AUTOMATION IS MY SALVATION:

EIGHT STUDIES FOR AUTOMATIC PIANO

Thesis

Submitted in partial fulfillment of the

requirements for the

Degree of

Master of Fine Arts in

Electronic Music and Recording Media

Mills College, 2010

By

Seth Horvitz

Approved by:

_______________________________
James Fei
Director of Thesis

_______________________________
Chris Brown
Head, Music Department

__________________________
Sandra C. Greer
Provost and Dean of the Faculty
Reading Committee:

______________________________
Chris Brown

_______________________________
Fred Frith
Table of Contents
1. INTRODUCTION ........................................................................................................ 6
2. HISTORY ...................................................................................................................... 6
2.1 Prehistory ............................................................................................................................. 6
2.2 Rise and Fall ......................................................................................................................... 8
2.3 Rediscovery......................................................................................................................... 10
2.4 Nancarrow Considered...................................................................................................... 11
3. INFLUENCES............................................................................................................. 13
3.1 Tenney................................................................................................................................. 13
3.2 Ligeti ................................................................................................................................... 14
4. REFLECTIONS.......................................................................................................... 17
4.1 Disklavier vs. Player Piano ............................................................................................... 17
4.2 The Sound of the Piano ..................................................................................................... 17
4.3 The Basic Shape ................................................................................................................. 18
4.4 The Harmonic Series ......................................................................................................... 18
4.5 Hand-made Algorithms..................................................................................................... 19
4.6 Time..................................................................................................................................... 19
4.7 Emergence .......................................................................................................................... 20
4.8a Human Music, Living Music........................................................................................... 21
4.8b Living Music, Live Music................................................................................................ 22
4.9 Electronic Music ................................................................................................................ 22
4.10 Immersion......................................................................................................................... 23
4.11 The Visual......................................................................................................................... 23
4.12 The Concert ...................................................................................................................... 24
5. DESCRIPTIONS ........................................................................................................ 25
CLASS A: IDEALIZED SYMMETRICAL FORM [14, 4] ................................................. 26
CLASS B: CONSTRUCTED BINARY FORM [1, 2] ......................................................... 26
CLASS C: INTUITIVE LINEAR FORM [21, 99] ............................................................... 26
CLASS D: INTUITIVE TRANSFORMATIONAL FORM [13, 29] ................................... 26
Note on order and numbering:............................................................................................... 27
5.1 Study No. 14: Arch Study for the Highest Eight Notes .................................................. 27
5.2 Study No. 4: Sixteen Diatonic Glissandi Moving at Harmonic Rates ........................... 29
5.3 Study No. 2 : An Approximate Series of Approximate Harmonic Series..................... 32
5.4 Study No. 13 : Echoes ........................................................................................................ 36
5.5 Study No. 21 : Bells ............................................................................................................ 39
5.6 Study No. 1 : Octaves, Systematically Filled and Folded ............................................... 39
5.7 Study No. 29 : Tentacles .................................................................................................... 42
5.8 Study No. 99 : Strumming Machine................................................................................. 45
6. CONCLUSION ........................................................................................................... 47
APPENDIX A : OVERVIEWS...................................................................................... 49
APPENDIX B : PHOTOGRAPHS................................................................................ 54
BIBLIOGRAPHY ........................................................................................................... 56
 EVERYONE SHOULD BE AWARE:
THE BEST TRADEMARK IN THE WORLD – IS ‘LIFE’

Beware of Forgeries!

-Sergei Yutkevich, The Eccentric Manifesto, 1922

 6

1. INTRODUCTION
The main purpose of this thesis is to illuminate the methods, processes,

philosophy, and influences at work in Eight Studies for Automatic Piano, composed for

the Yamaha C7 Disklavier® Mark III. Given the unique historical role of this instrument,

the discussion will be preceded by a brief overview of the history of mechanical music,

up to and including its earlier cousin, the pneumatic roll-operated player piano. This

history will highlight the aims of composers, especially those who saw the "potential to

create rather than recreate music"1 with mechanical instruments, culminating in a brief

consideration of Conlon Nancarrow's work. The second part of the thesis will discuss two

works that have been pivotal in my approach to the "automatic" piano: James Tenney's

Spectral Canon for Conlon Nancarrow (1974), and György Ligeti's Continuum for

Cembalo (1968). This section also considers the influence of Ligeti’s dynamic,

interactive approach to generating form, specifically in regard to his Études for piano. A

third section provides a series of personal reflections and mini-essays regarding my

working process, considering both the pragmatic and philosophical implications of the

ideas at hand. Finally, a detailed description of each work will be provided.

2. HISTORY
2.1 Prehistory

As Arthur W. J. G. Ord-Hume, scholar of mechanical music, notes: "The

production of music without attendant performing skills is no recent phenomenon. From

the earliest times, the creation of a musical instrument went hand-in-hand with the

1
Henry Wong Doe, “Musician or Machine: The Player Piano and Composers of the Twentieth Century,”
(Ph.D. Diss., The Juilliard School, 2006), 2.
 7

creation of a non-participant means of playing."2 Throughout their history – a history

longer than popular knowledge might lead us to believe – automated musical instruments

have rarely provided creative inspiration for composers. Prior to the 20th century, with

few exceptions, these musical machines served either as a method of documentation and

reproduction, or as a novel way to "perform" existing musical compositions. Indeed, the

magic of seeing a machine play "by itself" was often a large part of the attraction.

Secondary sources refer to the existence of an entire mechanical orchestra in the

third century B.C., built for the emperor of China during the Han Dynasty,3 though few

details are known. Later inventions commonly made use of a revolving barrel or cylinder

that could be "programmed" by manually arranging a set of wood or metal pins on its

surface. The earliest documented instrument of this type was a water-powered barrel

organ named "Banu Musa," dating from about 875 AD,4 while the oldest functioning

instrument to survive dates from 1502, and is currently on display at the Burgmuseum in

Salzburg. Over the centuries, artisans and engineers developed countless variations of

this basic cylindrical device, providing instruments for street performers as well as fine

collectibles for royalty.

In the 18th century, mechanical organ and musical clock builders worked closely

with composers such as Haydn, Mozart, and Beethoven, "although it is difficult to say

whether they were personally interested or rather obliged by their patrons to write for

2
Arthur W. J. G. Ord-Hume "Cogs and Crotchets: A View of Mechanical Music," Early Music, 11.2 (Apr.,
1983): 168.
3
See Roy Porter, “automata,” <http://www.answers.com/topic/automata-theory> (6 December 2010) and
press release from “Automata: Contemporary Mechanical Sculpture,” Chazen Museum of Art
<http://www.isthmus.com/theguide/details.php?event=233789> (6 December 2010), among many others.
4
Charles B. Fowler, “The Museum of Music: A History of Mechanical Instruments,” Music Educators
Journal, 54.2 (Oct., 1967): 45.
 8

these instruments."5 Haydn clearly considered the creative potential of these instruments,

writing at least thirty original works exclusively for mechanical organ, yet his

compositions do not vary greatly from his other work apart from the use of larger leaps

and faster embellishments6.

While much of the music written for early mechanical instruments has been lost,

we can be fairly certain that composers of the time did not intend to invent new forms or

styles of music around these instruments, but instead made use of instruments’

capabilities to either extend the virtuosity of human performance or to achieve a more

precise document of their work than sheet music could provide. These aims persisted

until the early 20th century, when a handful of composers first began to consider a music

that would derive its creative potential directly from the idiosyncrasies of the machine.

