All Problems of Notation Will Be Solved By the Masses
In the 1960s and '70s musicians devised innovative forms of notation and protocol to liberate themselves from aesthetic and social conventions. Today's digital devotees of code based production and improvisation are continuing this tradition, argues Simon Yuill*
Of all the art forms supported and enabled through FLOSS, ‘livecoding' has emerged as the one which most directly embodies the key principles of FLOSS production in the creation and experience of the work itself. In livecoding, the artwork is expressed in software code that is written and re-written live during its performance. Many livecoding artists write their own software tools to support this. Alex McLean's ‘feedback.pl' was one of the first such tools. It is a simple Perl script that continuously reads and executes an extract of its own code displayed in a text editor. This code defines various algorithms from which music is generated. During performance this is re-written by the performer, changing the musical structure and effectively improvising with the code. A projection of the performer's desktop makes this visible, thereby emphasising how the code and the changes made to it are integral to the work and to the audience experience of it. The material and formal relationships between code and music are therefore discernible, even though many audience members may be unfamiliar with programming languages themselves. To some extent this is comparable to witnessing a performance on an acoustic instrument such as a guitar or clarinet. While we may not understand how to play such instruments ourselves, we can relate the gestures of the performer to the sounds that we hear and thus acquire a sense of the relation between the sound and its material production. This contrasts sharply with previous forms of electronic music performance, such as those of Jean Michel Jarre and Todd Machover, in which interface devices are presented on stage often simulating and referring to acoustic instruments. Livecoding dispenses with such ‘fetishes' and is unashamed to expose the bare materiality of its production. The unfamiliarity of presenting code as a raw material, however, results in something very different from that of the guitar or clarinet performance, and more akin to revealing the stage machinery in a Brecht play. It creates a virtue by exposing something that is normally concealed.
While livecoding initially developed as a form of music, it is not restricted to this. Dave Griffith's ‘fluxus' and Tom Schouten's ‘PacketForth' are tools for creating visual works, the first based on a 3D graphics engine and the second a video processing system. Some existing tools, such as ‘SuperCollider', ‘Chuck' and ‘Pure Data' have also been used for livecoding work. In fact, any programming language or tool that can execute code on the fly can potentially be used for livecoding. The concept has also been extended into other forms of work. ‘Social Versioning System' (SVS) enables multi-player simulation games to be created and coded live, with new code distributed among the players as a game evolves. Ap's ‘Life Coding' is a large scale performance combining software coding, circuit bending and conference-style spoken presentations.
Livecoding Aesthetics
There are two key aspects of livecoding that embody FLOSS principles. Firstly, the way it makes the continual re-writing of code itself a primary mode of artistic production, and, secondly, its presentation of the ‘work' itself as an open-ended mutable piece of code rather than as a static discrete artefact. As opposed to most non-digital and new media art which is presented solely as a commodity to be consumed, livecoding makes its own materials and practices of production available to others. Livecoding emphasises the FLOSS principle of code-based production as a form of production that is itself ‘live' and living, that enables the possibility of production by others for their own purposes.
This ‘enabling the possibility of production by others' is often continued beyond the performance not only in the use of FLOSS-style distribution, but also in the conscious use of workshops as a means of presenting works and teaching the skills used in their creation. This pedagogic aspect extends to the prominence given to technical meetings and development workshops in artist-run festivals such as Piksel and MAKEART, or groups such as Dorkbot and OpenLab, and into the creation of dissemination platforms and projects such as pure:dyne and FLOSS Manuals. The often ad-hoc, workshop nature of many livecoding performances and projects themselves is an extension of the livecoding ethic of sharing and making materials generally available. In the case of the ap events that are staged over extended periods of 12 hours or more, this includes participants learning and adapting the tools of the performance as it takes place. On a smaller scale, the London OpenLab group host ‘drumming circle' performances in which anyone can join in with their own algorithms and code, constructing and developing a collective, rhythmic work, as well as performances that start from one piece of code that is rewritten by successive performers. Rather than something marginal or extraneous to the ‘art', the idea of the workshop has been absorbed as an integral aspect of livecoding aesthetics.
Livecoding is not the sole or even dominant form of practice pursued by all those involved in FLOSS-related arts. What all practitioners involved in these projects do share, however, is a commitment to the broader notion of ‘live code' as a mode of production and a common preference for a workshop aesthetic. It is also within these more ‘pedagogic' practices that artistic production within FLOSS meets with other aspects of the FLOSS world, and specifically the political and socially engaged practices emerging from hacklabs and hackmeets.
