FROM DEBUSSY'S STUDIO: THE LITTLE KNOWN AUTOGRAPH OF "DE RÊVE", THE FIRST

OF THE "PROSES LYRIQUES" (1892)

Author(s): Denis Herlin and Peter Bloom
Source: Notes , September 2014, Vol. 71, No. 1 (September 2014), pp. 9-34
Published by: Music Library Association

Stable URL: https://www.jstor.org/stable/43672861

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Music Library Association is collaborating with JSTOR to digitize, preserve and extend access to
Notes

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
 FROM DEBUSSY'S STUDIO: THE LITTLE-
KNOWN AUTOGRAPH OF DE RÊVE, TH
FIRST OF THE PROSES LYRIQUES (1892
By Denis Herlin

S
m
c
r
sole completed opera, based on the eponymous play by Maurice
Maeterlinck. Evidence of this experimental attitude is to be found in the
existence of more than one finished manuscript for the very same
mélodies . With the exception of some of the Anettes , the differences
among such sources concern less the structure of the songs than the re-
finement of their writing for voice and piano. Such in particular is the
case of the Proses lyrìques , for three of which we possess, in addition to the
autograph manuscripts that served the engraver of the published score,
four further manuscript versions.1 With special attention to the specific
case of "De rêve," the first of the four Proses lyúques , it is my intention
here to offer close examinations of some of the relevant texts and logical
explanations of the reasons that led Debussy to modify them. But before
looking into the nature of his transformations, I should like first to de-
scribe the circumstance in which Debussy undertook the composition of
this cycle of mélodies.
In 1892, when he began to conceive a new song cycle entitled Proses
lyúques, Debussy was hardly a well-known composer. Even though he had
won the Grand Prix de Rome - the highest honor that a graduate of the
Paris Conservatoire could receive - at the relatively early age of twenty-two,
the young man, still known at the time (1884) as Achille, was determined

Denis Herlin is directeur de recherche at the IReMUS (Institut de recherche en musicologie), a subdivi-
sion of the CNRS (Centre national de la recherche scientifique), headquartered at the Bibliothèque na-
tionale de France, in Paris. He is author of several major catalogs, including the Collection musicale
François-Lang (1993) and the Catalogue du fonds musical de la bibliothèque de Versailles (1995), and he is co-
editor (with Sylvie Bouissou and Pascal Dénécheau) of the ongoing Catalogue thématique des œuvres de
Jean-Philippe Rameau (3 vols, published). With François Lesure, Herlin collected and edited the first com-
plete edition of Debussy's letters - Claude Debussy, Correspondance 1872-1918 - a volume brought to
completion by Herlin after the untimely death of François Lesure in 2001. In 2002 Herlin became
editor-in-chief of the Œuvres complètes de Claude Debussy. From 2009 to 2011 Denis Herlin was president of
the Société française de musicologie.
1. For a complete description of these different manuscript sources, see below. The various manu-
script versions of the Proses lyrìques will appear in the appendix of the Œuvres complètes de Claude Debussy ,
série II/4, ed. Denis Herlin (Paris: Durand, forthcoming in 2015).

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
10 Notes, September 2014

not to follow the traditional tr

the time.2 Returning to Paris i
years at the Villa Médicis, hom
not attempt to find regular pro
tal. He rather followed a somew
that François Lesure, in his cel
propriately calls "les années boh
ularly enjoyed confabulation
which had long been hotbeds o
educated man, he had tremen
become as seriously interested
poetry and philosophy. The fre
ciety provided great nourishme
indirectly to his pursuit of new
Afternoon of a Faun and the Str
year as the Proses lyHques , all t
Piano, Debussy's oeuvre in its en
lectual stimulation he experien
In 1911, in a letter to his fr
looked back upon those years
friend, because in the sad 'facto
refreshed my memory of the
Quartet, a period which did n
nonetheless a truly golden age."
When Debussy took up the Pro
experienced composer of art so
student at the Conservatoire, h
songs we have from his pen, m
ing his lifetime, with the consp
of the poem by Theodore de Ban
would see publication, in June 1
poets who were popular at the
de Lisle as well as Paul Bourge
poems he set in those years.

2. In December 1889, Debussy modified h

Debussy. In March 1892, he definitively aba
1918), ed. François Lesure and Denis Herlin (P
3. François Lesure, Claude Debussy: Biograp
2003), 97-107. An earlier edition, with abbrev
4. Correspondance, after 15 March 1911,
représente en ce moment votre mot a rafra
pas extrêmement doré, mais c'était tout de m
5. Paris: Bulla (Société Artistique d'Editions

