Pub - Debussy and The Fragment Chiasma 18 PDF

Debussy and the Fragment
Chiasma 18
General Editor
Michael Bishop
Editorial Committee
Adelaide Russo, Michael Sheringham,
Steven Winspur, Sonya Stephens,
Michael Brophy, Anja Pearre
Amsterdam - New York, NY 2006

Debussy and the Fragment
Linda Cummins
Cover design: Pier Post
The paper on which this book is printed meets the requirements of

‘ISO 9706: 1994, Information and documentation - Paper for documents -
Requirements for permanence’.
Le papier sur lequel le présent ouvrage est imprimé remplit les prescriptions
de "ISO 9706:1994, Information et documentation - Papier pour documents -
Prescriptions pour la permanence".
ISBN-10: 90-420-2065-2
ISBN-13: 978-90-420-2065-8
©Editions Rodopi B.V., Amsterdam - New York, NY 2006
Printed in The Netherlands
CHIASMA
Chiasma seeks to foster urgent critical assessments focussing upon joinings

and criss-crossings, single, triangular, multiple, in the realm of modern French
literature. Studies may be of an interdisciplinary nature, developing connections
with art, philosophy, linguistics and beyond, or display intertextual or other
plurivocal concerns of varying order.
Michael Bishop
Halifax, Nova Scotia
Rather than solid frames, some less than perfect aesthetic objects have permeable
membranes which allow them to diffuse effortlessly into the everyday world.
In the parallel universes of music and literature, Linda Cummins extols the
poetry of such imperfection. She places Debussy’s work within a tradition
thriving on anti-Aristotelian principles: motley collections, crumbling ruins
real or fake, monstrous hybrids, patchwork and palimpsest, hasty sketches,
ellipses, truncated beginnings and endings, meandering arabesques, irrelevant
digressions, auto-quotations. Sensitive to the intermittences of memory and
experience and with a keen ear for ironic intrusion, Cummins draws the
reader into the Western cultural past in search of the surprisingly ubiquitous
aesthetic of the unfinished, negatively silhouetted against expectations of
rational coherence. Theories popularized by Schlegel and embraced by the
French Symbolists are only the first waypoint on an elaborately illustrated
tour reaching back to Petrarch. Cummins meticulously applies the derived
results to Debussy’s scores and finds convincing correlations in this chiasmatic
crossover.
Anja Pearre
Halifax, Nova Scotia 2006
This page intentionally left blank
ACKNOWLEDGEMENTS
I would like to thank Professors Matthew G. Brown, now of the

Eastman School of Music, University of Rochester, and Jan Herlinger
and Adelaide Russo of Louisiana State University, co-directors and
minor professor respectively, for invaluable guidance and generosity
with their time during the completion of the dissertation on which this
book is based; and since. The Graduate School of Louisiana State
University, the Bourse Chateaubriand, and the Camargo Foundation
provided indispensable financial support for research and writing, for
which I am most grateful. Dover Scores kindly provided permission to
use their publications as models for all musical examples. Thanks also
to Alaric Haag, who responded with characteristic élan to a request for
cover art; to Anja Pearre, whose conscientious and thorough editing
saved me from many infelicities of expression; and to Christa Stevens,
for seeing the manuscript expeditiously through the press. Finally, I
thank my family and friends for their unflagging support and
encouragement.
CONTENTS
Introduction 11
Chapter 1
Ruins of Convention; Conventions of Ruin 21
Chapter 2
Beginnings and Endings 63
Chapter 3
Arcadias and Arabesques 95
Chapter 4
The Sketch 117
Chapter 5
Auto-Quotation 135
Chapter 6
Preludes: A Postlude 151
Bibliography 171
INTRODUCTION
“Gifted as he was, there was nothing cosmic or inspired about

Debussy’s work, as is proved by the fact that he never produced
anything incomplete.”1 These words of the Berlin critic Adolf
Weissmann, written in the early 1920s—dated, opinionated, no doubt
nationalistically prejudiced—echo Friedrich Schlegel’s description of
the Romantic literary work from his often-quoted Athenaeum
Fragment 116: “the romantic kind of poetry is still in the state of
becoming; that, in fact, is its real essence: that it should forever be
becoming and never be perfected.”2 Weissmann’s equation of the
inspired with the incomplete identifies him as an heir of German
Romanticism and indicates the extent of the influence held by certain
aspects of that movement more than a hundred years after Schlegel’s
writings gave it direction.
Stefan Jarocinski, writing in the 1960s with a different
agenda—the promotion of Debussy as a creator of the modern in
music—continued to rely on the authority of the incomplete.
Justifying his praise of Debussy’s music by the presence of that very
element Weissmann found missing, he reiterated the Romantic ideal
of the forever becoming: “Sa musique ne commence ni ne finit. … Sa
forme n’est pas close. … [Elle] se forme, se renouvelle sans cesse …”
[His music neither begins nor ends. Its form is not closed. It forms
itself, it renews itself without ceasing.]3
Other critics of Debussy’s time agreed with Weissmann’s
negative assessment of the composer, but for the opposite reason,
decrying Debussy’s music because of its fragmentation and referring,
for instance, to “fragments of the tonal wreck,”4 “la décomposition de
1
Adolf Weissmann, The Problems of Modern Music, trans. M. M. Bozman (London:
J. M. Dent and Sons Ltd., 1925), 171. Originally published as Die Musik in der
Weltkrise, 1922.
2
Friedrich Schlegel’s Lucinde and the Fragments, ed. Peter Firchow (Minneapolis:
University of Minnesota Press, 1971), 175.
3
Stefan Jarocinski, Debussy: impressionnisme et symbolisme (Paris: Éditions du
Seuil, 1970), 74-75; my translation.
4
Louis Elson, Boston Daily Advertiser, March 4, 1907, quoted in Nicholas
Slonimsky, Lexicon of Musical Invective: Critical Assaults on Composers since
Beethoven’s Time, 2nd ed. (Seattle: University of Washington Press, 1969), 93 (of La
Mer, “We clung like a drowning man to a few fragments of the tonal wreck …”).
12 Debussy and the Fragment
notre art … et la ruine de notre être” [the decomposition of our art …

and ruin of our essence].5 Critics closer to our own time also find
fragmentation in Debussy’s compositions, calling attention to
“fragments of melody,”6 “fragments of counterpoint,”7 etc. Unlike
those critics of Debussy’s time, they side with Jarocinski and single
out these features for praise as harbingers of the future, comparing
them to developments in modern art through their choice of terms like
“collage citations”8 and “mosaic technique.”9 Yet they ignore or
marginalize Debussy’s ties to the Romantic tradition he inherited and
confronted. We have reached an imbalance: recent scholarship has
focused predominantly on Debussy’s ties to modernism,10 while
virtually ignoring his ties to fragmentation in music of the Romantic
period.
Although frequently considered the property of early German
Romanticism on the one hand or of the twentieth century on the other,
the fragment, the incomplete entity, has been recognized and
consciously employed in Western art and literature at least since the
fourteenth century. Scholars are now devoting more serious study to
the role of the fragment in other chronological periods; as a result, a
history of the fragment is currently being constructed. This history
offers an alternative and balance to the long-held Aristotelian
5
Camille Bellaigue, Revue des Deux Mondes, Paris, 15 May 1902, quoted in
Slonimsky, Lexicon, 90-91 (“Aucun n’est mieux qualifié que l’auteur de Pelléas et
Mélisande pour présider à la décomposition de notre art. La musique de M. Debussy
tend à la diminution et à la ruine de notre être.” [No one is better qualified than the
composer of Pelléas et Mélisande to preside over the decomposition of our art. The
music of M. Debussy leads to the emaciation and ruin of our essence.]).
6
Paul Roberts, Images: The Piano Music of Claude Debussy (Portland, Oregon:
Amadeus Press), 172 (of “Et la lune descend sur le temple qui fut”).
7
William Austin, Music in the Twentieth Century from Debussy through Stravinsky
(New York: W. W. Norton and Sons, 1966), 45 (of the “Ballade de Villon à s’amye”).
8
Glenn Watkins, Pyramids at the Louvre: Music, Culture, and Collage from
Stravinsky to the Postmodernists (Cambridge: The Belknap Press of Harvard
University Press, 1994), 298.
9
Michael L. Friedmann, “Approaching Debussy’s ‘Ondine’,” Cahiers Debussy, 6
(1982): 22.
10
Most characteristically, perhaps, in the often-quoted remark of Pierre Boulez: “ …
one can justifiably claim that modern music began with L’Après-midi d’un faune.”
Stocktakings of an Apprenticeship, trans. Stephen Walsh, collected and presented by
Paule Thévenin (Oxford: Clarendon Press, 1991), 267.
Introduction 13
perception of unity as the ultimate goal of art—a unity analysts and

critics construct by marginalizing disorder and fragmentation, by
concentrating on and valuing what binds together rather than what
pulls apart, and consequently by weakening and devaluing the very
powerful statements made by fragments prior to the Romantic era.11
In music criticism, scant attention has been paid to fragments
except in discussions of nontonal music of the twentieth century and
studies of sketches and works left incomplete at a composer’s death.
This omission is unfortunate in that it, too, distorts our view of the
modern fragment by placing it in a vacuum, without prior history.
More recently the importance of the fragment in the aesthetics of
Romantic music has been traced by a number of specialists, most
significantly Charles Rosen and John Daverio.12 They have
demonstrated how Romantic ideas of the fragment were developed in
the music of Chopin, Schumann, Liszt, Brahms, Wagner, Strauss, and
others. These studies have made the phenomenon of fragmentation in
music more visible and have provided a more nuanced view of the
function of disorder, disruption, and incompletion in its history. This
background can allow us to begin to see Debussy’s music from a more
balanced perspective, to understand how he exploits the tension
between whole and fragment.
The composers mentioned in the preceding paragraph were a
part of Debussy’s musical heritage: he knew their music, studied it,
performed it.13 He placed himself in the company of Chopin and
Schumann, writing in a letter to his publisher Jacques Durand,
11
On these points, see Lawrence D. Kritzman, preface to Fragments: Incompletion
and Discontinuity (New York: New York Literary Forum, 1981), vii.
12
Charles Rosen, The Romantic Generation (Cambridge: Harvard University Press,
1995); John Daverio, Nineteenth-Century Music and the German Romantic Ideology
(New York: Schirmer Books, 1993); also Ramon Satyendra, “Liszt’s Open Structures
and the Romantic Fragment,” Music Theory Spectrum 19 (Fall 1997): 184-205.
13
For works Debussy played in competitions as a student, John R. Clevenger,
“Achille at the Conservatoire,” Cahiers Debussy 19 (1995): 3-35; “Debussy’s Paris
Conservatoire Training” in Debussy and his World, ed. Jane Fulcher (Princeton:
Princeton University Press, 2001), 299-361. Debussy edited Chopin’s works for
Durand (1915-1917); he transcribed Schumann’s Six études en forme de canon, op. 56
for two pianos (Durand, 1891).
Avez-vous joué les Images...? Sans fausse vanité, je crois que ces
trois morceaux se tiennent bien et qu’ils prendront leur place dans
la littérature du piano... (comme dirait Chevillard), à gauche de
Schumann, ou à droite de Chopin... as you like it.
[Have you played the Images … ? Without false vanity, I think

these three pieces work well and will take their place in piano
literature … (as Chevillard would say), to the left of Schumann or
to the right of Chopin … as you like it.]14
In 1904 Laloy introduced a concert with a talk intended to show

