Elisavet Kiourtsoglou - Specialized Research AI at MIT

Elisavet Kiourtsoglou
13 rue Clavel, 75019, Paris, France
ellikiour@gmail.com

An Architect Draws Sound
and Light: New Perspectives
on Iannis Xenakis’s
Diatope and La L´egende
d’Eer (1978)

Abstract: This article examines the creative process of Diatope, a multimedia project created by Iannis Xenakis in
1978 for the inauguration of the Centre Pompidou in Paris, utilizing analytical research of sources found in several
archives. By interpreting Xenakis’s sketches and plans, the article elucidates for the first time the spatialization of La
L ´egende d’Eer, the music featured in the Diatope. The findings of the research underline the importance of graphic
and geometric representation for understanding Xenakis’s thinking, and they highlight the continuity and evolution
of his theory and practice over time. Furthermore, the research findings present the means by which this key work of
electroacoustic music might be spatialized today, now that the original space of the Diatope no longer exists.

Iannis Xenakis (1922–2001) was trained as a civil
engineer at the Polytechnic University of Athens
(NTUA) and became an architect during his collab-
oration with Le Corbusier (1947–1959). Later on,
Xenakis devoted himself mainly to the composition
of music, with the exception of purely architectural
projects sporadically conceived for his friends and
for other composers. Examples of these include
Franc¸ ois-Bernard M ˆache’s house (1967–1974), the
Scherchen Auditorium (1961), Karen and Roger
Reynolds’ house (1994), his proposal for the archi-
tectural competition organized by La Cit ´e de la
Musique in Paris (1984), and his utopian city of La
Ville cosmique (1964). Nevertheless, he maintained
a particular interest in space throughout his artistic
activities. Beyond elaborate spatializations for his
works of music, he also conceived a series of works
called polytopes, where sound, light, and space were
merged to offer a multisensory experience.

This article examines the creative process of Di-
atope, Xenakis’s most complex polytope, conceived
and constructed in 1978 for the inauguration of
the Centre Georges Pompidou in Paris. What was
distinct about this project was the fact that Xenakis
worked—almost concurrently—on the architecture
of the pavilion, the music of La L ´egende d’Eer (along
with its spatialization), and the light show. These

Computer Music Journal, 41:4, pp. 8–31, Winter 2017
doi:10.1162/COMJ a 00437
c(cid:2) 2018 Massachusetts Institute of Technology.

distinct levels of Diatope—architecture, music, and
light—were neither interdependent elements of the
work nor merely disparate artistic projects merged
together. Xenakis presented a four-dimensional
spectacle based upon analogies between diverse
aspects of space–time reality. To achieve this he
used elementary geometry (points and lines) as a
conceptual tool, based on the premise that this
kind of representation could be abstract enough to
simultaneously refer to architectural forms and to
the distance covered by light and sound.

In this article, I focus on the pavilion’s archi-
tecture and the music’s spatialization, which have
remained hitherto undiscussed in related analysis
(e.g., Harley 1998; Sterken 2001; Barrett 2002; Harley
2002; Solomos 2006), in light of new findings from
several archives, specifically, the Collection famille
Xenakis, the archives of the Centre Pompidou, and
the Archives du CEMAMu and the Audio Collection
Iannis Xenakis (the last two both housed in the Bib-
lioth ´eque Nationale de France). Because the music
of La L ´egende d’Eer has already been the object of
extended analysis, I will not analyze it here; neither
will I analyze the light show, other than with regard
to its connection with the other elements of the
event. I begin by looking at the architecture of the
Diatope pavilion and then turn to the spatialization
of La L ´egende d’Eer; next I will discuss the light
show presented inside the Diatope pavillion; finally,
the article concludes by underlining the continuity
and the evolution of Xenakis’s thought and practice.

Computer Music Journal

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

Figure 1. Earlier versions of
the Diatope project (a and
b) and its ﬁnal form (c).
(Reprinted with kind
permission of the

Collection famille Xenakis
from X[A] 11-12,
X[A] 11-3, and X[A] 11-1.
All rights reserved.)

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The Architecture of Diatope

In 1974, two years after the success of the Polytope
de Cluny, the president of the Centre Pompidou,
Robert Bordaz, commissioned Xenakis

to create a project of light and sound designed to
animate the plaza and the Center of Beaubourg’s
plateau (its facade and possibly some of the ex-
terior parties) for its inauguration (Bordaz 1974).

Diatope, initially called Polytope de Beaubourg,
was the sixth of Xenakis’s polytopes: Polytope de
Montreal (1967), Hibiki Hana Ma (1970), Polytope
de Persepolis (1971), Polytope de Cluny (1972),
Polytope Mondial (1974, not completed), Diatope

du Centre Georges Pompidou (1978), Polytope de
Myc `enes (1978), Polytope de Mexique (1979–1981,
not completed), and Polytope d’Ath `enes (1985,
not completed). It went through three preliminary
versions before its final form of 1978.

According to Xenakis (1974), the early versions of
Diatope presented audacious ideas: using the city’s
sirens to diffuse music and antiaircraft lights to
illuminate the most important buildings of Paris,
including the new edifice; covering the center’s
facades with a network of flashing lights and
“wrapping” them with music; and even constructing
a three-dimensional spider’s web of flashing lights
hanging from the center’s facade and the neighboring
buildings (see Figure 1). The fourth and final version

Kiourtsoglou

consisted of two different locations for the spectacle:
an open-air and an enclosed space. The first location
would include two parallel matrices of flashing
lights perpendicular to the Centre Pompidou,
within which people could walk, watching light
events computed using mathematical theories,
and the second location would house the musical
performance. Over and above financial concerns,
reservations were expressed by the architects
designing the Centre Pompidou, Piano and Rogers
(1974), who feared that Xenakis’s extravagant project
would constitute “a very important visual break as
well as a physically disturbing congestion for any
kind of exterior activity,” and continued that it was
not “integrated into the conception” of the new
building. It was the enclosed space that was finally
chosen for construction.

