machines for music: a condensed history of generative music

When asked "What is Generative music?" a comparison with the music of wind chimes can be a helpful answer. Given the potential complexity of some Generative music systems this could seem dumbed down but it is in no way a glib response. Wind chimes have an elegant simplicity that captures the essence of Generative music and allows one to extrapolate and imagine how such a system might work at larger scales. Alan Dorin writes that wind chimes represent something unique within this musical style:

The "infinite" that is so often discussed in Generative music is not endless in every sense of the term. In these works, infinite may characterize the potential length of performance or the perceived variety of melodic and textural development in the piece, to name a few. It does not suggest that comprehensive musical knowledge has been encoded as a simple computer program that will spin out tune after tune. Music like all of the arts benefits from constraints. They have the counterintuitive capability to increase creative potential and variety. While wind chimes are perhaps an extreme case of creative economy, they show that there can be a seemingly infinite variety of beauty and interest produced through very simple means. Other musical precursors reveal this in different ways as a result of their strengths and productive limitations.

Cybernetician Stafford Beer's definition of an algorithm is "a comprehensive set of instructions for reaching a known goal" (Beer 1972: 305). The idea of an algorithm was first introduced in the ninth century by Abu Ja'far Mohammed ibn Musa al-Khowarizmi (Cope 2000). Composers have been employing algorithms since the 1026. Guido d'Arezzo (995-after 1033) developed a systematic means to pair pitches with the vowel sounds in the words of a liturgical text (Toop 2001; Roads 1996). Years later, Philippe de Vitry (1291-1361) (Cope 2000), Guillaume de Machaut (1300-1377), and Guillaume Dufay (1400-1474) (Roads 1996), are all known to have used algorithmic techniques in various ways to combine the rhythmic, pitched, and textual material of motets. In 1660 Giovanni Andrea Bontempi wrote New Method of Composing Four Voices, by means of which one thoroughly ignorant of the art of music can begin to compose, in which he proposed various systematic means of composition for, as the title suggests, uninitiated musicians (Cope 2000).

In the eighteenth century Mozart is often the most-recognized for composing Musikalisches Würfelspiele (musical dice games), but Haydn, C.P.E. Bach, and Johann Philipp Kirnberger (Cope 2000) were also involved in composing these chance-based, musical parlor games.

David Cope, a composer and expert on Algorithmic music, continues his historical discussion of algorithmic precedents in western art music to include Johann Joseph Fux, whose rules regarding counterpoint were influential to Bach, Mozart, and Beethoven, among others (Cope 2000). He cites many other musical situations where an algorithm or some system of constraints has been employed in nearly all forms of composition leading up to the modernist serial approach of Pierre Boulez and the aleatoric techniques of John Cage. Cope considers indeterminate techniques, compositions created on performance instruments, and the rules of music theory all to be a kind of algorithm. This enormously inclusive claim, he believes, "…helps diffuse the usually destructive segregation…" (Cope 2000: 15) between composers who do and composers who do not use algorithms in their compositions. Cope's intentions to find common ground between composers with different methodological views are noble because they serve to break down some of the barriers in the discourse of music composition. However, they do not help communicate the nuances that make various kinds of algorithmically-based musical works unique due to the artistic intentions of their creators.

Beer's definition of an algorithm includes, "…a known goal," which means the destination or result of a process is specified in some detail at the outset of the operation. This definition speaks to the written musical work itself. It reflects the historical context of these techniques within western art music, and situates algorithmic composition as one "musical species" evolved from the seminal pieces identified here. In his book, The Brain of the Firm, Stafford Beer makes a clear distinction between algorithm and heuristic, which is defined as "a set of instructions for searching out an unknown goal by exploration, which continuously or repeatedly evaluates progress according to some known criterion" (Beer 1972: 306 emphasis added). He provides the example that if you were to help someone reach the top of a mountain covered by clouds, the heuristic "keep going up" (Beer 1972: 69) will get them there. The differences between algorithm and heuristic were outlined by Brian Eno in his essay Generating and Organizing Variety in the Arts as a means to help distinguish the differences of approach between traditional western art music and Experimental music. His statements serve as an excellent pivot-point in the history of Generative music.

