Wolfgang von Schweinitz

Iannis Xenakis’ Contributions

to 20th Century Atonal Harmony
and to the History of Equal Temperament
The Aristoxenean Sieve Theory, and the Ptolemaic Alternative
Some Thoughts on Our Perception of Musical Intervals 1

In his influential orchestral piece Metastaseis (1953-54), Iannis Xenakis

introduced a number of great new sonic entities, exciting innovations
that radically expanded the vocabulary of music. I would like to mention
just two of them here: complex glissando textures (with up to 46 different
simultaneous glissando speeds) and dense chromatic or micro-chromatic
tone clusters. These are major contributions to 20th century atonal
harmony, because they serve to neutralize our ear’s refined capability of
perceiving harmonic relations by creating the impression of noise (which is
generated by an accumulation of masses of tempered pitched sounds).

In the very next piece, Pithoprakta (1955-56), Iannis Xenakis organized

his sound masses by introducing the notion of probability, and here the
basic musical parameter of pitch “is important only in the global sense of
conveying registral boundaries (e.g. high versus low register, wide versus
narrow range, fixed versus evolving placement).”2 Here the pitches are no
longer composed according to any kind of either tonal or atonal harmonic
counterpoint; instead they are determined by a probability algorithm
defined at a higher structural level.

I think that György Ligeti was very impressed and influenced by Xenakis’
new polyphonic musical textures. In 1962 he presented and analyzed
Metastaseis and Pithoprakta in his lecture at the International Summer
Courses for New Music in Darmstadt. When I studied with him about
twelve years later, however, he criticized Xenakis for “not having a sense
of harmony”, referring to these pieces. Like Witold Lutoslawski, Luigi
Nono and others, György Ligeti was interested in developing and
reintroducing some harmonic principles to overcome the “impasse of the

1
This essay is a free reconstruction of a talk on January 30, 2011 at the CEIAT festival
‘ Iannis Xenakis in Los Angeles’ in the Museum of Contemporary Art.

2
James Harley. Xenakis : His Life in Music (New York and London: Routledge, 2004) 14.

1
chromatic scale”, as Xenakis called it;3 and I believe that Ligeti wasn’t yet
aware of the fact that Xenakis had also been working on solving the same
problem – in his own unique and profound way – at the same time since
the mid-1960s. Xenakis began by first studying Aristoxenos’ Elementa
Harmonica to analyze the structure of Ancient Greek Music, and from
this basis he then devised his “Sieve Theory”,4 which became a powerful
compositional tool in the creation of much of his music since the mid-
1960s. I regard this invention as another substantial contribution to the
methods of 20th century atonal harmony.

It would be exciting to analyze Xenakis’ scores in order to assemble a

comprehensive compilation of all the various pitch sieves (or scales) that
he used in his music – the remarkable counterpoint of the postlude in
Nomos Alpha, the bright and diatonic introduction of Jonchaeis, or the
stunning fireworks of Ata. In this essay, I will discuss Xenakis’ Sieve
Theory in a more general way, focusing on the historical relevance of his
invention. I believe that it was a major step, perhaps even the final step
towards what I’d like to call the historical fulfillment of our Twelve-Tone-
System of Equal Temperament.

This can perhaps be explained by a quick glance at the function and the
history of this tuning system. In Western music, Equal Temperament
was first devised as a standard tuning system for fretted instruments
only. The lutenist and composer, Vincenzo Galilei (circa 1520-1591), the
father of the astronomer Galileo Galilei, calculated and defined the size of
the tempered semitone as a practical approximation to the string length
ratio 18:17 (circa 99 cents), which was “convenient for fretting the lute”.5
But there was a general consensus amongst musicians during the 16th
and 17th century that this rather dissonant temperament with its harsh
major thirds and sixths is inferior to Meantone Temperament and not an
appropriate and desirable tuning system for keyboard instruments like
the organ and harpsichord.6 As the creative urge to also explore the more
distant keys had become increasingly irresistible, experimental keyboard
instruments with split upper keys on multiple registers were built in Italy
to complete a circular modulating Meantone System (with as many as 31

