The Biomechanics of Character Types in Javanese Dance

Luis Hernandez-Barraza, PhD, Chen-Hua Yeow, PhD, and Miguel Escobar Varela, PhD

Abstract iomechanics is a technical the tool kit of biomechanics is perti- The aim of this study was to describe and discipline that explains body nent for generating greater knowledge compare the different character sub-types motion through physical con- of this fundamental human form of of Javanese dance from a biomechanical ceptsB including angles, moments, expression. Through the present study, perspective. One professional dancer was energy consumption, power, and we aim to contribute an analysis of Ja- asked to repeat a basic motion (standing- ground reaction forces,1-4 and its ap- vanese dance to the growing literature up) according to the movement rules plications encompass fields such as on the biomechanics of dance. that pertain to six character sub-types sports, rehabilitation, and the analysis Javanese dance has a long history; (humble-refined, proud-refined, humble- of daily activities. Recently, potential it is perhaps rooted in the shamanistic strong, proud-strong, monkey, and bird). A motion capture system consisting of applications have expanded to include dances of the pre-Hindu era, which seven infrared cameras with a sample the analysis of dance and other forms were further shaped by interaction rate of 100 Hz and two force plates with of artistic expression. These studies with Indian-influenced dance and a sample rate of 1,000 Hz were used to are usually aimed at improving teach- music during the Hindu-Buddhist capture kinematics and kinetics. There ing, skill enhancement, and injury period.12 However, classical dance were significant differences in the bio- prevention.3 developed more actively from the 18th mechanical values we calculated for each Although biomechanics has de- Century onward, partially due to the character sub-type: range of motion, an- monstrably increased dance knowl- patronage of the courts of east and gular velocity, and ground reaction forces. edge, only a handful of dance genres central Java.13 Dance is very common The refined sub-types (humble-refined have been analyzed from a biome- today in various settings, and one of and proud-refined) showed the lowest val- chanical perspective; most commonly the best known forms is Sendratari ues at the knee joint for range of motion 2,5-11 and the lowest ankle, shoulder, and wrist ballet, Irish dance, and flamenco. Ramayana (the dance we used for this angular velocities. This result suggests Hence, very little is known about the study), a form that developed in the that low values in these measurements great diversity of dances around the 1960s on the foundation of classical are related to the smooth movements of world from a biomechanical perspec- Javanese dance. refined sub-types. These measurements tive, even though dance constitutes In all forms of classical Javanese help describe and contrast the motion one of the oldest and most widespread dance, character types are important. patterns of Javanese dance, contributing forms of cultural expression and ex- Sometimes the audience might be both to the scientific analysis of Javanese hibits many differences across cultures unfamiliar with the specific narrative dance and the application of biomechan- and geographic locations. We believe being depicted, but they are still able ics to the study of dance more generally. to infer certain things about the story by paying attention to the character Luis Hernandez-Barraza, PhD, and Chen-Hua Yeow, PhD, Department of types presented. Character types are Biomedical Engineering, National University of Singapore and Advanced Robotics Center, National University of Singapore, Singapore. Miguel Escobar Varela, divided into three major groups: putri PhD, Department of English Language and Literature, National University of (female), alusan (masculine-refined), Singapore, Singapore. and gagahan (masculine-strong). There are many sub-divisions of these Correspondence: Miguel Escobar Varela, PhD, Department of English Language major types, depending on the specific and Literature, National University of Singapore, Block AS5, 7 Arts Link Singapore kind of dance depicted. The character 117570, Singapore; [email protected]. types are identified by their makeup

