Home / Writings / le marteau sans maître by Pierre Boulez

The nature of commentary in music
as seen in the "Bourreaux de solitude" cycle
of le marteau sans maître by Pierre Boulez


Copyright (c) 2003 by Paul Nelson, all rights reserved.


Movement VIII of Pierre Boulez's le marteau sans maître is one of four movements of the work which are grouped together as the "Bourreaux de solitude" cycle. These movements are as follows:

bourreaux de solitude                                           movement VI

commentaire I de «bourreaux de solitude»            movement II

commentaire II de «bourreaux de solitude»           movement IV

commentaire III de «bourreaux de solitude»         movement VIII

Looking at these movements several questions immediately leap to mind. First, how is it possible for one piece of music to "comment" on another? Doesn't something as factual as commentary require words? What aspects of the "Bourreaux de solitude" are the subject of the commentary? Is it the music, the poem, or other extra-musical concerns?

As to the nature of commentary in music, a quote from Boulez provides a clue:

Even chorale variations present a problem. Although the original religious text has gone, a direct link with it remains in the syllabic disposition and the periods of the music. As commentary, it heightens the implicit significance of each verse, thus forming the classical example of the complex esoteric.1

Using this clue, we arrive at a possible method for musical commentary:  That each commentary "heightens" or demonstrates contrasts in some aspect of the original "Bourreaux de solitude" movement, and by doing so comments on that aspect.

In order for a commentary to be relevant, two rules must be met. First, the commentary must be clearly linked to the original "Bourreaux de solitude" movement. For example, it is impossible to consider the first movement of John Corigliano's first symphony (for example) to be a commentary on "Bourreaux de solitude" since the two movements are too different to be compared.

Second, the contrast in the subject of the commentary must be clearly demonstrated by showing the subject in two contrasting sections.

All three commentaries satisfy these two rules of contrasting commentary. In Commentary I, the subject is the presence and absence of the vocal line, as represented by the flute. In Commentary II, the subject is the virtual representation of the poem in its absence as seen in the pauses and pacing of the movement. These first two commentaries both contain two clear sections in which the subject is shown contrasted first one way and then the other.

In Commentary III, the subject is the use of intuition to construct a movement versus more systematic means, and the close relationship between intuition and the vocal line in the original "Bourreaux de solitude". Commentary III is not self-contained however. Instead it presents its subject in contrast against the first two commentaries.

The remainder of this paper will first discuss "Bourreaux de solitude" to lay the foundation for analysis. Next, each of the three commentaries will be discussed individually. Each commentary will be compared to the original "Bourreaux de solitude" to establish a connection between the two, and then the subject of the commentary will be identified and discussed.

"Bourreaux de solitude"

Movement VI, "Bourreaux de solitude," is the foundation for the entire "Bourreaux" cycle in le marteau sans maitre. In his article, "Speaking, Playing, Singing," Boulez declares that "the three Commentaries on 'Bourreaux de solitude' form a single large piece directly linked, from the formal point of view, to 'Bourreaux de solitude'."2  He further points out that the entire ensemble is never used continuously, except in this movement.3

But even though all instruments are used continuously in this movement, the resulting surface texture is marvelously transparent. The instruments are sparingly used, and the density of attacks is kept low. This gives the movement a sustained, gliding quality.

Overall, the movement has three layers of sound. The first layer is the voice, which surfaces first in measure 14 as an unaccompanied soloist, occasionally trading off with the instruments until measure 24 (see figure 1). The second time the voice enters, it is more integrated with the instrumental background (mm 45-67). According to Boulez, this cooperation between soloist and instruments is unlike other movements in marteau where the voice and the instruments struggle for primacy:

This antimony will be resolved in 'Bourreaux de solitude' into a total unity of voice and instruments, linked by the same musical structure, the voice emerging at intervals from the ensemble in order to enunciate the text. 4

The second layer of sound comes from the sustaining instruments, the viola, flute, and occasionally the vibraphone. The third layer of sound comes from the instruments with a sharper attack: the xylophone, guitar, and dampened vibraphone.