connections between Debussy and Schumann, both of whom he called
musical intuitives.15 Recent scholarship has illuminated the
relationship of Debussy to these musical forebears as well. Roy Howat
catalogs many instances of Debussy’s conscious or unconscious
modeling on the works of Chopin in the Preludes, the Etudes, L’Isle
joyeuse, and others.16 Rosen writes that Debussy
loved the music of Schumann and understood its capricious

imagination: he was able to adapt some of the unresolved
sonorities and the fleeting phrases that Schumann himself
distrusted into a new musical style, one almost totally liberated
from the academic requirements of the past—to realize, in short,
both the destructive and creative elements of Schumann’s art in
his own way.17
Ultimately, Rosen observes that the classical hierarchy of

genres
had been shaken to the ground by Schumann and Chopin, shored

up by Brahms and Wagner for a time, until the early years of the
twentieth century. This is certainly one of the reasons why the
14
Claude Debussy, Correspondance 1884-1918, ed. François Lesure (Paris: Hermann
éditeurs des sciences et des arts, 1993), 204; trans. Roger Nichols, Claude Debussy,
Debussy Letters, ed. François Lesure and Roger Nichols (London: Faber and Faber,
1987), 158. The phrase “as you like it” is written in English in the original letter,
dated 11 September 1905.
15
François Lesure, Claude Debussy (Paris: Klincksieck, 1994), 254.
16
Roy Howat, “Chopin’s Influence on the Fin de Siècle and Beyond,” in The
Cambridge Companion to Chopin, ed. Jim Samson (Cambridge: Cambridge
University Press, 1992), 246-283.
17
Rosen, 706.
Introduction 15
work of Schumann and Chopin has more affinity with the style of
Debussy and the early twentieth century than it does with the
music of the last half of the nineteenth.18
Rosen’s statements parallel literary comparisons between

French Symbolism and early German Romanticism that began at least
as early as 1891.19 The specific connections are often vague, and are
argued by scholars with various perspectives; nevertheless, Henri
Peyre claims: “It happens that a family of like minds finds itself thus,
a hundred years and a thousand miles removed from one another.”20
Ernst Behler and Roman Struc write of Friedrich Schlegel:
With this conversion to mythology and symbolism, Schlegel

[Friedrich] opened a path which consequently led to Nietzsche’s
early aesthetics and to French Symbolism of the 19th century.21
Lilian Furst expands this comparison in an extended work, noting

“parallel tendencies [that] develop within the framework of diverse
national traditions,” singling out as features common to Symbolism
and the Frühromantik “idealism, freedom of the imagination, flight
from formal plasticity, obscurity, emphasis on the musical qualities of
18
Rosen, 699-700.
19
Jean Thorel, “Les Romantiques allemands et les symbolistes français,” Entretiens
politiques et littéraires 3 (1891): 85-109. Peyre and Furst report this early comparison.
20
Henri Peyre, What is Symbolism? trans. Emmett Parker (Tuscaloosa, AL:
University of Alabama Press, 1980), 86-88; originally published as Qu’est-ce le
symbolisme? (Paris: Presses universitaires de France, 1974).
21
Ernst Behler and Roman Struc, introduction to Dialogue on Poetry and Literary
Aphorisms by Friedrich Schlegel (University Park, PA: The Pennsylvania State
language, religious feeling.”22 A comparison of texts by German

Romantics and by writers often labeled Symbolist reveals striking
similarities. Verlaine, calling on the infinite possibilities of the
incomplete, wrote of Villiers de l’Isle-Adam in Les Poètes maudits
that his poetic dramas “were ‘like cathedrals and revolutions,’ always
unfinished and always begun again.”23 A. W. Schlegel’s statement
The arts should be brought together again, and bridges sought

from one to another. Perhaps columns shall come to life as
paintings, paintings become poems, poems become music.24
foreshadows Baudelaire’s notion of “Correspondances” in the first

two stanzas of his sonnet, labeled by Henri Dorra “a preliminary
manifesto” of French Symbolism:25
La Nature est un temple où de vivants piliers
Laissent parfois sortir de confuses paroles;
L’homme y passe à travers des forêts de symboles
Qui l’observent avec des regards familiers.
Comme de longs échos qui de loin se confondent

Dans une ténébreuse et profonde unité,
22
Lilian Furst, Counterparts: The Dynamics of Franco-German Literary
Relationships 1770-1895 (Detroit: Wayne State University Press: 1977), 8, 106-7. In
her study, Furst notes other writers who compare early Romanticism and Symbolism,
including Albert Béguin’s influential L’Âme romantique et le rêve (Paris: Corti,
1937), which focuses on a “lineage of interior romanticism.” (The Béguin quote is
from page 328.) There is no space here to summarize Furst’s argument adequately,
nor to take into consideration the numerous questions it raises: the timing and spread
of the Frühromantik’s texts and theories in France, the effect of those intervening
years of French Romanticism in the formation of Symbolism proper, Wagner’s role in
France and Germany, and the unresolved problem of just what poets and writers
constitute the movement labeled Symbolism. [On this mare’s nest, see Laurence M.
Porter, The Crisis of French Symbolism (Ithaca: Cornell University Press, 1990),
especially Chapter I.
23
Edgar Wind, Art and Anarchy, 3rd ed. (Chicago: Northwestern University Press,
1985), 119, n. 70.
24
A.W. Schlegel quoted by Guido Adler, Handbuch der Musikgeschichte, vol. 2
(Tutzing: Hans Schneider, 1961), 865. Translated in Donald J. Grout and Hermine
Weigel Williams, A Short History of Opera, 4th ed., (New York: Columbia University
Press, 2003), 418.
25
Henri Dorra, ed. Symbolist Art Theories: A Critical Anthology (Berkeley:
University of California Press, 1994), 10.
Introduction 17
Vaste comme la nuit et comme la clarté,

Les parfums, les couleurs et les sons se répondent.
[Nature is a temple where living pillars

Sometimes unleash a confusion of words;
Man traverses it through forests of symbols
Which observe him with knowing glances.
Like extended echoes which mingle far away

In a mysterious and profound unity,
Vast as the night and as light,
Perfumes, colors, and sounds answer each other.]26
Henri Dorra speculates that the source for Baudelaire’s image of

“symbol-covered pillars” may have been a passage by Joseph Görres,
“Priests based the great principles of all cosmogony on [the holy
books]—a great, powerful, and noble row of pillars, which keep on
appearing in all myths, unchanged.” Görres was a colleague of
Friedrich Creuzer, who had been Friedrich Schlegel’s student. Dorra
names Creuzer, Joseph Guigniaut, and Pierre Leroux as particularly
important to the transmission of German Romanticism (specifically
theories of the symbol presented in the writings of Herder, Friedrich
Schlegel, Kant, Schiller, Goethe, and Moritz) to nineteenth-century
France.27
Certain aspects of Baudelaire’s correspondences find concrete
expression in his writings on the visual arts, inspired by his
association with Delacroix, and present interesting connections to
Walter Pater’s notion that all art aspires to the condition of music, a
sentiment shared by both Romanticism and Symbolism, and to ideals
of Romantic distance illuminated by Berthold Hoeckner. Hoeckner
begins his essay “Schumann and Romantic Discourse” with a quote
from Novalis: “Philosophy is prose. Its consonants. Distant
philosophy sounds like poetry—because every call into the distance
becomes a vowel …” Hoeckner then sums up: “Novalis’s use of
26
Charles Baudelaire, “Correspondances,” Les Fleurs du Mal, Oeuvres complètes, ed.
Claude Pichois (Paris: Éditions Gallimard, 1975), 1:17; trans. Glenn Watkins,
Soundings: Music in the Twentieth Century (New York: Schirmer Books, A Division
of Macmillan, Inc., 1988), 66.
27
Dorra, 9-10. Görres quotation: Joseph Görres, Mythengeschichte der asiatischen
Welt (Heidelberg: Mohr and Zimmer, 1810), 2:644.
distance captures the imagination of the musician: distant philosophy

sounds like poetry. Dying away into the distance, prose turns into
poetry, speech into vocalise, language into music.”28 What Novalis has
seen distance do to language, Baudelaire has seen it do to painting:
The appropriate way to determine whether a painting is melodious

is to look at it from a distance so as to be unable to comprehend
its subject or its lines. If it is melodious, it already has a meaning
and has taken its place in the repertory of memories.29
Dorra points out similarities to Delacroix’s aesthetic theories:
There is an impression that results from a certain arrangement of

colors, lights, shadows, and so forth. It is what one might call the
music of painting. Before you even know what the painting
represents, … when you are too far away from it, … you are
conquered by this magical accord.30
In the distance, everything becomes music.

“In Frankreich geschah etwas, was in Deutschland ausblieb.
Novalis’ Nachfolge ist in Frankreich viel echter und legitimer als in
Deutschland selbst.” [In France there took place what failed to happen
in Germany. Novalis’s succession is much more genuine and
legitimate in France than in Germany itself.]31 Werner Vordtriede’s
statement, quoted by Furst, mirrors Rosen’s notions of Debussy as
inheritor and continuer of Schumann. If the bridge from Novalis’s
blue flower to the Mallarméan azur has a parallel in music, if Peyre’s
“like minds” and Furst’s “parallel tendencies” can extend to music,
then we should look at Debussy’s work for the remnants of the
Romantic fragment, for the fragment that, as Harries notes, “suggests
28
Berthold Hoeckner, “Schumann and Romantic Distance,” Journal of the American
Musicological Society 50 (1997): 55-56.
29
Baudelaire, “Salon de 1846,” Oeuvres complètes, 2:425, quoted in Dorra, 3.
30
Delacroix, “Réalisme et idéalisme,” Oeuvres littéraires (Paris: Crès, 1923), 1:23-
24, quoted in Dorra, 3.
31
Furst, 108, quoting Werner Vordtriede, Novalis und die französischen Symbolisten
(Stuttgart: Kohlhammer, 1963), 182.
Introduction 19
our position in a historical continuum.”32 The purpose of this book is

to begin to examine Debussy’s music in such a historical continuum.
This examination involves the observation of techniques of
fragmentation in literature—techniques that writers have reinterpreted
and recast to their own use at least since the fourteenth century—and
the identification of parallels in musical structures. Drawing on
several studies of literary fragmentation, Chapter 1 provides an
overview of the literary fragment through selected examples of some
of the most famous and most influential fragment works, from
Petrarch’s Canzoniere through Sterne’s Tristram Shandy to Nerval’s
Aurelia. Writers create their fragments by breaking the particular
norms or conventions of their own time and place, but many of the
techniques and strategies of fragmentation remain the same through
the centuries: to omit the necessary, to add the superfluous, to
combine the incongruous, to exaggerate out of proportion, to put
things in the wrong order.
Fragment writers build on the work of previous creators of
fragments, as Harries points out;33 thus from the breaking of
conventions arise not only fragments but also conventions of
fragmentation: the essay collection whose founder was Montaigne; the
lyric cycle whose creator was Petrarch; the digressive, nonlinear
narratives of Rabelais and Cervantes that so influenced Sterne,
Diderot, and Schlegel; in whatever guise, in whatever period, these
genres are built on the inadequacy of perfection. Writers continue a
tradition of fragmentation because fragments make powerful
statements—statements of denial, of loss, of failure, or simply of the
ridiculousness of life. Thus Petrarch turns to the fragment to
acknowledge the inadequacy of words to capture his love for Laura,
and Schlegel the inadequacy of literary forms to confront his
Absolute. These and others became models, inspirations, for writers of
Debussy’s time as they adapted strategies and techniques of
fragmentation to new concerns, or different views of old ones.
Writers imitated past fragments, even as they responded to
their own artistic climate; Debussy did the same. Chapters 2-5
32
Elizabeth Wanning Harries, The Unfinished Manner: Essays on the Fragment in the
Later Eighteenth Century (Charlottesville: University of Virginia Press, 1994), 13.
33
Harries, 91.
examine strategies of fragmentation in individual works written either