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 1 1

Verlaine, in 1882, when Verlaine was by no means as well-known as he

later became.6 Indeed, at that time Verlaine belonged to the category of
what we might call the "politically incorrect" poet, with a scandalous rep-
utation. After setting five of the poems from Verlaine's Fêtes galantes of
1869, and six of the poems from his Romances sans paroles of 1874 (which
Debussy published in 1888 under the title of Anettes ), the composer
abandoned Verlaine and turned his attention to another poet whose
work he held dear: Charles Baudelaire. From Les fleurs du mal, Debussy
selected five poems, and worked on their settings for some two years,
from 1887 to 1889, at the same time that he was working on his mystical
and even pagan oratorio La damoiselle élue , on a text by the Pre-
Raphaelite poet Dante Gabriel Rossetti.
Debussy's Cinq poèmes de Charles Baudelaire is a cycle in which one
clearly hears the influence of Wagner, something that is hardly surpris-
ing in view of the fact that Debussy made pilgrimages to the Wagner fes-
tival in Bayreuth in the summers of both 1888 and 1889. The Baudelaire
settings present certain qualities that stand in distinct contrast to those
of his earlier mélodies , qualities that we also hear in the Proses lyHquesr. first,
a piano part of a density that is almost orchestral - and Debussy did in-
deed orchestrate "Le jet d'eau," the third of the Cinq poèmes, in 1907; sec-
ond, compositions of uncommonly grand dimension, most notably "Le
balcon," the first of the Cinq poèmes, of a duration of 131 measures! ("Le
jet d'eau" is 97 measures long.) Published by subscription in March 1890
and issued in a run of 150 copies, of which 50 were printed on special
Holland paper, the first edition of the Cinq poèmes de Charles Baudelaire
carries no indication of the name of the publisher and no indication of
the place of publication. The graphic layout of the publication is thus to-
tally unusual for its time: it is of a large format (37 cm x 28.5 cm); its
cover is printed on simulated parchment; its titles are printed in blue,
golden yellow, and brown; its margins are wide and its lettering is spa-
cious; and it is bound in such a way that each new song begins on a fresh
page. In short, this is an edition whose physical and visual properties well
demonstrate the aesthetic delicacy and refinement that preoccupied
Debussy throughout his lifetime.
Before setting to work on the Proses lyňques , Debussy returned to Ver-
laine in 1891 for two cycles of three mélodies each: first, the opening series
of Fêtes galantes (whose poems he had thought about as early as 1882): "En
sourdine," "Fantoches" (2d version), and "Clair de lune" constitute a
cycle that would not see publication until 19037; second, a cycle inspired

6. The first poem by Verlaine that Debussy set to music is his first version of Fantoches:, the manuscript
is dated 8 January 1882.
7. Paris: Fromont, 1903.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
 12 Notes, September 2014

by Verlaine's collection called

belle," "Le son du cor s'afflig
these, too, not published until
compose the four Proses lyrique
a totally new kind of adventur
write the poetry that he wou
knowledge, neither Fauré nor S
poetry for musical settings of the
Many years later, in March 1
Fernand Divoire - "What shou
Silly poetry? Free verse? Prose
cepts that were important to h
work on the Proses lynques : "Of
there is not all that much. Wh
when one finds it, one ought no
rich, classic poetry, cannot be set
itself, which becomes rather an
rhythmic prose, however, the m
own path."10 If I quote Debussy
one of the particular qualities of
Debussy deliberately chose to u
all metric constraints, whether
Alexandrines, or uneven lines of
and Baudelaire. Indeed, by asser
over classical poetry, Debussy u
the Symbolist aesthetic. For one
vestigated was precisely the libe
ditional meter. It is not by chan
lynques appeared in a magazine
("political and literary intervie
zine opened its first issue with
from the pen of the Franco-Am

8. Paris: Hamelle, 1901.

9. What Debussy says is false, because Régnier was set to music by both Albert Roussel {Quatre poèmes,
op. 3, 1906: "Le départ," "Voeu," "Le jardin mouillé," "Madrigal lyrique"; Quatre poèmes, op. 8, 1907:
"Adieux," "Invocation," "Nuit d'automne," "Odelette"; La menace, op. 9, 1908) and Gabriel Fauré
( Chanson , op. 94, 1906). Even Raoul Bardac, Debussy's student and future stepson, set Régnier to music
( Tel qu'en songe, 1901: three songs dedicated to his mother, Debussy's future wife; and Trois mélodies,
1910).
10. Claude Debussy, Monsieur Croche et autres écrits, rev. ed., ed. François Lesure (Paris: Gallimard,
1987), pp. 206-7: "Les vrais beaux vers, il ne faut pas exagérer, il n'y en a pas tant que ça. Qui en fait au-
jourd'hui ? Mais quand il s'en trouve, il vaut mieux ne pas y toucher. Henri de Régnier, qui fait des vers
pleins, classiques, ne peut pas être mis en musique. Les vrais vers ont un rythme propre qui est plutôt gê-
nant pour nous. . . . Avec la prose rythmée, on est plus à son aise, on peut mieux se retourner dans tous
les sens." There are various editions and English translations of these essays.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 13

with Henri de Régnier, was one of the founders of Entretiens. But at the
beginning, in April 1890, the founders had no intention at all of publish-
ing collections of poetry. It was not until issue 28, of July 1892, that
Régnier and Vielé-Griffin decided to introduce poetry into the maga-
zine, in a small way, with La mort , by Emile Verhaeren. Then, after
Laforgue and Whitman, we find in December 1892 Pour Varrà, by Marius
André, a French poet writing in langue d'oc (Occitan), and the first two
Proses lynques , "De rêve" and "De grève" by C. A. Debussy (see fig. 1 ) .
According to Louis Laloy, it was Henri de Régnier who went over the
texts of Debussy's first poems, as we learn from a passage in Laloy s La
musique retrouvée (1928), where the author notes that Régnier told him
that Debussy had specifically asked about the text of the Proses lynques'. "It
was not by chance that, immediately on returning from Rome, he alone
among his compatriots felt compelled to seek out the finest writers and
to consult Henri de Régnier, for example (who later told me the story),
on the texts of his Proses lynques , when Catulle Mendès was there and per-
fectly happy to assist."11
The title that Debussy gave to the four-song collection, Proses lynques , is
a clear indication of his heightened concern for texts appropriate to mu-
sical setting, for he in no way wished to be enslaved by someone else's
versification. On the contrary, he wanted to find rhythmic prose that fit
his own intimate, lyrical impulses. His mistrust of others' verse structures
surely explains, for example, the aversion he felt to Rodńgue et Chimène,
the pompous verse libretto by the then celebrated writer Catulle
Mendès. Nonetheless, before turning to the subject of Pelléas et Mélisande
in May and June of 1893, 12 Debussy did indeed agree to write the music
for Mendès's libretto, and he completed three acts of the score. But he
finally abandoned work on the opera in July 1893. Further proof of his
antipathy to traditional poetic verse comes in a letter to his friend Pierre
Louýs, with whom he had begun an opera entitled Cendrelune (Cinde-
rella): "My dear friend, first of all, thank you," Debussy wrote: "And now
let me tell you that, as far as music is concerned, I prefer rhythmic prose
to rhymed and metrical poetry. Rhythmic prose comes over in a more
lively fashion, and does not limit the composer in any way. Besides, music
and poetry are like two songs that try, but fail, to get themselves in tune.

