2. Inside and Outside Performance

Although recent scholarship in social aesthetics and musical mediation has drawn attention to the social dimensions of musical activity across practices, writers such as Georgina Born have maintained that ‘there is something singular about improvisation in that improvised performances are marked by and enable degrees of openness, mutuality and collaboration that are heightened and intensified when compared to scored works, and that necessitate participants’ real-time co-creation and negotiation of social-and-musical relationships’ (Born Reference Born, Born, Lewis and Straw2017: 41). Yet as Born herself points out elsewhere, the microsocialities of performance are only one of several avenues by which the social enters music (Born Reference Born, Born, Lewis and Straw2017: 34 and passim).

A characteristic example can be found in the preparation process for a 2003 performance of Oliveros’s Four Meditations for Orchestra, as recounted by Barbara Rose Lange. The performance was a collaboration between two groups of musicians from different cities with vastly different experience and training who expressed conflicting ideas about how to perform the piece (Lange Reference Lange2008: 44). The Four Meditations are notated using a combination of verbal instructions and graphic indications, allowing for a wide range of possible interpretations. As Lange explains, the interpretive differences between the two groups nearly led to the cancellation of the performance until Oliveros herself eventually intervened on an electronic message board the two groups were using to communicate (Lange Reference Lange2008: 55). Oliveros did not, however, simply explain the vague language of the work to head off the conflict; instead, she allowed the discussions to continue autonomously, merely expressing her hope that the differences in approach between the two groups could be resolved.

The ambiguous language in the score, combined with the performers’ commitment to different interpretive ideals, would inevitably result in conflict on some level. But Oliveros saw this ambiguity as an opportunity for the performers to communicate and resolve their differences. Her immediate concern was not the manner in which the groups were interpreting her instructions, but rather whether they were actually collaborating. This humanistic concern is at the heart of Oliveros’s project, and goes a long way towards explaining the powerful emotional reactions people often report after participating in her performances and workshops (Alarcón and Herrema Reference Alarcón and Herrema2017). At their best, they are rewarding social experiences as well as musical ones.

As the example of the Four Meditations shows, however, these meaningful social interactions often occur outside of the performance. A further distinction can be drawn between those interactions that stem from social roles that are laid out explicitly in the score, and those that emerge implicitly. For example, in several of the Sonic Meditations, social roles are defined in the performance instructions and taken on consciously. One of the Sonic Meditations is conceived as a greeting to others, another is an exchange of energy, and in several, performers are explicitly designated as transmitters or receivers. Nonetheless, most social interactions emerge implicitly, either through the process of preparation or during the performance, as participants relate to and rely upon one another in unexpected ways to complete a task whose realisation is subject to multiple interpretations and must be, to some extent, improvised. In addition, such activities often draw upon skills, knowledge and abilities that are not typically valued or acknowledged in the context of performing art music. As the following examples illustrate, Oliveros’s use of technology can be understood as an extension and intensification of these social dynamics present throughout all her work.

3. In Memoriam: Nikola Tesla, Cosmic Engineer

In 1969, Oliveros composed a work to be the musical accompaniment to Merce Cunningham’s dance Canfield entitled In Memoriam: Nikola Tesla, Cosmic Engineer. The basic premise is a task: the musicians must work together to determine the resonant frequencies of the performance venue using a number of sound-making implements, including a bugle, a cap gun, clickers and a slide whistle. The manner in which they achieve this task, however, is improvised in real time, while the dance is in progress. Intended for performance on tour with the Cunningham company, the work is designed to be heard in different venues, resulting in unique findings with each performance. In a typical realization, the musicians walk through the space making various test sounds and listening, all while communicating with one another using microphones routed through the house PA system.

The performance materials comprise a two-page typewritten score that divides the performance into three sections, each described with characteristically open-ended verbal instructions. In the first part, Oliveros asks the performers to discuss the acoustics of the space and plan their approach to finding the resonant frequencies, enumerating over a dozen possible topics for discussion, from ‘the absorptive quality of the audience’ to ‘resonance’ to ‘the weather’ (Oliveros Reference Oliveros1969). In the second part, performers are asked to ‘search for the resonant frequency of the space’ by physically exploring it, before undertaking a dramatic aural verification of their findings in the final section. During the search, Oliveros asks the performers to ‘test the environment’, share aloud their ‘discoveries and theories’ and ‘devise simple practical experiments which prove or disprove’ these theories. Between the discussion, theories and structure of each experiment – along with the various roles to be assumed or assigned – there is much to be improvised and negotiated.