3
Bálint András Varga. Conversations with Iannis Xenakis (London: Faber and Faber,
1996) 93, 94, 96.
4
Iannis Xenakis. Formalized Music: Thought and Mathematics in Composition (Hillsdale,
NY: Pendragon Press, 1992), Chapter VII: Towards a Metamusic.
5
J. M. Barbour. Tuning and Temperament: A Historical Survey (East Lansing: Michigan
State College Press, 1951; Mineola, N.Y.: Dover Publications, 2004) 8, 57; source text:
Vicenzo Galilei. Dialogo della musica antica et della moderna (Florence, 1574).
6
This is clearly evident from history itself. For what other reasons could there have
been for restricting the use of Equal Temperament to the fretted lutes and guitars?
See also: Mark Lindley. Temperaments, New Grove Dictionary of Music and Musicians
(London: Macmillan Publishers Limited, 1980) volume 18, 664-666.

2
tones in the octave), in which the tempered wholetone (i.e. the meantone)
was divided into five ‘dieses’. But the technical difficulties of building,
maintaining, tuning, and playing such fantastic instruments kept them
as exceptions, and they could not enter into common practice. By the
end of the 17th century, musicians therefore began to experiment with
many different modifications of the traditional Meantone Temperament,
and these irregular keyboard temperaments with all their expressive
microtonal diversity became highly popular. Johann Sebastian Bach’s
own favorite irregular keyboard temperament, which has recently been
discovered (right on the top of Bach’s handwritten title page of The Well-
Tempered Clavier) by Bradley Lehman,7 was, as it seems, the very most
subtle and refined attempt to represent all twenty-four major and minor
keys on a standard keyboard with only twelve pitches within the octave,
while also maintaining a complete compatibility with the tone system of
Extended Meantone Temperament, which was common practice for non-
keyboard instruments throughout the 18th century.

Like his contemporaries, Wolfgang Amadeus Mozart thought in terms of

Extended Meantone Temperament until the end of his life, differentiating
between the small chromatic semitone (i.e. the difference between a major
and a minor third, like d-d# ) and the large diatonic semitone (i.e. the
difference between a fourth and a major third, like d-eb ), and thus
between at least nineteen different pitches per octave: c-c#-db-d-d#-eb-e-
e#-f-f#-gb-g-g#-ab-a-a#-bb-b-b#-c.8

When Jean-Philippe Rameau changed his opinion and started advocating

Equal Temperament in 1737,9 he triggered a vivid discussion about its
pros and cons in France, but the Broadwood piano company in London
did not adopt Equal Temperament until 1846.10 Once the discussions
about the sound of this keyboard temperament had finally subsided, it
did not take long before it was in fact regarded as the “new modern tone
system”. Given the immense popularity of the piano and its central role
in our music education, even the theory of tonal harmony was eventually
understood in terms of Equal Temperament. The acoustical explanations
and critical comments published in 1863 by the great physicist Herman
von Helmholtz in his pioneering book On the Sensations of Tone as a

7
Bradley Lehman. “Bach’s extraordinary temperament: our Rosetta Stone” in Early
Music, volume XXXIII No. 1+2 (Oxford University Press, Feb. and May 2005).
8
See the papers of Mozart’s pupil Thomas Attwood, published as part of the new Mozart
Edition, series X/30 (Kassel: Bärenreiter, 1969).
9
Jean-Philippe Rameau. Génération Harmonique (Paris, 1737); English translation:
Deborah Hayes. Rameau’s “Theory of Harmonic Generation”: An Annotated Translation
and Commentary of “Génération Harmonique” by Jean-Philippe Rameau (Stanford, 1968).
10
Ross W. Duffin. How Equal Temperament Ruined Harmony (and Why You Should Care)
(New York and London: W.W.Norton & Company, 2007) 106.

3
Physiological Basis for the Theory of Music 11 did not have an immediate
impact on the general school of thought amongst musicians.

But the major German conductor, Hans von Bülow (1830-1894), a close
friend of Wagner, Liszt and Brahms, used to call this new tone system of
Equal Temperament “the piano lie” 12 all his life. And in 1911 the eminent
thinker, Arnold Schoenberg, commenced his forward-looking Theory of
Harmony with a discussion of the harmonic series of the overtones. In this
context he called the standard tone system of Equal Temperament “a
compromise between the natural intervals and our inability to use them”
and “a truce made for an indefinite period of time”.13 Only ten years later,
he invented his dodecaphonic method for an atonal pitch organization
that is conceptually based on the specific virtue of the equal-tempered
twelve-tone system, i.e. its unrestricted “transposability”.14 The atonal
twelve-tone music, as well as the serial music of the 1950s, was literally
composed for this tone system, and it may therefore be regarded as
commencing its historical fulfillment. For the first time in its long history,
Equal Temperament was not a compromise but the perfectly adequate
tuning system for this new kind of music.

