Symposium on Computer Music Composition

Author(s): Curtis Roads, Marc Battier, Clarence Barlow, John Bischoff, Herbert Brn, Joel
Chadabe, Conrad Cummings, Giuseppe Englert, David Jaffe, Stephan Kaske, Otto Laske, JeanClaude Risset, David Rosenboom, Kaija Saariaho, Horacio Vaggione
Reviewed work(s):
Source: Computer Music Journal, Vol. 10, No. 1 (Spring, 1986), pp. 40-63
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/3680297 .
Accessed: 19/02/2012 20:36
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to Computer Music
Journal.

http://www.jstor.org

EditedbyCurtisRoads,withMarcBattier,
on
Computer
JohnBischoff,Herbert Symposium
ClarenceBarlow,
Conrad
Brun,JoelChadabe,
Cummings, Music
Composition
Englert,DavidJaffe,Stephan
Giuseppe
Kaske,OttoLaske,Jean=Claude
Risset,
DavidRosenboom,
and
KaijaSaariaho,
Horacio
Vaggione

Introduction
From the very first research in music composition
with computers carriedout by LejarenHiller and
his associates in the mid-1950s, the computer has
offered enormous potential to the composer. Computers are among the most malleable tools ever developed by human beings, and in the three decades
since that early research, many hundreds of composers have adaptedcomputers to their own musical needs.
Articles in Computer Music Journaland other
publications' point to the broadapplication of computers in musical tasks, especially to sound synthesis, live performance,and algorithmic or procedural composition.
This symposium is the product of a questionnaire
sent in 1982, 1983, and 1984 to over 30 composers
experienced in the computer medium. The questionnaire contained 21 questions. Composers were
asked to respond to at least five of them. The composers were also invited to submit scores and other
graphics that describe their work. These fourteen
composers respondedto the challenge:
Clarence Barlow (Cologne, West Germany)
Marc Battier (Paris,France)
JohnBischoff (Oakland,California USA)
Herbert Briin (Urbana,Illinois USA)
Joel Chadabe(Albany,New YorkUSA)
Conrad Cummings (Oberlin, Ohio USA)
Giuseppe Englert (Paris,France)
David Jaffe(Stanford,California USA)
Stephan Kaske (Munich, West Germany)
Copyright ? 1985 by Curtis Roads.
1. See Curtis Roads,ed., 1985, Composersand the Computer,
published by William Kaufmann,Inc. (LosAltos, California)for
articles on nine prominent composers.

40

Otto Laske (Boston, Massachusetts USA)

Jean-ClaudeRisset (Marseille, France)
David Rosenboom (Oakland,California USA)
KaijaSaariaho(Paris,France)
Horacio Vaggione(Paris,France)
At the time of their responses, the composers ranged
in age from 22 (Kaske)to 66 (Brfin).They live in the
USA, France,and Germany,although their countries of origin include Italy, India, Argentina, and
Poland, as well.
The views representedhere cover a wide spectrum of opinions and attitudes, edited lightly for
publication. Because of the way the responses were
gathered (by written questionnaire), it is important
to consider each reply by itself, that is, not as a response to the previous composer'sreply. No composer saw the responses of the other composers.
The orderof responses was determined by the editor.
What was the most important part of your musical
training?
David Jaffe:The most important parts of my traditional education were the practical experience of
playing violin and mandolin in both classical and
improvisational contexts, conducting, and composing and hearing my music. The music of eastern
Europe,particularlyJewish music, which I learned
from my father, has had an important influence on
my style, as have my many years as a bluegrass
musician.
As for my theoretical background,one experience
stands out above the rest. At Bennington College, I
had the opportunity of witnessing and participating
in the spatial and orchestrational experiments set
up by Henry Brant.The spatial and orchestrational
planning that has gone into my recent computer
music pieces can be traced back to this training.
Computer Music Journal

Herbert Briin: Listening to music in concerts and

on records in the company of friends.
Jean-ClaudeRisset: I consider my piano studies as
the most crucial part of my traditional music training. This may look paradoxicalfor a composer, especially one who seldom uses real-time computer
facilities. I shall explain why.
The first reason is personal. I had a remarkable
piano teacher, Robert Trimaille, who demandedand
obtained very much from his piano students. Without the intense musical experience and the secure
feeling of professionalism I received through his
training, I am not sure I would have daredto venture into the profession of music. Also, working on
the piano repertorywas a thorough and active introduction to a large body of music. To be revived in
performance,works have to be studied both in their
details and their overall form. A great deal of musicianship is demandedof the teacher here. Robert
Trimaille and Huguette Goullon were admirable
guides to me.
Having to realize the pieces in sound-with ten
fingers-implies deep learning (in both head and
body) about phrasing, contrast, and the use of registers. One has to steer the sound, to understandthe
correlates of musicality in performance,to realize
illusions such as playing legato on a piano. Such
knowhow is invaluable training for the realization
of pieces through computer synthesis, where all
these aspects must be handled by the composer.
But I must also mention that studying composition was very enlightening, especially orchestration
with Andre Jolivet, who had a deep feeling for the
idiosyncracies of the instruments. Studying harmony and counterpoint is very useful to gain a
thorough understandingof western tonal music,
while offering the opportunity to try one's hand in a
domain with explicit rules and criteria. Of course,
these criteria correspondto styles of the past, and it
is debatable whether the study of harmony and
counterpoint is a must for the composer of today. It
does seem worth studying at least one established
musical language or system.

What was the least important part of your traditional musical training?

Herbert Briin: The universally accepted, academically perpetuated, consumer-oriented routine of accepting the consequences of composition as if they
were the properties of the composed music. I had,
alas, to suffer it, but I did not ever believe in it.
What was your most important educational
experience?
Marc Battier: My most important educational experience was the first computer music class in the
music department at the University of ParisVIII,
Vincennes. The class, taught by PatrickGreussay,
was mostly directed toward artificial intelligence
techniques in music, using languages such as Lisp.
Before you worked with computers, what was your
main compositional medium?
Marc Battier: I had the chance to study computer
music early in my student days, back in 1969. Before that, however, I had intensive experience in the
practice of traditional tape music, mostly musique
concrete. Aside from working with computers, I
have continued this electroacoustic activity. These
days, I consider the two media as integrated with
one another.
HerbertBriin: Instruments in chamber ensembles.
David Jaffe:Beforeworking with computers I wrote
for a wide variety of instrumental ensembles. I continue to write instrumental music along with my
computer music. I preferwriting for large groups.
However, since performancecommitments from
large ensembles are difficult to procure, much of
my music has of necessity been for chamber ensembles. I find that the two media-computer music and instrumental/vocal music-complement
each other. Instrumental music continually reminds one of the depth and richness of expression
that is possible with real instruments played by
skilled performers. Computer music allows expression of compositional ideas that would be difficult
to realize with performers.
I have mixed feelings about the combination of
live performers and computer sound. I have written

Roads

41

Fig. 1. Herbert Brfin'sMu-

tatis Mutandis33, com-

position for interpreters,

with ink graphics drawn
by a plotter under control
of a computer programmed
by the composer.

~~x

VzI
z
IDED[D?

cS

r] ,L$LD
$+

~f$

32Xxit

+X mt

I~L~El

Clarence Barlow: In 1971 I attempted to realize a

five-minute stochastic piece using an addingmachine and random number tables. Six months would
have been necessary but for my sudden idea of employing a computer. Within a week of my first Fortran lesson, I had the piece.
John Bischoff: The power of the computer to carry
out proceduresand its general lack of innate musi-

cality allow a composer to add structure from the

groundup in making an instrument. Therefore, the
composer has a chance to experience more clearly
the operation of those structures. Forexample,
Jim Horton in Berkeley,California has developed
numerous melody-generatingsystems that he and
others have listened to extensively over the past
five years. One can almost hear melodic spinning
wheels turning in these programs.They do not just
simulate a broadlyrecognizable musicality (a sequencer does this instantly), but rather they try to
build an original musical entity from the bottom
up. This is unique to music by computers.
Otto Laske: I turned to the computer because of an
inner necessity in my compositional thinking, expecting to find new planning resources. There was
also an outer necessity of having my ideas realized.
Actually I turned to computers before I had access
to one, on account of the kind of precompositional
work I was doing. I was always highly dissatisfied
with "writing music from left to right," a procedure
that seemed to restrict my intuition to lower-level
processes since it was predominantly bottom-up. In
short, the computer permitted me to explore highlevel planning as well as bottom-up (event-driven)
elaboration of musical structure. Unfortunately, the
majority of programmedtools in existence today are