The Music for Merce box set of recordings includes a 10-minute excerpt of the second part (the ‘search’) of a 1972 performance featuring John Cage, David Tudor, Gordon Mumma and Jean Rigg, an administrator with the Merce Cunningham Dance Company (Oliveros Reference Oliveros2010b). As the dancers execute Cunningham’s choreography on stage, the musicians can be heard conversing through the PA system, directing one another in different parts of the theatre and sharing their observations while making various sounds and listening to the results. The process is inevitably marked by moments of negotiation, disagreement and uncertainty. For example, at the beginning of the excerpt, Rigg announces that she is ‘at the top of the balcony to your right’, prompting Tudor to clarify exactly whose ‘right’; at another point, Cage explains to Tudor what he means by a musical ‘tattoo’ after asking Mumma to play one on his bugle. Still later, Rigg narrates her position to Tudor as she moves across the space with the slide whistle. Tudor listens, asking clarifying questions, and eventually asks Rigg to return to the centre of the space and change the quality of the test sounds she is making.

In Don Ihde’s theory of technological mediation, communication technologies typically fall under the rubric of ‘embodiment relations’, which are characterised by a low contrast between mediated and unmediated perception and a tendency towards the transparency of the technology itself (Ihde Reference Ihde1990: 78–9). Yet the use of the house PA system throughout the performance definitively shapes the interaction between the performers, and ultimately draws attention to its own limitations. As all four musicians are conversing through the same system projected through loudspeakers into a reverberant space, they must negotiate who speaks when so that they may be understood, all while observing Oliveros’s request that ‘particularly interesting points should be broadcast into the PA system via wireless microphone’. The Music for Merce recording dramatises the dense texture of voices, test sounds and ambient ringing that the performers must parse while moving between completely different spaces in and around the theatre, often unable to see one another.

That this task is essentially social in nature is transparent both to the performers and to the audience. For example, one can hear moments when the conversation becomes unintelligible because of crosstalk, such as when Jean Rigg says ‘excuse me’ in order for her question to be heard. It is the kind of conversation that might occur during a rehearsal, now grafted into performance. The audience’s reaction to the performers speaking is audible in the recording, and knowing that the audience could hear them undoubtedly affected how the performers chose to communicate. Cage, for instance, responds to Mumma’s careful description of the tuning of the horns at his disposal by knowingly deadpanning, ‘I don’t care what key they’re in.’ By dismissing wholesale what would be a significant matter in most musical contexts – clearly riffing on his anti-establishment reputation – Cage elicits laughter from the audience.

As mentioned previously, the score stipulates that the search is preceded by a planning phase, in which all four musicians gather in the orchestra pit or another suitable space to discuss the acoustics of the hall and strategise how they will go about determining the resonant frequencies. As Oliveros writes in the score, ‘Hypothesize, criticize, air opinions, [and] compare to other spaces, real or imagined.’ The performers speak through the PA system during this section as well. The public nature of this conversation effectively dissolves the barrier between performing and preparing to perform, while also allowing the audience to hear the performers’ assessments of the space and to compare what is planned with what actually happens during the search.

In the final section, the musicians must estimate the resonant frequencies of the room based on their testing in the previous section. The musicians then set oscillators amplified through the PA system to those frequencies and gradually turn up the volume. According to Oliveros, the process is intended to emulate Nikola Tesla’s research in mechanical resonance. Tesla performed a series of experiments in which he attempted to determine the resonant frequencies of buildings and, according to some accounts, produced a minor earthquake in his New York neighbourhood in 1898. Oliveros writes that, ‘if the search for the resonant frequency has been successful, then the frequency of the generators selected by the musicians can cause the performance space to add its squeaks, groans, and other resonance phenomena to the general sounds’ (von Gunden Reference Von Gunden1983: 64).