2.2 Rise and Fall

First introduced in the late 19th century, the pneumatic roll-operated player piano

ushered in a new era for the mechanical instrument, but its rapid rise and fall, in addition

to heavy resistance from detractors, preempted composers from fully exploring its

creative potential. The player piano possessed expressive and technical capabilities far

surpassing anything that came before. Yet, perhaps even more than some instruments of

the 18th century, it was not invented with creative use in mind. Instead, its role in popular

culture can be more accurately described as a precursor to the phonograph, and by

extension, the entire recording industry. The instrument marks the first effort to mass

market existing musical compositions for playback within the home, and for this purpose,

it achieved an astounding level of commercial success. By 1925, sales of player pianos

5
Wong Doe, 9.
6
Ibid., 8.
 9

had overtaken those of standard pianos.7

The popularity of the instrument piqued the interest of several prominent

composers of the time, including Paul Hindemith, Igor Stravinsky, Ernst Toch, and

George Antheil. For the first time, composers began to think seriously and deeply about

the machine’s idiosyncratic qualities, and how to exploit them for aesthetic purposes. The

excitement (and confusion) surrounding the player piano's new creative potential is

captured well in this review of a festival organized by Hindemith in 1926:

A brilliant Welte-Mignon played. IT played. The hall was bathed in light from
some invisible source, and you could have heard a pin drop as Hindemith adjusted
the levers of the instruments, which was to replay three worthy pieces by Ernst
Toch. The piano begins to play: musical studies for the instrument, toccatas with
row upon row of unplayable chords, with a velocity that the most accomplished
virtuoso could never even approach, with a precision far beyond the capability of a
human being, with a superhuman power of sound, with a geometrical clarity of
rhythm, tempo, dynamics and phrasing, which only a machine can really bring out.
What a wonderful box of tricks! The piano reaches the end of the composition and
the audience hesitates. Should one applaud? There is, after all, no-one sitting there.
It is only a machine. At last just a little applause, which becomes ever more
enthusiastic. Cries of “Encore!” And indeed, the piano repeats faultlessly, with the
same precision as the first time through.8

The excitement was high, but short-lived. Concerts of this kind were still

relatively rare, and reactions were often hostile, placing outspoken supporters like

Hindemith on the defensive9. Throughout the 1920s, Stravinsky expressed profound

interest in the creative potential of the instrument, yet he only managed to compose one

short and rather underwhelming solo piece for it, the Etude pour Pianola (1921). Antheil

called for the use of sixteen synchronized pianolas in his grand spectacle, Ballet

Mechanique (1924), but technical difficulties later forced him to re-orchestrate. With the

7
Arthur Ord-Hume, “Player Piano,” In The New Grove Dictionary of Music and Musicians, 2d ed. Edited
by Stanley Sadie, Vol. 25 (London: Macmillan Publishers Ltd., 2000).
8
Paul Hindemith quoted in Rex Lawson, “Hindemith: Toccata for mechanical piano.” Pianola Journal 9
(1996): 19-20.
9
Wong Doe, 46.
 10

onset of the Great Depression coinciding with the growing popularity of smaller and

cheaper devices such as the phonograph and radio, the player piano experienced a sudden

and rapid commercial decline, and the instrument’s ardent supporters quickly lost

interest.

2.3 Rediscovery

Nearly two decades after the player piano's commercial demise, an American

composer by the name of Conlon Nancarrow, while living in Mexico as a political

refugee, decided to devote the better part of his life to developing a radically new

compositional framework around the instrument. Inspired by Henry Cowell and

Stravinsky's rhythmic ideas, Baroque counterpoint, canon form, and jazz, Nancarrow's

work covers an astonishing range, yet by sheer originality of approach (not to mention

the unique sound of his modified player pianos) his oeuvre carries an instantly

recognizable signature. While most of his work was composed in relative isolation, a

surge of interest in the 1980s has produced a wealth of scholarship, and today, he is

commonly cited as one of the most important composers of the 20th century.

Until he discovered the player piano, Nancarrow struggled to realize his ideas. His

earliest compositions for human performers were met with almost unanimous frustration.

The rhythmic and temporal relationships involved were too complex for humans of the

time to deal with, both mentally and physically10. After several disappointing rehearsals

and cancelled performances, Nancarrow discovered Henry Cowell's highly prescient

treatise New Musical Resources, which contained this offhand recommendation:

Some of the rhythms developed through the present acoustical investigation could
not be played by any living performer; but these highly engrossing rhythmical

10
Roger Reynolds, “Conlon Nancarrow: Interviews in Mexico City and San Francisco", American Music
2.2 (Summer 1984): 1-24.
 11

complexities could easily be cut on a player-piano roll. This would give a real
reason for writing music specially for player-piano, such as music written for it at
present does not seem to have, because almost any of it could be played instead by
two good pianists at the keyboard.11

In 1947, with the help of a small family inheritance, Nancarrow traveled to New

York from Mexico to acquire a player piano and a punching machine12 then proceeded to

spend the next several decades bringing the potential of Cowell's statement to life. While

living in relative isolation from the greater musical world, Nancarrow developed a

massive body of work, making exclusive use of the piano's mechanical properties to

realize complex temporal relationships. In doing so, he completely overshadowed the

relatively limited experiments of previous composers.

2.4 Nancarrow Considered

Although a full analysis of Nancarrow's work is beyond the scope of this thesis,13

his compositional techniques—explained here—serve as a reference point for Eight

Studies for Automatic Piano.

Rather than compose with melody, harmony, or even sound per se, Nancarrow

used time itself, in fact multiple times, as the structural building blocks for his

compositions. The paradox here is that time, in order to be experienced musically, must

be realized as sound—in this case, the sound of a piano. And in this regard, Nancarrow

was never entirely satisfied:

Actually, at the beginning what I would have liked to have had would have been a
harpsichord, a mechanical harpsichord 'cos that's the sound I like because of the
clear divisions. A nineteenth-century grand piano is not built for contrapuntal

11
Henry Cowell, New Musical Resources. (New York: Cambridge University Press, 1996), 64.
12
Wong Doe, 61.
13
For in-depth analysis, the reader is directed to Kyle Gann's nearly definitive 1995 volume The Music of
Conlon Nancarrow.
 12

music. It's built for rich, harmonic sounds.14

In an effort to create a sharper, more discrete sound from his piano, Nancarrow

experimented with various modifications, eventually settling on a solution that involved

attaching steel and leather straps to the hammers. While his experiments with timbre were

fairly extensive, they were less about exploiting the sonic potential of the piano and more

about the composer's "desire for clarity and precise definition of notes within the

texture."15 Without such clarity, the distinct temporal entities in Nancarrow's work would

meld together, a result he definitely wished to avoid.

Seeking to explore "temporal dissonance" from the start,16 Nancarrow's first study

was a two-part canon setting tempos of 210 and 120 against each other (a ratio equal to

7:4, or a purely tuned minor seventh in the realm of pitch). Over the years, he explored

increasingly complex ratios and rhythmic relationships, culminating in the use of the

irrational numbers e and pi in Studies 40 and 4117. Nancarrow sometimes spent a year or

more punching a single piano roll before he was able to hear the results of his

calculations. As he himself noted, the element of surprise was an exciting part of the

process.18

While respecting the tremendous scope and influence of Nancarrow’s creative

vision, the player piano works discussed here do not emulate his compositional style. In

contrast to Nancarrow’s work, the material in Eight Studies for Automatic Piano

14
Conlon Nancarrow, interview by Natalie Wheen, In The Pianola Journal: Journal of the Pianola
Institute. West Wickenham, Kent, England: Vol. 3 (1990): 6.
15
Wong Doe, 64.
16
Ibid., 68.
17
Ibid., 70.
18
“As a matter of fact, after I finish punching a piece and before I put it on, you have no idea how excited I
am. What is going to happen? The first time I listen, it's just a matter of how well I did what I thought I was
doing. Naturally, the more it comes out the way I thought I did it, the better I feel. It's as simple as that.”
(Reynolds, 16).
 13

originates from extremely simple temporal building blocks and a minimum of advance

planning. Rather than seeking absolute clarity of articulation, Eight Studies embraces the

piano’s built-in palette and attempts to craft new sonorities from the perceptual fusing of

different “lines.” The following section will discuss works by two composers that have

been more important than Nancarrow in shaping Eight Studies. Both composers praised

Nancarrow, but departed from his compositional techniques in significant ways.

3. INFLUENCES
3.1 Tenney

The impetus for composing Eight Studies for Automatic Piano came as a direct

result of hearing James Tenney's Spectral Canon for Conlon Nancarrow (1974). Tenney's

piece sounded completely alien to me, almost unreal, and certainly unlike any other piano

music that came before it, including Nancarrow's. What struck me most was just how

easy it was to forget that I was hearing a piano, or for that matter, a composition at all.

The piece sounded more like a "fact of nature,"19 but one that constantly confused, rather

than confirmed my intuitive sense of time. Despite an obvious and precise pulse, the

piece draws attention to various perceptual entities that emerge from a process, entities

that stretch, multiply, and overlap in several directions at once.20

If the above description sounds suspiciously like Steve Reich's idea of "Music as a

Gradual Process,”21 it comes as no coincidence; Reich even quoted Tenney in his seminal

essay. Reich stated that "whether a musical process is realized through live human

19
Larry Polansky, “The Early Works of James Tenney," Soundings #13 (1983): 225.
20
For further discussion of Spectral Canon, see section 5.2 (Study No. 4).
21
Steve Reich, “Music as a Gradual Process,” in Writings on Music 1965-2000 (New York: Oxford
University Press, 2002), 34.
 14

performance or through some electro-mechanical means is not finally the main issue."22

However, after the disappointing 1969 premiere of his “Phase Shifting Pulse Gate” – an

electronic instrument developed over the course of a year in collaboration with engineers

from New Jersey Bell Laboratories23 – Reich abandoned automated music altogether. He

declared: "In any music which depends on a steady pulse, as my music does, it is actually

tiny micro-variations of that pulse created by human beings, playing instruments or

singing, that gives life to the music."24 By contrast, Tenney's piece depends absolutely on

the precision of the machine for its "life." As noble as the challenge to "play" it might be,

any variations introduced by a human performer would simply muddy the experience.

