee or IRB approval. This research did not involve the use of personal data, fieldwork, or experiments involving human or animal participants, or work with children, vulnerable individuals, or clinical populations.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Social Science and Humanities Research Council of Canada, (grant number Insight Development Grant 430-2021-00384).

ORCID iDs

Robin Mailly

Craig Turner

Felipe Verdugo

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analyzed during the current study.

References

Adams

L. M.

Turk

D. C.

(2018). Central sensitization and the biopsychosocial approach to understanding pain. Journal of Applied Biobehavioral Research, 23(2), e12125. https://doi.org/10.1111/jabr.12125

Allsop

(2007). Investigating the prevalence of playing-related musculoskeletal disorders in relation to piano players’ playing-techniques and practising strategies [Master’s thesis, University of Western Australia] UWA Repository. https://research-repository.uwa.edu.au/files/3236092/Allsop_Lili_2007.pdf.

Bigand

Parncutt

(1999). Perceiving musical tension in long chord sequences. Psychological Research, 62(4), 237–254. https://doi.org/10.1007/s004260050053

Castellano

Mortillaro

Camurri

Volpe

Scherer

(2008). Automated analysis of body movement in emotionally expressive piano performances. Music Perception: An Interdisciplinary Journal, 26(2), 103–119. https://doi.org/10.1525/mp.2008.26.2.103

Chong

H. J.

Kim

S. J.

Yoo

G. E.

(2015). Differential effects of type of keyboard playing task and tempo on surface EMG amplitudes of forearm muscles. Frontiers in Psychology, 6, https://doi.org/10.3389/fpsyg.2015.01277

Côté

J. N.

(2014). Adaptations to neck/shoulder fatigue and injuries. In Levin

M. F.

(Ed.), Progress in motor control (Vol. 826, pp. 205–228). Springer New York. https://doi.org/10.1007/978-1-4939-1338-1_13

Coutts

A. J.

Gomes

R. V.

Viveiros

Aoki

M. S.

(2011). Monitoring training loads in elite tennis. Revista Brasileira de Cineantropometria e Desempenho Humano, 12(3), 217–220. https://doi.org/10.5007/1980-0037.2010v12n3p217

Dahl

Bevilacqua

Bresin

Clayton

Leante

Poggi

Rasamimanana

(2010). Gesture in performance. In R. I. Godøy, & M. Leman (Eds.), Musical gestures (pp. 36–68). Routledge.

Dalla Bella

Palmer

(2011). Rate effects on timing, key velocity, and finger kinematics in piano performance. PLoS ONE, 6(6), e20518. https://doi.org/10.1371/journal.pone.0020518

10.

Dannenberg

R. B.

(2006). The interpretation of MIDI velocity. ICMC.

11.

Davidson

J. W.

(2007). Qualitative insights into the use of expressive body movement in solo piano performance: A case study approach. Psychology of Music, 35(3), 381–401. https://doi.org/10.1177/0305735607072652

12.

Degrave

Verdugo

Pelletier

Traube

Begon

(2020). Time history of upper-limb muscle activity during isolated piano keystrokes. Journal of Electromyography and Kinesiology, 54, 102459. https://doi.org/10.1016/j.jelekin.2020.102459

13.

Desmyttere

Hajizadeh

Bleau

Begon

(2018). Effect of foot orthosis design on lower limb joint kinematics and kinetics during walking in flexible pes planovalgus: A systematic review and meta-analysis. Clinical Biomechanics (Bristol, Avon), 59, 117–129. https://doi.org/10.1016/j.clinbiomech.2018.09.018

14.

Ding

(2024). Biomechanical and machine learning approaches to automating the identification of musical styles and emotions through human motion analysis. Molecular & Cellular Biomechanics, 21(2), 397. https://doi.org/10.62617/mcb.v21i2.397

15.

Downs

S. H.

Black

(1998). The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. Journal of Epidemiology and Community Health, 52(6), 377–384. https://doi.org/10.1136/jech.52.6.377

16.

Drake

Palmer

(1993). Accent structures in music performance. Music Perception, 10(3), 343–378. https://doi.org/10.2307/40285574

17.

Drew

M. K.

Finch

C. F.

(2016). The relationship between training load and injury, illness and soreness: A systematic and literature review. Sports Medicine, 46(6), 861–883. https://doi.org/10.1007/s40279-015-0459-8

18.

Ewald

Klerings

Wagner

Heise

T. L.

Stratil

J. M.

Lhachimi

S. K.

Hemkens

L. G.

Gartlehner

Armijo-Olivo

Nussbaumer-Streit

(2022). Searching two or more databases decreased the risk of missing relevant studies: A metaresearch study. Journal of Clinical Epidemiology, 149, 154–164. https://doi.org/10.1016/j.jclinepi.2022.05.022

19.