Hacklabs & Hackmeets
Hacklabs are voluntarily run spaces providing free public access to computers and the internet. They generally make use of reclaimed and recycled machines running GNU/Linux and, alongside providing computer access, most hacklabs run workshops in a range of topics from basic computer use and installing GNU/Linux software, to programming, electronics, and independent (or pirate) radio broadcast. The first hacklabs developed in Europe, often coming out of the traditions of squatted social centres and community media labs. In Italy they have been connected with the autonomist social centres, and in Spain, Germany, and the Netherlands with anarchist squatting movements. Hackmeets are temporary gatherings of hackers and activists in which skills, tools and knowledge are exchanged and projects developed. Among the first hackmeets were those in Italy in the 1990s. There are direct connections between many of these and artists working with FLOSS. The dyne:bolic project (from which pure:dyne evolved) partly developed through the Italian hackmeets and Dutch hacklabs. RampArts hacklab in London has provided a meeting point for the local OpenLab group, and in Barcelona, spaces such as Hackitectura and Riereta have supported several FLOSS-based art and political projects. Not all artists working with FLOSS and livecoding necessarily share the politics of the hacklabs scene, nor do all hacklab participants necessarily look upon their own activities as art-related, and some are, sometimes rightly, sceptical of artistic involvement in what they do. Hacklabs, however, have been absolutely fundamental to the development of FLOSS in recent years, especially in Europe and South America, and have provided a clear political and ethical orientation in contrast to the somewhat confused and often contradictory political and social perspectives articulated in the other communities and contexts of the wider FLOSS world.
If livecoding is one of the most emblematic artistic manifestations of FLOSS, hacklabs have become one of its most emblematic social forms. While the two may not occupy identical trajectories, they nevertheless overlap and compliment one another in many significant ways. Central to this is their shared principle of ‘enabling the possibility of production by others'. This is an issue of distribution, not simply at the level of product, in the way a piece of software can be easily distributed for example, but at the level of practice. The practice itself is inherently distributive, for it integrates the distribution of knowledge on how to produce into that which it produces. While this allows for possibilities of collaborative production, it should be seen as distinct from collaboration in itself. For whereas a practice that is collaborative coheres the production of many under a single goal, thereby directing the disposition of their labour, a practice that is distributive enables the disposition of labour by others under their own direction. This is enabled in the output of production as notation, as code that not only creates a product, but enters into an active life beyond its initial implementation.
Notational Production
Notational production is not unique to software. The emergence of livecoding as an initially musical activity reflects the engagement with notational production that has characterised many different musical traditions. The application of computer code to the construction of sound is, in one sense, simply one more episode in this process. Livecoding works from within a particular relation between notation and contingency. The specificity of code is opened towards the indeterminism of improvisation. In this respect livecoding not only adds to the evolution of notational production within music but echoes a particular period where a similar relation between notation and contingency came to the fore. This was a period in which the ‘free playing' of experimental jazz developed by the likes of John Coltrane, Ornette Coleman and Sun Ra, met with the ‘open' compositional systems of the avant-garde that had been developed by John Cage, Karlheinz Stockhausen, Earle Brown and others. Just as FLOSS brings together two related, yet differing, ethics of software production (‘Free Software' and ‘Open Source'), we might describe this music as Free Open Form Performance (abbreviated as FOFP). ‘Free playing' was a term preferred by Coleman and other jazz musicians who rejected the use of the term ‘improvisation' on the grounds it was often applied to black music by white audiences to emphasise some innate intuitive musicality that denied the heritage of skills and formal traditions that the black musician drew upon. ‘Open' comes from Umberto Eco's ‘Poetics of the Open Work', an essay from 1959 which was among the first to survey and analyse the experiments with aleatoric, indeterminate and partially composed works that were emerging in the classical avant-garde. By the late 1960s these two strands of development had crossed over. Jazz composers such as Coleman and Anthony Braxton consciously worked with classical avant-garde instrumentation and musical structures. Meanwhile, the Scratch Orchestra adopted the collective social form of outfits such as the Art Ensemble of Chicago. Experiments with notation were significant for many of these musicians and composers, but for the Scratch Orchestra, the exploration of notational production was a cornerstone of the project.
Scratch History
The Scratch Orchestra grew out of a series of public classes in experimental music that Cornelius Cardew and other composers had been running in London in the late 1960s. These began at the Anti-University on Rivington Street continuing at Morley College, a workers education centre set up in the 19th century. It was here that the original members of the Scratch Orchestra first came together: Cardew, Michael Parson, Howard Skempton and people attending their classes. The foundation of the Orchestra was officially announced in June 1969 through the publication in the Musical Times of ‘A Scratch Orchestra: draft constitution' written by Cardew. The constitution defines the Orchestra as:
[...] a large number of enthusiasts pooling their resources (not primarily material resources) and assembling for action (music-making, performance, edification).