Inspired by Tenney, I set out to make piano music that depends entirely on the

precision of the machine—music that deliberately avoids the micro-variations introduced

by human performers, yet is nonetheless full of "life." Rather than relying on human

hands for "micro-variation," Eight Studies exploits those micro-variations that exist in

acoustic sound itself, and in human perception, in a way that requires machine precision

to be realized. The machine, working hand in hand with various compositional processes,

becomes an equalizer; if we know that the music is being reproduced with exact,

superhuman precision, we are free to focus our perception on other aspects of the

resulting sound. In deference to Reich's original essay, this music shifts the listener's

attention "away from he and she and you and me outwards towards it."25

3.2 Ligeti

22
Ibid., 34.
23
K. Robert Schwarz, “Music as a Gradual Process Part II", Perspectives of New Music, 20.1-2 (Autumn,
1981 - Summer, 1982): 227.
24
Ibid., 228
25
Reich, 34.
 15

Remarkably, the second work to directly influence Eight Studies for Automatic

Piano was written for a human harpsichordist. Hungarian composer György Ligeti's

Continuum for Cembalo (1968) embodies his lifelong fascination with malfunctioning

machines, yet it was composed more than a decade before his own revelatory discovery

of Nancarrow.26 While Nancarrow went to great lengths to make his mechanical piano

sound more like a harpsichord in an effort to achieve clearer divisions between notes,

Ligeti sought to create a continuous, gradually shifting texture from the harpsichord, and

to use a human performer to bring out the instrument’s machine-like qualities:

It had never occurred to me to write for harpsichord, but… it suddenly came to me

that the harpsichord was really like some strange machine… I also remembered
that a harpsichord was most typically an instrument with a non-continuous sound,
the twang of the string is of short duration, followed by silence. I thought to
myself, what about composing a piece that would be a paradoxically continuous
sound...but that would have to consist of innumerable thin slices of salami? A
harpsichord has an easy touch; it can be played very fast, almost fast enough to
reach the level of continuum, but not quite (it takes about eighteen separate sounds
per second to reach the threshold where you can no longer make out individual
notes and the limit set by the mechanism of the harpsichord is about fifteen to
sixteen notes a second). As the string is plucked by the plectrum, apart from the
tone you also hear quite a loud noise. The entire process is a series of sound
impulses in rapid succession which create the impression of continuous sound.27

Eight Studies for Automatic Piano reflects a similar child-like fascination with machines,

human perception and the sonic characteristics of instruments, including those aspects

that are normally thought of as superfluous or undesirable.

In Continuum, along with other pieces in his so-called "pattern-meccanico"

26
“After the few player piano studies of Nancarrow I listened to, I affirm with all my serious judgement
that Conlon Nancarrow is the absolutely greatest living composer.” (György Ligeti to Mario di
Bonaventura, letter from 28 June 1980, quoted in Jürgen Hocker, “My soul is in the machine: Conlon
Nancarrow” In I sing the body electric: Music and technology in the 20th century (Hofheim: Wolke, 2000),
94.)
27
György Ligeti, Ligeti in Conversation (London: Ernst Eulenburg Ltd., 1983), 31.
 16

style,28 Ligeti creates slowly evolving textures from rapidly repeating rhythmic figures—

as he describes them, periods of "mistiness" contrasted with periods of "clearing up."29

The listener's perception of the passing of time is shaped by the rate of change of the

rhythmic figures rather than tempo or meter. Similar, though not identical techniques are

employed in two of my Eight Studies, Nos. 13 and 29.

Ligeti’s dynamic, interactive approach to generating form has provided additional

inspiration for Eight Studies for Automatic Piano. Even when there is a strict conception

guiding Ligeti’s work, the material goes through extensive testing and modification

before reaching its final state30. The role of the composer, then, is to act as a dynamic

intermediary between form and content. The best way to describe this working method is

that of a continuous feedback loop between the composer, the material, and the

instrument. My relationship to the piano is strikingly similar to Ligeti’s, with the vital

insertion of a computer added to the loop:31

I lay my ten fingers on the keyboard and imagine music. My fingers copy this
mental image as I press the keys, but this copy is very inexact: a feedback emerges
between idea and tactile/motor execution. This feedback loop repeats itself many
times, enriched by provisional sketches: a mill wheel turns between my inner ear,
my fingers and the marks on the paper. The result sounds completely different from
my initial conceptions: the anatomical reality of my hands and the configuration of
the piano keyboard have transformed my imaginary constructs. In addition, all the
details of the resulting music must fit together coherently, the gears must mesh.
The criteria are only partly determined in my imagination; to some extent they also
lie in the nature of the piano – I have to feel them out with my hand.32

28
Jane Piper Clendinning, “The Pattern-Meccanico Compositions of Györgi Ligeti,” Perspectives of New
Music, 31.1 (1993): 192.
29
Ligeti, 1983.
30
One notable exception is Poeme Symphonique (1962), a Fluxus-inspired "happening" for 100
metronomes.
31
Ironically, the description refers to Ligeti's Études pour Piano, a series of works for human performer,
inspired directly by Nancarrow's machine music.
32
György Ligeti, CD liner notes, Gyorgy Ligeti Edition 3: Works for Piano, SK 62308, 1996.
 17

4. REFLECTIONS
4.1 Disklavier vs. Player Piano

Eight Studies for Automatic Piano was composed specifically for the Yamaha C7

Disklavier Mark III, a modern-day player piano. An important difference between the

Disklavier and its earlier cousin is that it can be controlled directly and immediately by a

computer using the MIDI (Musical Instrument Digital Interface) protocol, rather than

requiring the composer to punch a paper roll in advance. This affords the opportunity to

engage in an intimate, dynamic relationship with the instrument, instantly hearing and

evaluating even the smallest modifications to a work.

The Disklavier cannot exceed sixteen simultaneous notes, but I found this

limitation to be more of a blessing than a curse. It provided a nice round number to guide

the compositions, a number that is handily divisible into four groups of four, or two

groups of eight. I also came to realize that from the listener's perspective, the difference

between sixteen simultaneous notes and eighty-eight simultaneous notes is not as great as

it might seem on paper.

4.2 The Sound of the Piano

Conlon Nancarrow never really liked the sound of the piano that much. He felt it

was, in some ways, inappropriate for his work and tried hard to make his instrument

sound more like a harpsichord. James Fei, after a few meetings discussing my piano

music, revealed that he did not actually like the sound of the piano either, mainly for its

weighty historical associations. I happen to like the sound of the Yamaha C7 Grand Piano

Disklavier very much, although I am also interested in finding ways to make it sound

unlike a piano, without resorting to the use of leather strips, metal screws, or ping pong
 18

balls.

4.3 The Basic Shape

I prefer to start with something simple: a set of octaves, a repeating note, a steady

glissando, occasionally a very short melody, maybe an interval or two or a rhythmic

shape. I make everything else by repeating this "basic shape," transposing it, copying it,

rhythmically displacing it, and layering it. Every once in a while, I push something in or

out of place so that it sounds better, or different. This is a little bit like Arnold

Schoenberg's notion of the “grundgestalt,”33 though I doubt he would have liked this

music very much.

4.4 The Harmonic Series

A string, pulled taut (as on a piano), has the natural tendency to vibrate at a rate

proportional to its length. It also has the tendency to vibrate at twice that rate, three times

that rate, and so on. Each successive "harmonic" will sound a little closer in pitch to the

one that preceded it, and will generally be a little quieter, with variations determined by

the physical material of the string, the material it is connected to, and the surrounding

acoustic space.