11. Louis Laloy, La musique retrouvée 1902-1927 (Paris: Librairie Plon, 1928), 121: "Ce n'est pas le
hasard qui, dès son retour de Rome, lui avait fait rechercher, seul de ses camarades, l'élite des écrivains,
consulter par exemple Henri de Régnier, qui me l'a raconté depuis, sur son texte des Proses lynques,
quand Catulle Mendès était là, tout prêt à l'entreprendre."
12. See Denis Herlin, " Pelléas et Mélisande aux Bouffes-Parisiens," in Pelléas et Mélisande cent ans après:
Études et documents, ed. Jean-Christophe Branger, Sylvie Douche, and Herlin, pp. 41-57 (Lyon: Symétrie,
2012).

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
 14 Notes, September 2014

môsEs LYÊumms

Vw' A Ràyuoxd Bomoroa.

' -.4 T.:Btoào«p.
- *;♦
La irait a de» do
Et lei vieux arbre
A oelie qui vient d
Maintenant navrée ! petite» fille» wjtent de l'éeote,
A Jamais navrée I Mtmi les firmaron» de leur robe
IÜin'ontpas sa M faire signe..... Soie ▼«*** teiste I

Tontes tTontes
Lestes? ont imtté k®*
t ont nuage»#
imtté gràvis voyagenrs
Se eoneertenfirar moãufak ôi*ge
le moãufak
Les Folles' c'est on load vraka^trop grt™
Semant leur rire au gazon " grêle» * **"' " çâS^
" * **"' " jje savent plos où -
An* brises frótense» Car yoW la méchante a verse
La caresse oharmetwe Froos-f roo* d» iwom en volée»
De* hanches fleurissantes I Sol* verte affolée f
Hélas 1 de tont œci, pins rien qa'un blanc frisson

L. & ,0M!
ÄÄÄS la f
Maintenait temia I A ce tiède et blane baiser.... ^
A jamsis terni» I A ce tiède et blane baiser.... ^
Le» chevillera »ont morte aur 1« «tamia dn Grâall Paíg ~ìm riea
- Ä .. . . . . , Pin» qne les cloehes attardées
^Ï&sîssï - Ä .. . . . . , arrus» qne
ìSiìfm!!
An twnp où
D'étrange» lssépé^chantaient
soupirs s^loipour
ťél« Vea t s^loi pourarbre«
Ellea!
Mon ime! c'est da rêve ancien qui ťétreint! •

Fig. 1. Debussy, "De rêve" and "De soir

(December 1892), pp. 269-71 (th

And even in those rare cases wh

bad pun."13
This search for a kind of prose that was at once rhythmic and lyrical
is mentioned again, during an interview with Victor Segalen that took
place on 8 October 1907: "Gabriel Mourey's prose is not very lyrical;
many passages fail to call forth musical ideas."14 And yet another year
later, he wrote again to Segalen, speaking of Orphée-Roi , that "sometimes
the rhythm [of the writing] was more literary than lyrical." "To better ex-
plain myself, I would cite, if I had the patience to do so, so many pages in

13. Correspondance, 10 April 1895, p. 249: "Ami: d'abord merci. . . . Puis; j'aime mieux la prose rythmée
que les vers (au moins pour la musique) ça rebondit davantage et on n'est plus arrêté par des tas de con-
sidérations. D'ailleurs, la musique et les vers, c'est deux chansons qui cherchent vainement à s'accorder,
même dans les cas très rares où cela s'accorde, ça fait l'effet d'un mauvais calembour."
14. Correspondance, p. 2201: "La prose de Gabriel Mourey n est pas tres lyrique, et beaucoup de pas-
sages 'n'appellent' pas précisément la musique."

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 15

Chateaubriand, Victor Hugo, and Flaubert, where we find flamboyant

lyricism but, in my view, no real music. This is a fact that the literary
people prefer not to admit, because it is in the end a mystery that is
inexplicable."15
That Debussy chose the title Proses lynques for his self-authored collec-
tion of mélodies is thus highly significant: the words were chosen first and
foremost for their power to call forth music. Indeed, as Debussy finished
work on the collection, he was beginning the composition of Pelléas -
another excellent instance of "prose lyrique," if we are to believe what
the composer said to Segalen on 6 May 1908. Segalen had told the com-
poser that "only you can offer me guidance on this matter, because, with
Pelléas , you have had a great deal of experience with prose lyrique ."
Debussy replied, "no, not experìence , not at all. I had instinct , and that was
everything."16
Each of the four songs of the Proses lynques - "De rêve," "De grève,"
"De fleurs," and "De soir" - is dedicated to a close friend. "De rêve" is
dedicated to Vital Hocquet, a sometime poet who wrote under the pseu-
donym of Narcisse Lebeau and who was otherwise a plumber!17 Debussy
had met the man when he was playing at the cabaret Le chat noir in the
1890s, and even served as a witness to his marriage in 1892. In November
1892, when he was completing the manuscript of "De rêve," Debussy
inscribed a dedication to Hocquet on a copy of his Cinq poèmes de Charles
Baudelaire (1890). The text of the dedication, in red ink, is in the form of
an in-joke based on a curiously underhanded exchange the two had had:
"To Vital Hocquet, who swiped my fish, but for whom I retain a very
special friendship" (fig. 2). 18

