DOI: 10.1111/musa.

12141

MARINA SUDO

More than Meets the Eye: Derivations and Stratification in

Boulez’s ‘Don’ (1962)

The poetry and aesthetics of Stéphane Mallarmé were an important creative

inspiration for the early and middle-period works of Pierre Boulez (1925–2016).
After previous explorations in Livre pour quatuor (1948–9), the unrealised plan
of Un coup de dés, and the Troisième sonate (1957–9), Boulez proceeded to Pli
selon pli: Portrait de Mallarmé (1957–62, revised in 1982 and again in 1989)
for soprano and orchestra, which would present not only a synthesis of his
study of Mallarméan thought, but also a synthesis of his creative trajectory
from the late 1940s on.1 The Pli selon pli cycle, comprising five movements –
‘Don’ (1960, first version, for voice and piano; 1962/1989, second version, for
voice and orchestra); ‘Première improvisation sur Mallarmé “Le vierge, le vivace
et le bel aujourd’hui”’ (1957/1962); ‘Deuxième improvisation sur Mallarmé
“Une dentelle s’abolit”’ (1957); ‘Troisième improvisation sur Mallarmé “À la
nue accablante tu”’ (1959/1982); and ‘Tombeau’ (1959/1960/1962) – exhibits
a complex network of material based on filiation planning. This underlying
filiation, whether achieved by the quotation of excerpts or by the manipulation
of the fundamental serial materials, provides us with a remarkable example
of Boulez’s compositional dialectic between the systematic development of
structural ideas and their flexible realisation in sound. This is, indeed, a central
problem faced by composers throughout the history of Western music: how
should one treat and reformulate the basic material in an imaginative way whilst
still providing a solid structural network?
The present article aims to shed light on Boulez’s planning and realisation
of material in the genesis of ‘Don’ by focusing on the composer’s renewed
treatment of material through processes of derivation. In the first part, it will
be shown that the work’s structure is conceived primarily in a symmetrical
and complementary relationship with ‘Tombeau’, though this is veiled by a
number of additional operations. The second part examines Boulez’s selection of
materials, which suggests interactions between the planning and the realisation
of materials during the compositional process.

Music Analysis, 39/iii (2020) 359

© 2021 The Authors.
Music Analysis © 2021 John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK
and 350 Main Street, Malden, MA 02148, USA
360 MARINA SUDO

From ‘Tombeau’ to ‘Don’: Construction of a Symmetrical and

Complementary Structure
On 17 October 1959 Boulez’s Tombeau à la mémoire du prince Max Egon zu
Fürstenberg was premièred as the opening work on the Saturday evening concert
of the Donaueschingen Festival. This was the first version of what was shortly
to become the last movement of Pli selon pli. At the end of December 1959,
the composer outlined in a letter to Karlheinz Stockhausen his plan not only
to revise this first ‘Tombeau’, but also to compose an extensive cycle based on
poems by Mallarmé that would include his three ‘Improvisations sur Mallarmé’,
which he had already written: ‘I have enlarged the plan of my Mallarmé, to the
extent of making a long work that will finish Mallarmé for me for some time
until I recover a taste for the coup de dés – which will be an essay in total synthesis
[…] At this moment I’m in the process of expanding the Tombeau (if I dare risk
this macabre comparison) […] to the point of turning it into a mausoleum!!’2 It
was around this time that Boulez started to conceive the composition of ‘Don’,
which initially came to fruition in a version for soprano and piano that was
premièred in Cologne in June 1960 as part of the first appearance of Pli selon
pli with all five movements. Despite the ‘completion’ of the Pli selon pli cycle,
this version of ‘Don’ for soprano and piano was withdrawn soon afterwards and
in 1962 replaced by the orchestral version.3 In another letter to Stockhausen
written sometime in November 1960 or thereabouts, Boulez declares, ‘I’m really
making Don [je fais vraiment le Don].’4 Logic dictates, therefore, that it was
after the première of Pli selon pli in June 1960 that the composer started the
full-scale composition of the orchestral version of ‘Don’, which he conceived
in a thoroughgoing relationship with ‘Tombeau’, essentially for the sake of
aesthetic proportions; the cycle now starts with gift/birth (don) followed by three
‘Improvisations’, and it ends with death (tombeau). The first version of ‘Don’
can thus be viewed as a stopgap first movement composed for the 1960 concert.
This intricate chronology makes the position of the first movement clear
in Boulez’s planning. As a summary of the whole cycle, ‘Don’ embraces
the other four movements, which were more or less finished by the time of
this movement’s composition,5 similar to the relation of the ninth and last
movement of Le Marteau sans maître (1952–5), ‘Bel édifice et les pressentiments,
double’, to the others. The summation is concretised first by quotations
from the three ‘Improvisations sur Mallarmé’ and secondly by a symmetrical
and complementary relationship between ‘Don’ and ‘Tombeau’. Although the
former is a perceptible presentation of the composer’s vision, the latter, by
contrast, involves no more than an underlying structure. In the following
paragraphs, I will concentrate on this structural correlation between the two
movements.
The principle of symmetry between ‘Don’ and ‘Tombeau’ is manifested at the
earliest stage of sketching. Based on the composer’s plan for ‘Fin de Tombeau =
Commencement de Don (Joyce Fin[n]egans Wake)’,6 the same ffff tutti chord

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 361

Fig. 1 Schema illustrating the symmetry between the six subsections of the coda
of ‘Tombeau’ and six sections of ‘Don’ (including the unrealised section e). The
central bar line represents the symbolic axis of the symmetry, i.e. the identical chord
sounded at the beginning of ‘Don’ and at the end of ‘Tombeau’

is sounded at the beginning of ‘Don’ and the end of ‘Tombeau’. This fully
chromatic twelve-note chord (see Fig. 1) appears with the exact same registral
disposition in the two movements, indicating the opening and closing points of
the cycle. It is on this very chord that the six-section structural framework of
both movements is based, as represented in Fig. 1 by the letters a–e for ‘Don’
and a–f for the coda of ‘Tombeau’. All six sections of the two movements are
paired in inverse order, forming symmetry with respect to the axis represented
by the tutti chord. (Section f of ‘Don’, shown in brackets in Fig. 1, is not realised
in the final composition; it appears only in the first sketch.)
It is on the basis of this framework that the composer developed the filiation
from ‘Tombeau’ to ‘Don’. Its overall plan was drawn up using the multiplication
tables (blocs sonores) shown in Ex. 1, which had already been employed in
the cycle ‘L’artisanat furieux’ of Le Marteau sans maître but was originally
conceived for the twelve-voice Oubli signal lapidé (1952). Multiplication is one
of the Boulez’s most important compositional techniques, one that he uses in,