Membership was open to anyone, regardless of musical ability. Many visual artists, such as Stefan Szczelkun, joined and brought with them an interest and experience of art happenings and urban intervention works. Through these, and more conventional concerts, the Orchestra aimed to ‘function in the public sphere' presenting works developed by the group. The constitution outlined various forms of activity that the Orchestra would follow in creating these works. One of the most important activities was the writing of ‘Scratch Music'. Each member of the Orchestra had a notebook, or ‘Scratchbook', in which they would write small works that could be combined into larger ensemble pieces. The constitution emphasises that these Scratch Music pieces should be an active process of experimentation with different notational forms: ‘verbal, graphic, musical, collage, etc.'. By 1972 a clearly defined process for the development of Scratch Music had emerged. Each piece was originally performed by its author, the scores were then exchanged and performed by other Orchestra members, providing a kind of ‘peer review' critique of the pieces. ‘Scratchers' were asked to write no more than one new piece per day, but encouraged to keep a ‘regular turnover', so that there was a tight feedback loop between writing and performing.
From the very beginning the Scratch Orchestra took a conscious decision to make all their notations freely distributable, stating that the Scratch Music works were without copyright. One of their first collections of scores, published in 1969 and called Nature Study Notes: Improvisation Rites, replaced the conventional copyright notice with the following:
No rights are reserved in this book of rites. They may be reproduced and performed freely. Anyone wishing to send contributions for a second set should address them to the editor: C. Cardew, 112 Elm Grove Road, London, SW13.
While rejections of copyright restriction were nothing new - both the situationists and the folk singer Woody Guthrie had placed anti-copyright notices on their works - it is notable that the Scratch Orchestra also encouraged others to modify and add to their scores, stating that these may be incorporated into the next version.
Noise Interrupts
The works in Nature Study Notes are all textual instruction pieces. Few of them describe ways of making sound however, and instead focus on various social interactions that construct and play with power relations among the performers. Some are like party games:
Form a standing circle. Nominate a leader, who stands in the circle with eyes blindfolded. The remainder of group rotate slowly around him/her. ... When the leader is touched, he forfeits his role and so doing shouts ‘Porridge'.
Others like generative automata:
Each person entering the performance space receives a number in order. Anyone can give an order (imperatively obeyed) to a higher number, and must obey orders given him by a lower number. No. 1 receives his orders from the current highest number (the most recently entered player); the highest number can give orders only to No. 1.
Many of the scores in Nature Study Notes set up small scale ‘operating systems', simple organisational structures that enable other works to be produced within them. The notion of the performance as an operating system is one that ap have taken up in their ‘Life Coding' project. Adapting mechanisms from computer systems, the interaction of performers is dictated by ‘interrupt' signals connected to actions defined in look-up tables. In conventional computers, the interrupt mechanism enables signals from peripheral devices such as mice, keyboards or network cards to enter into the operating system. When an interrupt signal is received, the computer selects a response action by matching an identifier code for each signal against a look-up table of programmed routines known as ‘interrupt handlers'. In this way pressing keys on a keyboard or moving the mouse can change the course of events currently in action. The interrupt creates a vector between the internal operation of the central processing unit (CPU), the domain of notational operations, and the contingency of the outside world. As Edsger Dijkstra, one of the inventors of the interrupt system, noted:
It was a great invention, but also a Box of Pandora. Because the exact moments of the interrupts were unpredictable and outside our control, the interrupt mechanism turned the computer into a nondeterministic machine with a non-reproducible behaviour, and could we control such a beast?
The interrupt breaks the closed linear unfolding of the Turing Machine, enabling programs to be stopped, altered and restarted. This enabled the development of languages that could be executed as individual statements one step at a time, giving rise to shell commands (the basic text-based commands used in the UNIX terminal) and the read-evaluate-print-loop (sometimes ‘read-eval-print-loop' or REPL for short) that forms the basis of interactive programming languages such as Lisp. The interrupt and read-eval-print-loop lie at the heart of any livecoding program and all UNIX-derived operating systems. In his notes for the first release of Linux, Linus Torvalds wrote: ‘interrupts aren't hidden'. It is here where contingency and notation meet, but it is here also that the possibility of error enters. However, rather than treading lightly for fear of a crash, for some the error carried on an interrupt signal is a positive, productive opportunity. This is not restricted to computer interrupts. During rehearsals, Sun Ra would deliberately interrupt and trick his performers. The ‘errors' this produced, however, were not mistakes but rather forms of evolution:
There are no mistakes. If someone's playing off-key or it sounds bad, the rest of us will do the same. And then it will sound right.