While I am interested in the acoustic principles of the harmonic series and often

take them into account when arranging notes, I tend to apply the idea in the way a toddler

might, one number at a time, counting in order. Perhaps even more important to my work

than any specific scientific or acoustic implication is the revelation that the series, at its

core, is nothing more than the natural numbers: 1, 2, 3, and so on. And yet, from the

33
"Whatever happens in a piece of music is the endless reshaping of the basic shape... There is nothing in a
piece of music but what comes from the theme, springs from it and can be traced back to it; to put it still
more severely, nothing but the theme itself’ (Arnold Schoenberg, quoted in Michael J. Schiano,
“Grundgestalt,” in New Grove Dictionary of Music and Musicians (Vol. 25), 2nd ed., Stanley Sadie, Vol.
25, (London: Macmillan Publishers Ltd., 2000).
 19

simplest applications of this series to musical data (adding, subtracting, multiplying or

dividing), one can easily derive both astonishingly elegant complexity and

incomprehensible garbage.

4.5 Hand-made Algorithms

I have been told that my music is algorithmic, although I don't really think of it

that way. I don't use any math other than simple addition, subtraction, multiplication and

division. I copy something (often a repeating figure), paste it next to itself, and then

change it a little. Then I do it again and again, changing it by the same amount each time,

and listening all the while. If it doesn't sound good, I might just start over. Or I might

copy half of all the copies and put them somewhere else, then change that a little. Then

do that again and again (listening all the while). I might make something go backwards or

turn upside-down, each time listening to how it sounds. Sometimes, when I am away

from the piano, I get an idea for how to make a very interesting composition by doing a

slightly more complicated thing over and over, but only occasionally do these ideas turn

out to sound interesting. I avoid using equations, because I never want the music to get

too far away from my ear.

4.6 Time

I prefer to focus on simple temporal relationships, or at least closely related ones.

In other words, I don't want things to drift too far apart. I like to create tension between

the feeling of speeding up and the feeling of slowing down. For example, a single phrase

might consist of progressively longer notes (giving the impression of slowing down), and

that phrase could be repeated a number of times. Meanwhile, the overall tempo could

speed up. In cases like these, it is difficult for a listener to tell what is speeding up and
 20

what is slowing down, but both directions can be sensed simultaneously from the same

material. Another example to try: start with a steady pulse, gradually allowing it to slow

down. When the pulse has slowed to half its original speed, start another pulse at double

the speed (equaling the speed of the first pulse at the beginning). This process could

continue for quite a while, making for an interesting effect. Similar processes occur in the

traditional music of Indonesia and James Tenney's Spectral Canon for Conlon

Nancarrow.

4.7 Emergence

I don't so much want the listener to hear my compositions as I want them to hear

the sounds, patterns, textures, and forms that emerge from them, as a result of the

interactions between elements. If it is unclear whether these entities are forming in the

music itself or the mind of the listener, all the better.

The concept of emergence can be applied to simple ideas in tonal music, such as a

chord arising from multiple notes, or to aural phenomena such as acoustic beating and

difference tones. An example in the visual realm is the moiré pattern, exploited at length

in the Op Art of the 1960s and countless "psychedelic" videos. Moiré patterns can also be

observed in daily life, for example when light passes through folds in silk fabric, or in

nature, when the feathers of a parrot overlap. We tend to understand the world by making

connections between closely related things, but sometimes these attempts just makes us

more confused.

Often, as a result of the application of strict processes in my music, the visual

view of the notes resembles a moiré. I am especially excited if listening to the music

inspires a similar feeling. While I have no scientific explanation to prove it, I can be
 21

fairly certain that other human beings, outfitted with similar perceptual apparati, will

experience something similar to what I experience when listening, depending on the type

and amount of cultural baggage that goes along with it.

4.8a Human Music, Living Music

Music composed by humans or listened to by humans is always, by definition,

human, but the "life" of music need not resemble human form, nor must it be "warm."

Cold-blooded animals are alive, as are trees. One might also say the planet itself is alive,

even the cosmos. Why stop at a scientific definition of life when evaluating something as

non-scientific as music? The life of music depends on human life; nonetheless, it cannot

be sufficiently described in terms of its resemblance to humans or other known creatures.

With this statement, I do not mean to invalidate any resemblances (which can surely aid

in our appreciation and understanding of music), but to argue that the emergent life of the

work itself, the life that we can never adequately describe in words, is what paradoxically

confirms its identity as a "living" work of art.

Many people have argued that the "human element" is missing in automated

and/or electronic music not involving human performance, and that this is a bad thing.

Milton Babbitt had this to say in defense of electronic music without human performers:

Never has music been more human. It begins with a human composer who tries to
realize his conception of the musical work in every sense. Except for the
intervention of the speakers, the only limitations on the other side are those of the
human perceptor.34

Babbitt is right in one sense, music without human performers provides a more direct

(but not necessarily better) way for a single composer to communicate with an audience

(that is, unless s/he assumes the dual role of composer and performer). With regard to the

34
Milton Babbitt, live interview by Charles Amirkhanian, 1984, transcribed by the author. Original audio
available at <http://archive.org/details/SOM_1984_11_15> (6 December 2010).
 22

human performer, I would argue that certain compositional approaches lend themselves

to interpretations by a sentient performer while other types simply do not. Both are

equally valid, and both are equally "human."

4.8b Living Music, Live Music

It is all too common that musicians attempt to bestow "life" upon their music by

finding a way to make it "live." We live in an age where most popular music is many

times removed from the classic, caveman's notion of live music, generated in real-time by

musicians. Perhaps it is time to admit that most of these efforts are just for show (and

there's nothing wrong with a good show).

Steve Reich has said that the life of his music originates in the "micro-variations"

introduced by human performers. While performers have certainly added life to his

music, I disagree that its life originates there. Admittedly, there are many cases where the

performance of music by humans causes a composer to rethink his or her approach —

indeed, reshaping the life of the music. There are also many cases where the life of music

is inexorably linked to human performance—for example, the African music that inspired

Steve Reich. In all cases, music “comes to life” when it is experienced by a human

listener, but the performer of the music need not be alive.

4.9 Electronic Music

My background is in electronic music. I started with synthesizers, drum machines,

turntables, and tape recorders, before pianos or even notes. Part of my approach to the

piano involves simulating certain common effects in electronic music such as the

crossfade, the echo, the loop, and the splice. It turns out this is not very difficult when

using a computer-based sequencer to compose.

 23

Where there are aggregates of notes in my work, I treat the piano like a sound

designer might approach a synthesizer, meticulously shaping attacks, sustains, decays,

and velocities in non-real-time, and paying close attention to the resulting acoustic

interactions, so that the boundaries between notes might either dissolve or combine to

create the perception of something new. In the end, it strikes me that the difference

between acoustic and electronic sound may be smaller (and more complicated) than it

seems.

4.10 Immersion

Generally speaking, I don't want to create a situation in which the listener must

mentally sift through the notes, apart from his or her direct experience, in order to

appreciate the work (although she or he is welcome to do so). Nor do I want the listener

to relate what they hear to specific emotional events in their life or the life of the

composer (ultimately, this cannot be controlled, but it is worth making an effort). Rather

than spending extra time sifting through the musical information on an intellectual level

or looking for easily recognizable feelings of nostalgia (a type of immersion I prefer to

avoid), the listener should attempt to become immersed in the impressions that the sounds

(and the accompanying visual patterns) make on their senses.

4.11 The Visual

Often, the music I make generates striking visual patterns when viewed as a

"piano roll." I use the computer's piano roll like a canvas, manually moving chunks of

notes around on a two-dimensional plane. Seeing the notes arranged in this way (as

opposed to traditional Western notation) helps me understand the structures that they

combine to create. In some cases, these structures are also clearly visible on the piano
 24

keys—always a pleasant surprise. Sometimes I have an idea that generates a beautiful

visual pattern on the piano roll, but when I send that pattern to the piano, it just sounds

like a bunch of notes. I save those ideas for later, but I don't let other people hear them.

4.12 The Concert

The idea of watching a piano play by itself on stage is still disconcerting to many,

even after all these years. Just a few years ago, Rex Lawson, pianola expert and personal

friend of Conlon Nancarrow, wrote:

It was clearly the memory of a lifetime to stand in Nancarrow's studio and marvel
at the sounds emanating at high volume from the nearby upright piano, no doubt
accompanied by a gleeful expression on the face of the composer. It is quite
another matter to sit in serried ranks in a smart concert hall and listen to a piano
with no-one sitting at the keyboard. Human audiences need human performers. For
five minutes it is quite fun to see the keys going up and down on their own, but the
excitement soon fades. Reproducing piano concerts need entertaining and careful
presentation. 35

While I agree that careful presentation is essential, I tend to disagree with this general

sentiment.