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
362 MARINA SUDO

Ex. 1 Multiplication table, with boxes marking the principal materials in ‘Tombeau’
(sections b–f) and italicised letters (top left) by the author. The sections marked at
the bottom of the table are also by the author. Transcript of a sheet from the pencil
manuscript; PSS, Pierre Boulez Collection, Mappe F, Dossier 2a, 5 (omitting minor
details). This page of the sketch has been published in the facsimiles editions of Le
Marteau sans maître (Decroupet 2005, p. 83) and ‘Tombeau’ (Piencikowski 2010, p.
50)

a b c d e

for example, the Troisième sonate (1955–7), Structures II (1956–61), Domaines

(1961–8) and cummings ist der dichter (1970). It is also the subject of a number
of earlier studies.7 Before examining the structural relationship between ‘Don’
and ‘Tombeau’ in detail, in the next few paragraphs we will briefly look at the
mechanism of pitch-class multiplication in Ex. 1.
The example shows five columns represented by the Greek letters Αλ (for
alpha), Βη (beta), α (gamma), ε (delta) and Επ (epsilon), each of which
contains five lines: Μυ (mu), Νυ (nu), ι (chi), ΐμ (omega) and ι (pi). In
each column the first line, Μυ, comprises five fundamental pitch classes, which
are deduced from different partitionings of the basic twelve-note row, Eb–F–D–
C#–Bb–B–A–C–Ab–E–G–F# (e.g. 24213 for ΜυΑλ, 42132 for ΜυΒη, etc.). All
other complexities in the lower lines result from the multiplication of one set of
fundamental pitch classes by another set. In the basic process of multiplication
(Ex. 2a), first the pitch classes of the multiplicand are transposed to those of
the multiplier. All resultant pitch classes are then added up to a superset with
higher density, duplicated pitch classes being eliminated. The extracted table
of the fifth column, Επ (Ex. 3), shows its multiplicative outline: although five
fundamentals in the first line are represented as the italicised letters a–e, letter
combinations (ba, cb, dd, ec, etc.) identify their two composite fundamentals,
that is, the multiplier and the multiplicand. In order to obtain the notated pitch
classes in the table, however, another transposition needs to be conducted as a
part of the multiplicative process – Stephen Heinemann (1998) calls it ‘complex
multiplication’ as opposed to the aforementioned ‘simple multiplication’. This
additional operation has been more thoroughly investigated in recent studies by

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 363

Ex. 2 (a) Simple multiplication of ba

(b) Complex multiplication of ba based on the transposed multiplier (T8 )

Ex. 3 Vertical commentary on ed/ee in section e of ‘Tombeau’

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
364 MARINA SUDO

Fig. 2 Transpositional matrix based on the five lowest notes in the first line (Μυ) in
Ex. 3 (adapted from Losada 2008 and 2014)

Catherine Losada (2008 and 2014). What distinguishes her contribution is her
insight into the structural trigger of the transpositional scheme. Based on her
research on Boulez’s sketches, she contends that ‘the scheme associated with
complex multiplication is intimately tied to the partial ordering in register of
both factors and products of the multiplication operations’ (2014, p. 90).8 The
partial ordering mainly involves ‘anchor notes’, which are the lowest notes of
the five partitioned fundamentals in Μυ. In Ex. 3, for instance, the anchor
notes are D, Bb, Ab, E and F#. The intervals of these five notes (ic8, ic10, ic8
and ic2) are subsequently utilised to generate the transpositional scheme for the
lower lines (Fig. 2). When multiplying two sets of the fundamental pitch classes,
multipliers’ anchor notes are successively transposed by these intervals, T8 , T10 ,
T8 and T2 , as seen in the case of ba, exemplified in Ex. 2b. Note that transposed
anchor notes lead to the transposition of the complete sets of multipliers’ pitch
classes. This is why the matrix of anchor notes in Ex. 3 works as a determinant
of the register of the resulting complexities: most of the transposed anchor notes
are set in the lowest position, on which the other unordered pitch classes are
superimposed (compare Ex. 3 and Fig. 2).9
Such a generative process leads us to an essential feature of the multiplication
table: the presence of multifold isomorphic relationships spread throughout
the table. This will form a crucial point of the discussion when I turn in the
second part of this article to Boulez’s unique selection of materials. For now,
whilst concentrating on how the composer used this multiplication table in
his compositional strategy, I return to the symmetrical structure of ‘Don’ and
‘Tombeau’.
This symmetrical structure was initially pursued through a positional
relationship of the materials employed in each movement. In order to organise
the pitch structure in each paired section of ‘Don’ and ‘Tombeau’ (see again

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 365

Fig. 1), Boulez used the same set of complexities from the same column; for
instance (see again Ex. 1), the material from the fifth column, Επ, used in section
e of ‘Tombeau’ is also employed in section b of ‘Don’. From the shared set of
materials embraced in the relevant column, the ‘principal’ pitches in ‘Don’ are
selected to complement those of ‘Tombeau’. In other words, the structure of
all the sections in ‘Don’ derives entirely from the position of the main sound
blocks assigned to the five sections of ‘Tombeau’ (b–f). For this reason, the
pitch organisation for ‘Tombeau’ needs to be revealed before the compositional
process of ‘Don’ is examined.
In Ex. 1 five sets of one or two blocks (enclosed in boxes) show the principal
materials for ‘Tombeau’ (section b–f). These outlined blocks present only the
fundamentals, and ‘commentaries’ on them enabled the composer to deduce a
new set of materials.10 The concept of the commentary has nothing to do with
an additional technical operation but rather implies an extensive reading of the
existing table. It is a device that provides the composer with a wider range of
operable materials without weakening his firm conception of the movements.
A new set of materials is often deduced from the unique position of certain
blocks within the respective columns. For example, section e of ‘Tombeau’ shows
how the operation of a ‘vertical commentary’ works in his selection of pitches
(see Ex. 3). In this section, the composer applies this vertical commentary to
the bottom blocks ed and ee of ιΕπ as fundamental materials (G#/B, Bb), as
highlighted in Ex. 3. The commentary results in the expansion of the available
materials, with the ten blocks in the broken-line box derived vertically from the
original bottom two; these blocks are used in a symmetrical trajectory from 1 to
10 in the resulting music of the section e of ‘Tombeau’ (bars 536–548).
It is these ten blocks shown in Ex. 3 that form the basis for section b of ‘Don’.
The composer chose a series of blocks in accordance with ‘complementary
horizontal commentaries in vertical order [of ‘Tombeau’]’,11 but with a reverse
trajectory. Let us consider the details with a reductive diagram (Fig. 3a). In the
opening of section b of ‘Don’, the set of four blocks in the third line is used.
It begins with block ca and proceeds horizontally up to cd; structurally, this set
of blocks corresponds to ce in ‘Tombeau’, that is, the tenth block in Ex. 3. The
reference block ce for ‘Tombeau’ being avoided in ‘Don’, the composer adopts a
similar strategy to the replacement of the structural pitches in a subsection called
‘Glose’ from ‘Trope’, the second movement of the Troisième sonate (O’Hagan
1998). The same principle is applied in the next phrase of ‘Don’: the set of four
blocks in the fourth line (da, db, dc and dd) is deduced from de in ‘Tombeau’. In
the succeeding phrase, a further operation of ‘vertical preparation’ is introduced
in the third set of blocks (Fig. 3b), which corresponds to ee for ‘Tombeau’. With
this operation, the four blocks at the bottom (ea–ed) are combined with the upper
enclosed blocks, and these original four appear only as background materials.
Thus, the outlined three sets of materials are assigned with high (1), middle (2)
and low (3) registrations, as shown in Ex. 4.

