XAS by Iannis Xenakis

for saxophone quartet


This piece by Iannis Xenakis, commissioned by the Rascher Saxophone Quartet was composed in 1987. It is written for the traditional quartet instrumentation of soprano, alto, tenor, and baritone saxophones.

The title is an anagram of SAX: S (the single unbalanced letter of the title) is written backwards, in order to create a mirror image. One also might note a play on the name of XenAkiS (i.e. highlighting the outermost letters and the center of his name). since Xenakis never wrote for sax before, giving part of his name and reversing the inventor's name in the title is a kind of appropriation of the saxophone for his own use, for the use of his own music.

This is also the only piece Xenakis wrote for the saxophone, since his concerto project was never composed.

When listening, the work exudes a sense of power mixed with rudeness, despite a small instrumental ensemble. Note the main use of fff dynamic, which probably contributes greatly to this particular perception. The formal clarity is particularly noteworthy and the general architecture is simple enough (see the summary table realized below). The total theoretical duration of the piece is 8'32 ", but Xenakis states that tempi must evolve according to the acoustics of the space, and offers a typical duration of 9'00".

The composer only briefly uses new techniques: some quarter tones and simple multiphonics (almost clarinets' broken sounds) punctuate the piece; on the contrary, most "attack modes" and tone colors (aside from extreme dynamics) are absent, evaded, or left to the performer's choice. However, we note that he adopts the principle of a very extended high tessitura (a minor sixth on soprano, a minor ninth on alto and tenor, and up to an eleventh on baritone above the theoretical limits of the respective instruments). On the contrary, the low C on the baritone is omitted. It seems that the composer followed the wish of the dedicatees, specialized in the high register, but the old instruments the Rascher Quartet plays (dating from the first half of the century) do not have this extension, which is now common on modern instruments.

But the exploration of the unheard by Xenakis is not only based on the development of particular techniques (the effect of which would soon be outdated), but also in an unconventional use of already known principles. Seeking to identify, through the study of this piece, the practical implications of Xenakis' work on the notions of order and disorder, I had to define a rating scale of the degree of order from O = -5 to +5, essentially based on psycho-acoustic sensations. This scale is based on several concepts:

a) Degrees of "mathematical" complexity of the parameters, for example :

     - Melody : from joint to disjoint

     - Harmony: distribution of aggregates from narrow and regular to wide and irregular

     - Rhythm: from regular flow to chaotic

     - Dynamics : from stable to moving

  1. -"Orchestration": Regular / in opposition / every man for himself

b) Degrees of auditory perception

-to identify the prevalence of a given parameter in relation to another, in the case of opposite treatments (e.g. a joint melody with dynamic movement)

-to determine the relative degree of instability of each parameter with respect to others in the juxtapositions (e.g. opposite and very disjointed pitches in sixteenth notes seems less ordered than a line of semi-joint thirty-second notes).

The writing of this work came after the recording I made as part of Xasax, it goes without saying that I consider this study as the analytical application of the work of the performer, a view of the piece illuminated from the inside out.

The measure numbers indicated refer to the published Salabert score, time information (t) to the recording made by Xasax, modular saxophone ensemble (EROL Records CD 7019, track 7).

The Material :

As in most pieces of his same composing period, Xenakis uses two "out-of-time sieves" as base material: scales A and B, which do not repeat at the octave. Both scales are complementary with the exception of a single note: F# 2.

A and B scales are present throughout the whole of the work, and are only temporarily "forgotten" during the introduction, coda, and central random passages, and are explored in several ways in the piece (here classified by degree of order).

Sieves A and B appear first in the form of parallel and joint scales (e.g. introduction and bars 50-53, t = 4 '10 ", O = +5.), which is similar perceptually to the highest degree of order in the score. It is interesting to note that Xenakis chose the center point (bar 50 of this 100-bars piece) to adopt this particularly stable process, while the previous passage was an example of a totally random cloud (i.e. the lowest level of order, or O = -5).

