Professional Documents
Culture Documents
joseph salem
Pierre Boulez composed several related pieces during the latter half of the 1950s. These pieces all
share both serial structures and wholesale notational borrowings. Importantly, when Boulez reused
notated material, his focus shifted from executing organizational processes to nuancing the contour,
Keywords: Pierre Boulez, serialism, revision, improvisation, Pli selon pli, Strophes, L’Orestie.
I
n the following, I guide the reader through a number of of these changes extend to Boulez’s later efforts as a composer,
sketches to demonstrate the prominent role of revision in including his continued revisions to Pli selon pli decades after
a specific selection of Pierre Boulez’s works from the late its premiere.
1950s.1 Virtually all of these sketches relate to what I have de- For at least the last thirty years, scholars have used the ex-
scribed elsewhere as the L’Orestie genus, named after Boulez’s pertly catalogued collection of Boulez manuscripts at the Paul
withdrawn 1955 theatrical work for the Madeleine Renaud Sacher Stiftung in Basel, Switzerland (henceforth PSS) to
and Jean-Louis Barrault theater company.2 My primary goal is demonstrate the inter-workings of Boulez’s compositional pro-
to demonstrate overt, notated connections among several com- cess by exploring his sketches.4 While the nature of these stud-
positions separated by just a few years, spanning from L’Orestie ies varies wildly, virtually all of them share a tacit
in 1955 to Strophes in 1957 to Pli selon pli (itself spanning acknowledgment that most of Boulez’s works are interrelated,
from 1957 to 1962 and beyond), several of which have never either through the use of similar row materials, shared sketches
been discussed previously. A secondary goal is to characterize a or matrixes, or the reuse of actual notated material. Indeed, the
particular moment in Boulez’s compositional development, vast majority of Boulez’s compositional efforts could be de-
one that may represent the most significant change in his crea- scribed as “cumulative,” such that the musical ideas or sketches
tive process during his formative years as a composer.3 of one work are explored in a “first” composition, only to be
My argument is straightforward, if also speculative: Boulez re-explored in additional compositions years or decades later.
conceived certain musical ideas while composing L’Orestie (in- Although uncommon in the prevailing discourse, I describe
cluding ideas shared with other works from the same year); he these and other practices as “revisions,” and I claim that this
revised some of these ideas for use in Strophes, and then he re- practice is a fundamental component of Boulez’s mature com-
vised these ideas again for use in Pli selon pli. Throughout, positional process.5 My reasons for this go beyond the termi-
Boulez’s sketches provide evidence of his creative process be- nological: I aim to shift the conversation from observing
yond his serial matrixes. Thus, the relationships between these material changes as a teleological path toward a single, desired
works provide evidence of Boulez’s changing priorities as a outcome, to questioning the conceptual role of revision in
composer during the 1950s. Furthermore, the stylistic effects Boulez’s compositional process, including related questions of
intent, authorship, and ontological stability. Hence, unlike
1 All references to specific sketches follow this format: PSS (Paul Sacher Brice Tissier’s recent spatial summation of the interrelatedness
Stiftung), BS (Boulez Sammlung)/ x (Mappe), y (Dossiers), or PSS, BS/
G,1c,2c, for example.
2 Salem (2014). This research is the result of several months of primary- 4 Representative examples include selections of the recent Pierre Boulez
source investigations at the Paul Sacher Stiftung in Basel, Switzerland and Studies, Campbell and O’Hagan, eds. (2016), virtually all of Leleu and
was funded in large part by that organization and the US Fulbright Decroupet (2006), the extended studies of Decroupet (1994), Dal Molin
Organization. I am grateful to both institutions for their generous (2008), Losada (2014), O’Hagan (1997, 2007, 2017), Salem (2014, 2015,
support. 2017), Strinz (1999, 2003), and Tissier (2012), as well as shorter studies
3 The following studies consistently highlight the powerful mixture of such as those by Edwards (1993, 2006) and Piencikowski (1993, 2013b).
ideas—compositional and otherwise—in works from this period, espe- Older studies correspond to either a private relationship with Boulez
cially in Boulez’s overt reference to non-musical literary and artistic influ- (Bennett [1986], Bradshaw [1986], Piencikowski [1985]) or the opening
ences of these and related works: Adamowicz (2015), Bradshaw (1986, of the archive (Bonnet [1987], Edwards [1989], Bradshaw [1996]).
1996), Breatnach (1996), Campbell (2010), Edwards (1989, 2006), 5 Piencikowski (2016) uses the word “revision” in ways that parallel my pre-
Goldman (2010, 2011), Guldbrandsen (1997b, 2016), Nattiez (2004), vious research with him, although in maintaining his usual virtuosic word-
O’Hagan (1997, 2017), Piencikowski (1985, 1991, 1993, 2002, 2010a, b, play, Piencikowski ultimately suggests “perpetual alternation,” “re-
2013a), Stacey (1987), Leleu and Decroupet, eds. (2006), Tissier (2012), revision,” and “spiral” as possible replacements for mere revision (all his
and, most recently, Campbell and O’Hagan, eds. (2016). references are to outside texts with deeper reflexive meanings).