Copyright © 2019 J. Michael Ryan Publishing, Inc. 104 https://doi.org/10.12678/1089-313X.23.3.104 Journal of Dance Medicine & Science • Volume 23, Number 3, 2019 105 and costume, but also by the kind of movement that is common for each character type. For example, refined sub-types are said to have relatively smooth and less angular motions. Although qualitative descriptions and notations are available14 the biomechanical analysis of character sub-types has not previously been attempted. In this study we focus on the following character sub-types, which are the ones used in Sendratari Ramayana (see Figs. 1 through 6 for the starting and ending positions): Humble-refined (impur) and proud- Figure 1 Humble-refined: left, initial position; right, final position. refined (kagok-kinantang), which are derived from the refined (alu- san) type. Humble-strong (kambĕng) and proud-strong (kalang-kinantang), derived from the strong (gagahan) type. Animal sub-types, which are the monkey (kambĕng-dhĕngklik) and the bird (Jatayu). It should be noted that Jatayu is the proper name of a specific character, and thus we capitalize it in writing. There are certain motions that are specific to a single character type and some that are shared across characters. Figure 2 Proud-refined: left, initial position; right, final position. For this comparative biomechanical study we selected the jumĕnĕng mo- tion (where the character stands up from a kneeling position), since this motion is performed by all character sub-types. For consistency, we wanted a single dancer to perform all motions according to the different character sub-types. Male dancers learn to per- form all character sub-types as part of their training, but not the female character sub-types. The objectives of the study were: 1. to describe the motion patterns of the different character sub-types in terms of biomechanical parameters, and 2. Figure 3 Humble-strong: left, initial position; right, final position. to compare motion patterns across the character sub-types to identify differ- reaction forces (GRF) for anterior- extremity injuries that might affect ences between them. We hypothesized posterior, medio-lateral, and vertical the dance biomechanics was recruited that, compared to the strong and components; and 4. joint moments. from Java, Indonesia. His anthropo- animal sub-types, the refined sub-types metric data were recorded (Table 1) would exhibit lower values in all the Methods and he gave informed consent before following measurements: 1. range of Subject Recruitment and participation according to the uni- motion (ROM) for both lower and Preparation versity’s Institutional Review Board upper body; 2. angular velocities for the One professional male Javanese dancer guidelines. The dances are always upper and lower body joints; 3. ground with no history of upper and lower performed barefoot, and we kept this 106 Volume 23, Number 3, 2019 • Journal of Dance Medicine & Science

plates (AMTI, Watertown, Massachu- setts, USA) were used to obtain GRF data at a sampling rate of 1,000Hz. The force plates were synchronized to the motion capture system, and both were calibrated according to the manufacturer’s recommendations be- fore the dance trials were conducted.15 Experimental Protocol The dancer was instructed to perform the jumĕnĕng (standing up motion) as it corresponds to the six different character sub-types described ear- Figure 4 Proud-strong: left, initial position; right, final position. lier. In the initial kneeling position (jĕnkĕng) the right knee is bent on the floor with the foot placed in a dorsiflexed position while the left knee is slightly flexed with the foot on the ground. Afterward, the dancer stands up, making a full extension of his right knee with the left knee in a slightly flexed position. Although the initial position of the lower limbs is roughly similar in all character sub-types, the final position varies in each case (Figs. 1 through 6). The dancer performed a total of three tri- als per character sub-type. After each performance the dancer was given 2 Figure 5 Monkey: left, initial position; right, final position. minutes of rest. Data Collection and Processing Vicon Nexus 1.8.5 and Polygon 3.5 software were used for data collec- tion and processing, respectively. The kinematic data were smoothed using a Woltring filter with a mean square error of 20 mm2.15-17 All kinetic (GRF and joint moments) and kinematic (joint angles and angular velocities) data were averaged among the three trials in each character sub-type set. The joints of interest for upper extremity were: shoulder, elbow, and wrist; and for lower extremity: Figure 6 Bird: left, initial position; right, final position. hip, knee, and ankle. The ROM was calculated from the difference be- condition in the study. Therefore, shoe employed to collect kinematic data at tween the joint angle at maximum variations do not affect the dancer’s a sample rate of 100 Hz. Forty-one extension and the joint at maximum performance, either in the experiment reflective markers (14 mm diameter) flexion for the sagittal plane. The joint or in real-life situations. placed by an experienced instructor angular velocities were calculated as were attached to the dancer following the first derivative of the joint angles Instrumentation the full body Plug-In-Gait Marker with respect to time in the sagittal A motion-capture system (Vicon MX, model (Vicon, Oxford Metrics, Ox- plane. Joints moments were obtained Oxford Metrics, Oxford, UK), con- ford, UK) to facilitate capture of the through use of the Vicon Polygon 3.5 sisting of seven infrared cameras, was dancer’s motion. Two embedded force software. Journal of Dance Medicine & Science • Volume 23, Number 3, 2019 107

