Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
di studi musicologici
7 · 2010
offprint
PISA · ROMA
FA B R I Z I O S E R R A E D I TO R E
MMXIV
Direttore / Editor
Paolo Fabbri
Consulenti / Advisers
Anselm Gerhard · Roger Parker
Ellen Rosand · Philippe Vendrix
*
«Musicalia» is a Peer-Reviewed Journal.
The eContent is Archived with Clockss and Portico.
IMMAGINI DI GIOVENTù.
SAGGI SULLA FORMAZIONE
E SULLE PRIME OPERE DI PIERRE BOULEZ
on the high F ; Chant : 19, very daring, but very skillful – the best chant. ». 1
ten in a telegraphic style that precludes developed conclusions, his criteria for a well-
constructed realization can be deduced from his criticism of mishandled modulations,
the abuse of pedal tones and cadences, parallelisms, pretense, and exercises without
rhythmic or formal unity. In far more subjective language, Dutilleux praised students
who had used the proper harmonic style and ‘clear’, ‘beautiful’ harmonies. These were
shared values inculcated by Conservatoire instruction, understood by the jury, but
more difficult to put into words. 3
1 « Basse : 17 T[rès]B[ien] mais attention à la tessiture de l’alto fa aigu ; chant : 19 très osé mais très habile
– meilleur chant. » (Basel, Paul Sacher Stiftung [henceforth pss], Sammlung Henri Dutilleux). In French
grading systems based on a total of 20 possible points, any grade above 15 is generally considered quite
good. Four premiers prix were awarded to the male candidates of 1945, three of them with six votes from
the eight member jury : Serge Baudo, from Jean Gallon’s class ; Pierre Boulez, from Olivier Messiaen’s class ;
and James Moreau, from Marcel Samuel-Rousseau’s class. The fourth premier prix, awarded with five votes,
went to Pierre Villette, also a student of Jean Gallon. An additional honor went to the first named of the
premiers prix, the Prix Lavignac, and was awarded to Baudo over Boulez in a vote that split the jury five to
three. Three seconds prix went to Michel Fano and Charles Chaynes, both students of Gallon, and Pierre
Houdy, a student of the recently appointed Maurice Duruflé. Archives nationales de France, Pierrefitte-sur-
Seine (henceforth an), aj37 538 (here and elsewhere, when not noted otherwise, all translations from the
original French are the authors’ own).
2 Receiving a premier prix after just one year of study was truly a feat. Most premiers prix were the result
of multiple attempts. Among the other premiers prix of 1945, Baudo received his premier prix after two con-
cours, Fano after three, Vilette after four and Moreau on his fifth and final try. an, aj37 537, 538, 558, 559.
3 In 1945 Dutilleux also evaluated the Conservatoire’s harmony students in the exam that determined
whether or not they would be admitted to the concours. Here Boulez also received the highest praise : « Basse,
«musicalia» · 7 · 2010
32 yves balmer · christopher brent murray
Boulez stands out in the history of Messiaen’s short-lived tenure as harmony profes-
sor. In five complete years of instruction, only three students from Messiaen’s class re-
ceived premiers prix : Sylvie Valès and Yvonne Loriod in 1943, and Pierre Boulez in 1945. 1
dents were among them, nearly 50%. Considering that the number of harmony classes at the
Conservatoire was six until 1934 [...] and five thereafter, it makes a rather impressive average for
one professor, don’t you think ? 3
gram ». 5 In making this assertion, Boivin cites the laudatory recollections of Messiaen’s
former students, notably Boulez, and infers that Messiaen’s Vingt Leçons d’Harmonie.
Dans le style de quelques auteurs importants de “l’Histoire Harmonique” de la musique depuis
very good, Chant, extraordinary, extremely pretty writing in the instrumental style. » (« Basse T[rès]B[ien] –
Chant extraordinaire – extrêmement joli d’écriture style instrumental. » (pss, Sammlung Henri Dutilleux).
1 Messiaen 2008. Messiaen au Conservatoire, ed. by Anne Bongrain, Paris, crec, 2008, pp. 29-33.
2 Having begun his career as a solfège professor, Jean Gallon became a professor of harmony in 1919, a
position he held for two years beyond his official retirement in 1945. See Jean Gallon’s personnel dossier.
an, aj37 482.
3 Noël Gallon, handwritten letter to Claude Delvincourt, 24 July 1945 : an, aj37 482 : « Entre 1920 et 1945, il
a été décerné 123 premiers prix d’Harmonie : sur ce total, 58 de ses élèves furent les bénéficiaires, c’est-à-dire
près de 50%. Le nombre des classes d’harmonie ayant été de 6 jusqu’en 1934 [...] puis de 5, cela fait, pour un
seul professeur, une assez jolie moyenne, n’est-ce pas ? ».
4 Messiaen officially studied with Gallon from 1920 to 1925, but he had begun learning harmony at an
early age. In Nantes, during the 1918-1919 school year, the young musician worked from treatises by Henri
Reber and Théodore Dubois in private lessons with Jean de Gibon, harmony professor at the local conser-
vatoire in Nantes.
Messiaen continued his private studies with Noël Gallon upon his arrival in Paris in the fall of 1919. He
entered the Paris Conservatoire in the fall of 1920 as a student in the preparatory piano class of Georges
Falkenberg, and was excused from taking solfège classes by Conservatoire director Henri Rabaud. Messiaen
audited Jean Gallon’s harmony class during the 1920-1921 academic year before joining it as an official stu-
dent in the fall of 1921. In contrast with the relative ease and rapidity with which he managed to garner his
later prizes in fugue and accompaniment, Messiaen would spend the maximum five years in Gallon’s class
(1921-1926) without managing to receive the premier prix in the year-end concours. It should be observed that
Messiaen was still very young the year of his final concours in harmony ; at seventeen, he was still younger
than most of the students entering the harmony class for their first year of study. Somewhat surprisingly,
this lack of a premier prix did not prevent him from being nominated as a harmony professor in March 1941.
See Peter Hill, Nigel Simeone, Olivier Messiaen, transl. by Lucie Kayas, Paris, Fayard, 2008, pp. 28-29, cit-
ing the unattributed article Hommage à un maître disparu, Jean de Gibon, « Echo du pays de Redon », 26 janvier
1952, and Anne Bongrain, Messiaen 2008. Messiaen au Conservatoire, Paris, crec, 2008, p. 118.
5 Jean Boivin, La Classe de Messiaen, Paris, Bourgois, 1995, p. 32 : « la classe de Messiaen dépassait déjà
aen’s class. Boivin credits Messiaen with personally reforming Conservatoire tradition
through an increased use of score analysis and reference to specific historical styles. 2
Other sources contradict this position. In Boulez’s homework, for example, lessons
from traditional harmony manuals greatly outnumber texts by Messiaen (these include
both lessons from the Vingt Leçons and other unpublished lessons written by Messiaen).
More importantly, the innovative nature of the Vingt Leçons cannot be assumed from
the title alone. Closer examination reveals traditional harmony lessons with an ambigu-
ous integration of the notion of ‘harmonic style’. In general, the Vingt Leçons cannot
be said to differ much from texts by Messiaen’s predecessors. Moreover, the manner in
which historical styles were referred to in Conservatoire harmony classes continues to
be an ill-defined subject requiring further research and nuance.
Keeping in mind that Boulez’s exceptional talent may mean that his homework and
notes do not represent a norm, this article will propose some working hypotheses on
Messiaen’s harmonic instruction based on Boulez’s experience in his class. After a re-
view of the sources used in Messiaen’s class and a closer look at the historical expecta-
tions for realizing basses données and chants donnés in Conservatoire harmony classes,
we will analyze notes and examples from Boulez’s papers to partially reconstruct Mes-
siaen’s class and its priorities. Special attention will be given to the influence of Jean
Gallon and Messiaen’s « formules d’harmonie » (« harmony formulas »). Taken together,
these examples lend support to the notion that although Messiaen’s instruction em-
ployed direct references to the repertoire, it remained well within the boundaries of
Conservatoire traditions.
2. Sources
The archival primary sources relating to Boulez’s studies in Messiaen’s harmony class
can be divided into four categories : Messiaen’s collection of annotated harmony trea-
tises and manuals, Boulez’s realizations, a set of harmonic formulas composed by Mes-
siaen for use by his students recopied by Boulez, and Conservatoire records kept at the
Archives nationales de France, including teaching evaluations, jury reports, and procès-
verbaux for exams and concours. 3
The harmony treatises and manuals used by Messiaen, often annotated in his own
hand, are conserved in the Médiathèque Hector-Berlioz of the Paris Conservatoire. 4
The lessons gathered in these manuals were traditional sources used in harmony in-
struction at the Conservatoire. We know that Messiaen realized some of these lessons
in concours during his student days in Jean Gallon’s harmony class (1921-1926), and his
annotations indicate that he assigned the same texts to his students. The composer’s
library is the first demonstration that his teaching materials remained solidly anchored
1 Olivier Messiaen, Vingt Leçons d’Harmonie. Dans le style de quelques auteurs importants de “l’Histoire Har-
monique” de la musique depuis Monteverdi jusqu’à Ravel, Paris, Leduc, 1939. For a discussion of the Vingt Leçons
and Messiaen’s unannounced inclusion of elements of his own style in the final lesson, see Yves Balmer,
Christopher Brent Murray, De l’harmonie à la composition : Messiaen prophète de son propre style, in Hori-
zons de la musique en France 1944-1954, ed. by Laurent Feneyrou and Alain Poirier, Paris, Vrin, forthcoming.