A comparable process is the use of very tight canons (at the sixteenth note and in unison): first appearing in the piece in bar 10 (t = 0 ' 53 ", O = 4.) The result is quite similar to the joint scales, but since it is obtained by a shift of instrumental entrances, the impression of verticality is blurred at the beginning of every phrase and on each held note. Here, the estimated order degree is +4 (or even +3 when the rhythmic values ​​are unequal, as in bar 25.) Both sieves are also the basis of "calls" (i.e. order: bar 9,  t = 0 '47 " O = 2), pointed, iterative, usually high-pitched homorythmic pointed rythms .

. The same type of rhythmic writing technique is used in the figures of the "stuttering" solo alto saxophone (bar 22). Aggregates, built on the rhythmic patterns of "Calls," generally consist of four joint notes from the same sieve (e.g. bar 9). We find the same principle of vertical construction, but sometimes with a mixture of sieves A and B to form complex "chords." In these passages Xenakis desired to hear an ‘Organ’-like sound. The four instruments must then play in a homogeneous and continuous way despite the great leaps of range and respiratory needs of the performers (e.g. bars 73-81, t = 6 '08 ", O = -1).

In this example, the long held notes follow a rhythmically agitated and complex section, creating a sense of calm, despite extremely loud dynamics (fff).

In this type of playing, compact chords are commonly used in opposite registers (bass/treble in bar 31, t = 2'37 ", O = -2 and +2).

The construction of aggregates is particularly interesting here, since it opposes, on equal note values, regular and ordered "chords" in the treble register.

These aggregates of consecutive even or odd degrees of sieve B (which would sound like major seventh chords) are against the more complex superpositions of the two sieves (BABA) in the low register, and with opposing dynamics (fff against mf).

In the two examples above, I analyzed the distribution of notes of the two sieves (A in bold, B regular)  in bars 31/32 and 73/81: one may refer to this analysis for a better understanding of these "harmonies" of Xenakis.

With "independent lines" (e.g. bars 19-21, t = 1 '37 ", +1 to -1), short agitated 32nd note values ​​are emphasized by the random movement of the four instruments, apparently lacking any relation. In the example, Xenakis nevertheless follows a general strategy that brings the saxophones from an extreme high to extreme low register: individual lines are driven by random processes, creating a striking mass effect, increased by vehement dynamic and punchy staccato. The degree of disorder is created by the simultaneous use of both sieves, the instruments grouped in pairs (A: soprano and tenor, B: alto and baritone).

This writing sometimes leads to a form of "Ataxia" (e.g. bars 54-58, t = 4 '31 ", O = -4), a principle of controlled delay between different voices by the use of highly complex rhythmic layering. Xenakis manages to create an extreme illusion of disorder: the next step in this direction is the realization of totally random events. The advantage of this method lies in the possibility of brief rhythmic (and sometimes melodic) stabilization (mm. end 57-early 58). Bars 66 to 73 provide an interesting special case of “fixed” ataxia (t = 5 '35 ", O = -3), where the outer voices are opposed to the medium in a alternative 14 notes against 15 during 8 bars: the effect obtained oscillates between a paradoxical impression of simultaneous stability and instability, amplified by the regular or animated use of dynamics.

Finally, the use of totally random "Clouds," (e.g. mm. 40-49, t = 3 '26 ", O = -5) images of the most complete disorder, where the rhythms are at the discretion of space distribution in the score, where the pitches are freed from the rule of the sieves.

In the quicker  figures, thus played and perceived as equal, Xenakis uses fragments of sieves A (soprano) and B (alto), but borrows momentarily from other scales (e.g. C, tenor and baritone).

As if to emphasize the "alien" aspect of this passage, it is also here that we find the only F# 2 of the piece (with the exception of the Coda, where it appears twice in the baritone part).