42

Computer Music Journal

for such combinations, as in my piece May All

Your Children Be Acrobats, in which eight gui-

tars, a sopranovoice, and stereo computer sound

are combined. However, one factor plagues all attempts at combining computer sound with traditional instruments: the discrepancybetween the
projection of an unamplified instrument and of a
loudspeaker is so pronounced that the two seem in
completely different worlds. Many composers have
handled this problem by amplifying the instruments. I consider this solution inadequate because
it is, in effect, lowering the instruments to the level
of the speaker. I believe EdgardVaresemay have had
the right idea in his composition Deserts, in which
he avoids ever combining the taped and live sounds.
Why have you turned to the computer?

not sophisticated enough to support a fully interactive way of working on all compositional levels.
Joel Chadabe:Long ago I wrote instrumental
and vocal music, mostly chamber music. Then I
worked with analog electronic system. In about
1975 I started working with computers because
my interests at that time, as now, lie in performance
with electronic systems. Computers have the significant advantageof exact repeatability from performance to performance,and the setup time,
because they do not require patching or tuning,
is short.
The main reason I like to work with computers
when composing is that I can compose while in the
presence of sound, and, in my case, in the presence
of the functioning system. Since my compositions
are functioning systems that operate with performer
interaction, I begin with a crude model of the finished system, something like a first draft of its
operations as well as the sounds it makes, and then
I refine it until it's ready.If I had to work with a
non-real-time computer system, I am not sure I
would want to use it. My primarymotivation in
composing is to be able to experiment with sound
and musical process, and the quick response of a
real-time system is a prerequisite to successful and
enjoyable experimentation.
Herbert Briin: I had been waiting for it. I turned
composer of music only after I barely surfaced
from the helpless depression of a haunted victim in
1942. All my music attempts to reflect, by analogy,
social configurations and relations that I preferto
those I see. Soon, however, I discovered that my
analogies kept referringto a "not yet reality" that
could only be reached if it were true that people
have to change so that "our society as is" could
function better. While not denying the potential
of that vision, I dislike it, because it would support
fascism and totalitarianism. It is thus under the rigorous dialectics of a pregnant contradiction that I
continue writing for instruments. At last the computer enables me to begin experimenting with compositions that by analogy point to social processes
where it is the structure that changes in order to
preserve the variety of human temperament by
guaranteeing the possibility of every human being's
contentedness.

Kaija Saariaho:I had graduallystarted to work in

my compositions with independent processes associated with differentmusical parameters.I became
increasingly interested in the nature of "process"as
well as timbre as a musical parameter.In computers
I saw a means of entering inside sound concretely
in orderto control timbre, and finding a vocabulary
for describing the different factors that comprise
musical color. It was also a means of continuing my
research on musical processes in an especially suitable environment.
Stephan Kaske: The reason I thought it would be
necessary to use computers in my music was my
despair about a composition for chamber ensemble.
It became impossible to survey all the structural
lines or developments of the composition. I needed
a helping hand that could keep the structural organization under my control. I realized that a computer programcould do the job for me, so I wrote
a very inefficient Pascal programon an Apple II
computer. Then I wanted to improve my computer
music programmingknowledge, so I attended a
course at M.I.T. My compositional problem was not
solved there either, but with digital sound synthesis
I was seduced to think about an aspect of music
that I had tended to underestimate until then: timbre. Unfortunately, the composition for chamber
ensemble was never completed.
Jean-ClaudeRisset: I always had a certain vivid interest in timbre. I was intrigued by the potential of
certain timbres to best express certain musical virtualities. I enjoyed composing for traditional instruments-and still do-but I was disappointed
by [analog]electronic music. I felt it opened a
wide sonic field, but it did not seem to me to offer
enough control to composers, who had to, to some
extent, rely on ready-madeobjects or processes.
I was fortunate to work with Max Mathews on
developing the musical use of the computer in
1964-65 and 1967-69. Although it was not easy to
explore the possibilities of computer synthesis of
sound, this exploration was rewardingbecause
everything could be capitalized upon and replicated.
The computer providedrefined control over sound.
It also helped in the application of compositional
processes to sound structure. This was the answer
to my more or less conscious urge to compose the
Roads

43

Fig.2. ExcerptfromJeanClaudeRisset'sDialogues
(1975)for instrumentsand
computer-generated
tape.

?eA.

OLD
- - , *I-- --Q-0-

iI

I'

or5

sounds and give a functional role to timbre. To state

it more explicitly, these are the projects I wanted to
tackle with the computer:
Experiment with the design of my own constraints, instead of having to come to terms
with instrumental or electronic constraints
Assemble a personal pa tte of lively sounds, endowed with some characteristic of identity, but
also very ductile-thus susceptible to intimate
transformationsthat preserve certain characteristics and alter others (e.g., the inharmonic
tones in my piece Inharmonique)
Create a flexible sonic world that could diverge
from the instrumental world but also merge
with it in subtle ways (I tried this in
my pieces Dialogues, Mirages, Profils, Passages, and L'AutreFace)
Suggest an illusory world, as John Chowning
demonstrated so convincingly, by playing directly, so to speak, on perceptual mechanisms,
thus unveiling perceptual "primitives" (cf.
the decomposition of pitch and rhythm in
my pieces The Little Boy, Mutations, and Moments Newtoniens)
David Jaffe:I would not say I "turned"to the computer since I continue to write instrumental music.
44

6400

However, it can be said that I "turnedaway"from

analog electronic music. I had done work with
analog electronics (an old Moog synthesizer) and
had been frustratedby the lack of precise control
over both the individual sounds and the progression
of sounds.
I was first introduced to computer music by Joel
Chadabe. Computers provide the potential to make
any sound. Thus, they must be able to produce that
subset of those sounds that can be called "vital"
and "expressive."Realizing this potential is another
matter entirely. Generally speaking, it is quite difficult to synthesize electronic sounds that rival the
sounds of nature in complexity and interest. The
more I work with computers, the more I have come
to appreciate the richness of acoustic instrument
sounds as well as the subtlety of phrasing and
tone production imparted to those sounds by gifted
players.
Is composing computer music significantly

differentfromcomposingtraditionalvocal or

instrumental music?

Kaija Saariaho:Generally speaking, the computer

is a tool for working out ideas, like a pencil. In any
Computer Music Journal

Fig. 3. Page 10 from Kaija

Saariaho'sVerblendungen
(1984) for 35 instruments
and tape.

2
fAtf

I_____~Pt'

-o-------

-4 ....

----?

I"F-

__

__________r==f_

__

_____

L
_
_..... .. ...___ __.. . .. - -4
....
___

Prf

1k4):
_

olr

ZS:

Oq14

- '

~l1
90p

;_ .3

ArC4

irr
..

Im
......

.....

Roads

45

case, the conditions under which people have to

work with computers differfrom the demands of
composers with traditional instruments. Forexample, the use of time is completely different when
you are composing intensively with paperand pencil, when you are dependent on your own motor facilities, than when you are using a slow timeshared
computer system, in which the computer rules
your time. Under these conditions, composers can
very easily lose contact with their original inspiration in the jungle of algorithms and bugs.
On the other hand, computers widen the working
field and open up possibilities for contact with musical material that can be more concrete and inspiring than ever. Comparework with timbre using
computers and using an orchestra. With computers,
composers can enter into the sound. Colors can be
built from the inside of a sound, and one can test
mixings and change things very freely. When writing experimental instrumentations for symphony
orchestra, the composer is very happy if the first
performancejust interprets the original idea. The
time span between the moment of composition and
the performance(final listening) is at best several
months. It is also much more difficult to maintain a searching mind within the confines of the
orchestra as an institution and traditional music
circles in general than it is in computer music studios, where a curious and open attitude is a basic
requirement.
My work with computers includes much more
planning than my work with instrumental music.
The search for material takes a much longer time,
since I try to understandthe possibilities that the
computer offers and find musical ideas that are idiomatic to this medium. My awareness of different
musical parametershas grown as well, since aspects
of performanceand interpretation must be included
in the work, if living music is desired. Here I have
noticed that the final mixing process replaces the
interpretation of instrumentalists. Hence, I should
have the same objective as they, namely an ana-

out on a piano back in your studio. Do we gain anything by thinking of it as different?We only impoverish our sense of connectness to the rest of music
making.
Giuseppe Englert:The specific exegencies of the
computer to the composer/programmerare: (1) in
the case of loudspeakermusic-the knowledge of
acoustical phenomena and their mechanism that
has to be created and (2) in the case of instrumental/vocal music [composed using a computer]-the
knowledge of what has to be formalized with respect to interpretation by the performer.

lytical yet sensitive approach to my material.

