Body and Mind in the Pianist s performance

Alma Mater Studiorum University of Bologna, August 22-26 2006 Body and Mind in the Pianist s performance Isabella Poggi Department of Education Sciences University Roma Tre Rome, Italy poggi@uniroma3.it ABSTRACT The paper presents a model of the motor, cognitive and emotional processes taking place in piano performance, and an annotation scheme for the analysis of the pianist s body movements. By applying the scheme to a concert and a rehearsal, the paper finds similarities and differences in the pianist s body movements depending on the musical stracture of the piece and on the performance situation, and finally presents a lexicon of the pianist s body movements. Keywords Music, pianist, body movements, pianist s mind, motor action, cognitive processes, emotional processes INTRODUCTION In the domain of relationship between body and music, interaction and bodily behaviour linked to music performance has been analysed with methods of conversational analysis (Haviland, 2005; Streeck & Oshima, 2005; Duranti, 2006), while the gestures of classic orchestra conductors were analysed in terms of their visual metaphors In: M. Baroni, A. R. Addessi, R. Caterina, M. Costa (2006) Proceedings of the 9th International Conference on Music Perception & Cognition (ICMPC9), Bologna/Italy, August 22-26 2006. 2006 The Society for Music Perception & Cognition (SMPC) and European Society for the Cognitive Sciences of Music (ESCOM). Copyright of the content of an individual paper is held by the primary (first-named) author of that paper. All rights reserved. No paper from this proceedings may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information retrieval systems, without permission in writing from the paper's primary author. No other part of this proceedings may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information retrieval system, without permission in writing from SMPC and ESCOM. (Boyes Braem & Braem, 2004). Previous works have started exploring the expressive role of the pianist s body during music performance. (Caterina et al., 2004), and stressed how deep and manifold is the relationship between body and mind in this activity, thus pointing how further investigation is welcome in such an area. In this work I propose a theoretical model of the pianist s motor, cognitive, emotional, and communicative process, along with an annotation scheme of the pianist s behaviour during piano performance, that allow to make research in this topic and to provide evidence to the proposed model. THE PIANIST S MIND What happens in a pianist s mind while s/he is playing? I propose a model of the pianist s mind, namely a hypothesis about the bodily and mental processes that occur when a pianist is playing. While playing the piano, three kinds of processes take place in the pianist s mind and in the pianist s body: cognitive, emotional and motor processes. Cognitive processes Cognitive processes encompass processes of attention, perception and memory. Attentional processes include, for example, the pianist being in concentration on the music to play, that is, his focusing of attention aimed at performing at his best. On the contrary, the pianist can sometimes be quite relaxed, for example when he masters the piece to play very well: in this case he can suspend his tension and concentration. Attention is also implied, for example, in caution, when the pianist has to play very softly, and so he must behave like in touching something fragile, something not to be spoiled with too vigorous a touch. While the pianist is playing, auditory, visual and tactile perceptive processes are at work. Visual perception may be focussed on the score, when the pianist is not playing by ISBN 88-7395-155-4 2006 ICMPC 1044

heart, but it can also be aimed at seeing the keys, especially for difficult intervals. Auditory perception is indispensable to start at the right moment, when playing with other musicians while, if oriented on the pianist s own playing, it serves as a feedback about how playing goes on. Tactile perception is finally also important as a feedback device. Memory processes relevant for the pianist are numerous and of various kinds. Auditory and visual images play an important role: auditory images obviously cncern the sounds to play, while visual images may concern two kinds of contents. On the one hand, the pianist may have a visual image of the score of the piece to play; but at the same time he can have images of objects or places that inspire his music, like objects, faces, or landscapes: in brief, any kind of visual or synesthaetic image linking sound to vision. Moreover, the pianist holds spatial and tactile memory about where the keys are or how they are felt by hands. Finally, linked to these, the procedural memory of hands and feet movement is central in the pianist s cognitive processes, since it governs his motor processes. Emotional processes In a pianist we can distinguish felt emotions and enacted emotions. The former are those the pianist really feels before, during or after his performance, that are due to aspects