ArticlePDF Available

EFFECTS OF A PIANIST'S BODY MOVEMENTS ON LISTENERS' IMPRESSIONS

Authors:
Article

EFFECTS OF A PIANIST'S BODY MOVEMENTS ON LISTENERS' IMPRESSIONS

Abstract and Figures

We often experience different impressions when we watch the recorded video of a musical performance, and when we only listen to the music. Several studies have revealed that visual information such as facial expressions or concert dress communicated performers' intention to the listeners. Several psychologists have investigated body movements in particular and showed that there is a relationship between performers' body movements and listeners' perception of intended expressivity. Emotional impressions of listeners, however, have not been considered in connection with body movements This paper aims at clarifying how a performer's body movements affect listeners' emotional impressions. To investigate this, 91 observers were presented with two recordings of a professional pianist ("Etude Tableaux Op.39-1" and "Prelude Op.32-5" by Rachmaninoff) in three modes (sound only, vision only, and sound and vision). They rated their impressions on 31 scales, as well as their liking of the respective recording. The results showed that the effects of visual information differed depending on the strength of the impressions produced by the music. Impressions that were not strongly induced were significantly affected when visual information was added. This possibly happened because the auditory information alone was difficult for the listeners to rate, and visual information made rating easier. The effects of visual information, however, were not necessarily positive. Listeners' liking for the sound-with-vision mode comes near the liking for the visual mode. Thus, sound with vision might not be preferred to sound only, if the visual information is disliked. To sum up, we revealed that body movements affect listeners' impressions. By taking a performer's "expression" into consideration in the future, we aim to create a model of the processes involved in musical impressions formation.
Content may be subject to copyright.
The inaugural International Conference on
Music Communication Science 5-7 December
2007, Sydney, Australia
http://marcs.uws.edu.au/links/ICoMusic
Proceedings of ICoMCS December 2007 Page 143
EFFECTS OF A PIANIST’S BODY MOVEMENTS
ON LISTENERS’ IMPRESSIONS
Haruka Shoda
1
Toshie Nakamura
2
, Maria Raluca Draguna
2
, Satoshi Kawase
2
, Kenji Katahira
2
and Shoko Yasuda
2
1
School of Human Sciences, Osaka University, Japan
2
Graduate School of Human Sciences, Osaka University, Japan
ABSTRACT
We often experience different impressions when we watch
the recorded video of a musical performance, and when we
only listen to the music. Several studies have revealed that
visual information such as facial expressions or concert
dress communicated performers’ intention to the listeners.
Several psychologists have investigated body movements
in particular and showed that there is a relationship
between performers’ body movements and listeners’
perception of intended expressivity. Emotional
impressions of listeners, however, have not been
considered in connection with body movements
This paper aims at clarifying how a performer’s body
movements affect listeners’ emotional impressions. To
investigate this, 91 observers were presented with two
recordings of a professional pianist (“Etude Tableaux
Op.39-1” and “Prelude Op.32-5” by Rachmaninoff) in
three modes (sound only, vision only, and sound and
vision). They rated their impressions on 31 scales, as well
as their liking of the respective recording.
The results showed that the effects of visual information
differed depending on the strength of the impressions
produced by the music. Impressions that were not strongly
induced were significantly affected when visual
information was added. This possibly happened because
the auditory information alone was difficult for the
listeners to rate, and visual information made rating easier.
The effects of visual information, however, were not
necessarily positive. Listeners’ liking for the
sound-with-vision mode comes near the liking for the
visual mode. Thus, sound with vision might not be
preferred to sound only, if the visual information is
disliked.
To sum up, we revealed that body movements affect
listeners’ impressions. By taking a performer’s
“expression” into consideration in the future, we aim to
create a model of the processes involved in musical
impressions formation.
Keywords
impression, body movement, piano performance
1. INTRODUCTION
The rapid technological changes enable us to enjoy music
through both auditory and visual channels. When we watch,
we often have different impressions than those we have
when we only listen to the music. Hargreaves et al.
[1]
pointed out that the processes involved in musical
communication include many factors. Visual information
appears to be one of these factors.
Earlier studies have revealed that visual information such
as facial expression or concert dress was useful for
communicating performers’ intention to the listeners
(Ohgushi
[2]
; Griffith et al.