For the premiere of Eight Studies, I took special care to craft an immersive

concert experience, removing extraneous sensory input in order to draw attention to the

abstract nature of the piano keyboard and its associated sound. By controlling the lighting

with a single spotlight and providing a video projection of the light reflected on the keys,

I was able to display the keyboard as a simple row of lines and dots (see Appendix B for

photo documentation). I organized the program order so that it would be diverse enough

to avoid fatigue, yet clever enough to illustrate a compositional thread. At the beginning

of the performance, I entered the stage to press the “play” button on the Disklavier, but

this was my only intrusion into an otherwise entirely mechanical performance. A “live”

35
Rex Lawson, “Compositions for Pianola – Conlon Nancarrow” The Pianola Institute History Page,
<http://www.pianola.org/history/history_nancarrow.cfm> (6 December 2010).
 25

performance of this kind involves plenty of real-time interaction, but the focus of such

interaction shifts from the actions of a human performer to the connection between the

instrument/object/performer and the perceptual faculties of the audience. To some extent,

the audience is forced to confront their own opinions concerning the role of the

composer, performer, and themselves in a concert setting like this.

Mechanical musical instruments are able to repeat their performances ad

infinitum, with very little variation, and without asking questions. This fact makes it easy

to adapt such performances to an installation setting, where the audience is allowed to

come and go at will. In some ways, the concert setting is similar; we are also allowed to

come and go at will, but it is generally assumed that we will stay and experience the

whole thing together. In that sense, Eight Studies benefits from a hint of benevolent

subjugation.

5. DESCRIPTIONS
Ideally, the listener should be able to enjoy these works without a prior

understanding of the technical processes, technologies, or conceptual framework at play.

I hope the experience of listening will pique the listener's curiosity about these aspects

while creating a pleasurable confusion of the senses. While not a required precursor to

listening, the descriptions included here may enhance the understanding and appreciation

of the work.

The eight works presented here can be divided into four thematic classes, based

on approach and outcome. Regarding the application of transformational processes to the

musical material, the classes proceed from most to least strict. The level of strict process

is also reflected in the amount of technical versus non-technical language used in each

description. These classes do not constitute a pre-ordained strategy, but were formed only
 26

in retrospect.

CLASS A: IDEALIZED SYMMETRICAL FORM [14, 4]

The highest level of symmetry and objective process is maintained, the final result

resembling an idealized mathematical form. Intuitive decisions are limited mainly to

high-level parameters. Gradual global tempo shifts are commonly employed.

CLASS B: CONSTRUCTED BINARY FORM [1, 2]

This form consists of two sections: an exposition and a development. In the exposition, a

basic, repeating shape is introduced and systematically layered, transposed, and

rhythmically offset against itself. After all layers have been introduced, the development

begins by incrementally shortening the length of each repeating shape, producing a

rhythmic phasing process. The exposition is generally handled in a strict manner, while

the development may be altered intuitively, eventually determining how the piece will

end. Both shifting and steady global tempos are commonly employed.

CLASS C: INTUITIVE LINEAR FORM [21, 99]

Intuitively generated elements are introduced linearly, above a steady pulse. No

systematic transformations occur with the exception of gradual increases or decreases in

velocity. Works of this type generally proceed at a steady, perceptible tempo.

CLASS D: INTUITIVE TRANSFORMATIONAL FORM [13, 29]

Systematic transformations of a basic, repeating shape are applied intuitively, sometimes

haphazardly. Suggestions of form and structure are led by the outcome of the

transformations. The composer listens to the materials and follows their lead, splitting,

splintering, and chiseling away. This form emphasizes intervals, scales, and rhythmic
 27

development, and generally proceeds at a steady, perceptible tempo.

Note on order and numbering:

The order of works presented here is not chronological, but corresponds to the order in

which they were originally presented in concert and will be presented as an album on the

LINE record label36 in spring of 2011. The numbering system of the studies is

chronological, based on the order in which the idea for the piece was first explored, but

not necessarily completed.

5.1 Study No. 14: Arch Study for the Highest Eight Notes
Tempo range: 5-750 BPM
Meter: 1/1
Duration: 4:54
Class: A

Three simple limitations were imposed in order to generate the material of this

piece. The first was to use only the highest eight notes of the piano, the second was that

each note should pulse at a rate which is related sub-harmonically (whole number

fraction) to a "base" rate, and the third was that each pulse would not remain at a steady

velocity, but would be in a constant state of either "fading in" or "fading out" according

to a set of gradual velocity curves.

The first limitation was imposed in order to make use of the unique sonic qualities

of the highest notes. The short strings of the highest octave are less spectrally complex,

yet more confusing to the ear, than the rest of the strings. In isolation, they take on a

distinctly non-pianistic character, with a much higher ratio of hammer noise to tone.

Because of the relatively pure, almost sine-like character of their tonal components, they

are also more subject to interference from room acoustics.

36
See www.12k.com/line/
 28

The use of steady pulses and gradual velocity curves are meant to simultaneously

blur the perception of tempo and tone. The piece begins at the lowest possible velocity,

making hardly any sound but a slight noise from the hammers. The global tempo begins

at a rate so slow that it is nearly impossible to perceive rhythm, and gradually increases

until it nearly becomes a continuous texture. During the first third of the piece, the pulses

fade in and out systematically, overlapping to form perceivable polyrhythms. As the

global tempo speeds up, the slower pulses fade away, one by one, creating an interplay

between the global tempo and the relative tempos of the individual pulses. This effect,

also explored in Studies 1 and 4, can be likened to a Shepard tone37 in the realm of tempo.

The piece approaches its crescendo as the global increase in tempo accelerates and takes

over any relative perception of speed. At its peak, only two pulses are left - they are the

highest two notes as well as the slowest two pulses, whose rates are closest to each other

(at a ratio of 7/6). It is at this point that the two pulses briefly coincide, illustrating the

close perceptual connection between polyrhythm and phasing.

After the climax, the piece proceeds in exact retrograde. While the global tempo

ascends in a pseudo-logarithmic fashion during the first half, it descends at a nearly linear

rate. This particular curve was chosen in order to generate the smoothest possible feeling

of motion from beginning to end.

37
“A Shepard tone, named after Roger Shepard, is a sound consisting of a superposition of sine waves
separated by octaves…This creates the auditory illusion of a tone that continually ascends or descends in
pitch, yet which ultimately seems to get no higher or lower.” Wikipedia,
<http://en.wikipedia.org/wiki/Shepard_tone> (6 December 2010).
 29

Figure 14-1. mm. 1-16

Seven different pulse rates are present in the piece, as follows:

One pulse every sixteenth note (F6)

One pulse every two sixteenth notes (F#6)
One pulse every three sixteenth notes (G6)
One pulse every four sixteenth notes (G#6)
One pulse every five sixteenth notes (A6)
One pulse every six sixteenth notes (A#6)
One pulse every seven sixteenth notes (B6)

The highest note, C7, begins pulsing at a rate of one pulse every sixteenth note

and incrementally slows down as it approaches the climax at m. 66, intersecting briefly

with B6 (the second-highest note). The B6 forms the spine of the piece, being the only

voice to follow a single, unbroken, and perfectly symmetrical crescendo and decrescendo,

peaking at the climax.

5.2 Study No. 4: Sixteen Diatonic Glissandi Moving at Harmonic Rates

Tempo range: 5-990 BPM
Meter: 1/1
Duration: 2:59
Class: A

In this piece, sixteen diatonic glissandi (using only white keys) begin

simultaneously, starting at the top-most note (C8) and ending at the bottom-most note

(A0). The first glissando moves down at a rate of one note per measure (52 whole notes,

52 measures total), the second moves at a rate of two notes per measure (52 half notes, 26
 30

measures total), the third at a rate of three notes per measure (52 "third" notes, 17 1/3

measures total), and so on, up to the sixteenth glissando. All glissandi are played legato,

so that each note is held until the following note sounds (important mainly for the visual

accompaniment). The entire process occurs simultaneously in retrograde inversion (and

played staccato), forming an inverted mirror image along the axis of the first glissando.

This axis forms because the first glissando, spanning the entire length of the piece, is its

own retrograde inversion.

The tempo of the piece begins at 5 beats per minute, rising gradually to 990 beats

per minute at the mid-point, then decreasing again to 5 beats per minute according to the

curve shown in Appendix A. The tempo curve was designed to bring attention to the

emergent patterns that form at the beginning and end of the piece (see Figure 4-1).

Figure 4-1. Final five measures (chromatic version depicted to enhance clarity)

Unlike the other studies in this series, Study No. 4 was conceived of in a flash and

immediately sketched out as a mathematical shape on paper. The idea was simply to start

sixteen identical processes in motion, moving at successive harmonic rates. Because the
 31

piano keyboard is nothing more than a set of 88 buttons, the simplest process I could

imagine was a glissando, moving from the top of the keyboard down to the bottom.