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
366 MARINA SUDO

Fig. 3 Derivation of materials for section b of ‘Don’

(a) Horizontal commentary on the block ce
(b) Horizontal commentary on ee with vertical preparation (the reductive diagram
corresponds to Ex. 3)

(a) (b)

Ex. 4 Transcript of a sheet from the pencil manuscript of ‘Don’ (PSS, Pierre Boulez
Collection, Mappe G, Dossier 3b, 7), omitting minor details (italicised letters at the
bottom are the author’s)

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 367

As these examples suggest, the sketch plan of ‘Don’ was written alongside the
outline of materials for ‘Tombeau’. It does not mean that the two pieces were
composed in parallel, but rather that Boulez used the structure of ‘Tombeau’
as a reference for the composition of ‘Don’, thus allowing him to design the
complementary positional relationship based on shared sets of materials in the
multiplication table. In this respect, the blueprint of ‘Don’ is derived entirely
from the structural outline of ‘Tombeau’, though with varying interpretations
of complementarity for each paired section. Furthermore, Boulez builds up a
number of other underlying links between the two movements, not only by using
the positional relationship mentioned above, but also by engaging in a behind-
the-scenes reworking of the elements of ‘Tombeau’. This is especially adapted in
sections c and e of ‘Don’, whose generative processes provide good examples of
Boulez’s method of derivation.
Sections b and d of ‘Tombeau’ (bars 530 and 532–535), which correspond to
sections e and c of ‘Don’ (pp. 28–38; pp. 14–21 of the published score), have
the same instrumentation and the same last line from Mallarmé’s ‘Tombeau’
(1897): guitar and horn, with the soprano melismatically singing ‘un peu
profond ruisseau/calomnié’. It is in this trio that Boulez explores the veiled
link that runs from ‘Tombeau’ to ‘Don’. In these two sections of ‘Tombeau’,
the horn and soprano parts make up a two-voice linear ensemble, whose pitch
organisation is built on the blocks in the table. In contrast, the guitar part is more
improvisational and does not adhere strictly to the table of materials. Boulez
does not, however, use this instrumental and vocal material directly in ‘Don’.
Instead, as is described more fully below, the original material is recomposed
within a new system. At the same time, this process of reinterpreting ‘Tombeau’
will also be integrated with the concept of positional complementarity, which
eventually results in a double correlation between the two movements.12
Section e of ‘Don’ consists of antiphony between the two main instrumental
groups of woodwinds/horns and strings/brass, augmented by improvisational
passages for piano, harp and xylophone. The construction of these three groups
is derived respectively from the voice, horn and guitar parts in ‘Tombeau’. In
the piano-harp-xylophone group, there is an acoustic allusion to characteristic
aspects of the guitar part in ‘Tombeau’ through analogical imitation of
rhythm and articulation (for example, arpeggiated chords, rapid passages, single
sustained notes, etc.). However, the voice-and-horn texture provides a source
of inspiration for the woodwind/horn and string/brass instrumental groups at a
more fundamental level. The following description will concentrate on how the
woodwind/horn group is structurally derived from the vocal line of ‘Tombeau’.
The derivation process starts with a reduction of the vocal part of ‘Tombeau’
(bar 530), using the block be and its vertical commentary (e, ce, de, and ee) in Bη
(Ex. 5). They are combined with two levels of qualitatively different durations,
as shown in Ex. 6: a series of long values for the fundamental block be (D#, E
and D), and different groups of interpolated short values for the four blocks of
commentary. It is from this outline that Boulez builds up, for section e of ‘Don’

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
368 MARINA SUDO

Ex. 5 Vertical commentary on the block be for section b of ‘Tombeau’

with its five component subsections (A–E), a new system for the woodwind/horn
strata. To obtain the resulting series of sound complexes for each subsection
shown in Ex. 7, the composer goes through several reconstructive processes.
First, as a framework for successive operations, the four blocks obtained by
horizontal commentaries on the fundamental block be (ba, bb, bc and bd) are
assigned respectively to subsections A–D, and secondly, the four blocks in the
vertical commentary (ee, de, ce and e) are assigned to subsection E (blocks
printed in boldface in Figs 4a and 4b). All these blocks are ‘distributed in 6
long durations as in the original’ to become the top notes of the complexes in
each subsection (represented as semibreves in Ex. 6).13 As a result, the set of
long values originally attached to the fundamental block be in ‘Tombeau’ are, in
‘Don’, re-spread out on the component pitches of the blocks derived from the
aforementioned commentaries. The rest of the top notes (the black noteheads
in Ex. 6) are deduced from the ‘original’ blocks by further procedures as follows

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 369

Ex. 6 A representation of Boulez’s reduction of the vocal part of section b of

‘Tombeau’ (bar 530) as part of his preparation for the woodwind/horn strata in
section e of ‘Don’

Ex. 7 Transcript of a sketch from ‘Don’ (PSS, Pierre Boulez Collection, Mappe G,
Dossier 3b, 5), omitting minor details; the numbers in the first line are the author’s.
(The pitches in the original are notated either in black, blue or green, the various
colours implying the combinatorial process of the blocks, as described in the main
text.)

