See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/317790128

Mathematics and Design in the Music of Iannis

Xenakis

Chapter · January 2012

CITATION READS

1 2,458

1 author:

Rob Wannamaker
California Institute of the Arts
21 PUBLICATIONS   530 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Xenakis View project

James Tenney View project

All content following this page was uploaded by Rob Wannamaker on 22 June 2017.

The user has requested enhancement of the downloaded file.

 Mathematics and Design in the
Music of Iannis Xenakis
Robert Wannamaker

Even brief biographies of Xenakis seldom fail to mention the application

of mathematical techniques to composition in his music of the 1950s and
1960s, and a voluminous literature has developed around these mathemat-
ical aspects of his work.1 The key pieces of this were penned by the com-
poser himself and were collected in his seminal book Musique Formelles.2 A
significant portion of the extensive secondary literature comprises exegeses
of Xenakis’ mathematical writings, rendering them more accessible to di-
verse readers and expanding upon their foundations.
I am not herein concerned to supplement this technical literature,
as interesting as much of it is to me. Instead, I propose to take a step back
from it in order to gain perspective upon the evolving role and significance—
the place—of mathematical methods within a body of work with which such
methods are so often associated. Given Xenakis’ use of mathematical proce-
dures to mechanically generate certain musical details, I will devote particu-
lar attention to his exercise of discretion in combination with such methods.

1
A non-exhaustive list of such techniques (and representative compositions) in-
cludes geometry (Metastaseis, 1954), statistics and probability theory (Pithoprakta,
1956; Achorripsis, 1957; the ST series of pieces, 1956–62), game theory (Duel,
1959; Stratégie, 1962), set theory (Herma, 1961; Eonta, 1964), group theory (No-
mos Alpha,1966; Nomos Gamma, 1968), and cellular automata (Horos, 1986). A
theoretical survey of these techniques is provided in Benoît Gibson. The Instrumen-
tal Music of Iannis Xenakis: Theory, Practice, Self-Borrowing (Hillsdale: Pendragon
Press, 2011). An extensive bibliography can be found at http://www.iannis-xenakis.
org/xen/index.html (accessed July 1, 2012).
2
Iannis Xenakis. Musiques Formelles = Revue Musicale n° 253–54 (Paris: Editions
Richard-Masse, 1963). Translated and expanded in Iannis Xenakis. Formalized
Music: Thought and Mathematics in Composition, 2nd Ed. (Harmonologia Series
No. 6) (Hillsdale: Pendragon Press, 2001).
1
2 XENAKIS MATTERS

Found Structures and Their Transcription

Xenakis displayed interests in science and literature from a young
age. He entered Athens Polytechnic Institute in 1940, completing in 1947
a degree in civil engineering.3 It is there that he likely would have first en-
countered many of the mathematical ideas that he would later explore fur-
ther on his own. Arriving in Paris following World War II, his mathematical
facility helped to garner him employment in the studio of the pioneering
modernist architect Le Corbusier. By 1954, he had been promoted from
doing engineering calculations to creating original architectural designs.4
1954 also witnessed the completion of the composition that is
usually identified as his first mature work: Metastaseis for an orchestra of 61
instrumentalists. Therein Xenakis took the radical step of adopting the ar-
chitectural draftsman’s geometrical methods to music composition. Figure 1
reproduces a famous sketch for one brief segment of the work. It comprises
four distinct figures, each comprising a set of overlapping straight lines.

Figure 1 Xenakis’ graphical compositional sketch for Metastaseis (1954) show-

ing string glissandi, mm. 309–14 (from Xenakis, Formalized Music, 3.)

3 
Nouritza Matossian. Xenakis ( Lefkosia, Cyprus: Moufflon Publications, 2005), 28.
4 
Matossian. Xenakis, 66.
WANNAMAKER 3

Each figure is a projection onto the two-dimensional page of a portion of

a three-dimensional geometrical construction called a doubly ruled surface
(specifically, a hyperbolic paraboloid). Figure 2 shows a single generic pro-
jection.

Figure 2 Planar projection of a portion of a hyperbolic paraboloid.