According to Xenakis (1978), working on this
last version, he had to answer the question of what
would be the appropriate shape of an architectural
shell to house a musical event. In 1958, while
he was working with Le Corbusier on the Philips
Pavilion for the World’s Fair in Brussels, he had
suggested the use of surfaces of double curvature
(hyperbolic paraboloids and conoids) as a solution
to that question. For Diatope, he also faced a
second question: What three-dimensional shape
has the highest volume to allow for the laser show
while occupying the least surface area? Xenakis
considered the answer to that geometrical problem
to be the sphere. Indeed, in his archives we find a
few sketches of a spherical space wherein Xenakis
would probably house Diatope (see Figure 2). “But
the sphere, in itself beautiful, is bad acoustically
and less rich in tactility than others forms of double
curvature” (Xenakis 1978, p. 9). (It seems that this
was not an obstacle for Karlheinz Stockhausen,
however, when he imagined and realized a sound
distribution in the spherical concert hall of Expo ‘70
in Osaka.) After this experiment, Xenakis turned
again towards hyperbolic paraboloids, knowing their
reverberational qualities. In the Diatope pavillion,
however, he did not use precast concrete, as at the
Philips Pavilion, but designed a metallic framework
that bore a red plastic membrane, because Diatope
was also intended to travel around the world as

the “Centre Pompidou’s cultural ambassador” (see
Figure 3).

From spring 1974 to June 1978, the engineering

offices of Groupe Arcora and of Esmery-Caron
were charged with designing and constructing the
architectural shell for Diatope, following Xenakis’s
preparatory sketches and collaborating with him
for the final plans. To fix the red tent overlapping
the three hyperbolic paraboloids of the metallic
structure, 134 blocks of pig iron would be borrowed
from the French army. The structure would appear
at the end of February of 1978 at the plaza outside
the Centre Pompidou.

Regarding the acoustics of this tent, on 20
June 2017 I interviewed Marc Malinowsky, the
civil engineer from the construction office Arcora
responsible for final construction of the Diatope’s
shell. Malinowsky stated that the selected double-
coated PVC fabric (red outside, gray inside, 900 g/m2)
was a pure “sieve” with no acoustical function, its
primary purpose being to offer a dark environment
for the light show. This membrane let all of the low-
frequency sounds of the three daily performances
of La Leg `ende d’Eer escape to the environment,
and although high-frequency sounds were slightly
attenuated, the ultrasonic frequencies coming out of
the tent attracted stray cats, who watched the show
from the upper openings of the Diatope’s structure.
“Only sharp sounds above 500 to 1000 Hz were
likely to be attenuated to a level close to 20 dB,”
Malinowsky explained, “but under 250 Hz, the
mass law remained unavoidable and a maximum
attenuation of about 3 dB would be expected.”
Because of the great amount of high-frequency
sound, performed very loudly as Xenakis was
prone to do for all of his electroacoustic works,
Malinowksy recalls that mothers were covering
their children with coats during the performance and
neighbors were constantly complaining about the
noise. Environmental sounds penetrated the red tent
equally easily, but as the Diatope pavilion was placed
on the inclined plaza outside the Centre Pompidou,
at the crossroads of a pedestrian zone and away from
the main Parisian traffic arteries, this fact did not
have a great effect on the sound performance. In
addition, owing to the initial acoustic permeability

Computer Music Journal

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

d
n
i
k
h
t
i

w
d
e
t
n
i
r
p
e
R

(

.
4
7
9
1

r
e
b
m
e
c
e
D
3
2
,
e
p
o
t
a
i
D
e
s
u
o
h
o
t
d
e
d
n
e
t
n
i

s
i
k
a
n
e
X
e
r
e
h
w
e
c
a
p
s

l
a
c
i
r
e
h
p
s

f
o
s
e
h
c
t
e
k
S
.
2
e
r
u
g
i
F

)
.
d
e
v
r
e
s
e
r

s
t
h
g
i
r

l
l

.
1
–
2
1

]

[

A
X
m
o
r
f

s
i
k
a
n
e
X
e
l
l
i

m
a
f
n
o
i
t
c
e
l
l
o
C
e
h
t

f
o
n
o
i
s
s
i

m
r
e
p

Kiourtsoglou

Figure 3. The Diatope on
the plaza of the Center
Pompidou, 1978. (Photo by
Bruno Rastoin, Collection
famille Xenakis. All rights
reserved.)

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

of the membrane inside Diatope, little reflection
or reverberation was produced except for specific
places on the surfaces of double curvature (e.g., the
parabola’s focal point). Malinowsky also recalled the
perforated acoustic panels of 15-cm-thick fiberglass,
suspended from the metallic structure to provide
sound absorption.

While the Diatope’s shell was under construction,

La L ´egende d’Eer was presented on 11 February
1978 by the Westdeutscher Rundfunk (WDR) at the
planetarium in Bochum, Germany, as part of WDR’s
festival Musik der Zeit IV (Xenakis 1995; Solomos
2006, p. 162). According to the official invitations
found in the archives, the Parisian premiere of the
son et lumi `ere show was programmed to take place
on 14 June 1978. Diatope would not be opened to
the public, however, until the following month,
owing to the countless technical problems relating
to the automation program (Matossian 2005, p. 276).

Diatope would stay in Paris from July until January
1979, and afterwards it would be hosted by Das
sommerlange Blumenfest in Bonn from 2 May to 28
October 1979.

Owing to the high expense of transporting the
show and of the show’s realization, it would not be
easy for Xenakis to find another city to host Diatope,
despite his intense efforts. The architectural shell
and the electronic equipment would, after Bonn, be
returned to Paris and stored in a public warehouse
at La Villette. Meanwhile, Xenakis would try to give
Diatope new life through extensive correspondence
with numerous organizations and cities (Athens,
Lyon, London, Strasbourg, Bordeaux, Chicago,
La Defense in Paris, and the Fondation d’Arc et
Senans). It wasn’t until 1981 that Diatope would
finally be bought by the Regional Office of Culture of
Provence-Alpes-C ˆote d’Azur, for the symbolic sum
of FF 1,000, to be installed at Marseilles. Despite the

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

Computer Music Journal

initial plan, the cost of installation was prohibitive,
and the Diatope structure would stay unused in a
regional warehouse in Vitrolles. Later, the metallic
structure would be sold and the rest discarded
(Sabatier 1984; Collection famille Xenakis, X[A] 12-
6; supplemented by personal correspondence with
Jacky Sabatier).

The Spatialization of La L´egende d’Eer

At the same time, Xenakis was working on the
music. For the 46 minutes of the seven-track tape of
La L ´egende d’Eer (whose title was inspired by Plato’s
Republic) Xenakis (1978, p. 10) used the following
families of sounds:

1. Instrumental sounds coming from Japanese

and African instruments

2. “Concrete” sounds, such as scraping on

cardboard or the clash of bricks

3. Sounds computed and generated by mathe-

matical functions (probabilities)

According to notes apparently written by Cornelia
Colyer, Xenakis’s assistant for computer program-
ming during this period, the computed sounds
were

written on digital tape in numerical form, then
converted into sound by a 16-bit D/A converter
at the CEMAMu laboratory in Paris. The
resulting sounds were mixed together, track by
track, in the Electronische Music Studio [sic] at
the WDR, Cologne (Collection famille Xenakis,
X[A] 13-1).