Foster

Daines

Hector

Snyder

A. C.

Welsh

(1996). Athletic performance in relation to training load. Wisconsin Medical Journal, 95(6), 370–374.

20.

Furuya

Altenmüller

(2013). Flexibility of movement organization in piano performance. Frontiers in Human Neuroscience, 7, https://doi.org/10.3389/fnhum.2013.00173

21.

Furuya

Altenmüller

Katayose

Kinoshita

(2010). Control of multi-joint arm movements for the manipulation of touch in keystroke by expert pianists. BMC Neuroscience, 11(1), 82. https://doi.org/10.1186/1471-2202-11-82

22.

Furuya

Aoki

Nakahara

Kinoshita

(2012). Individual differences in the biomechanical effect of loudness and tempo on upper-limb movements during repetitive piano keystrokes. Human Movement Science, 31(1), 26–39. https://doi.org/10.1016/j.humov.2011.01.002

23.

Furuya

Goda

Katayose

Miwa

Nagata

(2011). Distinct inter-joint coordination during fast alternate keystrokes in pianists with superior skill. Frontiers in Human Neuroscience, 5, https://doi.org/10.3389/fnhum.2011.00050

24.

Furuya

Kinoshita

(2007). Roles of proximal-to-distal sequential organization of the upper limb segments in striking the keys by expert pianists. Neuroscience Letters, 421(3), 264–269. https://doi.org/10.1016/j.neulet.2007.05.051

25.

Furuya

Nakahara

Aoki

Kinoshita

(2006). Prevalence and causal factors of playing-related musculoskeletal disorders of the upper extremity and trunk among Japanese pianists and piano students. Medical Problems of Performing Artists, 21(3), 112–117. https://doi.org/10.21091/mppa.2006.3023

26.

Gabbett

T. J.

(2016). The training—injury prevention paradox: Should athletes be training smarter and harder? British Journal of Sports Medicine, 50(5), 273–280. https://doi.org/10.1136/bjsports-2015-095788

27.

Gabrielsson

(1999). The performance of music. In Deutsch

(Ed.), The psychology of music (2nd ed., pp. 501–602). Academic Press.

28.

Goebl

(2017). Movement and touch in piano performance. In Müller

Wolf

S. I.

Brueggemann

G.-P.

Deng

McIntosh

Miller

Selbie

W. S.

(Eds.), Handbook of human motion (pp. 1–18). Springer International Publishing. https://doi.org/10.1007/978-3-319-30808-1_109-1

29.

Goebl

Palmer

(2013). Temporal control and hand movement efficiency in skilled music performance. PLoS ONE, 8(1), e50901. https://doi.org/10.1371/journal.pone.0050901

30.

Goubault

Verdugo

Pelletier

Traube

Begon

Dal Maso

(2021). Exhausting repetitive piano tasks lead to local forearm manifestation of muscle fatigue and negatively affect musical parameters. Scientific Reports, 11(1), 8117. https://doi.org/10.1038/s41598-021-87403-8

31.

Héroux

Fortier

M.-S.

(2015). Expérimentation d’une nouvelle méthodologie pour expliciter le processus de création d’une interprétation musicale. Les Cahiers de la Société Québécoise de Recherche en Musique, 15(1), 67–79. https://doi.org/10.7202/1033796ar

32.

Jensenius

A. R.

Wanderley

M. M.

Godøy

R. I.

Leman

(2010). Musical gestures: Concepts and methods in research. In R. I. Godøy, & M. Leman (Eds.), Musical gestures: Sound, movement, and meaning. Routledge.

33.

Juslin

P. N.

Västfjäll

(2008). Emotional responses to music: The need to consider underlying mechanisms. Behavioral and Brain Sciences, 31(5), 559–575. https://doi.org/10.1017/S0140525X08005293

34.

Kok

L. M.

Huisstede

B. M.

Voorn

V. M.

Schoones

J. W.

Nelissen

R. G.

(2016). The occurrence of musculoskeletal complaints among professional musicians: A systematic review. International Archives of Occupational and Environmental Health, 89(3), 373–396. https://doi.org/10.1007/s00420-015-1090-6

35.

Krumhansl

C. L.

(2002). Music: A link between cognition and emotion. Current Directions in Psychological Science, 11(2), 45–50. https://doi.org/10.1111/1467-8721.00165

36.

Leman

Maes

P.-J.

(2015). The role of embodiment in the perception of music. Empirical Musicology Review, 9(3–4), 236. https://doi.org/10.18061/emr.v9i3-4.4498

37.

Ling

C.-Y.

Loo

F.-C.

Hamedon

T. R.

(2018). Playing-related musculoskeletal disorders among classical piano students at tertiary institutions in Malaysia: Proportion and associated risk factors. Medical Problems of Performing Artists, 33(2), 82–89. https://doi.org/10.21091/mppa.2018.2013

38.