Conrad Cummings: It doesn't sound different. The
environment where composition goes on is different but there is also a difference between working
in your notebook on a camping trip and pounding it

Horacio Vaggione: Electroacoustic music (including

concrkte and electronic music) has opened up a vast
area of sound discoveries by means of direct manipulation of tape. These discoveries were inaccessible
and even unsuspected in the framework of tradi-

46

Computer Music Journal

Whathave the practitionersof computer

music learned from the practitioners of past

electroacoustic(electronic,musiqueconcrete)

music?

Jean-ClaudeRisset: Not enough. I am thinking in

particularof the knowhow of musique concrete
concerning sound classification and transformation,
textures, contrast, and sound distribution via loudspeakers.
Herbert Briin: Here I speak for myself only: (a)
a waveform is a sometime thing, (b)durations of
sound phenomena are of supreme importance, (c)
any steady state is a risk, and (d)parallel motions
of differentparametersor components or attributes
may easily turn musical events back into acoustical
events.
John Bischoff: We have learned at least three things:
(a) that musical sounds can be found in the unforeseen operating margins of a system; (b)that the
imperfections of an electronic instrument may musically parallel the involuntary noises of an acoustic instrument, and therefore will add to the music
rather than detract from it; and (c) that a fruitful
approachto a new technology is to search for the
qualities inherent in the technology itself. These
qualities will emerge and gain meaning apartfrom
any likeness to past musical conventions.

Fig. 4. "Thesefigures show

the position of each sound
(symbolized by a letter) in
a timbral space defined by
two coordinates: a vertical

axis for spectral energy

and a horizontal axis for
onset attack. A system of
circular coordinates has
been superimposed to

order the transitions that

take place between the
timbres." From Fractal
(1983) by Horacio
Vaggione.

]E;

K
D

"B

ZD--H
E

,CC

I
E
L
G

tional instrumental music. The analog voltagecontrolled techniques of the 1960s enlarged the
possibilities of classic electronic studios by permitting the definition of time-variant functions and by
multiplying the means of access to analog modules.
Towardthe end of the 1950s, however, Max Mathews began to develop digital sound synthesis. With
the exponential growth of computer technology,
one was able to go further and further into the possibilities of composing sounds on the micro level.
Today,practically all sound manipulations relevant
to electroacoustics are possible with digital means.
It is even possible to transform concrete sounds
through analog-to-digital conversion and to do this
more thoroughly (through spectral analysis) than
any analog technique could do. This being the case,
the most serious electroacoustic studios are now in
the process of acquiring digital technology. One
should not speak of a "break"between electroacoustic and computer music, but of continual
growth of a generalized "loudspeakerart" in which

various techniques are bridgedso as to make the

current technological environment of composition
extremely flexible.
As examples of current development in this
area, one could cite "intelligent audio editors" as
well as the advent of digital modules that will
quickly replace purely analog generating and soundprocessing devices in studios and on stage. However, what urgently remains to be done to improve
the effectiveness of the composing environment is
to attack the problem of the loudspeaker itself.
Loudspeakersremain far behind in their ability to
produce audibly the timbral subtleties elaborated
by other elements in the audio processing chain.
Marc Battier: Since electroacoustic music was
mostly good at processing sounds and has developed
many techniques for this purpose, computer music
has incorporatedthese tools. What electroacoustic
music gains is a flexibility unheard of before, the
ability to create sound-processingsystems impossible to build with analog means. I am currently
Roads

47

Fig. 5. Block diagram of

the system used to realize

On BeingInvisibleby
David Rosenboom.

Runningcrosscorrelationsystem
with short-term
memory

Stimulus
Adaptive

filter

Gate:thresholdand
level sense
with timing
parameters

I Fourieranalysis

Response

Associative
_

Brain

memory

array

Coordinating

,array

computer

Sensereceptors

Probabilistic
synthesis

hI

Synthesis

control
system

system

working with banks of filters, up to 40 bandpass

filters and banks of transposition devices (like
Harmonizers2)of several dozens of units, not to
mention banks of several hundredoscillators (not
only playing stored waveforms but also processing
natural sound). This accumulation of resources
brings new ideas into the practice of electroacoustic music.

Clarence Barlow: I am able to allow myself to envisage more elaborate algorithmic compositional
structures than formerly (providedthe musical context demands these).

How has your method of working changed since

you began using computers?

2. A Harmonizeris a commercially availabledevice that can perform time-rate changing--shifting the frequencyof a signal up or
down without changingits duration.-Ed.

Marc Battier: I find many answers in the psychoacoustics domain, as well as unexpected and exciting questions. We have both acousticians and
psychoacousticians at IRCAM.The link with musicians is strongerwith the latter, and several musical pieces have been written after psychoacoustic
experiments have been carriedout (for example,
timbral studies and studies of spectral fusion). We
know that the computer can play any sound, only
we don't know how to describe them to the computer. Cooperationbetween musicians and acousticians is of the utmost importance in this activity. I
am working on a piece that makes use of data from

48

Computer Music Journal

Herbert Briin: My method of working with instruments has not changed. There I continue to be
the structure who stipulates the system whose
changes of state I compose. I have, however, added a
method of working with computers. Here I compose the structure which generates the system
whose changes of state it composes.

Is any scientific branch (e.g., acoustics,

psychoacoustics) relevant to your current
compositional concerns?

spectral transposition and fusion studies, for tape

and brass quintet.
Herbert Briin: Acoustics, cybernetics, economics,
linguistics, information theory, some mathematics,
and aesthetics are all relevant to my current work.
Joel Chadabe:In the recent past, system theory has
been a great interest of mine, indeed a necessary
interest in developing the concepts of interactive
composing. At this point, artificial intelligence is
of great interest to me, particularlyas a route to
developing more interesting musician/machine
interactions.
David Rosenboom: Yes, especially psychoacoustics,
information processing in the brain and nervous
system, perception, cognitive modeling, mechanisms of attention and states of consciousness,
physiological aspects of performanceand musical
proprioception, experimental aesthetics, and the application of the methods of psychobiology to aesthetic experience-these are all very important
to my work.
Jean-ClaudeRisset: The field of psychoacoustics
(relating to the physical structure of sound and its
aural effect) is relevant to computer synthesis of
sound in general. It affects my compositional thinking in several ways. Forinstance, I am interested in
devising sonic structures so as to be able to bias
perception to organize them in one way or another,
both in the simultaneous and successive case, depending on the fine adjustment of certain parameters. For example, one might adjust the parameters
to favor analytic perception-analyzing and segregating, or synthetic perception-grouping and clustering. This relates to a branch of psychoacoustics
explored by Bregman,Warren,Wessel, McNabb,
Chowning, and McAdams. I am also interested in
the issue of categorical perception. Are established
categories necessary to differentiation?If so, how
can one teach (or learn) new categories? These are
vital questions if one wants to develop music by
structuring aspects of timbre. I also use the computer to set up illusory situations in my composi-

cillators playing very close frequencies is a direct

transposition of multiple-ray interference in the
Fabry-Perotinterferometer.
Clarence Barlow: The sciences of acoustics and psychoacoustics are valuable to me for increasing the
audible relevance of a composition (like a knowledge of good orchestration).I also often have to resort to algebraand other branches of mathematics
in orderto optimize (or indeed realize) my compositional processes. I am also interested in phonetics
and linguistics.

tions, as I did in the endless progression of "hotfudge sundaes" of pitch and rhythm-a branch explored by Shepard, Deutsch, Chowning, and myself.
I also want to use physical models that are unexploited in sound. For example, "phasing" with os-

this theorizing is highly procedural. Since traditional (including twentieth-century) music theory
has been so consistently declarative, it has rarely
addressed itself to problems of real music. What
happens when a primarily declarative theory is

Is any aspectof contemporarymusic theory

relevantto yourwork?
HerbertBriin:Everyone.
MarcBattier:Yes.Empiricalresearchon new tech-

niques of playing traditional music is useful to me

in sound synthesis, as well as the electroacoustic
treatment of sound. The vocabulary of describing
sound (enhancedby Xenakis, Boulez, and others) is
important in dealing with the new possibilities in
the creation and articulation of musical material.
David Rosenboom: I am especially interested in
those aspects of contemporarymusic theory that
attempt to achieve broaddescriptive and analytical
power when applied to the "music of the whole
earth." By this I mean theories of music that are
stylistically nonspecific. These tend to emphasize
the scientific study of music from the point of view
of perception and what might be termed aesthetic
information processing. I am particularlyinterested
in the work of theorists like JamesTenney on temporal gestalt perception, David Wessel and John
Grey on timbre, Diana Deutsch on musical perception, D. E. Berlyne and Paul Vitz on experimental
aesthetics, Manfred Clynes on morphological contour elements in expressive action, and numerous
others I cannot list here.
Otto Laske: I believe every composer is by necessity also a "music theorist," but for the composer

Roads

49

Fig. 6. Opening of Psyche's

Act I, Scene 2 aria in Conrad Cummings's opera
Eros and Psyche, for vocalists, orchestra, and
computer sound.