Table 1 Anatomical Measurements of the Javanese Dancer often indicates differences among characters. Body Measurement Side Units Body Measurement Side Units Weight N/A 67.00 kg Shoulder Offset Left 0.042 m Angular Velocities Height N/A 1.650 m Right 0.042 m The peak joint angular velocities of the Leg Length Left 0.850 m Elbow Width Left 0.068 m different sub-types are summarized in Figures 11 and 12. Interestingly, Right 0.840 m Right 0.066 m analysis of the peak angular velocities Knee Width Left 0.093 m Wrist Width Left 0.052 m indicated that humble-refined and Right 0.094 m Right 0.049 m humble-strong showed the lowest val- ues when compared to the strong and Ankle Width Left 0.061 m Hand Thickness Left 0.032 m animal sub-types at the ankle, knee, Right 0.062 m Right 0.036 m shoulder, and wrist joints. Regarding the left ankle joint, humble-refined exhibited low peak values compared Statistical Analysis (such as the left knee) the ROM is to the strong and animal sub-types, One-way (character sub-type) ANO- actually greatest in proud-refined while proud-refined showed lower VA, followed by Holm-Sidak post (Fig. 7). Our results indicate that values compared to the bird sub- hoc testing, was used to compare the the left knee exhibited a significantly type (Fig. 11). The refined sub-types peak vertical GRF, ROM, angular larger ROM (p < 0.05) for proud- showed the lowest significant values at velocities, and joint moments of the refined compared with humble- the right knee joint: humble-refined character sub-types. All significance strong, proud-strong, and monkey and proud-refined produced lower levels were set at p = 0.05. (Fig. 7). Regarding the right knee peak values compared to the strong joint, proud-strong showed a higher and animal sub-types (Fig. 11). The Results ROM than humble-refined (Fig. 8). refined sub-types also had the low- Range of Motion In terms of the upper body, proud- est value at the upper body joints: The peak ROM of the different strong displayed lower peak ROM humble-refined showed lower value sub-types are displayed in Figures than monkey for the right elbow joint compared to proud-strong at the left 7 through 10, which only show the (Fig. 9). For the right wrist, humble- shoulder joint while humble and kinematic and kinetic parameters refined had a lower peak ROM than proud showed lower peak values than where statistically significant results monkey and bird, and proud-refined monkey and bird. Humble-strong also were found. Qualitative descriptions showed a lower value than bird (Fig. had lower peak values than the animal of Javanese dance suggest that in 10). Similar to humble-refined, sub-types (Fig. 12). refined sub-types (such as proud- humble-strong displayed lower val- refined) the joints stay closer to the ues compared to monkey and bird Ground Reaction Forces body.14 However, our biomechanical (Fig. 10). The wrist is a meaningful In terms of kinetics parameters, our analysis shows that in some cases joint, since the position of the hand results indicated that the GRF was

Figure 7 Mean peak ROM of the left knee joint. *Significant Figure 8 Mean peak ROM of the right knee joint. *Significant difference p < 0.05. difference p < 0.05. 108 Volume 23, Number 3, 2019 • Journal of Dance Medicine & Science

Figure 9 Mean peak ROM of the right elbow joint. *Significant Figure 10 Mean peak ROM of the right wrist joint. *Significant difference p < 0.05. difference p < 0.05.

A B

Figure 11 Mean peak angular velocity of the left ankle joint (A) and right knee joint (B). *Significant difference (p < 0.05).

A B

Figure 12 Mean peak angular velocity of the left shoulder joint (A) and right elbow joint (B). *Significant difference (p < 0.05). Journal of Dance Medicine & Science • Volume 23, Number 3, 2019 109