2 Boivin, op. cit., pp. 38-39.
3 See Yves Balmer, Christopher Brent Murray, La classe de Messiaen : retour aux sources, in Une Mu-
sique française après 1945 ?, ed. by Emmanuel Ducreux and Alain Poirier, Lyon, Symétrie, forthcoming.
materials also underline that the Conservatoire’s harmony concours were a longstand-
ing insider’s affair. The recipients of premiers prix had to learn to realize texts that met
norms defined by previous winners and their judges.
Messiaen’s annotations of traditional harmony manuals occasionally include dates
that exercises were assigned as homework or practice exams. In some cases, Messiaen
indicates the composer (Bach, Mozart, Franck, Fauré) in whose style the lesson was
conceived. In some examples, Messiaen also wrote an alternative realization alongside
the solution in the teacher’s manual – an act that might be interpreted as a sign of
Messiaen’s open-mindedness or a simple demonstration of a certain degree of flexibil-
ity within the Conservatoire system. This flexibility can be related to Dutilleux’s com-
mentary on Boulez’s chant donné realization from the 1945 concours. Boulez had written
something unexpected and daring, an inventive and viable alternative deemed just as
acceptable as more conventional realizations.
The exercises realized by Pierre Boulez during his year in Messiaen’s class mainly
correspond to texts in Messiaen’s former library, although there are also a few lessons
by authors from Messiaen’s own generation not found in his collection, as well as a
handful of chants and basses donnés by Messiaen himself. 2 Out of about eighty harmony
lessons on one hundred and forty pages of exercises, fourteen are texts by Messiaen.
Five come from his Vingt Leçons and nine others remain unpublished. 3
Boulez’s harmony exercises are always realized in open score using soprano, alto,
tenor and bass clefs. The order in which the texts were realized is difficult to establish,
but given the few dates that are scattered throughout, they do not seem to have been
kept in chronological order. It is possible, nevertheless, to distinguish between work
written entirely in pencil that was probably done in class, and homework, carefully
recopied in ink. These differences also indicate two modes of thought – the rapidity
needed to finish work done in class and careful reflection and perfection realized at
home. 4 Messiaen’s hand is often present in the form of comments and corrections,
1 Charles Lenepveu, Cent leçons d’harmonie : recueil comprenant 50 leçons (basses et chants donnés) avec la ré-
alisation de l’auteur / par Ch. Lenepveu,... 30 leçons de concours du Conservatoire / réalisées par les élèves de la classe
d’harmonie (femmes) de M. Ch. Lenepveu. 20 leçons inédites avec la réalisation des auteurs / de MM. Th. Dubois, Er-
nest Guiraud et Henri Fissot, Paris, Lemoine, 1902 ; Jean and Noël Gallon, 60 exercices et thèmes d’harmonie,
Paris, Heugel, 1929 ; Jean Gallon, 60 exercices et thèmes d’harmonie. 2ème série, Paris, Heugel, 1933.
2, 2010, p. 309.
4 In a few exercises, the text to be realized is written in ink, with the realization in the other three voices
in pencil. In one instance, Boulez even took the time to recopy in ink an entire corrected realization that
he attributes to Messiaen. This realization is of an alterné by Jean Hubeau. An alterné is a lesson whose text
opens with a basse and ends with a chant – these were often used for the shorter mise en loges of exams. See
pss, Sammlung Pierre Boulez, Mappe A, Dossier 2a, 3, folio 3.
boulez and messiaen ’ s harmony class 35
Messiaen, written in a four-part texture appropriate for harmony lessons (see Figure 1). 1
The formulas’ labels often make reference to the harmonic procedures they employ and
the composer or work they are based upon. 2 The formulas appear to have functioned
or rhythmic motives from the repertoire 4 that could be reworked and reused in their
own creations, it is revealing, if not surprising, to find him adapting this technique to
the needs of the traditional harmony classroom. 5
ples from the repertoire in the realization of academic harmonic exercises, as well as
1 The harmony formulas are found in the Paul Sacher Stiftung, Sammlung Pierre Boulez, Mappe A,
Dossier 2a, 1, folio 1 = Microfilm 577.
2 Not all of the formulas bear labels making reference to a particular work, but this does not necessarily
mean that they were not adapted from the repertoire as well. In Messiaen’s correction reminding Boulez of
the first formula, he indicates its mixed stylistic references, (« Carmen, Lakmé, Manon »), something not pres-
ent in Boulez’s recopied version of the formula. The formulas are carefully recopied in ink on five sheets of
12-stave manuscript paper in folded folio format, a set of sheets which seem to have been carefully chosen to
contain the whole, no more, no less, suggesting that they may have been copied (or recopied) in one sitting.
3 See for example the passages from Debussy transformed in examples 189-192 and 224-226 of Olivier
Messiaen, Technique de mon langage musical, Paris, Leduc, 1944.
4 Also often referred to by Messiaen as « formules ».
5 On Messiaen’s encouragement to his students to use thematic dictionaries see Olivier Messiaen,
Yvonne Loriod-Messiaen, Analyses des œuvres pour piano de Maurice Ravel, Paris, Durand, 2003, p. 11 : « J’ai
souvent conseillé à mes élèves de se constituer un dictionnaire mélodique pour les étudier, les amplifier,
en faire leur suc et leur style. De même pour le dictionnaire rythmique. ». For similar confirmations of this
practice see Jean Barraqué , Écrits, ed. by Laurent Feneyrou, Paris, Sorbonne, 2001, p. 154, and Vincent
Benitez, A Conversation with Composer Gerald Levinson about Olivier Messiaen, consulted at oliviermessiaen.
net/sitedocs/papers/Levinson.pdf, p. 30.
6 See Boivin, op. cit., p. 37 : « Henry retained the odd memory of Messiaen singing Massenet’s Manon
and Léo Delibes’ Lakmé in class... [...] Strangely, with the exception of Debussy, Pierre Boulez did not retain
the same names as Henry : “I remember the Lieder of Schumann very well, but Messiaen didn’t like them
himself – and mélodies and other pieces by Fauré, that he didn’t like either, but which he analyzed in order
to show us how their languages worked. He analyzed many works by Debussy [...]”. To this, Boulez added
Die Meistersinger by Wagner, an opera that Messiaen liked in particular » (« Henry garde également le curieux
souvenir d’un Messiaen chantant en classe Manon de Massenet ou Lakmé de Léo Delibes... [...] Curieuse-
ment, à l’exception de Debussy, Pierre Boulez n’a pas retenu les mêmes noms : “Je me souviens très bien de
Lieder de Schumann – mais Messiaen n’aimait pas ça lui-même – et de mélodies, de pièces de Fauré, qu’il
n’aimait pas non plus, mais qu’il analysait afin de nous montrer les mécanismes du language. Il a beaucoup
analysé des œuvres de Debussy [...]”. A cela, Boulez ajoute Les Maîtres chanteurs de Wagner, un opéra que
Messaien affectionnait particulierement »).
When Christopher Murray presented findings related to these at the Messiaen Festival at La Grave on 5
August 2010, Pierre Boulez was in the audience and affirmed that he recalled the harmonic formulas from
Manon.
36 yves balmer · christopher brent murray
Fig. 1. Messiaen’s harmony formulas 1 to 5 as recopied by Pierre Boulez (Basel, Paul Sacher
Stiftung, Sammlung Pierre Boulez, Mappe A, Dossier 2a, 1, folio 1 = Microfilm 577, 150).
boulez and messiaen ’ s harmony class 37
the composer’s creative practice of reworking pre-existing music as both as a composer
and pedagogue. They show one way that Messiaen might have suggested his students
approach writing ‘in the style of ’ while remaining within the confines of a traditional
harmony lesson : by borrowing fragments of melody and harmony from the repertoire
The reason for the order in which the formulas are presented is not clear. Take for
example the variety among the first five formulas seen in Figure 1 : an exceptional reso-
lution, a pedal, the Neapolitan sixth chord, and two sequences. These devices crop up
1 In his interviews with Antoine Goléa, Messiaen claimed he was « the first person, in a harmony class,
to talk about Orfeo and the madrigals of Monteverdi, about the Brandenburg Concerti and Mass in B minor
of Bach, about Mozart Quartets and the Schumann Novelettes, about Debussy’s Images and Ravel’s Histoires
naturelles » – he was almost certainly not the first person to mention these works such a setting – although
some of them are indeed transformed in his harmony formulas (Antoine Goléa, Rencontres avec Olivier
Messiaen, Paris, Julliard, 1961, pp. 242-243).