Concerning the spatialization of La L ´egende

d’Eer, Xenakis makes a brief reference:

The music is on a tape of seven tracks. Each
track is distributed over eleven high-quality
loudspeakers spread under the shell of Diatope.
The distribution, static or kinematic, is realized
by a special computer program (Xenakis 1978,
p. 11).

Here the sound’s diffusion apparently followed
the principles described in the 1958 article “Notes
pour un ‘geste ´electronique,’ ” where Xenakis (1971,

pp. 143–150) explained the notions of static and
kinematic stereophony in an electroacoustic piece.
In static stereophony, each track plays over a specific
loudspeaker and the spatial effect is produced by
the simultaneous performance of all the tracks. On
the other hand, kinematic stereophony is made up
of gradual fade-outs and fade-ins of each track as it
moves from one loudspeaker to the other, having
constant or varying speed. Here the spatial effect
is produced by the movement of the sound of each
track.

This use of complex sound movements, in
addition to the piece’s duration (at 46 minutes,
unusually long for Xenakis) and the requirements
of the light show, demanded computer automation.
This was partially practiced by Xenakis in his
earlier polytopes. At that time, the electronic
infrastructure was still a tool, whose function was
worth mentioning:

The recorded music of the seven tracks is dis-
tributed automatically by the program-score, in
continuous movements, over the eleven high
quality loudspeakers. The command is passed
by a nine-track digital tape player that decodes
an “image” of the totality of simultaneous
commands (about 2,000) every 1/25 of a second,
which are distributed by cables to their desti-
nation into the space. The performance, with
a duration of 46 minutes, involves 140,500,000
binary commands (Xenakis 1982).

Nouritza Matossian (2005, p. 270) adds:

The distribution of the sound as well as the
relative changes of volume in the seven inde-
pendent channels was also digitally coded on
the command tape. Any combination of outputs
could be directed to any one or combination of
channels.

Colyer described the automatic production of
the performance in some of her notes addressed to
Xenakis:

The production of the spectacle:

A. 7-track music tape as described before
B. 9-track computer-programmed and

computer-generated digital tape:

Kiourtsoglou

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

flash+laser score written by you, coded in
parametric form and submitted as data to our
computer program.

Program is executed and generates the
numerical information corresponding to score
and writes it onto the tapes, separating it into
blocks of information, one block being the
amount of information needed for the 1/25
sec of running the spectacle. So, 9,000 blocks
produce 45 minutes of spectacle.

The distribution of each track to any one of
the loudspeakers is computer-programmed ac-
cording to a distribution score and the necessary
information stored with the flash-laser com-
mands on the digital tape of the spectacle, that
is, the second tape necessary for the production
of the spectacle (Collection famille Xenakis,
X[A] 13-1).

This means there was a seven-track tape of
music accompanied by a nine-track digital tape of
commands (which concerned light paths, sound
intensity, and spatial distribution). Daniel Teige,
in a letter to me, explained that an eighth track
accompanied the seven-track tape of music and
served as a time-code trigger to the second, digital
nine-track tape. This same eighth track appears in
the version provided today by ´Editions Salabert in
their publication of La L ´egende d’Eer (1977, for eight-
track, 1-in., 39-cm/sec magnetic tape). Teige also
suggested that a similar technology had already been
used by Xenakis for the Polytope de Cluny (cf. Teige’s
film, available at https://youtu.be/KpWGLJODI30).
In the English text, La l ´egende d’Eer, New Version,
probably also written by Colyer, we find a similar
description:

The visual composition itself, as well as the
distribution of the seven-track sound tape gen-
erated by a computer program which, by con-
trolling the more than 2,000 above-mentioned
elements 25 times a second during the 45 min-
utes making up the presentation, translates the
visual score and the plan of sound distribution
as created by Xenakis into numerical form,
computing some 135,000,000 numbers and
storing them on the tape (Collection famille
Xenakis, X[A] 13-1).

According to this passage, we may assume
that Xenakis probably made a plan of the sound
movements for the seven tracks of La L ´egende
d’Eer before Colyer transformed them into digital
form and added them to the computer program that
controlled the whole performance.

In Paris, Xenakis spent the period from September
1975 to September 1976 recording sounds computed
and generated from mathematical functions in
the studio at CEMAMu (Centre d’Etudes de
Math ´ematique et d’Automatique Musicale). The
tapes, labeled either “Function Unit” or “Logistic
Rerun” followed by consecutive numbers, contain
these sounds, and the dates are spoken at the
beginning of each recording (Audio Collection
Iannis Xenakis).

On 7 September 1976, Xenakis accepted an invi-
tation from Wolfgang Becker, director of the Studio
for Electronic Music at the WDR in Cologne and
in charge of music production, to use that studio’s
facilities from 24 January to 24 February 1977,
assisted by the sound engineer Volker M ¨uller and
the American composer James Whitman (Collection
famille Xenakis, X[A] 11-13). Beside mathematically
generated sounds, Xenakis brought to the WDR
studio sounds he had already used in some of his
polytopes (Solomos 2006, p. 174), and together with
Volker M ¨uller he produced a family of sounds called
“M ¨uller” that open and close the piece. James
Whitman performed a double-bass improvisation
that was also recorded and used in La L ´egende d’Eer
(Solomos 2006, p. 176; Friedl 2015, p. 112). These
materials underwent sound manipulations such as
filtering, transposition by altering tape speed, and
addition of echo and reverberation. They were then
merged into seven different monophonic tracks,
which were combined in three different settings,
offering different versions of the piece. For the first
setting (the one Xenakis took back to Paris) a one-
to-one synchronization of the seven monophonic
tracks was created on an eight-track tape, saving the
eighth track to control the spatialization and the
light show (Friedl 2015, p. 114).

On 30 January 1977, when Xenakis was probably

already at the WDR, in a document showing an
early “score” of Diatope’s three different elements
(flashing lights, lasers, and music), in which words

Computer Music Journal

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

describe their temporal evolution, the composer
noted in Greek for the first five minutes of the
music: “galactic dust, white noise, . . . stochastic
sounding from the upper loudspeakers” (Collection
famille Xenakis, X[A] 11-7, p. 11).

Later, on 8 February 1977, in a document describ-

ing the evolution over time of the seven tracks,
Xenakis also noted in Greek: “On the upper four and
then circle,” referring to the path the sound had to
follow between the sixth and seventh minutes (Col-
lection famille Xenakis, X[A] 11-7, p. 14). According
to the last two documents, it seems probable that Xe-
nakis arrived at WDR to compose and mix the tracks
of La L ´egende d’Eer already having an idea about the
loudspeaker positions in the space of Diatope.