Through every point on such a surface there passes two distinct

straight lines each of which lies entirely on the surface. Such straight lines in
three-dimensional space engender straight lines in a planar projection.5 This
makes it easy to produce such projections on paper with the aid of a straight-
edge, as Xenakis apparently did in this sketch. As the Figures demonstrate,
the visual impression of depth is retained in such planar projections.
The composer’s addition of conventionally notated pitch-ranges
and measure numbers to the sketch establishes a correspondence between
the two spatial dimensions in which these projections appear and the audi-
tory dimensions of pitch and time. Xenakis transcribed this sketch into his
conventionally notated score, the lines being rendered as string glissandi,
as indicated by the abbreviated names of instruments appearing in the
figure. Such a rendering may seem “natural” in part because such a cor-
respondence between physical and perceptual dimensions also exists on
the staves of conventional musical scores. In other words, on a staff, time

5
Strictly speaking, this assumes that no lines are perpendicular to the plane of
projection, since the projections of such lines would be points.
4 XENAKIS MATTERS

also unfolds from left to right and higher pitches are represented higher
on the page (albeit at discrete locations). Furthermore, the unbroken vi-
sual continuity of each line in the sketch suggests a similar continuity in
pitch-time (i.e., a glissando), and the constant slope of each line suggests
a similarly constant rate of change in pitch. The directness of this analogy
between graphical and musical figures engenders an auditory image whose
vividness apparently attracted the composer from not only a technical but
also an aesthetic perspective.

At the time of writing Metastaseis…I possessed some visual fantasy

(straight lines, for instance, that is, glissandos, came naturally for me)
which I could transform into auditive fantasy, and vice versa.6

In this way the doubly ruled surface served as a structural “found object”,
with musical imagination and aesthetic judgment playing roles in its se-
lection and in the choice of particular procedures for its translation into
sound.
The apparent exercise of compositional imagination and judg-
ment is further underscored by recognition that not all aspects of the sketch
are preserved in its musical transcription. The impression of depth—real
for a hyperbolic paraboloid but perceived in planar projections—has no
direct sonic counterpart. Dynamic and spectral brightness, however, may
easily impart the impression of source proximity. Thus the orchestrated
diminuendo as instruments drop out in the first half of the passage, and
the counterpoised orchestrated crescendo at its conclusion, may collude
with registral changes to evoke an added dimension in which sound-masses
move away from and towards the listener, respectively.7 These aural impres-
sions, however, do not correspond in their specifics to the impressions of
perspectival depth evoked in the graphical sketches. Instead, the potentially
impoverishing loss of one impression of dimensionality is compensated by
the evocation of a quite different one. Thus, while transcription of the dou-
bly ruled surface into musical notation may have constituted a useful recipe
for mechanically determining certain musical details, that procedure did
not fully determine the perceived character of the passage. Envisaging that
character would have demanded additional compositional insight.

6
Bálint András Varga. Conversations with Iannis Xenakis (London: Faber &
Faber, 1996), 47–48.
7
Even if the impression of receding and approaching sound-masses is not con-
sciously registered by a listener, I think that a corresponding affective response
may be evoked.
WANNAMAKER 5

Furthermore, numerous compositional variables were left to be

determined following Xenakis’ establishment of a correspondence between
graphical and musical dimensions. Instrumentation, for instance, required
additional decisions that would have entailed both practical and aesthetic
considerations. The scoring for strings obviously facilitated the execution
of continuous glissandi. Moreover, the timbral homogeneity of the string
choir across instruments and registers helped to preserve the aural cohe-
sion of each of the four paraboloid “surface sections”, which represent the
sort of composite sound-masses that are a hallmark of Xenakis’ early style.
Timbral heterogeneity would have promoted disintegration of these ge-
stalts into component parts.
In contrast with the extended massed string-glissando passages
that open and close Metastaseis, the passage in Figure 1 occupies only a
few seconds, perhaps because it belongs to and caps a very active “devel-
opmental” section of the music.8 Particularly developmental in character
is the quasi-contrapuntal disposition of its four sound-masses. The passage
begins with the entrance of a low mass spanning the bass register and con-
tracting upwards towards unison. This is followed by a varied imitation
in the upper register, which gradually contracts and curls downwards to
meet the tail of the first.9 Almost dovetailed with this is an upper-register
upward-expanding sound-mass, suggesting a retrograde-inverted diminu-
tion of the first mass. Finally, this is imitated in the low register by a rap-
idly expanding mass, with the rapid registral expansions of these last two
similar masses concluding simultaneously to impart a dramatic conclusion
to the passage. The overall dramatic design of this excerpt thus involves a
dynamic, textural density, registral compass, and rate-of-change that each
decreases towards the middle of the passage and then increases rapidly
towards its conclusion. This sonic boomerang recedes, dissipating the mu-
sical intensity generated by the preceding section’s dense activity, before
returning to create a “cliffhanger ending”, which in turn is followed by an
abrupt “jump-cut” into Metastaseis’ concluding section.
While the internal structures of the four sound-masses are gov-
erned by the ruled-surfaced projections of the sketches, this is not the case

