Looking Beyond the Score: The Musical Role of Percussionists’ Ancillary Gestures

[1] The written score frequently serves as the basis for our efforts to understand music’s structure, content and meaning. Although scores capture many important aspects of a composition, certain elements are difficult to analyze and impossible to understand without accounting for the way in which the acoustic signal (represented abstractly by the score) is perceived. For example, although they share identical intervallic relationships, our perception of the two triads in Example 1 differs markedly. The higher one sounds “sweet” and “pure” whereas the lower sounds “rough” and “muddy.” This appreciably different listening experience is attributable not to the structure of the triads themselves, but rather the structure of the perceptual system—in particular the differential sizes of critical bands in low vs. high frequency ranges.⁽¹⁾ This is but one example of how information beyond the score (in this case the listening apparatus used to experience sound) shapes music listening. Other factors are subtle and more difficult to recognize, yet also play a crucial role. For example, the next section summarizes ways in which a performer’s body movements routinely shape the listening experience. Given mounting evidence documenting the musical importance of gesture, we believe that efforts to understand music can and will benefit from exploring this perspective.

[2] Our goal in this paper is to build a case that (1) gestures used by performers play a meaningful role in music perception even though they are not represented in the score and (in some cases) do not have acoustic consequences, (2) new research tools coupled with the traditional techniques and methodology of music perception allow for precise analysis of these gestures (with a degree of rigor traditionally reserved for “the notes on the page”), and (3) much as the thoughtful analysis of a score can be insightful in understanding a musical composition, analysis of a performer’s body movements can also be informative. Although much of this article will focus on one particular class of gestures (those used by percussionists), this issue is broadly relevant for all musicians (in addition to artists, scholars, and critics). Therefore before delving into a focused review, we will discuss the relationship between gesture and music broadly, as well as commonalities between the use of gesture in music and dance. The second section will then focus on one particular type of ancillary gesture used to overcome acoustic limitations of the marimba. The third and fourth sections will explore the perceptual basis for this phenomenon, and the fifth will review efforts to analyze and deconstruct the gestures themselves. The final section will discuss future directions for this line of work, as well as general implications of these approaches for musical research.

1. The Use of Gesture in Music

[3] Interest in the role of gestures in music is a vast topic, one that has seen significant research attention in recent years. The term “gesture” itself can be interpreted in multiple ways. At a basic level, it may refer either to a particular segment of music (see Hatten 2004) or to a physical motion used by performers. The focus of this paper is on the latter. Wanderley and colleagues (2005) distinguish further between two classes of physical gestures—effective gestures that are required for sound production, and ancillary gestures, which are not necessary for the creation of sound. Ancillary gestures have been previously referred to as either expressive movements (Davidson 1993) or body language (Dahl and Friberg 2007). These movements are often thought to be of secondary importance, given that they lack significant acoustic consequences and their production is rarely consciously/actively regulated by performers. However, there is significant stability in these gestures across performances by a single musician (Wanderley 2002); and as discussed in the next subsection, they are capable of systematically affecting an audiences’ listening experience.

1.1 Ancillary Gestures in Music

[4] Ancillary gestures can play a profound role in music listening—despite their lack of acoustic consequences. For example, judgments of tension and phrasing in the second of Stravinsky’s Three Pieces for Clarinet Solo differed significantly when participants watched the performer rather than listened to the audio alone (Vines et al. 2006). This phenomenon is not limited to clarinetists—emotions including happiness, sadness and anger can be readily communicated through gestures on a number of instruments (Dahl and Friberg 2007). In fact, in some cases these distinctions may be more clearly discerned through a visual, rather than an auditory presentation (Davidson 1993). This significant role of vision carries important implications for the evaluation of performers—seeing the body movements of marimbists playing in an expressive style can affect ratings of audience interest (Broughton and Stevens 2009).

[5] The role of visual information is limited not only to judgments of high-level characteristics such as musical expression or performance quality, but also extends to low-level characteristics such as judgments of pitch intervals (Thompson et al. 2005), pitch accuracy (Gillespie 1997), and loudness (Rosenblum and Fowler 1991). For example, seeing a cellist’s bowing and plucking motions affects perceptual ratings of the concurrent note’s timbre—i.e., bowed vs. plucked (Saldaña and Rosenblum 1993). Therefore the visual information associated with a performer’s physical gestures plays an important role in shaping the musical experience on several levels (for a comprehensive review see Schutz 2008).