[3]
). Although visual information
has many components, herein we focused on body
movements. Some empirical investigations have suggested
that musicians’ movements communicated perceptual
information (e.g., Davidson
[4]
; Ohgushi
[2]
). Additionally,
Davidson
[5]
showed that a relationship existed between
movement size and performance expression (i.e.,
“Deadpan” expression induced small movement size;
“Exaggerated” induced large size).
Although the relation between a performer’s body
movements and listeners’ perception was examined, the
connection between musicians’ movements and listeners’
emotional impressions has been so far neglected. Juslin
[6]
suggested that music communicates emotion; a performer
(or a composer) expresses his/her emotional intention, and
listeners can perceive it. Thus, body movements appeared
to be connected with such emotional communication.
In light of this, we attempted to clarify the relationship
between a performer’s body movements and listeners’
impressions. In the present study which represents a first
step in this direction, we examined how listeners’
impressions changed when visual information was added.
2. METHOD
2.1 Stimuli
We audio- and video-taped performances of a professional
pianist who played two pieces (“Etude Tableaux Op.39-1”
and “Prelude Op.32-5” by Rachmaninoff; from hereon we
shall refer to them as “Etude” and “Prelude”). The pianist
was instructed to perform them as if he was in a concert
hall.
As the focus of our research was body movements, the
performer’s facial expressions were not recorded since it
has been suggested that they have a significant influence
on the listeners’ impressions. A sample of the recorded
scene is given in Figure 1.
Three types of stimuli were created: sound only mode (S),
vision only mode (V), matched sound and vision mode
(SV). The length of the stimuli was 3’ 04” (Etude) and 2’
48” (Prelude).
The inaugural International Conference on
Music Communication Science 5-7 December
2007, Sydney, Australia
http://marcs.uws.edu.au/links/ICoMusic
Proceedings of ICoMCS December 2007 Page 144
Figure 1: Sample of the visual mode. The white marks on
the performer’s body were used for the measurement of his
body movements.
2.2 Observers
Ninety-one students (41 male, 50 female) of Osaka
University served as observers. Their age ranged from 18
to 24 years (mean = 19.4 years, SD = 1.63). All of them
underwent three tests: S, V and SV modes.
2.3 Apparatus
The tests took place in a sound proof room of Osaka
University. In the V and SV conditions, a projector
(EPSON, ELP-735) was used to present the stimuli. The
visual content was shown from a separate room through a
translucent screen on a quadplex window (Figure 2).
Figure 2: Experimental setting.
2.4 Procedure
The tests occurred at least 1 week apart. The S test took
place first, followed by the V test. The last test (SV) was
carried out about one month later. This was done in order
to prevent memory of the renditions affect the ratings.
The procedure was the same for all tests. First, two
renditions were presented to the observers. Each stimulus
was rated by marking the appropriate number on
nine-point monopolar scales for 31 adjectives (Table1). 24
of the adjectives were selected from Hevner
[7]
, 2 from
Davidson
[4]
, and 5 from a preliminary interview with the
pianist. Liking of the renditions was rated on a nine-point
bipolar scale (9: most liked, 1: least liked). In the case of
SV, the observers were asked to rate only the sound.
Each test lasted approximately 40 to 60 minutes.
Factor Adjective
Elation Cheerful (H), Joyous (H), Bright (H),
Exhilarated (H), Humorous (H),
Gorgeous (I)
Tranquility Quiet (H), Leisurely (H), Tranquil (H),
Tender (H), Dreamy (H), Graceful (H),
Quick* (I), Longing (H)
Magnificence Majestic (H), Lofty (H), Robust (H),
Passionate (H), Emphatic (H),
Solemn (H), Exaggerated (D)
Gloom Depressing (H), Doleful (H),
Whimsical (H), Restless (H), Dark (H),
Heavy (I), Frustrating (I)
Monotony Deadpan (D), Metrical (I), Sober (H)
Table 1: Five factors were extracted based on factor
analyses. The letter in ( ) represents the source of the
adjectives, i.e., H for Hevner
[7]
, D for Davidson
[4]
and I
for “interview with the pianist”. *Factor loading for
Quick was negative. In computing the factor mean, the
mean for Quick needed to be subtracted from 10.