MacRitchie

(2015). The art and science behind piano touch: A review connecting multi-disciplinary literature. Musicae Scientiae, 19(2), 171–190. https://doi.org/10.1177/1029864915572813

39.

Mailly

Turner

Traube

Dal Maso

Verdugo

(2024). Embodiment of music expression through muscle activity in expert pianists: A case study. Journées d’Informatique Musicale, Marseille, France, May 2024. ⟨hal-04661288v1⟩.

40.

Mann

Paarup

H. M.

Søgaard

(2023). Effects of different violin playing techniques on workload in forearm and shoulder muscles. Applied Ergonomics, 110, 103999. https://doi.org/10.1016/j.apergo.2023.103999

41.

Massie-Laberge

Cossette

Wanderley

M. M.

(2019). Kinematic analysis of pianists’ expressive performances of romantic excerpts: Applications for enhanced pedagogical approaches. Frontiers in Psychology, 9, 2725. https://doi.org/10.3389/fpsyg.2018.02725

42.

McCrary

J. M.

Ascenso

Savvidou

Schraft

McAllister

Redding

Bastepe-Gray

Altenmüller

(2022). Load and fatigue monitoring in musicians using an online app: A pilot study. Frontiers in Psychology, 13, 1056892. https://doi.org/10.3389/fpsyg.2022.1056892

43.

Oikawa

Tsubota

Chikenji

Chin

Aoki

(2011). Wrist positioning and muscle activities in the wrist extensor and flexor during piano playing. Hong Kong Journal of Occupational Therapy, 21(1), 41–46. https://doi.org/10.1016/j.hkjot.2011.06.002

44.

Page

M. J.

McKenzie

J. E.

Bossuyt

P. M.

Boutron

Hoffmann

T. C.

Mulrow

C. D.

Shamseer

Tetzlaff

J. M.

Akl

E. A.

Brennan

S. E.

Chou

Glanville

Grimshaw

J. M.

Hróbjartsson

Lalu

M. M.

Loder

E. W.

Mayo-Wilson

McDonald

, … Moher

(2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, n71, https://doi.org/10.1136/bmj.n71

45.

Repp

B. H.

(1995). Acoustics, perception, and production of legato articulation on a digital piano. The Journal of the Acoustical Society of America, 97(6), 3862–3874. https://doi.org/10.1121/1.413065

46.

Repp

B. H.

(1998). A microcosm of musical expression. I. Quantitative analysis of pianists’ timing in the initial measures of Chopin’s Etude in E major. The Journal of the Acoustical Society of America, 104(2), 1085–1100. https://doi.org/10.1121/1.423325

47.

Rousseau

Barton

Garden

Baltzopoulos

(2021). Development of an injury prevention model for playing-related musculoskeletal disorders in orchestra musicians based on predisposing risk factors. International Journal of Industrial Ergonomics, 81, 103026. https://doi.org/10.1016/j.ergon.2020.103026

48.

Sakai

(2002). Hand pain attributed to overuse among professional pianists: A study of 200 cases. Medical Problems of Performing Artists, 17(4), 178–180. https://doi.org/10.21091/mppa.2002.4028

49.

Sforza

Macrì

Turci

Grassi

Ferrario

V. F.

(2003). Neuromuscular patterns of finger movements during piano playing. Definition of an experimental protocol. Italian Journal of Anatomy and Embryology, 108(4), 211–222.

50.

Shields

Dockrell

(2000). The prevalence of injuries among pianists in music schools in Ireland. Medical Problems of Performing Artists, 15(4), 155–160. https://doi.org/10.21091/mppa.2000.4030

51.

Shoda

Adachi

(2012). The role of a pianist’s affective and structural interpretations in his expressive body movement: A single case study. Music Perception, 29(3), 237–254. https://doi.org/10.1525/mp.2012.29.3.237

52.

Thio-Pera

De Carlo

Manzoni

D’Elia

Cerone

G. L.

Putame

Terzini

Gazzoni

Bignardi

Vieira

(2022). Are the forearm muscles excited equally in different, professional piano players? PLOS ONE, 17(3), e0265575. https://doi.org/10.1371/journal.pone.0265575

53.

Thompson

M. R.

Luck

(2012). Exploring relationships between pianists’ body movements, their expressive intentions, and structural elements of the music. Musicae Scientiae, 16(1), 19–40. https://doi.org/10.1177/1029864911423457

54.

Traube

Moulin

Verdugo

(2017). Controlling piano tone by varying the “weight” applied on the key. The Journal of the Acoustical Society of America, 141(5), 3874–3874. https://doi.org/10.1121/1.4988663

55.

