Analysing the First Electroacoustic Music of Iannis Xenakis

5th European Music Analysis Conference, Bristol, April 2002

Makis Solomos
Université Montpellier 3, Institut Universitaire de France
[email protected]

XENAKIS’ ELECTROACOUSTIC MUSIC AND ANALYSIS

The number of electroacoustic works produced by Iannis Xenakis represents only a

slight percentage of his overall output: about one-tenth, including the few mixed-media pieces
coupling instruments with tape. Nevertheless, their importance is enormous at least for two
reasons. First, like Stockhausen or Berio, Xenakis was one of the firsts to compose
electroacoustic music in the historical sense of the word, filling the gap between the two
antagonistic trends of the early fifties: the concrète and the electronic approach. Second,
unlike Berio, who after some early works stopped composing for tape, Xenakis continued to
do so; and unlike Stockhausen who did follow the dramatic evolution of this relatively new
musical genre, Xenakis also significantly contributed to its technology.
We can classify the 16 electroacoustic works Xenakis composed into three categories
according to the techniques they imply. This classification also corresponds to three periods in
his production:
1) The most important reaches from 1957 to 1977. These works use concrète and synthesised
sound, and thus belong to the electroacoustic genre in the historical sense of the word
mentioned above. This category can be further subdivided into:
a) works composed in the studios of GRM (Groupe de Recherches Musicales) between
1957 and 1962: Diamorphoses (1957), Concret PH (1958), the tape part of Analogique
(Analogique B, 1959), Orient-Occident (1960) and Bohor (1962)1.
b) after the break with Schaeffer (see below): works composed in various studios between
1969 and 1977 and intended for multimedia productions, namely the famous
polytopes: Hibiki Hana Ma (1969-70), Persepolis (1971), Polytope de Cluny (1972),
La légende d'Eer (1977) —in this category we can add the tape of Kraanerg
(1968-69).
2) The four works composed for/with the UPIC, the musical “drawing board” developed by
Xenakis and his team at CEMAMu2, beginning in 1975: Mycènes alpha (1978), the tape of
Pour la Paix (1981), Taurhiphanie (1987) and Voyage absolu des Unari vers Andromède
(1989).

1 Two other works have been withdrawn by Xenakis: Vasarely (1960) and Formes rouges (1961).
2 Cf. Gérard Marino, Marie-Hélène Serra, Jean-Michel Raczinski, “The UPIC System: Origins and
Innovations”, Perspectives of New Music vol.31 n°1, 1991, p. 258-269.
 2

3) The two works composed with the program GENDYN which Xenakis completed in 1991
at CEMAMu with the help of Marie-Hélène Serra3: Gendy3 (1991) and S709 (1994).
We see in this classification a clear evolution where Xenakis always finds himself in the
centre of the musico-technological discourse concerning electroacoustic, electronic, tape
music, etc.4. His first pieces contributed, as stated, to the birth of electroacoustic music in its
historical sense. With the UPIC, Xenakis was looking for a new way to ease the access to
music-making with the computer — of course his method (the drawing board) has its
advantages and disadvantages. Finally, through the music of the third period, Xenakis
explored the limits of algorithmic techniques applied to music.
Until today, Xenakis' electroacoustic production has not aroused the interest of many
analysts. For the first category, we have studies of Agostino Di Scipio, Makis Solomos and
Stefania de Stefano5. For the second one, there is only one analysis, of Ronald J. Squibbs6.
And for the third, we have the analysis of Agostino Di Scipio and Peter Hoffmann 7. The
reasons of this luck of interest are not to be found in the particularity of Xenakis’
electroacoustic music, but in the difficulty of analysing electroacoustic music in general. With
the first category of works, the difficulty is the absence not only of scores, but also of any
documents on their composition. With the UPIC pieces, there is a “score”, but which is not
helpful as are instrumental scores. Finally, with the GENDYN compositions, we can work
with the program, but that doesn’t mean that we can answer easier to analytical questions.