The first incarnation of the piece used all 88 keys, but was later changed to use

only the white notes. This decision was made due to the fact that this piece, more than

any of the others, depends on the visual display of the keyboard as an accompaniment,

and the white keys allow for the emerging patterns to be displayed in a completely

symmetrical fashion. The downside of this decision is that the 'Shepard tone' effect

created by the overlapping cascades of notes is diminished, since the (Western) ear

naturally hears the notes as members of the major scale (though for some listeners, this

situation may be preferable).

The main inspiration for this piece came from the "Rhythmicon" instrument

(conceived of by Henry Cowell in the late 1920s and built by Leon Theremin in 1931)

and James Tenney’s Spectral Canon for Conlon Nancarrow (1974), itself inspired by the

Rhythmicon. Cowell explored the idea of correlating pitches with rhythms extensively in

his writings, and the Rhythmicon marks his first and only attempt to realize this idea in a

mechanical instrument. The Rhythmicon uses a 17-key piano-style interface which

can produce up to sixteen different rhythms—a periodic base rhythm on a selected

fundamental pitch and fifteen progressively more rapid rhythms, each associated
with one of the ascending notes of the fundamental pitch's overtone series. Like the
overtone series itself, the rhythms follow an arithmetic progression, so that for
every single beat of the fundamental, the first overtone (if played) beats twice, the
second overtone beats three times, and so forth. Using the device's keyboard, each
of the sixteen rhythms can be produced individually or in any combination. A
seventeenth key permits optional syncopation. The instrument produces its
percussion-like sound using a system, proposed by Cowell, that involves light
being passed through radially indexed holes in a series of spinning 'cogwheel' discs
before arriving at electric photoreceptors.38

38
Wikipedia <http://en.wikipedia.org/w/index.php?title=Rhythmicon&oldid=374970998> (6 December
2010).
 32

In my opinion, James Tenney's 1974 piece constitutes the most sublime realization of

Cowell's original idea to date, making use of 24 pure harmonics on a custom-tuned piano

and adding a brilliant twist: the same ratios that connect pitch and rhythm are also applied

to a logarithmic arc of increasing and decreasing durations.

My own piece throws out the notion of pitch-rhythm correlation altogether,

instead focusing on the narrower application of harmonic ratios to repetitive pulses.

While directly correlating pitches and rhythms may provide worthwhile material for

organizing compositions, I do not believe the connection between them is inherently

perceptible (even if it contributes to a fantastic effect in Tenney's piece). While it is

certainly possible to smoothly transition between these two timescales (demonstrated

famously in Stockhausen's 1960 piece Kontakte), it is only at or near the boundary

between them that the perceptual connection is significant39.

5.3 Study No. 2 : An Approximate Series of Approximate Harmonic

Series
Tempo range: 55-300 BPM
Meter: 9/4
Duration: 6:05
Class: B
Note: Sustain pedal continuously engaged

39
For an in-depth discussion of this issue, the reader is directed to Curtis Roads' book Microsound.
 33

Figure 2-1. Basic shape (durations drawn for emphasis; continuous use of pedal makes held notes irrelevant)

Like Study No. 1, the basic shape of this study is a simple procession of

ascending and descending octaves (see Figure 2-1). The difference here is that the

durations increase incrementally throughout the measure, causing each self-contained

measure to feel as if it is slowing down, even as the global tempo speeds up (mm. 1-19).

As in Study No. 1, a systematic process of layering occurs: while the original shape

repeats, a transposed copy of the repeating shape is introduced at measure two, a third at

measure three, and so on, until sixteen overlapping layers have been created. However,

the entry point of each layer never occurs exactly at the beginning of the measure (with

the exception of the first), but is staggered in sixteenth note increments both before and

after the measure line. The increments cause an ascending set of sixteenth notes to

"grow" from each note of the original shape (see Fig 2-2).
 34

Figure 2-2. “Sprouts” forming from the basic shape

Since the transpositions of the layers are added in the order of an approximate harmonic

series (octave, octave plus fifth, two octaves, two octaves plus major third, and so on),

each note of the basic shape can be thought of as "sprouting" its own harmonic series.

This process can be contrasted to the exposition section of Study No. 1 in which the

delayed entry points cause a systematic, statistical filling of horizontal space rather than a

"sprouting" from the notes of the basic shape.

As a result of this process of layering, chromatic clusters of notes form in the high

range of the piano. While technically varying from the harmonic series of the lowest note

(E0), these clusters nonetheless add to a natural sense of the overtone spectrum because

of their distance from the base note (since, as overtones get higher, they also get closer).
 35

Put in more universal terms, any note in the highest range of the piano will naturally lie in

close proximity to the overtone series of a note in the lowest range. For this reason, the

piece feels “harmonic” throughout despite extensive chromaticism in the high ranges.

At the end of the exposition section, each layer is shortened by exactly the length

its entry was offset, then "looped" again. As a consequence of this process, the notes of

all sixteen layers line up in perfect vertical columns for exactly one repetition (Fig 2-3).

At this point, the global tempo is at its slowest (55 BPM), giving the listener time to

ponder this (quite literal) turn of events. In the measure that follows, the shape of the

exposition is reversed. The tempo ramps back up to a steady 100 BPM as the process

continues, and the shapes move gradually out of phase with one another. The discretely

identifiable cascades in the exposition now fold into slowly evolving, interlocking

patterns. It should be noted that in contrast to the strictness of Study No. 1, this study

utilizes much more intuitive "tinkering." After approximately ten measures of the

development, various pieces of the puzzle are copied, pasted, and carved away to

generate mini-recapitulations of the "turning point" at measure 20 in different ranges. The

piece ends with another subtle intrusion by the composer, hinting at the relative minor

key of c#.
 36

Figure 2-3. “Turning point”

5.4 Study No. 13 : Echoes

Tempo: 100 BPM (fixed)
Meter: 4/4
Duration: 3:54
Class: D

Where perfect or ideal symmetry defines Studies No. 4 and 14, this piece embeds

layers of imperfect symmetry. A resemblance may be seen to natural forms whose growth

is guided by ideal principles but whose final shape is inevitably distorted by the

complexity of its surrounding environment. While the rhythm is marked by an incessant

stream of sixteenth notes, the repetition of patterns at times gives the impression of being

suspended or frozen in time, drawing attention to subtle changes in the texture. This

phenomenon is especially noticeable during the middle section of the piece, where the

scalar and rhythmic clarity of the introduction (and conclusion) are intentionally blurred.
 37

As in earlier studies, a “basic shape” is used to generate all of the material using

only a few transformational techniques, but the manner in which those techniques are

applied is less strict. Also, the basic shape is not made deliberately sterile, but instead

consists of an intuitively contoured melody in the whole-tone scale (see Figures 13-1 and

13-2). In the opening measures, the original melody is not heard alone, but is

immediately accompanied by seven superimposed copies of itself (see Figure 13-3).

These copies are transposed and offset in a manner similar to Study No. 2, resulting in the

sprouting of clear and regular arpeggios from each note of the original melody.

Figure 13-1. Basic shape (score)

Figure 13-2. Basic shape (piano roll)

 38

Figure 13-3. Basic shape (layered)

It is from the material in these first four measures that the rest of the piece is generated,

using the same basic techniques of copying, pasting, transposing, and offsetting.

Additionally, inversions and retrograde inversions are applied on both a small and large

scale. In many cases, the transformations are applied iteratively, causing them to overlap.

For example, an inversion could be applied to a section of material which itself includes

one or more transitions between inverted and non-inverted, transposed and layered

material. As a result of the obsessive application of these processes, I began to hear an

echoing of phrases which I sought to amplify by applying smooth, linear velocity curves.

Using this simple technique, I was able to simulate the classic electronic effects of fading

in, fading out, and crossfading between clusters of repeating and/or related phrases.

Similar to Study No. 4, the overall shape of this piece is one of inverted

symmetry. That is, at the second half of the piece is an approximate retrograde inversion

of the first (see Appendix A). But the symmetry is never quite perfect. With each

transformation, it was necessary to reassess the material and its relationship to what came

before and after. Subtle alterations were made at every step of the way, influencing the

material that would then be used for further transformation

 39

5.5 Study No. 21 : Bells

Tempo: 80 BPM (fixed)
Meter: 1/1
Duration: 5:24
Class: C
Note: Sustain pedal continuously engaged

A quiet, single, steady pulse at A4 divides this piece with machine precision like a

perforated line. A quiet, unsteady melody appears above, and another below, chiming

together in two-part counterpoint, with just enough space between the chimes to

contemplate each as if it were alone, as if each simultaneity was not a chord but a single

sound. Yet, inevitably, connections are made. Movement is perceived. Lines are drawn.