10. Aria

, t- ca.52

19

Adagioj=ca.52

1 semplice.dolce

None

band.

'9

comes,

none

none comes,

none

comes,

vIc.pizz.

none

comes,
o b.

. ..

f1.

bsn,

comes

,,

none comes, none comes, none comes,

none comes,

none

comes.
--

, ,,

comes,

They ad - mi- re

me_

but pass me

by,

,.none

they
ad=

Smtr

.. . , .
.....,

..
W"'..."'
', . :

-re
m, m.

.."'.

by
bu.pass.me

"..
. .
. ....
. ".'' . ... ,

ad mi me
they
.re

" .. : . .
....."..- ,'.....

butassmeby.

bsnI.

-,.

. M.,
.....

50

Computer Music Journal

adthey

turned into a procedureis shown by the so-called

Princeton School of the 1960s.
The aspect of American music theory that is relevant to my work is strictly methodological-it is
the insistence on explicitness. However, in a procedural approachto theory (as I have pursued it since
1970), explicit means programmed. And what can
be programmedare procedures,not mere databases
(except where they are part of a knowledge base). In
brief, there are aspects of (American)music theory
that are of methodological interest. Unfortunately,
its protagonists are a little too afraidof their own
courage and thus restrict the application of their
methodology to safe topics, such as "goodold" pitch
classes.
How does computer music relate to the musical
tradition? Is it a continuum or is it a turning point?
Conrad Cummings: I hope the scientific mystique
of Modernism is passing in music as it has already
passed in architecture and the visual arts. The premise that music must be reinvented, free of its hindrance from the past, challenging its listeners to
enter a new and unprecedented world-I well remember how exciting that was. Central to its implementation was the notion that art must look to
science. Computer music came of age at the very
end of this premise's hegemony, in the early 1960s.
Like an incredible amphibian, it's been left on dry
groundas Modernism crested and receded. We'reall
out here on the sand, finding a new life in an environment very different from the one that spawned
us-and we're surviving splendidly!
Modernism has no use for the vernacular.What
a surprise that Modernism's child-computer music-thumbs its nose at distinctions between high
art and popular art. Digital synthesis meets Star
Warsand reaches millions!
No, computer music itself is not a turning point.
It is the late product of one era launched into and
thriving in another. It's the knee joint. No better
place to see Modern becoming Post-modern.
Clarence Barlow: I do not think that computer music exists as a separate aesthetic entity. If you mean
digital electronic music, this is just a new subset

of electronic music. (Now that was new!) Structural

composition existed long before the advent of computers. In short, the computer enables us to do what
we did before but much more efficiently, as does the
printing press.
John Bischoff: Applying computer technology to
music is bound to generate a new branch on the
tree of musical traditions. It seems likely that computers can aid us in creating new notions of what
it means to be musical. On the other hand, certain
ideas absorbedfrom recent experimental music
practices are relevant to working with computers.
John Cage'sinsights in regardto control in music
seem particularlyhelpful. The League of Automatic
Music Composers, of which I am a member, makes
music that is startlingly original yet its organizational structure is nonhierarchicaland cooperative.
The group is organized as a network of microcomputers running simultaneous and independent music programs.These programscontinually exchange
information along various paths. The rich and
unforeseen music that often arises from such a
situation deepens one's trust in democratic musical
arrangements.
Joel Chadabe:Formany composers, computer music seems to be an extension of the tradition, with
the computer used as a surrogateperformer.Forme,
however, it makes possible a technique that I call
interactive composing, that I consider a significantly new and rewardingway of working. [See
Computer Music Journal8(1):22-27, 1984.]
Stephan Kaske: Computer music is both a turning point and a continuation. On the one hand it
is a logical succession of a musical tradition that
searched for more precise control of compositional
structure and timbre, and that tried to introduce
noninstrumental sounds into music. On the other
hand, certain streams of contemporarymusical
thought lead to the automation of musical process.
This will be a turning point, even if traces of automated composition can be found in music history,
since the composer will have to say goodbye to the
myth that creation is identical with the creator.
The composer's way of thinking will presumably
change dramatically.
Jean-Claude Risset: A priori, computer music does
not have to relate to musical tradition. The com-

Roads

51

puter is seemingly neutral, although some things

are easier to do than others. But tradition has great
weight in music, where one deals with the fuzzy
norms of collective expression. Tradition is heavily
present in the Weltanschaung of everyone, including the composer-through the composer's training-and in the skills and habits of performers.It is
present in the "history,"the "mindset" of the listener, who categorizes and discriminates (or does
not discriminate). The listener's mindset is especially present in the perception of pitch and timbre,
where discrimination can be severely impaired
by excessive reference to previously established
categories.
Yet I believe that computer music (at least in
some of its many trends)is indeed a turning point.
It helps escape some traditional constraints, especially the constraints of mechanical systems for the
production of sound. It also offers new ways of dealing with inescapable tradition. Other aspects of
computer music can be regressive, as I point out in
my answers to some of the other questions.
Giuseppe Englert: Computer music, to mark a
turning point in musical tradition, has to satisfy
two conditions: (1) the musical concept of a piece
requires the use of a computer and (2) this necessity
is perceivable to the listener. We have alreadywitnessed two events that have shaken tradition: the
appearanceof electricity and electronics-loudspeaker music, and the introduction of new compositional categories, like indeterminacy,randomness,
and probabilities. These two "revolutions" have
deeply affected musical life, and have partially
masked the influence of computers on musical
thinking. The presence of computers is not completely accepted on the musical scene. For a long
time, it had to be justified by the imitation of tradition. The "turning point" is, for most people, not
really visible yet, but it will be.
David Rosenboom: I believe that the introduction
of computers to the world of music has changed
and will change nothing that is fundamental to music as an art form. What changes music is ideas, not

tion in extremely important ways. The great Greek

thinkers from the island of Samos did not have
computers made of silicon. They did, no doubt, manipulate symbols by whatever means were at their
disposal and the computer is, let us not forget, primarily a manipulator of symbols. It is the rest of
electronic and electromechanical technology
that translates these symbols into some physical
manifestation.
It has been said that Galileo changed astronomy
through the development of the telescope. He must,
however, have had an idea of what to do with it and
he basically demonstrated the truth of Copernican
theory, created previously without telescopes.
One is reminded again of Einstein's simple requirements, merely for pencil and paper.The Futurists
expanded our musical awareness to include the
realm of noise, and Cage helped us to understand
silence. Neither requiredthe development of computers. Computers have helped to expand humanity's reach. It is up to human beings alone to expand
their minds.

tools. It is true that the computer has provided us

with marvelous tools for thought development and
has opened up a vast new sound palette for our exploitation. It will aid us in our growth and evolu-

level. Once a network of sound sources is determined, one proceeds to define logical models of interaction between the sources. This is one of the
most interesting aspects of computer composition:

52

ComputerMusic Journal

What new musical concerns have been introduced

through your work with computers?
Horacio Vaggione:I am interested in generating
timbral polyphonies: complex events producedby
many simultaneous sound sources. In my work, algorithms for sound synthesis create groups or families of sound files that are digitally mixed so as to
produce complex textures, fused timbral entities,
stream segregationprocesses, or large constellations
of tiny fragments of sound materials. Each sound
synthesis algorithm contains instructions for executing micrological procedures.For example, an
algorithm can control the degree of fusion or of
spectral parsing. It can control the speed of transformation of various sound parameters,or it can control the interpolation or exchange of values between
several groups of parameters.In this way of working, the composition begins on the microspectral