The fact that there are clearly more efficient ways to determine the resonant frequencies of a given space, or to get a space to vibrate, underscores the extent to which the completion of the task per se is not a primary concern for Oliveros. In performing the piece, it becomes obvious that the complexity of the task laid out in the score requires careful listening and an awareness of space. Oliveros’s choice of technology obstructs or at least complicates the task, demanding closer coordination and sensitivity from the performers without explicitly asking for these things. Daniel Belgrad relates the awareness required of successful improvisers to the practice of ‘deutero-learning’, also known as second-order learning, or simply ‘learning how to learn’. As Belgrad explains, successful learning is best achieved by using methods that support the content itself. For instance, telling a student ‘don’t believe everything you hear’ would seem to contradict the fundamental lesson imparted by the phrase (Belgrad Reference Belgrad, Lewis and Piekut2016: 291–3).

In Nikola Tesla – just as in the preceding example of the Four Meditations – the qualities of awareness and sensitivity Oliveros aims to instil in the participants emerge by necessity from the social aspects of the process. Undoubtedly, the experience for the participants would have been much different had Oliveros specified a particular way of achieving the task. Even though the task might still have been achieved, the participants would not have engaged in the improvisatory social negotiations of which, for Oliveros, music is the ‘by-product’. In other words, the first-order learning – how to complete the task – would have been incentivised at the expense of second-order learning – how to learn how to complete the task collectively. The second-order learning was, in turn, mediated by technology, which shaped the way the participants interacted with one another due to the nature of the medium and the audience’s ability to listen in on the process.

4. Telematic Performance

In 1991, Oliveros conducted her first ‘large-scale musical transmission’ or telematic performance, a term used to describe live networked performances that occur over geographical distance (Oliveros Reference Oliveros2010a: 192). This first performance was conducted over a single telephone line between solo performers in six cities across the United States, augmented with a ‘video’ feed that amounted to a still image updated every five seconds. Although telephone lines are an effective medium in that they transmit signals with very little latency, they are ill-suited to multichannel signals. In this instance, while all performers were able to connect with one another as in a conference call, only the line with the loudest sound at any given moment was passed through to the other performers.

On the face of it, this is a serious limitation: each participant could never hear all of the participants at the same time. Nevertheless, Oliveros was quick to identify how the technological medium shaped the music-making, writing, ‘Since the telephone line would grab the loudest signal the improvisation was based on sensitivity to give and take’ (Oliveros Reference Oliveros2010a: 192). Ironically – or perhaps not – this ‘limitation’ actually reinforced fundamental aspects of Oliveros’s aesthetic by demanding greater awareness and sensitivity from the musicians so that all might be heard throughout the performance. The telephone system reifies the roles of transmitter and receiver described in her early text scores, autonomously switching between them based on the loudness of each performer.

This fundamental difficulty in determining whether certain technological obstacles should be conceived as aesthetic limitations persisted throughout the developmental trajectory of telematic performance over the last three decades of Oliveros’s career. Oliveros switched to internet-based telematic performances in the late 1990s, which allowed her to blend multiple audio streams simultaneously. The tradeoff, however, was the introduction of latency: initially, an eight-second audio delay. Again, this was initially confronted as a limitation, but the delay would quickly become, in her words, a ‘structural element of our improvisation’ (Oliveros Reference Oliveros2010a: 193). According to participant reports, this manifested in two ways: either using the delay to create ‘shifts in the rhythmic structure’, or an aesthetic choice to avoid synchronised cues altogether (Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe Reference Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe2009: 7–9).

From the mid-2000s, Oliveros began to use Internet2, a non-commercial network supported and maintained by research and educational institutions. Even as this new platform greatly reduced latency, participants still reported a significant learning curve. In 2007, Sarah Weaver collaborated with Oliveros in one of the first Internet2 performances as a conductor using Soundpainting, a sign language for creative improvisation. She conducted both local and remote ensembles over video with a delay of approximately one quarter of a second. Weaver writes:

Weaver’s observation emphasises that the qualities of the technological medium have repercussions for the way the music is both perceived and played. Attending to the latter by focusing on performers’ creative responses to altered perceptual conditions in turn helps to clarify aspects of existing theories of technological mediation. As Jason Robinson writes, establishing a groove across a latent network ‘challenges musicians … to improvise toward a negotiated, dialogic rhythmic relationship through expressive microrhythms’ (Robinson Reference Robinson, Siddall and Waterman2016: 108). Robinson contends that experienced improvisers can navigate network latency in a number of ways, whether by ‘[feeling] their way toward compelling rhythmic relationships’ in pulse-based music or, in music without a pulse, improvising so as to ‘effectively [obfuscate] the presence of disruptive time lags between networked sites’ (Robinson Reference Robinson, Siddall and Waterman2016: 108–9).