15. Correspondance, 28 August 1908, p. 1111: "Pour mieux m'expliquer je vous citerais - si j'en avais la
patience - des pages de Chateaubriand, V. Hugo, Flaubert que l'on trouve flamboyantes de lyrisme, et
qui ne contiennent - à mon avis - aucune sorte de musique. C'est un fait que les littérateurs ne voudront
jamais admettre, parce qu'il est plein d'un mystère qui ne s'explique pas."
16. Correspondance, p. 2205: "D'ailleurs vous seul pouvez me guider là-dessus, puisque vous avez, avec
Pelléas, une grande expérience de la prose lyrique." Debussy - "Non pas d'expérience, pas du tout. De
l'instinct, voilà tout."
17. See Steven M. Whiting, Satie th e Bohemian: From Cabaret to Cancert Hall, Oxford Monographs on
Music (Oxford; New York: Oxford University Press, 2002), 105.
18. The original reads: "à Vital Hocquet / qui me déroba mes poissons, mais pour qui je garde une
spéciale amitié." In his souvenirs, published with the assistance of Marius Richard ("Souvenirs confiés à
Marius Richard," La liberté, 11-13 December 1931), Hocquet explained the meaning of Debussy's dedica-
tion: "At the rue de Londres [where Debussy lived between June 1892 and July 1893], Debussy's friend
Lebeau [that is, the author of these souvenirs] permitted himself to make off with a Japanese silk hanging
that he still remembers today. It was painted with a stream, tree trunks, and carp [the "poissons" or "fish"
of Debussy's dedication]. Lebeau admitted his theft, and after a discussion, it was agreed that in ex-
change for the Japanese silk he would give Debussy an umbrella! Happy to have the umbrella, Debussy,
in addition to the silk, gave Lebeau a copy of the Cinq pobnes de Baudelaire [with the dedication we have
cited]." This copy was sold at auction, at the Hôtel Drouot, on 7 November 2012 (No. 133). Debussy also
offered Lebeau a photograph of himself, with a dedication on the back that he set down after the
printed indication "procédé inaltérable" (platinotype processing). Taking those words literally ("a
process that is immutable"), Debussy wrote: ". . . ainsi que mon amitié à V. Hocquet" (like my friendship
for V. Hocquet), and he signed: "CLAd. Debussy Août [18]91."

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
 16 Notes, September 2014

Fig. 2. Debussy, dedication to Vital Ho

of the Cinq poèmes de Charles Baudel

The three other songs are dedi

support to the composer betw
leged witnesses to the genesis
to the composer Raymond Bo
Prelude to the Afternoon of a
Jeanne Chausson, the wife of
(no. 4) is dedicated to the painte
law, with whom Debussy main
wrote to Chausson, at the begin
"finished the last of the Proses l
first of all because it gives m
cause I wish not to leave our close circle of friends."19 We know that the

two final songs, "De fleurs" and "De soir," were completed during t
summer of 1893, but the genesis of the first two is less clear. Their com
pletion must nonetheless be contemporary with the publication of the
poems of "De rêve" and "De grève" in the Entretiens politiques et littérair
of December 1 892.

19. Correspondance, 3 September 1893, p. 156: "Je viens de terminer la dernière Prose lyrique, dédiée à
H. Lerolle, d'abord pour me faire plaisir, ensuite pour ne pas sortir d'un cycle d'amitié."

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 17

Let me now briefly describe the known manuscript sources for the
Proses lynques. We have one manuscript of the four songs, the "engraver's
copy" or Stichvorlage , that is on deposit in the Robert Owen Lehman
Collection at the Morgan Library in New York City. The title page of the
manuscript bears the autograph indication "premier cahier / de proses
lyriques / C. A. Debussy."20 Folio 2r bears the dedication: "à Henry
Lerolle. / pour l'assurer de mon / amitié dévouée / Claude Debussy"
(to Henry Lerolle, to reassure him of my devoted friendship).21 In addi-
tion to this handsome manuscript, we have further sources for three of
the four mélodies. A manuscript of "De rêve" is preserved in the collec-
tions of the Juilliard School, which I shall comment on in detail. Folio lr
serves as the title page, marked in black ink, "Premier Cahier de Proses
Lyriques / (1) / Claude Debussy"22 (fig. 3). The manuscript earlier be-
longed to the composer Marcel Labey, who taught at the Schola
Cantorum in Paris, and who was a close friend of Vincent dTndy.23
To my knowledge there is no further source for "De grève," the second
of the four songs of the Proses lynques , but there are further sources for
the third, "De fleurs," which Debussy offered to Jeanne Chausson on her
name day, 24 June 1893. One manuscript is preserved in the Musashino
Ongaku Bunko, in Tokyo.24 On folio 2r Debussy inscribed in black ink a
lovely dedication to Jeanne Chausson: "à Madame E. Chausson / pour sa
fête et pour rendre respectueusement / hommage au charme qu'elle
met à / être Madame Chausson" (to Madame E. Chausson, on her name