Much of my subsequent work was devoted to the

the creation by programmingof specific and highly
differentiatedfields of relationships. These relation- realization of this goal. Though many of the early
ships can be based on any kind of model: statistical, experiments were beautiful examples of artistic
manifestation, we are only now reaching the point
ergodic processes, arbitrary,psychoacoustic prinin the development of intelligent instruments that
etc.
ciples,
allows the realization of a significant portion of
HerbertBriin: After the "mixtures" of timbres of
that early vision. There is still much to do, but the
instruments and the "composition" of timbres in
results are encouragingand the vision is still intact.
the electronic music studios, the computer now inThe second areaI mentioned previously, namely,
vites "transformations"of timbres. Not the concern
extended musical interfaces to the human nervous
is new, but its practical significance: in addition to
system, is certainly related to real-time algorithmic
changes of timbre we can now almost compose the
timbre of changes and the timbre of change.
composition. It could really be considered a subDavid Rosenboom: Two areas of my work with
category,one in which the input structures include
the intelligent processing of electrical signals recomputers have opened up significant musical concorded from the brain or other parts of the nervous
cerns, at least for me. The first involves the use of
system.
algorithmic compositional techniques in real time,
Charles Ives said earlier in this century that
live performance.The second involves extended
musical interface with the human nervous system.
someday music would be made by direct connection to the human brain. In 1927, the physiologist
Since my earliest work with computers I have
E. D. Adrian reportedon the effects of listening to
been concerned with real-time algorithmic comthe audible manifestation of brainrhythms we came
position. The great speed with which even early
an
was
execute
instructions
could
object to call alpha waves. In 1965, Alvin Lucier took the
computers
next step by creating his Music for Solo Performer
of great awe and inspiration. My interest in electronic music, beginning in the 1960s, has always
using alpha waves. Since that time, many composstudent
live
ers, kinetic artists, sculptors, performanceartists,
performance.During my
emphasized
and others have explored the world of bioelectronic
limmusic
facilities
was
to
electronic
access
days,
ited to the "classical" studio. Modular, voltagesignals.
These signals have been the subject of my recontrolled synthesizers were on the horizon but
search since 1968, and have, of course, revealed
were not yet widely available. I did, however, have
an enormously rich and complex coding of human
the good fortune to come in contact with the work
activities. Perhapsmy most complex work in this
of LejarenHiller at the University of Illinois.
area is On Being Invisible. In it a feedback loop is
This led me to an expansion of the notion of
created wherein the performerand the performer's
to
include
what
performanceand improvisation
nervous
or
system become like complex circuit elenormally would be called compositional "prein
a large system. Sometimes they play the
ments
compositional" activities. To be able to animate
of actions, sometimes they play a
of
initiator
role
compositional processes at will, as an option availmore
able instantly to the performingmusician, seemed
passive processing role in a system with a life
its
own.
simply fantastic. To be sure, disciplined improvisa- of
tion involves the animation of compositional proIn a performanceof On Being Invisible a comcesses in the performer'smind and even in the
puter begins by generating sound, either by means
of a stochastically controlled music programor a
collective mind of the performinggroup. Adding
this new kind of process to the possibilities already stored, preprogrammedcomposition. Also inside
the computer is a model of perception. All the comavailable, however, was very exciting. Moreover,
with suitable inputs, these processes could be made puter's sonic output is analyzed according to this
model of perception, which attempts to make preto react to the activities of the performer,which
might change from performanceto performance,or dictions about the structural significance of the
sonic events as they will be perceived by the listo the internal workings of a performinggroup.
Roads

53

tener. Additionally, the computer records and analyzes transient brain signal events, known as eventrelated potentials (ERPs)and coherent waves (alpha,
beta, delta, theta, etc.). Recent research has indicated that peaks contained in the ERPwaveform
and their trends of growth and decay are significantly correlatedwith the salience of the stimulus
to the subject, as well as to other psychological parameters. Analysis of the coherent waves provides a
context for the interpretation of these events. The
computer attempts to obtain confirming or nonconfirming information from these brain signals as to
its own predictions of the perceived structural significance of given sonic events.
In one mode of performance,a confirmation results in an increase in the probability that the kind
of sonic changes associated with the confirmation
will occur again. A nonconfirmation results in a decrease in probability of such an event.
The sonic events are dealt with on several hierarchically related levels of musical structure (reminiscent of the hierarchical Meta Hodos systems
described by JamesTenney). Changes in the sound
parameters(pitch, loudness, timbre, etc.) occur
according to contextually sensitive weighting
schemes that take into account the recent history
of the parameter,its rate of change, and other factors. Since many of the relevant brain signals are
significantly affected by the performer'sshifts
of attention, this work has been described by
LarryPolansky as "an attention-dependent sonic
environment."
Stephan Kaske: I have been fascinated by the control of timbre one has with digital techniques, and
this has extended into my instrumental works as
well. But the more I work with computers, I realize
that my actual way of thinking compositionally
hasn't changed much. I still spend a great deal of
time figuring out musical structure without a computer, in particularthe temporal organization of a
piece.
Programmedmusic that doesn't use a huge database or knowledge base typically results in rather
boring compositions, since the overall organization
is very linear. That's partly because the user interface of many computer music systems forces one to
punch in all those little notes and numbers-sound

events-one after the other, be it with Cmusic,

Music V, or Music 11. Only if there was an intelligent computer music system that enabled me to
work out structural ideas interactively, would new
concerns be introduced into my music.

54

Computer Music Journal

Traditionally, computer music synthesis has been

a relatively difficult task for anything beyond the
simplest of effects. New digital instruments make
synthesis much easier than it has been. Do you feel
this will have a positive or a negative effect on the
musical scene?
John Bischoff: This question brings up some common computer music assumptions: (1) computer
music should be primarily concerned with timbre
(an idea that stems largely from Europeanserial
music); (2) given an interest in timbre, one would
necessarily turn to digital synthesis techniques.
Will the greateravailability of digital synthesis be
positive or negative? Who can tell? Any musical
feature that is made dominant and effortless by a
new technological advance is the first thing one
should reevaluate.
Herbert Briin: It will have a positive effect on the
musical scene. The more people can do what they
want to do, the more dignified becomes the critical
discussion of what they did.
Jean-ClaudeRisset: Certainly making computer
music has, in the past, been a difficult task, and it
still is. However, there is always a risk in making
tools "easier,"that of limiting their power and
making them stereotyped. It is a difficult challenge
to design digital instruments that are easy to use
yet which preserve the diversity of possibilities inherent in the computer. Many digital synthesizers
are difficult to reconfigure,and they provide a limited palette of sonic possibilities that is hard to escape-hence, sonic clich6s. Avoiding such clich6s
was one reason for going to the computer in the
first place.
Real-time operation is hard to resist. It may entail
a less thoughtful approach,and trial-and-erroron
real-time systems is not guaranteedto lead you
where you want to go. The technical demands of
real-time synthesis still impose limits on sound

richness. Some synthesizers can recorda natural

sound (e.g., a note from a trombone) and transpose
it in pitch. While this makes it easy to generate
scales from a sound, such scales sound very mechanical-a turnoff for many listeners.
Hence, the effort to make synthesis easier may
lead to a musical regression-as was the case with
most uses of analog synthesizers comparedwith the
previous practice of electronic music before synthesizers were invented. It remains a tough but
worthwhile challenge to make the musical potential of the computer bloom. We must improve the
interactivity and real-time possibilities of computers, but we must also improve our input languages
and information transmission.
Conrad Cummings: It happened with the Moog
synthesizer already.Composer X: "These sounds
that we worked so laboriously to generate-we
can't use them anymore because they're in every
video game." Modernism was inherently elitist. We
knew the way of the future, and we would teach it
until everyone else saw that it was right. Putting
the music of Modernism in a video game is not
cheapening or perverting it, it is unselfconscious
guerilla warfareon the highest level. You want to
show us the right way to use your sounds? Well
thanks, but we'll use your sounds our own way!
Ease of access and ease of use lead more people to
use the tools for more varied ends. Nothing could
be healthier for the continuing vitality of our musical life.
Stephan Kaske: Did the introduction of the
pianoforte have a negative effect on the musical
scene? Or the first sine wave generator?If the only
virtue of music producedusing computers was the
capability of generating new timbres, then computer music would be a poor show. The introduction of inexpensive digital synthesizers like the
YamahaDX series is releasing composers from the
obsession of creating new timbres. I suppose it will
have a positive effect on the scene in that it will
help many composers who had been seduced by the
rather peripheral aspect of sound synthesis to get
back to the real thing called music.
David Rosenboom: I feel this is a decidely positive
development. The proliferation of accessible, powerful new tools can only increase the probability of