The role of the percussion is to fill in the gaps and provide a sense of pulse. Generally the pulse of the movement is clear, with the exception of mm 45-120 (and the end) where the sense of pulse is obscured somewhat. According to Boulez:

… [the percussion] appears in only one of the cycles, 'Bourreaux de solitude', in which it marks the time… When the other instruments are playing, the percussion is silent; but as soon as there is a pause, the percussion fills it with one or several strokes, according to the length of the pause. The percussion thus plays a complementary part, filling with indeterminate pitches the void left by the determinate pitches - a kind of architectural time game.5

The instruments and percussion are thus combined to create a gentle, rocking pointilistic background. This background is controlled using a technique called "Pitch-Duration-Association"6 which associates a pitch value with a particular note duration and volume level. Therefore, along with all 12 notes in the chromatic scale, one can also expect a wide range of note durations (from a single sixteenth note to a dotted half) and dynamics (from pp to ff with an accent). Boulez carefully gives the longer note durations to the sustaining instruments (flute, arco viola, vibraphone) and the shorter durations to the others. This method gives rise to the instrumental background found in the "Bourreaux de solitude".

In general, the thin background texture is created by having the instruments articulate notes at different times, rather than together. There are two exceptions to this. First, the xylophone and (to a lesser extent) the guitar often articulate their notes with someone else, rather than alone. This makes them disappear into the texture, as if they were simply sharpening the attack of the other instruments.

Second, every two to three measures several instruments will articulate at the same time (examples: mm 2, 6, . . . 34, 38, 41, etc.). These simultaneous articulations create "sign-posts" for the listener (and for the players, one suspects) which poke out above the rest of the texture. This has the effect of breaking up the texture to provide variety and help make the movement more dynamic.


Figure 1:  Graph of 'Bourreaux de solitude'

Figure 1 shows many of the indicators within the movement which help to articulate the structure of "Bourreaux de solitude". These indicators include:  1)  the voice, which has two major sections (14-24 and 45-67),  2) the flute, which uses flutter tongue twice in the movement, once before each voice entrance, 3) the viola, which punctuates the voice parts with occasional harmonics (shown with small circles, "o", above), and  4) the fermate, which occur in groups of three near the beginning and the end of the piece.

Using these indicators, the movement appears to be in two parts, each containing an instrumental introduction, a voice part, and an instrumental "echo". Part one starts on measure 1 and part two starts around measure 38.

The idea of an "instrumental echo" in marteau is described by Boulez:

The role of the voice is in fact extremely variable and ranges from primacy to absence, between direct expression of the text and expression of the poetic world that the text evokes. This might be described as intellectual drama prompted by reading the poem and the echoes that it creates in a world that is, properly speaking, interior.7

This is emphasized by the notion of the "poem as reflection":

The time taken to read a poem is a single, exact datum: but musically speaking there are two ties, one for the poem as action and one for the poem as reflection. . . .  8

One can hear these "instrumental echoes" or "reflections" quite clearly in measures 24 and 25, where the descending minor-third line in the voice is clearly echoed in the voice and guitar in measure 25. In addition, there is a wide spaced ascending line which occurs in the voice in measures 15 and 21 that is echoed in the instruments many times (viola mm 25-26, 30, 46; flute in mm 26, 37-38, etc.).

The fermate in the movement also emphasize the idea of echoes. The first time they occur as part of the vocal expression itself along with a series of rallentandi and slower tempo changes. The second time, however, they occur after the vocal expression, as if indicating where the voice line might have paused had it continued to participate.

Finally, it is important to note that there are two aspects of the piece which are consistent throughout the work and therefore are not used to articulate form. The first is the instrumentation, and the second is dynamics, which consistently varies from pp to ff as part of the PDA system.

Commentary I

In Commentary I, Boulez is commenting, quite explicitly, on the relationship between the voice and the textured background. While talking about movement V ('Bel edifice et les pressentiments') of marteau, Boulez states:

As the voice assumes an anonymous role, the flute takes its place in the forefront of the scene and, as it were, takes over the part of the voice. Thus the roles of voice and instrument are gradually reversed by the disappearance of the verbal text. This is an idea that I find valuable, and I should describe it as the poem being the center of the music though it is in fact absent from the music - just as the shape of an object is preserved by lava even when the object itself has vanished - or like the petrification of an object which is both RE-cognizable and UN-recognizable [sic].9

Boulez further states that the voice and the flute are linked by "the performer's breath, and the fact that both are monodic 'instruments'."