of the performance itself. The latter are emotions he should feel to impress them into the music he is playing, and through this, to transmit them to the audience. In this sense, he must recite emotions, feel as if he were feeling some. Not only a professional pianist, but also an amateur one, is like an actor on the stage, who must induce emotions in himself to transmit them through music. Two types of felt emotions can be distinguished: process emotions, the emotions felt during and about the very process of playing. Some of them are positive, like relaxation or flow, the feeling of being doing something beautifully and easily. Other emotions may be negative, like tension or fear of making mistakes in performance. These emotions, can have different distributions, for example, according to the pianist s expertise and habit to perform in public. outcome emotions, the emotions felt about the outcome of one s playing. One can feel disgust or shame for making a mistake, or else pleasure, satisfaction or ecstasy for how sweet it sounds. Enacted emotions can be: meaning oriented emotions, the ones the pianist must simulate or induce in himself in order to exhibit them through music: for instance, feeling sad in order to be able to play a sad music. movement oriented: the emotions the pianist needs to simulate because their expression is linked to a particular movement or manner of movement. Typically, for example, a pianist may frown when he has to play very loud notes: frowning is an expression of anger, and anger makes you so strong and energetic as to touch keys with more strength. Here the emotion, and its consequent expression, is functional to the quality of sound. Actually, research on emotions tells us how in an emotional process body and mind are strictly intertwined so that any event in the former backfires on the latter, and viceversa. So, the feeling may determine the muscular action (Wallbott,, as well as, possibly, the other way around. Motor processes In the pianist s body, motor processes take place, relative to the movements to perform and to the manner in which they should be perfomed. Music is produced by body movements; among the parameters of music, melody, rhythm and harmony are produced by hand movements (which tunes are touched by fingers and when), while parameters like timbre, tempo, expression and intensity can be better seen as determined not by the hand movements themselves, but by the way in which they are perfomed their manner of movement. In a pianist, the way in which motor processes work in producing music is determined by the combination of cognitive and emotional processes. But the Pianist does not move only hands and feet; of course, they do the bulk of the job; but also his/her trunk, head, face, mouth, eyes and eyebrows often move during playing: they participate in the production of sound and of its quality, and at the same time they express the Pianist s cognitive and emotional processes. THE PIANIST S BODY I have proposed a hypothesis about the pianist s mental and bodily processes. But how could we test a such model? Do the processes we hypothesize really occur in his/her mind? And what are the relationships among them? How do cognitive, emotional and motor processes determine and interact with each other? The pianist s body behaviour can be a cue to what is going on in his/her mind; in other words, if we found a way to read the pianist s body we could have some evidence to test our hypotheses. Of course, to this goal the relevant thing is not to analyse the movement of hands and feet, which are doing the technical job; what it is most revealing to analyse are the movements and postures of his/her head, face and trunk. But how can the pianist s body behaviour be transcribed and analysed? In the last fifteen years, also in order to reproduce communicative behaviour in Artificial Agents, a whole research area developed aimed at creating schemes for the annotation of multimodal communication. One such annotation scheme is the score of multimodal communication proposed by Poggi (2006): a procedure to describe, analyse and classify signals in different modalities: verbal, prosodic-intonational, gestural, facial and body behaviour. ISBN 88-7395-155-4 2006 ICMPC 1045

Different variants of the score were used to analyse a number of bodily behaviours: gesture, gaze, eyebrows, touch, speech gesture relationship, the orchestra conductor s face and so on. In this work I present a variant of the musical score that can be used in analysing the multimodal behaviour of music players. By analysing some fragments of a pianist s body behaviour, I will show how this variant of the score can be a useful tool in research about a musician s cognitive, emotional and motor processes. A SCORE FOR THE ANALYSIS OF THE PIANIST S BODY In order to investigate some of the relationships among cognitive, emotional and motor processes in the pianist s mind and body, and thus to have a first test of the model proposed, a research was carried on about a pianist s performance. The performance of the pianist Marcella Crudeli was videotaped and analysed