Figure 13 Peak value of medial-lateral ground reaction force Figure 14 Peak value of anterior-posterior ground reaction force (ML GRF) of the left leg. *Significant difference (p < 0.05). (AP GRF) of the right leg. *Significant difference (p < 0.05). different for each sub-type (Figs. 13 proud-strong at the anterior-posterior other hand, proud-strong sub-types and 14). Humble-strong showed a component, and 4. joint moments often represent ogres that require a significantly higher value compared were not affected significantly by the smaller ROM to give the appearance to monkey for the left leg at the diverse sub-types. of strength. Although hand gestures medial-lateral GRF component. At It is interesting, and counterintui- can sometimes play important roles the right leg for the anterior-posterior tive, that proud-refined exhibited the in character differentiation, for this GRF, humble-refined and proud- biggest ROM at the left knee, given study we only considered wrist mo- refined displayed lower peak values that it is considered a refined sub- tions. Our results indicate that the compared to the strong and animal type. In contrast, the most refined animal sub-types (bird and monkey) subtypes. This is in keeping with the sub-type, humble-refined, did show showed bigger ROM than humble- perception that humble-refined and the lowest ROM at the right knee. strong and humble-refined at the right proud-refined are more “balanced” This difference between humble- wrist joint, but both humble-refined and symmetric than the rest of the refined and proud-refined calls for a and humble-strong had lower ROM character sub-types. more nuanced understanding of what at this joint. Thus, we conclude that “refinement” means, and it is this kind the ROM of this joint cannot be Joint Moments of new knowledge that biomechanics used to distinguish between refined With regard to joint moments, our can offer. and strong sub-types, although it is results did not show any significant The upper body plays an important helpful in the human versus animal differences between the character role in Sendratari and offers more differentiation. sub-types. clues about characterization than Taken together, our results indi- the lower body. Our results showed cate that the movements that look Discussion that the right elbow and wrist were more realistic and refined to Javanese The purpose of this study was to ana- two joints that had significant differ- audiences (the refined sub-types) lyze and compare elements of Javanese ences across all character sub-types. often require a greater ROM. This dance from a biomechanical perspec- Traditionally, the wrist and elbow are is again counterintuitive, but not il- tive. Our hypothesis was that refined emphasized in qualitative descriptions logical from a traditional perspective. character sub-types would exhibit of Javanese dance,14 but there is no More refined and realistic sub-types lower values for 1. ROM, 2. angular previous mention of the prevalence of are made to look supple by moving velocities, 3. GRF, and 4. joint mo- the left side over the right for differen- their joints constantly, whereas strong ments. This hypothesis was only par- tiation purposes. The right elbow data sub-types project strength by remain- tially verified, as the key findings were: showed that the movements of proud- ing in larger frames (i.e., obtuse and 1. humble-refined showed the lowest strong and monkey were significantly straight angles) that do not require a peak ROM at the right knee and different since the dancer exhibited big ROM. right wrist joints, 2. humble-refined greater ROM for the monkey sub- Moreover, in terms of joint angular and proud-refined showed the lowest type than for proud-strong. This velocities, our results indicated that peak angular velocities at upper and makes sense from a dance perspec- the refined sub-types showed lower lower joints, 3. humble-strong had the tive, as the monkey sub-type requires values compared to strong and ani- highest peak ground reaction forces larger flexion-extension movements mal sub-types at the upper and lower at medial-lateral components and to imitate simian motion. On the body joints. The right knee was a key 110 Volume 23, Number 3, 2019 • Journal of Dance Medicine & Science lower body joint in terms of kinemat- Vertical force is an important ferences across the sub-types that had ics (but, interestingly, not the left parameter, as it is related to the force not previously been explored from a knee). Proud-strong, humble-strong, that the dancer needs for standing up. biomechanical perspective. For ex- and bird showed the highest angular However, our results did not show ample, the refined sub-types tend to velocities, while humble-refined and any significant difference across the exhibit lower angular velocities and proud-refined showed the lowest. vertical force needed by different sub- lower anterior-posterior GRF values, These findings are consistent with types, suggesting that all sub-types especially for humble-strong, which traditional knowledge. Similarly, used similar vertical components to was the sub-type with the highest the upper body joints of the refined stand up, even when refined sub-types number of significant differences. sub-types showed the lowest peak required less stabilization. The low values observed in this sub- values. The low peak values at the Interestingly, the results did not type suggest that the kinematic and shoulder and elbow joints suggest show any significant difference in kinetic values were affected by the that refined types require smoother peak joint moments for left and right different sub-types. Refined subtypes movements compared to the more lower body joints. A lack of significant displayed low peak values of kinetic aggressive movements of strong and results in the joint moments could and kinematic parameters, suggesting animal sub-types. The low values on suggest similar muscle activity across they require the dancer to move his the peak angular velocities suggest that all sub-types, but further studies are joints in a constant, smooth manner. the motion for refined sub-types tends needed to analyze the muscle pattern This makes sense from a traditional to be smoother. Conversely, strong of all character sub-types. perspective, but the biomechanical and animal sub-types, which tend analysis provides a more specific vo- to be more aggressive, display higher Limitations cabulary with which to describe move- angular velocities. Although we were able to quantify and ment differences across sub-types. Ground reaction force is an im- describe the biomechanical differences Acknowledgments portant parameter in biomechanical of the diverse sub-types, several limi- analysis, and in the case of Javanese tations should be acknowledged. For Funding for this project was provided dance it can help to quantify the example, our analysis was constrained by the NUS Start Up Grant FY2015. dancer’s force distribution for each to the sagittal plane, and further stud- References character sub-type. Regarding the ies might well examine the frontal and medial-lateral components, the results transversal planes, as well as muscle 1. Chang M, Halaki M, Adams R, et al. An exploration of the perception of suggest that, for the refined sub-types, electrical activity. The resulting analy- dance and its relation to biomechani- the peak values differ considerably be- sis could be useful to dance educators. cal motion: a systematic review and tween humble and proud. Our results As noted earlier, this study collected narrative synthesis. J Dance Med Sci. only showed a significant difference data from a single dancer. This could 2016 Sep;20(3):127-36. between humble-strong and monkey. be considered a limitation, but since 2. Midgett VD, O’Bryant HS, Stone Medial-lateral forces describe how the basic movements are standard- MH, Johnson RL. Effects of arm the dancer leans his body toward the ized in the tradition, we assume that position on hang time during a grand center of gravity. Our results demon- the movements would be the same if jeté. Kinesiol Med Dance. 1993 strate that strong sub-types require the performed by any other traditionally Spring/Summer;15(2):3-22. dancer to lean his body with greater trained dancer. Another limitation 3. Wilson M, Kwon YH. The role of force. Contrary to our expectations, is that we only used one movement biomechanics in understanding ĕ ĕ dance movement: a review. J Dance the animal sub-types showed the low- (standing-up, or jum n ng). Future Med Sci. 2008 Sep;12(3):109-16. est peak values of medial-lateral GRFs. studies might look at more move- 4. Winter DA. Biomechanics and Motor Perhaps these sub-types required low ments and at geographical variations Control of Human Movement. New medial forces as a strategy to keep of the same dance. Further studies York: John Wiley & Sons, 2009. the body balanced. At its anterior- could also build on this foundation 5. Bejjani F, Halpern H, Nordin M, posterior components, GRF showed to analyze historical and geographical et al. Spinal motion and strength that the refined sub-types displayed changes across Javanese dances, using measurements of flamenco dancers the lowest peak values compared to biomechanical quantification together using 3D motion analyzer and cybex the strong and animal sub-types. This with other modes of historical and II dynamometer. In: Groot G, Hol- result suggests that the refined sub- descriptive analysis. lander AP, Huijing PA, et al. (eds): types produced low values as a com- Biomechanics XI-B. Amsterdam: Free Conclusion University Press, 1988, pp. 925-930. pensatory strategy when performing 6. Bennell K, Khan KM, Matthews B, delicate motions. Lower values on the This report described the kinetic and et al. Hip and ankle range of motion anterior-posterior components might kinematic differences of character and hip muscle strength in young fe- indicate a better body weight distribu- sub-types in the Javanese dance, male ballet dancers and controls. Br J tion to generate a high stabilization on Sendratari Ramayana. The findings Sports Med. 1999 Oct;33(5):340-6. the upper and lower body. identified small but meaningful dif- 7. Imura A, Iino Y. Comparison of low- Journal of Dance Medicine & Science • Volume 23, Number 3, 2019 111