8
It also prepares the ensuing conclusion of the work by reducing the instrumen-
tation to string glissandi alone.
9
It appears that some lines in the tail of the second sound mass do not belong to
the same doubly ruled surface as its body. These lines may represent a small por-
tion of an additional surface (or surfaces), or they may have been freely added.
6 XENAKIS MATTERS

for their contrapuntal disposition with respect to one another. In other

words, while the composer elected to specify certain local musical details
(massed glissandi) according to a particular geometrical guideline, larger-
scale formal features were determined according to different principles.
More generally, if distinct hierarchical levels of formal segmentation coex-
ist in a musical passage—as is usually the case—a compositional principle
operating at one level need not necessarily operate at others.10
Figure 3 furnishes another illustration of this point. It represents a
graphical sketch for part of Xenakis’ next composition, Pithoprakta (1956),
which is scored for string orchestra, two trombones and percussion. Again
pitch and time are represented on the vertical and horizontal dimensions
respectively. Each line represents a pizzicato string glissando. The rates of
pitch-change for these glissandi are randomly distributed according to
Maxwell-Boltzmann statistics, which provide the classical model for the
distribution of molecular velocities in a gas.11 The analogy is vivid—it is
easy to imagine these glissandi as representing the trajectories of sonic mol-
ecules. On the other hand, the evolution of the glissando-cloud over the
course of the excerpt exhibits a structure that cannot be ascribed to the
mathematical model—this gas apparently passes though some elaborate
externally imposed plumbing! In turn, the ensemble rises, falls and com-
presses, bifurcates, and finally recombines into a broad, static, mid-register
sound-mass as a conclusion to its registral adventures. As in the previ-
ous example from Metastaseis, local musical details generated according
to particular mathematical guidelines are sculpted on a larger formal scale
in order to yield a brief drama in sound. In Xenakis’ own words regard-
ing Pithoprakta, “the macroscopic configuration is a plastic modulation of
the sonic material”.12 His choice of the word “material” seems significant,
suggesting that it awaits a “form” imposed through “plastic modulation”.13

10
In the context of Western music of the Classical era, successive levels of hierar-
chical formal segmentation might correspond to notes, motives, subphrases, phras-
es, periods, sections, etc. In modern music, a less loaded terminology of “temporal
gestalt levels” might be adopted; see James Tenney with Larry Polansky. “Temporal
Gestalt Perception in Music” in Journal of Music Theory, 24 (1980) 205–41.
11
The analogy between string pizzicati and gas molecules is limited, since in a
classical gas the molecules change direction when they collide with one another
or with the sides of their container, whereas in this passage from Pithoprakta the
duration of glissandi is constant within any single instrumental line (see Xenakis.
Formalized Music 15).
12
Xenakis. Formalized Music, 15.
13
In practice, specific details of the larger-scale form were presumably decided
WANNAMAKER 7

Figure 3 Graphical compositional sketch for Pithoprakta (1956), mm. 52–59

(from Xenakis, Formalized Music, 18–21; the gap in the figure corresponds in
duration to a portion of the graph that is missing from that published version).