through a variant of the score of multimodal communication specifically built for the analysis of pianists behaviour. The videorecordings were taken during a rehearsal and a concert where Marcella Crudeli was playing the piano concert K 488 by W.A.Mozart, with the orchestra Res Musica conducted by M o Fabrizio Santi. The pianist s multimodal behaviour was analysed through the score presented in Table 1. Later the results of the analysis were shown to the pianist, who provided some further comments and generally agreed about the interpretation of the data given by the researcher. In this and the following Sections I present the structure of the annotation scheme and the categories that were singled out to classify the pianist s body movements. In the subsequent Sections I overview the first results of some analyses carried on Marcella Crudeli s data by using the scheme illustrated. Bar Table 1. The multimodal score of the pianist s body 1 100 1. notes B E D# C B A# A G 2. time 1.20 3. trunk Shoulders raised and backward 4. head Shakes head slowly 5. eyebrows Frowns, then raises inner eyebrows 6. eyes Squeezes eyes, then opens eyes 2 Goal / Meaning I retract, I do not want Sorrow No no = I reject sorrow I am worried + I am sad Help play sharp A (# A) 3 Goal Type / MeaningType CP Mea E CP Mea E Mea E + Mea E HM 7. mouth The analysis is carried on by making reference to the score of the music performed and by examining fragments bar by bar (sometimes even note by note). In the first column you write the bar number, and in the lines 1 through 7, respectively, the following information: 1. the notes of that bar, either in musical notation or with their names; 2. the time in the tape at which the passage occurs; 3. 7. the modalities taken into account for the analysis; in this study, respectively, trunk, head, eyebrows, eyes and mouth. For each modality, the analysis is written in the first three columns. In col 1. you write a description of the movements performed by that part of the body; in col. 2, an interpretation of that movement: if you judge it a noncommunicative action, you simply write its goal, while if it seems to have expressive or communicative goals, you write down the information it aims to provide, that is, its meaning. An action is defined as communicative if it has a conscious, unconscious or tacit goal, or a goal determined by social or biological function, of having some other Agent assume some belief [5]. An action is defined an expressive action if is: a) a communicative action; b) providing information about the Sender s mental states (not about the external world); and c) whose Sender is not aware either of his goal of communicating or of the signal produced or of both. An example of communicative action occurs, for esample, when the Pianist, during the rehearsal, nods to the orchestra players to praise them for how it is playing (Marcella Crudeli is a famous pianist, while the orchestra is one of young and not yet famous players, so she has an encouraging attitude towards them). Instead, a case we classify as expressive action in one in which she frowns while concentrating before starting: a non deliberate signal of a cognitive state of attention. Finally, when she moves her head up and down rhythmically to accompany the rhythm of the music she is playing, this counts as a non-communicative action. In col. 3 you classify the goal or meaning of col.2 in terms of a typology of the movements performed (see Table 2 below). The analysis of each movement provided in col. 2 and 3 can be written in two lines, since you can write, when necessary, both the literal meaning or the apparent goal of each movement, and its indirect meaning, or its superordinate goal. For example, this is the analysis of head behaviour in Table 1, row 4. At bar 100, the pianist shakes her head slowly (col. 1), like if saying no no, a performative of rejection; but what she rejects, as confirmed by the context, namely the meaning I am sad (line 5, col.2) provided by the raised inner eyebrows (line 5, col.1), lets you infer that she feels sorrow; sorrow is well something one rejects. So, in ISBN 88-7395-155-4 2006 ICMPC 1046

line 4, col.2, the literal meaning is I reject, but the indirect meaning is I feel sorrow. The analysis of these two meanings in line 4., col. 3, classifies, respectively, the literal meaning as CP (a Communicative act, namely a Performative), and the indirect meaning as a MeaE (Meaning oriented emotion): the rejection the pianist is communicating is a way to express her sorrow, which thus she induces in the sounds she is playing. On the right of the first three columns, new bars can be written down and analysed in the same way. Moreover, one can add other lines underneath to provide a parallel analysis of the multimodal behaviour of another pianist at the same bar, or of the same pianist in rehearsal vs. concert. This allows to compare the different behaviours and to state if the same body behaviour systematically co-occurs with the same hand movements (the same notes to play) across performances and across pianists. BODY MOVEMENTS DURING