38 yves balmer · christopher brent murray
elsewhere in the series as well. 1 Some groups of consecutive formulas are drawn from
the same composer, but there is no apparent guiding logic to the arrangement of the
whole. In addition, the length of the formulas varies greatly, ranging from just three
or four chords to as many as twenty measures. Perhaps the formulas were created over
the course of the school year (or were leftovers from a previous school year). They may
also represent traces of Messiaen’s analyses of works from the repertoire that pertinent
to preparing specific exercises in a related style that were then recopied in the weeks
before the exam. 2 The possible adaptation of formulas for use in particular exercises
might also explain their inconsistent levels of transposition, which will be discussed
further below.
The Conservatoire’s continuity of tradition is no doubt partly due to the fact that the
institution had long hired its former star pupils, often prix de Rome recipients, as pro-
fessors. 4
Maurice Duruflé, who studied with Gallon at the same time as Messiaen and was
nominated to teach harmony at the Conservatoire in 1943, just two years after Mes-
siaen, 5 leaves a useful recollection of Gallon that details the technical prerequisites for
entering a harmony class. Duruflé also hints at the different techniques required to real-
ize a basse donnée or a chant donné.
In those days the harmony professors personally recruited their students without an official
exam. [...] I was dazzled by the lessons that my new classmates brought to class. Basses données
in the polyphonic style of Bach with subject, countersubject, imitations, and the return of the
subject in different voices, always accompanied by the countersubject. It was a veritable prepara-
tion for writing fugues. We were already quite distant from the Dubois treatise. The chant donné,
to which I felt particularly attracted, also astonished me by its distinct harmonization and the
perfection of the realizations. 6
1 « 1). Résolution exceptionnelle de l’accorde de 6te » ; « 2). Pédale médiaire » ; « 3). 6te napolitaine. (Bach,
Kyrie | Messe en si min.) » ; « 4). Marche (imitée de “Manon” 4e acte) » ; « 5). Marche (Franck-Ravel) ».
2 Although we are in touch with Boulez and other former harmony students of Messiaen in hopes of
receiving some further clarifications on these documents, this information was not available at the time of
publication.
3 John Robert Clevenger, Debussy’s Paris Conservatoire Training, in Debussy and His World, ed. by Jane
Fulcher, Princeton, Princeton University Press, pp. 299-361.
4 For a development of this point see Maguerite Sablonnière, Le Conservatoire de musique de Paris
pendant l’entre-deux-guerres, doctoral diss., École nationale de chartes, 1996, p. 109, and Rémy Campos, Au-
rélien Poidevin, « Nous entrerons dans la carrière... » Le prix de Rome, concours d’entrée des compositeurs dans
la profession ? (1906-1968), in Le Concours du prix de Rome de musique (1803-1968), ed. by Julia Lu and Alexandre
6 Maurice Duruflé, Souvenirs et autres écrits, ed. by Frédéric Blanc, Paris, Séguier, 2005, p. 34 : « Il faut
dire qu’à cette époque, [1922] les professeurs d’harmonie recrutaient leurs élèves personnellement, sans
examen d’admission officiel. [...]. J’étais ébloui par les travaux qu’apportaient mes nouveaux camarades.
boulez and messiaen ’ s harmony class 39
The basse donnée was meant, in part, to test the student’s contrapuntal capacities. Imi-
tative writing was key. Students were expected to derive their realization from motivic
elements already present in the basse, pointing out what they had accomplished by
marking points of imitation with brackets and letters. George Caussade’s harmony
treatise shows that the idea of composing a coherent form was not a concept unique
to Messiaen’s class, contradicting the intimations of Boulez’s interviews with Jean
Boivin.
The student should also, as much as possible, seek the most rational development for the pro-
posed text by using imitations. A composition of any real significance almost always begins
with a rhythmic or expressive main idea. There then follows a development of that main idea,
of varying length, which we most often hear again in the conclusion. | This essential first idea
must be put to good use over the course of the realization (moving through all of the voices, as
much as is possible). The idea will be reproduced, whole or fragmented, in the original key, in
a neighboring key (or even in a key distant from the principal key of the composition). These
reproductions are called imitations. 1
Caussade also advises that the student carefully analyze the text before attempting a
realization, noting that difficulties most often arise from an insufficient analysis. 2 He
indicates that this analysis should include the identification of the main key areas and
modulations. This technique puts an emphasis on the thematic unity of the realization
and the integration of systematic analysis in the act of composition.
More often than not, basses données contained a contrapuntal trick that the successful
student was expected to discover (the similar motivic treatments found in the collec-
tions of premier prix realizations published annually by Heugel further underline this
point). 3 Caussade distinguishes between three sorts of basses données : basses provid-
ing the opportunity for fugue-like imitation in which a subject-like motive is passed
the four voices ; basses containing the possibility for extended invertible counterpoint
(sometimes one half of the bass could be presented in against the other and vice-versa) ;
and basses termed ‘purely expressive’ but in which the student was still expected to
make the most of possibilities for imitative writing. 4
Des basses données dans le style polyphonique de Bach avec thème, contre-sujet, imitations, retour du
thème dans les différentes parties vocales, toujours accompagné du contre-sujet. C’était un véritable travail
de préparation à l’écriture de la fugue. Nous étions très loin du traité de Dubois. Le chant donné, vers le-
quel je me sentais particulièrement attiré, fut également un étonnement pour moi par la distinction de son
harmonisation et la perfection de sa réalisation. ».
1 Georges Caussade , Technique de l’harmonie, Paris, Lemoine, 1931, vol. 1, pp. 258-264 : « L’élève devra
aussi, le cas échéant, rechercher le développement le plus rationnel du texte proposé, par l’emploi des
Imitations. Une composition d’une certaine importance débute, presque toujours, par une idée principale,
rythmique ou expressive. Vient ensuite un développement, plus ou moins étendu de cette idée principale qui
nous retrouvons, le plus souvent, à la conclusion. | Cette idée première, essentielle, doit être mise en valeur
au cours de la réalisation, [alternativement dans toutes les parties, autant que possible.] L’idée sera donc repro-
duite, entière ou fragmentée, dans le ton initial ou dans un ton voisin [ou même encore, dans un ton éloigné
du ton principal de la composition]. Ces reproductions se nomment Imitations. ». The passages in italics are
Le chant donné
The chant donné could usually be realized in a freer style, something Dubois terms « style
libre » or « style moderne » in contrast with the « style rigoureux » of the basse donnée : 1
The student must first allow himself to be permeated by the style and nature of the chant and
direct himself to finding it. Any harmonic artifice can be used in a chant donné, as long as the
character of the chant allows for it. There are broad margins for developing music taste and
sensitivity. 2
Caussade emphasizes that the chant donné may incorporate imitative elements if ap-
propriate, but that, in general, the form is « essentially melodic ». Still, as with the basse
donnée, the ideal harmonization was composed of four elegantly constructed voices.
The key then, to the chant donné, was to find the ideal progression of harmonies sug-
gested by the given text, harmonies that were as ‘natural’ as possible. A number of
former Gallon students, including Messiaen himself, indicate that the harmonization
of a chant donné was deduced from the melodic contents of the chant itself, in a quest
for what was habitually termed the harmonie vraie. The notion profoundly marked
Messiaen’s musical thinking 3 and even appears in Boulez’s writing. In his article, In-
trace of Boulez’s traditional Conservatoire studies, and proof of the term’s common
acceptance. 4
The harmonic vocabulary of such exercises could focus on a range of periods and
composers, from the harmonies of Bach, Mozart, Schumann, Franck, and Massenet,
up to and including modal or chromatic texts in the style of Fauré. Like Dubois and the
Gallon brothers before her, Messiaen’s former harmony classmate, Yvonne Desportes,
emphasized the importance of respecting the text’s historical style when realizing basses
or « chants ‘libres’ ». 5
The format of Boulez’s harmonic instruction based on traditional basses and chants
donnés had been in place since the institutional changes introduced by Ambroise Thom-
as in 1878. 6 These reforms most notably included the separation of harmony study
into written harmony and exercises in harmonization at the piano. The division of the
1 Théodore Dubois, Traité d’harmonie, théorique et pratique, Paris, Heugel, 1921, p. vi.
2 Idem, Notes et études d’harmonie pour servir de supplément au traité de H. Reber, Paris, Heugel, [1890], p. 160 :
« L’élève doit d’abord bien se pénétrer du style, de la nature du chant, et se diriger vers cet examen. Tous les
artifices harmoniques peuvent trouver leur place dans un chant donné, si le caractère de ce chant le permet. Le
sentiment et le goût ont donc ici un champ large et vaste pour se manifester et se développer. ».