in direct response to literary fragments or to the images and aesthetics
of fragmentation that remained powerful symbols in his society. Of
particular importance are parallels in Debussy’s music to musical
fragmentation in the works of Schumann and Chopin. Following the
model of Rosen and Daverio, comparisons with literary fragmentation
are based in part on the substitution of the tonal system for a linear or
serial concept of time. Topics covered include the mutilation of
beginnings and endings and the analytical uncertainty they can
present; the translation of the arabesque as digressive narrative
structure to tonal harmony and the accompanying subversion of
notions of Aristotelian beginning, middle, end; the continuing
fascination with the sketch perceived as the result of immediate and
unmediated inspiration; the use of quotation to pull the listener/reader
out of the work itself, interrupting progression even as layers of
meaning expand. In conclusion, the final chapter investigates open and
closed collections, taking Debussy’s Preludes for piano as a case study
and comparing them to a cabinet of curiosities, a holder of souvenirs
and memories.
CHAPTER 1
RUINS OF CONVENTION; CONVENTIONS OF RUIN
A portent therefore, does not arise contrary to

nature, but contrary to what nature is
understood to be. Portents are also called
“signs,” “monstrosities,” and “prodigies”
because they seem to portend and to point out,
to demonstrate, and predict future happenings.
—————Isidore of Seville1
Man’s fascination with the fragment is ancient, his uses of it

as varied and rich as imagination. From broken bits of reality—from
disassociated heads, horns, wings, and feet, stitched and glued
improbably together—he has fashioned both his gods and his
monsters; from potsherds and ruins he has reconstructed his history
and previewed his fate; and from shreds of artistic convention he has
confronted Aristotelian ideals of perfection with the power of the
incomplete, countering harmony, order, and unity with discord,
disjunction, and deformity.
The fragment in Western European art and literature never
strays far from its kinship with ruins and monsters, with the broken
and the malformed. Mimicking decay and deformity, artists through
the centuries have turned the familiar into the unfamiliar, creating not
only the monstrous creatures cavorting on the margins of medieval
manuscripts and carousing in the intergalactic Star Wars bars at the
boundaries of the science fiction universe, but also literary fragments,
many now canonical, whose deformity has often been made perfectly
clear by their authors: Petrarch titled the rime sparse (scattered
rhymes) of his Canzoniere “fragments” (Rerum vulgarium fragmenta),
Montaigne described his Essais as “monstrous bodies,”2 Rabelais
admitted that his reader would find little perfection in Gargantua’s
1
Isidore of Seville, Etymologiarum sive originum, Libri XX, XI.iii/2-3, “De
Portentis,” trans. William D. Sharpe in “Isidore of Seville: The Medical Writings,”
Transactions of the American Philosophical Society, new series, 54, pt. 2 (1964):
51-54. All subsequent quotations from Isidore are from this source.
2
Oscar S. Kenshur, Open Form and the Shape of Ideas: Literary Structures as
Representations of Philosophical Concepts in the Seventeenth and Eighteenth
Centuries (Lewisburg: Bucknell University Press; London and Toronto: Associated
University Presses, 1986), 41, citing Michel de Montaigne, Essais (Paris: Garnier,
1962), 1:198.
story, Cervantes opened Don Quixote with endorsements as fake and

as incomplete as the mutilated documents on which he claimed to base
his tale, Diderot equated the unfinished statue with the ruin3 and chose
the fragment to illustrate the failure of system in his De
l’Interprétation de la Nature, Schlegel purposefully titled his works
fragments, Victor Hugo compared his ideal theater to the composite
creatures of mythology,4 while Eliot declared, “These fragments I
have shored against my ruins.”5 Such fragments may cause
interpretive and analytical problems for the critic, the analyst, and the
audience. They foil expectations; they defy clear analysis; they do not
fit neatly into prescribed buckets, not quite filling the measure or
spilling over into other categories.
The desire to complete the fragment, to identify the whole that
has been somehow deformed, is compelling. When writers, artists, and
composers intentionally make fragments, identifying the model they
fragment is a necessary step for the audience toward classifying such
works, locating them in a tradition, and understanding their meaning
and purpose. These conventions do not function as an underlying
structure in individual works, merely hidden by surface fragmentation
and waiting to be uncovered; they are simply the measures against
which the fragment can be recognized. In the arts, convention
ultimately defines the whole, the model, and thus the perception of the
fragment. Though the “normal” or familiar forms of the natural world
and the concept of time as linear progression may serve as the most
deeply rooted models of perfection, many broken versions of these
ideals become conventions in their own right. A portrait painted from
the waist up represents a fragment of the human subject, but does not
strike the viewer as incomplete because it does not violate artistic
convention; let the artist extend the subject’s hand outside a painted
frame, as Bartolomé Esteban Murillo has done in a self-portrait c.
1670 (London, National Gallery), and he calls attention to the hidden
3
Anne Betty Weinshenker, “Diderot’s Use of the Ruin-Image,” in Diderot Studies 16
(1973): 328 n. 55, citing Diderot, Observations sur la Sculpture et sur Bouchardon in
Oeuvres complètes, vol. 8, ed. J. Assézat and Maurice Tourneux (Paris: Garnier,
1875-77), 43.
4
Victor Hugo, Cromwell (Paris: Garnier-Flammarion, 1968), Préface, 69-71.
5
T. S. Eliot, The Wasteland V/431 in T. S. Eliot, The Complete Poems and Plays,
1909-1950 (New York: Harcourt, Brace and Company, 1958).
Chapter 1: Ruins of Convention; Conventions of Ruin 23
part of the image and to the artificiality of the convention of portrait

painting itself.6 The bust, rounded edges neatly displayed on its
pedestal, represents a fragment of the human subject, but is
unsurprising because its presentation has become a convention. Igor
Mitoraj’s Tindaro Screpolato, no bust but a gigantic partial head—a
fragment of a fragment—with purposefully mutilated surface, rests
alongside a path in the Boboli Gardens; a modern remnant of some
imagined antique colossus, it defies convention and calls attention
both to its deliberately incomplete state and to its kinship with the
ruin. Heads remain more appropriate than other body parts.
Grandville’s sketch of a robot artist chiseling a gigantic thumb titled
“Le Doigt de Dieu” caricatured the rage for fragments in the
nineteenth century; more than a hundred years later, César’s gigantic
Thumb now rising from the pavement of La Défense is still
unconventional, still makes a powerful statement.7
Conventions define, but conventions can change over time,
and the perception of fragmentation in a work depends, in part, on a
familiarity with the convention on which it was built, or rather, from
which it was broken. Western European literature has a history of
conventions based on unity and completion; it also has a history of
conventions that have grown from attacks on those concepts of
perfection, conventions based on fragmentation. The following brief
survey of literary fragments, taken from studies of this subject, will
begin to show how, and with what intent, each writer manipulated the
conventions he inherited. It will reiterate the thought of current
scholarship stressing that the literary fragment was not a Romantic
invention or discovery, and will document Elizabeth Harries’ assertion
6
At least three reproductions of this image available online show only Murillo’s head
and chest—typical of portraits—cutting off the painted frame and the extended hand;
perhaps this is due only to matters of space.
7
For more on this subject, see Charles Rosen and Henri Zerner, Romanticism and
Realism: The Mythology of Nineteenth-Century Art (New York: The Viking Press,
1984), 82.
that fragment writers build on the work of previous creators of

fragments.8 It will also begin to show the extent to which writers have
used the fragment, the variety of its forms and settings, and the issues
and concerns that revolve consistently around it.
Petrarch
“Voi ch’ ascoltate in rime sparse il suono / di quei sospiri …”

[You who hear in scattered rhymes the sound of those sighs …].9 So
begins Petrarch’s Canzoniere (final version 1374), called Rime sparse
(Scattered Rhymes)—sometimes considered one of the first works of
modern literature.10 Referring to Petrarch’s Latin title for the work,
Rerum vulgarium fragmenta, Robert Durling believes this may have
been the earliest documented use of the term “fragment” to describe a
work of (presumably Western European) art.11 When Petrarch gave
this title, he was not only commenting on the impossibility of
confining love—and, of course, his response to love—to logical
narrative; he was also describing the form of his work. In his time,
there was no tradition for presenting a chronological narrative through
short, lyric forms. Lyric collections generally grouped poems by
type—sonnet, canzona, ballata—or by poet.12 There was, however, the
tradition of the prosimetrum, a genre that alternates lyric poetry with
prose commentary.13 Dante’s Vita nuova, a principal model for the
Canzoniere, belongs to this genre.14 Dante documented his love for
Beatrice through a series of lyric poems, each preceded and followed
8
Harries, 13.
9
Petrarch’s Lyric Poems: The “Rime Sparse” and Other Lyrics, trans. and ed. Robert
M. Durling (Cambridge: Harvard University Press, 1976), Poem #1, stanza 1, and
translation, 36-37.
10
Harries, 14.
11
Durling, introduction to Petrarch’s Lyric Poems, 26.
12
Poems were grouped by genre; Italian manuscripts separated sonnets, ballate,
canzone. Durling, introduction to Petrarch’s Lyric Poems, 13-18, citing E. H.
Wilkins, Life of Petrarch (Chicago: University of Chicago Press, 1961), no page
number given.
13
Peter Dronke, Verse with Prose from Petronius to Dante: The Art and Scope of the
Mixed Form (Cambridge: Harvard University Press, 1994).
14
Durling, introduction to Petrarch’s Lyric Poems, 9-10.
by a prose narrative explaining the poem’s structure and the

circumstances of its creation. The work as a whole presents a
continuous, if, as Dronke notes, sometimes tenuous, narrative.15
Petrarch’s Canzoniere also documents a love affair, his love for Laura,
but Petrarch, rather than containing his poems within prose frames,
“scattered” his narrative across the pages, telling his story only
through isolated, lyric moments. Each perfect poem is a fragment of
an imperfect work, a work that is itself only a fragment of the
prosimetrum. Petrarch’s collection of fragments became a convention,
a genre with many progeny, the lyric (or sonnet) sequence.16
The fragment thrives on the tension between its imperfect,
fragmented state and the perfect whole from which it was broken, the
source of its fragment identity. As Harries stresses, from form to
subject to surface detail, fragmentation permeates this work at every
level,17 but always as the counterpart to elements that create an
expectation of unity. Laura unifies, pervades the collection as subject
and inspiration, but that unity is thwarted by the poetic images that
consistently echo fragmentation: copious references to individual parts
of Laura’s body 18 and to parts of the poet’s own body; references to
Acteon, Orpheus, and other well-known myths of dismemberment and
dispersal; frequent use of words and images of fragmentation, such as
scattering;19 and many quotations and auto-quotations that turn the
reader’s attention to other works, as the Canzoniere reaches out
beyond itself, fragmenting itself.
15
Dronke, 97. Dronke notes the kinship of Dante’s prose sections to the troubadour
razos, short prose notes in some troubadour manuscripts that recount—with varying
degrees of accuracy—the details of the poet’s life and the circumstances of the
composition of his poetry.
16
As Durling reports, C. S. Lewis credited Petrarch with the invention of the sonnet
sequence (more accurately the lyric sequence since Petrarch included not only sonnets
but also madrigali, ballate, sestine, canzone) by omitting the prose narrative found in
the Vita nuova. Durling, introduction to Petrarch’s Lyric Poems, 9-10.
17
Harries, 14-17.
18
This listing of body parts in praise of a woman is called blason. Petrarch did not
invent the blason, but his use of it was so often copied that it became a common
device associated with his poetry.
19
Harries, 177 n. 4, quoting Nancy J. Vickers, “Diana Described: Scattered Woman
and Scattered Rhyme,” in Critical Inquiry 8 (1981): 85-97. Vickers gives a tally of 43
occurrences of forms of the Italian verb spargere, to scatter.
Petrarch presents a chronology, writing 365 poems plus a final