In Commentary I, the flute clearly substitutes for the voice. It rises in and out of the texture, in much the same way that the voice did in "Bourreaux de solitude," and then in the second half of the movement, starting in m. 54, it is absent altogether until it comes back in m. 104 for a short coda.

Note that, unlike the "Bel édifice" movement, the flute texture in the first half of Commentary I is more integrated with the instrumental texture, as the voice was in the "Bourreaux de solitude" movement. Another similarity can be found in the percussion, which maintains the pulse, fills the gaps in the instrumental texture, and helps glue the various parts of the movement together.

Figure 2:  'commentaire I de «bourreaux de solitude»'

In addition to his use of the flute, Boulez has made the structure of this movement quite clear with several additional devices (see figure 2), including:  1) double bar lines between sections, 2) a change in dynamic patterns (the second half has a louder mixture of dynamics than the first half), 3) use of fermate (the second half has them, the first does not), 4) the tempo indications (faster for the second half), and 5) the choice of percussion (the second half uses bongos, the first half snares).

The general shape of Commentary I is shown in figure 3. While this shape is clearest in commentaries I and II, I believe that this same general shape controls all of the movements in the "Bourreaux de solitude" cycle. In the original "Bourreaux de solitude" movement, the peaks in Part I and Part II coincide with entrances of the voice.

Figure 3: General Shape of the Movements in the "bourreax de solitude" Cycle

In figure 3 the coda section shows a "?" because its shape varies from movement to movement (and may even be absent). In the original Bourreaux movement it was relatively flat, although with a reduced tempo. In Commentary I the coda clearly intensifies to the end by means of he density and the average volume, which seems to anticipate the denser textures found in Commentary II.

To emphasize the commentary on presence and absence, Boulez elevates the instrumental texture when the flute is silent. This is done in several ways. First, the dynamics are louder on the whole, ranging from mp-mf at the beginning and rising to f-ff in the middle. This contrasts with the flute section which has a consistent pp-mf range that is most similar to the "Bourreaux de solitude" movement which has a pp-ff range throughout. Second, the percussion switches from snare drums to bongos. Put together, these changes make for a more assertive instrumental texture for the second half.

As a final observation on the instrumental texture, consider the viola pizzicato and the xylophone which often articulate notes at the same time, but not so often as to create a homorhythmic effect. This has the effect of blending the instruments together to make a more consistent texture. If, for example, they always articulated at different times, then they would be heard as two independent voices. Rather, since they often play together, they become a single instrument with a wider timbre range.

Commentary II

Commentary II is also divided into two parts which are clearly marked and differentiated. However, rather than commenting on the presence or absence of the solo line (represented by the flute in Commentary I), the key characteristic in Commentary II is the change in pacing, from a large number of very small fragments in the first part to fewer, larger and slower fragments in the second part.

Regarding fragmentation, Boulez states "The poem as reflection . . . may be submitted to a kind of fragmentation or distortion from its original form. . .".10  The poem itself is, of course, not actually present in Commentary II, but then Boulez does not need it to be there. The same effects of fragmentation can occur using the background texture (which now occupies the foreground in the absence of the voice or flute) and the use of fermate.

In the original "Bourreaux de solitude" movement the voice part frequently stopped and started, either with fermate (in the first section) or by simply dropping in and out (in the second section). Commentary II recalls and emphasizes these pauses in the original voice part, by over emphasizing (to the extreme) the fermate in the first part, and including similar pauses (although shorter in number with smaller pauses) in the second part (see figure 4).

Figure 4:  Graph of Movement IV
'commentaire II de «bourreaux de solitude»'

The vibraphone in the first part and the viola in the second part further connect Commentary II with the "Bourreaux de solitude". Both of these instruments play a dual role by producing both sustained and non-sustained sounds, and Boulez seems to be explicitly commenting on this aspect of the instrumentation. For example, when played pizzicato the viola is placed somewhere between the guitar (plucked strings) and the xylophone (minimal sustain). When played arco it is placed closer to the flute and the vibraphone.