At the end of 1977, the Groupe Arcora, one of the

firms responsible for construction of the Diatope
shell, produced two plans (dated 21 November) with
the position of the loudspeakers inside the shell (see
Figure 4). According to these plans, the first four
loudspeakers were to be hung from the metallic
structure at different heights, each one forming a
specific angle with the ceiling, as shown in Figure
4a. The other seven loudspeakers are not shown in
these plans, suggesting that they may have been
positioned on the floor. These two plans can explain
the only sketch published to date, drafted by Makis
Solomos (2006, p. 167) and reproduced in Figure 5,
particularly regarding the spatialization of the seven
tracks, where ten of the numbered loudspeakers
appeared inside a curve as dots, while one (number
3) lies outside the curve. This curve is revealed to
correspond to the projected curve labeled QBPAQ of
Diatope’s shell, shown in Figure 4b. The confusion
here is caused by Xenakis’s use of principles from
architectural drawing. Loudspeaker 3, as shown in
cross section (Figure 4a), falls outside the curve
QBPAQ when it is projected onto a plane (cf. Figure
4b, where this loudspeaker is labeled P1), but it
remains inside the Diatope’s shell. I am going to use
these two plans as basis for the diagrams of sound
movement presented later in this article.

While writing the English summary for Diatope,

Colyer also confirmed the spatial placement of
the loudspeakers: “The seven-track sound tape
is distributed over eleven 75-watt J. B. Lansing
loudspeakers suspended on two levels throughout

the shell” (Collection famille Xenakis, X[A] 13-1).
Obviously, here Colyer refers to the upper-level
loudspeakers and the floor-based loudspeakers.

Another eight-track version of La Legende d’Eer
was produced in Germany for the unofficial premiere
given in the Bochum Planetarium on 11 February
1978. At that time, Xenakis did not have access
to the automation software that he would create
for Diatope in Paris four months later, so the spa-
tialization in Bochum was in great part improvised
and manually coordinated. The so-called Bochum
Version was an eight-track version, in which

Xenakis spatialized four of the mono tapes to
four different tracks (1, 3, 5, 7) by hand, using
quadrophonic effect generators EMS QUEG.
James Whitman spatialized two more tapes
to the other four tracks (2, 4, 6, 8), using two
quadrophonic joysticks, and Volker M ¨uller spa-
tialized the last material tape to all eight tracks,
using a normal fader box. This created the illu-
sion of continuous rotating movements possible
for all the spatialized materials, even when they
could not be projected onto each track. This
final spatialization had to be recorded on an
eight-track tape without interruption (Friedl
2015, p. 114).

Is there a reason for such a grouping of mono-
phonic tapes into different tracks (for instance, to
produce a specific spatial phenomenon) instead of
a simple division into odd and even tracks? Not
having access to the material of each tape, it is
difficult to reach a conclusion. Volker M ¨uller’s
recollection of the Bochum performance, however,
gives us a hint as to how this manually controlled
spatialization was actually attained: Several people,
positioned on different levels of the building (owing
to lack of space), operated the seven tape machines
while using an intercommunication system to syn-
chronize their actions (Friedl 2015, p. 115). This
means that in the Bochum version the spatialization
already consisted of producing rotating movements
of sound diffused from loudspeakers and tape ma-
chines placed at different levels in space. As I will
show shortly, this was also the main principle be-
hind the spatialization of the Diatope version of La
L ´egende d’Eer.

Kiourtsoglou

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

Figure 4. Cross-section of
the Diatrope’s space with
loudspeaker positions
viewed from the side (a)
and from above (b). The
four suspended
loudspeakers shown in

these plans are labeled P1
(corresponding to
loudspeaker 3 in the
sketch by Xenakis in
Figure 5), P2 (loudspeaker
4), P3 (loudspeaker 1), and
P4 (loudspeaker 2).

Original plans by Groupe
Arcora, redrafted by the
author. (Source: Archives
du Centre Pompidou, Box
92037/744.)

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

The sketches and plans in the Xenakis archives

detail this meticulous elaboration of the sound
movements of the seven tracks over the eleven
loudspeakers inside the Diatope’s space. Not

knowing whether these same movements were
produced during the final performance does not
minimize the value of this preparatory work for
the reconstruction of the spatialization of La

Computer Music Journal

Figure 5. Sketch by
Xenakis as published by
Solomos (2006, p. 167)
concerning the
spatialization of La
L ´egende d’Eer with eleven
loudspeakers, labeled with

Arabic numbers, and
seven tracks, indicated by
Roman numerals (a). The
positions of upper-level
loudspeakers as given in a
reproduction of Archora’s
plan (b). Color-coded

versions available at
http://www
.mitpressjournals.org
/doi/suppl/10.1162/COMJ
a 00437.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

L ´egende d’Eer inside Diatope that I am attempting
here.

I first studied Xenakis’s plan, shown in Figure 6,
describing the sound movements for the first 4(cid:2)50(cid:2)(cid:2)
of the seven tracks over the eleven loudspeakers
(almost one ninth of the piece’s total duration
of 45 minutes), where Arabic numerals are used
for the loudspeakers and Roman numerals for the
tracks. As shown in Figure 7, track I starts at
loudspeaker 5 at full volume and at the 40(cid:2)(cid:2) mark
passes to loudspeaker 1. It then passes successively
to speakers 2, 3, 4, 5, and so on until 2(cid:2)25(cid:2)(cid:2), when
it passes to the loudspeaker 6 and stays there until
4(cid:2)15(cid:2)(cid:2), to then pass on to loudspeakers 5, 1, 2, 3, 4,
and 5. It then remains at loudspeaker 5 until 4(cid:2)50(cid:2)(cid:2).
Does this mean that the sound repeats the same
movements every 4(cid:2)50(cid:2)(cid:2)? It is difficult to say for sure
without access to the eighth track, which controlled
the whole performance but is now inaccessible.
In this same document, Xenakis distinguishes the
two distinct types of sound spatialization by using
different graphic representations. A red horizontal
line represents the continuous reproduction, for
example, of track 1 by loudspeaker 5 or 6 (static

stereophony); whereas red points, when joined by
diagonal lines, symbolize the loudspeakers used for
the sound movement (kinematic stereophony).