3 This can be seen in Messiaen’s evaluations of his own students where he often praises their ability to
find what he termed the « harmonie vraie » as well as in his theoretical writing on composition in Messiaen,
a serial harmonization of a Bach chorale on the grounds of some supposedly right harmony that a melody
might contain. » (Idem, Incidences actuelles de Berg [1948], in Idem, Relevés d’apprenti, ed. by Paule Thévenin,
Paris, Seuil, 1966, p. 239 : « Je ne repousse pas le principe d’une harmonisation sérielle d’un choral de Bach à
cause de la prétendue harmonie vraie qui se trouverait enclose dans une mélodie. »).
5 Yvonne Desportes, Guide servant d’appendice aux traités d’harmonie, Paris, Ricordi, 1961, p. 20.
6 John Robert Clevenger, The Origins of Debussy’s Style, doctoral diss., University of Rochester, 2002,
pp. 99-116.
boulez and messiaen ’ s harmony class 41
theoretical from the practical may have, in some cases, contributed to a certain abstrac-
tion of harmonic instruction through formulas and contrapuntal techniques developed
without much contact to music as it was practiced, a « musique à usage interne » to bor-
Although the traditional chants donnés or basses données of Messiaen’s and Boulez’s
harmonic educations were a product of this tradition, they were not necessarily
understood as being completely cut off from broader references to historical style.
A lesson’s stylistic references were rarely pointed out in words, but many sources
indicate that identifying the style was a key part of realizing the exercise. Yvonne
Desporte’s Guide servant d’appendice aux traités d’harmonie bears the trace of the same
concerns for style as the souvenirs of other former Gallon students. She also sum-
marizes the various styles that harmony students could be expected to recognize and
reproduce :
The style to be adopted in a harmony lesson is of great importance. One must, upon reading
the text, be inspired by the author, or more specifically, by the period that seems to have been
proposed and consciously restrict oneself to the language of that epoch. If the text is classi-
cal and contrapuntal, one’s mind will turn at once to J. S. Bach, without making any inroads
toward Franck or Ravel. Errors of style are worse than errors in voice leading. In the style of
Mozart, passing tones and neighbor tones will be less abundant than in the style of Bach, but
one would do well to include appoggiaturas resulting from double neighbors or suspensions ;
the style should be amiable and graceful. | In the style of Franck, beyond the modification of
the theme, which we will speak about in the chapter on basses libres, the harmony plays a very
important role, due to continual and unexpected modulations. | It is truly through the style of
his realizations that the student can show his taste and musicality. 2
Paul Tortelier’s recollections of Gallon’s harmony class are also illuminating in this
respect :
Errors of taste were also a serious matter. If you realized a text in the religious style of the six-
teenth century but inserted a sensual harmony in the style of Schumann, [Gallon] would say,
« This makes me think of a priest, who, in the middle of his sermon, pauses to apply makeup in
front of a mirror. ». 3
In his text on how the Conservatoire’s classes were organized, director Henri Rabaud
(director of the Conservatoire during Messiaen’s own studies) notes that harmony pro-
fessors are free to choose their teaching materials, as long as « their instruction leads
1 Christian Corre, Harmonie : Le Traité in Ecritures de la musique, Paris, puf, 1996, p. 82.
2 Desportes, Guide servant d’appendice aux traités d’harmonie, cit., p. 20 : « Le style à adopter dans un
devoir d’harmonie a une très grande importance. Il faut, en lisant un texte, s’inspirer de l’auteur, ou plus
exactement de l’époque qu’il semble proposer et s’y cantonner avec beaucoup de scrupules. Si le texte est
classique et contrapuntique, on pensera à J.-S. Bach sans incursion dans le domaine de Franck ou Ravel. Les
erreurs de style sont plus graves que les fautes d’écriture. Dans le style de Mozart, les notes de passage et
les broderies seront moins touffues que dans le style de Bach ; par contre on mettra des appogiatures, issues
de doubles broderies ou retards, le style sera aimable et gracieux. | Dans le style de Franck, en dehors de la
modification des thèmes dont nous parlerons au chapitre des basses libres, l’harmonie joue un très grand
rôle, en raison de ses modulations continuelles et inattendues. | C’est vraiment dans le style de ses réalisa-
tions que l’élève pourra prouver son goût et sa musicalité. ».
3 Paul Tortelier, Autoportrait : en conversation avec David Blum, Paris, Buchet-Chastel, 1986, p. 77 : « Faire
des fautes de goût était aussi une chose grave. Si vous aviez réalisé un texte dans le style religieux du xvie
siècle en y insérant une harmonie sensuelle dans le style de Schumann, [Gallon] disait alors : “Ceci me fait
penser à un prêtre qui, au beau milieu d’un sermon, se mettrait du fard devant un miroir.” ».
42 yves balmer · christopher brent murray
young people to know and understand the languages of masters of the past, of Bach
and Mozart, as of Fauré and Debussy. ». 1
Although some authors have interpreted the Vingt Leçons d’Harmonie as a manifes-
tation of Messiaen’s desire to reform harmony instruction, it should be pointed out
that Messiaen never openly criticized the traditional Conservatoire education he had
received. Publishing a harmony textbook so early in his career was an unusual move,
particularly for somebody like Messiaen who had never actually taught harmony in an
institutional setting. 2 The publication of the Vingt Leçons can be interpreted a strategic
decision by Messiaen to improve his chances of being nominated for one of the Con-
servatoire teaching positions that he knew would soon be vacant. 3 Although Messiaen
presented his manual’s emphasis on historical style as something new, professors like
Gallon, Dupré, and Koechlin had long encouraged their pupils to develop an awareness
of the historical development of tonal harmony and harmonic grammars of different
periods. 4
Messiaen composed many of the Vingt Leçons using borrowed and altered material
from pre-existing works in the repertoire, 5 a practice which we have demonstrated else-
where as key to his own compositional techniques, particularly during the 1940s. Musi-
cal borrowing, whether identifiable or not, does not seem to have guaranteed stylistic
cohesion in the Vingt Leçons. Lessons 7 and 8 in the style of Mozart are perhaps the most
evident examples of this. It is difficult to find much in these lessons that is particularly
1 Henri Rabaud, Preface to Le Conservatoire National de Musique et Déclamation, 1930, p. 2 (repr. from the
July 1928 issue of the « Revue des deux mondes » : « Aucun traité n’est officiellement adopté par le Conserva-
toire. Aucun n’est imposé aux professeurs, aucun ne leur est interdit. Ils sont libres de recommander à leurs
élèves l’étude des ouvrages qu’ils trouvent les meilleurs, à condition que leur enseignement aboutisse à faire
connaître et comprendre aux jeunes gens le langage des maîtres du passé, de Bach et de Mozart, comme
de Fauré et de Debussy »). This imperative, may not have always been at the center of every harmony pro-
fessor’s concerns as Conservatoire directors thought it should be. This is suggested in a letter to Charles
Koechlin written by Rabaud’s successor, Claude Delvincourt, shortly after Messiaen’s nomination in 1941.
Delvincourt underlines that Messiaen’s nomination will uphold an approach more closely based on the ap-
prenticeship of historical styles, ostensibly the approach with which he was educated in Gallon’s own class.
See the letter from Claude Delvincourt to Charles Koechlin, 17 April 1941, cited in Charles Koechlin 1867-1950,
Correspondance, « Revue musicale », 348-350, 1982, p. 106 : « A harmony class is vacant and I’ve nominated Mes-
siaen, who is young and gets it. These people have to be made to understand that there is nothing arbitrary
about the rules of harmony or counterpoint, and that the rule means nothing if it is not in accordance with
the style. » ; « Une classe d’harmonie est vacante : j’y fait nommé Messiaen, qui est jeune et qui pigera. Il faut
faire comprendre à ces gens que la règle d’harmonie ou de contrepoint n’a absolument rien d’arbitaire, et
qu’une règle ne vaut qu’en fonction d’un style. ».
2 There is however evidence of Messiaen having given harmony lessons to fellow soldiers preparing
concours for musical ensembles in the army during his military service as well as teaching lessons to private
students during the 1930s.
3 In 1941, Messiaen described the Vingt Leçons in the following terms : « I currently have, in press, a series
great musical masterpieces in order to understand different languages, to find in this work the source an ap-
plication of different rules and even to develop a personal language if any potential is there. » (Peter Hill,
Nigel Simeone, Olivier Messiaen, New Haven, Yale University Press, 2005, pp. 109-110).