Certainly, as Peter-Paul Verbeek points out, following Ihde, ‘mediation always strengthens certain specific aspects of the reality perceived and weakens others’ (Verbeek Reference Verbeek and Achterhuis2001: 128). Yet in an artistic context in which subjectivity of perception is highly valued, the altered sense of time and unison downbeat described by Weaver is not necessarily a negative development, even if these changes are far removed from what a musician might experience on stage as part of an orchestra, for instance. As Robinson concludes, ‘rather than simply striving to re-create the real, I hope that telematics opens new creative strategies aimed at understanding time, the body, and dialogism’ (Robinson Reference Robinson, Siddall and Waterman2016: 109). This observation resonates strongly with an understanding of Oliveros’s use of technology as a way to facilitate new kinds of social interactions. At the same time, the fact that telematic performance can stimulate new, extended forms of listening jibes with the objectives of Oliveros’s related practice of Deep Listening (Oliveros Reference Oliveros2010c).

Telematic performance participants reported experiencing a ‘composite’ awareness in which they responded differently to information presented visually or aurally (Oliveros et al. Reference Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe2009: 7). For example, video provides access to facial expressions and other kinds of body language, which was reported by many participants to be helpful in matching the feel of the music between sites. At the same time, the video latency – often greater than that of the audio – required that participants ignore or compensate for time-specific synchronised gestures perceived visually. For instance, another participant in the 2007 performance, bassist Mark Dresser, reported that at one point he could see Weaver, as conductor, preparing an ensemble cutoff with her group. He writes, ‘recalling that our video software, iChat, had a greater latency than the audio, I followed the aural cues from Sarah rather than visual cues, resulting in a unison cutoff’ (Oliveros et al. Reference Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe2009: 15).

The very concept of cueing someone in a telematic performance, however, is not straightforward, as Weaver explains:

The question of space also pertains to the acoustic field experienced at each location, which is itself a result of collaboration that occurs during the preparation process. Bringing performers at different sites into a shared, virtual acoustic space, for example, was eventually realised through a collaboratively developed system called ViMiC, or Virtual Microphone Control. In this system, each performer is close-miked, and each location has a separate central array of microphones (Braasch, Peters, Oliveros, Van Nort and Chafe Reference Braasch, Peters, Oliveros, Van Nort and Chafe2011). Analysis software is used to compare the close-miked signals with the signals detected by the array, suggesting the location of each performer. The close-miked signal is then transmitted to the appropriate loudspeakers in the remote venue, typically positioned among the performers, so that participants experience the musicians from both sites performing together in a virtual space.

These examples illustrate a fundamental tenet of technological mediation theory, namely how subject and object relationships are co-constituted (Verbeek Reference Verbeek and Achterhuis2001: 130–1). Just as the remote performers are defined by their representation through audio and video technology, one’s awareness and creative strategies are shaped by their perception of the sounds and gestures of others. This observation, in turn, resonates with Tracey McMullen’s description of Oliveros’s goals pertaining to improvisation. As McMullen writes, ‘Deep listening in improvisation could be understood as a practice of dissolving the self by putting deep focus on the other-in-front-of-us and coming to understand the interrelatedness and co-arising of supposed “other” and “self”’ (McMullen Reference McMullen2010: 9). Both the use of technology to recreate the experience of co-presence, as with ViMiC, and the ‘composite’ awareness engendered by latency and audiovisual asynchrony draw attention and awareness to the contingencies of musical and social interaction, prompting participants to take new, external factors into account as they improvise. At the same time, this process of learning how to perceive strongly resembles the second-order learning proposed by Belgrad.

Reflecting on the 2007 performance, Dresser remarked that the most interesting aspects of the process were on the ‘social and human level’, as opposed to the technological (Oliveros et al. Reference Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe2009: 16). It goes without saying that telematic performances present a variety of obstacles that must be overcome through negotiation and cooperation between many individuals and institutions that range from the technologically arcane, such as ViMiC, to more mundane matters such as scheduling rehearsals across different time zones. In many cases, much of this process also takes place telematically. Dresser writes that telematic communication promotes a ‘unique sense of group sharing’ and that in moving between different tools to communicate, such as instant chat, Google Docs and Skyping, he experienced the different ‘tempos’ of these platforms more profoundly (Oliveros et al. Reference Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe2009: 13, 16). The preparation process, for him and others, instilled a kind of ‘telematic etiquette’ that informed how the musicians interacted while performing, and also generally deepened the connection and ‘depth of collaboration’ they felt with the other participants (Oliveros et al. Reference Oliveros, Weaver, Dresser, Pitcher, Braasch and Chafe2009: 11).