If we now transpose the sound movements of all
seven tracks from this document into the Diatope’s
space (the curve labeled QBPAQ in Figure 8a), we
observe that, during the first seconds of the piece,
tracks I, II, IV, and V are playing on the upper
loudspeakers (1–4) and then move down to the
floor-based loudspeakers (5–11), forming circles. On
the other hand, tracks III, VI, and VII move over
the floor-based loudspeakers, forming circles. These
tracks do not use the upper loudspeakers, except
the fourth, which is the lowest of the suspended
speakers according to the plans drafted by Arcora, as
mentioned earlier. One can imagine how a spectator
might have felt, listening to sounds moving in circles
or traveling throughout the space of the Diatope,
while thousands of small lights were twinkling
and two powerful lasers were cutting through the
darkness: a human, helpless and amazed by this
genesis of the cosmos.

At the upper side of the document in Figure 6,
under the time axis (measured in seconds), we find

Kiourtsoglou

Figure 6. Spatialization of
the seven tracks over the
eleven loudspeakers by
Xenakis. A color version is
available at http://www
.mitpressjournals.org

Figure 7. Spatialization of
the ﬁrst track over the
eleven loudspeakers by
Xenakis, corresponding to
that shown in the Figure 6.
Plan transcribed by the

author. A color-coded
version is available
at http://www
.mitpressjournals.org
/doi/suppl/10.1162/COMJ
a 00437.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

Figure 6

Figure 7

the relative block of information (1/25 sec) needed
for automation of the performance. For the first 70
sec, 1 mm of graph paper corresponds to 1 sec, such
that every 5 mm we find 128 blocks (5 sec should
actually be 125 blocks). From 70(cid:2)(cid:2) to 140(cid:2)(cid:2), 1 mm
corresponds to 3 sec, such that every 5 mm we find

384 blocks (3 × 128). Finally, from 140(cid:2)(cid:2) until 290(cid:2)(cid:2),
1 mm corresponds to 0.75 sec, such that every 5 mm
we find 96 blocks (0.75 × 128). Could this possibly
mean that the density of the events taking place
was changing over time? Some of these numbers
are used by Xenakis to indicate the sound’s position

Computer Music Journal

Figure 8. Transcription of
the ﬁrst 4(cid:2)50(cid:2)(cid:2) for the seven
tracks over the eleven
loudspeakers as shown in
Figure 6, viewed from
above (a) and from the
side (b). Drawing by the

author. A color-coded
version is available
at http://www
.mitpressjournals.org
/doi/suppl/10.1162/COMJ
a 00437.

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

in space, its loudness, or even the light event to
be executed by the computer program at that point
in time. Otherwise, what else can the note “starts
1793 + bring 10 up” in track III mean, other than
an instruction to start executing the information
stored inside the 1,793rd 1/25-sec block (concerning
loudness or other elements of the performance, such

as light configuration) and to go from loudspeaker
10 to loudspeaker 5 (referring to sound movement)?
It can be supposed from the presence of these
numbers that this version of the spatialization is
almost final, because the computer program was
set up in the last four months before the Parisian
premiere.

Kiourtsoglou

Figure 9. Document by
Xenakis showing
graphically the gradual
entrance of the seven
tracks of La L ´egende d’Eer.
A color-coded version is
available at http://www

D
o
w
n
o
a
d
e
d

f
r
o
m
h

t
t

:
/
/

d
i
r
e
c
t
.

i
t
.

e
d
u
/
c
o
m

j
/

a
r
t
i
c
e
–
p
d

f
/

4
1
4
8
2
0
0
5
1
6
4
/
c
o
m
_
a
_
0
0
4
3
7
p
d

b
y
g
u
e
s
t

o
n
0
8
S
e
p
e
m
b
e
r
2
0
2
3

In this same document referring to the first
4:50, we can observe the gradual entrance of the
seven tracks at the beginning of the piece. This
information is also available on the “graphic score”
used for the synchronization of the tracks as
published by Solomos (2006, pp. 165–166). Another
document in Xenakis Archives, reproduced in
Figure 9, graphically shows this same gradual,
linear entrance of the seven tracks. In this last
document, the starting speaker for each track is
indicated along with the number corresponding
to the relative 1/25-sec package of information of
the computer program. Track I starts immediately
after the beginning of the piece whereas the other
tracks start gradually, being silent at the beginning.
Here again, for the first minutes of the piece,

tracks I, IV, and V were diffused from the upper
loudspeakers, realizing Xenakis’s intentions for the
first minutes of La L ´egende d’Eer, as mentioned
earlier: “galactic dust, white noise . . . stochastic
sounding from the upper loudspeakers, noise with
articulation” (Collection famille Xenakis, X[A] 11-7,
p. 11). Indeed, we could suppose that the “metallic
bars” (Xenakis’s description of the sound used
during the first minutes of the piece commented
by Solomos [2006, pp. 168–170]) when they were
transmitted from the upper loudspeakers, could
produce the sensation of “galactic dust” to the
public.

A version of the computer program, mentioned
earlier, used for the automation of the sound move-
ments of seven tracks over the eleven loudspeakers,

Computer Music Journal

Table 1. Sound Score

108

0.0

2748

II 4

14 TRACK1

2 >0

2< −5.12 2 >1
5
2
1
2
5
2
9
2
5
2
5
2
4
2

2< 6 2 6 10 6 1 5 2 2 2 2 2 1 2 1 108 45 2703 II4 8 TRACK2 2 >0

7< −5.12 2 2 2 2 2 2 2 2 51 70 2678 II4 8 TRACK3 2 2 2 1 4 11 4 10 4 7 4 6 10 6 2 5 1 5 1 5 1 5 1 2 2 2 2 2 2 2 2 5 1 5 1 5 1 5 80 9 2 9 2 6 2 6 2 >0 <10 −5.12 2>10 <5 7 11 8 2 2 1 7 11 8 8 4 110 2 2 2 8 4 8 9 5 10 2 30 2>2