4 Yves Balmer, Edifier son œuvre : Genèse, médiation, diffusion de l’œuvre d’Olivier Messiaen, doctoral diss.,
cantilenas »).
boulez and messiaen ’ s harmony class 43
Mozartian, apart from their harmonic vocabulary on a very generic level. Messiaen’s
realization, with its extreme degree of imitation based on themes taken from the basse
donnée has far more in common with Conservatoire harmony class tradition than any-
thing Mozart ever put to paper. This would suggest that following Yvonne Desportes’
description of Mozart’s style – a particular use of specific non-harmonic tones and
a general « amiable, pleasant tone » – a lesson could be understood as ‘in the style of
Mozart’ even if it sounded very little like Mozart’s actual music. The notion of style in
Conservatoire harmony lessons may have been linked to particular techniques, without
necessarily constituting a recognizable pastiche. 1
In spite of their titles’ references to musical style, many of the Vingt Leçons started
out their lives as Conservatoire exam and concours pieces written by Messiaen in the
1930s. Many had their styles imposed or at the very least ‘pointed out’ only when they
were re-adapted and integrated into the Vingt Leçons, proving that the texts in and of
themselves, were hardly revolutionary. This is the case of lessons 13 (basse donnée « in the
which were composed by Messiaen as a chant donné and basse chiffrée for the Conserva-
toire’s accompaniment exams of 1935. 2 Similarly, Messiaen’s texts for the 1938 concours
in accompaniment later became lesson 4 (basse donnée « in the style of a Prelude by J.S.
Bach ») and lesson 17 (chant donné « in a style part-Chabrier, part-Debussy »). 3 In spite of
some material links to the pre-existing musical sources, we should reiterate that style
in the Vingt Leçons is a matter of broad references to historicized harmonic vocabulary
in line with Conservatoire traditions, not the assimilation and pastiche of specific com-
posers taught by later harmony manuals like that of Bernaud and Desportes. 4
In this interview with Jean Boivin from May 1987, Messiaen claims no particular origi-
nality for his harmony class and its practices. Indeed, what choice did he have ? If he was
ary and 31 March 1937 – the stylistic indication does not appear until their publication in the Vingt Leçons.
3 �������������������������������������������������������������������������������������������������
Médiathèque Hector-Berlioz, Msc. 235. Three other unpublished manuscripts by Messiaen in the Con-
servatoire’s collection of exam and concours pieces, do make reference to particular musical styles, but the
date of these compositions is unfortunately unknown. Given that on the eve of publishing the Vingt Leçons
in 1938 Messiaen was still not indicating the style of his concours pieces would seem that they post-date
the 1939 Vingt Leçons. All three manuscripts are catalogued under Msc. 237, Basse chiffrée « Un peu vif, Style
Rameau », Continuo chiffré « Vif et décidé, Style Bach » and Chant donné « Très modéré, ému, comme un sou-
4 Yvonne Desportes, Alain Bernaud, Manuel pratique pour l’approche des styles de Bach à Ravel, Paris,
Billaudot, 1988.
5 Boivin, op. cit., pp. 33-34 : « Je n’ai rien changé aux habitudes du Conservatoire lorsque j’ai pris la classe
d’harmonie. J’ai fai faire aux élèves des leçons d’harmonie, tout simplement, des basses données et des
chants donnés, écrits à quatre voix, comme on faisait à l’époque, sur quatre portées, avec trois clés d’ut et la
clé de fa. [...] les devoirs étaient tout de même de style classique, c’est-à-dire que ça ne dépassait pas Fauré
au point de vue du langage. ».
44 yves balmer · christopher brent murray
to fulfill his job description, he had to prepare his students to do well in the concours, and
in this, there was little margin for innovation.
The harmony class met for four hours three times a week, with students being ex-
pected to complete a lesson as homework between each of the classes. 1 Although the
attendance sheets signed by professors to indicate their presence were not kept for
the 1944-1945 school year, those for January to July 1942 were, and these documents, in
combination with information from Boulez’s own recollections allow us to draw a few
simple conclusions about how the class worked. 2 Generally class was held on Monday,
Wednesday, and Friday – although Messiaen would, on rare occasions, move a class to
another day, sometimes even Sunday. Messiaen also seems to have been rather assidu-
ous in comparison to other professors. In 1942 he continued to teach well into July even
though the concours had already taken place in mid June.
Two interviews contain useful recollections from Boulez about Messiaen’s harmony
class. 3 In his interview with Jean Boivin, Boulez described a typical class, fleshing out
the physical description of his homework with its division between exercises done in
class and at home :
Here is how he went about it. We had a first class and handed in our homework. In the second
class, which was in the middle of the week, he dictated the text of a new exercise (he dictated
the chant donné or the basse donnée which we were expected to take down by ear). We realized
this lesson in class and he corrected it. [...] The exercise had the double function of developing
our ear and cultivating our mastery of the material without being able to consult an instrument.
Moreover, he tried to have us compose our realizations ; he tried to have us find a musical idea
that would make the realization coherent. We were learning, partially, to compose. 4
In the following excerpt, taken from his 1996 interview with Olivier Mille, Boulez talks
about the difference of Messiaen’s class – a problematic piece of testimony given that
he did not regularly attend other harmony classes. It seems most prudent to interpret
his description as a comparison with the elementary harmony class he took with Dan-
delot the previous year, (a class meant to get students through a basic harmony treatise
and prepare them for the upper level classes), or the counterpoint classes with Simone
Plé-Caussade that Boulez famously abandoned out of frustration with their academic
abstraction. It should also be considered that Boulez seems to speak of what interested
him the most in Messiaen’s harmony instruction, the aspects he felt to be exceptional,
perhaps passing over the more mundane instruction that prepared students to realize
chants and basses in the style expected for the concours.
I ought to say that frankly, in my opinion, he was the only truly creative professor in the classes
d’écriture 5 of that period. The only one. His class, furthermore, was truly exceptional – it didn’t
dictait le chant donné, ou la basse donnée, que nous devions repérer à l’oreille). Nous faisions le devoir
en classe, et il le corrigeait. [...]. L’exercice servait en même temps à développer notre oreille et à cultiver
notre propre domination du matériau, sans références à un instrument. En outre, il essayait de nous faire
composer un devoir, il essayait par exemple de nous faire trouver une idée d’écriture qui allait rendre la chose
cohérente. Nous apprenions partiellement la composition. ».
5 These Conservatoire classes focused on writing music and included harmony, fugue and composition.
boulez and messiaen ’ s harmony class 45
fit within the mold of the other typical classes d’écriture. [...] I think for him it was essential to
communicate his knowledge, as it is necessary for all young people. I think his instruction in
those days really allowed students to see his creative process. The communication was certainly
very direct. It wasn’t just a conventional [...] classe d’écriture. He based his exercises, and we have
the reflection of this in his book La Technique de mon langage musical, on musical texts, which was
not often the case with other professors. The other professors based their instruction on a sort of
harmonic abstraction that had nothing to do with direct sources whereas he, to the contrary, gave
us exercises where one had to have a sense of composition and creation. When he gave us a text
to realize he always told us that we had to invent an idea or several ideas in order to give the re-
alization a sense. [...] In particular he gave us homework that was to be realized in a certain style,
and he analyzed pieces by a composer, [...] and gave us a text that he had composed himself. 1
We have already explained that although Messiaen gave students chants and basses he
had written himself, these exercises were far outnumbered by texts from traditional
harmony manuals. Moreover, we have also noted that Messiaen’s own lessons ‘in the
style of ’ were themselves quite similar to traditional Conservatoire harmony lessons.
The other harmony teachers at the Conservatoire may have made fewer references to
‘real music’ than Messiaen, but Boulez did not attend the four other harmony classes,
which most certainly encouraged their students to identify the appropriate style of
harmonic vocabulary to be used and realize chants and basses by finding a key idea and
developing it to create a well-constructed composition. These practices were indeed
the bread and butter of Conservatoire harmonic instruction and the key to writing a
premier prix-worthy concours.
5. Boulez ’ s Homework
Boulez’s homework displays a broad variety of annotations in Messiaen’s hand, ranging
from simple appreciations of his work to criticisms of exercises that he judged unat-
tractive. Such exercises were partially erased and rewritten – usually by Boulez, but at
times by Messiaen himself. One example will be explored in depth below. Boulez was
mostly immune to basic technical problems, but there are occasional instances of direct
octaves, unresolved sevenths, false relations and problems with tessitura. Messiaen’s
commentaries are more often directed at the framing of modulations and endings –
noting harmonies that might work better or possible adjustments in timing. In one case
his praise goes so far as to indicate « 1er prix », perhaps following a mock exam.