hymn to the Virgin, but it is a framework that is fictive and echos the
tension between lyric collection and chronological narrative.20 The
work has no climax or turning point, as Harries observes: Laura’s
death provides neither the closing nor a climax, but triggers simply
repetitive situations and emotions. The number 365 (plus the closing)
does not indicate a progression through the year; it serves simply to
limit the collection, to close it and frame it in a symbolic reference.
Harries believes we may account for this lack of progression by
accepting the notion that Laura was a fictional character: if Laura and
the poet’s relationship with her are a fabrication of the poet’s mind,
his attitudes toward her and his feelings for her need not change so
drastically after her death.21
Why did Petrarch “scatter” his rhymes? Perhaps to imply the
impossibility of grasping love in its totality, the necessity of breaking
that experience into manageable pieces. Mazzotta writes, “desire
knows only shreds and fragments, even if plenitude is its ever elusive
mirage.”22 Why then did Petrarch bind his poems as he did, creating
more than a simple collection? Perhaps to restore Laura and to restore
himself in the only way he could. If the poems correspond to the body
pieces of the hero in certain myths of dismemberment, then only by
being gathered can they either be laid to rest (Orpheus) or reborn
(Osiris).23 Yet as Petrarch must have known, collection and
preservation are ultimately impossible:
For Petrarch the term [fragment] expresses the intensely

self-critical awareness that all integration of selves and texts is
relative, temporary, threatened. They flow into multiplicity at the
touch of time, their inconsistencies juxtaposed as the successive
traces of a subject who dissolves and leaves only words behind.24
20
For more on this topic, see Mazzotta, The Worlds of Petrarch (Durham: Duke
University Press, 1993); Durling, introduction to Petrarch’s Lyric Poems. Petrarch
copied Laura’s obituary on the flyleaf of his copy of Virgil.
21
Harries, 18, and also on questions of Laura’s existence, Durling, introduction to
Petrarch’s Lyric Poems, 4.
22
Mazzotta, 78.
23
Harries, 15-17.
24
Durling, introduction to Petrarch’s Lyric Poems, 26.
As a reader of classical manuscripts, Petrarch knew both the

widening holes in parchment and the widening gap between the
author’s intent and the reader’s interpretation. Had he mutilated his
own manuscript, not only to reflect the inadequacy of telling but also
the inadequacy of reading? Perhaps as he assigned to the reader the
task of creating a narrative from his fragments, he acknowledged the
inability of any reader to recreate accurately, regardless of information
given or withheld; at the same time, he paralleled his own inability to
recreate accurately from memory. Petrarch depicted memories, stored
and retrieved like snapshots, but memories are far from perfect.
Petrarch worked over the poems of the Canzoniere for more
than forty years, collecting, reviewing, reorganizing, and refining his
fragments and his memories. Mazzotta speaks of the palimpsests of
memory,25 memory writing over reality, memory creating reality,
memory creating memory. Perhaps Petrarch wrote and rewrote as
much for himself as for posterity or audience, for forgetting is also a
form of ruin.
Montaigne
Like Petrarch’s Canzoniere, Montaigne’s Essais also establish

a new genre: predecessor of the Essays of Bacon, the Pensées of
Pascal, the Pensées, Maximes et Anecdotes of Chamfort, they lead to
the aphoristic writings of Friedrich Schlegel and the Romantics.26 Like
the sonnet sequence, the Essais create a genre whose progeny may be
more easily recognized today than many of the genres Montaigne
incorporated in its creation. He presents a collection of separate units
of prose bound and published together, a unit defying classification in
any contemporary genre, lacking any unifying narrative.27 He used as
his models the classical maxim and epigram collections, and
25
Mazzotta, 4.
26
For a discussion, Philippe Lacoue-Labarthe and Jean-Luc Nancy, The Literary
Absolute: The Theory of Literature in German Romanticism, trans. Philip Barnard and
Cheryl Lester (New York: State University of New York Press, 1988), 40.
27
Donald M. Frame, preface to Michel de Montaigne, The Complete Essays of
Montaigne, trans. Donald M. Frame (Stanford: Stanford University Press, 1965), 2.
particularly the annotated Adages of Erasmus.28 These collections do

not normally imply any narrative and are not associated with
autobiography, but that is precisely what Montaigne’s preface tells the
reader that his work is intended to be: a self-portrait, complete with all
his defects, for the use of his friends and relatives to remember him
when he is dead. Stressing their autobiographical nature, Northrop
Frye calls the Essais a short form of the genre of confession, missing
only the continuous narrative of those longer works.29
Within this fragment form, as Gilbert Highet notes,
Montaigne draws on a wide-ranging collection of topics belonging to
the many other genres he deforms and mingles in his work: the moral
treatises of Plutarch and Seneca, though longer, have similar titles,
presented in a varied list of topics that range from lofty ideals to the
functioning of the human body: Montaigne writes “Of Idleness,” “Of
Liars,” “Of Cannibals,” “Of the Power of the Imagination,” “Of
Drunkenness,” “Of a Monstrous Child,” “Of Three Good Women,”
“Of Cripples,” “Of Coaches,” “Of Smells.” In addition Montaigne
also appropriates the apology (“Apology for Raymond Sebond,” II/12,
the longest of the essays), letters (Seneca’s treatises sometimes take
the form of letters to friends; see Montaigne’s “Of the Education of
Children” written to Madame Diane de Foix, Comtesse de Gurson),
and even a type of subjective literary self-portrait whose origin
Montaigne attributed to the Roman Lucilius (see “Of Presumption).30
Montaigne peppers this mix with quotations, attributed and
unattributed, to such an extent that they become a part of the texture of
the work, interwoven strands, Montaigne addressing the past. (Highet
reproduces a list of 51 authors quoted by Montaigne; included are 110
quotations from Plato, 116 from Vergil, 148 from Horace, 149 from
Lucretius, 312 from Cicero, and 398 from Plutarch.31) Digressions and
interpolations are also part and parcel of the texture; note that the
28
Gilbert Highet, The Classical Tradition: Greek and Roman Influences on Western
Literature (Oxford: Oxford University Press), 192.
29
Northrop Frye, Anatomy of Criticism: Four Essays (Princeton: Princeton University
Press, 1990), 307. Frye also writes, “Montaigne’s scheme is to the confession what a
work of fiction made up of short stories … is to the novel or romance.”
30
Highet, 191-92.
31
Highet, 188-90, from P. Villey, Les Sources et l’évolution des essais de Montaigne
(Paris, 1908).
essay “Of Friendship” begins with a quotation comparing his writing

to the decorative grotesque (given below).
Montaigne was as aware of the fragmentation and deformity
of his Essais as was Petrarch of his Canzoniere. As Petrarch knew the
inadequacy of the forms of art, inventions of man, to contain the
perfection of love, Montaigne recognized the impossibility of
confining the imperfections of the human mind in the perfect forms of
art. He wrote:
If I had written to seek the world’s favor, I should have bedecked

myself better, and should present myself in a studied posture. I
want to be seen here in my simple, natural, ordinary fashion,
without straining artifice; for it is myself that I portray. My
defects will here be read to the life, and also my natural form, as
far as respect for the public has allowed.32
In a vivid description of their structure, Montaigne described

his hodge-podge as an image that anticipated one of Friedrich
Schlegel’s literary categories alternately labeled arabesque and
grotesque. He compares his Essais to the Renaissance adaptation of
decorative frescoes discovered in the late fifteenth-century excavation
of Nero’s “golden house” in Rome: a work made of parts, mere
decoration, a grotesque: 33
As I was considering the way a painter I employ went

about his work, I had a mind to imitate him. He chooses the best
32
Donald M. Frame, preface to The Complete Works of Montaigne: Essays, Travel
Journal, Letters by Michel de Montaigne, trans. Donald M. Frame (Stanford, CA:
Stanford University Press, 1957), 2.
33
They were labeled grotesque because viewers had to be lowered through tunnels
into the excavated rooms—hence the Italian root of grotesque, grotta, meaning cave.
Goethe’s description of them prompted Friedrich Schlegel’s adoption of the term for
digressive narrative. Wolfgang Kayser, The Grotesque in Art and Literature, trans.
Ulrich Weisstein (Bloomington: Indiana University Press, 1963), 19. The function of
these grotesques was decorative: profuse linear designs formed borders around a
series of small portraits. The content was fanciful: vegetation merged into both
composite creatures—creatures made of assembled parts that do not belong together
in nature—and other constructions that defied natural laws. The style of the grotesque
was adopted by Renaissance artists, notably Raphael, and subsequently spread
throughout Europe. Geoffrey Galt Harpham, On The Grotesque: Strategies of
Contradiction in Art and Literature (Princeton: Princeton University Press, 1982), 29.
spot, the middle of each wall, to put a picture labored over with all
his skill, and the empty space all around it he fills with grotesques,
which are fantastic paintings whose only charm lies in their
variety and strangeness. And what are these things of mine, in
truth, but grotesques and monstrous bodies, pieced together of
divers members, without definite shape, having no order,
sequence, or proportion other than accidental?
A lovely woman tapers off into a fish.

HORACE
I do indeed go along with my painter in this second

point, but I fall short in the first and better part; for my ability
does not go far enough for me to dare to undertake a rich, polished
picture, formed according to art.34
Montaigne’s preface announced his Essais as autobiography.