20. The indication "premier cahier" shows that as early as 1892-93 Debussy intended to compose a
"deuxième cahier" of Proses lyriques, which he would entitle Nuits blanches. He began work on these songs
in 1898, but of the five projected mélodies only two were completed, and these were not published during
his lifetime. See Denis Herlin, "Des Proses lyriques aux Nuits blanches ou Debussy et la tentation poétique,"
in La note bleue: Mélanges offerts au Professeur Jean-Jacques Eigeldinger, ed. Jacqueline Waeber, pp. 299-328
(Bern: Peter Lang, 2006).
21. New York, The Morgan Library, D289.P966. This manuscript, formerly in the collection of the
Lerolle family, was purchased at auction in Neuilly-sur-Seine on 22 June 1999 (No. 94); it consists of six-
teen folios of music of twenty-eight staves, 35 cm x 27 cm, with the lozenge-shaped, embossed stamp of
"Lard Esnaut / Paris / 25, rue Feydeau." The musical notation is in black ink from fol. 2v to fol. lOr, and
from fol. 1 lr to fol. 12v, and at fol. 13r.
22. New York, The Juilliard School. This manuscript, formerly in the collection of Marcel Labey, was
purchased at auction in Paris on 28 May 2003 (No. 578). It consists of six folios of music on paper of
twenty-eight staves, 35 cm x 27 cm, with the lozenge-shaped, embossed stamp of "Lard Esnaut / Paris /
25, rue Feydeau." The musical notation is in black ink, from fol. 3r to fol. 5v, and carries additions in
both red ink and black pencil. The manuscript may be seen at: http://www.juilliardmanuscriptcollection
.org/composers.php#/works/DEBU (accessed 14 May 2014).
23. We do not know how Marcel Labey (1875-1968) came into possession of this manuscript. A stu-
dent at the Schola Cantorum, Labey became professor of piano at that institution and assisted Vincent
d'lndy in organizing its concerts. He was also general secretary of the Société nationale de musique. In
April 1903 he invited Debussy to participate in a concert at the Schola Cantorum. See Correspondance,
p. 724.
24. This manuscript, formerly in the collection of the Chausson family, was offered for sale by
Musikantiquariat Hans Schneider, Tutzing, catalog 225 (1978), no. 57. It consists of five folios of
music of twenty-eight staves, 35 cm x 27 cm, with the lozenge-shaped, embossed stamp of "Lard Esnaut /
Paris / 25, rue Feydeau." The musical notation is in black ink from fol. 3r to fol. 5r, and carries additions
in red ink. This manuscript differs frequently from the Stichvorlage, particularly in the piano part.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
18 Notes, September 2014

Fig. 3. Title page of thejuilliard manu

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 19

day, and in respectful homage to the charm she displays as Madame

Chausson). Another - which also includes "De soir" - is preserved in the
collections of the Bibliothèque nationale de France.25 This one was of-
fered by Debussy to Marie Fontaine "en hommage à sa voix si délicieuse-
ment musicienne" (in homage to her deliciously musical voice), with the
dedication inscribed on the upper right-hand side of the title page,
which also bears the mention: "1- Cahier de proses lyriques. / proses =
3 = 4 = / Claude Debussy." (Marie Fontaine was the third daughter of
Philippe Escudier, she was Jeanne Chausson 's sister and Madeleine
Lerolle's sister, and she was the sister-in-law of the composer Ernest
Chausson and the painter Henry Lerolle).26 These mélodies are thus obvi-
ously connected to a circle of friends that turned around the families of
Chausson, Lerolle, and Fontaine. Comparing the three manuscripts to
the Stichvorlage at the Morgan Library, we see that while they offer com-
pleted versions of the songs, they all differ in important ways from the
Stichvorlage , as becomes apparent on close observation.
Nevertheless, on first hearing, it is difficult to discern the differences
between the version of "De rêve" preserved in the Juilliard manuscript
and the version traditionally sung and recorded, which is the version
published by Fromont, in 1895, with a lovely art-nouveau cover that
clearly evokes the "ennui si désolément vert" (the desolate green bore-
dom) of no. 3 of the Proses lyriques. Fromonťs publication is based on the
Morgan Library Stichvorlage. The Juilliard manuscript seems to include
two measures more than the Stichvorlage : the first added measure comes
at the end of the song, where Debussy prolongs the resonance of the
final chord not by means of a fermata but by means of an added chord,
which is furthermore disposed differently from the chord in the
Stichvorlage (see fig. 4a, mm. 100-101; and fig. 4b, m. 99).
The second added measure is found between mm. 47 and 48 of the
Stichvorlage , where Debussy reiterates (in the Juilliard manuscript) th
opening two measures of "De rêve" (see fig. 5a) . In the Stichvorlage , th
"introductory" measure, as it were, occurs only once (see fig. 5b). B
close examination of the Juilliard version reveals other differences as we
In the Stichvorlage (m. 47), Debussy removed the left-hand octaves on th
second, third, and fourth beats. In order to play the Juilliard version, th

25. Paris, Bibliothèque nationale de France, Département de la Musique, Ms. 8642. This manuscript,
formerly in the collection of Marie Fontaine, consists of fourteen folios of music of twenty-six stav
30 cm X 40 cm, cut in half, with the lozenge-shaped, embossed stamp of "Lard Esnaut / Paris / 25
rue Feydeau." The musical notation is in black ink from fol. lv to fol. lir, and carries a few additio
in black pencil. The manuscript may be seen at: http://gallica.bnf.fr/ark:/12148/btvlb857222
.r=Debussy+proses+lyriques.langFr (accessed 14 May 2014).
26. To Henry Lerolle himself, Debussy offered a sketch of "De soir," which is also preserved in the col-
lections of the music department of the Bibliothèque nationale de France (Ms. 19183).