er limb kinetics during vertical jumps N, Frankel V. Dynamic loading 15. Yeow CH, Lee PVS, Goh JCH. An in turnout and neutral foot positions on flamenco dancers: a biome- investigation of lower extremity by classical ballet dancers. Sports chanical study. Hum Mov Sci. 1989 energy dissipation strategies during Biomech. 2017 Mar;16(1):87-101. Oct;8(5):503-13. single-leg and double-leg landing 8. Lin CF, Su FC, Wu HW. Ankle 12. Parto FS. Recent history of Java- based on sagittal and frontal plane biomechanics of ballet dancers in nese classical dance: a reassessment. biomechanics. Hum Mov Sci. 2011 relevé en pointé dance. Res Sports Contemporary Theatre Review. Jun;30(3):624-35. Med. 2005 Jan-Mar;13(1):23-35. 2001;11(1):9-17. 16. Burnfield JM, Powers CM. The role 9. McGuinness D, Doody C. The 13. Hughes-Freeland F. Embodied Com- of center of mass kinematics in pre- injuries of competitive Irish munities: Dance Traditions and dicting peak utilized coefficient of dancers. J Dance Med Sci. 2006 Change in Java (vol. 2). Oxford, friction during walking. J Forensic Mar;10(1&2):35-9. UK: Berghahn Books, 2008. Sci. 2007 Nov;52(6):1328-33. 10. Shippen JM, May B. Calculation 14. Sudarsono. Wayang Wong in the 17. Lewis A, Stewart C, Postans N, of muscle loading and joint contact Yogyakarta Kraton: History, Ritual Trevelyan J. Development of an forces during the rock step in Irish Aspects, Literary Aspects, and Char- instrumented pole test for use as a dance. J Dance Med Sci. 2010 acterization. Ann Arbor, Michigan: gait laboratory quality check. Gait Mar;14(1):11-8. University Microfilms International, Posture. 2007 Jul;26(2):317-22. 11. Voloshin AS, Bejjani FJ, Halpern 1983.