He began writing them in 1578, published the first edition in 1580;
and continued writing and publishing them until his death in 1592.35 A
true “work in progress,” they were presented in a series of editions
each containing new essays. They raise the expectation of a
chronological presentation and of an autobiographical account; yet
according to Montaigne, they are not organized, merely collected, and
not even chronology structures them:
… at each new edition, so that the buyer may not come off
completely empty-handed, I allow myself to add, since it is only
an ill-fitted patchwork, some extra ornaments. … Thence,
however, it will easily happen that some transposition of
chronology may slip in, for my stories take their place according
to their timeliness, not always according to their age.36
Countering the expected chronology is a deformed

chronology; countering the expected autobiographical account is a
34
Montaigne, “Of Friendship,” The Complete Essays of Montaigne, I:28, 135.
Harpham, 26, reports that this quotation is from Horace’s Ars Poetica.
35
Revised and enlarged editions appeared in 1582 and 1588; his final additions were
included in a posthumous edition of 1595, which was in all likelihood heavily edited
by Montaigne’s adopted daughter, Marie de Gournay, and his friend Pierre de Brach.
Richard L. Regosin, “Montaigne and his Readers,” A New History of French
Literature, ed. Denis Hollier (Cambridge: Harvard University Press, 1994), 249.
36
Montaigne, Essais III:9, Complete Works, 736.
series of disconnected wanderings of the mind. As the works

themselves form a grotesque around the absent center/subject—
Montaigne himself—so the individual writings form a grotesque
around the central topic announced by each title.
Our actions are nothing but a patchwork. … We are all

patchwork, and so shapeless and diverse in composition that each
bit, each moment, plays its own game.37
Anthony Grafton writes, “The introspective, wide-ranging Essays of

Montaigne sometimes resemble a set of commentaries set loose from
the texts they originally applied to.”38 Where Petrarch omitted the
frames, Montaigne has omitted the center—and what Montaigne
omitted, what he could not form according to art, or address directly,
was himself, the center portrait: “I am myself the matter of my
book.”39
Like Petrarch, Montaigne’s fragments may acknowledge the
inadequacy of memory and the ruin of time on both the physical
document and its interpretation. According to Rendall, Montaigne
claimed he had a bad memory, distrusted it, and subsequently chose
this format—nonlinear, following no system—over a discursive unity
that required accurate recall.40 “If my mind could gain a firm footing,”
he wrote, “I would not make essays, I would make decisions.”41 His
own research reinforced the belief that the text is not incorruptible,
that the future would change both its interpretation and its physical
state. After studying a Virgil manuscript in the Vatican Library, he
observed, “This Virgil confirmed me in what I have always judged,
that the first four lines that they put in the Aeneid are borrowed: this
book does not have them.”42 If Montaigne wrote in fragments to
acknowledged inevitable ruin, perhaps he also wrote in fragments to
37
Montaigne, Essais II:1, Complete Works, 243-44.
38
Anthony Grafton, The Footnote: A Curious History (Cambridge: Harvard
39
Frame, preface to Complete Works by Montaigne, 2.
40
Steven Rendall, “In Disjointed Parts/Par articles décousus,” in Fragments:
Incompletion and Discontinuity, 74-80.
41
Montaigne, Essais, III:2, Complete Works, 611, quoted in Dan Engster, “The
Montaignian Moment,” Journal of the History of Ideas, 59/4 (1998), 633.
42
Montaigne, “Travel Journal,” Complete Works, 950-51.
protect his work. Does the mangling of the already imperfect seem
less drastic than the ruin of perfection?
Like Petrarch, Montaigne wrote to preserve. If he chose his
fragmented forms to depict the random workings of the human mind,
of his mind specifically, then he collected to preserve that mind, to
preserve himself. He wrote for his friends and family so that they
might “keep the knowledge they have had of me more complete and
alive,” but perhaps he also wrote, as did Petrarch, to preserve himself
in his own memory, to create and recreate from that memory he so
mistrusted, a sense of who he was, the “matter of his book.”43
Neither Petrarch nor Montaigne could communicate, using an
ordered and complete text, the overwhelming nature of his subject—
for Petrarch, love; for Montaigne, his own identity. Both used
fragments to approach telling what could not be told. Their fragments
are separated by the space of the inexpressible and bound together by
what both the authors and readers know is missing, by a lack as
significant as the physical separation, the silent spaces between, that
remain the physical marker of form and format.
Both the lyric cycle of Petrarch and the essays of Montaigne
exist in a balance between the expectation of a narrative chronology
and its negation. Viewed as either the deletion of narrative from
longer forms (literally stripped from the prosimetrum or merely
omitted from the confession), or the addition of implied narrative to
forms that carried no such implication (the lyric collection and the
collection of aphorisms), both play on genre expectations. The essay
collection and the lyric cycle have become accepted genres, but they
are genres that retain their fragment heritage, genres in which the
blank spaces still imply missing material, still challenge the reader to
seek some overarching plan—even while they still invite the reader to
select passages at random, rather than to read straight through, to read
as Steven Rendall reports that Montaigne read his books, leafing
through now one, now another, without order or plan, by disconnected
fragments.44
43
See also Rendall, 73, for his discussion of Wayne Booth’s point that the Montaigne
presented in the Essais is also a fictional construct.
44
Rendall, 75.
Given the sheer numbers of their imitators and the acceptance

of the genres they created, these works of Petrarch and Montaigne
may now appear less radical, their fragmentation less extreme, than in
their own time. The opposite may be true for the works of Rabelais
and Cervantes, whose discontinuous narratives provided models for
Diderot, Sterne, and ultimately Schlegel and the Romantics; these may
present analytical problems to readers and critics today because the
works follow narrative norms that are no longer familiar, norms that
differ from the linear chronological model (from Aristotelian
expectations of beginning/middle/end, or what Harries describes as
temporal continuity and narrative flow45). As both Edwin M. Duval
and Carroll B. Johnson point out, an apparent lack of coherence and
order in the works of Rabelais and Cervantes can best be understood
as a transformation of the compositional logic, the norms and
conventions, of genres common to Classical, medieval, and early
modern literature: the epic, the epic quest, and the related epic or
chivalric romance. As Duval notes of Rabelais, these works are most
successfully understood when anachronistic notions of literary form
are put aside; the exaggeration and fragmentation of those unfamiliar
narrative conventions led later readers to view the books as even more
radical and extreme than perhaps originally intended.46
Rabelais
In 1534 and again in 1535, Rabelais was in Rome as private

physician to Cardinal Jean du Bellay. While the Cardinal joined the
ranks of French antiquarians quite literally seizing for France “the
glory that once was Rome,” dragging into exile artifacts to adorn
France’s gardens and palaces,47 Rabelais continued to dismantle the
45
Harries, 111.
46
Edwin M. Duval, The Design of Rabelais’s Pantagruel (New Haven and London:
Yale University Press, 1991), xiii-xvi; Duval, “Rabelais and Textual Architecture” in
A New History of French Literature, ed. Denis Hollier (Cambridge: Harvard
University Press, 1994), 154-59; Carroll B. Johnson, Don Quixote: The Quest for
Modern Fiction (Boston: Twayne Publishers, 1990), 71.
47
Eric MacPhail, “Antiquities and Antiquaries,” in A New History of French
Literature (Cambridge: Harvard University Press, 1994), 209-11.
literary traditions of the epic quest and romance for the diversion of
the French reader, creating a narrative of literary fragments often as
disjointed and as jumbled as the hodge-podge of artifacts being carted
off to Paris. Gargantua, one volume of Rabelais’s chronological
history of the lives and exploits of the fictitious giant Gargantua and
his son Pantagruel, opens with an address to the reader, a poem that
clearly states, “Vray est qu’icy peu de perfection.”48
In this book and the others in the series, Rabelais confiscated
the goal-directed quest of the epics of Homer and Virgil, the story line
of the medieval epic romance (the birth, education, and exploits of the
hero), and the episodic and digressive narrative fostered by both
conventions.49 He then peopled that mixed tradition with his own
bizarre creatures and situations, substituting giants (with rather coarse
habits) for the likes of the noble warrior Aeneas and a detailed
description of Gargantua’s baby clothes for the depiction of the shield
of Achilles.50
More than mere parody of character and situation, his works
parody the very structure of the genres, amplifying the norms and
conventions that contradict linearity—the in medias res beginnings,
the expected digressions, exaggeration of minor events, fusion of
history and folk tale—exaggerating but retaining enough scraps of
convention to avoid total chaos, to allow the work to remain
comprehensible against the model whole, the source of its fragments.
Gargantua begins with a fragment, as Harries has pointed
out.51 The narrator, Alcofribas Nasier (an anagram of François
Rabelais) has discovered (or perhaps invented) an account of
Gargantua’s lineage, taken from a book whose excessive, alliterative,
and contradictory description is a microcosm of the text itself: “un
48
Rabelais, Gargantua, Oeuvres complètes, ed. Mireille Huchon with François
Moreau (Paris: Éditions Gallimard, 1994), 3.
49
Duval, The Design of Rabelais’s Pantagruel, 1; “Rabelais and Textual
Architecture,” 157. For a discussion of medieval parody, Mikhail Bakhtin, Rabelais
and his World, trans. Hélène Iswolsky (Bloomington: Indiana University Press, 1984).
50
Rabelais, Gargantua, “Comment on vestit Gargantua,” 24-27; Homer, The Iliad,
trans. Robert Fagles (New York: Penguin Books, 1991), Book 18, “The Shield of
Achilles,” 482-84. Rabelais was inspired by a giant of folk legend named Gargantua;
Bakhtin, Rabelais and his World, 341-42.
51
Harries, 20-24, elaborates.
gros, gras, grand, gris, joly, petit, moisy livret” [“a fat, greasy, thick,
gray, pretty, tiny, moldy little book”] that is so worn it is impossible to
make out all the letters. He has translated it “practicant l’art dont on
peut lire lettres non apparentes” [“practicing that art by which one
reads invisible letters.”] At the end of this book, Alcofribas Nasier has
found a fragment whose beginning lines have been partially eaten by
rats and moths or “aultres malignes bestes” [“other malign beasts”];
the mutilated opening is reproduced in the text:
a i ? enu le grand dompteur des Cimbres

v sant par l’aer, de peur de la rousée,
‘ sa venu on a remply les Timbres
’ beure fraiz, tombant par une bousée
= uquel quand fut la grand mere arrousée
Cria tout hault, << hers par grace pesche le.
Car sa barbe est pres que toute embousée
Ou pour le moins, tenez luy une eschelle.>> 52
Alcofribas performs the job of the audience for any

fragmented work: he attempts to reconstruct a whole from the partial.
Rabelais thus begins with the mutilated text, questioning the accuracy
of writing, of telling, and of interpretation. Perhaps he also admits the
difficulty of making a beginning, the need for some outside marker to
answer the question, “why am I writing this?” His story has now been
validated by the outside world—and yet that beginning is torn and
unsure, just as any beginning must be.
Rabelais’s characters are monsters, giants out of context in the
human world, “uncompleted in the sense of the overcompleted,
overfinished,” the whole plus more, added material that should not be
present.53 Physical deformity and character deformity abound in all the
books.54 Not only the physical descriptions of Rabelais’s characters,
but many of their situations as well, are reminiscent of the Roman
52
Rabelais, Gargantua, “De la généalogie et antiquité de Gargantua,” 11; other
sources use various spacings and symbols to represent the varmin-chewed holes.
Harries, 23, notes that, as part of Rabelais’s joke, the complete lines in this and the
rest of the fragment make no more sense than the incomplete ones.
53
David Williams, Deformed Discourse: The Function of the Monster in Mediaeval
Thought and Literature (Montreal: McGill-Queen’s University Press, 1996), 113.
54
Lawrence D. Kritzman, “Rabelais’s Splintered Voyage,” in Fragments:
Incompletion and Discontinuity, 53-70; Kritzman includes an extensive discussion.
grotesques. Gargantua’s mother, for instance, eleven months pregnant,

ill from having eaten too much tripe, has taken an astringent that
constricts all her sphincter muscles. As a result, Gargantua cannot be
born in the normal fashion.
By this mishap were loosened the cotyledons of the matrix,

through which the infant sprang up into the vena cava; and,
climbing up by the diaphragm up above the shoulders, where the
said vein divides in two, took the route to the left, and came out
through the left ear.55
Though recalling the birth of Athena from the head of Zeus,

this image perfectly illustrates Bakhtin’s description of Rabelais’s
writing, filled with forms “interwoven as if giving birth to each
other.”56 The image would also find company with the decorative
grotesques recalled by Montaigne when describing the lack of order,
coherence, and logic in his own Essais.57 Edwin M. Duval uses this
same image, and this same reference from Horace, to describe modern
reaction to the convoluted structure of Rabelais’s works:
… their inconsistencies and internal contradictions, their episodic