These examples from Metastaseis and Pithoprakta underscore the

importance of imagination and aesthetic discretion as adjuncts to Xenakis’
technical procedures in his early compositional designs. While mathemati-
cal structures may have served to define certain analogous musical ones,
and while the quantitative nature of mathematics may have facilitated the
mechanical generation of many local musical details, musical imagination
(“auditive fantasy”) and judgment crucially informed the selection of par-
ticular structures and the procedures for their sonification. Furthermore,
many compositional decisions were left undetermined by the analogy be-
tween mathematical and musical structures, and were made separately by
the composer according to other criteria. These included crucial choices
regarding orchestration, register, dynamics, pacing, duration, and formal
design at various scales.

Automation and Discretion

Xenakis’ next composition following Pithoprakta was Achorripsis
(1957) for orchestra, which he described as an example of “free stochastic
music”.14 Here Xenakis extended the probabilistic approach of Pithoprakta
“to make music with fewer elements (i.e., fewer than a ‘mass’) and yet with

before those of local features (glissandi) in order that the latter could conform to
the former, but the metaphorical “plastic modulation” of an inchoate preexistent
“material” is compelling from a listener’s perspective.
14
Xenakis. Formalized Music 29–38.
8 XENAKIS MATTERS

the same method, that is, stochastically…”.15 In Achorripsis, not only was lo-
cal material (such as the disposition in pitch and time of individual attacks)
determined by mathematical methods, but these also informed the larger-
scale organization of the composition, which is represented in Figure 4.

Figure 4 Xenakis’ formal diagram for Achorripsis (1957) (adapted from Xe-
nakis, Formalized Music, 28). The numbers of “events” indicated in the legend
actually indicate event densities rather than tallies of individual events.

Achorripsis is divided into twenty-eight contiguous sections, each

fifteen seconds in duration. Each of these sections is populated with sounds
drawn from seven different instrumental sound-types (piccolo/clarinet
tones, oboe/bassoon tones, string glissandi, percussion sounds, string piz-
zicati, brass tones, and arco strings). Within each section, each sound-type
is assigned a particular density-value, which may be zero. In order to deter-
mine these density-values, collections of such values for each sound-type
were determined in advance such that these collections exhibited speci-
fied probability distributions. The assignment of these density values to
particular sections was then made in accordance with constraints derived
from probability models, although these constraints left some latitude for
the composer’s discretion. For instance, the section of overall highest den-
sity occurs near the work’s conclusion, and is immediately preceded by a

15
Varga. Conversations with Iannis Xenakis 78.
WANNAMAKER 9

nearly silent section.16 This dramatic succession is neither prescribed nor

proscribed by the composer’s stochastic procedures as he describes them
and apparently serves a climactic function.
Such allowances for composerly choice are further restricted in
Xenakis’ later ST series of pieces (1956–62).17 The constructional meth-
odology used in these pieces was largely similar to that used in Achorrip-
sis, but they were algorithmically composed with the aid of a computer.18
The succession of sections and their sonic contents were thus determined
for the most part automatically, although Xenakis would sometimes re-
assert his compositional discretion by reordering or omitting sections as
he manually transcribed the computer’s output into conventional musical
notation.19

When I used programs to produce music like ST/4, ST/10 or ST/48,

the output sometimes lacked interest. So I had to change. I reserved
that freedom for myself….My approach is similar to Le Corbusier’s,
because he pitched his sights higher than the rules.20

Certain formal constraints imposed by the algorithm nonetheless

remain discernible in the music. The internal hierarchy of formal segments
in Xenakis’ free stochastic compositions comprises (1) individual sound-
events at the most local level and (2) statistically uniform sections demar-
cated by differences in the proportions of the various sound-types. The
composer’s limited discretionary interventions may have been intended to
promote the perception of some larger-scale formal segments, but these
are less clearly differentiated from one another, so this larger-scale articu-
tion is less salient if perceptually definite at all.21 In any event, attempts at