244
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 245
L’Orestie Genus
typical Boulez’s works, I categorize the composer’s various clarifies that a certain fundamental sketch (including specific
forms of self-borrowing as manifestations of revision more pitch material) was used as the basis for multiple works. In
specifically, with individual works or families of works repre- all cases, it means that the works in the L’Orestie genus
senting different types or applications of revision in his com- share actual notes and structures with one another that
positional process.6 In effect, Boulez is not merely revising evolve over time. This is in contrast to works that could re-
materials, he is reusing ideas in order to refine, expand, and/or late merely by title (the Mallarme Improvisations) and typi-
develop them as a catalyst for creative development, often us- cal rewrites to parts and expression markings of a specific
ing specific types of revision depending on changing contexts piece (Don of 1962 and the revised Don of 1989), although
or needs. This hypothesis marks his reuse of material with in- some members of the genus are also occasionally related in
tent, allowing one to reconsider the coherence of Boulez’s own these ways, too. Finally, while the L’Orestie genus includes
aesthetics by highlighting his willingness to change, adapt, and far more works than the four I discuss below, other works
alter previous works. It also applies pressure against his use of will only be occasionally mentioned as I focus on the specific
organicist metaphors to stress the genetic relationship between ties between L’Orestie, Strophes, “Don” for piano, and vari-
part and whole, microcosm and macrocosm.7 ous elements of the Improvisations sur Mallarme from Pli
Boulez’s use of revision results in a multitude of composi- selon pli.9
tional families: here, I focus on only a few works to provide a
demonstration of the sheer variety of revisionist tactics used by DEFINING THE “DNA”
Boulez and how they may impact our view of him as a serial
composer.8 Drawn from the category of revisions I call Boulez’s own jargon is useful for defining what constitutes the
“genetic relations,” Example 1 (the L’Orestie genus) represents DNA of his works. Given his penchant for literature and—
a group of works that all share common “DNA.” In some especially during the later 1950s—the poetry of Stephane
cases, this means that actual notes and themes are borrowed Mallarme, Boulez opted to label a number of side sketches
from one work for use in another; in other cases, it merely “sonnets,” just as he labeled his resulting work for flute
Strophes.10 Together, these terms reify an amorphous set of
6 Tissier (2012), Salem (2014, 2016). Both studies were aided by the over-
sight of Robert Piencikowski. 9 Throughout this text, I will use < “Don” for piano > to designate the first
7 Boulez’s aesthetics are not an object; they are a living, breathing body of version of this work for piano and voice, composed and performed in
writings and thought. Nonetheless, certain themes show up consistently 1960, and < Don > to designate the later, quite different orchestral move-
from his earlier writings (Stocktakings, Orientations), through his later ones ment featured in the premiere of the Pli selon pli in 1962 and revised for
(the Collège de France lectures), with one primary theme being the rela- subsequent performances of that work.
tion of part to whole, microcosm to macrocosm. This corresponds to 10 These sketches are discussed at length by Guldbrandsen (1997b, 2011,
Boulez’s analyses of Webern’s music. See Goldman (2011) for a broad 2016), as well as in his dissertation (1997a); see also Adamowicz (2015)
consideration of this perspective. and Salem (2014). While much ink has been spilled on Boulez and
8 I discuss several families of compositions and related types of revision at Mallarme, I trace a particular moment of Boulez’s fascination with the
length in Salem (2014, 47–71). poet in the year 1957, as relayed in his private correspondence with
246 music theory spectrum 41 (2019)
poetically inspired conceptual and material relations. Strophes significance and a procedural one, since the work relies in-
seems to have been commissioned as an ensemble piece by creasingly on the “sonnet” material toward its close.
Wolfgang Steinecke in late 1956 or early 1957. At least three Example 2 provides a schematic outline of Strophes based
commentators suggest the work was intended for flute and cham- on alphanumeric labels derived from L’Orestie; this table also
ber ensemble, and a letter from Wolfgang Steinecke implies as includes a schematic outline of yet another work, “Don” for pi-
much.11 However, extant sketches provide little indication that ano, relative to these same sources. The abbreviations provide
Boulez ever seriously attempted to complete the ensemble a clear, bird’s eye view of Strophes, including the cyclic permu-
parts—a few pages appear ambitious in some respects, but they tations of its first fifteen sections (beginning with strophe 1a),
demonstrate only the beginnings of any such compositional proj- as well as the different permutations of the extended coda (be-
worked hard to finalize the production by the following sum- L’Orestie stems from his experience at the theater, including
mer. While not remarkably successful (the piece received only the need to adjust his usual compositional methods to accom-
a handful of performances), it remains Boulez’s only attempt modate the demands of changing rehearsal schedules, dramatic
at a substantial dramatic work for the stage.17 cuts and additions, and the abilities of various singers. Boulez’s
L’Orestie is significant to my investigation for two reasons: sensitivity to these concerns undoubtedly expanded his reper-
first, a small portion of it forms the basis of Strophes and other toire of compositional processes to include a variety of new
subsequent works; and second, it features a host of composi- sketch procedures. Chief among these are the use of vocal par-
tional methods that challenge the very limits of how we define ticelle, block harmonizations, a shift from dodecaphonic to
Boulez’s brand of serialism.18 It is also significant that smaller pitch-class sets, and the beginnings of recurring me-
lodic signals.19 Indeed, Boulez even seems to have invented an
17 An incomplete version of the play was premiered at the Sixth Bordeaux
entirely new form of pitch matrix to allow for the easy distilla-
Festival, 26–28 May 1955. A complete version was performed in Paris on tion of non-dodecaphonic rows and sets, including the blocs
5 October 1955, and once again for a Paris revival (4 January 1962). See sonores (or chord-like stacks of pitches) used throughout the
Steinegger (2012, 147) and O’Hagan (2007). Campbell (2016) and Salem work.
(2014, 2018) also provide biographical context leading to the composition
of L’Orestie based on Boulez’s travels and correspondence.