Another conceptual reference to the tempered twelve-tone system with its

equal semitones was achieved around the same time during these early
1920s in the atonal music of Edgard Varèse, which fully exploited the
dissonant sonority of stacked equally-sized semitones, – and later in the
“cluster music” developed during the late 1950s and early 1960s by
Iannis Xenakis, Luigi Nono, György Ligeti, Krzysztof Penderecki, and
others.15 Yet another remarkable summit in the history of Equal
Temperament was established by its microtonal extension in the work of
Alois Haba and Julian Carillo, and by Ivan Wyschnegradsky, who devised

11
Herman von Helmholtz. Die Lehre von den Tonempfindungen als physiologische
Grundlage für die Theorie der Musik (Braunschweig, 1863); English translation by
Alexander Ellis. On the Sensations of Tone as a Physiological Basis for the Theory of
Music, translated from the edition of 1877 (New York: Dover Publications, 1954).
12
Anonymous. Testimonials, Etc. (Translations) Relating to the “Enharmonium,” Invented
by Shohé Tanaka. London, 1891 (The Virtual Laboratory, Max-Planck-Institute for the
History of Science, Berlin, ISSN 1866-4784 - http://vlp.mpiwg-berlin.mpg.de).
13
Arnold Schönberg. Harmonielehre (Vienna, Universal Edition, 2nd edition, 1922);
English translation by Roy E. Carter (Berkeley: University of California Press, 1978).
14
In the tone system of Equal Temperament, any pitch collection (e.g. motif, chord, or
twelve-tone-row) can be precisely, i.e. literally transposed. This is not generally possible
in other tuning systems (e.g. Just Intonation, Meantone Temperaments, or irregular
keyboard temperaments, like that of Francesco Antonio Vallotti, for example).
15
For example, Iannis Xenakis: Metastaseis, Pithoprakta, and often again much later,
e.g. in Ata; Luigi Nono: Canti di vita e d’amore ; György Ligeti: Apparitions, Atmosphères,
Requiem ; Krzysztof Penderecki: Anaklasis, Threnody to the Victims of Hiroshima,
Fluorescences.

4
“ultra-chromatic scales,” as he called them (e.g. various sequences of
equal-tempered quartertones and semitones),16 and thus employed an
equivalent of the Sieve Theory before Xenakis had even formulated it.

To end my summary of the historical context of Iannis Xenakis’ Sieve

Theory, I would like to mention two unique pieces by James Tenney from
1985.17 Written for pedal harps and guitars, i.e. for those non-keyboard
instruments that are most intimately tied up with the tuning system of
Equal Temperament, these two compositions are profound investigations
into the harmonic potential of the tempered “72-set”, as James Tenney
called it, which divides the octave into 72 discrete equidistant pitches.
Using the 72 pitches of this “twelfth-tone sieve” with the aim of better
approximating as many just intervals as possible, he was already working
with a very different mind set than Iannis Xenakis – looking at Equal
Temperament; not any more from within, but from above or beyond, so to
speak, as a protagonist of next generation tonal thought, once again
conscious of harmonic relations between the musical tones. This is an
important distinction, and it shall be developed as we proceed.

Iannis Xenakis was fully aware of the conceptual bond between his Sieve
Theory and the tone system of Equal Temperament, which was generally
taken for granted in its heyday fifty years ago – in fact, just before the
much more relaxed sound of Meantone Temperament was rediscovered.
In his article Towards a Metamusic, in which he first presented his Sieve
Theory in 1967, Iannis Xenakis says:

[…] the tempered diatonic system—our musical terra firma on which all
our music is founded—seems not to have been breached either by
reflection or by music itself. This is where the next stage will come. The
exploration and transformations of this system will herald a new and
immensely promising era.18