PIANO PLAYING A first analysis over Crudeli s data allowed to state a typology of the possible functions of the pianist s body movements, to be used in the classification of col. 3. All in all, the types of actions or meanings provided by the pianist s trunk, head and face are the following (Table 2): Communicative acts. Sometimes the pianist, even during performance, communicates something to specific people, and since music prevents verbal communication, this is done through gaze, head movements or facial expression. In some cases only the performative of a communicative act is conveyed (CP), and the propositional content is to be understood from context: for example Marcella Crudeli, with a head nod, eyebrows raised and a smile, praises the orchestra. But sometimes also the propositional content of the communicative act is explicitly conveyed by trunk or face (CPc): with fast movements she incites the orchestra to go faster. Another type of communicative action is one giving information on the musical structure of the passage being played. For example, M.Crudeli raises her trunkwhen the musical phrase is finished, or leans on the right when she starsts a new theme. This reminds us of the metadiscursive function of trunk movements in speech (Condon & Ogston, 1971): also in verbal discourse a change of posture communicates a topic change. Table 2. Goals and meanings of the Pianist s body movements Communicate Performative CP Performative + Content CPc Musical Structure CMs Express or E/C Process Emotions PE communicate E/C Outcome Emotions OE Emotional E/C Meaning Emotions MeaE Processes E Movement Emotions MoE Express Cognitive Express Attention EA Processes Express Perception EP Express Memory EM Help Motor Help Melody HM Action Help Rhythm HR Help Harmony HH Help Manner of Help Tempo The Movement Help Timbre Hti Help Intensity HI Communication and expression of emotional processes. Given our definition of expression, information about the music can only be communicated, while information about the pianist s emotions or cognitive processes can be either communicated (if the pianist deliberately and consciously signals it) or expressed (if it leaks instinctively or without awareness). Process emotions (PE) and outcome emotions (OE), since they are felt, can be either expressed or communicated. Among enacted emotions, those meaning oriented (MeaE) will presumably be communicated, if they are consciously simulated, while movement oriented emotions (MoE) may be simply expressed: the pianist might be conscious of the goal of making that movement, not of simulating the emotion in order to make that movement. An example of Process emotion in Crudeli s data is her shaking her head rhythmically and smiling, showing mirth about her own playing; she displays an Outcome emotion of pride and satisfaction about how she played by raising external eyebrows and smiling. By raising inner eyebrows she expresses a Meaning oriented enacted emotion of sorrow, and she frowns enacting anger a Movement oriented emotion when playing loud notes. Expression of cognitive processes. The pianist can express her mental states of attention (EA), perception (EP), and memory (EM). Lowered eyebrows show attention and concentration; head down with ear close to the tunes shows the pianist is listening to the sounds played; a signal of memory processes (presumably typical of novice pianists, not of professionals in concert, might be eyes up while trying to remember. Accompaniment of hand motor actions. In some cases head, face and trunk movements do not have a communicative or expressive import: they are simply movements. But since they are often synchronous and analogous to the movements of the pianist s hands, they even seem to help them to move better. An objection to this claim is that, were this so, pianists that move their body more would be better pianists. Which is not. Actually, different schools of piano technique exist, some encouraging body expression, others composure; but also people of different personalities may be more or less keen to body motion, and yet be equally good pianists. However, saying that one can be a good pianist even without so much moving his body does not exclude that moving ISBN 88-7395-155-4 2006 ICMPC 1047

head or trunk at the same time as hands does help. After all, if work songs helped Negroes to work better, why should body movements not help hand movements? We classify a body movement as helping hand motor actions when trunk, head or face movements are analogous to the movements of the hands. Within the movements that directly help the hand motor actions, we can distinguish those that help the actions aimed, respectively, at producing melody, rhythm, and harmony. Movements that help melody (HM) exploit a spatial analogy: direction and intervals of trunk head and face movements are analogous to those of hands. For example in Crudeli s data, at bar 209, the pianist s right hand makes a circular movement to play the tunes E, #G #F M; at the same time, her head makes an analogous circular movement. Or again, in playing an acciaccatura from an E to