18 “Serialism” remains difficult to define in the abstract or in relation to
Boulez’s long and varied oeuvre. Whittall (2008) provides both a pithy ab- 19 I provide details of these in Salem (2014, 2015); similar topics related to
stract definition and a subsequent book of specific examples. Longer con- this specific repertoire include O’Hagan (2007), Guldbrandsen (2016),
siderations relevant to this period include Grant (2001) and Iddon (2013). and Piencikowski (1993).
248 music theory spectrum 41 (2019)
Example 3 is a transcription of one of these matrixes.20 My matrix used in Le Marteau sans maı̂tre. In this new style of ma-
annotations highlight a few basic features of this matrix: (1) it is trix, the above features distort some of the traditional values of
not dodecaphonic in the traditional sense; (2) it features overt dodecaphonic composition, including the implicit goal of equal
pitch duplications; (3) it is divided into five vertical columns; pitch-class representation throughout the matrix. For example,
and 4) the bottom row is actually a repeat of the first, further while the matrix is driven by basic transposition, Boulez does
complicating its relationship to typical (dodecaphonic) matrixes. not transpose the row based on the first note of each series,
The significance of this matrix in Boulez’s musical develop- but by one of two other possible operations. One could shift
ment cannot be overstated. When perusing the Boulez the second note up and over to the next row’s first note, result-
Collection at the PSS from beginning to end, it represents a ing in many of the rows actually starting on the same note
clear shift from the “true” dodecaphonic rows used in virtually (three start on A[ or G], three on F\, two on B[). If one in-
every earlier work published by Boulez, including his Sonatine, stead shifts the last note up and over to the next penultimate
his early cantatas (Le Soleil des eaux, Le Visage nuptial), note, one can trace a Boulezian diagonal from the upper right
Polyphonie X, Structures, and even the advanced blocs sonores corner downward, this time with no chromatic duplications in
the operation (making this the more likely source operation,
even if the results are identical).21 Meanwhile, random col-
20 The original sketch appears in BS/G,1b,1 and has been transcribed in var-
umns are used to create vertical stacks of tones (blocs sonores),
ious publications, beginning with Piencikowski (1993). Boulez used other
pitch matrixes in L’Orestie, including one likely borrowed from an earlier
“lost” work (Symphonie mecanique, 1954) that probably became the basis 21 Piencikowski favors tracking the last note diagonally. In other works, such
for another later work (the Troisième sonate). Piencikowski (2016, 107) as Poesie pour pouvoir (1958) Boulez uses transpositional schemes based
also suggests that the lost Symphonie concertante (1947) may have been an on rotating intermediate notes, such as the second note of a series, which
additional source for these works and Doubles. creates similar diagonal relationships; this makes both readings tenable.
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 249
such as those found in the box crossing rows I, J, K, and A in beginning of strophe 1a from Strophes. The shared gestures
the bottom right corner of the matrix. Readers familiar with correspond to individual cells of row A of the matrix in
Boulez’s “multiplication” method—such as the one used for Example 3. There is no mistaking the duplicated pitch ma-
the composition of Marteau—will recognize how immensely terial, although it is also quite clear that Boulez reworks the
simplified this procedure is in the L’Orestie matrix for generat- material into new, unique gestures that highlight the techni-
ing similar chord-like structures.22 Piencikowski has previously cal, registral, and dynamic strengths of the flute. Examples
demonstrated how these blocs sonores were used for a handful 4(c–d) highlight similar relationships between L’Orestie and
of primarily harmonic moments in L’Orestie, as well as in strophe 1b. As above, Boulez continued to (mostly) respect
Strophes, “Don” for piano (1960) and Eclat (1965).23 Here, I the original contour of the gestures, but here he changed the
explain how similarly sourced material was used as melodic fod- order of their appearance. Any such examination of the
der in L’Orestie and Strophes, “ Don” for piano, and the three rhythms, gestural contour, and register of the two works, no
Improvisations sur Mallarme from Pli selon pli, all with an ap- matter how superficial, will likely support that these revi-
propriate emphasis on how this relates to Boulez’s developing sions resulted from a direct conversion from one score to
lyricism at this time. the other, without any need to reference the original pitch
matrix.25
Boulez varied the literalness of his transcriptions further as
GENETIC ENGINEERING
he composed Strophes, making it increasingly difficult to ac-
count for borrowings from L’Orestie in later sections.
A few specific passages from L’Orestie provide ample evidence
Furthermore, the borrowings change in a fundamental way be-
of how Boulez reuses musical material among these formative
ginning with strophe 1g. With this strophe, Boulez seems to
projects, although it is important to remember that because
have tired of using basic surface permutations, instigating the
Boulez used multiple and varied forms of borrowing between
search for a new process-based method. In place of the sponta-
these works, these passages represent examples, not archetypes.
neous transcriptions above, Boulez devised a new and distinct
Examples 4(a–d) compare L’Orestie, Les Choephores 4 (Choir
combinatory process for the second half of the work.26
C), and Strophes to demonstrate how the above matrix is first
However, the resulting process is still firmly based on the
converted into counterpoint for winds in L’Orestie before being
pitches and durations of the L’Orestie score rather than its or-
revised into melodic writing for solo flute in Strophes.24
ganizational matrix.
Example 4(a) includes writing for flute, cor anglais, and
Examples 5(a–c) demonstrate the “informal” formation of
harp from L’Orestie, while Example 4(b) displays the
this new process. First outlined by Boulez in 1963 and later
described in relation to Strophes by Erling Guldbrandsen, the
process at work compresses the polyphonic instrumentation of
O’Hagan highlights yet another function of these types of transpositional
schemes in relation to “spirals” in the piano sonatas (2017, 54–55).