3. RESULTS
3.1 Factor Analysis
For better understanding of the relations between
impressions, a factor analysis was computed. The oblique
solution (promax) contained the five factors shown in
Table1. We termed them Elation, Tranquility,
Magnificence, Gloom and Monotony. Because Cronbach’s
alpha factors were much higher than 0.8, we computed the
mean of the components for each factor as the
representative value of impressions.
Sound Proof Room
quadplex window
translucent screen
projector
observers
Sound Proof Room
quadplex window
translucent screen
projector
observers
3.2 Impressions
Figure 3 shows the mean values of impressions in each
mode. A repeated measures ANOVA with mode as the
within-participants factor was used for comparison.
Tukey’s test revealed differences between the three modes.
Sound. In S mode, the impression which was strongly
induced was termed “Specific”. On the other hand, the
value not highly rated was termed “Non-specific”.
“Specific” impressions represented the emotional features
of the pieces.
Figure 3 shows that Magnificence and Gloom were
“Specific” and Elation and Tranquility were
“Non-specific” impressions for Etude. The reverse is
applicable to Prelude.
Vision. Visual impressions were significantly different
from those in S and SV modes for all factors except
Monotony.
Sound & Vision. Impressions of SV were significantly
different from those in S mode for “Non-specific”. On the
other hand, “Specific” impressions were similar in most
cases. Only Magnificence in Etude differed significantly;
The inaugural International Conference on
Music Communication Science 5-7 December
2007, Sydney, Australia
http://marcs.uws.edu.au/links/ICoMusic
Proceedings of ICoMCS December 2007 Page 145
however, the significance level was larger than that for
“Non-specific” impressions.
Monotony impressions were similar in all modes.
Figure 3: Means of impressions in the three modes.
3.3 Liking of Stimuli
There was no significant difference between S and SV
modes in terms of liking of the renditions. As the liking
rating of SV was close to that of V, we may consider that
SV rating was affected by the visual information.
Figure 4: Means of liking.
4. DISCUSSION
4.1 Visual influence on impressions
The existence of significant differences between S and SV
modes may be attributed to the difficulty of rating
“Non-specific” impressions. Further, the fact that SV
impressions are close to V impressions only for
“Non-specific” impressions showed that visual
information affects ratings of impressions produced by the
music in SV mode.
On the other hand, “Specific” impressions did not differ
between S and SV mode and impressions of V were rather
different from those of the other modes. This may be due to
the fact that such impressions are easier to rate.
**
***
***
***
***
**
***
***
*
***
***
1
2
3
4
5
6
7
8
9
Elevated Dainty Glorious Gloomy Prosy
S
V
SV
Elation Tranquility Magnificence Gloom Monotony
(a) Etude Tableaux Op. 39-1
***
***
***
***
***
***
***
***
***
***
1
2
3
4
5
6
7
8
9
Elevated Dainty Glorious Gloomy Prosy
S
V
SV
(b) Prelude Op.32-5
***p<.001, **p<.01, *p<.05
“Non-Specific”
“Specific”
Impression RatingsImpression Ratings
Elation Tranquility Magnificence Gloom Monotony
“Specific” “Non-Specific”
**
***
***
***
***
**
***
***
*
***
***
1
2
3
4
5
6
7
8
9
Elevated Dainty Glorious Gloomy Prosy
S
V
SV
Elation Tranquility Magnificence Gloom Monotony
(a) Etude Tableaux Op. 39-1
***
***
***
***
***
***
***
***
***
***
1
2
3
4
5
6
7
8
9
Elevated Dainty Glorious Gloomy Prosy
S
V
SV
***
***
***
***
***
***
***
***
***
***
1
2
3
4
5
6
7
8
9
Elevated Dainty Glorious Gloomy Prosy
S
V
SV
(b) Prelude Op.32-5
***p<.001, **p<.01, *p<.05
“Non-Specific”
“Specific”
Impression RatingsImpression Ratings
Elation Tranquility Magnificence Gloom Monotony
“Specific” “Non-Specific”
Therefore, the effects of body movements differed
depending on the type of impressions: impressions that did
not reflect the emotional features of the music were
affected by visual information; those which reflected such
features were not influenced when visual information was
added.