16
As actually scored, this section contains a very low event-density rather
than “no events”.
17
Xenakis begin work on the pieces of the ST series in 1956, completing each
of them in 1962. They are ST/48,1-240162 for orchestra, ST/10, 1-080262 for
ten instruments, ST/4, 1-080262 for string quartet (arrangement of ST/10),
Amorsima-Morsima (ST/10-2) for ten instruments, Morsima-Amorsima (ST/4,
2-030762) for piano, violin, cello and double bass, and Atrées (ST/10, 3-060962)
for ensemble.
18
James Harley. “Computational Approaches to Composition of Notated Instru-
mental Music: Xenakis and the Other Pioneers” in The Oxford Handbook of Com-
puter Music (ed. Roger T. Dean) (New York: Oxford University Press, 2009) 119.
19
James Harley. Xenakis: His Life in Music (New York: Routledge, 2004) 29.
20
Varga. Conversations with Iannis Xenakis 201.
21
One possible hierarchical sectionalization for Achorripsis is proposed in Ronald
10 XENAKIS MATTERS

large-scale formal design would have contended with the nature of the pre-
defined compositional algorithm, which made no provision for segmental
hierarchies comprising more than two levels. Furthermore, the sense of
clear directed motion within formal segments, which was often evident at
multiple hierarchical levels in Metastaseis and Pithoprakta, is largely absent
in the free stochastic music.22 Instead, the content of each section is statisti-
cally uniform because each sound-type is assigned a constant density value
within that section.
Notwithstanding the crucial importance of the ST series within
the history of algorithmic composition,23 Xenakis was dissatisfied with the
musical output of the algorithm, as attested by the modifications he made
during transcription. By the time of Atrées (ST/10, 3-060962) (1962)—
the last piece in the ST series—these modifications extended to many local
musical details.24 Xenakis subsequently abandoned the algorithm, and in
ensuing works his procedures generally involved an increasing role for his
discretion with regard to compositional shaping at all formal levels.25 This
was the case even when he imported structural principles from mathemat-
ics, which remained a prevalent and evolving feature of his work through-
out the next decade.26

Squibbs. “The Composer’s Flair: Achorripsis as Music” in Proceedings of the

International Symposium Iannis Xenakis (Athens, May 2005) (Makis Solomos,
Anastasia Georgaki, Giorgos Zervos, eds.) (www.iannis-xenakis.org, 2006).
22
Occasional exceptions may occur when the duration of an individual sound-
event is significant compared to the duration of the section in which it appears,
as is the case for the lengthy descending scalar line that begins at m. 222 of
ST/10 (and m. 224 of ST/4).
23
The works in the ST series represent some of the earliest examples of com-
puter-generated algorithmic compositions; see Peter Manning. Electronic and
Computer Music, Revised & Expanded Ed., (New York: Oxford Univ. Press, 2004)
201–03.
24
Harley. Xenakis, 29–30. Xenakis’ ultimate attitude towards the ST pieces is not
entirely clear from the documentary record. The last quote above suggests that
he was satisfied given the modifications he made to the algorithm’s output. Else-
where, he expressed dissatisfaction with the series; see “Iannis Xenakis, interview
by Harry Halbreich” on Iannis Xenakis, Volume 5: La Légende d’Eer, Mode 148
DVD (1995), c. 9 min. In any event, they remain exceptional within his oeuvre on
account of their unusually high degree of systematization.
25
Xenakis’ final application of the ST algorithm appears to have been the generation
of certain materials used in Eonta (1963–64) for piano and brass quintet; see Varga.
Conversations with Iannis Xenakis 100–02.
26
It should be noted, however, that during this period Xenakis also produced a
WANNAMAKER 11

Xenakis’ Technical Procedures in Perspective

The solo piano work Evryali illustrates how far this evolution in
Xenakis’ procedures had progressed by 1973. Five basic types of sound-ma-
terial populate the composition: silences, cloud-like stochastic materials,
wave-like gestures traversing the registers of the instrument, fanfare-like
patterns of repeated pitches projecting static chords, and arborescences,
which are (predominantly conjunct) melodic lines that ramify into po-
lyphony.27 Figure 5 provides a graphical score excerpt exhibiting the last
two of these material types.