22 The terms blocs sonores and “multiplication” come from Boulez and first 25 Although I prefer the word revision, Szendy (2008) provides an apt con-
appear in his writings in the early 1950s. Boulez’s multiplication method ception of transcription and its ability to provide a window into how
and more typical blocs sonores matrixes have been explored recently by others “hear” or “listen” to musical works that complements my descrip-
Losada (2008, 2014), Scotto (2014), and Salem (2018). tion of Boulez’s efforts during the compositional process.
23 Piencikowski (1993). In addition, Losada has begun investigating the use 26 It is possible that Boulez used a loosely standardized mode of transcription
of blocs sonores in Eclat as well (SMT Vancouver paper, 4 November based on the nomenclature/instrumentation of each strophe (1,2,3 versus
2016). a,b,c). My point is to highlight that changes to register, dynamics, and
24 I follow O’Hagan (2007) in my titles for this work. contour do not appear to have been preplanned or serialized.
250 music theory spectrum 41 (2019)
example 4(c). Another “ strophe” from L’Orestie, Les Choephores 4 (Choir C) (partial diplomatic transcription of PPS, BS/G, 1c, 2c)
example 4(d). Example 4(c) repurposed as “ strophe 1b” in Strophes (partial diplomatic transcription of PSS, BS/G, 2c)
L’Orestie into a single line of music.27 In short, this process L’Orestie to guide a new distribution of the same pitch classes
uses the original durations of individual, polyphonic notes in as recurring notes in a single melodic line. This is easiest to see
in the hard-edged rectangular boxes of Example 5: a retro-
grade reading of the pitch classes in Example 5(a) corresponds
27 See Boulez (1963 French, 1971 English), Examples 55–57, and to the pitch material in the Example 5(b), but the order, note
Guldbrandsen (1997b, 2011, 2016).
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 251
values, and attacks change from one to the next. Example 5(c) role of what I call Boulez’s first “unnamed sonnet,” which
highlights how these rewrites were then used for the ornamen- appears as a side sketch in the early generative materials for
tal inserts in strophe 1g (the squared boxes relate to Strophes and is the first result of the serial process described
Example 5[b] box F while the round corners relate to box E). above.29 Boulez converted just one more unnamed sonnet in a
In effect, Boulez turns the three-voice texture of L’Orestie, re- similar fashion before formalizing this procedure to create the
hearsal 14, into a series of melodic gestures. Since the details of four labeled sonnets.30 Example 3 places each of these by
this process are quite complex and have been elaborated on else- name among the completed Strophes score as unnamed sonnets
where, I will not dwell on them here; suffice it to say, the con- 1 and 2, and “named” sonnets AB, BC, AC, and ABC, which
version from polyphony to melody follows a rule-based serial are the same labels used by Boulez and others for these previ-
system. Nonetheless, other non-trivial aspects of the notation ously acknowledged and specifically “named” sketches.
that affect musical gestures, such as dynamics, grouping, slurs,
and register, were adjusted by the composer after the fact.28 29 The sketch is catalogued as PSS BS/G, 2a. It is not labeled and is situated
Despite previous research into this process, Examples among other side sketches.
5(a–c) are diplomatic transcriptions all of which are shown 30 The named sonnets appear in folder G, 2b, 1–3; it is likely that the con-
here for the first time (to my knowledge). They highlight the version of the second “unnamed sonnet” was undertaken directly in the
score without a separate side sketch, probably because the original score
features homophonic simultaneities that were easier to convert due to their
28 See Guldbrandsen (1997b, 2011, 2016). durational equivalences.
252 music theory spectrum 41 (2019)
Sonnets AB–ABC are named for how they combine and per- premeditated, that his serial designs organically integrate local
mute different parts of the same three-voice texture. For exam- and large-scale formal attributes prior to notating the work, and
ple, when Boulez formalized the sonnet-making process, he that the serial organization of his works often left him with little
selected a new passage of three-voice music (now L’Orestie, re- control over the notated results of those same serial processes.
hearsal 16), but opted to label the three staves as A (flute), B Looking more closely at Example 3 helps to put the various
(cor anglais), and C (harp); he then combined these using the “sonnets” in relief against the composition of Strophes more
polyphony-to-melody process above, first using only flute and broadly. I mentioned that Strophes was commissioned, that it
cor anglais, then cor anglais and harp, and so on, to create son- reused recently composed material, and that it remains incom-
nets AB, BC, AC, and ABC. Boulez even included a neat lit- plete: while the flute part exists as a performance score, the
tle graphic in the score of L’Orestie, reproduced here as original idea of the work was likely more in the style of the
Example 6—no doubt a thoughtful consideration for archiv- later Domaines (for clarinet and ensemble). That time pressures
ists, allowing us to imagine and confirm the network of musi- also played a role in these outcomes is supported by the fact
cal connections.31 The named sonnets are also the only ones that—although Boulez often began the composition of his
with added annotations for the movements of Pli selon pli works with prose outlines that list the musical features or tech-
(discussed below).32 niques to be exploited, the number of sections or parts in the
The formalizing of the process is a genuinely astounding form, and/or other details related to dynamics, register, etc.—
fact. It implies that Boulez actually shifted from putting notes no clear prose outlines appear in the early sketches for
on a score backward to formalizing a new serial process while Strophes.33 Similarly, Strophes has no row sheets not shared
in the process of composing; and that he accomplished this while with L’Orestie: while there is an elaborate matrix of blocs
still achieving the prescribed goal of creating a specific serialized sonores, I find no obvious relation between this sheet and the
outcome that would result in the desired endpoint for a work solo flute part.34 Together, these details suggest a change in
that was already mostly notated. This order of events is contra- behavior for the composer. I speculate that Boulez rushed to
dictory to the usual impressions of Boulez’s compositional pro- complete the work after abandoning the overly complicated
cess, including that his formal designs were mostly idea of drafting ensemble parts, even to the point of skipping
the steps related to planning and organizing the (alternative)
31 Graphics like this pop up in various places in ways that immediately direct work for solo flute. Then, he borrowed materials to provide
the attention of the scholar to see certain possibilities over others, with
two notable ones appearing in a later number for L’Orestie and in the early 33 Comparatively rough sketches exist for potential movements and instru-
materials for Don for orchestra. I critique this aspect of sketch studies in mentation for the hypothetical ensemble work, but these have little rela-
Salem (2016). tion to the work for solo flute.