structure, their radical open-endedness—conspire to suggest to
postclassical readers not a coherent and ordered composition but
something akin to those grotesque composite images described by
Horace, which start as a beautiful woman and end as an ugly fish
…58
The forms Rabelais’s writings take are as varied and absurd as

his characters and their situations. A partial list includes pointless
insertions; quotations and, more often, misquotations; long poems;
series of clichés and maxims, strung together with no connectives; and
whole chapters that consist of nothing but lists, often in double
columns:
55
Rabelais, The Complete Works of François Rabelais, trans. Donald M. Frame
(Berkeley: University of California Press, 1991), 20.
56
Bakhtin, 32.
57
Montaigne, “Of Friendship,” Essais, I:28 in Complete Works, 135, quoted above.
58
Duval, The Design of Rabelais’s Pantagruel, xiii-xiv.
Then, when the green cloth was laid out, they brought out plenty
of cards, plenty of dice, and enough boards for checkers or chess.
There he played:
Flush, Primiera,
Grand slam, Robber,
Trump, Prick and spare not,
One hundred, The spinet,
Poor Moll, The fib,
… [continuing for 206 more entries to end the chapter]59
In the end, Gargantua does complete his quest, but the

achievement of his goal is compromised: the book ends not at the
point of his military victory, but with a disagreement about the
meaning of another found text, this one engraved on a bronze plate.
Gargantua ends as it began, with an undecipherable fragment text.
Framed by the uncertainty of time on the written word—on the
physical document and on its interpretation—Rabelais’s story states
clearly what Petrarch’s Canzoniere only hints: the ravages of time
work on both document and meaning. The physical condition and
circumstances of found manuscripts—damaged, incomplete, nearly
illegible, only partly decipherable, author or authors unknown—bring
to the forefront questions of origin, originality, authenticity, and the
power of language to communicate over time.
Gargantua began with a fragment; it also began with an
ending—the fragment text reproduced was found at the end of the
moldy book of Gargantua’s lineage; the story ends with another
fragment text. Rabelais thus questions the possibility of any real
beginning or ending. Gargantua completes his quest, but it is only the
end of a task, the end of another in an ongoing series; the episodic
form and content of the epic and the romance ever invite new
episodes, unending challenges, sequels continuing even after the
contrived “death” of the hero, since tales from the past may always be
told, though never accurately and always through interpretation.
Rabelais made fragments for another reason: to make people
laugh. Borrowing from the comic traditions of his time, as Mikhail
Bakhtin has shown in Rabelais and His World, he confronted literary
traditions and the world around him, offering his criticism and
commentary through grotesque deformity of genre and of character.
59
Rabelais, Complete Works, 50-51.
His giants are as out of proportion in this world as his digressions are
out of proportion in the epic. Where other writers will take the
grotesque ornament complex and the single grotesque creature to
express horror, the evil side of the unnatural, Rabelais and his
followers will use them to make the reader laugh, to make fun—even
cruel fun—of the conventions of literature and society.
Cervantes
Cervantes, too, contains Don Quixote’s grotesque parody of

knighthood and quest, his windmill enemies and magical barber’s
bowl helmet, within fragments of genres he inherited. Cervantes
directly confronts his principal model in his introduction, proclaiming
that he writes a satire of chivalric romance. As Kenshur notes,
Cervantes plays on the components of the romance that were singled
out when critics compared the romance to the ideal epic: though both
are episodic, multiple plots and digression are the elements that
Cervantes highlights and expands,60 the elements he takes to excess.
Don Quixote is monstrous, as over-filled with digression as Gargantua
is over-sized.
The physical fragment and its defects continued to provide
props and structural models for later digressive narratives.61 Where
Rabelais began Gargantua with a text missing the beginnings of lines,
Cervantes opens Don Quixote with a poem written in versos de cabo
roto, verses in which the final unstressed syllable of each line is
omitted, for the reader to supply; the poem is the first of his fake
preliminary endorsements. In Don Quixote, Cervantes retains the
illusion of a dependence on the physical fragment as the narrator, Cid
Hamete Benengeli, tells Don Quixote’s misadventures through a
framework of incomplete found manuscripts, suffering the appropriate
deterioration of age and munching vermin, missing endings or
beginnings, written in exotic languages that must be translated, falling
60
Kenshur, 68 n. 5.
61
Harries, especially Chapter I.
into the hands of the narrator completely by chance, in strange

marketplaces.62
In addition to these old manuscripts that supposedly generated
the story of Don Quixote, other texts appear—texts that Cid Hamete’s
story itself has generated: an unauthorized, printed version of the first
manuscript and a spurious sequel result in Don Quixote’s being
followed about by two different versions of himself, reduced to
wearing a sign declaring his identity. Cervantes breaks Don Quixote
onto fragments by duplication, just as surely as Petrarch made
fragments of Laura by breaking her body into separate parts. The
Don’s duplicates and the manuscripts that are the sources for them are
fragments of replication and destroyers of true origin. They raise
questions of authenticity that remain problematic today—questions
Borges addresses in “Pierre Menard, Author of the Quixote” and Andy
Warhol with his multiple Marilyns.63
Within this frame of fragmented and duplicated texts,
Cervantes draws on an encyclopedic array of literary fragmenting
techniques: temporal paradoxes and inconsistencies, mixed styles,
borrowing and quotation, repetition, multiple perspectives, authorial
intrusion, digressions that range from interpolated stories to related
asides—and the famous digressions that condemn digression.
Boy, boy,” said Don Quijote at this point, in a loud voice, “go
straight ahead with your story, and do not go curving off at a
tangent; for it requires much proof and corroboration to bring a
truth to the light.64
The chivalric romance, the mores of Renaissance Spain, mean

little to us today; our Don Quixote is still a sympathetic rebel, but now
he sings his impossible dreams to full orchestral accompaniment. Yet
common themes do traverse the centuries, and if a foolish old dreamer
can still touch our hearts, a rambling story can still lose our attention.
One paperback edition of Don Quixote advertises itself on the title
62
The narrator first relies on a manuscript that ends in the middle of a battle. The
outcome can be related only after the discovery of a second manuscript that picks up
at the point the first left off. A third manuscript, appropriately chewed, must
eventually be found to allow Cid Hamete to finish the story.
63
Harries, 24-26, gives a more complete analysis of this aspect of Don Quixote.
64
Trans. Kenshur, 57.
page as “an abridged version designed to relate without digressions

the principal adventures of the Knight and his Squire.”65
Why does Cervantes write fragments? Kenshur believes that
for Cervantes, order was founded on the subordination of all detail to
truth: pertinence to truth was the criterion for inclusion or exclusion.
In his view, when Cervantes/Cid Hamete realizes that the truth cannot
be known, he is faced with an unsolvable problem: he has no basis for
excluding or including anything. Digression, the representation of
disorder and the inability to know, becomes his only option.
Ultimately, “Cervantes’s subject is not reality itself, but the human
efforts to comprehend it,”66 and ultimately, that has been, and will
continue to be, a driving force behind such fragmented works.
Cervantes, like Petrarch, Montaigne, and Rabelais, also writes
fragments to confront the impermanence of life and of the written
word. Like the medieval monsters that must inhabit only the margins
of manuscripts or the edges of maps, Don Quixote exists on the
fringes of acceptable society, and Cervantes’ story exists on the
fringes of acceptable narrative. When ancient, monstrous creatures are
brought into the light of modern thought, they seem silly and useless;
when Cid Hamete’s ancient manuscripts are brought out of their dark
hiding places, they crumble; when Don Quixote is brought back to his
real life, brought back in from the edge, he dies. As Martínez-Bonati
notes, there is in Don Quixote a layer that is “disquieting, pessimistic,
and tragic, where the ultimate reality of the body creeps in, with its
inevitable corruption of all happy dreams.”67
There is also the layer that, like Rabelais, makes us laugh.
Unlike Gargantua, Don Quixote is deformed more in mind than in
body: the juxtaposition of his abnormal interpretation of sensory data
against a “normal” interpretation is often the source of grotesque
comedy. Don Quixote makes us laugh, but unlike Gargantua, he often
brings us laughter that is bittersweet, tinged with sadness and pity.
Rabelais and Cervantes would have us laugh because laughter is often
65
Cervantes, Don Quixote of La Mancha, trans. and ed. by Walter Starkie (New York:
The New American Library, 1957. My emphasis.
66
Kenshur, 67.
67
Félix Martínez-Bonati, “Don Quixote” and the Poetics of the Novel, trans. Dian
Fox (Ithaca: Cornell University Press, 1992), 225.
the only response to the impossible. They send their deformed,

monstrous characters to face improbable situations; they send their
deformed and monstrous books—scraps pieced together, seams
showing, full of holes—to confront their own improbable monsters:
the impossibility of representing and of comprehending reality, time
and the reality of the ruin, the straitjacket of conformity, the ultimate
futility of the written word, the blank spaces that neither inspiration
nor craft will fill.
These digressive narratives become models for writers in the
centuries that follow, as each uses form, or the disruption of form, to
address the issues of time, place, and personal need for expression.
Northrop Frye, in fact, gives this type of digressive narrative a place
among his four genres of continuous prose; he labels it anatomy and
notes its frequent combination with his other categories: novel,
confession, romance. He traces a lineage forward from Rabelais
(where he believes it is combined with the novel) through Swift and
Sterne, to Huxley and into the twentieth century. Tracing backward,
he moves from Rabelais through Erasmus, Lucian, Apuleius, and
Petronius to Menippean satire, a form named for Menippus, a
third-century BCE author whose writings alternated prose and verse.68
Thus Menippean satire was a subcategory of prosimetrum, but, as
Peter Dronke notes, later definitions of this form, and later works in
this category, omit or downplay the mixture of prose and verse in
favor of other elements: mixed tone (serious and comic, high and low)
and linguistic styles (macaronic texts, slang, exotic words, coinages,
prayers, baudy jokes), the fantastic as well as paradox and contrast.
These juxtapositions, these combinations of disparate elements, create
the literary grotesque; they are the means by which such forms
undermine what Dronke calls the established decorum of genres and
diction, the means by which such works question and undermine
authority.69
Thus the prosimetrum, an example of mixed genre if
measured by classical notions of genre purity, appears as the ancestor
68
Frye, 308-312, and Dronke, 1. All but fragments of the writings of Menippus are
lost. See Dronke for a brief summary. Though Frye’s categories have met with
criticism, this is not a lineage he developed, but one well-known to the
nineteenth-century; see below.
69
Dronke, 1-6.
of both the digressive narrative and the disconnected fragment, works

that, left in pieces or overfilled, providing too much information or too
little, can incorporate in their overall structure all the markers of the
fragmented text: the mutilated beginnings and missing endings, the
holes, the extraneous pages erroneously bound, the sections
mistranslated and misunderstood, the quotations, the digressions, even
the fragments of perfection scattered about.
As mentioned at the beginning of this section, contemporary
readers were familiar with the plot outline and the episodic nature of
the models Rabelais and Cervantes used, even made fun of; these
writers could weave their infinitely diverse digressions throughout
their books, yet to some extent avoid for their own readers the level of
chaos later readers and critics may perceive when they compare these
narratives to the linear, chronological model. The perception of the
balance between continuity and disruption, the identity and meaning
of the fragment, changes for the reader who is not familiar with the
model the author has mutilated. Yet, as Cervantes’s diatribes on
digression clearly show, Aristotle’s classification of episodic plots as
inferior and plots with beginning, middle, and end as superior, was
already predominant.70 Modern readers may read and identify these
works by their variance from conventions that, despite our familiarity
with stream of consciousness works and the open novel, still rely on
Aristotelian ideals. At bottom, regardless of genre, the strongest model
for a story, against which it will be measured positively or negatively
or neutrally, was given by Don Quixote: “tell it straight through,” not
in isolated and disconnected moments lacking narrative glue, and not
in narratives that lack proper beginnings and endings, that are filled
with holes and stuffed with unrelated material. Conventions exist, or
come into being, that incorporate these abnormalities, thus making the
sense of fragmentation less sharp, but rarely do they erase that sense
of what narrative ought to be when the reader is led to expect it.
70
Martínez-Bonati, 106.
Seventeenth- and Eighteenth-Century Fragments
The seventeenth and eighteenth centuries produced rich stores