We can sense the energy of his optimistic belief in the remaining musical
potential of the tempered tone system that was yet to be exploited. There
is no trace any more of the profound, fundamental skepticism Arnold
Schoenberg had articulated about half a century earlier, before he
proceeded on the basis of his courageous decision to fully embrace the

16
Charles Amirkhanian. Interview with Ivan Wyschnegradsky, recorded 1976 (radiom,
otherminds.org) http://radiom.org/detail.php?et=interview&omid=AM.1976.06.04.
17
James Tenney. Changes: Sixty-four Studies for Six Harps (1985) and:
Water on the Mountain... Fire in Heaven for 6 electric guitars (1985); see also:
James Tenney. “About Changes: Sixty-four Studies for Six Harps” (Perspectives of New
Music 25, 1987).
18
Xenakis. Op. cit. 182.

5
tempered system anyhow, by composing the dodecaphonic atonal music
it asked for.

In order to introduce his Sieve Theory, Iannis Xenakis first proposes in

Towards a Metamusic the valuable new categories called “outside-time”,
“in-time”, and “temporal”, and then gives a comprehensive account of the
outside-time structure of Hellenic and Byzantine music. He follows the
ancient Greek source texts (mainly the brilliant above-mentioned treatise
Elementa Harmonica by Aristoxenos), albeit with a radically formalized
modern terminology, which is entirely based on Equal Temperament by
using the twelfth-tones of the 72-set (which Xenakis calls “Aristoxenean
segments” ) as a general unit of measurement.

In fact, Aristoxenos briefly touches upon this twelfth-tone in his treatise

in order to define the small difference between the third-tone (‘chromatic
diesis’) and the quarter-tone (‘enharmonic diesis’). Aristoxenos discusses
the magnitudes of the intervals between the four notes of the tetrachord
in the diatonic, chromatic and enharmonic genera or modes (with their
various microtonal “shadings”) by specifying the possible pitch ranges for
the two movable middle notes (lichanos and parhypate) of the tetrachord
meson; and he immediately adds the remark:

For it must be understood that the lichanoi are unlimited in number.

Wherever you arrest the voice in the range of the lichanos will be a
lichanos, and no place in the lichanos range is empty or incapable of
receiving a lichanos.19

This means that the melodic function of this second-highest note of the
tetrachord still remains the same within the genus (or mode) when its
intonation is slightly altered within its permissible pitch range. And
somewhat later, in Book II, he explains:

Each of the genera moves with what perception apprehends as its own
characteristic movement while using not just one division of the
tetrachord, but many. Thus it is clear that the genus can remain
constant while the magnitudes change, since up to a certain point the
genus does not change when the magnitudes do, but remains the same:
and while the genus remains constant it is reasonable to suppose that
the functions [dynameis ] of the notes do too.20

We can imagine that the intonation of the two movable middle notes of
the tetrachord was indeed performed with much greater flexibility than

19
Quoted from Greek Musical Writings, Volume II, Harmonic and Acoustic Theory, edited
by Andrew Barker (Cambridge, New York, Melbourne: Cambridge University Press,
1989) 144.
20
Ibid. 162-163.

6
the schematic representation of the note positions in Xenakis’ summary
suggests – whereas the pitches of the fixed outer notes were certainly not
tempered at all, but tuned in perfect fourths, fifths, unisons and octaves.
Regarding the intonation of consonances and dissonances, Aristoxenos
says near the end of Book II:

Of the magnitudes of intervals, those of the concords appear to have

either no range of variation [topos ] at all, being determined to a single
magnitude, or else a range which is quite indiscernible, whereas those of
the discords possess this quality to a much smaller degree. Hence
perception relies much more confidently on the magnitudes of concords
than on those of discords.21

Aristoxenos was the son of a musician and a disciple of Aristotle, and in

his revolutionary endeavor to write up the first full-fledged account of the
“science concerned with melody” he avoided even the slightest reference
to the impressive knowledge about the physics and the mathematics of
sound accumulated by the previous generations of Pythagoreans. They
had interpreted sound as the result of an impact and as movement of the
air, studied the vibrations of strings at different speeds and analyzed the
magical mathematical ratios between various string lengths and their
associated pitches. Rejecting their theories and methods, especially their
precise “language of the ratios” (as Harry Partch has called this powerful
Pythagorean conceptual tool),22 as irrelevant or misleading “extraneous
territory” for the practicing musicians, Aristoxenos introduced a new and
very different set of practical musical concepts (some of which he must
have adopted from the musicians who had taught him). By discussing
musical intervals in terms of their sizes, i.e. by focusing the attention of
the ear and the mind on the melodic distances between the notes, and
not on their harmonic relationships like the Pythagoreans, Aristoxenos
inaugurated a second school of thought amongst musicians. They were
henceforth (and even to this day!) divided into these two opposing groups
called the Pythagoreans and the Aristoxeneans.