XENAKIS’ FIRST ELECTROACOUSTIC COMPOSITIONS

To introduce the analysis of Diamorphoses and Orient-Occident — in fact rather a

graphical representation as an attempt for their analysis —, I shall briefly recall the famous
quarrel between Xenakis and Pierre Schaeffer. Xenakis was probably present at the first
concert of musique concrète in 1950, at a time when he was studying with Olivier Messiaen

3 Cf. Marie-Hélène Serra, “Stochastic Composition and Stochastic Timbre: GENDY 3 by Iannis Xenakis”,
Perspectives of New Music vol. 31 n°1, 1993, p. 236-257.
4 Note the relative disinterest of Xenakis in mixed (tape and instrumental) music: only 3 out of the 16
compositions are of that kind: Analogique, Kraanerg and Pour la Paix.
5 Cf. Agostino Di Scipio, “Compositional Models in Xenakis’s Electroacoustic Music”, Perspectives of New
Music 36 n°2, 1998, p. 201-243; Agostino Di Scipio, “The problem of 2nd-order sonorities in Xenakis'
electroacoustic music”, Organised Sound 2 n°3, 1997, p. 165-178 (extended as “Clarification on Xenakis: the
Cybernetics of Stochastic Music”, in Makis Solomos (ed.), Présences de Iannis Xenakis / Presences of Iannis
Xenakis, Paris, CDMC, 2001, p. 71-84); Makis Solomos, A propos des premières œuvres (1953-1969) de I.
Xenakis, Ph.D. dissertation, Université Paris 4, 1993, p. 263-272 (the graphic transcription of Diamorphoses and
Orient-Occident presented in the present paper are issued from this study); Stefania de Stefano,
“Spettromorfologie e articolazione strutturale in Diamorphoses (1957) di Iannis Xenakis”, in M.C. De Amicis
(ed), Atti del Congresso di Dittatica della musical elettronica, L'Aquila, Instituto Gramma, 1998, p. 131-133.
6 Cf. Ronald J. Squibbs, “Images of Sound in Xenakis' Mycenae-Alpha”, in Gérard Assayag, Marc Chemillier,
Chistian Eloy (ed.), Troisièmes journées d'informatique musicale JIM 96 = Les cahiers du GREYC 4, 1996, p.
208-219.
7 Cf. Agostino Di Scipio, op. cit. (the two articles); Peter Hoffmann, “Analysis through Resynthesis. Gendy3 by
Iannis Xenakis”, in Présences de Iannis Xenakis, op. cit., p. 185-194; Peter Hoffman, Makis Solomos, “The
Electroacoustic Music of Xenakis”, in Proceedings of the First Symposium on Computer and Music, Corfu,
Ionian University, 1998, p. 86-94. For the programm GENDYN, cf. also Peter Hoffmann, “Implementing the
Dynamic Stochastic Synthesis”, in Gérard Assayag, Marc Chemillier, Chistian Eloy (ed.), op. cit., p. 341-347.
 3

and composing music in the spirit of Bartók. In 1953, he tried to get access to Schaeffer's
studio. Thanks to a recommendation by Messiaen, he met Schaeffer in 1954. During four
years he was more and more involved in the projects of this group. In 1958, when the
“inventor of musique concrète” attempted to align his colleagues, the quarrel started. In 1959,
Schaeffer spoke very roughly about Analogique B. The latest clash took place in 1963, when
Xenakis proposed the use of mathematics and the computer in the studio. On the refusal of
Schaeffer, Xenakis left the group.
We still do not know exactly what happened in this historical dispute. I invited
Françoise Delalande to speak about that in the symposium Présences de Iannis Xenakis
(Paris, Radio France-CDMC, January 1998). But his paper8 is very general, either because he
didn’t search for documents (and there are documents in the INA/GRM’s archives!), either
because he didn’t want to speak about that. It would be very interesting, though, to know what
were the arguments of Schaeffer against the use of computers because it is well known that,
some years later, he changed his mind. Maybe we will know one day, when somebody will be
assigned the task to transcribe the kilometres of tapes with the recordings of the GRM
discussions of that time… Nevertheless, it is clear that Xenakis and Schaeffer could not find
common grounds: Xenakis spoke much about abstraction but was very pragmatic in the end,
whereas Schaeffer spoke much about “concreteness” while his theoretical thought tended to
be very abstract (his Traité des objets musicaux is actually an ambitious project to found a
phenomenology of music9).
It has often been said that the composers of the fifties were much influenced by their
electroacoustic experience, transferring their results in this domain to their writing for
instruments. David Ewen e.g., wrote that Xenakis “explored the possibilities of simulating
electronically produced sounds and sonorities with conventional instruments” 10 . Hugues
Dufourt, probably thinking of his own music, repeated the same statement11. However, this is
not true. It may be that electroacoustic practice made composers like Ligeti, Stockhausen or
Berio discover radical new ways of conceiving music in general — and, consequently, they
applied these concepts in their instrumental music. But Xenakis' case is similar to Varèse's,
who wrote a radically new music before the introduction of the new technology, a music
which was no more composition with sounds but composition of sound. Xenakis developed
this concept already in his orchestral works Metastaseis (1953-54) and Pithoprakta (1955-56),
i.e. before his electroacoustic experience. Even more so, his first electroacoustic pieces 12
continue the musical thinking introduced by the two orchestral works mentioned above, while
the major part of his instrumental production of that time (1956-62) 13 opens a kind of
parenthesis. Trying to systematise and to implement the stochastic techniques introduced in
bars 52-59 of Pithoprakta, the compositions Achorripsis (1956-57), Analogique A (1958) and