(see blocks enclosed in boxes in Figs 4a and 4b). Four vertical blocks for
subsections A–D and four blocks (ea, b, cc and dd) for subsection E are derived
through a double procedure of selection, combining the remaining four lines and
the four columns in Bη.14 Thus, below these disposed top notes, the composer
builds up several other notes that ultimately form the entirety of the complexes.
All of these resulting agglomerations are also logically constructed on a certain
set of blocks from the multiplication table, essentially related to ‘Tombeau’.
This will become even more evident when we consider certain details from
subsection E.
In subsection E, the original blocks (ee, de, ce and e) are not ‘described’ in
the same way as the semibreve linear notes in subsections A–D;15 rather, they
maintain their shape as a vertical complex. These four blocks, labelled by the
composer as α, β and γ (see subsection E in Ex. 7), function as important

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
370 MARINA SUDO

Fig. 4 Derivation of fundamental blocks for (a) subsection A–D and (b) subsection
E shown in Ex. 7

(a) (b)

Ex. 8 Three different modes of multiplication based on the multiplicand α and the
top note of A1 from Ex. 7

subsets for multiplying the top notes in all subsections A–E. Note that the
blocks e (G and F#) and ce (D and C#) are related by transposition (T7 ) and
therefore share the same label γ . Using these sets of blocks as multiplicands of
the top notes, Boulez executes an additional multiplicative operation in three
different ways in order to change the density of the resulting complexes. In
subsection A, for example, all of the complexes are deduced by the application
of multiplier α, but the number of component notes varies. There are three
different levels of density: complexes with three, four and five notes. As shown
in Ex. 8, the first step is simply the multiplication of multiplier α with pitch

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 371

C (first semibreve of line A in Ex. 7 = A1), which results in a new block (α1:
C/Db/B). The same process is used to deduce A2, A3, A4 and all other blocks
containing three blocks. The second and third steps demonstrate the process
of ‘multiplied multiplication’: if we apply the same operation of multiplication
on these resulting notes C, Db and B, we obtain two other blocks (α2: B/C/Bb
and α3: Db/D/C). By combining blocks α1 and α2, we obtain a four-note block
(C/Bb/Db/B), for example A5, A6 and A9, and adding the blocks α1, α2 and α3
results in a five-note block (C/Bb/D/Db/B), for example A7, A8 and A11. Boulez
thus acquires more possibilities for materials by means of three different types
of multiplication based on the three fundamental multipliers α, β and γ in all
five subsections. This operation finally leads us to all resulting agglomerations
shown in Ex. 7, which are combined with the rhythmic series notated above in
each subsection and finally put into place in the final score.
The compositional system of ‘Don’, as demonstrated so far, is made up of
minute deductive elaborations of the original materials. Thus the fully realised
material in the score never appears in the same form as the material that acts
as its starting point; there is a considerable distance between the two. It is
indisputable that the creative process of ‘Don’, as is often the case with Boulez’s
works, is based on gradual deductive steps from elaboration to mise en place of the
prepared material. At first glance these generative steps might appear to follow
a simple trajectory. Yet these stages cannot really be described as unidirectional;
rather, they present a continual back-and-forth in his compositional practice.
Some sketches of ‘Don’ suggest that the composer reflected on the possibility
of sound realisation at the earliest stages of planning by carefully selecting
the blocks in the multiplication table. In other words, Boulez’s elaboration of
material is not always a result of procedural thinking, but may come about
through a more subtle conflict between system and freedom. The next section
will focus on these strategic selections of material.

Commentaries as a Preparation for mise en place: Quotation of

‘Improvisation III’ and Introduction
Apart from the symmetrical relationship with ‘Tombeau’, another Boulez tactic
for synthesising Pli selon pli was to quote excerpts from the three ‘Improvisations
sur Mallarmé’ in sections b and d of ‘Don’.16 Whereas the symmetrical
relationship, as shown, makes use of underlying and therefore imperceptible
filiations, these quotations are tangible compositional gestures.
Five instrumental fragments from ‘Improvisation III “À la nue accablante
tu”’ are quoted in section b, as are three vocal fragments from all three
‘Improvisations’ in section d. The quoted fragments are readily identified in
the published score of ‘Don’, since each is enclosed in a box.17 At the level
of perception, furthermore, the appearance of these quotations is often marked
out by their stillness. Locating their point of origin in the three ‘Improvisations’

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
372 MARINA SUDO

is more complicated, however, since the fragments are to varying degrees

reformulated. For instance, although one of the instrumental quotes from
‘Improvisation III’ – a section called ‘Hétérophonie’, for two xylophones – in
section b of ‘Don’ is almost exact, it differs from the original, which has here
been separated into several parts and inserted in a different order.18 Other
quotations are more approximate. The third quotation on pages 8–10 of ‘Don’
(rehearsal numbers 9–11 in the revised version), for example, is taken from pages
38–39 of ‘Improvisation III’ but retains only the outline figuration and dynamics,
the details of pitch and ornamentation having been greatly modified. The vocal
quotations in section d of ‘Don’ show a similarly varied approach: although
the quotations from ‘Improvisations II and III’ are fairly straightforward, that
from ‘Improvisation I’ differs considerably from the original in that the text is
fragmented and the melodic lines reconstructed.
Another aspect of the quotations is equally important in section b. Beneath the
instrumental fragments from ‘Improvisation III’, Boulez inserts sustained piano
sounds. These derive from the multiplication table, his particular selection of
materials being strongly associated with the network of isomorphs therein. As
previously discussed, this multifold isomorphic network is a natural outcome
of the multiplicative process. In fact, what is included in this network is not
only isomorphs but also a considerable number of identical blocks with the same
notes. For example, Pascal Decroupet and Jean-Louis Leleu (2006, p. 189) point
out that six sets of identical blocks are symmetrically situated in relation to the
diagonal from top right to bottom left in the fifth column of Ex. 9. This is a
predictable consequence of (1) the same pairing in inverse relation between
multiplicand and multiplier (e.g. bc/cb or bd/db) and (2) the transpositional
scheme deriving from the five anchor notes. In Fig. 2 the same pitch classes
are situated diagonally, matching the placement of the identical blocks in Ex. 9.
Note that the transposed anchor notes –mostly the lowest notes – determine the
register of the resulting multiplications. It can therefore be said that the network
of identical blocks shown in Ex. 9 is entirely a result of the transpositional outline
in Fig. 2. What is striking is that these series of identical blocks function not only
as a potential network in the structural level, but also as a perceptible musical
gesture in the final composition of ‘Don’. Boulez exploits this characteristic of
the multiplication table with the intent of creating a specific sound beneath the
quotations.
The following description demonstrates an example from section b, where
Boulez introduces a quotation from ‘Improvisation III’. As previously discussed,
the structural outline of the first half of section b is designed according to the
principle of complementarity with ‘Tombeau’, but in the second half Boulez
sketches a series of blocks that form ‘diagonal commentaries’ for the background
piano part of the quotation. First of all, cc, in column Επ, which is the last
block used in the previous part of section b, is chosen as a starting point to
select four other blocks that together form a left diagonal line, that is, one
that descends diagonally from right to left, cc–db–ea–e–bd (Fig. 5a). These five