Monotony impressions did not show any variation between
the three modes for both renditions. In this case,
impressions for S and V modes were very similar, this
probably leading also to a similar impression rating in the
SV mode. Other pieces of music need however to be
examined in this respect.
4.2 Liking and Disliking Tendency
In Etude, SV mode was most liked by observers. In Prelude,
however, the most liked mode was S. This shows that the
effects of visual information are not always positive.
However, as the difference between S and SV was not
statistically significant for both renditions, it is also
possible that visual information does not notably
contribute to the liking of a musical piece.
4.3 Controversial Points and
Prospects for Future Research
This research aimed at providing some insight into
impressions produced by audiovisual stimuli; however,
several problems came into focus. The listeners’
impressions appear to change depending on many various
factors, such as situations, contexts, listeners’ individual
characteristics, amongst others, as Hargreaves et al.
[1]
suggested in their reciprocal feedback model of musical
response. Therefore, bringing light to the relationship
between the performer’s body movement and the listeners’
impressions requires a more thorough control of these
factors.
*
***
4
5
6
7
8
Et ude Op. 39- 1 Prelude Op.32- 5
S V SV
Etude Op. 39-1 Prelude Op.32-5
***p<.001, *p<.05
9
1
7
6
5
Liking
Disliking
*
***
4
5
6
7
8
Et ude Op. 39- 1 Prelude Op.32- 5
S V SV
Etude Op. 39-1 Prelude Op.32-5
***p<.001, *p<.05
9
1
7
6
5
*
***
4
5
6
7
8
Et ude Op. 39- 1 Prelude Op.32- 5
S V SV
Etude Op. 39-1 Prelude Op.32-5
***p<.001, *p<.05
9
1
7
6
5
Liking
Disliking
Additionally, the music we used in this experiment was
only classical music played by only “one” performer. We
may need to use other music genres, as well as other
performers as stimuli. Further, in this study both performer
and listeners were Japanese; as Folkestad
[8]
, implied that
impressions differ with listeners’ national identity,
responses of listeners with various cultural backgrounds
need to be examined.
Observers in our study were 18-24 year old students, and
their musical experience varied. However, there were no
highly trained musicians among them. Different
impressions may be generated in skilled listeners. Children
who have little musical knowledge may also have different
impressions. Such aspects need consideration as well in
this research.
Further, the analysis of acoustical features (i.e., tempo,
level) and the measurement of body movement remain to
The inaugural International Conference on
Music Communication Science 5-7 December
2007, Sydney, Australia
http://marcs.uws.edu.au/links/ICoMusic
Proceedings of ICoMCS December 2007 Page 146
be implemented. Through such analyses, an objective
description can be obtained.
Shoda et al.
[9]
showed that a performer’s intended
expressions (i.e. “Deadpan”, ”Artistic”, ”Exaggerated”)
affect listeners’ impressions in the presentation of musical
sound only. By applying the performer’s intended
expression to visual stimuli, we intend to investigate the
relationship between intended expression, body
movements, and impressions.
Earlier studies revealed the mechanism of musical
behavior. For example, Hargreaves et al.
[1]
created a model
of musical communication and its relevant determinants;
Juslin
[10]
proposed a Brunswikian lens model for
communication of emotion in music performance. In this
direction, our final target is to create a model of the
processes involved in musical impressions formation that
includes visual information, as most models proposed in
musical psychology have neglected that. This study was
the first step towards this target. In addition, this model can
contribute to various fields, such as therapeutic
communication or jam-session system.
5. CONCLUSION
To investigate the effects of a performer’s movements on
the listeners’ impressions, we asked observers to rate two
renditions in three modes (i.e., Sound only, Vision only,
Sound and Vision). The effects were different on each
impression. If an impression was strongly induced in S
mode, S and SV exhibited similar impressions. In addition,
the SV impressions are close to V impressions when the S
impressions were NOT highly rated. It was also shown that
visual information affected the liking/disliking of the
musical piece. We consider these findings to be the first
step towards creating a model of musical impressions
formation in relation to visual information provided by
performers.
6. REFERENCES
1. Hargreaves, D.J., MacDonald, R.A.R & Miell, D.E.
2005 How do people communicate using music? In
D.E. Miell, D.J. Hargreaves and R.A.R MacDonald
(Eds.), Musical Communication. New York, Oxford
University Press. Pp. 1-25.