The operating system of Ra's Arkestra incorporated such ‘noise' and restructured itself in the process. This ‘noise' is not simply that of unmusical sound, but also, in the sense that Jacques Attali adapts from information and systems theory, any material that is not recognised by an existing system and is therefore opposed to ‘information' which is material that has value or significance in a given system. Attali describes the evolution of musical styles as one in which an existing system of music becomes exposed to ‘noise' that at first disrupts it, but then, through incorporation restructures it and gives rise to a new system. In the voyage of the Arkestra, systems would collapse and be reborn on a daily basis.
Schooltime Compositions
This power over systems was not limited to the Demiurge or intergalactic jazz master. During the same period in which the Scratch Orchestra was re-inventing music from the ground up, a group of children at Muzzey Junior High School in the US were experimenting with their own improvised notation systems. These children were not writing music however, but teaching themselves to program computers. They were part of the first LOGO Lab, a project initiated by Seymour Papert, a researcher from the MIT Artificial Intelligence Laboratory. LOGO was a simple programming language that directed an entity called a ‘turtle'. The turtle could either be an on-screen virtual character or a small robot that was instructed to move around their terrain (screen or floorspace) and that could draw a trail on its path. LOGO Lab students developed their own programs in which the turtles would act out drawings or spatial exercises. In so far as LOGO expresses a series of potential actions out of which a drawing emerges it is analogous to the notations of the Scratch Orchestra, which often did not express sound directly but rather actions from which sound could arise. As Cardew wrote in his notes to Treatise: ‘Notation is a way of making people move.'
Like the Scratch Orchestra, the LOGO Labs grew out of a conscious pedagogical interest in developing forms of collective, self-directed practical research. These were realised through semi-structured ‘improvisational' activities and used self-developed notational systems as a means of constructing, communicating and reflecting upon these activities. As the constitution makes clear, the Scratch Orchestra was a conscious exploration of what notation could be and how that related to attempts to establish another understanding of what the practice of music itself might be. This developed out of the pedagogic context of the Morley College classes, and, in a perhaps self-mocking gesture, the Orchestra's Nature Study Notes and Cardew's earlier Schooltime Compositions scores deliberately took the form of school exercise books.
Papert believed that programming was a skill that should be available to everyone not as a ‘technology' - a mechanism for manufacture abstracted from human labour - but as a means of conceptual exploration. There are political parallels between the two projects. Papert's approach to computing was influenced by his previous involvement in radical left-wing politics - in the 1950s he had been involved in the group running Socialist Review in London. The LOGO Lab concept combined insights from Jean Piaget's and Lev Vygotsky's psychological studies of child development with the non-schooling principles of Ivan Illich. It advocated an approach in which, ‘the child programs the computer rather than the computer being used to program the child.' Papert also argued that the design of a programming language could reflect a particular political and ethical position. He criticised BASIC, another language originally designed for teaching programming, as demonstrating
how a conservative social system appropriates and tries to neutralise a potentially revolutionary instrument.
Although the Scratch Orchestra did not develop out of a defined political program, it nevertheless acted as a context for the development of a politicised arts practice, informed by both Marxist and anarchist tendencies. It was through the Scratch Orchestra that Cardew was to acquire a profound political awareness, applying an explicit Maoist perspective to his own practice, and leading to his involvement in founding the Revolutionary Communist Party of Great Britain (Marxist-Leninist). Echoing Papert's criticisms of BASIC, Cardew similarly criticised the institutionalised conservatism of much music notation, demanding instead that ‘all problems of notation will be solved by the masses.' For both Papert and Cardew, pedagogy was a two way street. The lab and the orchestra broke down distinctions between pupil and tutor, and placed learning in the context of self-directed production. In these ways they were forms of distributive practice.