another E an octave higher, as she moves her right little finger up, at the same time she moves her external right eyebrow up! In movements that help rhythm (HR), their rhythmical structure is analogous to that of hand movements: at bars 210 211, head shakes rhythmically every two quartines. In movements helping harmony (HH) an analogy holds with the visual image of harmony relations: the pianist, for example, plays a piece around the tonic and then, while playing it again a third lower, lowers her eyebrows. Within the movements helping the manner of movement, some help tempo, others timbre, others intensity. Movements helping tempo (HTe) use the same tempo of the hands; like in the fast head shaking that accompanies fast hand movements; to help timbre (HTi), the tension of body movements is analogous to that of hands: like in raising the external parts of eyebrows which helps a tense timbre; to help intensity (HI), the movement or position of trunk, head or face allows or favors the intensity of hand movements; like in raising whole eyebrows, that helps to play a delicate sound. The annotation scheme presented here was applied in two studies that I illustrate in the following Sections. DIFFERENT MUSIC, DIFFERENT BODY? The first study is an analysis of body movements aimed at discovering possible similarities and differences in different situations and different passages of piano performance. Thirty fragments were singled out from Crudeli s data to test if the categories found systematically recur across music passages. More specifically, each fragment from the rehearsal was compared to the corresponding fragment (the same bar of the Mozart piece) during concert performance; moreover, when some music passages recurred twice or more in the whole movement, many bars after, the first fragment was compared with its repeat; and both the first occurrence and the repeat were analysed in concert and rehearsal. Let us compare two fragments: one is taken from the main theme of the first movement, bars 67-70, lasting around 6 seconds; the other from the beginning of the second movement, bars 3-11, around 40 seconds. In the main theme of the first movement the fragment of bars 67-70 was compared with its repeat, at bars 206-209 (Table 3). The body movements are, respectively, 5 and 7 in the rehearsal and concert in bars 67-70, and 8 and 6 in bars 206-209. In the rehearsal of the main theme the only categories represented were Help Melody and Help Rhythm (both with 1 body movement, 20% of the 5 found), and Help Intensity, with 3 movements (60%). In the corresponding bars during the concert, Help Rhythm counts 1 movement (14% out of the 7), Help Intensity 2 (28%), but also Outcome Emotion (1, 14%) and Meaning Emotion (3, 42%) are represented. Also in the repeat, during rehearsal (bars 206-209), the only categories are Help Melody (4, 50%), Rhythm (2, 25%) and Intensity (2, 25%); in concert Help Melody counts 1 movement (17%), Rhythm 2 (33%) and Timbre 1 (17%), and 2 (33%) Express Attention appear. As to the second Movement, Table 4 shows the result of the comparison between Rehearsal and Concert performance. HM and HR are respectively represented 5 and 2 times (22% and 9%) in rehearsal and concert (13% and 5%), while HI occurs only once in the concert. 4 Expressions of Attention occur in rehearsal and 1 in concert (17% and 3%). Communication of Musical Structure appears here, whereas it was not present in the first movement: 3 times (13%) in rehearsal and 1 (3%) in concert. Meaning emotion movements appear 9 times (39%) in rehearsal and 29 (74%) in concert. The comparison of Table 3 and 4 appears quite revealing. A first interesting commonality is that both show more Table 3. Main Theme of the First Movement Main theme 67-70 206-209 R C R C n. % n. % n. % n. % OE 1 14 MeaE 3 42 EA 2 33 HM 1 20 4 50 1 17 HR 1 20 1 14 2 25 2 33 Hti 1 17 HI 3 60 2 28 2 25 Tot. 5 7 8 6 Table 4. Second Movement ISBN 88-7395-155-4 2006 ICMPC 1048

2 nd Movement 3-12 R C n. % n. % C Ms 3 13 1 3 MeaE 9 39 29 74 EA 4 17 1 3 HM 5 22 5 13 HR 2 9 2 5 HI 1 3 Tot. 23 39 or less the same categories of body movements: the movements that help motor action (Help Melody, Rhythm and Intensity) and the Expressions of Attention (EA) and of Meaning Oriented Emotions (MeaE) are present in both movements; the only difference is for Outcome Emotions and Help Timbre, that appear only in the first, and Communication of Musical structure, occurring only in the second. But let us note some other differences. The second movement (Table 4) shows a higher frequence of body movements in concert versus rehearsal, 39 versus 23, and in this it differs from the first movement (Table 3), where the total of body movements is the same in the two performances (13). On the other hand, movements expressing Meaning oriented Emotions (MeaE) occur only in concert, and only in the main theme (42%), in the first movement, while they are the most frequent in general, and far more frequent in concert than in rehearsal (74% vs. 39%) in the second movement This striking difference is congruent with the idea we have of the second movement