Figure 5 Graphical score for Evryali (1973), mm. 107–15, transcribed from
the published conventional score. Dots indicate individual attacks. Fanfare-type

number of significant choral dramatic works that do not specifically engage math-
ematical ideas, including Polla ta dhina (1962), Hiketides (1964), Oresteïa (1966),
and Medea Senecae (1967). This current in his work suggests that at no time did he
privilege mathematical procedures over all other compositional approaches.
27
This tally of material types agrees with Harley. Xenakis 80. Squibbs identifies
only four material types, regarding what I call wave-like gestures as stochastic-type
materials with hybrid features derived from what I have called fanfare-type mate-
rial (and which he calls “gamelan”-type material); see Ronald Squibbs. “A Method-
ological Problem and a Provisional Solution: An Analysis of Structure and Form in
Xenakis’ Evryali,” in Présences d’Iannis Xenakis (ed. Makis Solomos)(Paris: Centre
de Documentation de la Musique Contemporaine, 2001) 153–58.
12 XENAKIS MATTERS

materials appear at both the lower left and the upper right. Gray lines delineate
constituent branches within an arborescence-type material
The rhythms associated with individual pitches in the fanfare-
like materials have been analyzed by Squibbs as set-theoretic operations
carried out on residue classes.28 Xenakis refers to such structures as sieves
and applied them extensively as scalar pitch structures beginning in Nomos
Alpha of 1965–66.29 In Evryali they appear as both pitch reservoirs and—
in the fanfare-type material—rhythmic sequences.
Alongside the fanfare-type material’s mathematically derived
structures are arboresence-type materials, which are the product of
graphical sketching.

Varga What I don’t understand is how you developed the shapes them-
selves. When you wrote Evryali, for instance, did you draw the shape that
corresponded to the melodic pattern that you imagined or the other way
around?

Xenakis The drawing and thinking of the sound-image go hand-in-

hand, the two can’t be separated. It would be silly to leave out of ac-
count, when drawing, what will sound in reality. We have also to be able
to find on paper the visual equivalent of the musical idea. Any changes
and modifications can then be carried out on the drawing itself…30

The juxtaposition of fanfares and arboresences in Evryali makes

clear that Xenakis at this point in time regarded mathematical and more
“handcrafted” sources of musical material as equally valid and useful with-
in the context of a single compositional design. In fact this recalled his ap-

28
Ronald Squibbs. “An Analytical Approach to the Music of Iannis Xenakis:
Studies of Recent Works” (PhD diss., Yale University, 1996) 65–66, 147. Residue
class n modulo m is the set of integers that leave remainder n when divided by
m. For instance, residue class 1 modulo 3 is the set {…, 1, 4, 7, 10, …} because
all of these numbers leave remainder 1 when divided by 3. If integers represent
pitches in an equal-tempered scale, then a residue class (ordered by size) corre-
sponds to a sequence of pitches separated by identical pitch-intervals. If integers
represent evenly spaced points in time, then the residue class will correspond to a
regular pulse. The sets that Xenakis calls sieves are formed by subjecting such resi-
due classes to set-theoretic union, intersection and complementation operations.
For further discussion of sieves, see Gibson. Op. cit. 81-102.
29
Xenakis. Formalized Music, 194.
30
Varga. Conversations with Iannis Xenakis 90.
WANNAMAKER 13

proach in Metastaseis and Pithoprakta, wherein local details determined by

mathematical procedures served larger-scale formal designs of apparently
less-systematic origin. In Evryali, however, the outcomes of mathemati-
cal and non-mathematical procedures were now conspicuously mingled
on the same formal scale, further emphasizing that—as compositional re-
sources—they shared an equal footing. This approach contrasted with that
represented by Xenakis’ free stochastic compositions, wherein predefined
algorithms held a greater determining role at all formal scales.
The role of explicit mathematics in Xenakis’ music and writ-
ings would attenuate further in the coming years. Indeed, Xenakis
would later characterize his early work that was informed by math-
ematical ideas as follows:

All those years served as a kind of training. I can now work with
the theories intuitively—they’ve become an innate part of my think-
ing. Most of the time I don’t need rules or functions for composing.
They’re in my blood.31

It seems appropriate to conclude that—with very few excep-

tions—mathematics in Xenakis’ music generally served an overall compo-
sitional design that was not reducible to a mathematical formulation. This
fact is almost overshadowed by the prevalence of mathematically oriented
writing on Xenakis’ compositional techniques.