With his fundamental dedication to Equal Temperament, Iannis Xenakis

was of course a genuine Aristoxenean. Let us look at his opinion on this
important topic of the two languages:

Attention must be drawn to the fact that he [Aristoxenos] makes use of

the additive operation for the intervals, thus foreshadowing logarithms
before their time; this contrasts with the practice of the Pythagoreans,
who used the geometrical (exponential) language, which is multiplicative.
Here the method of Aristoxenos is fundamental since: 1. it constitutes

21
Ibid. 168.
22
Harry Partch. Genesis of a Music (Madison: University of Wisconsin Press, 1949).

7
 one of the two ways in which musical theory has been expressed over
millennia; 2. by using addition it institutes a means of “calculation” that
is more economical, simpler, and better suited to music; and 3. it lays
the foundation of the tempered scale nearly twenty centuries before it
was applied in Western Europe.

Over the centuries the two languages – arithmetic (operating by addition)

and geometric (derived from the ratios of string lengths, and operating by
multiplication) – have always intermingled and interpenetrated so as to
create much useless confusion in the reckoning of intervals and
consonances, and consequently in theories. In fact they are both
expressions of group structure, having two non-identical operations;
thus they have a formal equivalence.23

I think that Xenakis’ reasons for preferring the Aristoxenean method are
not that strong. This argument of “greatest convenience” has always been
used by the advocates of Equal Temperament, but I believe that it does
not suffice to secure the exclusive predominance of this tuning system in
the long run. “This reduction of the natural relations to manageable ones
cannot permanently impede the evolution of music; and the ear will have
to attack the problems, because it is so disposed.” [original spacing
maintained] says Arnold Schoenberg in 1911 in his Theory of Harmony,24
and he follows this thought all the way, in his radical and visionary
mind, saying:

Then our scale will be transformed into a higher order, as the church
modes were transformed into major and minor modes. Whether there will
then be quarter tones, eighth, third, or (as Busoni thinks) sixth tones, or
whether we will move directly to a 53-tone scale … we cannot foretell.
Perhaps this new division of the octave will even be untempered and will
not have much left over in common with our scale.

If we have an interest in developing new ways of dealing with the musical

parameters of pitch, timbre and tonal relationships, we must sharpen
our tools by combining the specific virtues of both languages now,
completely avoiding the least bit of confusion. Let us reconsider the old
Pythagorean method of reckoning intervals by their frequency ratios,
which precisely specify not only their melodic magnitude, but at the
same time, their harmonic complexity and their individual timbre which
is characterized by their partial unisons, combination tones and periodic
signatures. All of this useful data is represented by the two numbers
making up each ratio. Perhaps we may indeed appreciate these resonant
sounds and the wealth of subtle pitch inflections that will come about, if
we fully embrace the phenomena of psychoacoustics and give up the

23
Xenakis. Op. cit. 185.
24
Schoenberg. Op. cit. (German Edition: 22, English translation: 25).

8
habit of tempering our intervals – by adopting the basic principle of
microtonal just intonation. For what else could possibly have the
substance and power to be capable of effectively overthrowing the reign of
Equal Temperament? An exciting and alarming musical enrichment has
been achieved during the past few decades by creative efforts to reveal
the power and beauty of “noise music”. The sensation of noise arouses in
us the somewhat frustrating awareness that our remarkable capability of
harmonic perception can never be engaged by these sounds at all. But the
ear and the brain have a natural, innate desire to perceive and process
the pitches of musical tones, i.e. sounds with harmonic timbres.