1< 1 2 6 2 6 10 6 1 5 2 2 2 2 2 2 2 2 2 3 7 3 7 11 85 2 45 7 2 9 2 9 2 6 2 6 2 2 2 2 2 2 2 2 8 3 10 3 10 3 7 3 7 4 8 4 8 4 5 3 5 1 5 9 5 20 4 3 7 3 7 11 6 2 5 2 2 2 2 1 2 1 2 2 2 2 1 4 8 4 8 4 5 3 1 8 3 10 3 10 3 7 3 5 9 5 10 2 2 2 2 2 1 2 1 4 11 4 25 4 35 4 6 10 6 120 “Sound score” with movements of the seven tracks numerically codified, probably written by Cornelia Colyer (Archives CEMAMu, dossier 123). The meaning of the numbers and columns is discussed in the main text of this article. Original handwritten table transcribed by the author. can be found at the archives of CEMAMu, in a document named “sound score” (Archives CE- MAMu, dossier 123). This document could be the “program-score” mentioned by Xenakis. The author of this program is probably Colyer, despite the fact that her name is not mentioned anywhere in the document, based on a comparison of the graphic character of this document with texts found at the archives Xenakis and signed by Colyer. The author of this program distinguished two states of sound distribution. State 1 was described as: This could be what Xenakis called static stereophony in his article “Note sur un ‘geste Electronique’ ” (Xenakis 1958). On the other hand, state 2 was described by Colyer as: Given track moving from speaker A to speaker B. We add a fourth parameter when we want to change # sec for a to b. This new speed stays in effect until next change (ex. 2-5-6-2, that means the second track moves from speaker No. 5 to speaker No. 6 in 2 seconds). Given track to given speaker during given # of seconds at full alternation (ex. 1–5–10 that means the first track at the 5th loudspeaker for 10 seconds). This could be what Xenakis called kinematic stereophony in that same article. Next to these instructions, we find a table with the movements of the seven tracks encoded numerically, reproduced in Table 1. For example, Kiourtsoglou 21 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Figure 10. Transcription of the numerical “sound score” (Table 1) for the seven tracks over the eleven loudspeakers into Diatope’s space. Drawings by the author. Color-coded versions are available at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437 and at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437. l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 reading Table 1 horizontally, track I starts at state 2, meaning the sound moves from 0 (probably silence?) to speaker 2 in 5.12 sec. Then it passes to state 1, as the sound stays at speaker 2 for 30 sec and then moves on to speaker 1 in 5.12 seconds. Here we do not have a new fourth parameter, so we reuse the previous value for the parameter, following the programmer’s notes that “this speed stays in effect until the next change is mentioned.” All the commands of this table can be read in the same way. If we now transcribe the sound movements of the seven tracks from the above sound score of Table 1 into Diatope’s space, we obtain Figure 10. Here, tracks I, II, IV, and V are at the beginning, following the trajectory 1–2–3–4 many times. This means they are moving over the upper loudspeakers and then 22 Computer Music Journal Figure 11. The evolution of the seven tracks over the eleven loudspeakers for the ﬁrst 6 min, represented here on a two-axis system (loudspeakers versus time). A color-coded version is available at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437. (Reprinted with kind permission of the Collection famille Xenakis from X[A] 11-7, p. 27. All rights reserved.) l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 more or less forming circles over the floor-based loudspeakers. At the same time, tracks III, VI, and VII form circles on the floor-based loudspeakers without using the suspended loudspeakers (with the exception of speaker 4). Xenakis’s notes describing the first minutes of the piece also indicate these movements: “On the upper four and then circle” or “galactic dust . . . from the upper loudspeakers.” The numerical sound score of Table 1 may well be the transcription of the sound’s distribution plan (cf. Figure 6) made by Xenakis for the first 4’50”, since their resulting sound movements are quite similar (see Figures 8 and 10). In addition, their sound speed is almost the same (movement between two loudspeakers in 5 seconds for Xenakis’s plan and in 5.12 seconds for the sound score of Table 1). Nevertheless, we cannot say if the sound movements of the numerical sound score were kept unchanged, as in the same folder at CEMAMu’s archives we also find “sound tests” using slightly different rates for the sound speed and the duration of sound movements. This does not prevent us, though, from assuming that the movements described here could be very close to the ones conceived and followed during the whole or a part of the performance. In any case, there is sufficient evidence that Xenakis organized the sound movements of La L ´egende d’Eer inside the Diatope’s space with great precision. In concluding this section, it is interesting to see how the alternation of the loudspeakers for the above spatialization was conceived. A schema found in the archived materials serves here as a guide (cf. Figure 11). It is an A3-format page in size and represents the evolution of the Kiourtsoglou 23 Figure 12. Graphic representations of the numerical sound score for the seven tracks over the eleven loudspeakers (a) and sound movements for the ﬁrst 4(cid:2)50(cid:2)(cid:2) (b), both in two-axis systems (loudspeakers versus time). Plans drawn by the author. A color-coded version is available at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437. l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 seven tracks over the eleven loudspeakers for the first 6 minutes of the piece in a two-dimensional system (loudspeakers versus time). Likewise, we can geometrically represent the numerical sound score of Table 1 or the plan of 4’50” from Figure 6 in a similar two-axis system (see Figure 12). In all these schemata, the lines parallel to the time axis show state 1 of the sound (that is, its static distribution) while the lines forming an angle with this axis show state 2 of the sound (that is, its kinematic distribution). Moving backwards, similar schemata could be derived from various sketches of configurations of colored points made by the architect-composer on graph paper also found in the Xenakis Archives (X[A] 11-7, pp. 12, 25, and 26). As shown in Figure 13, seven series of points (in the original with distinct colors, probably representing the seven tracks) are organized diagonally in eleven lines (representing the eleven loudspeakers) on graph paper. (Each box of the graph paper represents a time unit, perhaps 5 seconds.) If, as in Figure 14, we join points of the same color by a line, we represent the sound movements of the seven tracks over the eleven loudspeakers over time, exactly as in the diagrams drawn with two axes in Figure 12. This simple graphic method, used to choose and connect loudspeakers and tracks, although different, may not be all that far from the sophisticated method of sieves used in Xenakis’s instrumental music from 1977 to 24 Computer Music Journal Figure 13. Seven series of points (colored in the original and probably representing the seven tracks) organized diagonally in eleven lines (representing the eleven loudspeakers) on graph paper. A color-coded version is available at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437. (Reprinted with kind permission of the Collection famille