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
20 Notes, September 2014

Fig. 4a. Debussy, "De rêve," Juilli

pianist almost needs a third h

closely at fol. 4r of the Juillia
chords in the left hand were s
added after the fact (see fig. 6)

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 21

Fig. 4b. Debussy, "De rêve," Stichvorlage (Morgan manuscript), mm. 88-99

eration of the opening measures could have been accomplished without

the support of notes in the bass, as we have at the opening of "De rêve,"
as well as at mm. 66 and 67 of the Juilliard version, where there is like-
wise no bass.27
How are we to interpret those chords notated in red ink? In my opin-
ion, they were added with a view toward the future orchestration of the
27. In the end, at m. 66 in the Stichvorlage (and thus in the published version), Debussy maintained a
whole-note F|, in octaves, in the bass.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
22 Notes, September 2014

Fig. 5a. Debussy, "De rêve," Juilliard

Fig. 5b. Debussy, "De rêve," Stichvo

(piano part only)

song. Such a hypothesis is supp

tom of the same folio, below
note D in octaves, with a trill, and with an indication - the number "1"
and the letter "V" - that stands for "premier violon." In a letter to the
eminent violinist Eugène Ysaýe dated 13 October 1896 - about one year,
that is, after the publication of the Proses lyrìques - Debussy announced
that he had "begun to orchestrate two of the Proses lyriques ."28 It may well
be that from the very beginning, Debussy had in mind an orchestral ver-
sion of "De rêve." This is even more intriguing in that we have so far dis-

28. Correspondance , p. 326: "J'ai aussi commencé l'orchestration de deux proses lyriques. . . ."

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 23

Fig. 6. Debussy, "De rêve," fol. 4r of the Juilliard manuscript, third and fourth systems
(New York, The Juilliard School)

Fig. 7. Debussy, "De rêve," fol. 4r of the Juilliard manuscript, fifth system,
last two measures (New York, The Juilliard School)

covered only one manuscript of the orchestration of the first twenty-five

measures of "De grève."29 There are several further orchestral indica-
tions in "De soir." These, notated in red ink, may be found in Debussy's
29. This manuscript, preserved in a private collection, was offered on 2 January 1923 by Emma
Debussy, the composer's second wife, to Roger-Ducasse.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
 24 Notes, September 2014

Fig. 8. Debussy, "De rêve," fol. 5r of

last two measures (New York, T

own copy of the first edition o

Be this as it may, Debussy did
trating the Proses lyriques until 1
To return to the markings ad
places, we must wonder if the
tion of the song. In mm. 87 an
we see a short motif that does
the characteristic rhythm of
thirty-second-note, in a gestur
fifth, this motive, with a kind
role through to the end of th
when the "golden helmets" of
mm. 63-65, when the knights
again in m. 80 and in mm. 83-
ined medievalism, and it serves
ultimate reminiscence without words. The fact that this motive is notated

in red ink may well indicate that Debussy conceived it for use only in th

30. This copy is preserved in a private collection in Paris.

31. The last time he mentioned the work in correspondence appears to have been in February 1901
in a letter to Pierre de Bréville ( Correspondance , p. 587); and in April 1901, in a letter to Blanche Mar
( Correspondance , p. 593). Still, a few years earlier Debussy seems to have deemed it useless to enlarge the
songs by means of any orchestral "fracas" whatsoever: "inutile de les augmenter d'un fracas orchestra
quelconque"! See Correspondance , letter to Pierre de Bréville, 24 March 1898, p. 394.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 25

Fig. 9. Debussy, "De rêve," Juilliard manuscript, mm. 84-89

orchestral version, to be played, perhaps, by French horns - a hypothesis

supported by the chordal writing of these two measures (see fig. 9).
After looking at the most visible differences between the Juilliard man-
uscript and the Stichvorlage , I wish now to consider another difference be-
tween the two, and that concerns the tempos. Figure 10 compares the
tempo indications of the two manuscripts. The Stichvorlage carries five
more tempo indications than the Juilliard manuscript, especially as con-
cerns rallentendos: in the Juilliard manuscript, there is one such indica-
tion, at m. 65; in the Stichvorlage , there are four, at mm. 42, 46, 59, and
64. Such additions of rallentandos must be viewed together with the
other tempo indications in the two manuscripts. In the Juilliard manu-
script, the initial tempo marking is "Lent. (Tempo Rubato)," replaced in
the Stichvorlage by the single word "Modéré." Presumably, modéré leads to
a tempo that is faster than lent and rubato. This explains the necessity of
such indications as retenu , très retenu , and plus lent , which occur in the

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
26 Notes, September 2014

Measure Juilliard Manuscript

_1

_6

30

42

46

47

50-52 Augmentez peu à peu = 49-5 1 Peu à peu animé et en augmentant

54

60

65

66

77

89

Fig. 10.
Juilliar

Stichv
Juilliar
en anim
both m
Close s
parison
dynam
manus
fourth
which
tave le
found
occur i
25, the
half-n
Debuss
blurre
shorte
uscript
mm. 9
relativ
of the
render
closer
elsewh
this son

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 27

ã
'C
£
3
¡iļļ ! ļ G
cri

-
-
»
c
a> ■
»
S
i i e

bo
""
^ ,
"5-
? *
,"**
"..' -
& 0
S
=?£
- ü
- ^
-~s ■
■ /Î " I ■ |l| 'Sí
1Ü ^
S5

'3.
â> 2 . rn S rn
i
^ co
P;
N S ^~ N
I/
.
, r1- ,1
^
^ ^tT
tT
£ S UflU-
1
S = .; EE U fi! E - I
1 ::■" .; m. « U.» -Ï
¿o
<u

£
-
a. -
- >•%
- c _' >•%
i & ;_ ^ * „&3i il.
' >-**" - SII. ~~
_ -w *Š• 3 ^ * * $ „ g T3
C

S ^ $ ü -w S ^ ; : i Ht - : • ^ s
2 . o - ' - 1 -v '* ìi '§ « Il - ( « ■- ^ a

as
"5. k-
o *>a;- -rc
'c -- «,v*-
w I Hi lí( »ili
* > : .%_ ■- «-.SP
> %_ : crcr •c
u