2. Ohgushi, K. 2006 Interaction between auditory and
visual information in conveyance of players' intentions.
Acoustical Science and Technology, 27(6), 336-339.
3. Griffith, N & Davidson, J.W. 2006 The effects of
concert dress and physical appearance on perceptions
of female solo performers. Proceedings of 9
th
ICMPC,
Bologna, University of Bologna. Pp. 1723-1726.
4. Davidson, J.W. 1993 Visual perception of
performance manner in the movements of solo
musicians. Psychology of Music, 21, 103-113.
5. Davidson, J.W. 1994 Which areas of a pianist's body
convey information about expressive intension to an
audience? Journal of Human Movement Studies, 26(6),
279-301.
6. Juslin, P.N. 2005 From mimesis to catharsis:
expression, perception, and induction of emotion in
music. In D.E. Miell, D.J. Hargreaves and R.A.R
MacDonald (Eds.), Musical Communication. New
York, Oxford University Press. Pp. 85-116.
7. Hevner, K. 1936 Experimental studies of the elements
of expression in music. American Journal of
Psychology, 48, 246-268.
8. Folkestad, G. 2002 National identity and music. In
R.A.R MacDonald, D.J. Hargreaves and D.E. Miell
(Eds.), Musical identities. New York, Oxford
University Press. Pp. 151-162.
9. Shoda, H. et al. 2007 Piano ensosha no hyogen no
chigai ni yoru choshusha no insho hyotei no henka
[Change in listeners' evaluation of impression by the
difference in performer's intentional expressions in
piano performance] Proceedings of the 5
th
meetings of
the Japanese Society for Cognitive Psychology, Kyoto,
Kyoto University. Pp. 71. (in Japanese)
10. Juslin, P. N. 1995 Emotional communication in music
viewed through a Brunswikian lens. In Musical
expression. Proceedings of the Conference of ESCOM
and DGM 1995, (ed. G. Kleinen), Bremen, University
of Bremen. Pp. 21-25.
7. ACKNOWLEDGEMENT
We would like to thank Takuya Watari for his
performance.
... In the study of Shoda et al. (2007), it was seen that visual information about performer has an important effect on listener; however, visual information does not always leave a positive effect. At this point, it can be stated that it is necessary to use gestures and body movements in harmony with composition. ...
Chapter
Full-text available
Abstract As one of the most important functions of transferring emotions, thoughts, or knowledge among people, communication has an important place for expressing the art of music, which is one of the special fields where verbal and nonverbal communications occur together. In this context, it is thought that the process of communication should be addressed carefully in theoretical and behavioral terms at the stage of instrumental music education, which is a sub-dimension of music education, and at the stage of performance that is carried out within the scope of this education. In this study, the elements of communication process were mentioned basing upon the general definition of communication. Further, a model of “Musical Communication Process” was created in order to discuss in the performing process that forms the main framework of the study. Musical communication process was examined in an extensive framework by means of descriptive analysis at all the phases of performing process, by presenting each of the elements individually and their relations among each other. It is thought that the ideas and recommendations obtained as a result of this study will be a guide for music educators and students within the scope of instrumental music education practices. It is further thought that these ideas and recommendations will contribute to acquisition of effective performing skills. Özet İnsanlar arası duygu, düşünce ya da bilgi aktarımının en önemli işlevlerinden biri olan iletişim, sözlü ve sözsüz iletişimin bir arada gerçekleştirildiği sayılı alanlardan biri olan müzik sanatının ifade edilmesinde de önemli bir yer tutar. Bu bağlamda, müzik eğitiminin bir alt boyutu olan çalgı eğitimi ve bu eğitim kapsamında gerçekleştirilen performans sergileme aşamasında iletişim sürecinin kuramsal ve davranışsal olarak dikkatle ele alınması gerektiği düşünülmektedir. Bu çalışmada, iletişimin genel tanımından yola çıkılarak iletişim sürecinin öğelerine değinilmiş ve çalışmanın ana çerçevesini oluşturan performans sergileme sürecinde ele alınmak üzere bir “Müziksel İletişim Süreci” modeli oluşturulmuştur. Müziksel iletişim süreci, performans sergileme sürecinin tüm aşamalarında, her bir öğesi kendi başına ve birbiriyle olan ilişkileri ortaya konularak betimsel analiz yoluyla geniş bir çerçevede incelenmiştir. Çalışma sonucunda elde edilen fikir ve önerilerin, çalgı eğitimi uygulamaları dâhilinde yararlanılmak üzere müzik eğitimcilerine ve öğrencilerine yol gösterici olacağı, etkili performans sergileme becerilerinin kazandırılmasında yarar sağlayacağı düşünülmektedir.