Xenakis from X[A] 11-7, p. 25. All rights reserved.) l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 1993. In this period, Xenakis used seives to choose the values of certain sound parameters, such as pitch and note duration (cf. Gibson 2011, pp. 81–114). It is interesting to add that Xenakis had been literally “drawing” sound movements ever the since “Notes pour un ‘geste ´electronique’ ” and Concret PH, which was diffused inside Philips Pavilion. While working on the spatialization with the Philips engineers, Xenakis conceived the sound as moving very quickly across a series of loudspeakers (points), placed along the lines of the hyperbolic paraboloids of the pavilion. Next to a sketch of the pavilion found in his diary (see Figure 15), Xenakis notes des ´eclairs de sons = droits d’un coup (“flashes of sound = lines all at once”). Here the lines of the hyperbolic paraboloids of the pavilion were colored differently to point out each of the loudspeaker groups used to create the impression of “sonic” lines. This sonic geometry, which is the basis of the spatialization of the Concret PH (or Interlude sonore, as it originally entitled in 1958), has been meticulously studied by Xenakis in three A3- format pages. This documentation, called Partition ster ´eophonique (Collection famille Xenakis), had been thought to be lost for over 50 years (cf. Tazelaar 2013, pp. 153–157; Kiourtsoglou 2016, pp. 220–225). The Light Show of Diatope For the light show, which was part of the perfor- mance that Xenakis imagined for Diatope, four Kiourtsoglou 25 Figure 14. The same point conﬁgurations as in Figure 13, joined by lines. Plan drawn by the author. A color-coded version is available at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437. l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 lasers were used (one red and three green) and 1,680 flashing electronic lights were distributed around a network of cables, affixed 50 cm below the metallic structure and following its parabolic form. Six glass columns, called a “well of light,” were placed inside the audience space, bearing some of the 400 reflect- ing mirrors (needed for the lasers’ trajectories). These mirrors were positioned in the form of a helix, and the rest were placed together with the lasers outside the security barrier. A false floor was created with glass blocks, lighting up the space and leaving about 60 cm below for cables. The electronic control panel was situated inside a truck near the tent, where a technician verified the flow of the performance. like lotus, anemones, spiders, wheels, dropper, omega, and so on, and sequences of flashing lights created glimmering “galaxies,” a “fluid star,” and “spirals” (see Figures 16 and 17). To study and calculate the duration of these light events, he used mathematical functions, “going from imaginary number (complex) to probability distributions” (Xenakis 1978, p. 12). Here too, Xenakis literally drew the laser trajectories and the configurations of the flashing lights. He used points (flashing lights) and lines (lasers) for the light show as he had done for the spatialization of the music, with points standing for loudspeakers and lines representing movement of sound. Xenakis composed the light show as a musical Nevertheless, as Dominique Druhen wrote, score, describing what each bulb and laser had to do every 1/25 sec. Using lasers, he created configurations of continuous shapes with names Music and light are independent in this perfor- mance. The way light is used and the routes 26 Computer Music Journal Figure 15. A sketch taken from one of Xenakis’s diaries. The handwriting at the bottom of the page reads des ´eclairs de sons = droits d’un coup (“ﬂashes of sound = lines all at once”), The groups of loudspeakers on the hyperbolic paraboloids of the Philips Pavilion form “sonic” lines. A color-coded version is available at http://www .mitpressjournals.org /doi/suppl/10.1162/COMJ a 00437. (Reprinted with kind permission of the Collection famille Xenakis from notebook No. 20, p. 20. All rights reserved.) l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 created by the [apparent] motion of light are programmed and automated, but the conflu- ences that are produced with music—which is also fixed in its production—are fortuitous and different at every presentation (Druhen 1995). This is consistent with Xenakis’s artistic atti- tude: It was not his habit to “translate” directly from one form of art into another, but rather to provoke similar or complementary sensations us- ing different media. The distinct levels of Diatope (music, architecture, and light show) were neither interdependent elements of the work nor merely disparate artistic projects merged together. Xenakis presented a four-dimensional theater based on analo- gies between diverse aspects of space–time reality. To achieve this, he used geometry (points and lines) as a conceptual tool, based on the premise that this Kiourtsoglou 27 Figure 16. Laser conﬁguration called “anemone” represented by lines, to be produced by reﬂecting the laser beam with mirrors. (Reprinted with kind permission of the Collection famille Xenakis from X[A] 11-6. All rights reserved.) l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 kind of representation could be abstract enough to refer both to architectural forms and to the distance covered by light and sound. It is interesting to point out that, twenty years earlier, Xenakis had disagreed with Le Corbusier’s use of images for a visual component for Le Po `eme ´electronique in the Philips Pavilion; the new art he was searching for was abstract and nonfigurative (Delalande, 1997, pp. 113–114). Xenakis would have to wait until the mid 1970s and his Polytope de 28 Computer Music Journal Figure 17. Flashing light conﬁguration using points. Produced by turning on the bulbs successively or all together. (Reprinted with kind permission of the Collection famille Xenakis from X[A] 11-10. All rights reserved.) l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Cluny and Diatope to realize this dream. For him, Diatope is “neither a ballet nor an opera” (Xenakis 1985, p. 182) but abstract theatre (in the broadest sense of the term), in the same way that the natural phenomena of storms and thunder are. Regarding Diatope as such, we can understand that the preex- isting texts accompanying the performance (Plato’s Republic; Blaise Pascal’s Pens ´ees; the Hermitic Cor- pus, attributed to Hermes Trismegistus; Siebenk ¨as, by Jean Paul; and Robert Kirshner’s article in Sci- entiﬁc American “Supernovae in Other Galaxies”) were not “translated” by Xenakis into light images or musical themes. Instead, they were presented to the public like the Diatope’s “arguments,” as they describe in another fashion some of the questions or sensations that music and light were evoking. Richard Barrett puts it well when he writes that the texts provide a selection of possible points of de- parture from which to enter the music, but also might draw the listener outwards from the music to some fascinating manifestation of human thought. While Xenakis himself is not normally at his most eloquent when giving verbal expression to his ideas (nor, I am sure, would be claim to be), he does not appropriate these texts to speak for him: They speak for themselves, but in doing so illuminate the music—not what it consists of or how it came about, but why (Barrett 2002, p. 79). Continuity and Evolution of Xenakis’s Creative Instinct In Diatope, Xenakis