E"5 í ••£ k-
> ta M£
■ *a;
* ^•
,£ita
^ "S M ■5 »fli
* -*_íT "S ^ <u
5/3

3
-» • tj u ¡*fcí C
So - tj - > ' .*&*■ Q_ 5/5 rsi Y lo "3 * : S **?] O «
S
I So - É ^ - S ' ^ .*&*■ ^ Q_ 1 5/5 rsi Ë I Y lo Ë Jí "3 * : '"^9 S **?] O Ë 5r ? I « 'S
.2
2¿ ^ CN T-" o 3
^2
nsoo
h lo
*-ro"^0
íN m °° £m
ro"^0 "? Orn
"> <? <D
£ h 'L s¿ rA eŇ CN CN ^ ^ ^ LO ^
^ rj- *- fN rn rt Tt II II II |' II
<4H
O
C/3
I nH i <U

.£
W
u
o
>
- C ^ <u
i Ł il . - ~ •S
5 M M i c "ni i : j§ c
D

í-
"5l _ . S / ï 'l
, ł»'M ' _ «/i "TS . i / "P
V

O i:; ~ v - „ ì Ž
cu
-
- ^_
- ' '¡.ili
-s ^ j -
! ;! <j ' ;. '
i «
. ,
i 71
, o
o -O
O _ m uj r' hs C/3

tfi i s j 1 1 I ; oo ļ > i n h <U

e ; !i iéi i E E '_i ļ J i' E -

U
C

fH
• :• ;
Hi0|
I íļ.**
I -
f«_ ~ i
v
u

«_
¿3
_ 'í"'R - í t'H ^ i ' » ~ <3 'i Tí
D. ' C 'u'
<U

i" 7§2»
§2»: :_: 4 i * _i*i I§^vá>
pi -25
>
«V

1 , £. 4 * * ( lí
u

<U
Q
3

^ 05
05 - .4- - -^1 o_
, o_ ( ^-J rs
^ i rsv1-iSi
^ v ^ "3«I
"3 «I ¿]!" " «Sr'
jT «Sr' i=-f"5s
r- *•- r- *•-"3
"3
^ ;îë!Ë o_ rs v ^ 1
C/3

.O
OJ

2f
Q

^ 00 io no CO ro £> ro rsT
aj^r-fŅ rņ ro un I' Oi
^ ^órń r'i rsT lo r^- rj- r^-Orj- bb
pt p-cN m ^ LO LO N K ai
S

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
 28 Notes, September 2014

This phrase ending also led to

most notably in the left hand
(see figs. 4a & 4b, above). At m.
third beats, and A# to C# o
Stichvorlage (m. 90) by a single
beats. Similarly, in m. 93, the cho
on each beat in the Juilliard m
(m. 92) by one chord (of the
two pianistic changes, which in
greater fluidity in playing the
hand part.
By contrast, we see the opposite phenomenon in mm. 94-95: in the
Stichvorlage , the left-hand chords are reinforced by the repetition of the
three middle notes G|, Bļ, and C# on the first two beats of the measures,
while in the Juilliard manuscript (mm. 93-94), the same notes are held
in a half-note chord over the first two beats of m. 93, and a dotted-half-
note chord over the three beats of m. 94. It should also be noticed that
Debussy changed the disposition of the chord in the Stichvorlage to F#,
G#, and BLļ on the third beat of mm. 94-95. In the Stichvorlage , the dy
namics are precisely noted, and extend from pp to pppp . These changes
it seems to me, may be seen as resulting from Debussy's pragmatic musi-
cianship: they urge greater precision, greater clarity, and greater appreci
ation of the poetic text.
As for the differences that appear in that text: most of them concern
matters of punctuation and capitalization. Nevertheless, the compose
does make a slight verbal change in mm. 14-15: in the Juilliard
manuscript, the line "A jamais navrée" (forever distraught) is twice
repeated, while at this place in the Stichvorlage , we read "Maintenant
navrée / A jamais navrée" (now distraught, forever distraught). Thi
small difference - the latter text being the one that was published in the
Entretiens politiques et littéraires - suggests that the Juilliard manuscrip
was completed and polished before December 1892. Also to be noted i
a modification of the piano accompaniment on the third beat of m. 14
and on the first beat of m. 15. Here Debussy opts in the end to double
the vocal line, as he does on the first two beats of m. 14.
There are also subtle changes in harmony that Debussy made in order
better to present the poetic text. Two passages in particular demonstrat
how Debussy tended to refine his writing as he revised the score for pub
lication. The first is at mm. 45-46, just before the return of the opening
theme, which we have already examined, with its chords notated in red
ink (see figs. 12a & 12b). Debussy reuses the same chordal harmony on
the first beat of mm. 45 and 46, except that in the Juilliard manuscript,

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 29

Fig. 12a. Debussy, "De rêve," Juilliard manuscript, mm. 43-46

Fig. 12b. Debussy, "De rêve," Stichvorlage (Morgan manuscript), mm. 43-46

what was a Bl?-major chord in m. 45 becomes a Bt-minor chord in m. 46,

with the shift from to Dl>, no doubt in order to underline the sense of
the word "rien" (nothing). Oddly enough, in the Stichvorlage , Debussy
sets down a D# in m. 46 in order, apparently, to cancel the effect of the
earlier Dk And indeed, in the published edition, the D in m. 46 carries
no accidental.