8 Cf. François Delalande, Evelyne Gayou, “Xenakis et le GRM”, in Présences de Iannis Xenakis, op. cit., p.
29-36.
9 Cf. Makis Solomos, “Schaeffer phénoménologue”, in Ouïr, entendre, écouter, comprendre après Schaeffer,
Paris, Buchet/Chastel-INA/GRM, 1999, p. 53-67.
10 David Ewen, Composers of Tomorrow's Music, New York, Dodd Mead and Co, 1971, p. 125.
11 Cf. Hugues Dufourt, “Hauteur et timbre”, Inharmoniques n°3, 1988, p. 69.
12 With the exception of Analogique B.
13 With the exception of Syrmos (1959) and Herma (1961).
 4

the 5 ST pieces (ST/48, ST/10, ST/4, Morsima-Amorsima and Atrées: 1956-62) abandon for a
while the concept of composing-the-sound in favour of a much more austere sound world. In
contrast, the electroacoustic pieces of this epoch (Diamorphoses, Concret PH,
Orient-Occident and Bohor) pursue and develop the sound concept even further.

DIAMORPHOSES AND ORIENT-OCCIDENT

Let us now focus on Diamorphoses and Orient-Occident. Diamorphoses provided an

opportunity for Olivier Messiaen to pay tribute to his former student by stating that “ce sont
de gigantesques toiles d'araignées dont les calculs préalables se muent en délices sonores de la
plus intense poésie”14, alluding probably to the passage of glissandi between 6:00 and 6:16.
But we don’t know if Xenakis made calculations. My hypothesis is that he didn’t, and that
everything in Diamorphoses was worked out empirically.
The aims of Xenakis in Diamorphoses are numerous and interrelated. We can mention it
as a study on:
1. noises
2. the duality continuity-discontinuity and especially of how continuity can be produced by
discontinuity
3. glissandi
4. variations of density and their logarithmic perception
5. stochastic processes
6. emergence of complex sonorities (sounds and forms) through 2, 4 and 5
7. dramatic form.
Here are Xenakis’ own words about Diamorphoses, for those analysts who would like
to search the ways it was composed:
-in Formalized Music, on the beginning of chapter 2, Xenakis writes: “Now we can rapidly generalize the study
of musical composition with the aid of stochastics. The first thesis is that stochastics is valuable not only in
instrumental music, but also in electromagnetic music. We have demonstrated this with several works:
Diamorphoses […], Concret PH […]; and Orient-Occident […]”15.
-the sound material is made of “décollages de jets, tremblements de terre, accidents de chemins de fer, ainsi que
des instruments de musique tels que cloches, percussion et vents”16.
-before beginning the composition, the aims were: “Mélanger des timbres en vue d’arriver à un corps sonore
pareil à un bruit blanc; étudier l’évolution des timbres, de la dynamique et du registre; faire des unissons
avec attaques seules, avec ou sans transposition; faire des chromosomes d’attaques”17.
-to the question of Varga: “Diamorphoses was the first among your electroacoustic compositions. What did you
learn in the course of its realization?”, Xenakis answers: “I drew several conclusions. For instance, that by
dense mixing one can obtain continuous sounds out of discontinuous ones. It also became clear that there is a
logarithmic relationship between the increase in density and its perception. I made the following experiment:
I recorded sounds on tape without regular rhythm, of irregular density. Then I copied the recording and
mixed it with itself. However, I edited the second tape in such a way that I avoided any repetition and echo:
the sequence of sounds didn’t correspond to that on the first tape. When I listened to the new recording I
asked myself: Can I hear the difference? And I found that I had arrived at the border of a change. I then