of fragments. Building on Renaissance predecessors, writers tailored
both the digressive narrative and the collection of disconnected units
to express contemporary concerns. These forms of fragmentation
provided the models that would open the way to the Romantic
fragment, models so numerous and rich that Harries applies one of
Schlegel’s famous aphorisms (“Many works of the ancients have
become fragments. Many works of the moderns are already fragments
at the time of their origin.”)71 to the seventeenth- and
eighteenth-century tradition rather than, as is the more common, to the
“broken, incomplete texts that proliferated in the Romantic period.”72
In contrast to writers who often used fragments to represent
chaos and disorder, some seventeenth- and eighteenth-century writers
used them to express belief in an underlying, though incompletely
understood, natural order. Cervantes, Petrarch, and Montaigne
fragmented to represent the impossible. Bacon, Diderot, and Hamann
chose both the form of the aphorism (disconnected and disjunct
thought) and its physical format (physical separation of short sections
on the page) to express concerns about man’s ability to organize and
to categorize the knowledge he acquires, but they also indicated a
belief in the existence of a system—a system that man could, through
examination and logic, come ever closer to understanding. To
encourage further investigation, they used the fragment and its ability
to encourage—or perhaps demand—a search for more.73
In “Knowledge Broken,”74 Kenshur compares Bacon’s
seventeenth-century Novum Organum and its eighteenth-century
counterpart, Diderot’s De l’Interprétation de la nature. Both are
presented in aphoristic format—in short sections separated by blank
spaces. Kenshur credits Montaigne’s Essais and the ancient Greek and
Roman collections of aphorisms as models for both. These more
71
Harries, 2, citing Friedrich Schlegel, Athenaeum Fragment 24.
72
Harries, 5.
73
See Harries and Kenshur for detailed discussions of the following.
74
Kenshur, 38-54.
recent works are concerned with the interpretation of the outside

world through scientific method, and with recording those
interpretations. Bacon, by his own admission, believed that the state of
man’s knowledge was incomplete; to express that incomplete
knowledge in a written form that was systematic and gave the
impression of completeness was misleading. Such a form did not
inspire further investigation. He turned to the authority of the ancients
for a solution:
But the first and most ancient seekers after truth were wont, with
better faith and better fortune, too, to throw the knowledge which
they gathered from the contemplation of things, and which they
meant to store up for use, into aphorisms; that is, into short and
scattered sentences, not linked together by an artificial method.75
Aphorisms represented “knowledge broken” and would

“invite men to enquire farther”; whereas a systematic presentation,
appearing to be complete, would “secure men, as if they were at
furthest.”76 For Bacon, the aphoristic format represented fragmentary
knowledge and stimulated further thought and investigation, but
according to Kenshur, it was primarily the format—the short
paragraphs separated on the page by white space—that Bacon
exploited. His ideas actually flow together and would make perfect
sense if written without the spaces.77
Diderot, in his concern with the way nature presents
information to us in disconnected fragments, turns not only to the
aphoristic format, but also to the aphoristic form; the sections of De
l’Interprétation de la nature do not flow together to make a cohesive
narrative.78 His first aphorism echoes Montaigne and a concern not
only with the way nature gives information, but also with the way the
brain stores and retrieves it.
Je laisserai les pensées se succéder sous ma plume, dans l’ordre

même selon lequel les objets se sont offerts à ma réflexion; parce
75
Kenshur, 41, quoting Bacon, Novum Organum (English trans.), Book 1, aphorism
86, in Works, ed. Spedding, Ellis, and Heath (Boston, 1861-64), 8:120-21.
76
Kenshur, 40, quoting Bacon, “The Advancement of Learning,” in Works, 6:292.
77
Kenshur, 44-45.
78
Kenshur admits there are other opinions.
qu’elles n’en représenteront que mieux les mouvements et la

marche de mon esprit.
[I shall let my pen record my thoughts in the same order in which

the objects have presented themselves to my reflection; thus they
will better represent the changes and the progression of my
mind.]79
Diderot’s statement also coincides with the attitude of another

fragment writer in Harries’s study, Johann Georg Hamann:
My remarks will be just as unsuitable for a specific treatise as they

would be for a book. I will pursue those thoughts that I come
across, and follow them at my convenience; … truths, principles,
systems are too much for me. Crumbs, fragments, whims, ideas.
Each to his own.”80
Hamann is more embedded in Christian thought and belief

than in scientific study, but Harries interprets his writings as yet
another example of the ability of the disconnected and the
unsystematic to reflect our inadequacy to comprehend and yet to
express the belief in an underlying greater order.81 Implying the whole
through the use of the fragment creates the same sort of tension
Petrarch created when he implied his narrative but refused to supply it,
the tension between what we can express and what we believe to exist.
As Frye noted (see above), the line of succession in the
digressive narrative continues in the eighteenth century with the works
such as Swift’s A Tale of a Tub and Gulliver’s Travels; Sterne’s The
Life and Opinions of Tristram Shandy, Gentleman, and Sterne’s great
model, Diderot’s Jacques le fataliste. All announce a narrative, by title
(a tale, a life) or opening—Jacques opens with the Master’s initial
request that Jacques tell the story of his loves—but these titles and
opening strategies are only setups, backdrops for narratives that thwart
the very expectations they raise. In stories often still revolving around
the requisite journey or quest, ruins and ruined manuscripts, the
79
Original and translation, slightly modified, from Kenshur, 41, recalling Montaigne,
Complete Works, 721, “And when shall I make an end of describing the continual
agitation and changes of my thoughts, whatever subject they light on …”
80
Harries, 36.
81
Harries, 36-37.
atemporal quality of memory, and the interruptions of daily life, these

writers question how, where, and when a story begins and ends; they
weave and interweave digressions that overpower, obscure, or (in the
case of Tristram Shandy) completely block the supposed main
narrative; and they interrupt the idea of the fiction itself by allowing
the author to intrude in the text, providing examples Schlegel will
draw on to develop his theories of Romantic irony.82
Walter Bagehot’s negative evaluation and description of
Tristram Shandy not only sums up Sterne’s debt to Rabelais, and in
part to the medieval romance tradition, but also illustrates the
importance given to the chronological model in narrative, the very
model these writers flaunt:
In Tristram Shandy especially there are several defects which,

while we are reading it, tease and disgust so much that we are
scarcely willing even to admire as we ought to admire the refined
pictures of human emotion. The first of these, and perhaps the
worst, is the fantastic disorder of the form. It is an imperative law
of the writing-art, that a book should go straight on. A great writer
should be able to tell a great meaning as coherently as a small
writer tells a small meaning. … But Sterne’s style is unnatural. He
never begins at the beginning and goes straight through to the end.
He shies-in a beauty suddenly; and just when you are affected he
turns round and grins at it. “Ah,” he says, “is it not fine?” …
People excuse all this irregularity of form by saying that it was
imitated from Rabelais. But this is nonsense. Rabelais, perhaps,
could not in his day venture to tell his meaning straight out; at any
rate, he did not tell it. Sterne should not have chosen a model so
monstrous. Incoherency is not less a defect because a foreign
writer once made use of it.83
Defects, disgust, disorder, unnatural, monstrous, incoherency:

these words describe the digressive narrative—negatively to Bagehot,
more positively to those who treasure the playfulness, the humor, the
disrespect in these parodies of the act of writing, the act of producing
a fiction.
82
Lloyd Bishop, Romantic Irony in French Literature From Diderot to Beckett
(Nashville, TN: Vanderbilt University Press, 1989), 31-32.
83
Walter Bagehot, Literary Studies, Vol. II, ed. Richard Holt Hutton (London:
Longmans, Green, and Co. 1879, reprint New York: AMS Press, Inc. 1973), 118-19.
Schlegel and the Romantic Fragment
The fragment and the fragmented lie at the heart of Friedrich

Schlegel’s critical and aesthetic theories, but as Raymond Immerwahr
has noted, Schlegel was a discoverer, not an inventor.84 In formulating
his theories, he drew on the works of the past, among others on
Shakespeare, Montaigne, Cervantes; he labeled Petrarch’s poems
“classical fragments of a novel,”85 and noted Diderot’s authorial
intrusions: “When Diderot does something really brilliant in his
Jacques, he usually follows it up by telling us how happy he is that it
turned out so brilliantly.”86 His preferred contemporary writers in their
turn had been influenced by the earlier fragments as well: Hoffmann’s
debt to Sterne is well known, Jean Paul translated Cervantes and
mentions both Rabelais and Cervantes as models for his Der Komet,
Tieck’s essays discuss works of Rabelais, Cervantes, and Sterne, and
so on.87
Schlegel’s fragmented works are among his most influential:
his collections of fragments—the Athenaeum Fragments, the Critical
Fragments, the Ideas—are aphoristic in form (disjunct thought) and in
format (physical separation on the page), and his only novel, Lucinde,
is a digressive narrative and a mixture of genres—an arabesque or
grotesque. Lucinde is autobiographical, written when Schlegel was 26
(in 1798-99); it follows the life and loves of Julius (Schlegel), ending
with his relationship with Lucinde (Dorothea Veit, whom Schlegel
84
Raymond Immerwahr, “Romantic Irony and Romantic Arabesque Prior to
Romanticism,” The German Quarterly 42 (1969): 665; Philippe Lacoue-Labarthe and
Jean-Luc Nancy, The Literary Absolute, translated by Philip Barnard and Cheryl
Lester (Albany, New York: State University of New York Press, 1988), 40. Originally
published as L’Absolu Littéraire (Paris: Éditions du Seuil, 1978).
85
Friedrich von Schlegel, Literary Notebooks, 1797-1801, ed. Hans Eichner (Toronto:
University of Toronto Press, 1957), 50. Quoted in Harries, 14, from Hans Eichner,
introduction to Friedrich Schlegel, Charakteristiken und Kritiken I (1796-1801), ed.
Hans Eichner, Kritische Friedrich-Schlegel-Ausgabe, vol. 2 (Munich: Schöningh,
1967), lvi, n. 4.
86
Friedrich Schlegel’s Lucinde and the Fragments, 143.
87
Steven Paul Scher, “Hoffmann and Sterne: Unmediated Parallels in Narrative
Method,” Comparative Literature, 28 (1976): 309-25; Eric A. Blackall, The Novels of
the German Romantics (Ithaca: Cornell University Press, 1983), 66-67, 276.
married) and the birth of their child. The work is far from a
chronological narrative or comprehensive story; it is a sequence of
atemporal images (only the central episode, “Apprenticeship for
Manhood,” is chronological) with virtually no plot, and is concerned
with the internal workings of the mind rather than external matters,
with reflection on events rather than events themselves. Schlegel tells
his story through a mélange of thirteen sections of unequal length:
narratives, dialogues, letters, visions, confessions, and character
sketches; the work thus tends toward his notion of the perfect, though
necessarily unattainable, novel that would combine all previous
genres.
Julius writes to Lucinde in the opening section of the novel,
“Confessions of a Blunderer”:
No purpose, however, is more purposeful for myself and for this