Training in Contingency
An element of the contingent was essential to this form of radical pedagogy. In Papert's eyes, one of the strengths of programming as a tool for learning was the attitude to error that it encouraged. Encountering error, in the form of bugs, was an inevitable and necessary part of programming, especially that particular practice of programming developed at MIT's AI Labs known as ‘hacking'. Papert pointed out that in conventional education errors had a purely negative connotation. When a student makes a mistake they are discredited for it, losing marks or being punished, thereby inculcating a fear of error, leading to an unwillingness to stray from conventional boundaries and take risks. For the hacker, conversely, what mattered is not whether or not a mistake is made but rather how creatively it can be responded to. As with the Arkestra, embracing error is a productive possibility. The embracing of error is reflected in documents such as HAKMEM. Short for ‘hack memo', this was a collection of code snippets and programming ideas distributed among the hackers within the AI Labs - contributors include Richard Stallman, James Gosling and Marvin Minsky. Many of the entries utilise possibilities discovered through bugs and inconsistencies within the PDP computers that the AI Lab worked on. Other entries suggest ways that a particular algorithm might be played with, encouraging people to mess around with it in what can only be described as a form of aesthetic code play. HAKMEM can be seen as the AI Lab's equivalent of the Scratchbooks exchanged between Scratch Orchestra members. Within the LOGO Labs, code was written and exchanged between students in a similar manner. Rather than planning out programs in advance, pupils would ‘improvise' with their code, responding to the how the turtle performed and modifying their programs accordingly. LOGO learning thereby operated through a similar feedback loop of coding-performing that livecoders such as Alex McLean identify as the basis of their practice and which builds upon the principle of the read-eval-print-loop.
Computers and programming languages present highly constrained environments that limit the possible varieties of interpretation that a particular notation may be subject to. The interpretation of notation by a human may be far less constrained. For Cardew this was a major concern in the development of new forms of notation, for it presented both a danger and an opportunity. The opportunity was that notations need not only encode existing patterns or defined systems of sound, but could also be proposals and provocations to create new ones. The danger lay in the fact that a trained musician, when confronted with an unfamiliar notation system, rather than responding to it directly, might fall back on their personal predispositions and ingrained habits. The performance may simply become the regurgitation of old clichés and formulas like that of the amateur jazz musician described by Adorno, unable to stray from the existing models to which he has adapted and subordinated himself. The trained musician approached a performance with a predefined system of producing sound against which the new notation was interpreted. What was novel in the new notation may simply be responded to as ‘error' or noise within that system and therefore avoided. New notations required performers with a similar attitude to that of the hacker and LOGO Lab student, one who could respond creatively to the unknown and unexpected. The performer, therefore, could not rehearse such music but rather ‘trained' for it like a martial art, developing ways of acting upon contingency. This similarly developed through a feedback loop of coding-performance that formed the basis of Scratch Music practice.
Through such feedback loops notation incorporates the experience of the contingent into future practice. What was the unexpected ‘error' of the past becomes preparation for unknown future possibilities. In absorbing this a notation records the historical development of a practice, capturing different versions of how things could be done, and enabling comparison, analysis and synthesis. In both the LOGO Labs and Scratch Orchestra, this process of versioning was consciously engaged in, with the evolving knowledge, intentions and standards of the practitioner community acting as a form of version control identifying those practices which are most current and those which are conflicting or tangential.
Black Notated Music
How a notation comes to be defined and how it is distributed are inherently political issues. This distribution extends beyond the publication of music scores and software code, in the form addressed by Scratch Orchestra's use of copyleft mechanisms. As Ornette Coleman recalls, the very visibility of notation within the production process, how it is revealed and concealed, both depends on and expresses particular relations of power:
I once heard Eubie Blake say that when he was playing in black bands for white audiences, during the time when segregation was strong, that the musicians had to go on stage without any written music. The musicians would be backstage, look at the music, then leave the music there and go out and play it. He was saying that they had a more saleable appeal if they pretended to not know what they were doing. The white audience felt safer.
The denial of notation described in this episode is a denial of the black musician's self-legitimation. If the use of a notation can help register the development of a practice, its history and self-reflection, then the denial of notation is a denial of this history and, therefore, a denial of the practitioner's basis for legitimation. It is from this perspective that Coleman distances his own practice from the idea of improvisation, for this form of ‘virtuosity' became the basis of a denial of legitimation. The ‘free playing' that he and other black jazz musicians promoted in the 1960s was not simply free in the sense of a break from conventional musical structure, but also free in breaking away from the condition of being ‘improvisers in a compulsory situation.' This led to the development of new performance venues, many situated directly within black communities, and of the conscious articulation of practice as a form of research. Lester Bowie of the Art Ensemble of Chicago adopted a scientist's white lab coat on stage to announce the performance itself as a site of radical experiment. Sun Ra encouraged his Arkestra by declaring: ‘You're not musicians, you're tone scientists'. Ra developed this concept further through the creation in 1967 of Ihnfinity Inc, a research corporation intended
to own and operate all kinds of research laboratories, studios, electronic equipment, electrochemical communicational devices of our own design and creativity ...
In St. Louis the Black Artists' Group set up a Training Centre to create a discussion forum for the local community that, alongside performances, rehearsals, and workshops, hosted regular meetings and debates about local issues. For Anthony Braxton the relation of notation to l