14 Olivier Messiaen, “Préface”, Revue Musicale n°244, 1959 p. 5.

15 Iannis Xenakis, Formalized Music, Stuyvesant (New York), Pendragon Press, 1992, p. 43.
16 This quotation is of Nouritza Matossian (Iannis Xenakis, Paris, Fayard, 1981, p. 148), but the words are
probably those of Xenakis.
17 Xenakis in idem.
 5

carried out these operations with three tapes. The result was that the density undoubtedly increased by one
step. In order to perceive an increase of the same magnitude I had to mix the three tapes once again, with the
help of three tape-recorders. The density was now nine times greater but, as I said, my senses perceived it as
an increase of only one step. Here is the proof of the logarithmic connection”18.
-in his interview by Delalande, Xenakis says: “Il y a des parties qui sont très construites, avec des petits glissandi
de cloches. J’avais des cloches merveilleuses qui produisaient des sons très intéressants ; je les avais
enregistrés et je les faisais glisser ; et puis ensuite je les mixais de manière conforme à des distributions de
probabilité pour obtenir des formes de son nouvelles, et intéressantes, bien sûr. Ca, c’est surtout la deuxième
partie de Diamorphoses. Donc ces préoccupations étaient à la fois abstraites, autant que possibles, et
globales, et aussi terre à terre, travaillant avec le matériau existant ; par exemple, j’étais très content de
pouvoir utiliser, de nouveau dans Diamorphoses, des bruits qui n’étaient pas considérés comme musicaux et
que, je crois, personne n’avait utilisés de cette façon-là avant moi. Je prenais des chocs de bennes, des choses
comme ça, des tremblements de terre enregistrés vite, et puis je les mettais ensemble pour essayer de
comprendre aussi bien leur nature interne, par opposition ou par similitude, et de les faire évoluer, et faire
passer de l’un à l’autre. Et ça, ça ne pouvait se faire qu’en travaillant sur le tas avec le matériau même. En
mettant ses mains dedans, dans ses intestins, et pas d’une manière abstraite. Mais d’autres parties, d’autres
passages de Diamorphoses étaient faits de manière plus abstraite comme dans d’autres pièces, comme
Analogique B”19.
As for Orient-Occident, one of the first histories of electronic music ever written refers
to it as a major masterpiece for tape20. And it is worth pointing out that it happens to be one of
Xenakis' “easier” works, probably due to the narrative project behind its composition.
Orient-Occident was conceived as music for a short film by Enrico Fulchignoni,
commissioned by UNESCO. The music traces the film's development which relates the
passage from one civilisation to another, from prehistoric times to Alexander the Great.
Although it is certain that Xenakis did not compose an “illustrative” music, some of his
chosen sonorities are quite suggestive. For example, the highly reverberated atmosphere
toward the end (beginning at 8:00) seems to evoke the later civilisations of Antiquity, marked
by a special sensuality, as has been written by the author of the film21.