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 373

Ex. 9 Reproduction of Ex. 11 in Decroupet and Leleu (2006, p. 189)

blocks are vertically combined with other blocks, though with a progressive
reduction of the commentaries: the first contains a vertical set of five blocks,
the second four, the third three, and so on (see the blocks in boldface in
Fig. 5b). After this process a set of diagonal commentaries is applied again.
When developing these supplementary diagonals, Boulez changed his strategy,
and in the end he used only two different diagonals (be–cd–dc–ea–a/e–bd–cc–db–
ea). A possible reason for this selection can be found in the table itself, which
contains a number of symmetrical arrangements around cc, the central and
densest block. Ex. 10 shows numerous isomorphs (exact or by transposition)
among the blocks, resulting in a dense network of symmetrical relations within
the table: for example, bb, bd, dd, and db; bc, cd, dc, and cb; bd, be, de, ed,
and eb; c, ce, and ec; and finally, e and ee as well as a and ea. Considering the
harmonic detail, we observe that blocks eb and be consist of interval class 3
(ic3); in blocks db and bd, these ic3s are duplicated. In blocks dc and cd, the
duplicated ic3s appear simultaneously in two chromatic transpositions, and the
central block cc also consists of three chromatic transpositions of ic3s. Moreover,

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
374 MARINA SUDO

Fig. 5 Two stages (5a and 5b) in the construction of diagonal commentaries on block
c, used for the piano part underneath the quotation from ‘Improvisation III’

(a)

(b)

a and ea, with density 3, are composed of precisely these two intervals, ic3 and
one chromatic relation. On the other hand, the blocks ba, ca and da in the first
column do not reflect such exact symmetry around cc; ba and da, which are
simply identical by transposition, appear as incomplete variants of the type cd,
and ca is an incomplete variant of cc. All the blocks of the table are thus irrigated
by a single harmonic colouration (ic3), eventually combined with chromaticism,
which is simply a stylistic constant in post-Webernian music. Out of this network
of isomorphs Boulez finally selects the two diagonals be–cd–dc–ea–a and e–bd–
cc–db–ea, which are the two possibilities with the highest level of isomorphic

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 375

Ex. 10 Network of isomorphs in column Επ

relations among the blocks. In other words, the numerous recapitulated variants
of similar harmonies allow the composer to concentrate precisely on a sonic
limitation within the table in order to develop the structural stratum of this
section.
This particular selection of blocks is not an arbitrary decision on the part
of the composer, a fact that will become apparent when we see the flexible
manner in which Boulez realises these materials. Ex. 11 shows a passage for
piano including blocks be, cd, dc, eb, and a (first diagonal commentary; see
again Fig. 5b) over which a quotation from ‘Improvisation III’ is inserted.
The series of blocks in this commentary contains two groups of identical
chords (dc/cd and be/eb); the first set of identical blocks dc/cd, positioned on
either side of the sustained ec, is realised acoustically in a similar way, that

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
376 MARINA SUDO

Ex. 11 Transcript of excerpts from the pencil manuscript of ‘Don’ (PSS, Pierre
Boulez Collection, Mappe G, Dossier 3b, 7), omitting minor details. The boxes
marking the identical blocks in the piano part and the italicised letters above are the
author’s

is, with fixed registration, short duration and an ascending arpeggio. On the
other hand, the second set, be/eb, is treated as a unique block and combined
with the external notes E and Eb. It is a result of an additional operation of
‘chromatic complementarity’ (complémentaire chromatique), which is used to fill
the chromatic space between the original D and F. Three of the four resulting
notes are shared with block a (from the same diagonal commentary) and
placed in the same registration. The harmonic elements of this piano passage
are replicated in a quoted excerpt from the end of ‘Improvisation III’ played
by mandolin and guitar, but its rhythmic elements derive analogically from
‘Improvisation III’.19 In other words, whilst this section is based on the materials
from the table, it also involves a pre-defined compositional decision inspired by
the original text of ‘Improvisation III’.
Subsequent passages use the other sets shown in Fig. 5b, and in these too
a number of perceptibly similar sounds appear among the quotations from
‘Improvisation III’, with similar articulation of identical blocks and, for the
most part, fixed tessitura (several notes having two registrational possibilities).
This suggests that the composer intentionally selects these limited sets of
materials by diagonal commentary in order to lend coherence to the sound.
Interestingly, on one page of very early sketches, Boulez writes a plan for ‘right
diagonal commentaries’ (commentaires diagonales droite) for section b,20 which

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 377

was, however, unrealised. Although there is no way to know the exact reason
for this, it is conceivable that the more homogenous nature of the blocks in
the left diagonal lines, that is, the diagonals from top right to bottom left,
seemed to Boulez more appropriate when realising his sound image for the
quotation from ‘Improvisation III’. A further examination of the sketch material
supports this hypothesis; Boulez usually realises the outlined materials after
careful elaboration of the distribution of intervals inside blocks or in combination
with other parameters. Nevertheless, for this section of diagonal commentaries,
there are no such intermediate sketches, and it seems rather that he went directly
from the first sketch to drafting. Boulez’s ability to design the final sound
without elaborating the sketch in further detail can presumably be credited to his
meticulous preparation: the left diagonal materials were prepared well enough
for the composer to incorporate them into the sound image in his mind, even
though that image was not yet complete in every detail. When Boulez quotes
all three ‘Improvisations’ in section d of ‘Don’, he employs, as before, a limited
series of materials in column Βη, obtained by left diagonal commentaries that
contain a greater number of identical blocks. Diagonal commentaries thus serve
as a gesture for recalling the previous occurrence of the three ‘Improvisations sur
Mallarmé’ in ‘Don’.
Such a compositional act shows Boulez reflecting on two questions in parallel:
how to use the material according to his conceptual planning, and how to
arrange it for the sound image in his mind. In contemplating the acoustic
realisation (beyond the logical formalisation), the composer never hesitates
to make adjustments to or substitutions for rationally outlined materials.
This systematic approach to the multiplication table suggests that for him
it was both a starting point for further consideration of its potential and
an indispensable process for logically concretising his concept. This aspect
of Boulez’s composition is also evident in a final example, taken from the
introduction of ‘Don’.
In order to construct the opening (section a), the composer not only carefully
chooses the sound blocks from the multiplication table, but also uses a device
for modifying their original form. As already described, the beginning of ‘Don’
is paired with the end of ‘Tombeau’, the same chord being used both to open
and to close the cycle. Ex. 12 illustrates the complementary relationship between
the end of ‘Tombeau’ and the beginning of ‘Don’. Just before the final chord of
‘Tombeau’, the instrumental parts play the first two blocks in ΜυΑλ (a + b; see
again Ex. 1) between the two syllables of the concluding vocal line, ‘la mort’
(F#–D). The complementary three blocks c, d and e in ΜυΑλ are employed
in ‘Don’, played simultaneously with the opening chord but perceived only
in retrospect as a sustained echo. On the other hand, the two blocks used in
‘Tombeau’ (a + b) are also employed in ‘Don’ for the vocal line with the first
line of Mallarmé’s ‘Don du poème’ (1865): ‘Je t’apporte l’enfant d’une nuit
d’Idumée!’ (‘I bring you this child of an Idumaean night!’).21 It is with this vocal
phrase that Boulez synchronises two parallel instrumental lines in piano and