truly great works being created. Some of the finest

composers, particularlyyounger ones, cannot afford
or do not have access to the fruits of developments
in computer science. Of course, such proliferation
will also result in a great deal of boring and uninteresting work being created with these instruments. So what else is new? Nothing will change
in this regard.The proliferationof the piano has
resulted in great music and uninteresting music,
none of which can really be blamed on the piano
itself.
In addition, I might point out that the creation of
"great"works is not the only legitimate goal for the
use of these instruments. A vast amount of musical
activity by the people of our culture is undertaken
for the personal edification of themselves as individuals or their social groups. The evaluation of
musical works for their high cultural longevity is
an irrelevant activity for these persons. Their musical activity has its own legitimacy, even if its meaning is limited to a relatively small social sphere.
The people need rich and inexpensive resources for
their musical activity. This is an important point
and should not be overlooked by those primarily
concerned with "high"art.
Giuseppe Englert:Devices that aid composers in
certain tasks enable them to concentrate on other
tasks that are more important to them. But such
devices will impose limitations on composers or
pose unforeseen problems on them. Certainly musicians involved in live electronic music performance
welcome digital modules.
Digital synthesizers and signal processors have
an extremely wide dynamic rangewith low noise,
matching the capabilities of high-quality amplifiers
and loudspeakers. Analog tape is the weakest link
in the performancechain. Therefore,new techniques that allow musicians to dispense with analog tape will enhance the acoustical quality and add
liveliness to concert performances.
Marc Battier: Fora long time there have been works
for tape and instruments, developing the idea of a
mixed music, and there have also been works for
electronic instruments and orchestra (not to mention pieces in which the older Ondes Martinot or
Hammond organ has been used). At IRCAM we are
working toward an integration of traditional instru-

Roads

55

ments and electronics. In orderto achieve integration, we use several modes of interrelation between
the two worlds. We use digital sound processors, capable of sound synthesis and natural sound treatment in real time. The real-time processors can
respond to commands from a performeror conductor, and more generally to cues from a traditional
instrument. Thus it is responsive to gestures. Its activities can also be triggeredby sounds, after some
sort of pitch, octave, or amplitude threshold detection. More importantly, the sound quality and capabilities of modern sound processors are such that it
is not so much an instrument as it is a network of
sound activities. The positive aspect on the musical
scene can be viewed as a better connection between
the electronics and the instrumental performers,
the conductor, and the composer.
What do you think of attempts to automate or
simulate compositional processes?
Otto Laske: This question concerns a much maligned and even more misunderstood topic. The
issue is human musical planning. Forme, computer
programsfor composition are planning aids, regardless of whether they "automate"or "simulate"
cognitive processes. It is always the human composer who develops the meta-plan for the use of
such tools.
Although an individual's compositional processes
are, by nature, highly idiosyncratic, one would have
to be a solipsist in the sense of Schopenhauerto
deny that composers share a common cultural
context, including certain scripts and procedures.
(Schopenhauer,in good German fashion, recommended a beating as the only way to cure solipsism.
I don't know what the musical equivalent would
be.) The question is: How can we transferhuman
musical expertise to a computer and represent it
within the machine? How can we construct musical knowledge bases incrementally? How can we get
the machine to explain its musical reasoning to a
human being? There is nothing peculiar about musical expertise that would force us to use different
methods from those used in artificial intelligence
applications today to solve these very legitimate
problems.

56

Herbert Briin: The question ought to be investigated and politically analyzed. Forexample, is
artificial intelligence desirable if it triumphantly
simulates the human moron's submissive obedience
and ruthless efficiency? Furthermore,I cannot simulate compositional processes. I can, however, compose automated processes or processing automata.
Giuseppe Englert:Algorithms have been introduced by many composers at all times. There are
also compositions for which all attempts to discover rules or formulas have failed. Composition
rules are algorithms that can be traced in works of
more than one composer in a specific historical period. More interesting are individual algorithms
that a composer invents, eventually for only one
piece. The research made by Andr6 Riotte on compositions of J.S. Bach, Stravinsky,and Bartokreveals
astonishing facts.
For some of my works like the cantata Au jour
ultime liesse (1963) and the string quartet La
joute des lierres (1966) I have built strict rules and
mechanisms. These are algorithmic compositions
written long before I became interested in computers. My recent compositions are automated
to a large extent: Mutations Ocre-Violet (1982)
for NEDCO digital synthesizer lasts 30 minutes
and requires only a few manual interventions during performance.Babel (1981) for orchestraand
Ecorces (1982)for five instruments are pieces in
which pitch, duration, and articulation are calculated and printed by computer, with dynamics
addedby hand afterward.
The myth of automation (for power)has accompanied the intellectual life of mankind a long time.
Adam eating the Apple (Ho-ho! Coincidence?)Prometheus, Rabbi Loew-Golem, Faust-Homunculus, etc. The logical scheme behind the myth,
simplified, is as follows: "WhatI know I can describe. What I can describe I can reproduce(or simulate)." This represents three stages: knowledge
acquisition, description, and formalization. To fully
automate a composition process we have to know
all about what goes on in a composer's brain (and
other interior organs) in a given cultural context.
For the moment we have only partial knowledge of
the problems involved, which limits present expectations in automated composition. A final remark:
in all traditions or legends related to the myth of

Computer Music Journal

Fig. 7. Page 1 of Ecorces

(1982) for five instruments
by Giuseppe Englert.

sig
7771'i

S ax n...
t. ......

itte

.. ..

...

"

t>>"~1[

lat
remoloF...... t....

.......nge . a ticulation

automation, the simulation of man by artificial

means is finally condemned and punished. This
could explain why some of our colleagues become
irrationally angry when discussion turns to musical
automation.
David Rosenboom: I am very concerned with particular kinds of applications of these processes. I believe this to be an absolutely valid and interesting
pursuit and possibly, a new kind of music that can
be listened to with new ears and a new type of musical attention.

mul tipl"

s mblable

DavidJaffe:A goodcomposerdrawson a wealth

of practicalexperiencewith musicalmaterialsand
is versatilein a numberof techniques.Fora given
piece, a composerdevelopswhatevertechniques
areneededto producethe desiredexpression.When
workingwith a computer,a composerwho can programcan depicta musicalidea in termsof a program.The programcan be completelydeterministic
or haveprobabilisticelements.Evenif it is deterministic, it maybe sufficientlycomplexthat the
exact detailsof the outputcannotbe imaginedin
Roads

57

advance. However, this does not imply that the

composer does not have a clear general conception
of the result.
On the other hand, automation is not a prerequisite for quality, nor does it guarantee quality.
It is neutral. It is just a technique that can be used
toward artistic ends by an imaginative composer. In
my computer music, I have used a variety of techniques within a single piece that span a landscape
from completely automated to completely manual
composition. Automation can be implemented on
any level of the compositional process. Often I will
specify exact pitch and rhythmic material but have
automated systems controlling how these are used.
Sometimes automated systems can produce results that could not be attained with manual techniques. For example, in Silicon Valley Breakdown,
tempo, rubato, and phrasing are automated such
that groups of instruments can have wildly varying
rubatos but still "understand"where they are in the
music. Several schemes are used, depending on the
musical context. One scheme involves ensembles of
pseudo-instrumentalists, called voices. Each voice
follows its own tempo and rubato trajectory,yet
within the context of this high level of contrapuntal independence, keeps track, from moment to
moment, of the resulting harmonic combinations.
Based on what it "hears,"each voice or ensemble
of voices can alter its own or another voice's preplanned behavior. (This intervoice communication
is implemented via message-passing in the Pla programming language.)In this manner, a responsive
improvisational ensemble is created that nevertheless remains faithful to the precomposed plan.
Stephan Kaske: Every composer should know a
little about how he or she works, since intuition is
too nebulous a term for describing the compositional process. So a simulation of creative phenomena is definitely worth profound scientific investigation. But I wonder if I would be interested in
automating composition to such an extent that I
would be only peripherallyinvolved in the composition process.
We have to determine
positional design could
Taking the most recent
into account, we might

