Musicians\' expressive bodily movements can influence observers\' perception of performance. Furthermore, individual differences in observers\' music and motor expertise can shape how they perceive and respond to music performance. However, few studies have investigated the bodily movements that different observers of music performance perceive as expressive, in order to understand how they might relate to the music being produced, and the particular instrument type. In this paper, we focus on marimba performance through two case studies-one solo and one collaborative context. This study aims to investigate the existence of a core repertoire of marimba performance expressive bodily movements, identify key music-related features associated with the core repertoire, and explore how observers\' perception of expressive bodily movements might vary according to individual differences in their music and motor expertise. Of the six professional musicians who observed and analyzed the marimba performances, three were percussionists and experienced marimba players. Following training, observers implemented the Laban effort-shape movement analysis system to analyze marimba players\' bodily movements that they perceived as expressive in audio-visual recordings of performance. Observations that were agreed by all participants as being the same type of action at the same location in the performance recording were examined in each case study, then across the two studies. A small repertoire of bodily movements emerged that the observers perceived as being expressive. Movements were primarily allied to elements of the music structure, technique, and expressive interpretation, however, these elements appeared to be interactive. A type of body sway movement and more localized sound generating actions were perceived as expressive. These movements co-occurred and also appeared separately. Individual participant data revealed slightly more variety in the types and locations of actions observed, with judges revealing preferences for observing particular types of expressive bodily movements. The particular expressive bodily movements that are produced and perceived in marimba performance appear to be shaped by music-related and sound generating features, musical context, and observer music and motor expertise. With an understanding of bodily movements that are generated and perceived as expressive, embodied music performance training programs might be developed to enhance expressive performer-audience communication.

Self-reflective performance review and expert evaluation are features of Western music performance practice. While music is usually the focus, visual information provided by performing musicians\' expressive bodily behaviors communicates expressiveness to musically trained and untrained observers. Yet, within a seemingly homogenous group, such as one of musically trained individuals, diversity of experience exists. Individual differences potentially affect perception of the subtleties of expressive performance, and performers\' effective communication of their expressive intentions. This study aimed to compare self- and other expert musicians\' perception of expressive bodily behaviors observed in marimba performance. We hypothesized that analyses of expressive bodily behaviors differ between expert musicians according to their specialist motor expertise and familiarity with the music. Two professional percussionists and experienced marimba players, and one professional classical singer took part in the study. Participants independently conducted Laban effort-shape analysis - proposing that intentions manifest in bodily activity are understood through shared embodied processes - of a marimbists\' expressive bodily behaviors in an audio-visual performance recording. For one percussionist, this was a self-reflective analysis. The work was unfamiliar to the other percussionist and singer. Perception of the performer\'s expressive bodily behaviors appeared to differ according to participants\' individual instrumental or vocal motor expertise, and familiarity with the music. Furthermore, individual type of motor experience appeared to direct participants\' attention in approaching the analyses. Findings support forward and inverse perception-action models, and embodied cognitive theory. Implications offer scientific rigor and artistic interest for how performance practitioners can reflectively analyze performance to improve expressive communication.