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
378 MARINA SUDO

Ex. 12 Reduction of the end of ‘Tombeau’ and the beginning of ‘Don’ indicating the
blocks employed from the multiplication table (see. Ex. 1)

vibraphone. The pitches of these two instruments are conceptually derived from
the three blocks c, d and e in ΜυΑλ, which complement the vocal part. However,
Boulez does not perform the operation of commentaries from these blocks in in
ΜυΑλ, substituting the blocks c, d and e from a different column, ΜυΕπ. On one
page of sketches, Boulez notes an ‘excuse’ for using this alternative: that the last
three blocks in ΜυΕπ are ‘equal apart from B natural’ to those in ΜυΑλ.22 This
is a natural consequence of the multiplication table (see again Ex. 1): the five
complexes in each of the five columns of the first line (Μυ) in the multiplication
table differ from one another only in different partitionings of the basic twelve-
note row. In observing the three blocks c, d and e in ΜυΑλ and ΜυΕπ, it can
be seen that the total number of notes is six (2/1/3) and seven (4/2/1), and
their difference lies only in B. Thus, exploring column ΜυΕπ, Boulez uses the
vertical commentary to the fifth block (e) vertically back and forth (described
in the sketch as ‘aller/retour’) for piano, and to the fourth block (d) diagonally
and then downwards (‘diagonal/aller’) for vibraphone (Ex. 13). Even though the
commentary to the third block (c) is planned in the sketch, it is not realised. In
the commentary to d, furthermore, Boulez substitutes dc for bc, conceivably to
avoid having two identical blocks, bc and cb, appear in succession.
For the two commentaries used for the vocal line ‘Je t’apporte l’enfant d’une
nuit d’Idumée’ (in the piano and the vibraphone – the lowest staves in Ex. 12
– and enlarged on in the orchestra), Boulez adjusts the number of notes they
contain in order to match the twelve syllables of the text. First of all, in the

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 379

Ex. 13 Vertical commentary on block e for piano and diagonal/downward

commentary on block d for vibraphone

vertical commentary to e, the number of notes is increased by creating a structure

centred on be, the blocks de and ee and two notes from ce (Eb and C#, excluding E
and C) being situated either side (compare Ex. 12, piano, and Ex. 13, right-hand
boxed blocks). Secondly, the four blocks of the commentary to d arise through
an operation of filtering to reduce the number of notes from twenty to twelve;
the three blocks c, d and e in ΜυΑλ (E–F#–G–G#–A–C), which were used for the
echo, are employed as a device for this filtering23 (Ex. 14).24 Moreover, the range
of filtering gradually changes so that it does not excessively reduce the number of
available notes. In the first step dc is filtered by all six notes, removing the pitch
En; in the second the pitch G is removed from the echo filter so that only G# and
E are filtered from cb; in the third both G and A are removed from the filter, so
that only the pitch F# is filtered from ea. Even though Boulez draws the outline

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
380 MARINA SUDO

Ex. 14 Derivation process of the filtered notes for vibraphone (see Ex. 13)

of the introduction in the range of commentaries, he modifies it according to his

needs. Nonetheless, even the modifications are based on the table.
The analytical evidence in ‘Don’ demonstrates Boulez’s highly elastic manner
of treating his materials. The initial idea derives from a logical concept and is
constantly expanded and adjusted by various interpretations and explorations of
prepared materials. The sound blocks in the multiplication table are an extensive
‘collection of possibilities’ (Boulez 1963, p. 36) that work as a multifunctional
tool for concretising his concept of ‘Don’. Although the materials are certainly
fixed in register and ordered in the five columns and lines in the table, these
orderings are often modified and reconstructed through his explorations of new
potentialities of the material. The back-and-forth interaction between the stages
of planning and realisation, as shown by some sketches, suggests that Boulez
finally achieved a fully creative style of serial composition by the end of the
1950s. The original series no longer confine the composer by their restrictive
laws; rather, they serve as a fountain of creative ideas, stimulating his imagination
through their manipulation.

NOTES
I would like to thank Pascal Decroupet for his support and feedback
during the preparation of the present article. I also express my gratitude

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 381

to Robert Piencikowski for his valued advice during the course of my

research. Finally, I wish to thank the staff of the Paul Sacher Foundation,
especially Michèle Noirjean. Permissions to use my transcriptions of the
sketch manuscripts and the letters were obtained from the Paul Sacher
Foundation and the heirs of Pierre Boulez.
1. For the context and chronology of this work, see Albèra (2003a and
2003b), Bassetto (2003), Decroupet (2003), Durand (2003), O’Hagan
(2003), Piencikowski (2003) and the introduction to Piencikowski
(2010).
2. ‘J’ai agrandi le plan de mon Mallarmé, jusqu’à en faire une œuvre longue
qui me liquidera Mallarmé pour que temps jusqu’à ce que je reprenne
goût au Coup de Dés – qui sera essai d’une synthèse totale […] Je suis,
en ce moment, en train d’élargir le Tombeau (si j’ose cette macabre
comparaison) […] au point d’en faire un mausolée!! Il y a une dialectique
entre registres fixes et mobiles que j’espère réussir.’ Letter from Boulez to
Stockhausen, [20] December 1959, Paul Sacher Stiftung (PSS). Excerpt
translated in Piencikowski (2010), p. 33.
3. The genesis of the piano version of ‘Don’ is an example of Boulez’s
self-borrowings. ‘Don’ for piano uses elements from Strophes (1957),
which is originally derived from L’Orestie (1955); the composer then
reused materials from ‘Don’ for the composition of Éclat (1965). The
musical connection between these works is more throughly examined in
Piencikowski (1993), Edwards (2006) and Salem (2014), Ch. 6.
4. The complete context of Boulez’s letter is: ‘Je suis toujours dans les
Mallarmé./1. Je réalise un ensemble de versions fixes [de Tombeau]/2.
Je fais vraiment le Don/Je recompose l’improvisation I./Encore beaucoup
de travail. Ce sera finalement une œuvre qui va durer largement plus
d’une heure’ (‘I’m still in the Mallarmé./1. I realise a set of fixed versions
[of Tombeau]/2. I’m really making Don./I recompose the Improvisation
I./Lots of work remaining. It will finally be a work lasting for much more
than an hour.’) Letter from Boulez to Stockhausen, [late November] 1959
(PSS).
5. The re-orchestration of ‘Première Improvisation sur Mallarmé’ was also
taking place during the same period.
6. PSS, Pierre Boulez Collection, Mappe G, Dossier 3b, 4. Boulez’s reference
to James Joyce’s Finnegans Wake has a symbolic meaning in this sketch. In
Finnegans Wake the story starts mid-sentence, the first part being found at
the end of the book. The result is a closed circle analogous to that in Pli
selon pli.