1 Interview with Olivier Mille from 1996 available in the dvd published with Messiaen 2008, cit. : « Je dois
dire franchement, à mon avis, c’était le seul professeur vraiment créatif dans le domaine de l’écriture à cette
époque-là. Le seul. Sa classe, du reste, était une sorte de classe vraiment exceptionnelle, elle n’entrait pas
dans la moule des classes habituelles d’écriture. [...] Je pense que pour lui c’était essentiel de communiquer
son savoir comme tous les gens jeunes ont besoin de communiquer leur savoir. Je pense que son enseigne-
ment à cette époque-là a vraiment fait participer ses élèves à son stade de créativité. Et certainement la
communication était très directe. Ce n’était pas une classe seulement [...] d’écriture conventionnelle. [Il]
s’est basé, on en a d’ailleurs le reflet dans son livre La Technique de mon langage musical, on voit très bien qu’il
s’appuyait sur les textes, ce qui n’était pas le cas, souvent, des autres professeurs. Les autres professeurs s’ap-
puyaient sur une espèce de l’abstraction de l’harmonie qui n’avait plus rien à voir avec des sources directes
tandis que lui, au contraire, il nous donnait des devoirs [...] c’était des devoirs où le sens de la composition,
de la création s’exerçait. Il nous disait toujours, quand il nous donnait un devoir, même un devoir disons
traditionnel comme un chant donné ou une basse donnée, il nous disait toujours : il faut inventer une idée
ou inventer des idées de façon à ce que votre devoir ait un sens. [...] Et en particulier il donnait des devoirs
dans un certain style à faire, il analysait des pièces d’un compositeur, [...] et il donnait un devoir qu’il avait
rédigé lui-même. ».
46 yves balmer · christopher brent murray
The harmonic vocabulary of Boulez’s realizations varies from lesson to lesson –
adapting to the texts he was assigned – but there are certain features that appear com-
monly throughout the budding composer’s homework. These include a fulfillment of
the harmony lesson’s expectation for extreme contrapuntal density, motivic economy
and cohesion, and a marked preference (as will also be seen in Messiaen’s harmony for-
mulas) for long, conjunct chromatic lines that maintain their chromaticism to the point
of nearly rupturing the tonal discourse. These two very ‘horizontal’ qualities meant
that Boulez would avoid the criticism some students received during the concours, de-
scribing their work as « too vertical ». 1 Example 1 2 is one such example of such motivic
imitation (between the given soprano line and Boulez’s tenor line) and long chromatic
lines (here in the bass) among many found in Boulez’s homework.
Boulez’s realization of a chant alterné by Jean Gallon which Messiaen also happens to
have reworked in his own copy of Gallon manual, provides us with the occasion to
compare both basse and chant techniques as realized by Gallon, Messiaen, and Boulez. 3
Although Messiaen considered Boulez’s basse « très bien », he found the realization of
the chant « frightful, not in the style, in contradiction with the given text ». Boulez and
Messiaen’s hands can be seen in a fourth version incorporating aspects of Boulez’s ini-
tial work – Messiaen’s hand intervenes in the final five measures to write an appropriate
ending. He confirmed this new version of the basse as « très bien ». All four realizations
1 an, aj37 536. 2 pss, Sammlung Pierre Boulez, microfilm no. 577, p. 191.
3 See ibidem, pp. 169-170 for Boulez’s realization. Messiaen’s annotated copies of the student and teacher
manuals of Jean Gallon, 60 exercices et thèmes d’harmonie. 2ème série, Paris, Heugel, 1931 (pp. 53-55 of the
teacher’s manual for Messiaen’s alternative realization) are preserved at the Médiathèque Hector-Berlioz
under the shelfmark Rmb. 74 (1-2).
4 Carillons of various sorts abound in Messiaen’s music of the 1940s – from Visions de l’Amen to Vingt
regards to the song cycle Harawi written the year that Boulez studied in Messiaen’s class.
boulez and messiaen ’ s harmony class 47
(continues)
the text, whereas Boulez knows quite well he is expected to use the last phrase of the
bass as his opening melody – the ‘trick’ that every good harmony student had to notice.
Messiaen also eschews the individual entrances of the voices in the chant used by both
Gallon and Boulez, preferring a much slower harmonic rhythm, a relatively homopho-
nic treatment of the bottom three voices, and an abundance of major ninth chords.
Messiaen’s method of reimagining Gallon’s lesson is reminiscent of a realization
found in Boulez’s homework where Messiaen’s second chant donné from the Vingt
48 yves balmer · christopher brent murray
Ex. 2. Reductions of a harmony lesson by Jean Gallon : Chant alterné no. 34 from 60 exercices et
and Pierre Boulez, with corrections in the hands of both Pierre Boulez and Olivier Messiaen
in the final five measures in the latter (pss).
boulez and messiaen ’ s harmony class 49
Leçons, « Dans le style d’un Passepied de Rameau » marked « Vif et gracieux », is trans-
formed into « Comme une Pavane [Ravel] » and marked « Lent ». Boulez realizes Mes-
siaen’s lesson in the style of Ravel, not Rameau. The reworking of an existing melody
in a new harmonic style also happens to be one of Messiaen’s personal compositional
techniques, 1 one which he may have been encouraging his students to consider. Still,
it seems unlikely he would have gone so far as to recommend risking such alternative
realizations for the concours.
Although Boulez played along with the modal atmosphere of Gallon’s conventional
basse, he seems to have lost patience when he got to the chant. Or perhaps he just
wanted to play around. This is particularly evident at the point where the fourth voice
enters in measure 19 creating in a series of parallel fourths in another key, similar to the
opening counterpoint of The Rite of Spring. Even if Messiaen found this original real-
ization « affreux » he worked toward a compromise and retained something of Boulez’s
parallelisms in the final measures of his corrected ending – a middle ground between
the syrupy modality of the Gallon text and Boulez’s seemingly parodistic realization.
If Boulez’s realization was indeed a joke or provocation, Messiaen seems not to have
understood it as such given the care he spent rewriting the ending.
Messiaen’s harmonic formulas echo the ambiguity of style observable in the Vingt
leçons – in many cases their stylistic traits are linked just as much to the nature of their
melodies as to their use of signature harmonies associated with historical periods or
particular composers. In most of the formulas, the soprano is far more active than the
other voices. In fact, if a formula shows rhythmic variation, non-chord tones, phrasing
marks, dynamics, and articulations, these are overwhelmingly found in the soprano.
The other voices are often realized in long note values, often half or whole notes with-
out any further indicated nuances. In adapting the formula from its source, Messiaen
tends to alter the soprano the most. Thanks to its greater variety of movement, it is also
the soprano that often conserves traces of the melodic material of origin : contours,
rhythms, and use of non-chord tones. In this sense, the formulas correspond to the
techniques needed for realizing chants donnés. It is no wonder then, that so many for-
mulas are derived from Franck, whose chromatic voice-leading and quick modulations
to distant keys through alternative resolutions typify the harmonic language needed
to tackle many Conservatoire chants données. The formulas’ origins in late nineteenth
century repertoire has already been observed, but it should also be underlined that
a number of the formulas are drawn from passages that Messiaen transformed in a
very different way when composing his own music. In this sense, Messiaen does indeed
seem to be teaching his students to realize harmony lessons using techniques similar to
those he used to compose. 2
A first example can be found in the « vocal style » demonstrating the use of the Nea-
politan sixth chord from the second Kyrie of Bach’s Mass in B minor – Formula 3, « 6te
napolitaine (Bach, Kyrie, Messe en si min.) ». The formula, adapted from the first mea-
sure of the second Kyrie (see Example 3), is transposed down a half step from F# minor
1 ��������������������������������������������������������������������������������������������
Yves Balmer, Thomas Lacôte and Christopher Murray are preparing a forthcoming volume on Mes-
siaen borrowing and transformation of pre-existing musical material that will be published by Symétrie
editions. The previously cited articles co-authored by Balmer and Murray also discuss facets of this tech-
nique.
2 For more on this link see Yves Balmer, Christopher Murray, De l’harmonie à la composition : Mes-
Ex. 3. Olivier Messiaen’s harmony formula 3, « 6te napolitaine (Bach, Kyrie | Messe en si min.) »
as recopied by Pierre Boulez (pss) ; Bach, Mass in B minor, Kyrie ii, measure 1 ; Messiaen har-
mony formula reduced and transposed to the F-sharp minor to aid comparison.
to F minor. The meter, rhythms, bass-line and melody are all slightly altered. The rising
eighth-note passing motion on the third beat of the Messiaen formula is absent from
the first measure of the Bach model, but seems to have been taken from a similar mo-
tive in the third measure of Bach’s continuo (Ex. 4) – mirroring the manner in which
motives from the bass line were expected to be incorporated into the upper voices in
the realization of basses données.