Turner

Goubault

Maso

F. D.

Begon

Verdugo

(2023). The influence of proximal motor strategies on pianists’ upper-limb movement variability. Human Movement Science, 90, 103110. https://doi.org/10.1016/j.humov.2023.103110

56.

Turner

Mailly

Dal Maso

Verdugo

(2024). Impact of expressive intentions on upper-body kinematics in two expert pianists. Frontiers in Psychology, 15, 1504456. https://doi.org/10.3389/fpsyg.2024

57.

Turner

Visentin

Oye

Rathwell

Shan

(2022). An examination of trunk and right-hand coordination in piano performance: A case comparison of three pianists. Frontiers in Psychology, 13, 838554. https://doi.org/10.3389/fpsyg.2022.838554

58.

Turner

Visentin

Shan

(2021). Wrist internal loading and tempo-dependent, effort-reducing motor behaviour strategies for two elite pianists. Medical Problems of Performing Artists, 36(3), 141–149. https://doi.org/10.21091/mppa.2021.3017

59.

Verdugo

Begon

Gibet

Wanderley

M. M.

(2022a). Proximal-to-distal sequences of attack and release movements of expert pianists during pressed-staccato keystrokes. Journal of Motor Behavior, 54(3), 316–326. https://doi.org/10.1080/00222895.2021.1962237

60.

Verdugo

Ceglia

Frisson

Burton

Begon

Gibet

Wanderley

M. M.

(2022b). Feeling the effort of classical musicians—A pipeline from electromyography to smartphone vibration for live music performance. NIME 2022. NIME 2022, The University of Auckland, New Zealand. https://doi.org/10.21428/92fbeb44.3ce22588

61.

Verdugo

Kokubu

Wang

Wanderley

(2020a). MappEMG: Supporting musical expression with vibrotactile feedback by capturing gestural features through electromyography. In International Workshop on Haptic and Audio Interaction Design.

62.

Verdugo

Pelletier

Michaud

Traube

Begon

(2020b). Effects of trunk motion, touch, and articulation on upper-limb velocities and on joint contribution to endpoint velocities during the production of loud piano tones. Frontiers in Psychology, 11, 1159. https://doi.org/10.3389/fpsyg.2020.01159

63.

Vines

B. W.

Krumhansl

C. L.

Wanderley

M. M.

Levitin

D. J.

(2006). Cross-modal interactions in the perception of musical performance. Cognition, 101(1), 80–113. https://doi.org/10.1016/j.cognition.2005.09.003

64.

Wanderley

M. M.

Depalle

(2004). Gestural control of sound synthesis. Proceedings of the IEEE, 92(4), 632–644. https://doi.org/10.1109/JPROC.2004.825882

65.

Wanderley

M. M.

Vines

B. W.

Middleton

McKay

Hatch

(2005). The musical significance of clarinetists’ ancillary gestures: An exploration of the field. Journal of New Music Research, 34(1), 97–113. https://doi.org/10.1080/09298210500124208

66.

Wang

Nonaka

Abdulali

Iida

(2023). Coordinating upper limbs for octave playing on the piano via neuro-musculoskeletal modeling. Bioinspiration & Biomimetics, 18(6), 066009. https://doi.org/10.1088/1748-3190/acfa51

67.

Wong

G. K.

Comeau

Russell

Huta

(2022). Postural variability in piano performance. Music & Science, 5, 205920432211378. https://doi.org/10.1177/20592043221137887

Impact of Music Expression-Related Parameters on Pianists’ Kinematics and Muscle Activity: A Systematic Review

Abstract

Keywords

Introduction

Methods

Search Strategy

Eligibility Criteria

Study Selection

Quality Assessment

Data Extraction

Main Results

Risk Factors of PRMDs

Inter-Rater Agreement

Results

Search Results

Quality Assessment

Studies’ Characteristics

Impact of Performance Parameters on Pianists’ Kinematics

Technical Melodic Exercises

Isolated Keystrokes

Repertoire Excerpts and Mixed Tasks

Impact of Abstract Parameters on Pianists’ Kinematics

Impact of Performance Parameters on Pianists’ Muscle Activity

Isolated Keystrokes

Musical Excerpts

Ontological and Methodological Choices

Risk Factors of Playing-Related Musculoskeletal Disorders

Discussion

Impact of Music Expression on Pianists’ Gestures

Loudness and Articulation

Tempo

Type of Task

Abstract Parameters and Gestural Functions

Ontological and Methodological Differences of the Targeted Literature

Risk Factors of PRMDs and Considerations for Injury Prevention

Limitations

Conclusion

Footnotes

Acknowledgments

Action Editor

Peer Review

Author Contributions

Declaration of Conflicting Interests

Ethical Approval

Funding

ORCID iDs

Data Availability Statement

References