ANALYSIS OF DIAMORPHOSES AND ORIENT-OCCIDENT

For the graphical transcription (see figures 1 and 2) of these compositions in an attempt
for their analysis, I rely on perception rather than on a physical sound analysis (e.g. a
spectrogram). My interest was not so much the physical aspect of these pieces (neither was it
the physical sound sources) nor the technology used by Xenakis, but mainly two things:
1. the types of sounds used by Xenakis and their relationship with the “sonorities” of his
instrumental music. By “sonorities” I refer to my hypothesis that all Xenakis’ instrumental
music is in fact, regardless of what theories or technologies are used,

18 Xenakis in Bálint A. Varga, Conversations with Iannis Xenakis, London, Faber and Faber, 1996, p. 111. A
footnote adds: “It is a matter here, therefore, of a logarithm of base 3” (ibid., p. 218).
19 Xenakis in François Delalande, “Il faut être constamment un immigré”. Entretiens avec Xenakis, Paris,
Buchet-Chastel/INA-GRM, 1997, p. 39.
20 Cf. Herbert Ruscol, The Liberation of Sound. An introduction to Electronic Music, Prentice-Hall
International, 1972, p. 235.
21 Cf. Enrico Fulchignoni “Sur Orient-Occident, in Regards sur Iannis Xenakis, Paris, Stock, 1981, p. 260.
 6

composition-of-sound, and that it is always determined by three types of sonorities: sliding

sounds, static sounds and masses of punctual ones22.
2. the relationship between these sounds, i.e. the form of the pieces.
Thus, in the present transcriptions, which respect the two traditional dimensions of time
and pitch — of course the latter is, in the case of a music which uses almost exclusively noise
sounds, not precise at all — and which represent loudness by thickness, the most important
aspect lies in the types of sounds used: it is not the shape of the graphics which is important,
but their distinction and numbering (with the letters subdividing numbers).
I apply the first of the two above-mentioned goals to the transcription of Diamorphoses.
In this composition, the sounds are not very much differentiated because, as has been said,
here Xenakis is interested in the shades of noise. Only 15 sounds are clearly differentiated and
we can classify them into only five categories according to: a) their spectral composition:
graduation from white noise to quasi-sinusoidal sounds; b) their duration; c) the evolution of
their pitches: static or glissando; d) their density: from isolated sounds to sound masses. These
categories are:
-continuous noise: sounds a.1 to a.4
-percussive sounds: b.1, b.2
-more “pure” sounds: with sliding (sound c) or static with continuity (d.1) or interrupted by
silence (d.2)
-quasi-sinusoidal sounds organised in static small fields or small sliding (sounds e and f) or in
masses of high-pitched glissandi (sound g)
-punctual sounds isolated or in small masses: h.1, h.2 and i.
Concerning the relationships with the three above-mentioned sonorities of the
instrumental production, only sounds a, b and some e do not correspond to one of the three
types. The others correspond in the following manner:
-sliding sounds: sounds c, f and g
-static sounds: d.1, d.2 and some sounds e
-punctual sounds: h and i.
In contrast to Diamorphoses, Orient-Occident has many more different sounds: the
piece is much nearer to the concrète tradition than Diamorphoses —remember also the
narrative background of this composition. For this reason, here the interest of the analysis
rather lies in the second goal mentioned above: the relationship between the sounds. Their
typology is less important than the affinities that exist between sounds which seem very
different at first glance. The affinities can be established by:
-direct acoustic affinities: all the sounds that have the same number (1a-1b, 2a to 2g, etc.)
-indirect acoustic affinities: the pairs 2b-16, 2d-20, 4c-12a, 7-12a, 7-19, 8-10, 9-14, 17-22
-the possibility to derive one sound from another: this is only the case for sounds 17 and 18
which produce sound 21
-the association to the same well-known instrumental sounds: sounds 3, 4a, 5, 6, 7, 9, 17, 19,
22 and 23 are very near to percussion sounds

22 Cf. Makis Solomos, A propos…, op. cit., chapter 3 and 9 to 11; Makis Solomos, Iannis Xenakis, Mercuès,
P.O. Editions, chapter 5.
 7

-formal associations: sounds 3, 4a, 5, 6, 7, 8 and 9 enter simultaneously and contrast with
what precedes; sounds 11 and 13 succeed one another and have the same function; sounds
14 and 15 are always heard together
-associations full of imagery: of course here the affinities are very numerous: I mention only
the “oceanic” atmosphere which is established progressively toward the end of the
composition with sounds 17, 21, 22, 23, 24 and 25.
 8

Figure 1: Graphical transcription of Diamorphoses

 9

Figure 2: Graphical transcription of Orient-Occident