Similarly, the vibraphone also plays a dual role with the dampers either on or off. When the damper is off, the vibraphone is most similar to the guitar, and when the damper is on it is most similar to the xylophone. These observations when coupled with additional comments by Boulez 11 give rise to the following ordering of instruments: Voice, flute, viola-arco, vibraphone-undampened, guitar, viola-pizzicato, vibraphone-dampened, xylophone, and percussion.

In the first half of Commentary II, the vibraphone is conspicuously oscillating between dampened and sustained notes and as such it either emerges from the texture or is absorbed by it. This is especially apparent when notes or rests are held with fermate. In this way the vibraphone acts as a substitute for the voice part in the "Bourreaux", further connecting the movements together.

In the second half, the viola, which had been pizzicato before, is now playing arco, taking over the role of the sustained element of the work. Not only does this demonstrate the dual role of the viola, but it also serves to clearly mark the structure of the movement to the ear (see figure 4).

Again, percussion is used to fill gaps in the instrumental background and mark time, which, due to the fermate, is especially needed. Further, three types of percussion are used, all using struck metal: cymbalettes, cloche, and triangle. This "struck metal" percussion gives Commentary II a distinct sound that will be referenced later in Commentary III.

Commentary III

The first two commentaries contained internal contrasts. Each one had two parts and the contrast was highlighted between the two parts. In the third commentary, the contrast is not internal to the movement, rather it is between Commentary III as a whole and first two commentaries.

In the first two commentaries the background texture was elevated to the forefront as the flute (acting as a substitute for the voice) dropped out in the middle of Commentary I. Even when the flute was present, it was merged with the texture, as were the vibraphone and viola in Commentary II.

In contrast, Commentary III contains a present and assertive flute line, often playing cadenza-like or strident passages. The commentary here is two fold. First, the solo line which had been submerged before is highlighted while the background texture is thinned out and deemphasized.

Second, and perhaps more importantly, Commentary III is a commentary on the use of intuitive control versus the more systematic control of the prior two commentaries. The is considerable evidence to support this interpretation, as follows:

First, the flute writing is much more recognizable as an independent line, it clearly stands out above the sparser background texture. This can be seen in measures 20-33 and 110-117 where the flute is very sparsely accompanied, measures 81-83 where it is unaccompanied. This connection between the flute and the intuitive nature of Commentary III mirrors the treatment of the voice in the "Bourreaux de solitude" movement, where it also flexibly treated.

Second, the writing of all the instruments (except for percussion) contains several cadenza-like passages with considerable ornamentation (in terms of grace notes). This occurs in the flute in measures 25, 28-29, 77-83, and 110-114, the vibraphone in measures 84-88, and the xylophone in measures 87-93.

Third, the tempo fluctuates freely emphasizing the fluid nature of the movement. This is similar to the fluctuations in Commentary II, but without the regular occurrence of fermate. The original "Bourreaux de solitude" and Commentary I are considerably more restrained in tempo fluctuations.

Figure 5:  Graph of Movement VIII
'commentaire III de «bourreaux de solitude»'

Fourth, as has been noted by Wentzel12, Commentary III depends on the PDA system much less rigorously than the original "boureaux de solitude" movement (in which the PDA system is used throughout) and Commentary I (in which it is used for the entire first section). In Commentary III, clear PDAs have only been found in measures 71-76, and 93-95 (both of which are labeled as "languid" flute sections in figure 5). Of course, this argument is weakened somewhat by Commentary II, in which no clear PDA system has yet been discovered.

Finally, Commentary III contains a wider range of instrumental combinations and percussion instruments than the previous movements. The percussion uses the claves, the cloche, the bongos, and the maracas - all instruments with quite different timbres. Other interesting combinations include: 1) the distinctive use of vibraphone pedal in measures 84-89,  2) the xylophone solo in measures 89-92, 3) the opening instrumental texture which recalls the first two commentaries, and 4) the use of bongos plus maracas to close the movement.

Commentary III is linked to all of the other movements in the Bourreaux cycle. As discussed above, it is linked to the first two commentaries by contrast. Other links to the first two movements include:  1) the opening section (mm 1-20) which recalls the more active background textures of the first two commentaries, 2) the use of percussion: the bongos from the first commentary, and the struck metal percussion (the cloche) from the second commentary, and 3) Commentary III contains two parts, measures 1-51 and 51-end, a structure that is made clear by the use of percussion, as it was in the first two commentaries.

