A 365-Day Project

"We Are All Mozart"

A project to create
new works and change
the perception of the
music of our time.

Mantra Canon score pages Dennis

previous   October 13, 2006   next

Friday the 13th. Chance events. Strange dreams. Unexpected messages.

Philosophers for millennia have tried to plumb the source of creativity. What intrigues me is how we essentially identical bags of mostly water follow so many paths but arrive at the same activity -- to start the vibration of air that will tickle the hairs just inside the heads of other bags of mostly water. This is more than a 'where do you get your notes' question, but goes to the broader notion of how a composition concept comes into question, what materials we choose to answer it, and what methods we lean on for disentangling ourselves from the inevitable traps.

This thought path was triggered by an email from Alberto Vignani, a composer whose work I have heard many times online. On his website he writes:

I have of course also written music "by hand", and then used the computer only to render it. ... The reason why I started writing my own composers tool [...] was on one hand to experiment with generation and manipulation of musical structures at the software level. On the other hand, I wanted to automate as much as possible the generation of orchestral music by quickly creating lots of raw materials on which to choose/build later. To be fair, using algorithms to compose and samples to perform is not very unconventional today but one always hope to do things nobody else has already done.

Nick Didkovsky is another composer who uses this sort of 'assistive technology' to generate raw material. It's an excellent route because the parameters chosen can either reflect one's overall taste or challenge it, depending on the mood of the day, and the arid and fruitless paths one might follow with great effort (and so become psychologically invested in) can be abandoned without regret. Notebooks that might be filled with unsatisfactory sketches are instead electrons that vanish with the delete key.

Two anecdotes come to mind. David Gunn and I are both proud typists, but for some years he had a typing mania. He created stories and essays whose lines were exactly the same length, justified by verbal choice in the days of manual typewriters without an easy facility for right-justification that we take for granted on every bargain-basement computer. The improvised automatic-writing justified-typing constructions were marvelous:

With a friendly cup of tweedy tea at the ready, I gaze into a mirror,
but the reflection is empty, almost abandoned, and not at all like me

The corset is a wonderful device to combat one's accidental overstep-
ping of those predestined boundaries which gives mellowing a bad name

Incongruous patronage from a pelican sends Pavlov to his room ranting
about calumnicus man and how he'd like to share his grief with anyone

Bojoho, another cat of Penny's which is also gray, lolls under violet
skies, as the world comes unceremoniously to an end, and so does this

In the late 1970s, we were working together on tales he had written for the Loch Duck Radio Players, and each time there was a rewrite, David would take the script to his studio and retype it so it would be clean and perfect. It was a marvelous madness, but aside from the sense of clarity it afforded to his view of his writing, it was an enormous waste of time for the rest of us as the tape recorder sat on 'pause'. It was not long afterward that David finally capitulated to the word processor and its ability to format each page perfectly. Conceived with writing and formatting integrated, the commercial word processor was from the start a flexible working environment. My first one was The Electric Pencil in 1978, which, by adding only formatting codes that resembled SGML tags, became a typesetter (in the days before soft fonts).

The second anecdote (and really, I will get to the point) was a composition that began as dripping water in the night. The relationships among the dripping water pitches sorted themselves out, and pages of handwritten Markov chains were the first result. Clarence Barlow helped out with the computerized derivation of more pages of pitches based on those relationships, but then I set to work writing Dripworks, an orchestral work. It was being created in pen & ink. Some forty pages into the score, it became surreally clear that the piece was meaningless -- and poorly written. The future solution would seem to be a music program. But which one? There are engraving programs and performing programs and sequencers, and no integrated working environment. Each toolkit grew up in its own dominion.