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
382 MARINA SUDO

7. Important contributions focusing on the theoretical aspect include

Koblyakov (1977 and 1990), Heinemann (1998) and Scotto (2014). In
parallel to this, the musical potential of this technique has been widely
studied since Boulez’s sketch material became accessible; see e.g. O’Hagan
(1997), Pereira de Tugny (1999), Mosch (2004), Decroupet (2005),
Decroupet and Leleu (2006), Strinz (2006) and Losada (2008 and 2014).
The composer’s comments on this technique are also found in a limited
number of texts, e.g. ‘Possibly …’ (Boulez 1991, pp. 111–40; originally
published in 1952 in French as ‘Éventuellement …’ and thereafter inserted
in Boulez 1966, pp. 147–82) and in translation in Boulez on Music Today
(Boulez 1971).
8. Losada’s emphasis on the partial ordering in Boulez’s multiplicative
operation contrasts with both Heinemann’s and Scotto’s approaches,
which treat the multiplication as the products of unordered pitch-class
sets.
9. The register of the transposed multiplier in the table is not necessarily in
the theoretically ‘correct’ place; in some cases the anchor notes are placed
not in the lowest position but above some other pitches (e.g. cc, bc in
Ex. 3). However, Losada’s observation proves the structural importance of
the subseries of anchor notes, showing brackets around the anchor notes
in the sketch from Structure II (Losada 2014, pp. 97–8).
10. In the sketches of ‘Don’, Boulez notes the word commentaire as an agent by
which he enlarges the interpretation of basic principles. In this article, the
author uses the English translation, ‘commentary’.
11. A sketch of ‘Don’, PSS, Pierre Boulez Collection, Mappe G, Dossier 3b,
3.
12. For more detail of section c of ‘Don’, see Sudo (2016).
13. ‘[R]épartis comme l’original en 6 valeurs longues’, PSS, Pierre Boulez
Collection, Mappe G, Dossier 3b, 3.
14. A possible reconstruction of Boulez’s procedure reads as follows: whilst cc
and dd are selected as part of the diagonal, ea and b cannot be deduced by
this principle. To complete the diagonal, Boulez could have chosen a, but
bb is a part of the horizontal commentary in the former subsection B. It
results in the selection of eb as the only possibility in the second column;
however, since this block is identical to the structural starting point be, it
could not serve as commentary, so the selection of b and ea is, in the end,
the only solution that satisfies the rule of equilibrated distribution among
the lines and columns.

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 383

15. Boulez uses the term ‘description’ in a technical sense, meaning the
melodic ‘scanning’ of a block (see Boulez 1963, p. 126, and 2005, pp.
656–7). The idea of scanning (balayer) is also expressed by the composer
in Boulez (2005, p. 309).
16. Substantial investigations on ‘Improvisations sur Mallarmé’ include
Stoïanova (1973), Breatnach (1996), Brunner (1996a and 1996b),
Guldbrandsen (1997 and 2016) and Salem (2014, Ch. 8).
17. This is only the case in the 1962 version (UE 13614); see pp. 7–14 and 24–
7. In the revised score (UE 31538), the quoted fragments are thoroughly
integrated (see rehearsal numbers 6–19 and 35–7).
18. Compare p. 11 in the published score of ‘Don’ (1962 version) and
rehearsal number 8 in the revised 1982 version of ‘Improvisation III’ (UE
19521).
19. This passage is found at the very end of the 1959 unpublished version
of ‘Improvisation III’, which Boulez referred to when composing ‘Don’.
In the revised 1982 version, this mandolin and guitar part is followed by
another instrumental part (see pp. 89–90).
20. PSS, Pierre Boulez Collection, Mappe G, Dossier 3b. 2.
21. The English translation is from Mallarmé (2006), p. 27.
22. PSS, Pierre Boulez Collection, Mappe G, Dossier 3b, 3.
23. The term ‘filtering’ comes from the French filtrage, used by Decroupet
(2012, pp. 240–5) to illustrate the deductive system of the series in
‘Bourreaux de solitude’ from Le Marteau sans maître.
24. The echo sound by the three blocks is also used as a definition of ambitus
for all of sections a and b.

REFERENCES
Albèra, Philippe, 2003a: ‘Entretien avec Pierre Boulez’, in Philippe Albèra (ed.),
Pli selon pli de Boulez: Entretien et études (Geneva: Contrechamps), pp. 7–24.
, 2003b: ‘… l’éruptif multiple sursautement de la clarté …’, in
Philippe Albèra (ed.), Pli selon pli de Boulez: Entretien et études (Geneva:
Contrechamps), pp. 59–82.
Bassetto, Luisa, 2003: ‘Orient-Accident ? Pli selon pli, ou l’“eurexcentrisme”
selon Boulez’, in Philippe Albèra (ed.), Pli selon pli de Boulez: Entretien et
études (Geneva: Contrechamps), pp. 37–44.
Boulez, Pierre, 1963: Penser la musique aujourd’hui (Geneva: Gonthier);
published in English as Boulez on Music Today, trans. Susan Bradshaw and
Richard Rodney Bennett (London: Faber & Faber, 1971).

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
384 MARINA SUDO

, 1966: Relevés d’apprenti, ed. Paul Thévenin (Paris: Seuil); published in

English as Stocktakings from an Apprenticeship, trans. Stephen Walsh (Oxford:
Clarendon Press, 1991).
, 2005: Point de repère III: Leçons de musique, ed. Jean-Jacques Nattiez
(Paris: Bourgois).
Breatnach, Mary, 1996: Boulez and Mallarmé: a Study in Poetic Influence
(Aldershot: Scolar Press).
Brunner, Raphaël, 1996a: ‘“L’Improvisation III sur Mallarmé” de Pierre
Boulez: Éléments pour une mise en perspective’, Dissonanz/Dissonance, 50,
pp. 4–14.
, 1996b: ‘Quelques notes sur l’Improvisation III sur Mallarmé de Pierre
Boulez’, Mitteilungen der Paul Sacher Stiftung, 9, pp. 12–6.
Decroupet, Pascal, 2003: ‘Comment Boulez pense sa musique au début des
années soixante’, in Philippe Albèra (ed.), Pli selon pli de Boulez: Entretien et
études (Geneva: Contrechamps), pp. 49–58.
, 2012: ‘Le rôle des clés et algorithmes dans le décryptage analytique:
L’exemple des musiques sérielles de Pierre Boulez, Karlheinz Stockhausen
et Bernd Alois Zimmermann’, Société française de musicologie, 98/i, pp. 221–
46.
(ed.), 2005: Pierre Boulez, Le Marteau sans maître: Facsimile of the Draft
Store and the First Fair Copy of the Full Score (Mainz: Schott).
Decroupet, Pascal and Leleu, Jean-Louis, 2006: ‘“Penser sensiblement” la
musique: Production et description du matériau harmonique dans le
troisième mouvement du Marteau sans maître’, in Jean-Louis Leleu and
Pascal Decroupet (eds), Pierre Boulez: Techniques d’écritures et enjeux
esthétiques (Geneva: Contrechamps), pp. 177–215.
Durand, Pascal, 2003: ‘“Pli selon pli”, les modèles musicaux de Mallarmé’,
in Philippe Albèra (ed.), Pli selon pli de Boulez: Entretien et études (Geneva:
Contrechamps), pp. 83–100.
Edwards, Allan, 2006: ‘Éclat/Multiples et le problème de la forme musicale
dans les œuvres sérielles de Pierre Boulez’, in Jean-Louis Leleu and Pascal
Decroupet (eds), Pierre Boulez: Techniques d’écritures et enjeux esthétiques
(Geneva: Contrechamps), pp. 159–75.
Guldbrandsen, E. Erling, 1997: ‘New Light on Pierre Boulez and Postwar
Modernism: on the Composition of Improvisation I–III sur Mallarmé’,
in Søren M. Sørensen (ed.), In the Plural: Institutions, Pluralism, and
Critical Self-Awareness in Contemporary Music (Copenhagen: Department of
Musicology, University of Copenhagen), pp. 15–28.
, 2016: ‘Casting New Light on Boulezian Serialism: Unpredictability and
Free Choice in the Composition of Pli selon pli – Portrait de Mallarmé’,
in Edward Campbell and Peter O’Hagan (eds), Pierre Boulez Studies
(Cambridge: Cambridge University Press), pp. 193–220.
Heinemann, Stephen, 1998: ‘Pitch-Class Set Multiplication in Theory and
Practice’, Music Theory Spectrum, 20/i, pp. 72–96.