32 I provide transcriptions of all the sonnets and the original passages from 34 This sheet could resemble an attempt at starting the ensemble parts for
L’Orestie in Salem (2014). Strophes; see note 12 above.
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 253
the pitch and phrase structure of the work. Ultimately, he har- way prompts speculation on a slow, deepening focus on serial
nessed a new serial process only when inspiration finally hit procedures. Perhaps it was only after Boulez became increas-
midway through the project; he then used this new method as ingly absorbed in the new work for flute that he became more
the basis for an expansive coda and, ultimately, a new set of motivated to create an original serial process for this work,
works based on the conceptual cross-domain mapping of his resulting in a rather unique and disproportionate coda. These
“sonnets” to his Mallarme Improvisations.35 speculations on the gestation of Strophes highlight a distinct
This reading is supported by an additional fact: Boulez ear- and important trajectory for the work: that of a through-
marked the structure for Strophes in the actual manuscript for composed piece, in which aspects of the actual compositional
L’Orestie, treating the theatrical score itself as the basis of the process acted as feedback to influence the design, structure,
new design. Indeed, one version of the L’Orestie score contains and execution of the final product (this being in contrast to a
annotations that obviously refer to the “strophes” nomencla- more linear, spatial, or “organic” design). While these types of
ture.36 These annotations change between strophe 2d and 3d, speculations may seem intuitive to any working composer,
just as Boulez begins to incorporate the new serial process used having evidence of how such a feedback mechanism may have
in the later strophes and coda. Presenting the material in this functioned in Boulez’s working process provides unique
insights into his creative method, some of which may contra-
dict the “rhetoric of autonomy” harnessed by Boulez and so
35 Piencikowski (2016) outlines specific social circumstances and their rela-
tion to Boulez’s compositional choices and practices connected to chang-
many other contemporary composers.37
ing rehearsal practices with large orchestral groups during later decades.
36 As suggested by Example 3, these annotations were determined as much
by the phrasing of the score as by changes in instrumentation; similarly, 37 Wilson (2004) is now the classic argument for a skeptical approach to
new tempo markings correspond to the strophe assignments. how composers shape their own narratives over time, including elements
254 music theory spectrum 41 (2019)
[red]
[green]
abc
A
A b
B
In sum, changes in the conversion of material from For example, simpler, additive formal designs provided
L’Orestie to Strophes highlight changes to Boulez’s composi- Boulez with the ability to expand from within or without by
tional process more broadly. These are in contrast to the deep, making passages longer or by adding successive movements
thoroughly hierarchical organization that appears in the first without having to rewrite or reconceive the serial organization
movement of Structures, Book II (finished in 1956) and the of his form. Altering the structural coherence of Structures,
Troisième sonate, both of which occupied Boulez around the Book 2, Chapter 1, while maintaining its “organic” structural
same time. Individual movements from both works have been properties, seems like it would be a challenge after reading
extensively and convincingly analyzed by scholars to demon- Losada’s excellent work on the subject.40 It also seems like a
strate the abundance of macro- to micro-organizational con- wholly different kind of project than Boulez’s later expansions
nections.38 My sense, however, is that this type of deep, of Derive or Repons, or his other works based on the “Sacher”
meticulous serial planning of large-scale structures was taken or Originel series (or even his Structures, Book 2, Chapter 2,
to its limits during the late 1950s. Later works still feature composed roughly five years after the first movement and fea-
deep points of structural interconnection between micro- and turing a modular construction).41 Note also how Boulez has
macro-structural organization, but the ways and methods these expanded, contracted, and rewritten select parts, movements,
relationships are executed appear to become more modular and
less hierarchical than in these earlier works—changes that ap- relationships are executed appear to be more modular and flexible than in
pear in large part indebted to the procedures used in earlier works. Even later approaches to earlier works, such as the orches-
Strophes.39 trated Notations and revisions to . . . demonstrate a retroactive application
of this kind of formal thinking in which works are published and per-
formed in part rather than whole, and are expanded at liberty without
of their compositional processes. Boulez had a particular talent in this concern for their formal integrity as “wholes.” See Tissier (2012), Dal
area. Molin (2008, 2016), Goldman (2001, 2011), and Piencikowski (2016) for
38 See Losada (2008, 2014), O’Hagan (1997, 2006a, b, 2017), and Salem summaries on how Boulez’s approach to form developed at this time.
(2014). 40 Losada (2014).