Occasional techniques

Apart from writing processes developed throughout the piece as described above, Xenakis sporadically employed special techniques, particularly in the opening and final passages. Because of these stylistic breakthroughs, I allowed myself to isolate an "Introduction" (bars 1-9) and "Coda" (bars 88-100). In the initial part, in addition to the presentation of sieve A, acting as a sort of "herald" of the generative material used in the rest of the piece, Xenakis uses a variety of "Distortions" of pitches (e.g. bars 1-6 ) in order to change the base material. These techniques are generally related to sound acoustics: 1/4 tones, slow vibrato (or waves), and split sounds (multiphonics). But they can also sound quite "ordinary" in a classical acception, as in the case of the use of unison or trills : what is striking is their incongruity in Xenakis' writing. It is then their incongruity in the discourse of Xenakis that is striking.

In the coda, the writing of pitches is mainly unstructured, as Xenakis’ development takes the form of "Arborescences" a process that he had abandoned for several years. The writing is then essentially chromatic, based on rotations and permutations of intervals and simple cells (e.g. bars 88-94, t = 7 '11 ") in time and space. In the central part, Xenakis builds three short passages that I call "rarefactions".

They consist of moments of "rest" where one of the voices is highlighted, solo or accompanied by long tones, in which Xenakis chose to explore a specific aspect or instrumental technique. These breaks are particularly noteworthy since the four instruments are normally used together, in "compact writing," and are treated as roughly equivalent without undue concern for their "idiomatic" considerations.

On the contrary, these sections are representative of the specific nature of the solo instrument. The first is a stuttering solo by the alto sax in the low register (bars 22-24, t = 1 '51 "), during which each bar alternates the development of B, A, then A again, in an attempt at "melodism" that is quickly integrated into the canonic writing that follows. Second, entrusted to the soprano saxophone, is a kind of solo screaming in an extremely high register played at full power, supported by punctuation in the other three voices.

The third rarefaction is entrusted mainly to the baritone saxophone (bars 34-39), in the form of long low notes ​​in multiphonics ("split sounds"), and at a medium dynamic, before being joined by the other three (soprano, tenor and alto, using the same technique), and dissolves in the great central "aleatoric cloud".

A fourth kind of rarefaction starts in bar 81, with staggering (almost) inverted instrumental entries in the altissimo register (s, a, t, a, s), joined by the baritone holding a note in the low register (bar 85), which bursts into multiphonics (bars 86 and 87), recalling the process of the 3rd rarefaction. The successive entries here are: bar, ten, sop, alto. But, it does not indicate the predominance of a particular instrument relative to another: this passage seems rather to serve as a "bridge" between the recall of 34 and the final "chords."

The contrast between the three instruments in the altissimo register (because of its bias of extended tessitura) and baritone in the low register, desynchronised, is at the end of the work (e.g. bars 97-100, t = 8 '00 ") designed to generate the most curious of the "effects" in the piece: three (later two) altissimo notes played simultaneously generate an audible difference tone, then are doubled as an echo by the baritone at a very soft dynamic.

Order and form

Of particular note is the elegance of changes in the level of order in the piece. The first part of the work (bars 10-50) can be read as a journey from the highest order (canons at the sixteenth note in bars 10-16, O = 4) to complete disorder (“clouds” in measures 40-49, O = -5). I will not repeat here the details of the first half of this paper, which was discussed in the "Material" section, but I am interested mainly in relations between the two parts. The main break is placed exactly at the center of the piece (bar 50), where it contrasts a completely random section (rhythm and pitches) with parallel scales of Sieve A (O = 5). Note that the control of random processes allows the author to generate very convincing mass movements, as the transformation of an aleatoric cloud into joint scales is united by a progressive fixation on certain pitches (bar 50).

This major breakthrough is immediately followed by the passage containing maximum movement between order and disorder (e.g. bars 54-65, t = 3 '26 ", O = -4 to +4), which can therefore be considered the peak of instability. From there on, the end of the piece is conceived as a move towards stability, as the following two sections operate a return to a less severe disorder: firstly a partial disorder is stabilized (bars 66-73) the long ‘organ’-like sounds ensue (bars 73-81). This second part is a kind of blurred mirror centered around the physical center of the piece. Distorted images of the main sections of the first part can be found here in the form of "montages" in which the composer chooses to modify some of the original parameters. Among other examples, the ‘Organ’ tutti of bar 59 is in sixteenth notes, twice as fast as the "original" in bar 31. The independent lines of bar 60 are distributed over a wide ambitus and each voice remains stable, without mass movement, as in bars 19-21. The cloud in measures 61-62 remains fixed on Sieve B, whereas its model from bars 40-49 is generated from completely random pitches.