in classical music: it is a more emotional movement, one in which the pianist can give vent to all of her feelings. So here the number of body movements is a cue to the musical structure of the piece played. But further, another difference shows up from this data, in the social situation of the performance. In both the first and the second movement, Meaning oriented emotions are much more frequent in concert than in rehearsal; perhaps because in concert the pianist really wants to give the best of herself, and thus abandones herself to feel and/or express her enacted emotions in a very free fashion. Therefore, body movements are also a cue to differences in the social situation. A LEXICON OF THE PIANIST S BODY On the basis of the fragments analysed, it was also possible to single out a list of the body movements that recur during piano performance, thus constructing a first core of a lexicon of the pianist s body. Research on multimodal communication has shown how systematic and rule governed communicative behaviour is not only in the written and spoken modalities but also in bodily behaviour; to such an extent that it was proposed, and sometimes attempted, to write down lexicons of body communication systems. By lexicon I mean a list of signal-meaning pairs that are codified and then steadily represented in memory in one mind and (assumed to be) shared with other minds. But in a lexicon of the body the correspondences between signals and meanings are not necessarily or always one-to-one correspondences: in any lexicon, both of a verbal language and of a bodily communication system, polysemy and synonymy may hold, that is, in some cases one and the same signal may convey more than one meaning with meanings varying according to the context and also, sometimes, (almost) the same meaning may be conveyed by two or more different signals. So the guess is if it is possible to find out a lexicon of the pianist s body, just as the lexicons described for Italian symbolic gestures (Poggi, 2002 a), touch (Poggi et al., 2004) gaze (Poggi & Pelachaud, 2002; Poggi, 2006), performative faces (Poggi & Pelachaud, 2000), and the Lexicon of the Orchestra conductor s face (Poggi, 2002 b). The analysis of the fragments above allowed to single out a list of body movements of Marcella Crudeli, distinguished for categories and modalities (Trunk, Head, Eyebrows, Eyes and Mouth), along with their respective interpretations (Table 5). This is, so far, an idiosyncratic lexicon, that is, one specific of this pianist; nonetheless, in future research it could be assessed whether these correspondences, that look quite systematic in her, could be possibly hold also for other pianists. Moreover, this is a lexicon only in part, that is, only in a metaphorical sense. Strictly speaking, a lexicon is a list of correspondences in the communicative domain: a list of signal-meaning pairs, where the meaning is some information that the Sender has the conscious, unconscious or biological goal to convey to some Addressee. While in the case of the pianist, as we saw, many categories of body movements are only aimed at accompanying hand movements, perhaps at helping them, but not at communicating anything. Yet, what remains in this list is its being a list of systematic correspondences between body movements and their respective goals of communication, expression or motor action. As one can see, in some cases the categories count different possible signals, each with its specific meaning or function, and sometimes distributed across different modalities, sometimes performed by the same part of the body. For example, Crudeli exhibits four different Help Rhythm movements, two by head and two by eyebrows. Of the two eyebrow movements, one is very specific: a single raising and lowering of the eyebrows is used to count while maintaining a key over a point, while the rhythmical raising of the eyebrows helps concomitant stress and rhythm of the hands. Also in this lexicon, just as in a verbal one, we can find polysemic and polyfunctional items, as well as synonymical ones. A case of a polysemous item is the frown, which can express both concentration (EA) and anger (MoE). Synonymical cases, in which the same function is ISBN 88-7395-155-4 2006 ICMPC 1049

served by different items, are, for example, the function of helping to play an accidental (HM), which can be performed both by raising the internal parts of the eyebrows and by squeezing eyes; and that of helping delicacy (HI), which is achieved either by moving one s trunk up or sideward, or by raising the eyebrows. Table 5. The Lexicon of the Pianist s body Type M Body Movement Goal / Meaning C Ms E Raise eyes End of musical phrase PE OE MeaE H Repeated head shakes Merry, carried by rhythm E Look right and raise eyes Serenity M Mouth straight Boredom External eyebrows raised Proud M Smile Satisfied H Shakes head merrily Free and easy Int.eyebrows raised Ext.eyebrows raised Whole eyebrows raised Sorrow Artful and merry Free and easy E Eyes closed Inspired MoE Frown Angry EA HM HR T Frown Raise ext.eyebrows Whole eyebrows raised Wave trunk and