1.2 Ancillary Gestures as Music

[6] The twentieth century has seen the role of gesture in music reach new heights—particularly in music for percussion. For example, in the marimba solo Six Elegies Dancing (1987), composer Jennifer Stasack gives elaborate instructions on the motions to be used while performing—many of which have no acoustic consequences. The preponderance of compositions emphasizing gestures have led to a sub-genre of “theatrical percussion” capitalizing on the tight relationship between gestures, music for percussion, and perception (due to the large amount of physical motion required to play such instruments, music for percussion is a particularly fertile ground for exploring such connections).⁽²⁾

[7] Although it is gaining popularity in new music, this focus on gesture is far from “new.” John Cage uses gestures to great effect in a number of compositions such as Living Room Music (1976), which combines elements of percussion and theatre. In this composition, performers are called upon to play rhythms on “found objects” such as cups, bowls, books, and other items commonly situated in a living room. The variety of creative realizations of this score demonstrates the integral role of body movements, as performers frequently add gestures for reasons as much theatric as acoustic. The use of ancillary gestures as music raises interesting questions about links with another form of expression built upon the use of movement over time—dance.

1.3 Ancillary Gestures as Dance

[8] Like ancillary gestures in music, dance frequently occurs concurrently with acoustic information. In fact, some have gone so far as to define dance as “human movement that is formalized...to the accompaniment of music or other rhythmic sounds...” (Van Camp 1981). In a sense, musicians “dance” when performing, given that their ancillary movements accompany the music without affecting its acoustic characteristics. Therefore, one way of viewing dance is essentially as a special case of ancillary gestures in that it is a series of movements accompanying music not required for sound production.

[9] Dance choreography frequently reflects musical structure as dancers’ movements are generally designed to accompany and interact with concurrent musical events. Consequently, ratings of section breaks, tension, and emotion when viewing ballet show a strong similarity for the separate music and dance components (Krumhansl and Schenck 1997). Similarly, observers consistently rate dance as more realistic, creative and natural when the performance is generated to match the relative changes in musical features (Kim et al. 2009). These findings illustrate that movement can be used to communicate structural features of a musical composition, with particular movements conveying specific musical characteristics.

[10] Dance movements made in response to music listening (i.e., “music-induced movement”) not only react to the low-level temporal structure of music, but also reflect the rich hierarchy of temporal information by which it is organized. For example, when asked to move freely while listening, movement of the extremities tends to synchronize with faster metric levels whereas movement of the torso tends to synchronize with slower metric levels (Toiviainen et al. 2010). Therefore, analysis of dance in this context actually informs our understanding of musical structure as it lends insight into the way timing information is hierarchically organized in the minds of listeners.

[11] The relationship between the temporal structure of music and dance is so prominent that some have hypothesized they may have originated as a single system of communication (Hagen and Bryant 2003). Although we may never fully understand the evolutionary development of these domains, broad similarities between the two are consistent with the possibility of a common origin. In fact in many non-Western cultures, music and dance are intertwined into a single, multimodal experience. Although the two often remain distinct within the Western classical tradition, they are presented in tandem in artistic contexts such as ballet, opera and musical theatre. Additionally, the two frequently co-occur in popular music concerts and music videos, which often include elaborately staged sequences of movement in conjunction with the sound.

2. Ancillary Gestures as Performance Tools

[12] In addition to reflecting large-scale aspects of structure, ancillary gestures can also be useful in overcoming acoustic limitations of certain musical instruments. In other words, they can be used to accomplish perceptually that which is impossible acoustically. This is illustrated in a “musical illusion” resolving a long-standing debate amongst percussionists (Schutz 2009) concerning the relationship between the length of the physical gesture used to strike a marimba and the duration of the consequent note.

[13] Longtime New York Philharmonic percussionist Elden “Buster” Bailey observed that, “[When] sharp wrist motions are used the only possible results can be sounds of a staccato nature.... [When] smoother, relaxed wrist motions are used, the player will then be able to feel and project a smoother, more legato-like style” (1963). Others, such as Leigh Howard Stevens, are adamant that gesture length is irrelevant, arguing it has “no more to do with [the] duration of bar ring than the sound of a car crashing is dependent on how long a road trip was taken before the accident” (personal communication, 2004). On the surface, there appears to be merit to each of these competing points of view. A longer swing of the bat intuitively sends a ball farther. However, from the physicist’s perspective, motion after impact is not directly relevant to the acoustic consequences of the preceding event.

2.1 Gestures are Acoustically Inconsequential, but Perceptually Useful

[14] To explore this issue, renowned marimbist Michael Burritt (now Professor of Percussion at the Eastman School of Music) performed a series of “long” and “short” notes in a recital hall on a professional-grade marimba (see samples of these gestures in Example 2). An analysis of the acoustic information produced by these long and short gestures found no meaningful distinction between the two sounds. Therefore, his long and short gestures failed to create notes that were acoustically distinguishable. However, the following experimental research demonstrates that the failure of this gesture acoustically does not necessarily prohibit its success perceptually.