The notes get a little bit louder as time goes by, the decays overlap a little bit more,

interfering with each other just slightly. An octave is added above the pulse, then a fifth

above that, then another octave. But thanks to the patience and precision of the machine,

we are only able to perceive these changes in retrospect (unless we have our eyes open).

Repetition allows us to focus on details that would otherwise get swept away by the

music.

5.6 Study No. 1 : Octaves, Systematically Filled and Folded

Tempo: 40-990 BPM
Meter: 6/8
Duration: 5:57
Class: B

Like Study No. 2, the “basic shape” of this study is a simple procession of

ascending and descending octaves. Here, the notes move at a regular rate of one note per

measure (see Figure 1-1). All additional material is generated by repeating, duplicating,

and layering the basic shape. Each new layer is offset and transposed systematically in

order to fill both the horizontal (rhythmic) and vertical (tonal/chromatic) spaces between

the notes of the first layer.

 40

Figure 1-1. Basic shape

Figure 1-2. Basic shape, filled

After the introduction of 12 layers, the rhythmic pulse that began at one note per

measure has "filled in" to become a steady stream of sixteenth notes, and the ascending

and descending octaves become a perfectly filled chromatic scale. An interesting side

effect of this process is that the ascent of the original shape itself becomes a series of

stepped descents, while the descent of the original fills in to form a continuous line (see
 41

Figure 1-2). As new layers are introduced and the rhythmic density increases, the global

tempo decreases. More simply stated: as the rhythm speeds up, the tempo slows down.

A direct correlation is established between the horizontal and vertical offsets, with

each measure (12 notes in 6/8) corresponding to one chromatic octave (12 tones). The

second layer divides the octaves of the basic shape into perfect halves both vertically (as

two tritones) and horizontally (as half notes). As a result, the original octave shape is

transformed into a series of ascending and descending tritones, pulsing at twice the rate

of the original. The third and fourth layers, in turn, divide the tritones in half, producing a

series of ascending and descending minor thirds, pulsing at four times the rate of the

original. Here the formula gets slightly trickier, since minor thirds cannot be divided in

half according to the chromatic scale. However, the introduction of layers proceeds in the

most statistically even fashion possible.

At the end of the exposition section, each layer is shortened by exactly the length

its entry was offset, then "looped" again. As a consequence of this process, the notes of

all 12 layers line up in perfect vertical columns for exactly one repetition (Fig 1-3, note

the similarity to Fig 2-3). At this point, the global tempo is at its slowest (40 BPM),

giving the listener time to ponder this (quite literal) turn of events. As the shortened

layers begin to repeat, the tempo ramps back up to 990 BPM, and the direction of the

original shape is reversed. Complex phasing relationships develop as the looping

continues, the lines of notes stretching and folding over each other to form something

akin to a moiré pattern.

The phasing process continues unaltered for 672 measures. During this time, the

global tempo drops down to 300 BPM, then back up to 990, but the change is hardly
 42

perceivable by the listener, as the phasing patterns of notes now suggest dozens of

overlapping and ever-changing tempi. After 672 measures, several layers are selectively

removed, leaving behind the notes of a dominant seventh chord. For a moment, the

patterns sound "musical," perhaps even "bluesy," as our brains connect the familiar

harmonic content with the jagged, syncopated rhythms that exist only as a result of the

simplest mathematical combinations. As the tempo slows down further and the notes

begin to sustain, a few more layers are removed to reveal the original octave shape which

was there all along.

Figure 1-3. “Turning point”

5.7 Study No. 29 : Tentacles

Tempo: 110 BPM (fixed)
Meter: 3/4
Duration: 4:29
Class: D

This piece resembles Study No. 13 in approach, using strict processes to start

with, but allowing (and helping) them to bend and fracture in response to their

environment. Whereas Study No. 13 is structured symmetrically, this piece follows an

obstructed linear path. One might imagine swirling pools of water forming behind rocks
 43

in a stream, occasionally breaking free, merging with larger pools.

Again, the basic processes of copying, pasting, transposing, and offsetting are

central. However, this study does not utilize any inversion or retrograde transformations.

Instead, the large-scale form takes shape through an intuitively crafted process of

splitting, splintering, and chiseling of the material.

Figure 29-1. Structural overview + opening transformations

Figure 29-1 shows the process used to generate the first section of the piece

(which in turn is transformed to generate the remainder). This time, two distinct but

overlapping "seeds" are introduced during the first two measures. The four notes used are

Eb, Gb, Bb, and Db (Figures 29-2, 29-3, 29-4).

Figure 29-2. Seed 1

 44

Figure 29-3. Seed 2

Figure 29-4. Seed 1+2

These notes, in addition to their obvious identity as black keys, are significant for

forming exactly one half of an octatonic scale (defined by a series of alternating half-

steps and whole steps). The exact same configuration of notes, moved a tritone up or

down (to A, C, E, and G), will complete the scale (note that the same rule holds true for

other four note combinations within the octatonic scale, as long as the group itself does

not include a tritone). When sounded together as a chord, the sonority of these notes

resembles either a minor seventh chord or a major chord with an added sixth (depending

on the inversion). As the piece progresses, these fairly stable sonorities blur against the

surrounding texture, which itself is formed by shifted copies of the same sonorities. A

struggle ensues between the vertical sonorities and the horizontal/diagonal texture, which

resembles a series of overlapping, ascending scalar patterns, whose frequent use of half

steps and whole steps again recalls the octatonic scale.

In the rhythmic domain, the two seeds alternate between half notes and quarter
 45

notes, mimicking the alternating half steps and whole steps of the octatonic scale. The

second seed (Figure 29-3) contains within it a slight rhythmic irregularity: the high Bb

that begins the second measure is delayed by a sixteenth note. As the seed is layered, this

subtle shift is amplified.

The first section of the piece (mm. 1-60, sections A, A', A'') resembles a

sequential modulation, not moving towards a harmonic goal, but gradually spiraling,

swirling, breaking apart under its own weight. Each cycle of the "A" shape descends a

perfect fifth. At the moment the fourth cycle begins, the shape splits into two. For the rest

of the piece, there is a feeling that the same shape is repeating, yet it never returns in

exactly the same form. Instead, the tentacles that sprung up during the first section

differentiate themselves, occasionally grafting together with other tentacles or breaking

off completely. A contorted skeleton of a classical form appears, just barely, beneath the

bubbling activity, but before it comes into view, the layers melt away, one by one.

5.8 Study No. 99 : Strumming Machine

Tempo: 150 BPM (fixed)
Meter: 4/4
Duration: 12:32
Class: C

A 12-note arpeggio, forming an evenly dispersed harmonic series, pulses at a rate

of ten notes per second, 600 notes per minute, overlapping with itself every four notes

(see Figure 99-1). It begins quietly, played staccato. Over the course of the piece, the

velocity increases very gradually, as does the depression of the pedal, transforming the

ultra-precise, grid-like formation into a swarm of sound. Notes are added to both expand

and cloud the harmony. Just as the ear and mind have become fully accustomed to this

state of being, forgetting the notes, forgetting the piano, several layers are peeled away to
 46

reveal hidden formations.

During the performance of this piece, which is the longest and last in the concert

program, the lighting should be gradually reduced to complete darkness following

approximately the curve shown in Figure 99-3.

Figure 99-1. mm. 1-4

Figure 99-2. mm. 318-324

Figure 99-3. Lighting fade (indicated in red). Complete darkness should be reached at m. 321 (8m30s).

The style of repetitive playing in Strumming Machine is influenced by

Charlemagne Palestine's work (hence the title, a reference to his 1974 piece Strumming

Music), but my approach here differs in two notable respects. First, whereas Palestine's
 47

work is played by alternating between the two hands (starting with one note per hand and

gradually adding notes to create chords), the rhythm of this piece is created by a single,

overlapping arpeggio. Second, Palestine's “strumming” style requires a human performer

to feel the resonance of the piano strings and respond to them dynamically in order to

bring out clouds of overtones. In this piece, the precision of the machine is unflinching

and precise, yet the resulting sound inevitably becomes cloudy and unpredictable, simply

because the state of a vibrating string and its resulting sound can never be predicted with

complete accuracy, as long as it exists in the physical world.