58

which processes of comprofit from automation.

compositional resources
best concentrate on urgent

problems like the musician/machine interface. In

an improved music programmingenvironment, it
would certainly be useful to automate specific compositional procedures,depending on the composition technique being used. For example, if one is a
friend of stochastic music, why not have the computer generate the randomnumbers rather than
throwing dice or coins, like composers did in the
1950s? Or if one is obsessed by patterns, why not
have the computer generate them ad infinitum according to the composer's rules?
A complete automation of the composition process is of merely scientific interest. Among other
things, composition involves emotions, and for
some it is pure emotion. So complete automation of
the composing process calls for a computer with
feelings. I would not object to listening to a computer composition created from a programmed
model of emotions. Let's hear what Maestro Computer wants to tell us!
Marc Battier: Automated musical processes are a
general part of contemporarymusical thought. The
concept has been applied in instrumental music
(for example, Michel Philippot) and in electroacoustic music studios (more evidently in American and Belgian studios). Computers offer a systematic way of investigating this subject.
Horacio Vaggione:The role of the composer in
working with computers is to produce new musical
situations by programming.Automated processes
are an important part of this approach.However,
the composer is not limited to strategies like pure
determinism or pure stochastic processes. I am very
interested in creating compositional systems in
which the software is based on collections of autonomous musical objects, that is, modules that
contain some kind of specific knowledge and are
thus able to execute well-defined tasks. These modules can be made available permanently so they are
available to form various networks of functions. A
single message can activate any module, and the
module must respond by sending messages to all
relevant modules. Activating a module by sending
it a message accomplishes a specific musical task.
For example, one module might distribute sounds
in time according to a law of evolution on another
level, or direct the flux of sounds toward the inputs

Computer Music Journal

Fig. 8. A graphic representation (not a score) of two

sample phrases from John
Bischoff'sNext Tone,
Please (1985). Each circle
represents a three-tone
chord in the range speci-

fled to the left. The vertical

lines indicate simultaneous chords. Lines with
arrows indicate nearly
simultaneous chords. The
"v"symbols represent
regularly modulated pa-

rameters within the sound

of a chord, e.g., waveform
or filter changes. The
tempo is slow; for example, the first chord's
duration is about four
seconds.

"NEXT TONE, PLEASE"

John Bischoff
v v 1vvvvOvv

Sop.

Alto 1vvvvvv
2YVv

Tenor

vvj

vv v

VVVVVV vvv

vv
vjv-4v rvvv

v
VVVv

vey

vv

Bass
2

of an automatic digital mixer according to messages

received from other modules.
What are the dangers of computer music?
Giuseppe Engelert:There was a time when, by singing Verdi'snewest arias in the streets, people manifested their sympathy with the Italian independence
movement and gave moral support to the activists,
the Carbonari.Verdi'soperas constituted a danger
to the Austrian power. Times have changed; music
does not triggerrevolutions anymore. We have to
admit that computer music is not dangerous.
Marc Battier: The time when computer music
sounded more computer than music has gone. I see
no danger, except the dangerof being totally absorbed by computer programming.However, programming will be less and less associated with
computer music in the future, in that musical tools
will be offered to composers. These will partly fill
the gap between the composer's intentions and the
means of realizing them. The dangerwould be to
lose control of the development of these tools, and
as Phillipe M6nardused to say, let Radio Shack do
it all.
John Bischoff: Computer music systems of any
kind are so much more complicated than musical
instruments of the past that there is a tendency for
a composer to spend increasingly more time designing a piece and much less time playing it. This

point is important because traditionally,making

music has involved repeatedplaying and listening. I
don't see why it would be any different for computer music. How does one try out ideas for a piece
without actually defining and building the piece? In
computer music, once design decisions are made,
they are harderto change because of the large
amount of development time invested in them. A
related dangeris to get stuck in a perpetual design
state and never make it to reviewing or testing the
aesthetic assumptions one's work is based on.
Joel Chadabe:Computer music is dangerousto performing musicians who depend on commercial jobs
for living, because computers can produce acceptable orchestral sounds relatively inexpensively. The
same could be said of set designers who were put
out of work by computer graphics used in filmmaking. Overall, we're entering an age when
people's ideas of what is amusing is changing, and I
fear that the music literature that I grew up with,
and the method of its delivery (i.e., performancesin
concert halls) will seem increasingly less rewarding.
What are the worst cliches of computer music?
Herbert Briin: The drone and the loop. It is not
enough that they are the cheapest bragof "can-doism," they play a hapless tribute to just that which
holds them in freezing contempt: well-tempered
tonality.
Roads

59

Kaija Saariaho:Quite often computer music composers focus their ambitions on purely technical aspects, for example extremely complex algorithms
for composition or synthesis. Little attention is
paid to the fundamentally musical elements. This
lack of attention does not stem from a radical approach that searches for musical solutions for new
directions, but rather stems from a lack of interest.
The consequence is that what is heard is often musically conventional, and the solutions are banal.
Too many computer pieces are like audible games,
without any artistic content or depth. The worst
clich6 is a cold, technologically meaningless and
boring-soundingpiece that supposedly is made with
ingenious algorithms. This strongly contradicts the
searching spirit that is usual among computer music composers. Maybe the equipment has been too
elementary to enable composers to save their energy for composition after the tiring programming.
Probablyalso many computer music composers
have until now been more interested in technological aspects than music itself.
David Jaffe:The assumption that loudspeakerplacement is irrelevantand unimportant is counterproductive to the advancement of computer music.
Although there have been composers such as [D.]
Scarlatti who have written for only one instrument,
most composers since the seventeenth century have
written for a variety of musical forces. It will be a
pity if computer musicians forget this and write all
their music for four speakers in a square or two
speakers in the front of a room. I would like to see
more experimentation with nonstandardspeaker
placements and nonstandardspeakers. The idea
that speakers should be completely general is also
counterproductive.I would like to see idiosyncratic
"speaker-instruments"built to have a certain desirable sound and projection, in a manner analogous to
a fine violin. Perhapscomputer musicians will have
to become loudspeaker artisans.
If you could change some aspect of current
computer music practice, what would that be?

more portable, so that we could use them in my

remote province.
Otto Laske: The most important aspect I would
want to change is the way in which computer music is taught today. I would like to see the notion of
a "computer"interpretedmore broad-mindedly.A
comprehensive computer music curriculum that
deserves the name would have to include cognitive,
historical, technological, and scientific topics. It
would also have to include a "composition theory"
that discusses musical planning, as well as topics
relating to sonology (i.e., systematic orchestration
based on insights into the score). Artificial intelligence topics such as planning paradigms,expert
systems, and knowledge representations should be
included as a matter of course, on a par with digital
signal processing and software engineering.
At the present time, the limitations of "computer
music" in the very narrowsense are becoming quite
apparent.One knows a bunch of very idiosyncratic
sound-synthesis techniques, displayed in overlong
pieces, and they are giveaways. They classify a work
based on the techniques it uses. But that is why we
abandoned"electronic music"!
There is very little interest today in teaching
computer music in the broadsense of a computer
as a symbol manipulator (ratherthan a data processor), which would introduce a broadspectrum of related disciplines. This I would like to change.
Clarence Barlow: I would not want to change anything, but I wish all the same that I could be confronted with less music resulting from inscrutably
abstract, extramusically autonomous processes inaudible to me as a listener. I want to hear more music resulting in an obvious way from a musically
powerful idea, such as was desirable as a matter of
course before computers entered the scene.
What is your assessment of the state of computer
music in today's society?
Marc Battier: The French state radio has two pro-

Jean-ClaudeRisset: I would want to have the wonderful programsthat exist or are being developed be

grams devoted to culture and music. Computer music is often played on these programs, and also on
other private stations. There have also been several
educational programs. We may regret that these pro-