This very animated section is followed by a paradoxically calm passage that I call "fixed Ataxia" (bars 66-73, t = 5'35" O = -3) based on a particularly complicated rhythm of 14 against 15, the soprano and baritone on one hand, the alto and tenor on the other hand, each instrument playing alternately with his vis-à-vis. Supporting this rhythmic "unstable stability" are violent dynamic (p to fff) and tessitura changes. I think this passage is representative of a search for a "partial," or controlled, disorder. The following passage (bars 73-81, t = 6'08 ", O =. - 2) is another form of altered recall based on the organ sounds of bar 31, this time expanded over very long values, based on complex aggregates juxtaposing and superimposing the two original sieves (see musical examples). This section then connects to bars 81-87, which opens with the presence of shrill cries, followed by low sounds and multiphonics. This is both a reminder of the 3rd rarefaction and an anticipation of the end of the piece.

Note that the average size of the sections of the second part is larger than the first: their number is reduced accordingly, unless we decide to separate in small parts the section of "maximum movement" (bars 54-65) in each of its components, which seems to be contrary to the composer’s will.


Some passages are still at odds with the analysis: it is difficult to know if these are copying errors, the aesthetic choice of the composer, or a lack of finesse in this same analysis. Nevertheless I want to identify four of these "uncertainties":

Bar 5, the last note of the baritone saxophone: it makes more sense to read the note in treble clef (a), which would reconstitute a joint aggregate of sieve A merging perfectly with bar 6, whereas C in bass clef is borrowed from sieve B, which shouldn't appear before bar 8.

Bar 19, the baritone saxophone is written here without indicating 8va. It is also the only two notes (F and G, 2nd and 3rd notes of the 2nd beat) that foray outside of the assigned sieve …

Bars 24/25, linking between the last note of the alto saxophone and the first note of the soprano: the connection between these two notes seems to have slipped from one staff to another. Is this a question of relay? The alto sax indeed begins the canon in the next bar, after he has just played his stuttering solo.

Bar 31, all serious and unresolved chords are BABA except the last (7° eighth note) which is BBBA. If the tenor played a minor third lower, a Bb instead of the written Db, we could find a beautiful harmony of sieves. But that would also alas produce the harmony of a minor seventh chord, which is not quite adequate here...


Twelve years after its creation, XAS by Iannis Xenakis is (or has become) one of the most compelling pieces for saxophone quartet. Over time, one might even consider it as the first “serious” piece that was written by a major composer for this type of ensemble. This was often previously considered as ridiculous or in bad taste (similar to the brass quintet), as the saxophone was limited to its role in popular music and jazz by most major composers of the twentieth (as well as the nineteenth) century.

We can therefore thank Xenakis for his laudable non-conformism, which certainly contributed to the current passion for saxophone quartet. In fact, since the publication of XAS, the saxophone repertoire has been enriched with other prestigious and valuable pieces by Donatoni, Cage, Dufourt, and Aperghis, to name a few. These composers have also "dared" to confront the many unexplored possibilities of this rich formation of chamber music.

The saxophone quartets who decided to add XAS to their repertoire are rare - probably due to the necessary high technical and musical level required to play it –. But it should be noted that this is a piece that is frequently performed at saxophone quartet concerts. When the saxophone (the newest of the wind instruments, remember) and its performers have reached sufficient maturity not to be dissuaded by their first impressions, it is not unreasonable to suggest that this piece will become the major "classic" of the repertoire of the twentieth century for the instrument. Indeed, beneath a harsh and angular exterior, we find a generous and powerful work perfectly architected and teeming with musical ideas that even an unprepared public audience applauds.

Text by Serge Bertocchi, originally published in

«Présences de Iannis Xenakis» (CDMC),

translated with the precious help of Sean Fredenburg.