As a modern Pythagorean acoustician, James Tenney has revitalized the

language of the frequency ratios in the grand tradition of Helmholtz and
Partch, all within the creative light of cutting-edge psychoacoustics. He
supplied us with a universal conceptual model of our brain’s harmonic
perception, which he called “harmonic space”. 25 I believe that this infinite
multi-dimensional lattice representing all the non-tempered pitches is
the new concept that can succeed our simplified model of the circle of
twelve equally tempered fifths. All the musical tones in James Tenney’s
matrix are precisely arranged according to their mutual harmonic
relationships, according to the order suggested by the unisons of their
partials. There is also a pitch-height projection axis, which represents the
basilar membrane of the inner ear and runs at an angle of 45º to every
axis right through James Tenney’s harmonic space; and upon this
glissando continuum each non-tempered pitch, or pitch-point in harmonic
space, may specify its precise frequency position.

Thus any number of microtonal scales or pitch collections can be filtered

out of harmonic space. The intervals between adjacent notes will most
likely have many different magnitudes. Moreover, it will usually not be
the case that they are composite multiples of some general measurement
unit. Like Harry Partch’s 43-note scale, these harmonic pitch collections
display a subtle microtonal melodic order that resembles the irregular
patterns of colored lines in the filtered spectrum of light – and not those
on the ruler.26

Let us return to Aristoxenos. His most influential new concept was that
of the semitone, which must certainly have caused quite a stir among the
Pythagoreans. For they had divided the octave (2:1) into a fourth and a
fifth (4:3:2), which could then be subdivided into a minor and a major

25
James Tenney: John Cage and the Theory of Harmony, first published in
SOUNDINGS 13: The Music of James Tenney (Santa Fe: Soundings Press, 1984);
also available at www.plainsound.org.
26
E.g. the uniform intervals of 100 cents in Equal Temperament, or of some other “unit
of displacement” in a Xenakian pitch sieve.

9
third (6:5:4), and so on; so they knew both from experience and reason
that none of such tuned intervals with superparticular ratios can ever be
harmonically divided into two or several equal parts. A real semitone can
only exist in Equal Temperament, and we have seen that Aristoxenos
understood his new technical term as a handy approximation. Facing the
severe criticism of the Pythagoreans, he insisted on the assumption that
the fourth is made up of two and a half tones – despite the well-known
fact that the difference between the fourth (4:3) and the ditone (81:64),
the so-called leimma (256:243), is somewhat smaller than half a tone.

Yes indeed, we can distinguish many different semitones; each of them

has its own particular size – and its own particular harmonic function,
identity or meaning. Besides the Pythagorean leimma (which has a size of
90 cents and is constructed out of five consecutive fourths or fifths), we
have already encountered Mozart’s mezzo tuono grande, the large dia-
tonic semitone (16:15 = 112 cents), and Mozart’s mezzo tuono piccolo, the
small chromatic semitone (25:24 = 71 cents); but there are in fact more
than twelve different semitones between the largest (15:14 = 119 cents)
and the smallest (28:27 = 63 cents), and they may all be called a
semitone. They can all be performed in non-tempered just intonation –
within the appropriate musical setting that is informed by the mathe-
matical know-how of Pythagorean acoustics.

To conclude, I’d like to quote a few words from the other great Hellenistic
sage: Claudius Ptolemy (90-168 AD), whom Iannis Xenakis did not hold
in high esteem.27 This devoted Alexandrian mathematician, astronomer,
geographer and music theorist must have had an extraordinary ear and
musical feeling. In his enlightened treatise on music, which he called
Harmonics, he says:

For in everything it is the proper task of the theoretical scientist to show

that the works of nature are crafted with reason and with an orderly
cause, and that nothing is produced by nature at random or just
anyhow, especially in its most beautiful constructions, the kinds that
belong to the more rational of the senses, sight and hearing. To this aim
some people seem to have given no thought at all, devoting themselves to
nothing but the use of manual techniques and the unadorned and
irrational exercise of perception, while others have approached the
objective too theoretically. These are, in particular, the Pythagoreans and
the Aristoxeneans, and both are wrong. For the Pythagoreans did not
follow the impressions of hearing even in those things where it is
necessary for everyone to do so, and to the differences between sounds
they attached ratios that were often inappropriate to the phenomena, so
that they provided a slander to be directed at this sort of criterion by
those whose opinions differed. The Aristoxeneans, by contrast, gave most