39 Later works undoubtedly feature deep points of structural interconnection 41 Goldman (2011) offers the clearest discussion of the pervasiveness of the
between micro- and macro-structural organization, but the ways that these “Sacher” and “Originel” series in Boulez’s later works. I discuss important
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 255
and refrains between the original Explosante-fixe (1972) and comparing Example 7 with Example 8 reveals the transparent
the much later . . . explosante-fixe . . . (1991–93).42 The list relationship between the singular flute part in L’Orestie (re-
goes on. What all of these examples have in common is a per- hearsal 16) and the large, structural notes in strophe 1h. The
ceived value in combining serial materials with modular or ad- correspondences highlight that, even after formalizing his new
ditive forms, allowing Boulez to quickly draft, permute, and serial process, Boulez did not hesitate to reuse unadulterated
rework old, new, borrowed, or revised musical materials into material as the melodic basis for Strophes, just as he did at the
new or expanded musical works. They also share another fea- start of the piece. Things get more complicated as Boulez
ture: the general abandonment of creating brand new matrixes shifts to additional reuses of this material. The relationships
for each new work.43 In short, like the conversion from between L’Orestie and Strophes demonstrate how, in virtually
increasingly cryptic revisions that connect sonnets AB–ABC analyses of his works. Nonetheless, several recent studies have
to Boulez’s three Improvisations sur Mallarme from Pli selon examined its sketches, revealing layers of analytical detail about
pli, adding yet more evidence as to how borrowings shaped the its construction and Boulez’s corresponding compositional
creative preoccupations of the composer. process.46 The discussion that follows shares many
example 9(a). Excerpt from Strophes, strophe 3d (partial diplomatic transcription from PSS, BS/G, 2c)
example 9(b). Excerpt from “ Don” for piano, cahier 1 (partial diplomatic transcription from PSS, BS/G, 4b, 6)
258 music theory spectrum 41 (2019)
example 9(c). Excerpt from Strophes, strophe 1h, cycle 1 (partial diplomatic transcription from PSS, BS/G, 2c)
observations with these studies, although the analytical detail cases, Boulez uses the sonnet for pitch material, while the
is based upon my own investigation of the primary sources. rhythmic material is tied to the declamation of the text.
Boulez often made clear notes in the margins of his In these examples, it is easy to see the literal, unadulterated
sketches. This may be the most surprising and enlightening as- use of the sonnet material. Similarly, numerous commentators
pect of looking at his manuscripts: many cues are provided for have also pointed out the use of Boulez’s 1945 Notations (for
the observer, from how one serial process relates to another, to piano) in Improvisation I sur Mallarme, “Le vièrge, le vivace et
exactly what pre-composed sources he planned on using for bel aujourd’hui,” although rarely with specific comparisons.49
various parts of his Improvisations.47 Other scholars have dis- Indeed, they resemble the type of clear visual correspondences
cussed these relationships in part, with Guldbrandsen’s studies found between Strophes and “Don” for piano, with a similar
providing a clear argument regarding the source of the me- dedication to accurate pitch duplication, even when pitch
lodic/vocal content of the Improvisations and how it relates to names and accidentals are switched between the two works
certain trial-and-error revisions used to reorganize this material (from, say, a G] to an A[). Note, for example, the correspond-
between Strophes and the Improvisations.48 Below, I outline ences between Examples 11(a–b): these types of transcriptions
how Boulez’s procedures for reusing such materials changed account for almost all of the four sections of the Improvisation
and developed across the Improvisations, from an embrace of (not including the unpitched percussion parts that were added
reusing vocal materials borrowed from the “sonnets,” to new, during later revisions). Together, these examples highlight that
more elaborate reuses of other compositional materials in the Boulez used the same combination of transcription and revi-
instrumental parts of these works. sion found in Strophes—that is, transcription from one instru-
In the first Improvisation (composed during 1957, the same ment to another with little attempt to obfuscate the
year as Strophes), Boulez used his most basic “improvisatory” unpublished source—to compose the vast majority of his first
methods by creating a serialized reading of sonnet BC. He Improvisation, even as he combined these methods with more
employed prime and retrograde readings that usually begin at elaborate forms of serial derivation in the vocal part.
intermediate points within the sonnet, but otherwise faithfully Borrowings in Improvisation II sur Mallarme, “Une dentelle
follow the original pitch order. Examples 10(a–d) provide two s’abolit,” are less transparent than in Improvisation I.
of the more obvious uses of such pitch derivations. In both Guldbrandsen highlights how Boulez used a specific sonnet
from Strophes (here, sonnet AB) to provide the basic melodic
47 PSS Sammlung Boulez also includes many published and unpublished
relationships among his pieces, thanks to the meticulous efforts of the 49 Goldman (2011, 33) illustrates one such correlation. Tissier (2012) is very
archive’s staff. transparent when discussing these connections, whereas Jameux (1991)
48 Guldbrandsen (1997b, 2016). epitomizes more typical approaches.
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 259
example 10(b). Excerpt from sonnet BC, where a retrograde reading starting on F] provides the pitch material for the vocal part above
(partial diplomatic transcription of PSS, BC/G, 2b, 1–3)
material for the Improvisation; however, the use of the sonnet such instance of this borrowing. Here, the theatrical source
is both less consistent and more oblique in this case.50 What contains color-coded labels referring to each “row,” as well as
has been less discussed is that Boulez also felt compelled to yet pencil additions citing various instruments from the
again borrow instrumental parts for the Improvisation: in this Improvisation (vibe, piano, cloches, and harp all appear in the
instance, he used an excerpt from L’Orestie, Agamemnon 3 transcription).51 Ultimately, Boulez changed his mind as to
(Choir A) as the basic material. Examples 12(a–b) provide one
51 The use of such labels corresponds to the similar annotations made for the
50 Guldbrandsen (1997b, 2016). Strophes borrowings.
260 music theory spectrum 41 (2019)
what rows would be used for what instruments (in Boulez expanded the orchestration to “balance” the perfor-
Example 12[b], Ax1 ¼ piano, Bx1 ¼ vibes, Ax2 ¼ chimes, and mance of this movement against the massive forces of
Cx1 ¼ celesta), but the basic underlying process remains faith- Improvisation III. Once more, Boulez capitalized on the reuse
ful to his sketch: each row provides clearly defined pitch mate- of previously scored material to facilitate a greater focus on
rial and phrasal limits for the polyphony of the Improvisation. shape, color, and texture, resulting in the sumptuous and ex-
In fact, the polyphony was rich enough to be expanded twice: quisitely refined orchestration of the second Improvisation.