6. CONCLUSION
Mechanical musical instruments have existed for almost as long as musical

instruments themselves. Yet, it is only in the last century that composers have taken

seriously the potential for such instruments to inspire new forms and approaches to music

making. At present, the influence of machines and automated processes on the

composition, performance, and reception of music is vast and undeniable. Long before

the proliferation of electronic and digital instruments, the pneumatic roll-operated player

piano played a seminal role in the complex and evolving relationship between music and

machines.

Eight Studies for Automatic Piano lives in between the colorful past and

unpredictable future of mechanical music, its possibilities and its limitations, its potential

to liberate and to control. Placed in an immersive concert setting, this work advocates for

an expansion of traditional notions of live performance and musical “life.” Each piece

makes use of simple, computer-aided compositional processes to test the limits of human

perception while relying on one of the most recognizable musical sounds in the Western

world: the piano. By adhering to a set of machine-dependent parameters, I hope to raise

 48

questions about the relationship between nature, humanity, and mathematical precision.

As hard as any machine tries, it will never be able to draw a straight line. And if the

precision in this work seems inhuman, it is only to help us scrutinize our own humanity.

Building upon the practical, theoretical, and aesthetic contributions of Conlon

Nancarrow, Steve Reich, James Tenney, György Ligeti, and Charlemagne Palestine,

Eight Studies plants the seeds for an ongoing, focused and innovative approach to

composition and performance for automated piano. To the visionary composers listed

above, this work owes its gratitude: To Nancarrow for unraveling the notion of a single,

linear time into hundreds of jagged, jubilant threads. To Tenney for unraveling it again by

weaving those very same threads back together in perfect, crystalline order. To Reich for

reassuring us that we are not crazy for fixating on an empty swing as it gradually comes

to rest (and if we are, it is a good kind of crazy). To Ligeti for demonstrating that

"mechanical" is not just about machines, but is a life-giving metaphor with limitless

aesthetic potential. To Palestine for hearing and feeling the piano as an extension of the

body, and for showing that it is possible to invent a new kind of piano music without

writing anything down on paper, or calling oneself a composer.

And finally, to machines and humans everywhere for their enduring,

unpredictable obedience.
 49

APPENDIX A : OVERVIEWS
The following four pages display visual overviews of each piece, using the computer’s
virtual piano roll view. Each overview can be read according to the following key:

NUMBER

TITLE

TEMPO RANGE
DURATION
CLASS

TEMPO CURVE (WHERE APPLICABLE)

PIANO ROLL VIEW (NOTES)
NOTE VELOCITIES
50
51
52
53
 54

APPENDIX B : PHOTOGRAPHS
The following photos were taken at the concert premiere of Eight Studies for Automatic
Piano at Littlefield Concert Hall, Mills College, April 23rd, 2010.

Camera: Barton McGuire (contrast enhanced digitally)

Camera: Carson Whitley

 55

Camera: Carson Whitley

 56

BIBLIOGRAPHY

Babbitt, Milton, Interview by Charles Amirkhanian, 1984, transcribed by the author from
<http://www.archive.org/details/SOM_1984_11_15> (6 December 2010).

Busoni, Ferruccio, and Theodore Baker. Sketch of a New Esthetic of Music. New York:
G. Schirmer, 1911.

Clendinning, Jane Piper. “The Pattern-Meccanico Compositions of György Ligeti.”

Perspectives of New Music 31.1 (1993): 192-234.

Cowell, Henry, and David Nicholls. New Musical Resources. New York: Cambridge
University Press, 1996.

Emmeche, Claus, Simo Køppe, and Frederik Stjernfelt. “Explaining Emergence: Towards
an Ontology of Levels.”.Journal for General Philosophy of Science 28.1 (1997):
83.

Evans, Edwin. “Pianola Music.” The Musical Times 62 (1921): 761-764.

Fowler, Charles B. “The Museum of Music: A History of Mechanical Instruments.”

Music Educators Journal 54.2 (1967): 45-49.

Gabrielyan, Emin. “The Basics of Line Moiré Patterns and Optical Speedup” (2007),
<http://arxiv.org/abs/physics/0703098v1> (6 December 2010).

Gann, Kyle. The Music of Conlon Nancarrow. Music in the Twentieth Century. New
York: Cambridge University Press, 1995.

Glinsky, Albert. Theremin: Ether Music and Espionage. Music in American Life. Urbana:
University of Illinois Press, 2000.

Gregerson, Niels Henrik. “Emergence and Complexity.” In The Oxford Handbook on

Science and Religion, edited by Philip Clayton and Zachary R. Simpson. London:
Oxford University Press, 2006.

Griffiths, Paul. György Ligeti. Contemporary Composers. London: Robson Books, 1997.

Hicks, Michael. “Interval and Form in Ligeti’s Continuum and Coulée.” Perspectives of
New Music 31.1 (1993): 172-190.

Hocker, Jürgen. "My Soul Is in the Machine: Conlon Nancarrow - Composer for Player
Piano - Precursor of Computer Music." Music and Technology in the Twentieth
Century. (2002): 84-96.

Lawson, Rex. “Hindemith: Toccata for mechanical piano.” Pianola Journal 9 (1996): 19-
 57

20.

Ligeti, György. Gyorgy Ligeti Edition 3: Works for Piano. CD liner notes, SK 62308,
1996.

________. Ligeti, György, Péter Várnai, Josef Häusler, and Claude Samuel. Ligeti in
Conversation. London: Eulenburg Books, 1983,
<http://openlibrary.org/books/OL22640977M/Ligeti_in_conversation> (6
December 2010).

Maher, Ciarán. “James Tenney on Intention, Harmony and Phenomenology – A Different

View of the Larger Picture.” MusicWorks 77, 2000.

Nancarrow, Conlon. Interview by Natalie Wheen. The Pianola Journal: Journal of the
Pianola Institute 3 (1990): 6.

Ord-Hume, Arthur W. J. G. Joseph Haydn and the Mechanical Organ. Cardiff, U.K.:
University College Cardiff Press, 1982.

________. “Cogs and Crotchets: A View of Mechanical Music.” Early Music 11.2
(1983): 167-171.

________. “Player Piano.” In The New Grove Dictionary of Music and Musicians (Vol.
25), 2nd ed., edited by Stanley Sadie. London: Macmillan Publishers Ltd., 2000.

“Rhythmicon," Wikipedia <http://en.wikipedia.org/wiki/Rhythmicon> (accessed 6

December 2010).

“Shepard tone,” Wikipedia, <http://en.wikipedia.org/wiki/Shepard_tone>(accessed 6

December 2010).

Ostertag, Bob. “Human Bodies, Computer Music”. Leonardo Music Journal 12.1 (2002):
11-14.

Pinch, Trevor J., and Karin Bijsterveld. “Should One Applaud?” Breaches and
Boundaries in the Reception of New Technology in Music.” Technology and
Culture 44.3 (2003): 536-559.

________. “Sound Studies: New Technologies and Music.” Social Studies of Science
34.5 (2004): 635-648.

Polansky, Larry. “The Early Works of James Tenney” Soundings #13, 1983,
<http://eamusic.dartmouth.edu/~larry/published_articles/tenney_monograph_soun
dings/index.html> (6 December 2010).

Reich, Steve. “Music as a Gradual Process.” In Writings on Music, 1965-2000. Steve

 58

Reich and Paul Hillier. Oxford: Oxford University Press, 2002.

Reynolds, Roger. “Conlon Nancarrow: Interviews in Mexico City and San Francisco.”
American Music 2. 2 (1984): 1-24.

Roads, Curtis. Microsound. Cambridge, Mass: MIT Press, 2001.

Roig-Francolí, Miguel A. “Harmonic and Formal Processes in Ligeti’s Net-Structure

Compositions”. Music Theory Spectrum. 17.2 (1995): 242-267.

Schiano, Michael J. “Grundgestalt.” In The New Grove Dictionary of Music and

Musicians (Vol. 25), 2nd ed., edited by Stanley Sadie. London: Macmillan
Publishers Ltd., 2000.

Schwarz, K. Robert. “Steve Reich: Music as a Gradual Process Part II.” Perspectives of
New Music. 20.2 (1982): 225-286.

Sousa, John P. “Machine Songs IV: The Menace of Mechanical Music.” Computer Music
Journal 17.1 (1993): 14.

Steinitz, Richard. György Ligeti: Music of the Imagination. Boston: Northeastern

University Press, 2003.

Wong Doe, Henry. “Musician or Machine: The Player Piano and Composers of the
Twentieth Century.” Ph.D. diss., The Juilliard School, 2006.

Youtkevich, Sergei. The Eccentric Manifesto. Petrograd: State Publishing House, 1922.
<http://reality.gn.apc.org/ecc/ecc.htm> (6 December 2010).