The sense of community engendered through this process resembles what Georgina Born describes as the second order of social mediation in music: the ways in which music ‘animates imagined communities, aggregating its listeners into virtual collectives or publics’ (Born Reference Born2010: 232). For Born, this order of social mediation is ‘assembled specifically by musical practice’, but goes beyond the microsocialities of ‘practice’ per se, which characterises the first order. The preparation process of a telematic performance of even modest scale mobilises numerous individuals over weeks or months, most of whom will engage with technology that may have taken years to develop. Unlike performers whose time in the practice room is mostly spent alone, most of the preparation for a telematic performance requires constant communication and interaction with others, often remotely, extending the process of learning how to interact telematically far beyond the time of performance. In a way, the asymmetry of time between preparation and performance gives Oliveros’s observation about music as a ‘by-product’ a curiously literal valence.

Oliveros herself saw telematic performance as a way to effect social change, describing it as a ‘new, nonviolent development in global communication’, and as a tool to ‘connect with my many friends throughout the world and strengthen our relationship with global culture’ (Oliveros Reference Oliveros2010a: 194). Although telematic performance is in some sense a practical solution to having musical partners scattered around the world, it is also a catalyst for fostering a community built around Oliveros’s guiding humanistic principles. As the observations by Weaver and Dresser make clear, the idiosyncrasies of different technological platforms definitively shape how users interact by rewarding personal qualities such as sensitivity, awareness, patience and creativity. Furthermore, just as in Nikola Tesla, the technology itself is never transparent. Telematic technology is constantly drawing attention to itself in the form of latency and signal disruptions that may be caused by hardware or software at a different site. As resolving any such issue would require reaching out to distant colleagues, Oliveros’s approach ensured that even technological failure would result in the cooperative processes that lead to successful social dynamics.

5. The Adaptive Use Musical Instrument (Aumi)

In 2006, Oliveros and occupational therapist Leaf Miller conceived of a new electronic tool for improvisation that would come to be known as the Adaptive Use Musical Instrument, or AUMI. A note on the AUMI website describes the instrument as a way for ‘the user to play sounds and musical phrases through movement and gestures. This is an entry to improvisation that enables exploration of sounds ranging from pitches to noises rather than learning set pieces’ (AUMI n.d.). AUMI uses a camera that begins tracking movement automatically using a single cursor, but the user can also click a part of the screen corresponding to a head, finger, or shoulder, for instance, to reassign the music-making part of the body. When the object passes within determined zones superimposed over the camera view, it triggers a sound, ranging from a stock library of musical sounds to prerecorded samples and noises. Since the development of the first prototype in Max/MSP in 2007, many individuals have created or extended versions of AUMI for a variety of platforms. At the time of writing, AUMI is available as a desktop software program – the current version was developed by Ivan Franco and John Sullivan – and as an iOS app developed and maintained by Henry Lowengard.

One of the fundamental priorities of AUMI is allowing musicians across abilities to be able to use it to improvise, with a particular focus on providing access for people with limited mobility. Miller’s early use of the software at the Abilities First School in Poughkeepsie, NY, was formative in AUMI’s development trajectory. The software is designed so that users can get started quickly and play with minimal modification to settings, reflecting its usage among individuals who may require assistance. The monophonic nature of the instrument also makes it conducive to group settings, preventing any one performer from dominating the musical texture. As Lowengard has pointed out, without further intervention, the instrument maintains a consistent operational logic through a fixed correspondence of cursor and screen position, making it easy to grasp conceptually as an instrument (Lowengard Reference Lowengard2020).

AUMI is designed to adapt to not just anybody but also literally ‘any body’, in contrast with most instruments – electronic or otherwise – which exhibit specific body requirements (Stewart, Tucker, Williams and Haaheim Reference Stewart, Tucker, Williams and Haaheim2017: 5–6). Additionally, the intuitive nature of AUMI distinguishes it from other electronic instruments and interfaces, which often require extensive calibration and frequent interventions. Although AUMI allows for numerous adjustments of its tracking functionality and musical output, the relatively simple interface points towards other priorities.