literally “drew” the spatializa- tion of the music and the form of the light events. Points and lines represent, respectively, static or moving sounds and discontinuous or continuous light configurations. From working on graph paper and thinking of sound in terms of points and lines, Xenakis conceived a space. That geometrically Kiourtsoglou 29 defined sound-space was also one that could be transcribed by numbers; the light and sound config- urations were, it is worth remembering, converted into a digital automation program, in which every command was represented by a sequence of binary code. As early as 1958, according to Xenakis, the electronic, digital infrastructure was the defining characteristic of the artist of a new era: “a new con- ceptual conscience, the abstraction, and a technical infrastructure—the electronic—are changing the human civilization” (Xenakis 1971, p. 150). In conclusion, it can be assumed that the par- ticular architectural form of the Diatope space determined the music of La L ´egende d’Eer in a way that Xenakis had long sought: But I say also that the effect of the architectural forms has an almost tactile influence on the quality of music or of the spectacle within which they are presented. And this, beyond any considerations of the acoustics or of the optimal proportions of the spectacle or the sonic experience (Xenakis 1978, p. 9). In 1978, with Diatope, the composition La L ´egende d’Eer, and his text “Geste de lumi `ere et de son,” Xenakis was still working in an area he had begun twenty years earlier, in 1958, with the Philips Pavilion, Concret PH, and the text “Notes pour ‘un geste ´electronique.’ ” The close relationship between space and sound would also be a crucial question in his later project for the Cit ´e de la Musique in Paris in 1984. Xenakis’s imagination never ceased finding new ways to work on ideas that had preoccupied him since the very beginning. Acknowledgments I am grateful to Franc¸ oise and M ˆakhi Xenakis, who granted access to the Iannis Xenakis archives. I am thankful to Sharon Kanach, who pointed me to the CEMAMu Archives. A first version of this work was presented in the international symposium Xenakis: La Musique ´electroacoustique, held at Universit ´e Paris VIII in May 2012 and organized by Makis Solomos, whom I would also like to thank for his advice. The research presented here is part of my doctoral dissertation in architecture, entitled Le Travail de l’analogie dans la musique et l’architecture de Iannis Xenakis, which was submitted to Universit ´e Paris 8 in 2016. My doctoral work has benefited from the financial support of the A. G. Leventis Foundation, the Panayotis and Effie Michelis Foundation, and the Foundation for Education and European Culture (IPEP). I have also been a fellow of the Greek State Scholarships Foundation (IKY), in particular the “Program of IKY Scholarships by Individual Evaluation for the Academic Year 2011–2012 from Resources of the European Program ‘Education and Life Learning’ of the European Social Fund (ESF) and the NSRF, 2007–2013.” I would also like to thank Liam Cagney and Eleni Mouatsou for reviewing earlier versions of the English manuscript. Pierre Carr ´e has been a valuable interlocutor regarding acoustics. I am thankful to Doug Keislar, Editor of CMJ, for his suggestions and to Peter Castine for his editorial work. References Archives CEMAMu. Biblioth ´eque Nationale de France, D ´epartement de Musique. Paris, Site Richelieu. Archives du Centre Pompidou. Paris, Pole Archives. Available online at www.centrepompidou.fr /content/view/full/49272. Audio Collection Iannis Xenakis. Biblioth ´eque Nationale de France, D ´epartement de l’Audiovisuel. Paris, Site Tolbiac. Barrett, R. 2002. “Musica instrumentalis of the Merciless Cosmos: La L ´egende d’Eer.” Contempory Music Review 21(2/3):69–83. Bordaz, R. 1974. Letter to Xenakis, 10 May. Archives du Centre Pompidou. Paris. Collection famille Xenakis. Paris. Available online at www.iannis-xenakis.org/fxe/asso/archives.html. Delalande, F. 1997. “Il faut etre constamment un im- migr ´e”: Entretiens avec Xenakis. Paris: Buchet/Castel. Druhen, D. 1995. “A propos de La L ´egende d’Eer”. Liner notes to Iannis Xenakis 2: La L ´egende d’Eer. Gentilly, France: Auvidis Montaigne MO 782058, compact disc. Friedl, R. 2015. “Towards a Critical Edition of Electro- acoustic Music: Xenakis, La L ´egende d’Eer.” In M. Solomos, ed. Xenakis. La musique electroacoustique. Paris: Harmattan, pp. 109–122. 30 Computer Music Journal l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Gibson, B. 2011. The Instrumental Music of Iannis Tazelaar, K. 2013. On the Threshold of Beauty: Philips Xenakis. Theory, Practice, Self-Borrowing. Hillsdale, New Jersey: Pendragon Press. and the Origins of Electronic music in the Netherlands 1925–1965. Rotterdam: V2. Harley, J. 2002. “The Electroacoustic Music of Iannis Xenakis.” Computer Music Journal 26(1):33–57. Harley, M. 1998. “Music of Sound and Light: Xenakis’s Polytopes.” Leonardo 31(1):55– 65. Kiourtsoglou, E. 2016. “Le Travail de l’analogie dans la musique et l’architecture de Iannis Xenakis.” PhD dissertation, Universit ´e Paris VIII, Unit ´e mixte de recherche “Laboratoire Architecture Ville Urbanisme Environnement” (UMR LAVUE). Matossian, N. 2005. Iannis Xenakis. Nicosia, Cyprus: Mufflon. Piano, R., and R. Rogers. 1974. Letter to Robert Bor- daz, 22 October. Archives du Centre Pompidou. Paris. Sabatier, J. 1984. Letter to the president of the Centre Pompidou, dated 9 September. Archives du Centre Pompidou. Solomos, M. 2006. “Le Diatope et La l ´egende d’Eer de Iannis Xenakis.” In B. Bruno, A. Veitl, and M. Battier, eds. Musique, instruments, machines: Autour des musiques electroacoustiques. Paris: Universit ´e Paris IV, Musicologie, informatique et nouvelles technologies, pp. 161–196. Sterken, Sven. 2001. “Towards a Space–Time Art: Iannis Xenakis’s Polytopes. Perspectives of New Music 39(2):262–273. Xenakis, I. 1958. “Notes sur un ‘geste ´electronique.’ ” In I. Xenakis. 1971. Musique architecture. Tournai, Belgium: Casterman, pp. 143–150. Xenakis, I. 1971. Musique architecture. Tournai, Belgium: Casterman. Xenakis, I. 1974. “Les Polytopes de Beaubourg.” First published in I. Xenakis. 2006. Musique de l’architecture. Marseille: Parenth `eses, pp. 347–351. Xenakis, I. 1978. “La L ´egende d’Eer (Premi `ere version). Geste de lumi `ere et de son du Diatope au Cen- tre Georges Pompidou.” In Le Diatope: Geste de lumi `ere et de son. Paris: Centre Georges Pompidou, pp. 8–12. Reprinted in Xenakis, I. 2006. Musique de l’architecture. Marseille: Parenth `eses, pp. 353–356. Xenakis, I. 1982. “La Composition musicale est `a la fois d ´ependante et ind ´ependante de l’ ´evolution tech- nologique des syst `emes analogiques ou num ´eriques.” In Conf ´erences des journ ´ees d’ ´etudes, pp. 137– 155. Reprinted in I. Xenakis. 2006. Musique de l’architecture. Marseille: Parenth `eses, pp. 358–366. Xenakis, I. 1985. “Music Composition Treks.” In C. Roads, ed. Composers and the Computer. Los Altos, California: Kaufmann, pp. 171–192. Xenakis, I. 1995. La L ´egende d’Eer. Paris: Auvidis Montaigne MO 782058, compact disc. Xenakis, I. 2006. Musique de l’architecture. S. Kanach, ed. Marseille: Parenth `eses. Kiourtsoglou 31 l D o w n o a d e d f r o m h t t p : / / d i r e c t . m i t . e d u / c o m j / l a r t i c e - p d f / / / / 4 1 4 8 2 0 0 5 1 6 4 / c o m _ a _ 0 0 4 3 7 p d . j f b y g u e s t t o n 0 8 S e p e m b e r 2 0 2 3 Elisavet Kiourtsoglou image

Download pdf