Although some formulas reproduce the original pitch level of their sources, many
of them, like the formula 3, transposed from F sharp minor to F minor, do not. With
some examples (the adaptation of Frauenlieben und -leben in formula 38 and the numer-
ous adaptations of Fauré mélodies) one might imagine that Messiaen was working from
a score transposed for a particular vocal range, but this hypothesis is invalidated by
examples like the excerpt from the opening theme of Franck’s Symphony in D minor
boulez and messiaen ’ s harmony class 51
Ex. 4. Measures 1-3 of the continuo from the Kyrie ii of Bach’s Mass in B minor with the source
of the melodic motion in Messiaen’s formula indicated.
Ex. 5. Olivier Messiaen, harmony formula 28, « Schumann (Humoresque) » as copied by Pierre
Boulez (pss) ; Robert Schumann, Humoreske, measures 266-269 ; and harmony formula 28
1 The previous collaborations of the present authors as well as their forthcoming volume with Thomas
Lacôte go into greater detail upon this point. See also Yves Balmer, Christopher Murray, Repenser la ré-
ception de Claude Debussy par Olivier Messiaen, in Regards sur Debussy, ed. by Myriam Chimènes and Alexandra
Laederich, Paris, Fayard, 2013, pp. 501-516.
52 yves balmer · christopher brent murray
the genesis of the formulas as having been transposed to the key of particular lessons
for which they were intended to serve as models.
In formula 28 « Schumann (Humoresque) » (Ex. 5) Messiaen transforms music writ-
ten for the piano into a four-part instrumental texture comparable to a string quartet
and appropriate for use in a harmony lesson without transposition, change in meter,
or even much in the way of motivic or rhythmic alteration (the inner voices still enter
on the off beat), apart from the introduction of an imitative figure in the third measure
of the tenor part and a redistribution of scale degrees between the different voices. In
the second part of the example, Messiaen adapts the bass line to follow the pattern
established in the first two bars, creating a descending line where Schumann changes
directions and ascends.
The Schumann phrase from which the formula is derived moves through the rela-
tive major on the way to the dominant, but Messiaen stops short at the relative major
(B flat, via its dominant, F). Although the harmonic device at work is not signaled in
the label as it was in the previous example, the second measure’s ‘borrowed’ dominant
seventh chord in first inversion (or « accord de sixte et quinte diminuée » in French har-
monic parlance) seems to be the focus. This was and still is a harmony typically associ-
ated with Schumann in traditional Conservatoire instruction. 1
Formula 46, labeled, « Franck, Symphony. Wagner, Tristan. Quarte et sixte dans les
modulations lointaines. Lento, molto espressivo » (Ex. 6), is clearly based on the se-
cond part of the opening theme of the Symphony in D minor (measures 6-12), a
passage also marked Lento and espressivo. This is one of the most detailed formulas. Its
dynamics and phrasing (all applied to the soprano voice alone) are clearly based upon
Franck’s phrasing, although they introduce details not found in Franck’s score. For-
mula 46 is also representative of Messiaen’s occasional expansion and contraction of
his models. Although this example corresponds mostly one-to-one with the measures
of the Franck excerpt, the second measure of the Franck passage is expanded into two
measures in the formula. Finally, in spite of slight changes to the bass line and melo-
dy, most of Franck’s pitches and rhythms are present in Messiaen’s formula, at least
in residual form. Messiaen notably retains Franck’s signature appoggiaturas, adding
chromatic passing tones and developing the melody with rising sixths. The C sharp
leading tone resolving to the tonic on the first beat of Messiaen’s formula (which pre-
cedes Messiaen’s version of Franck’s descending motive quickened to eighth notes)
is related to the symphony’s germinal opening motive (Ex. 7). The three-note motive
at the beginning of the second measure of the formula in turn adapts the half step
appoggiatura in the second measure of the Franck excerpt, crossing it with a referen-
ce to the profile of the symphony’s three-note opening motive by replacing Franck’s
rising diminished fourth with a sixth.
1 Desportes and Bernaud also indicate a different passage of Humoresque as a prime example Schuman-
nian harmonies in their chapter on Schumann. « Favorite chord, [accord de sixte et quinte diminuée] with
or without the alteration of the fifth. Harmonic system : to affirm each degree of the scale, [Schumann]
borrows the dominant of each degree, using the first transposition. » (p. 19). Apart from Humoreske, most
of Desportes and Bernaud’s examples on Schumann’s style are reductions of Frauenlieben und -leben. Mes-
siaen’s following Schumann formula 38 « Schumann (Amours d’une femme) » is drawn from « Er der Herr-
lichste von Allen » of the same song cycle. These parallels suggest the possibility of common references
inherited from the harmony class of Jean Gallon : Desportes, Bernaud, Manuel pratique pour l’approche des
Ex. 6. Olivier Messiaen’s harmony formula 46 (« Franck, Symphonie. Wagner, Tristan [...] ») as
copied by Pierre Boulez (pss) ; Franck, Symphony in D minor, first movement, measures 6-12 ;
Symphonie, Wagner, Tristan [...] ») is unclear. Perhaps Messiaen showed his students
examples of similar chord progressions in Tristan und Isolde, or perhaps Messiaen’s me-
lodic variation on Franck peppered with rising sixths is was intended to refer to the
rising sixths that are characteristic of Tristan und Isolde, from the first measures of the
opening prelude to the Liebesverklärungs leitmotif. The chromatic filling-in of the for-
mula’s penultimate measure might also be of Wagnerian inspiration.
A final Example (formula 51 « Imité de Nell – Fauré »), shows Messiaen transforming
the texture of solo voice with piano accompaniment to an instrumental setting, chro-
matically filling the gaps in the voice-leading of the final two phrases of the song « Nell »
from Gabriel Fauré’s op. 18 (Ex. 8). Messiaen has transposed the music up a half step,
from G flat major to G major. The dynamics and phrasing of the formula reproduce
54 yves balmer · christopher brent murray
Ex. 8. Olivier Messiaen’s harmony formula 51 « Imité de Nell – Fauré » as copied by Pierre
Boulez (pss) ; Gabriel Fauré, « Nell », measures 32-40 ; the same formula in closed score and
those of Fauré, and while nearly all of Fauré’s melody and bass line pitches are present.
Messiaen complicates the texture with dissonant appoggiaturas and passing tones to
create long, conjunct lines that feel somewhat strange in comparison to their model.
Similar long chromatic lines are observable in Boulez’s homework. A number of other
formulas adopt a comparable technique of chromatic ‘filling-in’, particularly those in
the styles of Wagner and Franck (see the penultimate measure of the Franck formula
studied above, for example). Messiaen also uses this technique in his own compositions
boulez and messiaen ’ s harmony class 55
Ex. 10. Piano reduction, Pierre Boulez’s realization of a chant donné composed
by Olivier Messiaen (pss). Annotated « Stravinsky », Boulez’s realization uses elements
Mixed among the traditional exercises realized by Boulez are a few realizations that
fall outside the Conservatoire canon, including a setting that clearly refers to Stravin-
sky’s Firebird. Perhaps this sort of extra ‘treat’ was how time was spent in the weeks of
June and July after the concours was over (Ex. 10). 2 The text of the exercise itself was
originally marked « Milhaud » but this indication was erased, crossed out, and replaced
by « Stravinsky » in pencil and « C[hant] D[onné] O. Messiaen » in ink. It may be that this
chant donné by Messiaen was based on a melody found in the music of Darius Milhaud,
but it is clear that Boulez chose to harmonize the melody using elements borrowed
from the opening bassoon solo and its oboe accompaniment in the « Berceuse » of The
Firebird. His imitation even picks up points from Stravinsky’s orchestration such as the
pizzicato bass notes on the first beats of the opening measures. Boulez also varies the
rhythms of the descending chromatic motive in the alto voice in a manner similar to
Stravinsky’s oboe solo. The Stravinsky lesson is one of the few readily identifiable and
explicit examples of Boulez following Messiaen’s lead by borrowing directly from the
repertoire to realize a chant donné.
Although it is not entirely clear how Messiaen used the formulas in his harmony
class, it seems that their reworking of traditional music was intended to serve as an
example to his students. By making direct reference to the repertoire and rework-
ing fragments of pre-existing music, Messiaen also happened to be introducing his
students to his own creative techniques, proposing them as a means for realizing
traditional harmony lessons. We have shown that Messiaen’s harmony class, while
cleaving to Conservatoire traditions, also made explicit references to the repertoire
in connection to the realization of traditional lessons. But the novelty of these refer-
ences is thrown into doubt if one takes into account the lessons written in homage
to the legendary Jean Gallon upon his retirement. The 64 Leçons d’Harmonie offertes
en hommage à Jean Gallon par ses élèves prix d’harmonie entre 1919 et 1948 included lessons
like Jeanne Aïtoff ’s « Etude de réalisation sur les harmonies de la 1ère “Scène de la
Forêt” de Schumann » and Jean Vuillermoz’s « Etude de réalisation sur les harmonies
du prélude en ré mineur du Clavecin bien tempéré (1er livre) de J.-S. Bach ». Although
of his students from the exams that admitted students to the concours in May 1942, at
the end of Messiaen’s first complete year as a professor, are a stark contrast to those
of students in the class of Jean Gallon, singled out for their ‘natural’ and ‘musical’
1 ���������������������������������������������������������������������������������������������������
This transformation is analyzed in detail and compared to similar transformations elsewhere in Mes-
siaen’s work in the forthcoming volume authored by Balmer, Lacôte, and Murray.