For example, AUMI’s capacity to latch onto unexpected parts of the body, or trigger sounds unpredictably, has occasionally been frustrating to musicians experienced in using motion-tracking technology to perform. However, the extent to which AUMI makes music out of gestures not conventionally recognised as expressive is precisely where the unique role of technology comes into relief. Instead of using technology to extend conventional gestures of musicality or virtuosity, AUMI’s mediating role continually resists transparency: ‘when performing with the AUMI, the relationship between bodily gesture and auditory response does not always feel like a unidirectional flow in which a particular gesture reliably triggers a particular sound’ (Stewart et al. Reference Stewart, Tucker, Williams and Haaheim2017: 14). Yet this experience of uncertainty ultimately emphasises the extent to which, as Verbeek writes, ‘mediation does not simply take place between a subject and an object, but rather co-shapes subjectivity and objectivity’ (Verbeek Reference Verbeek and Achterhuis2001: 131). Just as in telematic performance, the subjectivity of the musician is just as much a product of the technological mediation as the objects they perceive.

This holds for the music produced collectively as well. AUMI users report that, ‘in most cases, the unexpected sounds and silences introduced into the performance by our interactions with the AUMI provided additional opportunities for performers to improvise. Participatory imprecision is valued in many musical cultures as being conducive for creating new and necessary social relations’ (Stewart et al. Reference Stewart, Tucker, Williams and Haaheim2017: 14). As Caleb Lázaro Moreno has written, ‘If we hold the AUMI’s early history close, then the AUMI is, first and foremost, designed to facilitate musical access for performers with severe physical disabilities. Its structural schematic and even its response-rate latency and sporadic visual mode of recognition ought to be treated as enabling functions rather than limiting ones’ (Moreno Reference Moreno2017). Just as with Oliveros’s telematic music, features of the technology that might elsewhere be considered limitations are precisely what makes its usage possible, especially within the communities for which it is designed.

Moreno argues that moments of technological uncertainty present participants with a unique opportunity to reconceptualise their approach to making music, writing that:

From this perspective, when AUMI responds to unintentional movements, such as bobbing the head or shrugging the shoulders, it is actually expanding the possibilities of musical expressivity, even for trained, able-bodied musicians. Similarly, just as telematic latency necessitates novel improvisatory strategies, AUMI’s idiosyncrasies shape the kind of music produced: its unpredictability, for instance, lends itself to music that is often pulseless, and equally open to noises as musical sounds. This unpredictability also works against the reification or ‘fossilisation’ of musical culture that, as Thor Magnusson argues, characterises most digital musical interfaces (Magnusson Reference Magnusson2009: 173). While according to Magnusson most digital instruments are ‘more likely to contain an expressive closure’, through its operational uncertainties AUMI seems more closely aligned with the open-ended ‘explorative openings’ Magnusson associates with an acoustic instrument (Magnusson Reference Magnusson2009: 173).

In a further affinity with Oliveros’s approach to telematic performance, the social interactions that occur outside of the performance highlight the community-building that distinguishes Born’s second order of social mediation in music. As Moreno writes, ‘What makes the AUMI a unique improvisational instrument is the tradition currently being co-created around its mechanisms in the Lawrence area, and the way these practices disrupt the borders between able-bodied and disabled musical production, participation, and communication’ (Moreno Reference Moreno2017). Just as intimate social interactions across distances emerge through the use of telematic technology, AUMI facilitates musical and social interactions between individuals that might not otherwise occur, illustrating how AUMI’s ‘benefits extend far beyond typical notions of disability accommodation, but actually challenge the lines that societies draw between ability and disability’ (Stewart et al. Reference Stewart, Tucker, Williams and Haaheim2017: 5). Sustained engagement with AUMI reveals a spectrum of bodily differences, as opposed to a normative distinction between ‘normal’ and ‘limited’ bodies (Stewart et al. Reference Stewart, Tucker, Williams and Haaheim2017: 15). By enabling individuals who are typically excluded from the performance of music to participate, AUMI problematises the structurally enforced social differences that characterise the third order of social mediation identified by Born (Reference Born2010: 232).