Performance methods

One initial aim of this project was that it should result in a repertoire for improvisation, as a foundation for music making. Not as a set of fixed clichés that can readily be produced when needed, but rather as a vocabulary of a language or coordinates in a musical space of possibilities to be explored. Repertoire in this context refers to a whole range of internalized know-how about how to use these speech sources to make music.

For the improviser, the process of developing and internalizing such knowledge includes the aspect of performance, both practically mastering the instrument, but also musically knowing one’s ways around the material at hand at a moment’s notice. Usually one can rely on an accumulated reserve of experience as a performer and improviser, but I found that for this project I was literally starting from scratch. Not only was I developing new approaches to making improvised music based on speech analyses, I was also building a new instrument and learning to play it at the same time. In retrospect it is not so strange then, that it took much more time than I had anticipated reaching the point where I could improvise and perform on a sufficient level. To get there I went through several rounds of exploring, learning, developing, rehearsing, performing, and revising. On a small scale these rounds typically took the form of a repeating creative loop:

||: Idea -> testing -> evaluation -> revision :||

On a larger scale, over the course of the whole project, the process of musical development through performance can be summarised with these stages:


During early instrument development, I made a series of simple studies to explore basic musical possibilities, usually focusing on only one element at a time. Such limited studies have been suggested as a useful method for developing and discussing isolated elements in artistic research (Schwab & Borgdorff, 2014). This early into the project it was not so much about discussing, but more about just identifying basic material. The results are not terribly interesting to listen to as music, but making these studies was an important first step in the process of defining and exploring material for later use.

Sound example: organ study, exploring ways to use different musical features from the same speech source to arrange multiple polyphonic layers, in this case three parts for different registers of a MIDI-controlled pipe organ, arranged and improvised in real time:


Improvisation as method for discovery

The next step was to play more freely with different elements in open-ended improvisations, which I recorded and evaluated. These improvisations were often of long durations with little attention paid to form or development, but served as a productive method for discovering interesting musical ideas as well as identifying general shortcomings of the instrument system.

Sound example: improvisation


Through systematically working through such open-ended improvisations, I was exploring new sonic and musical possibilities as well as different formal ideas, thereby gradually building and internalizing a vocabulary for improvising with this material and instrumental setup.

Repertoire development

The accumulated results from these improvised explorations gradually took the form of a repertoire of musical possibilities, like amassing building blocks or mapping out elements of a formal language. To better understand what this repertoire consisted of and what the relations between the different musical ideas were, I worked on different ways of how to categorize these ideas conceptually. Early on I had tried to define the musical ideas in this project in terms of continua between opposites, constituting dimensions in a space that could be explored and navigated. Although this was appropriate for some ideas, not all was best thought of as spatial dimensions. Working on other ways to organise the repertoire conceptually, I found that one productive approach was to view ideas in relation to a set of different categories of significance relating to the material, instrumentation or musical properties, serving as a framework for further musical exploration with these ideas:

Speech genre

Perception of time (narrative, interactive, static, cyclic)

Sound source / orchestration (voice, loudspeaker, instrument, acoustic, synthetic)

Formal organization: structured thoughts (speech) or structured sounds (music)

Traditional musical parameters: layers (polyphony), instrumentation, density, register, tempo, dynamics, continuity, etc.

Composing for improvisation

Starting to get a grip on what my materials were, I was still not able to maintain the control or overview needed to improvise freely with this repertoire. As a step towards greater improvisational control I composed sketches with graphic notation. This way I could work on formal issues on how to combine these elements and notions into coherent forms. The many ideas I had developed earlier through improvisation were then organised within the framework of these main categories of significance, and from that I could compose the broader form of a piece of music while still leaving the details to be improvised.

The composition sketched above consists of five main parts which explores musical associations of different speech genres, different aspects of time perception, as well as different musical ideas, orchestrations and soundscapes:

Part 1: LINES
Time concept: forward motion through a continuous stream of speech.
Source material / speech genre: social interaction in reality-TV.
Soundscape: Speaker mediated speech is abstracted and moved to acoustic instruments, different lines of rhythmical layers (syllables, accents, stress, phrases) orchestrated out to different instruments, more and more stylized like musical figures (smoothing tempo, pitch and adding diminutions and augmentations), before returning to the free flow of continuous speech.