Commentary III is also linked to the original "Bourreaux de solitude" movement:  1) The solo flute passages in Commentary III are a clear reference to the solo voice passages in the original "Bourreaux de solitude" movement,  2) maracas are used in both movements, 3) both movements contain sustaining instruments (vibraphone in Commentary III, viola and flute in the original "Bourreaux de solitude" movement) which act as a "middle" layer between the pointilistic background texture and the solo line, and 4) both movements have a similar relaxed pace.

Finally, both Commentary III and the original "Bourreaux de solitude" movement appear to use motives to help unify the pieces (see above for a discussion of the motives found in the "Bourreaux de solitude" movement). In Commentary III, there is conspicuous use of a "jumping" motive (for example, the flute in measure 78), where there is a quick leap away from and back to a note (leaping up or down). This could possibly be related to the original vocal line (measures 66-67 on the words "gra-nit ré…") from the "Bourreaux de solitude" movement. Looking through the notes of the vocal line in the Bourreaux movement, "gra-nit ré…" is the only time there is a conspicuous repetition of a note (the C#).

In summary, Commentary III accomplishes many different functions. First, it serves as a contrast to the prior two commentaries by putting the flute line (as a proxy for the voice) in the foreground instead of highlighting the instrumental texture. Second, the fluid form of the movement highlights the use of intuition over the more consistent application of systematic methods in the previous two commentaries. Finally, it summaries the entire "Bourreaux de solitude" cycle by referencing and integrating aspects of all the movements in the cycle.


Each of the three commentaries on the "Bourreaux de solitude" explore a different aspect of the notion of "centre and absence" in poetry and music. The first commentary shows how, when the poem is absent, the instruments rise to the foreground and how this background texture can be a beautiful work of art on its own terms. The second commentary demonstrates how music can be affected by the pauses and pacing of poetry even when the poetry is absent. The third commentary completes the discussion by lifting the vocal line to the centre, demonstrating by example and contrast how this affects our perception of the background.

Each of the commentaries operate under two basic rules. First, the connection between the commentary and the original "Bourreaux de solitude" movement is made clear. Second, the subject of the commentary is made clear by showing two contrasting sections.

With these commentaries, Boulez is pointing out parts of music which might otherwise be overlooked. We too often focus on the object and too often under appreciate what surrounds it. Lava can, in fact, be beautiful all by itself, as can a picture frame or the setting of a diamond. In the same way, Boulez's le marteau sans maître can be admired and enjoyed not only for the poetry and the setting of the words, but also for all of the beautiful textures and colors which support it.


1  Pierre Boulez, "Poetry - Centre and Absence - Music," from Orientations, Collected Writings by Pierre Boulez, edited by Jean-Jacques Nattiez, translated by Martin Cooper (Cambridge, Massachusetts: Harvard University Press, 1986), p. 185.

2  Pierre Boulez, "Speaking, Playing, Singing, Pierrot Lunaire and Le marteau sans maître," from Orientations, Collected Writings by Pierre Boulez, edited by Jean-Jacques Nattiez, translated by Martin Cooper (Cambridge, Massachusetts: Harvard University Press, 1986), p. 341.

3  Boulez, "Speaking, Playing, Singing," p. 340.

4  Boulez, "Speaking, Playing, Singing," p. 338.

5  Boulez, "Speaking, Playing, Singing," p. 340.

6  Steven D. Winick, "Symmetry and Pitch-Duration Associations in Boulez's Le Marteau sans maître," Perspectives of New Music 24, no. 2 (1986), p 282.

7  Boulez, "Speaking, Playing, Singing," p. 340.

8  Boulez, "Centre and Absence," p. 196.

9  Boulez, "Speaking, Playing, Singing," p. 339.

10  Boulez, "Centre and Absence," p. 196.

11  Boulez, "Speaking, Playing, Singing," p. 340.

12  Wayne C. Wentzel, "Dynamic and Attack Associations in Boulez's Le Marteau sans maître," Perspectives of New Music 29 (1991), p 142.