Because of this dominion-dominated field, the workflow is slowed -- but at the same time (yes, I'm getting there now!) its very fragmentation encourages composer-programmers and their kin to create new tools. Automatic 'poetry' generators like my own jophxo from 1978 were a novelty (succeeded by story generators and interactive fiction), but have (yet) contributed little to what is essentially a linear artform. But music occupies multiple simultaneous spaces of time and meaning, and unlike words whose grammar and syntax are not (pace, Gertrude Stein and E. E. Cummings) dominated by the arbitrary and the personal, musical meaning's abstraction levels are deeper and more eluctible than words, opening the door to individual arbitrariness.

On CNN today, David Cope was interviewed about algorithmic analysis of music (or however much can be covered in three minutes). It is interesting to consider how much easier it is to analyze and recreate verbal styles than musical ones, at least from an automated perspective. Our human analytical engine happily absorbs and recreates other styles; not so well does the digital device accomplish this, however effective David's approach may be with a sufficient body of consistent and coherent material.

Which returns to the moutons of source. If musical expression contains significant arbitrary drive, then arbitrary sourcing can be rewarding for creation. Cage understood this but perhaps not all of its implications within a non-Zen world, where the arbitrary or chance source can either contain the inherent motion of the universe (if that holds validity for the composer) or a narrowable choice set like the knurled ophthalmoscope knob tweaked until the image comes into focus.

So Beethoven may have found his scene in the brook or Stockhausen his formants in the Swiss mountains (or I may have tried to discern a composition by splattering fountain pen ink on dozens of sheets of music paper), but once digital tools were at hand, sources could be found within by analyzing photographs, calculating fractals, or observing the conchoid of Nicomedes. Programs such as The Voice convert image to sound for what Peter Meijer calls 'synthetic vision'. Midimage has control over levels of color, rhythmic sets, voices and repetitions by means of seed numbers. Alberto Vignani's ismidi extracts features from an existing file or pattern, and Clarence Barlow's Autobusk creates probabalistic Midi by a set of twelve variable parameters. There are hundreds more, dating from Karlheinz Essl's Lexikon-Sonate back before Charles Dodge's computer punch cards for the creation of Earth's Magnetic Field thirty-six years ago to the indeterminacy in Cage's Music of Changes in 1951 -- even if one fails to include medieval and ancient numbers compositions calculated by hand.

My wife Stevie and I have argued these points. As a midwife, her sense of humanism is highly refined and deeply felt. Working with generated sources strikes her not only as abdication of human commonality but also deeply male. Clues and cues that are offered by life are ignored for the twist of the knob or the punch of an <enter>. The mind and spirit are a far richer source, she would argue, if we would only attend to them with open eyes and heart. Change is unnatural, I would argue, and sometimes we need help, whether it's a brook or mountains or splattering ink or algorithms.

Sure, composition sourced in these arbitrary ways has the tendency to sound all the same -- like, say, minimalism or Baroque music or Gregorian chant or Classical-era music. But the essense for me comes from ways of inducing change, wherever they are found. And yes, sometimes it is easier to hand off control to a machine, but if a robotic Monk amanuensis -- abbey or Adrian -- were the alpha and omega of algorithms, there would be little gained but full-speed driving toward mediocrity, save that raising the level of noise may induce madness. No, it's about the undiscerned pattern, the unexpected mutation, the uncharacteristic sense and how to feel it instantly as it snaps our heads back. The ears and the mind and the memory form a marvelous and unpredictable machine themselves, influenced by parameters unknown to the digital computers.

In other words, as the computer streams a thousand suggestions our way, we do what the best-practice enterprise management programmers would call exception management. Artists are the best best-practice exception managers, finding exceptions in common human practice, the tears in the fabric of life, and from them weaving with new colors and textures and patterns, and in doing so, we disentangle ourselves from the seemingly inevitable to create a new inevitability.

mm. 273-279 from Dripworks
Measures 273-279 from the abandoned Markov-based orchestral composition Dripworks (1992).

Back to the Blog Index
Back to the WAAM Page
Back to my Home Page
Please Write to Me
Previous Day | Next Day

WAAM Info Feed RSS feed for this site