Time concept: accentuated present, interaction between performer and machine, music and speech.
Source material / speech genre: paralanguage, telephone conversations.
Soundscape: Live acoustic piano against speaker mediated speech. Develops from interactive dialogue through musical association to polyphonic texture: a crowd where each individual voice disappears. The sound of the crowd is then drawn out and abstracted into a static time chordal texture.

Part 3: SPACE
Time concept: still, static, stationary, open space rather than measured time.
Source material / speech genre: intimate thoughtful conversations.
Soundscape: Sparsely orchestrated parts across the room, blending both speech recordings, synthesized sound and acoustic instrument in a wide-open soundscape.

Part 4: RHYTHM
Time concept: cyclic, repetitive, ever present.
Source material / speech genre: representative, non-personal radio broadcasts.
Soundscape: acoustic drums. Isolated similar syllables make a rhythmic web accompanied by a metronome measuring the mean pulse. Switching between repetitions and narration and gradually adding more and more subdivisions it develops and culminates into a dense distorted electroacoustic soundscape, a raw mass of sound that stripped of any resemblance with speech still conveys some overall phrasal gestural shapes.

Part 5: PAUSES
Time concept: the spaces and gaps between utterances, between lines.
Source material / speech genres: dignified official speech from a head of state to its citizens, characterized by the downwards melodic finality of each statement and the long pauses between them.
Soundscape: spatial arpeggii in bells and acoustic cymbals fills in and sets music to the pauses, articulating the gaps between utterances and their imprint on the aural memory. The sound of a gong and the dark drone background adds to the ritual character of this speech genre.

Performance logistics

Preparing for improvisation by way of composition also made it much easier to rehearse the logistics of performing, and really concentrate on the practical aspects of controlling the new instrument. This is of great importance, as it is easy to underestimate the amount of motor memory training and neural rewiring necessary to properly internalise control over an instrument.

The first public performances were of this kind with an overall composed form.

Video: solo performance at Harpefoss Poetry Festival in 2016:

Improvisation and interaction

It was not until after these developments that I began to reach a level of control and internalisation where I could actually keep up with the ideas and opportunities of the moment and really start to improvise freely with this material and instrument setup. The main performance concept was still the solo improvisation setup, but now I could also try out experiments by bringing other musicians into the mix, exploring how this concept could work in interaction with other performers. One such experiment was a completely improvised duo set with vocal and electronics performer Tone Åse.

Video: duo improvisation with Tone Åse:

Another such experiment was an ensemble performance with a semi-composed piece for strings and percussion, where one issue was how to combine the improvised nature of my approach with the need for composing a piece and writing parts for the musicians. One option was to transcribe a piece in detail based on my current material and methods, arranged for that particular group of musicians. But then the reactive and interactive nature of the improvised interplay would disappear. Instead I wrote a score with general directions for improvisation within set limits. This is also how I usually conceptualize musical ideas when working with improvised music, also drawing on the tradition of indeterminate notation developed by composers like John Cage, Christian Wolff, Earle Brown, Pauline Olivieros a.o.

The piece has five parts, each about 3 minutes long each notated on one page:

Despite these experiments with different ways of approaching ensemble performances, it is the solo performance setup that has remained the main performance format throughout this long process of developing both a new way of making music, a new instrument and a new performance concept at the same time. Nevertheless, for the final presentation of artistic results, I also chose to include two alternative versions in addition to the solo format – the instrument setup as a self-playing interactive sound installation, and employing the instrument in a free improvised ensemble performance with six additional musicians. These experiments provide alternative uses of the system and point forward to possible future explorations of the ideas developed in this project. These performances are documented in the following chapter on final artistic results.


Schwab, M., & Borgdorff, H. (Eds.). (2014). The exposition of artistic research: publishing art in academia. Leiden: Leiden University Press.

← Previous page: Sound and orchestration Next page: Final artistic results