The products of the intelligence are so complex that it is impossible

to purify them in order to submit them totally to mathematical laws.
Industrialization is a forced purification. But you can always recog-
nize what has been made industrially and what has been made by
hand. Industrial means are clean, functional, poor. The hand adds
inner richness and charm. […]

Only one set of my works, the ST, came out of computer programs.
All the others are mostly handiwork, in the biological sense: adjust-
ments that cannot be controlled in their totality. If God existed He
would be a handyman.32

  Varga. Conversations with Iannis Xenakis 200.

31

32
Iannis Xenakis. “Xenakis on Xenakis.” Interview by Roberta Brown and John
Rahn in Perspectives of New Music 25/1 (1987) 23.
14 XENAKIS MATTERS

Xenakis’ early experiences with mathematics were as an engineer

and architect, whose work would have been understood and judged not on
the basis of his technical means but on what he made using them, and for
whom mathematical methods would ultimately have represented means.
In 1980, Xenakis-the-composer expressed a comparable outlook regarding
(mathematical) theories in the following comments, prompted by a ques-
tion about the disappearance of theoretical remarks from the prefaces to
his later scores:

In the past I developed theories and tried to compose in accordance

with them. Each theory was sound and unique.

Today I draw on them in a sporadic and sequential manner. Theories

now are dominated by the general approach, the architecture of the
composition itself.

Why no new theories? I don’t know. Perhaps because I concentrated

on constructing pieces, which should be architecturally more…I
don’t know how to put it.33

“How to put it” thus becomes the province of the critic and
analyst. The project of explaining why Xenakis deployed his materials
as he did (whatever their provenance)—or of at least better understand-
ing the particular impact of his music upon listeners—entails challenges
not encountered in the exegesis of his mathematical techniques because
it engages aspects of psychology, history and culture that (for now) resist
mathematical formulation. Nonetheless, for most listeners—even math-
ematically literate ones—such determinants will have a stronger effect on
their experience of Evryali than will the specific residue-class combinations
employed in its composition.34 These factors include the vivid distinctions

33
Varga. Conversations with Iannis Xenakis 199.
34
This is not to suggest that the study of particular residue-class combinations
has no place in a listener-oriented analysis of Evryali, since they engender par-
ticular scalar and rhythmic qualities. (Be that as it may, it is not clear that such
combinations are the best representation of Evryali’s scalar and rhythmic materi-
als for the purpose of elucidating their perceptible characteristics.) In any event,
as illustrated by the above explorations of other mathematically derived musical
configurations, such structures in Xenakis’ work only attain a definite musical
significance within a context that includes many other factors, which may be as
important or more so.
WANNAMAKER 15

between material-types, the obstinate oppositions into which they enter,

their occasional syntheses, the weighing of durations and contrasts, the
suspicion of a narrative subtext involving conflict, and the athletic virtuos-
ity demanded in performance.35 Consideration of Xenakis’ mathematical
techniques may sometimes inform the analysis of such psychologically sa-
lient features of his work, but it should not displace such analysis.
While it is thus important to maintain perspective upon the role
of technical procedures in the works of Xenakis (or of any composer), this
is not to deny the powerful historical and musical repercussions of his radi-
cal adoption of mathematical methods in his early compositions.

That’s my contribution to the development of music: I use ideas in

composing that are completely alien to music.36

While one may reasonably disagree that Xenakis’ contributions

reside solely within this narrow scope, his applications of mathematics
clearly represent a significant expansion of the range of possible mod-
els for musical structure beyond traditional ones. They also provide an
influential precedent for ongoing compositional explorations that are
informed by ideas drawn from non-musical disciplines. In order to com-
prehend the roles and impact of such imported concepts in real compo-
sitional instances, however, they must be understood as one contribution
to a holistic musical design.

35
One possible narrative reading based on the Greek myth of Perseus is offered
in Linda M. Arsenault. “Iannis Xenakis’ Evryali: A Narrative Interpretation,” in
Présences de Iannis Xenakis op.cit. 179–62. Regarding the challenges of Evryali
for performers, see Marie-Françoise Buquet. “On Evryali,” in Performing Xenakis
(ed. Sharon Kanach) (Hillsdale: Pendragon, 2010) 65–70.
36
Varga. Conversations with Iannis Xenakis 79.

View publication stats