27
Xenakis. Op. cit. 185.

10
 weight to things grasped by perception, and misused reason as if it were
incidental to the route, contrary both to reason itself and to the
perceptual evidence [to phainomenon ] – contrary to reason [logos] in that
it is not the distinguishing features of sounds that they fit the numbers,
that is, the images of the ratios [logoi], but to the intervals between them
[…]28

Iannis Xenakis’ Sieve Theory serves to arrange the distances between the
notes. Thus this method of composition is fundamentally related to that
of the Serial Music, which Xenakis never liked. Both concepts establish
an atonal pitch organization for the tone system of Equal Temperament,
where the musical tones are not tuned according to their inner structure
and according to their mutual tonal relationship, but in such a way that
they are equally far apart from each other.

A few pages later, Ptolemy says:

This shows, then, that we should not find fault with the Pythagoreans in
the matter of the discovery of the ratios in the concords, for here they are
right, but in that of the investigation of their causes, which has led them
astray from the objective; but we should find fault with the Aristoxeneans
since they neither accepted these ratios as clearly established, nor, if
they really lacked confidence in them, did they seek more satisfactory
ones – assuming that they were genuinely committed to the theoretical
study of music. For they must necessarily agree that such experiences
come to the hearing from a relation that the notes have to one another,
and further that where the impressions are the same, the differences are
determinate and the same.29

Ptolemy explains that the particular sensation is created by the specific

timbre of an interval between two musical tones, by the harmonic quality
of its tone relationship, i.e. by its overall compound sound which is made
up of all the partials and combination tones. That is what generates the
identity and thus allows the ear to understand the interval, even when it
is being transposed. For a transposed interval vibrates at different
frequencies and at another virtual fundamental, but the pattern of the
interferences between these frequencies is recognizably the same. He
continues:

Yet in what relation, for each species [of concord], the two notes that
make it stand, they neither say nor enquire, but as if the notes
themselves were bodiless and what lies between them were bodies, they
compare only the intervals [diastaseis] belonging to the species, so as to
appear to be doing something with number and reason. But the truth is

28
Quoted from Greek Musical Writings, Op. cit. 278-279.
29
Ibid. 293-294.

11
 precisely the opposite. For in the first place they do not define in this way
what each of the species is in itself (as when people ask what a tone is,
we say that it is the difference between two notes that comprise an
epogdoic ratio [i.e. 9:8]); instead, there is an immediate shift to yet
another undefined item, as when they say that the tone is the difference
between the fourth and the fifth […] And if we enquire after the
magnitude of the difference in question, they do not explain even this
without reference to another, but would only say, perhaps, that it is two
of those of which the fourth is five, and that this again is five of those of
which the octave is twelve, and similarly for the rest, until they come
back round to saying ‘…of which the tone is two’.30

Ptolemy’s last remark sounds rather like a parody of our now historical,
old-fashioned 20th century tempered thought, as displayed for example in
the formalized Pitch-class Set Theory, which does not even have a proper
name for the octave and calls a fourth the interval number “5”, the major
third “4”, the tritone “6” – and so on. This terminology is of course
perfectly adequate for the intervals of Equal Temperament, in which the
tempered ditone, for example, is defined by the 12th root of the octave
(2:1) to the power of 4. And also Iannis Xenakis’ idea of a general “unit of
elementary displacement,”31 upon which his Sieve Theory is based, is of
course a completely appropriate compositional tool for establishing a
conceptual order within the set of equidistant notes available in any kind
of Equal Temperament.

However, as Ptolemy says, the ear does not only perceive the magnitude
of an interval between two musical tones (i.e. their pitch distance), but it
also recognizes the timbre of their compound sound (i.e. their harmonic
distance, as James Tenney called it).32 This psychoacoustic phenomenon
is not taken into account on the basic conceptual level in either of these
compositional methods. Nevertheless Arnold Schoenberg’s Twelve-Tone-
Technique and Iannis Xenakis’ Sieve Theory have both come about out of
historical necessity, for it is the very principle of Equal Temperament
itself that is in fact subjecting the musical tones to this kind of mutual
alienation.33

30
Ibid.
31
Xenakis. Op cit. 194-200.
32
Tenney. Op. cit.
33
I thank Marc Sabat and Georgi Dimitrov for their invaluable help.

12