The concept of musicking is based on the notion that music does not exist as an object (Small, 1998). Putting to one side the (doubtlessly important) social and cultural aspects of musicking, it is essential to consider the physical and perceptual aspects of sound and music. Although there are many traditions for which notations and graphical representations are a central aspect of mu- sical understanding (Small, 1998, Thiebaut, 2010), music is fundamentally an aural experience. The aural nature of music represents a key distinction between CDMI and research in CSCW, which typically investigates activities that are primarily visual and spatial, such as producing tex- tual documents, drawings and diagrams. Compared to the large amount of research concerned with support for multi-user interaction with visual artefacts such as textual documents, there are relatively few examples of research focused on the design of systems to support the collabora- tive production of auditory media. This section considers the properties of sound as a medium, compares this to other domains for collaboration, and identifies the challenges sound poses for designers of collaborative musical systems. Although this section has been written with CDMI in mind, it has implications beyond the domain of music, and may be especially relevant for designers of systems to support other forms of auditory collaboration.
3.2.1 Temporal Qualities of Sound
Sound is intrinsically time-based; it is important to remember that all sounds, no matter how short, are perceived as a function of time (Roads, 2004). Sound is also ephemeral, and leaves no persistent trace. This means that unless a recording is made, there is no permanent record of a sonic event. This is quite different to visual media such as drawing and written documents, where symbols and marks remains in place after they have been created. Systems such as written notation have been developed to account for this. Written notation instructs a musician in the re-creation of a particular sequence of sounds, whilst time-line based sequencers allow sounds to be organised in time using graphical representations. In real-time musical interaction, however, there is no way to revise what has been previously contributed.
Human memory does however provide a means by which musicians can recall ideas between rehearsal and writing sessions Nabavian (2009), and memory also provides a mechanism for musicians to remember, recall and modify the musical contributions made by other members of an ensemble during improvisation Sawyer (2003). However, these instances of recall are unlike
the marks which can be made on a piece of paper or the words typed into an electronic document, as they are reproductions of past event, rather than persistent traces or records.
The time-based nature of musical interaction means anticipation and predictions of future events are important. This has been emphasised especially in the case of improvised styles such as jazz (Sawyer, 2003). In cultures of musical improvisation such as jazz, shared knowledge of musical styles can facilitate the predication of musical developments such as chord changes (Sawyer, 2003, Small, 1998). These predications may also be informed by musicians’ familiarity with the playing styles of their collaborators. However, the prediction of future events is excluded in the Workspace Awareness framework (Gutwin and Greenberg, 2002), as it is argued that de- signers are unlikely to be able to support these aspects of collaboration. Similarly, (Renaud et al., 2007) stated that the ability for musicians to anticipate each-other’s actions is difficult to support in a networked environment.
3.2.2 Spatial Qualities of Sound
Working with sound is unlike working with visual materials, as sound is pervasive and ephemeral and has entirely different spatial qualities to visual media. Unlike visual media such as drawing, sound does not linger or leave a persistent spatial impression. Sound is not directly malleable, and computer based interaction with sound is usually via a graphical or tactile interface. Where graphical interfaces are employed within software as a means of controlling the generation and manipulations of sounds, the interaction can be regarded as multi-modal. This is because the means of interaction operates in a separate modality to the output. By means of comparison, in a collaborative diagram editing context users are often able to directly manipulate a visual representation of the artefact they are producing.
The theory of communicative grounding (Clark and Brennan, 1991) notes that ‘indicative gestures’ such as looking, pointing and touching are important means by which interlocutors ar- rive at an understanding that they are both referring to the same object. The spatial properties of visual media have been observed to contribute to the management of tasks and conversation (Healey and Peters, 2007), and are an important aspect of establishing joint attention in collabo- rative domains such as document editing (Cohen, 2008). However, the ephemeral nature of sound means groups of people are unable to point at, gesture or orient around a sound in the same way as they can when working with visual artefacts.
design of collaborative interface features such as awareness mechanisms which rely on conveying information about the identity, location, source or ownership of a sound or musical contribution. For example telepointers have been used in collaborative environments to show where another user’s mouse currently resides as an indicator of their current (and possibly previous) actions and attention within a graphical document (Gutwin and Penner, 2002). However such visual information may not provide users with as much information about what another user is currently attending to in a collaborative music environment due to the disconnection between graphical representation and fleeting sonic output.
Coughlan and Johnson (2007) stated that musicians use many forms of representation to convey ideas and refer to aspects of the music, including playing their instruments, vocalising, gesture and verbal communication. The studies in this thesis also show that people use verbal imitation, descriptions, and reference to graphical representations to draw attention to aspects of the shared music. This illustrates the idea that a musical gesture is both an act of commu- nication and an aesthetic product in its own right (Bryan-Kinns and Hamilton, 2009, Gratier, 2008). Coughlan and Johnson (2007) argued that an understanding of how musicians represent and convey ideas is crucial to the design of new musical interfaces and software environments, while Nabavian (2009) noted that musicians often successfully collaborate while holding entirely different cognitive representations of the music they are co-creating.
3.2.3 Disruption
One property of the phenomenon of auditory masking is for loud sounds to conceal quiet sounds (Bregman, 1990). This means the generation of sound may cause verbal communication to be problematic during musical interaction as excessive sound volume may interfere with normal conversation. Equipment such as headphones may also cause verbal communication to be dif- ficult. In addition to sound level in the environment, individuals may have difficulty speaking whilst concentrating on the activities associated with playing their instruments or controlling in- terfaces, and the positioning of individuals may also restrict conversation. Where technologies such as computers are involved musicians may also be tied to a specific location, and furniture, stands and other items may make it difficult for musicians to move around.