Article
Full-text available
This chapter first sets out the purpose of the book, which is to explain the musical, social, and cultural processes that underlie the eventual realization of the acoustic performance event (by the composer, performer, arranger, and all others involved), the means by which they lead to the listener's response, and thence to short and long-term effects upon arousal level, cognition, emotion, and subsequent behaviour. It then discusses a reciprocal feedback model of musical communication. An overview of the subsequent chapters is presented.
Article
Full-text available
A literature review demonstrates the importance of visual information in performance. There is evidence that dress and appearance affect perceptions of musical elements of a performance. There are differences in men and women's dress that may put female classical soloists at a disadvan-tage that could affect their musical advancement. This study investigates the roles of concert dress and physical attractiveness on audience perceptions of performance quality, focusing on solo women violinists of the Western Classical tradition. Four female violinists were filmed playing in four states of dress: Jeans, a Nightclubbing dress, black Concert dress and Point-light condition (body movement is apparent but not physical appearance). Each clip was recorded in two conditions: both as the Per-former's own version and with a mastertrack dubbed over the top. The dubbed versions therefore had a constant mu-sical soundtrack. Fifteen male and 15 female participants (age range 17– 66 years) were asked to rate clips on six point scales in terms of Technical Proficiency, Musicality, Appropriateness of Dress and Attractiveness of Performer. Significant effects were found of Dress, Performer and Condition on participant perceptions, and an interaction of Performer and Dress was observed. This work confirms the importance of the visual information in performance and its effect on the formation of audience judgements on per-formers.
Article
This chapter reviews theoretical concepts and empirical findings on musical emotions. First, it examines the notion of music as a means of communicating emotion and presents some relevant evidence concerning the issue. Then it provides a working definition of emotions and some conceptual distinctions for the study of musical emotion. It reviews mechanisms through which music may express and induce emotions. Finally, it considers various objections to music-as-communication and provides an agenda for future research. The discussion is limited to Western music, especially classical and popular music from the 18th century to present day.
Article
Music performers usually play music with some intentions. They play musical notes elongated or shortened and loud or soft for their expressive performances. Furthermore, they move their bodies consciously or unconsciously to enhance expression. Singers often show specific emotions on their faces during singing before audiences. Visual information from performers plays an important role for listeners to understand the performers' intentions. Our studies in this field and a pioneering work by Davidson are reviewed.
Article
The study of music perception has focused almost exclusively on sound, ignoring the role of seeing the performer's body movements. Whilst anecdotes frequently refer to the importance of the performer's movements, there is scant psychological evidence to support this finding. The closest equivalent work in visual event perception research has shown that covert mental dispositions (for instance, an intention to deceive an observer) are specified in body movements, and therefore provide important information for the observer.. With these findings in mind, this article investigates the information conveyed by the movements of a musical performer when s/he is asked to play-the same piece in three different expressive manners. These performance manners are presented to observers in three modes: vision alone, sound alone and sound and vision together to investigate the relative contributions of the different perceptual modes. The results reveal that not only is vision a useful source of information about manner, but that it actually more clearly specifies manner than the other modes. These findings emphasise the need to consider visual as well as sound information in psychological enquiries into music perception.
Piano ensosha no hyogen no chigai ni yoru choshusha no insho hyotei no henka [Change in listeners' evaluation of impression by the difference in performer's intentional expressions in piano performance
  • H Shoda
Shoda, H. et al. 2007 Piano ensosha no hyogen no chigai ni yoru choshusha no insho hyotei no henka [Change in listeners' evaluation of impression by the difference in performer's intentional expressions in piano performance] Proceedings of the 5 th meetings of the Japanese Society for Cognitive Psychology, Kyoto, Kyoto University. Pp. 71. (in Japanese)