shoulders Moves trunk to left backward Concentration Attention Attention Precision Help arpeggio Help play low keys H Circular movement Help hands go and back E H External eyebrows raised Internal eyebrows raised Squeezes and then opens eyes Shakes head rhythmically every two quartines Shakes head at each stroke Raises eyebrows rhythmically Lowers and then raises eyebrows once Help high notes Help accidentals Help accidentals Help rhythm Help rhythm Help rhythm Help count during the point HH Lowers eyebrows Help repeat a third lower HTe H Shakes head very fast Help Tempo HTi HI Raises external eyebrows Help Timbre M Smart and cunning smile Help Timbre T Moves trunk to up left Help Delicacy H Shakes head at each stroke Help Forte Frown Raises whole eyebrows Help Forte Help Delicacy ISBN 88-7395-155-4 2006 ICMPC 1050

CONCLUSION Piano performance is a complex activity in which motor, cognitive and emotional processes occur, and the pianist s body can reveal much of these processes. But this is possible only if we can find a systematic correspondence between specific body movements and their functions. In order to this this paper presented an annotation scheme through which each movement of the pianist s trunk, head, eyebrows, eyes and mouth can be described, interpreted and classified. The scheme was applied to the analysis of fragments from Mozart concert K488, and it allowed, on the one hand, to show the differences in the pianist s body movement depending both on the social situation and on the piece musical structure, and on the other hand to write down a list of the pianist s movements with their interpretation, thus providing a first lexicon of the pianist s body. Research through the annotation scheme presented can go on in various directions. First by investigating if and how much the lexicon of the pianist s body is really shared by different performers; then by comparing different pianists playing the same pieces, or the same pianists playing different music, thus possibly discovering differences in the pattern of body movements that can distiguish experts from novices, or pupils of different school, or different musical styles. This will shed light on the relationship between body and mind in musical performance. ACKNOWLEDGMENTS Participation in the ICMPC9 was supported by HUMAINE (European Project IST- 507422). I am indebted to M a. Marcella Crudeli and to M o. Fabrizio Santi and the orchestra Res Musica for allowing me to videorecord and analyse their performance. REFERENCES Boyes Braem, P., & Braem, T. (2004). Expressive gestures used by classical orchestra conductors. In C.Mueller and R.Posner (Eds.), Proceedings of the Symposium on The Semantics and Pragmatics of everyday Gestures (pp.127-143). Berlin: Weidler. Caterina, R., Bonfiglioli, L., Baroni M., & Addessi, A.R. (2004). Mimic Expression and piano performance. Proceedings of the 8th International Conference on Music Perception and Cognitition, Evanston, Ill., August 04, 2004, cd edition. Duranti, A., & Burrell, K. (2006) Jazz improvisation: a search for hidden harmony and a unique self. Ricerche di Psicologia 3, 27, 71-101. Haviland, J. (2005). Batons in the hands, and batons in the air: Creating a gestural environment. 9 th International Pragmatics Conference, Riva del Garda, 10-15 July 2005. Poggi, I. (2002 a). Symbolic Gestures. The case of the Italian Gestionary. Gesture, 2, 1, 71-98. Poggi, I. (2002 b). The lexicon of the conductor's face. In P.McKevitt, S. O Nuallàin, & C. Mulvihill (Eds.) Language,Vision, and Music. Selected papers from the 8 th International Workshop on the Cognitive Science of Natural Language Processing (pp. 271-284). Amsterdam: John Benjamins. Poggi, I. (2006). Le parole del corpo. Roma: Carocci. Poggi, I., Cirella, F., Zollo, A., & Agostini, A. (2004). The communicative system of touch. Lexicon, alphabet and norms of use. In A.Camurri & G.Volpe (Eds.) Gesturebased communication in Human-Computer Interaction. Proceedings of the 5 th International Gesture Workshop, GW 2003, Genova, Italy, April 15-17, 2003 (pp.77-89). Berlin: Springer. Poggi, I., & Pelachaud, C. (2000 b). Performative facial expressions in Animated Faces. In J.Cassell, J.Sullivan, S.Prevost, & E.Churchill (Eds.), Embodied Conversational Agents (pp.155-188). Cambridge, MA: The MIT Press. Poggi, I., & Pelachaud, C. (2002). Signals and meanings of gaze in Animated Faces. In P.McKevitt, S. O Nuallàin, & C.Mulvihill (Eds.), Language,Vision, and Music. Selected papers from the 8 th International Workshop on the Cognitive Science of Natural Language Processing, Galway, 1999 (pp.133-144). Amsterdam: John Benjamins. Streeck, J., & Oshima, S. (2005). Hip-hop gestures. The conduct of M.C. Interacting Bodies Conference, Lyon, 15-18 June, 2005. Wallbott., H.G. (1998). Bodily expression of emotion. European Journal of Social Psychology, 28, 879 896. Condon, W.S., & Ogston, W.D. (1971). Speech and body motion synchrony of the speaker-hearer, in the perception of language. In D.H.Horton & J.J.Jenkins (Eds.), The perception of language (pp.150-184). New York: Academic Press. ISBN 88-7395-155-4 2006 ICMPC 1051