60

Computer Music Journal

sell insurance, and some programcomputers. The

most valuable commodity for a composer is time,
enough time to compose. The time must be steady
and must continue for a lifetime if the composer is
to have a chance of developing a mature style.
Horacio Vaggione:Apropos the subject of the "composer seduced into programming"I recommend the
readerto the article by Gareth Loy that appearedin
Perspectives of New Music 1980-81. Before offering a well-articulated panoramaof the dangersand
advantages,and of the technical and subjective
changes that can result from the interaction of the
composer and the machine, Loy cites this statement
by Harry Partch:"I am not an instrument builder,
but a philosophic music-man seduced into carpenManymusiciansinvolvedin the new musical
try." Of course, Loy speaks of positive seduction,
the
of
that music students experience for their instruhave
noticed
like
danger being
technologies
ments. He also points out the aesthetic aspects of
or another
"seduced"into programming
the practice of programming.Composers were the
extramusicalagenda.Do you see this as a
first
for
to use the computer for artistic purposes. The
problem yourself?
resemblance of composing and programmingis obGiuseppe Englert: Some extraordinarypianists have vious, since both deal with processes that evolve in
time defined by specific constraints. From a musibeen "seduced"into becoming composers. Percuscal point of view, however, the finality of programsionists have become conductors. This is not to
of
who
become
or
maming does not rest in itself, but in the musical
speak composers
managers
How
that
in
the
results that the composer can produce with his
fast
field
in
nipulators.
comforting
and aroundmusic, where so many disciplines inter- digital partner.
Composers "seduced"into programmingin the
mingle, one can be seduced by one activity rather
than by others! In my case, I still compose and pernegative sense are people who lose their need to
form music, and love programming.
produce music in orderto dedicate themselves exthis
as
a
I
don't
see
David Rosenboom:
problem par- clusively to the exploration of communication with
the
the machine. At this moment, we can say they are
drudgeryof
ticularly. I am often frustratedby
of
I
am
the
no longer musicians. But they can become good
drudgery copying
by
programming,as
acto
programmersif their interest takes them that far. In
parts from a score. I have learned, however,
I
the same way, this ex-musician programmercan becept both as necessary parts of musical activity.
medicome a fine collaboratorfor a composer that doesn't
can, at times, even transformboth into almost
understandcomputer science but who desires to
tative, creative disciplines. I enjoy very much the
creative aspects of programmingand creation of cir- work on certain ideas and musical images whose
characteristics (e.g., complexity) could only be accuitry, for both have led to many new musical concomplished by means of computers.
cepts and methods.
HerbertBriin: I wish I were a brilliant programmer.
David Jaffe:The problem is not a dangerof being
"seduced."The problem is that being a "composer" My respect for those who are good programmersis
is still not considered an honest profession in the
deep and affectionate. Nothing whatever can belong
United States of America, although in some circles
to an "extramusical agenda" once I have used it for

grams spend more time talking about the music

than playing new pieces. Also, due to the fact that
we have several music research centers in France,
computer music can often be heard in concert.
HerbertBriin: Not being a fame-backed composer I
can only assert with some evidence and full conviction that my six pieces in SAWDUSTare (a)credible
complimentary acknowledgments of the immense
gift presented to me by technology, (b)the most
radical display of "computer age composition" to
date, and (c) one of the successful attempts to restore living interest in the function of composed
music to contemporarylistener's society.

it is a fashionable hobby. Nearly all American composers have to support themselves doing something
other than music composition. Some teach, some

the composition and realization and implementation of a "piece of music."

Marc Battier: Generally speaking, I notice that those

Roads

61

composers who have had to write programsin order

to compose have a different view of computer music. At IRCAM,we have composers who program,
others who don't, and in between those who can
transcribe traditional scores into data for a program
such as the SCOREinput language. However, we all
know of musicians who have been completely
eaten up by the computer, and who have completely quit their musical activities. The other side
of the problem is the composer who only works
with tools developed by another composer, and is
thereby limited. Nevertheless, programmingis a
natural aspect of computing, and music is a field in
progress, so programmingnew musical models
seems now a part of contemporary musical
thought.
Joel Chadabe:I do not think composers have a monopoly on creativity. I know researchersand equipment designers who are more creative than many
composers. It is understandablethat at this point in
the development of computer music, the design of
the instruments themselves is a primary concern.
Many people are likely to become involved in this,
including composers. It is fascinating, and should
not be derided.In my own case, writing the PLAY
program,I have noted the satisfaction one feels in
doing something that others might find useful. But
outside of that excursion into general-purposesoftware development, I have not been tempted by
general-purposework. That might well change,
however, in the future.
Are you interested in combining your musical
works with other media?
Giuseppe Englert:Between 1976 and 1981 our
GroupeArt et Informatiquede Vincennes gave many
concerts, with great success. In these concerts, visual artists (Bret,Huitric, Nahas) displayed on a
video screen animated images realized in real time
by the COLORXsystem (L.Audoire) controlled by
a DEC LSI-11 computer. Accompanying the visuals
were musicians (M. Battier, G. Dalmasso, Holleville, Hunstiger, and me) playing on hybrid computer/synthesizers (computer-controlled analog
synthesizers), and then on the Synclavier I digital

62

computer/synthesizer. The relation between image

and music was improvised and not strict.
Since 1981 visual artists have become more ambitious concerning resolution and color, and they
are no longer satisfied with portable equipment.
The very high cost of renting video equipment for
concert spaces has become an obstacle to our combined performances.
The fact that the performanceof computer music
is not visually spectacular has deprivedus of the
support of television. This may partially account
for the difficulty of inserting computer music into
general musical life.
Joel Chadabe:The idea of interactive composing is
easily extendible to video, computer graphics, and
dance. Video images can be used in parallel with
the music. Since the performancedevice in an interactive composing system can be freely chosen, a
device that translates physical dance motions into
music information can allow dancers to be performers of music.
David Rosenboom: I have been involved with multimedia work a long time, certainly before my first
work with computers. Most recently, I have created
two works in this category. In the Beginning (The
Story) was written in 1980 for chamber orchestra,
film, and synthetic speech. A complex fabric of
music created with a model of proportionalstructures (pitch and rhythm), melodic shape contours,
stochastic selection processes, and other subjective
musical concerns was created in part with the aid
of a computer.
This music was combined with a text, a dialogue
of synthetic speech, and a film (photographedby
George Manupelli).The film depicted surrealscenes
of clay-coveredfigures acting in relation to the text.
The artistic subject concerns modeling behavior,
evolution, and the development of global human
consciousness.
The second work, Daytime Viewing, was created
over the period from 1979 to 1982 in collaboration
with the artist JacquelineHumbert. It involves music performedwith computer-aidedinstruments,
visual material created by mixing photographic,
drawing, and computer graphics processes, electronic processing of sung and spoken text, video,
fashion and costume design, and theatrical perfor-

ComputerMusic Journal

mance. The work is concerned with modern communication media, television, and images of women
in contemporary society.
Kaija Saariaho:In my piece Study for Life (1980) I
combined white light and dancer with electronic
tape and soprano. The light part is very precisely
scored. It represents in my mind a visual parameter
for some musical ideas in the piece, since all the
material on tape consists of sounds made with glass,
which in turn gave me associations of reflections,
different intensities, and shadings. After this piece I
have had many plans to continue work in this direction. For example, I would like to try to realize my
formal ideas with video. I see in video and in music
many common factors, the most important being
that they are both arts in time. In my compositional work I use much drawing, and I would also
like to try to realize these ideas in visual form.
Artistic experience can be used to enlarge several
senses, and the senses are naturally intertwined. In
my score for Study for Life I ask that the room be
filled with scents. I am also interested in multidimensional works of art, but in the abstract, strict
sense. Right now I am working with a spectacle,
where music is connected to actors' movements.
The amplified, well-controlled breathing of the
actors is part of the music, which also consists of
tapes and live processing of sound.

Otto Laske: Since 1980 I have repeatedly collaborated with a modern dance choreographer(my wife,
Peggy Brightman),and have come to appreciatethe
more than musical concerns that enter into such a
collaboration. I am particularlyinterested in works
where choreographerand composer use a common
plan but different computer programs(planning
aids) to accomplish it. An example of this way of
working is Windshadows (1982) for flute, dancer,
and mobile, with Peggy Brightman.
In Windshadows, the choreographyis based on
output from G. M. Koenig'sProjectOne program,
and the music is based on output from Iannis Xenakis's ST program.These programsact as planning
aids for designing and realizing a form. In Windshadows, both the dance and the music are based
on the idea of a sequence of events whose distribution in time and space increases up to a midpoint, and then returns to its initial state. To realize
this idea musically, I defined a form in five sections
for solo flute with the aid of the ST program.Peggy
Brightmanused Koenig'sprogramto yield a blueprint for each section that would correspondin
certain ways to my structure. This blueprint was
interpreted by the choreographer/dancerteam in
terms of Rudolf Laban'sEffort/Shapetheory of
movement.

Roads

63