first in this revision and again, just a few years later, when The gulf separating the effects of this orchestration from
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 261
usual, Boulez prepares far more material than he could possibly In Improvisation III, Boulez incorporates his most elaborate
exploit in the movement as it exists, reminding us that the procedures of the set to “improvise” the gestural contour of the
original (unrealized) plans for Improvisation III actually vocal line based on “improvised” transpositions and fragmenta-
stretched far beyond its current length.57 Coincidentally, this tions of material. These materials are themselves “improvised”
conversion process occurs in a later sub-movement (V of VI) riffs taken from a Strophes “sonnet,” which is itself a reduced
of the final Improvisation. Thus, like the first two “improvisation” based on a tune from L’Orestie. In each case,
Improvisations, it also involves a “sonnet,” but like the coda to “improvisation” here refers to a creative rethinking of material
Strophes, its execution suggests an in medias res development. based on preconceived music that is inherently limited by cer-
tain restrictions or stylistic guidelines—a definition that could
apply as easily to a jazz number as to a rock solo. Put another
57 In fact, Boulez began his prose outline with six sections; he then expanded
way, these are conceptual revisions of musical ideas and potenti-
this to ten sections. Ultimately, he seems to have used only one of these
ten sections (the first and shortest of them all) for the 1959 sub-move-
alities, not mere material revisions of scored work.
ment, although it is entirely possible that other material was reworked Boulez’s reuse of instrumental material is similarly reveal-
into other sections of the movement. ing. The sub-movement “Bulles” provides a good example.
264 music theory spectrum 41 (2019)
example 14(b). Incomplete diplomatic transcription of a trial-and-error realization of Improvisation III, “ A” vocal part, p. 3, original
1959 score (from PSS BS/G, 3e, 6b)
teasing the ever-expanding sonnet from pierre boulez’s musical poetics 265
Here, Boulez uses marginalia from a sketch for L’Orestie to “side sketch,” testing the pitch material in different forms and
new and intriguing ends, as shown in Example 15(a). In combinations. As with the vocal parts, certain grace-note pitch
L’Orestie, this numeric pattern helped rearrange harmonies content appears “ad lib”—whether or not it has been serially
(themselves likely borrowed from Symphonie mecanique of derived from some other place, these notes are not consistent
1955) across a polyphonic introduction.58 Boulez then tran- from one option to the next (despite the consistency of other
scribed the pattern to create a new, four-color sketch with pitch content). Remarkably, all of this content is derived from
added annotations in order to apply the pattern to “Bulles.” the six pitches in Example 16(a) in a virtuosic expansion that
Example 15(b) shows the conversion of this raw pattern into a exhausts the concept of “improvisation” at the heart of his
linear-spatial arrangement of harmonies, now labeled as series Mallarme project.
according to a large matrix borrowed from the Troisième sonate. While evidence of this type is known to most scholars of
In effect, these specific harmonies are now used in at least four Boulez’s sketches, only scholars who look across Boulez’s oeu-
different works, each with different modes of accentuation and vre discover these reiterative gems and, even then, one must be
articulation: not coincidentally, these same harmonies form the willing to challenge the composer’s meticulous sketches to ask
basis of other, future works as well.59 This evidence empha- what happens in between, or even conceptually before, his or-
sizes the pragmatism of Boulez’s revisions and self-borrowings, ganizational processes take hold. Each of Boulez’s
as the materials in question come from a single, elaborate bloc Improvisations “improvise” on borrowed material—not only
sonore matrix that provided the harmonic content for many borrowed series and matrixes, but actual, pre-composed mate-
works. rial that is sometimes labeled in his sketches. His stylistic evo-
Finally, Examples 16(a–b) illustrate the first and last parts lution during this period, both as a serial composer and as a
of a three-stage genesis of one of the bubbles. First, Boulez melodist, was thus indebted to his reuse of notated material,
converted the “series” label from Examples 15(a–b) into actual which served as a catalyst to liberate his thinking and to shift
pitch-class sets borrowed from earlier organizational sketches his compositional efforts away from organizational sketches to
for the Sonate (et al.). Example 16(a) also includes an arrange- actual notes that could be manipulated into more idiomatic
ment in the left margin that “spatializes” the content with col- musical gestures for new instruments and scores.
ors for the first, “open” score of Pli selon pli, although this Accepting this argument necessitates a new hermeneutics
material is rendered irrelevant in the “fixed” 1983 revision. for these works. This is particularly significant given recent
Nonetheless, these colors are relevant to the final sketch of this scholarship on Boulez. Edward Campbell, in particular, stands
bubble, reproduced as Example 16(b), in which grouping out as a lead investigator of how Deleuze’s philosophical dis-
structures are defined by colored brackets to indicate the multi- position may relate to Boulez’s musical one, with an emphasis
ple options for performance, each with its own gestural inflec- on reiterations of “virtual” themes as a fundamental component
tions. In between these sketches, Boulez created a large, dense of a slightly more postmodernist, rather than rigidly structural-
ist, Boulez.60 Central to Campbell’s many observations is the
58 I discuss the deep connections of these sketches with L’Orestie and other idea that Boulez often evokes his own form of “difference and
works in Salem (2014), especially in relation to Examples 6.4–12. repetition” in his use of virtual themes and gestures, where
59 A single matrix, usually associated with the Troisième sonate, also appears
to be the basis for Figures-Doubles-Prismes (r.1968), Domaines (r.1968), 60 See Campbell (2010, 2013) and Scherzinger (2008). For a broader appli-
and, by extension, Dialogue de l’ombre double (1985). See Tissier (2012) cation of Deleuze’s ideas to music, including a second contribution by
and Salem (2014), especially my “Troisième sonate Stack Genus.” Scherzinger, see Hulse and Nesbitt (2010).
