Playing with Constraints: Stylistic Variation with a Simple Electronic Instrument

The current ubiquity of embedded sensing and computing technologies, in concert with the facility of real-time digital signal processing, has fostered a more rapid pace of invention of new musical instruments than we have seen since electricity was tamed for widespread use. The state of interactive music performance has progressed significantly since John Cage, in 1937, lamented, “most inventors of electrical musical instruments have attempted to imitate eighteenth- and nineteenth- century instruments, just as early automobile designers copied the carriage” (Cage 1961, p. 3). This progress is, in large part, thanks to Cage’s own work and legacy. Yet, although the “desire to imitate the past rather than construct the future” (Cage 1961, p. 4) has diminished in some quarters, contemporary authors have called attention to the changing nature of creation and perception of performance with computer-based instruments in the absence of a strong correspondence to the acoustic instruments with which we are familiar. There remains a widespread desire to see virtuosic performance, in which there is not only evident mastery of the instrument, but also an individualized contribution from the performer (Schloss and Jaffe 1993; Schloss 2003; Dobrian and Koppelman 2006). In many cases, however, it is not obvious how such virtuosity can be achieved in an interaction with an electronic device or computer system, which may have programmed, random, or otherwise non-deterministic behavior. In computer music, it is frequently difficult to distinguish between the performer’s and computer’s contributions, or to know how the performance might differ in the hands of another person.

In this article, we address the individuation of performance with electronic instruments—what we refer to as style. We first derive a working concept of style as distinct from structure in an activity, which we propose as a useful framework for considering virtuosity and individuality in interactions with technology, including musical ones. Drawing on recent work that has explored the role of constraint—a limitation in an interface’s allowed, suggested, or perceived activities—in facilitating personally expressive actions and interactions (Norman 1999; Boden 2004; Hornecker 2005; Stokes 2006; Magnusson 2010), we posit an alliance between constraint and the development of style. The bulk of the article then describes a qualitative study that explores the emergence of personal performance styles in experienced performers with a novel, constrained electronic musical instrument.

The study aimed to embody aspects of a realistic situation within the new interfaces for musical expression (NIME) community: A performer must determine how to perform with a new instrument for which there is no established performance practice and no instruction manual. In particular, we sought to examine such a circumstance where the instrument in question offers a limited number of obvious physical controls. Apart from a small number of commercial controllers, it is a rarity in the NIME community for a substantial number of performers to take up a new instrument at the same time. Thus, in order to investigate individual variations, multiple copies of a new instrument were simultaneously distributed to performers who were asked to develop a performance in isolation from one another. The study was not a positivistic experiment that treated constraint as a control and style as a dependent variable; rather, drawing on phenomenological and “third paradigm” research in human–computer interaction (Dourish 2001; Bødker 2006; Harrison, Tatar, and Sengers 2007), it was a qualitative investigation of approaches to performance with this particular constrained and novel device.

Through observations of the performances at the end of the study period, we characterize performers’ approaches and outcomes—for example, whether they played the device in the most obvious way possible, or discovered and invented different ways of playing it. We describe attributes of the performances by which the participants individuated themselves, and show that in some cases these arose as an explicit response or reaction to perceived constraint. This analysis is complemented with detailed interviews and questionnaires in order to discover how situational, experiential, and other factors interacted with the design to give rise to the performances we observed, with particular attention paid to notions of...