Experimental music emerged from New York in the 1950s as "Sound come into its own" (Cage 1973: 68). In this movement, musicians John Cage, Morton Feldman, Earle Brown, and Christian Wolff shared a common determination:

Cage and those listed above were significant in getting the ideas behind Experimental music started. Others working in England, such as Cornelius Cardew, Gavin Bryars, and John Tilbury, were able to move some of these ideas out of traditional "music" performance venues and into art schools, galleries, and other accessible public places (Nyman 1999). Brian Eno, then an aspiring art student who experienced this dissolution of austerity first-hand, comments that this music was "…explicitly anti-academic…" in order to counter the more cerebral serial music of Stockhausen and Boulez that was currently the rage with other students from the nearby music college (Nyman 1999: xi). In some cases these works were admittedly written for non-musicians (Nyman 1999), but this music was in no way overly simple or childlike. Rather, as Cage expressed, it granted sound, performers, and listeners a great deal of freedom. The process of creation and the process musicians and audiences experienced during a performance was far more important than any sort of artifact or product.

Terry Riley's In C (1964) is a seminal work in both the Experimental and Minimalist music traditions. The piece consists of 53 melodic phrases (or patterns) and can be performed by any number of players. The piece is notated, but was conceived with an improvisatory spirit that demands careful listening by all involved in the performance. Players are asked to perform each of the 53 phrases in order, but may advance at their own pace, repeating a phrase or a resting between phrases as they see fit. Performers are asked to try to stay within two or three phrases of each other and should not fall too far behind or rush ahead of the rest of the group. An eighth note pulse played on the high C notes of a piano or mallet instrument helps regulate the tempo, as it is essential to play each phrase in strict rhythm (Riley 1964).

The musical outcome of In C is a seething texture of melodic patterns in which phrases emerge, transform, and dissolve in a continuous organic process. Though the 53 patterns are prescribed, the choices made by individual musicians will inevitably vary, leading to an inimitable version of the piece every time it is performed. Riley's composition reflects one of John Cage's thoughts on Experimental music, when he writes that the "experiment" is essentially a piece of music: "the outcome of which is unknown" (Cage 1973: 13). When performed, In C has indefinite outcomes and yet—like wind chimes—is always recognizable as In C due to the character of the musical material and directions that comprise the work.

Free and non-idiomatic improvisation (Bailey 1992), games-based improvisation (Zorn 2004), and other forms where personnel choices are enough to constitute a loose set of rules or organization (Warburton 2005) also share some common ideas with Experimental music. But in these forms the machines are human, and far more complex than anything discussed here. While algorithms can be a part of Generative music, there is often more involved in the process. Strict algorithmic techniques certainly set the foundation for some of the qualities that make Generative music what it is, but the trial and error of a heuristic approach also has great value while the seed of a new work is being created.

The overall aesthetic of Generative music is much more consistent with the casual, open, and more restless attitude of Experimental music. As with most histories, the past of Generative music is fragmented—an amalgam of technologic possibilities and musical aesthetics. Like the music itself, once these ideas have been blended together, the process of unfolding continues. Of particular interest to this thesis is how this history has affected the career of Brian Eno and the contributions he has made that allow this work to go forward in new directions.

brian eno & his contemporary musical machines

Through his recording studio interventions and Oblique Strategy cards, Brian Eno can be credited with transplanting Experimental techniques into Art Rock and popular music (Sheppard 2008). His tape delay experiments with Robert Fripp (1973-1974) and later solo project Discreet Music (1975) furthered an ongoing musical investigation into processes and systems, and among other circumstances, led Eno to pioneer Ambient music. Eno's first "official" Ambient album, Music for Airports (1978) makes use of tape phase techniques similar to those of Steve Reich in his pieces It's Gonna' Rain and Come Out. Eno commented that It's Gonna' Rain is:

The idea of a music-making machine has fascinated Eno throughout his career (Eno 1996; Darko 2009). Tape delay and phase systems mark the beginningof this ongoing process in his body of work. In the years that followed, while working in a more conventional art setting with video and light installations, he used looping cassette tapes to create a continuous, ambient sound world that could be heard throughout the environment where his work was experienced. With the standards set by contemporary computers this seems primitive, but it echoes the incredible "variety through simplicity" idea in Reich's work that was so initially inspiring.