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
 MORE THAN MEETS THE EYE: BOULEZ’S ‘DON’ 385

Koblyakov, Lev, 1977: ‘P. Boulez Le Marteau sans maître: Analysis of Pitch
Structure’, Zeitschrift für Musiktheorie, 8/i, pp. 24–39.
, 1990: Pierre Boulez: a World of Harmony (New York: Harwood Academic).
Losada, Catherine, 2008: ‘Isography and Structure in the Music of Boulez’,
Journal of Mathematics and Music, 2/iii, pp. 135–55.
, 2014: ‘Complex Multiplication, Structure, and Process: Harmony and
Form in Boulez’s Structures II’, Music Theory Spectrum, 36/i, pp. 86–120.
Mallarmé, Stéphane, 2006: Collected Poems and Other Verse, trans. E. H.
Blackmore and A. M. Blackmore (Oxford and New York: Oxford University
Press).
Mosch, Ulrich, 2004: Musikalisches Hören serieller Musik (Saarbrücken: Pfau).
O’Hagan, Peter, 1997: ‘“Sonate, que me veux-tu?”: an Investigation of
the Manuscript Sources in Relation to the Third Sonata’ (DPhil diss.,
University of Surrey).
, 1998: ‘“Trope” by Pierre Boulez’, Mitteilungen der Paul Sacher Stiftung,
11, pp. 29–35.
, 2003: ‘Pierre Boulez Pli selon pli’, in Philippe Albèra (ed.), Pli selon pli de
Boulez: Entretien et études (Geneva: Contrechamps), pp. 25–35.
Pereira de Tugny, Rosãngela, 1999: ‘Au commencement était l’ esquisse: à
propos de Constellation-Miroir de Pierre Boulez’, Musurgia 6/i, pp. 41–61.
Piencikowski, Robert, 1993: ‘“Assez lent, suspendu, comme imprévisible”:
Quelques aperçus sur les travaux d’approche d’Éclat’, Genesis: Revue
internationale de critique génétique, 4, pp. 51–67.
, 2003: ‘“Tombeau”, extrait de Pli selon pli de Pierre Boulez’, in
Philippe Albèra (ed.), Pli selon pli de Boulez: Entretien et études (Geneva:
Contrechamps), pp. 45–8.
(ed.), 2010: Pierre Boulez, Tombeau: Facsimiles of the Draft Score and the
First Fair Copy of the Full Score (Vienna: Universal Edition).
Salem, Joseph, 2014: ‘Boulez Revised: Compositional Process as Aesthetic
Critique in the Composer’s Formative Works’ (PhD diss., Yale University).
Scotto, Ciro, 2014: ‘Reexamining PC-Set Multiplication, Complex
Multiplication, and Transpositional Combination to Determine Their
Formal and Functional Equivalence’, Perspectives of New Music, 52/i, pp.
134–216.
Stoïanova, Ivanca, 1973: ‘Pli selon Pli: Portrait de Mallarmé’, Musique en jeu,
11, pp. 75–98.
Strinz, Werner, 2006: ‘Quelques observations sur des “objets retrouvées” dans
l’œuvre de Pierre Boulez’, in Jean-Louis Leleu and Pascal Decroupet
(eds), Pierre Boulez: Techniques d’écritures et enjeux esthétiques (Geneva:
Contrechamps), pp. 45–93.
Sudo, Marina, 2016: ‘De “Tombeau” à “Don”: Filiations par symétries voilées
dans Pli selon pli de Pierre Boulez’, Mitteilungen der Paul Sacher Stiftung, 29,
pp. 47–52.

Music Analysis, 39/iii (2020) © 2021 The Authors.

Music Analysis © 2021 John Wiley & Sons Ltd
386 MARINA SUDO

NOTE ON THE CONTRIBUTOR

MARINA SUDO completed her bachelor’s and master’s degrees in musicology at
the Tokyo University of the Arts. In 2015, supported by a stipend from the Paul
Sacher Foundation in Basel, she conducted a study of the manuscript sketches
and scores held in their Pierre Boulez collection. She is currently a doctoral
student in musicology at the University of Leuven, where she is exploring the
constructive potential of ‘noise’ in contemporary practice through the analysis
of works by Iannis Xenakis, Helmut Lachenmann, Peter Ablinger and Merzbow
(Masami Akita). Her research focuses on a new form of musical investigation
that combines score-oriented and sound-based analyses.

ABSTRACT
Based on a study of the sketch material held at the Paul Sacher Foundation,
this paper examines the derivation of materials in the 1962 orchestral version of
‘Don’, the first movement of Pli selon pli: Portrait de Mallarmé by Pierre Boulez.
Although the composer starts with a prepared serial multiplication table, he
exhibits a flexible and imaginative treatment of material between conceptual
planning, conceived initially in the relationship with the last movement,
‘Tombeau’, and its mise en place. Some sketches suggest that Boulez carefully
selects the material according to the sound image in his mind. The interaction
between planning and sound realisation shows, furthermore, that he did not view
the series as a restriction but rather as a fountain of creative inspiration.

© 2021 The Authors. Music Analysis, 39/iii (2020)

Music Analysis © 2021 John Wiley & Sons Ltd
Copyright of Music Analysis is the property of Wiley-Blackwell and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for
individual use.