Another interesting coloristic difference here derives from the arpeg-

giated chords on the third beat of m. 45 (a dominant-ninth chord with-
out root), which casts light on the word "plus" (more). In the Juilliard
manuscript, which prescribes a slightly slower tempo than the Stichvor-
lage, the dominant-ninth occurs in thirty-second notes played on th
third beat of the measure; in the Stichvorlage , the dominant-ninth occur
in grace notes played before the third beat, which is marked "plus retenu.
Finally, on the third beat of m. 46, in the Stichvorlage, Debussy adds a low
G that leads very smoothly to the low Ff at the beginning of m. 47.
The second passage occurs in mm. 60-62 (see figs. 13a & 13b). The
first difference we observe here is at the fourth beat of m. 60, where, in

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
30 Notes, September 2014

Fig. 13a. Debussy, "De rêve," Juil

Fig. 13b. Debussy, "De rêve," Stichv

the Juilliard manuscript, Debu

bottom of the keyboard. In th
rather less expressive Al> that
tated in the Juilliard manuscr
beat of m. 61 in the Juilliard
chord that occurs on the sec
chord in the third inversion w
Stichvorlage (m. 60) he decided
Bt-major chord. Finally, in m.
chords in the right hand from
script, to the lower register, c
(m. 61). To prolong the reson
quarter-note on low G, in the
the Stichvorlage .
What are we to make of these
always relate them to the tex

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 31

morts" (the knights who are dead). The change of key signature at m. 60
in the Juilliard manuscript introduces a passage of an intentionally dif-
ferent sonic color at the heart of which are the chords in the right hand
of the piano, at mm. 60 and 62, which contrast with the four quarter-
note chords that set the words "chevaliers sont [morts]." Here we are at
the epicenter of what Debussy was attempting to express in the musical
and poetic language he called prose lyriquey that is, a conspicuously close
concordance of music and text.

I conclude this look at a little-known manuscript with a comment on

the first three measures of "De rêve." In the first two measures of the

Juilliard manuscript (fig. 14a), we see phrase groupings in the left han
by the beat ; in the Stichvorlage (fig. 14b), we see phrase groupings by t
measure. This adjustment, I believe, is related to the tempo Debussy ima
ined for the song, which, as we have seen, he increased slightly, from
the "Lent (Tempo Rubato)" of the Juilliard manuscript to the "Modéré"
of the Stichvorlage. We also find an adjustment of the vocal part. In th
Juilliard manuscript, the first six syllables - "la nuit a des douceurs de
femme" (the night is as sweet as a woman) - carry tenuto markings. B
these are removed in the Stichvorlage. (We observe the same phenome-
non in m. 9.) Debussy no doubt feared that the singer might misunder
stand the meaning of this notation, which he uses only when notes are
repeated. In his sonic imagination, the notation indicates not that the sy
lable should be accentuated, but rather that it should be spoken, o
rather sung in a parlando style.
Finally in m. 3, we observe a significant modification of the piano part
in which Debussy abandons the broken fifths of the Juilliard man
script's left hand (Gl> to Dt; Dtj to A^) and replaces them in the Stichvo
lage with octaves, which opens the space in the bass. We also observe a
enharmonic change in the notation of the vocal part, where flats b
come sharps, presumably to facilitate the singer's approach to the crucia
word "d'or" in m. 4 - by diminished fourth in the Juilliard manuscrip
by major third in the Stichvorlage.
There are many other differences of detail that I would have liked t
consider were there space to do so. But even from a relatively brief ex
amination I think it is possible to draw certain conclusions about
Debussy's methods of work. From a musicological point of view, th
changes Debussy operated on these manuscripts prove that he always r
visited and revised with a very fine chisel. They prove as well that befo
publishing his music he preferred to age it and refine it, sometim
doing so by singing and playing himself (which he enjoyed doing
whether the work in question was a melodie , Pelléas et Mélisande , or even an
opera by Wagner), and sometimes doing so with the assistance of singe

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
32 Notes, September 2014

Fig. 14a. Debussy, "De rêve," Juill

such as Marie Fontaine, whom I

was his momentary fiancee at t
lyňques. Finally, the changes be
compositional personality, for a
performers in mind: he wanted

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
From Debussy's Studio: The Little-Known Autograph of De rêve 33

Fig. 14b. Debussy, "De rêve," Stichvorlage (Morgan manuscript), mm. 1-4

tice to the smallest and slightest inflections of his musical and verbal
texts.

I should like to express the hope that the study of these two sourc
from Debussy's studio will both renew and enrich performers' readin
and interpretations of this song, and render them more alert to the
anced indications that are added and subtracted by the composer

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms
34 Notes, September 2014

other repertory as well. Detailed

the sensibilities and perceptions
- Translated by

ABSTRACT

In the early part of his career, one of the areas in which Debu
ried out extensive experimentation was the art song - the mélodi
at first set poetry by Verlaine, Baudelaire, and Mallarmé, De
cided in the eighteen-nineties to write his own poetry for musica
taking inspiration from the free verse ("vers libre") of the Sym
Completed in the years 1892-93, the group of four melodies that
bolically entitled Proses lyňques was conceived at the moment at w
composer was completing both the String Quartet and the P
l'après-midi d'un faune . And it is not by chance that he comple
mélodies just as he was beginning work on the opera Pelléas et Mél
For the Proses lyňques we possess several manuscript versions, i
the autograph manuscript that served the engraver (the Stichvo
today on deposit at the Pierpont Morgan Library in New York.
first of the group, "De rêve," another autograph is preserved in
lections of the Juilliard School. Comparison of the Morgan and
manuscripts reveals differences not in the structure of the mél
much as in the subtle refinements Debussy made in the writing
and piano. Indeed these refinements reveal the experimental and
matic predilections of a composer anxious to offer his interp
musical text accurately reflective of his sonic universe.

This content downloaded from

193.54.67.95 on Wed, 09 Jun 2021 18:59:52 UTC
All use subject to https://about.jstor.org/terms