2 pss, Sammlung Pierre Boulez, Mappe A, Dossier 2a, 3, folio 7.
3 In the 1940s, harmony teachers submitted evaluations of their students to the administration in January
and again in April or May, just before the exams that admitted students to the concours.
boulez and messiaen ’ s harmony class 57
work. Messiaen’s students received a litany of criticism for clumsy or overly compli-
cated realizations that reads as much as an indictment of the newly hired teacher as
of his charges : « a bit complicated », « inventive, but anti-vocal », « a bit disorganized »,
« lacks clarity, lack of order in the soprano », « less adroit but just as complicated »,
these criticisms were attenuated in the following years’ exams, we have already noted
that Messiaen’s students fared relatively poorly in concours, garnering only three pre-
miers prix in his five complete years of instruction. Perhaps this is part of the reason
that Messiaen was nominated as an analysis and esthetics teacher during the 1947-1948
school year.
Messiaen’s notes on his students reveal the qualities he most admired. They are glob-
ally positive, with his strongest praise split between the terms he used to describe prom-
ising harmonists (« musician », « knows his trade », « finds the right harmony ») and his
designation of « future composers ». The latter group was, not coincidentally, composed
of the students (Pierre Henry, Maurice Leroux, Yvonne Loriod, and Serge Nigg) he
came to know better in private analysis classes (classes which Boulez has unhesitatingly
called « une classe de composition » 2). Although these students, were not always the
most promising in terms of their results in the harmony concours, Messiaen rarely hesi-
tated to lavish them with praise for their creative inclinations. Tables 2 and 3 provide a
sampling of both sorts of encouragement. These comments would seem to underline
how Messiaen the harmony teacher had his eye on those he felt would be future com-
posers – perhaps as much for their own benefit as for his own, given his desire to be-
come one of the Conservatoire’s new composition teachers, an ambition that was only
realized in 1966. 3 Although the harmony class was an important step on the road to en-
tering the composition classes, similar observations are not to be found in the notes of
the other harmony teachers, who limited themselves to describing students’ progress,
the quality of their ears, their musicianship, or ability to find the right harmony. The
case of Serge Nigg reveals Messiaen’s willingness to praise creative promise combined
with his impatience for Nigg’s unwillingness to follow the rules in his harmony lessons.
More importantly, Messiaen’s comments on Nigg reaffirm that he was well aware that
the harmony class existed to prepare and pass a concours, not to promote creativity for
its own sake.
Boulez’s rapid success as a harmony student means that only a single evaluation
from his time in Messiaen’s class survives, brief remarks on the promising future of
an extremely talented student : « Très musicien, est appelé à un bel avenir. ». The rest is
history.
Pierre Henry4 [May 1945] Extrêmement bien doué. Peut donner des résultats magnifiques
s’il est en forme.
[January 1946] Extrêmement doué. Musicien, entend l’harmonie vraie. Ma-
nifeste dans ses devoirs un tempérament particulièrement inventif.
[May 1946] Admirablement doué également. Egalement nature de compo-
stieur.
[Janvier 1947] en congé
[May 1947] Très bien doué, musicien sensible – a été en congé pendant 2
premiers trimestres.
Maurice Le Roux5 [April 1944] Nature enthousiaste ! Très doué et sensible. Futur compositeur.
[January 1946] Enthousiaste, cultivé, intelligent et remarquablement doué.
Un musicien, un compositeur.
[May 1946] Admirablement doué. Nature de compositeur.
[Janvier 1947] Compositeur – Fin musicien – Nature généreuse et sensible –
est appelé à un bel avenir.
[May 1947] Très musicien, très doué, nature généreuse – fait pour la com-
position –
(continues)
1 After having received a premier accessit in 1943 and nothing better in 1944, François Aubut is indicated as
on leave for the 1944-1945 and 1945-1946 school years. Messiaen 2008, cit., pp. 30-32.
2 An amusing play of words on Aubut’s name.
3 René Hanicot was kept in the harmony class an extra year in 1943-1944, but would not improve upon his
second prix of the previous year (ibidem, pp. 30-31).
4 In spite of Messiaen’s prescient praise, Pierre Henry fared poorly in concours and received no particular
distinction during the period from 1944 to 1947 when he studied harmony with Messiaen (ibidem, pp. 31-33).
5 After army service in the Division Leclerc interrupted his studies in 1944-1945, Le Roux received a deux-
ième accessit in 1946 before leaving Messiaen’s class in 1947.
boulez and messiaen ’ s harmony class 59
Yvonne Loriod [January 1942] Extrêmement douée. Très musicienne. Esprit particulière-
ment inventif et original. Doit faire la composition. Travaille avec une fer-
veur digne de tous éloges.
[May 1942] Très musicienne. Intelligente et réfléchie – fait souvent des trou-
vailles inattendues, originales et raffinées. Extrêmement bien douée, tem-
pérament de compositeur.
[January 1943] Merveilleusement douée ! et en plus extrêmement travail-
leuse – très musicienne ; tempérament harmonique et mélodique tout à fait
personnel.
[May 1943] Très travailleuse. Prodigieusement douée ! Tempérament in-
ventif et poétique. Futur compositeur.
Serge Nigg1 [January 1942] Très intelligent – Tempérament de compositeur – Esprit cul-
tivé – Travaille avec une grand régularité et a fait du progrès depuis octobre
–
[May 1942] Très doué, extrêmement travailleur. Recherche de contours ci-
selés et les harmonies rares. Esprit inventif, avide de nouveauté. Progresse
constamment. Tempérament de compositeur.
[January 1943] Tempérament de compositeur – Travaille avec régularité.
– Malheureusement réfractaire à toute loi ou code harmonique ! Avec des
dons aussi prononcés, c’est dommage...
[May 1943] Futur compositeur. Très doué. Malheureusement beaucoup de
mal à obéir aux règles, pourtant si utiles !
[Janvier 1944] Tempérament de compositeur. Une nature, une «personna-
lité» !
[April 1944] Sera certainement dans quelques années un de nos meilleurs
compositeurs !!
Table 3. Messiaen’s praise for ‘future composers’.
1 Like Pierre Henry, Serge Nigg would go on to have a successful career, but did not fare well in concours
during his studies with Messiaen from 1941 to 1944 (ibidem, p. 30).
c o mpos to in car atter e dan t e mon ot y p e d a l l a
fabrizio serr a editor e, p i s a · r oma .
s tamp ato e rileg at o n e l l a
t ipog r afia di ag n an o, ag n a n o p i s a n o ( p i s a ) .
*
Dicembre 2014
(cz 2 · fg 3)
Periodico annuale · A Yearly Journal
*
Amministrazione e abbonamenti · Administration & Subscriptions
Fabrizio Serra editore®
Casella postale n. 1, Succursale n. 8, i 56123 Pisa,
tel. +39 050 542332, fax +39 050 574888, fse@libraweb.net, www.libraweb.net.
*
Autorizzazione del Tribunale di Pisa n. 17 del 15.09.2004
Direttore responsabile: Fabrizio Serra
Per la migliore riuscita delle pubblicazioni, si invitano gli autori ad attenersi, nel predisporre
i materiali da consegnare alla Redazione ed alla Casa editrice, alle norme specificate nel
volume Fabrizio Serra, Regole editoriali, tipografiche & redazionali, Pisa · Roma,
Serra, 20092 (ordini a: fse@libraweb.net).
Il capitolo Norme redazionali, estratto dalle Regole, cit., è consultabile Online
alla pagina «Pubblicare con noi» di www.libraweb.net.
issn1 824-6206
issn elettronico 1826-7858
isbn 978-88-6227-740-2
SOMMARIO
IMMAGINI DI GIOVENTù.
SAGGI SULLA FORMAZIONE
E SULLE PRIME OPERE DI PIERRE BOULEZ
A cura di Paolo Dal Molin
duzione 15
Yves Balmer, Christopher Brent Murray, Pierre Boulez and Olivier Messiaen’s
Harmony Class 31
Luisa Bassetto, Ritratto del compositore come apprendista etnologo : Pierre Boulez
ment iv 95
Pascal Decroupet, Séries, timbres artisanaux et objets sonores : Pierre Boulez et la