The term generative is most closely connected to Eno's work with SSEYO Koan, a software application for composing Generative music. Eno's only Koan-based album was titled Generative 1 (Eno & SSEYO 2006). Koan (and Noatikl, its successor), have not gained much popular momentum as music production tools, but the techniques and aesthetics these applications enabled has come to be known broadly as Generative music. In the way that the name Experimental music contains a raison d'être, so does Generative music:

Generative music uses algorithms, but it is not Algorithmic music. For instance, an algorithm used by Eno in Music for Airports follows the instruction: play the note C every 21 seconds. The term generative communicates the idea of a computer processor (or other machine) doing the work to compute the iterative sequences, shifting permutations, and random routines that can also be used in this music. Unlike much Algorithmic music that focuses on producing a final, notated output or a music that can authentically match a predetermined style (Cope 2000), Generative music can be compared more aptly to a seed. Brian Eno has made this metaphor in talks and interviews, (Eno & Wright 2006; Toop 2001) saying that like a seed, something unique will grow out of this music. Neither the listener nor musician knows exactly what it will be, but just as one would not expect a daisy to sprout from tomato seeds, each has a general idea or range of expectations. A tomato plant will sprout, but it won't look exactly like the one next to it or like any others in the row. The generative musical experience, like Beer's destination at the misty mountain top, is there, but the details are uncertain and the ensuing journey rich in possibility.

This essay is excerpted from Amergent Music: behavior and becoming in technoetic & media arts

references

Bailey, D 1992, Improvisation: its nature and practice in music, Da Capo, New York.
Beer, S 1972, Brain of the Firm: the managerial cybernetics of organization, Allen Lane The Penguin Press, London.
Cage, J 1973, Silence: lectures and writings, Wesleyan University Press, Middletown, Conn.
Cope, D 2000, The Algorithmic Composer, A-R Editions, Inc., Madison, WI.
Darko, J, 25 July 2009, 'Artscape - Brian Eno in Conversation' (Video). Available from: <http://www.vimeo.com/5763066>. [5 July 2010].
Dorin, A 2001, 'Generative Processes and the Electronic Arts', Organised Sound, vol. 6, no. 1, pp. 47-53.
Eno, B 1976, Generating and Organizing Variety in the Arts, Available from: <http://www.ziorjen.com/archives/geo/eno.pdf>. [26 June 2009].
________. 1996, A Year With Swollen Appendices, Faber & Faber, London.
Eno, B & SSEYO 2006, Generative Music 1 by Brian Eno, SSEYO, Available from: <http://web.archive.org/web/20060323154716/www.sseyo.com/products/koancontent/genmus1.html>. [16 July 2010].
Eno, B & Wright, W 2006a, Playing with Time (audio recording), Long Now Foundation, San Francisco.
________. 2010, 'Will Wright and Brian Eno Play with Time' (video). Available from: <http://fora.tv/2006/06/26/Will_Wright_and_Brian_Eno>. [2 September 2010].
Nyman, M 1999, Experimental Music: Cage and beyond, 2nd edn, Cambridge University Press, Cambridge, U.K.; New York.
Riley, T 1964, In C, Other Minds, Available from: <http://www.otherminds.org/shtml/Scores.shtml>. [24 May 2005].
Roads, C 1996, The Computer Music Tutorial, MIT Press, Cambridge, Mass.
Sheppard, D 2008, On Some Faraway Beach: the life and times of Brian Eno, Orion Books, London.
Tamm, E 1995, Brian Eno: his music and the vertical color of sound, Updated edn, Da Capo Press, New York.
Toop, D 2001, 'The Generation Game', THE WIRE, no. 207, pp. 39-45.
Warburton, D 2005, 'Les Instants Composés', in Blocks of consciousness and the unbroken continuum, eds B Marley & M Wastell, Sound 323, London, pp. 96-116.
Zorn, J 2004, 'The Game Pieces', in Audio Culture: readings in modern music, eds C Cox & D Warner, Continuum, London, New York.