266 music theory spectrum 41 (2019)
polemical or not—regarding the potential of music to “express” polemicist, conductor, and overwhelming supporter of the
anything beyond “sauceless” concepts or forms.64 And yet, the genre, but also to reiterate his assertion for the continued ne-
ghost of Stravinsky has only encouraged our desire to search cessity of avant-garde composition across no less than sixty
and question the influence of a composer, his writings, and his years as an internationally respected musician. If there is one
spokesperson over our own interpretation of his works. Why thing I have learned in sorting through Boulez’s sketches, it is
not embrace the controversies surrounding Boulez in similar that the term “serialism” was expanded by his changing com-
ways? positional approach in ways that apply to many of the holdings
Finally, Boulez’s recent passing is also a call not only to at the PSS, from Stockhausen and Ligeti to Reich and so
reassess his contribution to modern music as an aesthetician,
many others. Whether or not these composers actually assimi-
lated Boulez’s compositional methods is another question for
64 I refer here to Stravinsky’s various writings on his Octet and what is de-
another day. There is certainly reason to believe, however, that
scribed as the beginnings of his turn to the “neoclassical” style, as summa- Boulez helped to reconceive the limitations of twelve-tone
rized in Taruskin and Gibbs (2013, 893–900). composition by exploding those limits from within, and that
268 music theory spectrum 41 (2019)
he did this not through Structures 1a and declaring Schoenberg Davis, Whitney. 1996. Replications: Art History, Archeology,
dead, but through his careful, continuous, and labored revision Psychoanalysis. University Park: University of Pennsylvania
of his own ideas to create a unique, immediately identifiable Press.
style from the most amebic of musical series. Decroupet, Pascal. 1994. “Developpements et ramifications de
la pensee serielle Pierre Boulez, Henri Pousseur, Karlheinz
Stockhausen 1951–1958.” Ph.D. diss., Universite de Tours.
works cited Donin, Nicolas and Jonathan Goldman. 2008. “Charting the
Score in a Multimedia Context: The Case of Paradigmatic
Adamowicz, Emily J. 2015. “A Study of Form and Structure Analysis.” Music Theory Online 14 (4). http://www.mtosmt.
Iddon, Martin. 2013. New Music at Darmstadt: Nono, travers ses ecrits. Ed. Jean-Jacques Nattiez, François Nicolas,
Stockhausen, Cage, and Boulez. New York: Cambridge and Jonathan Goldman. 73–88. Paris: Delatour France.
University Press. ———. 2013a. “Le franc-tireur et les moutons ou l’avant-
Jameux, Dominique. 1991. Pierre Boulez. Translated by Susan garde selon Boulez.” In Avant–gardes: Frontières,
Bradshaw. Cambridge, MA: Harvard University Press. Mouvements. Volume I: Delimitations, Historiographie.
Leleu, Jean-Louis and Pascal Decroupet, eds. 2006. Techniques 259–84. Paris: Delatour France.
d’ecriture et enjeux esthetiques. Genève: Editions ———. 2013b. “Structure, que me veux-tu?” In Theories de la
Contrechamps. composition musicale au XXe siècle. Volume I. Ed. Nicolas
Losada, Catherine. 2008. “Isography and Structure in the Donin and Laurent Feneyrou. 867–86. Lyon: Symetrie.
Techniques d’ecriture et enjeux esthetiques. Ed. Jean-Louis ———. 2008. The Cambridge Introduction to Serialism.
Leleu and Pascal Decroupet. 45–93. Genève: Editions Cambridge: Cambridge University Press.
Contrechamps. Wilson, Charles. 2004. “György Ligeti and the Rhetoric of
Szendy, Peter. 2008. Listen: A History of Our Ears. Trans. Autonomy.” Twentieth-Century Music. 1 (1): 5–28.
Charlotte Mandell. New York: Fordham University Press. Zenck, Martin. 2002. “L’Orestie – Marges. Zu den frühen rit-
Taruskin, Richard T. and Christopher H. Gibbs. 2013. The uellen musiktheatralen Projekten von Pierre Boulez.”
Oxford History of Western Music College Edition. New York: Mitteilungen der Paul Sacher Stiftung 15: 20–27.
Oxford University Press. ———. 2003. “Pierre Boulez’ Orestie (1955–1995). Das
Tissier, Brice. 2012. “Mutations esthetiques, mais continu- unveröffentlichte Manuskript der szenischen Musik zu
Music Theory Spectrum, Vol. 41, Issue 2, pp. 244–70, ISSN 0195-6167,
electronic ISSN 1533-8339. V C The Author(s) 2019. Published by Oxford
University Press on behalf of The Society for Music Theory. All rights
reserved. For permissions, please e-mail: journals.permissions@oup.com.
DOI: 10.1093/mts/mtz011