work, for my love for it and for its own structure, than to destroy
at the very outset all that part we call “order,” remove it, and
claim explicitly and affirm actually the right to a charming
confusion. This is all the more necessary since writing about our
life and love in the same systematic and progressive way we
experienced them would make this unique letter of mine
insufferably unified and monotonous, so that it would no longer
be able to achieve what it should and must achieve: namely the
re-creation and integration of the most beautiful chaos of sublime
harmonies and fascinating pleasure. I’m making use, therefore, of
my incontestable right to confusion and am inserting here—in
quite the wrong place—one of the many scattered pages I wrote or
scribbled on when I longed for you …88
In this short passage, already self-reflexive, Julius as

protagonist describes the work in which he appears.89 The form Julius
depicts—the chaos, the discrediting of order and temporal
progression, the right to digress “in the wrong place,” and the genre
mix that follows—perfectly describes Schlegel’s arabesque structure.
Yet as Eichner points out, the visual arabesque, despite its apparent
chaos, produced a coherent unit, and Schlegel considered this a
88
89
Blackall, 39 and Firchow, introduction to Friedrich Schlegel’s Lucinde and the
Fragments, 30; Firchow connects this and the many other such internal descriptions of
the structure of the work to Sterne’s use of Tristram as narrator.
requirement. Athenaeum Fragment 305 states, “so the grotesque …

loves the illusion of the random and the strange.”90 Behind the illusion
of Julius’s chaos Firchow describes the structure of Lucinde: six short
sections show Julius as he is, a central and much longer middle section
(the “Apprenticeship for Manhood”) tells how Julius came to be what
he is, and six final sections balance the first six to detail the direction
of his growth.91 A sort of order can be made of Lucinde’s fragments,
even though that order is best described via its deviations from the
expected chronological model it deforms.
Schlegel also relied on other tactics of earlier fragment
narratives: the found manuscript ploy and the value given the
unmediated intensity of the sketch are combined here as Julius’s
excuse to include “one of the many scattered pages I wrote or
scribbled on when I longed for you”; chance and the interruptions of
daily life (that sidetracked the entirety of Tristram’s life) intrude as
later Julius is interrupted “by rude and unkind chance,” which he
intends to mold to his purpose: another opportunity to expound his
theories of literature. Unlike the attitude toward memory taken by
other authors of fragmented works, Schlegel, at least as he speaks
through Julius, neither mistrusts it, as did Montaigne, nor seeks to
preserve it intact; rather, the knowledge that memory both ruins and
preserves—preserves but in an imperfect state, as memories
deteriorate with time, become mixed with other memories, are subject
to reinterpretation at the time of remembrance—becomes a tool, an
adjunct to creativity and imagination, allowing Julius to avoid
recording life in “the same systematic and progressive way we
experienced” and enabling him to accomplish his “re-creation and
integration of the most beautiful chaos of sublime harmonies and
fascinating pleasure.”92
Many discussions of Lucinde dwell on the work as
representative of Schlegel’s theories, but Lucinde is also a work of
literature, not just a work of theory. Schlegel chose the subject matter
and the form because they served each other. Lucinde is a love story
90
Friedrich Schlegel’s Lucinde and the Fragments, 205; for more see Hans Eichner,
Friedrich Schlegel (New York: Twayne Publishers, Inc., 1970), 64. My emphasis.
91
Firchow, introduction to Friedrich Schlegel’s Lucinde and the Fragments, 32.
92
Quotations from Friedrich Schlegel’s Lucinde and the Fragments, 45.
recreated in memory, not as real events occurred, not as laws of

narrative dictate, but remade by imagination, embroidered with
emotion, reordered and refocused, in the full overabundance and
confusion of love. Schlegel recalls writing Lucinde: “naively and
nakedly … in the thoughtlessness of youth, I made Lucinde reveal the
nature of love in an eternal hieroglyph.”93
It is not only internally that Lucinde is fragmented: Schlegel
intended to continue it with a collection of poems, and in its complete
version it was to have been one of a set of four novels.94 Thus it is one
of those many Romantic endeavors doomed to incompletion because
its scope exceeded either the ability of the author or the form chosen
to encompass it.95
Lucinde was not well received, partly for its blatantly
autobiographical sexuality but primarily for its form (or apparent lack
of form). When it was reissued in 1835, Heine called it “ludicrously
Romantic”; in 1870, Rudolf Haym labeled it an “aesthetic
monstrosity,” and his contemporary Wilhelm Dilthey wrote that it was
“morally as well as poetically formless and contemptible.” Though
reception is more positive today, Firchow writes of critics who are still
“slow and grudging” in giving approval. Born on the margins of
acceptable formal structure and acceptable content, Lucinde was
moved to the margins of literature (Schlegel even omitted it from the
publication of his complete works in 1823) and has lived there, with
other monsters, since its publication in 1799.
Schlegel’s collections of fragments were more successful and
more influential than Lucinde in his own time, and remain so today.
The German translation of Chamfort’s Pensées provided an immediate
inspiration, though Schlegel certainly had a rich store of other
93
On Incomprehensibility, Friedrich Schlegel’s Lucinde and the Fragments, 261.
94
95
See Balachandra Rajan, The Form of the Unfinished: English Poetics from Spenser
to Pound (Princeton: Princeton University Press, 1985); Marjorie Levinson, The
Romantic Fragment Poem: A Critique of a Form (Chapel Hill: The University of
North Carolina Press, 1986); and Thomas McFarland, Romanticism and the Forms of
Ruin: Wordsworth, Coleridge, and Modalities of Fragmentation (Princeton: Princeton
University Press, 1981) for discussions of such unfinished works and variations on
categories of the incomplete. Examples include Coleridge’s Christabel, Shelley’s The
Triumph of Life, and Byron’s Don Juan. The range of such works extends from
Chaucer’s Canterbury Tales to Pound’s Cantos.
examples including the aphorism collections of the ancients, the

essays of Montaigne, the writings of Pascal and La Rochefoucault, the
collection of “Fragmente” included in Herder’s Über die neuere
deutsche Literatur (1767). In his collections of fragments, Schlegel
retained the fragment format, but created a grotesque form that
combined aspects of the genres he inherited: the brevity of the
aphorism and its tendency to say much in few words, the accidental
incompletion of the aphorism collections of the ancients (lost or
mutilated manuscripts, deciphering texts so far removed from their
time and context, etc.), the lack of developed discourse or logically
presented evidence that marked the essay, the wide variety of topics
that might be covered in a collection of aphorisms or essays, and a
certain sense of unity that derives from the thoughts or personality of
the author rather than from the subject matter.96
In the Athenaeum Fragments in particular, Schlegel presented
a disconnected stream of ideas in widely varying lengths and forms,
some contributed by other writers (August Wilhelm Schlegel,
Schleiermacher, Novalis): in addition to individual fragments that are
in length similar to aphorisms, Schlegel included what he termed
condensed essays and reviews,97 and mini-dialogues posed as a single
question and a reply. In content, all are more open, less self-contained,
than previous examples. Sometimes this open quality is clearly
evident from the construction, as in Athenaeum Fragment 73,
beginning in mid-thought, ending with a question: “Might it not be the
same with the people as with the truth: where, as they say, the attempt
is worth more than the result?”98 In others, straightforward sentence
structure and brevity work at odds with content that constantly
questions interpretation—”The historian is a prophet facing
backward”—and in many, Schlegel’s terminology made the fragments
difficult to understand even in his own time. Rarely, if ever, does one
of Schlegel’s little hedgehogs, no matter how “isolated … and
96
This list is expanded from Lacoue-Labarthe and Nancy, 40.
97
98
complete in itself,”99 ever appear, as Michel Chaouli states, as a

“shrink-wrapped version of thinking.”100
Against the disorder of this bizarre mixture of incomplete,
miniature genres on wide-ranging topics, Schlegel raises the
expectation of plan, system—unity—through the tension he creates
between the aphorism as a complete statement and the incompletion of
the fragment, between the fragment’s right to stand on its own and its
need to reach out both to other members of the collection (some
containing content that will contradict, others that will confirm) and
outside the work itself to the external knowledge and experiences of
the reader. To the extent that each fragment is independent, it destroys
unity; to the extent that each relies on other members, or reaches out
toward them for completion, it raises the possibility of unity or
system; to the extent that each reaches outside the collection itself, it
again destroys unity, or posits it outside the collection itself. That
tension between part and whole questions the role of the fragment
within the collection and subsequently the identify of the collection
itself.
In Athenaeum Fragment 53 Schlegel wrote: “It’s equally fatal
for the mind to have a system and to have none. It will simply have to
decide to combine the two.”101 By stressing the interdependence of the
fragments, their reliance on each other for “mutual support and
clarification,” Eichner claims that, through the collections of
fragments, Schlegel has at least attempted the impossible—the
combination of system and non-system—by creating a “system of
fragments” (a phrase Schlegel uses to describe himself, see below).102
Firchow also claims these fragments form a system, a substitute for a
formulated system that appealed to Schlegel because it did not need to
exclude even the contradictory. He notes that despite the varied
subject matter, the collection is primarily literary and forms what he
calls an unusual and “curious form of criticism.” From these
fragments, “ruins and not complete edifices,” as Firchow labels them,
99
Athenaeum Fragment 206, Friedrich Schlegel’s Lucinde and the Fragments, 189.
100
Michel Chaouli, The Laboratory of Poetry: Chemistry and Poetics in the Work of
Friedrich Schlegel (Baltimore: The Johns Hopkins University Press, 2002), 55.
101
102
Eichner, 48.
the reader must construct a system that, in fact, never existed. From
this perspective they recall the fake ruins so prominent in the
eighteenth century, built to resemble something once whole, now
broken, but actually inviting the viewer to mentally reconstruct
something that never actually existed.103 Schlegel’s fragments are
simultaneously sketch and ruin, with the potential to inspire both
creation and recreation.
Just as Montaigne blamed his essays on the limitations of his
own mental ability, Schlegel, too, claimed the fragment as a result of
his own mental makeup: “I can give no other ‘echantillon’ [swatch] of
my entire ego than such a system of fragments because I myself am
such a thing.”104 Dorothea evidently described Schlegel’s entering a
“ceaseless stream of his thoughts” as brief statements in notebooks—
“unconscious processes and flashes of intuition.”105 Chaouli reports
thousands of entries in 150 notebooks,106 undeveloped ideas intended
for a theory of literary criticism that, as with the Lucinde project,
Schlegel never completed. It was from these notebooks that Schlegel
drew many of the fragments he included in published collections—
works that, themselves, resemble the sketchbook: a hodge-podge of
topics in genres of various lengths gathered together in no apparent
order, with no sense of opening or closure.
Schlegel’s theories, in part, were his justification of the
fragment and fragmented, a defense of structures that already existed
in literature, and that offered promise for literature of the future. He
turned to them for his own literary expressions because he believed in
their ability to accommodate both order and chaos—a chaos from
which order may be intuited but never completely realized. As Bishop
writes, the acceptance of the fragmentary is an attempt to transcend
contradiction rather than resolve it.107 Because the fragment itself is
never stable, the tension between part and whole allows the
irreconcilable conflict between the absolute and the relative,
perfection and reality. On a pessimistic note, Behler and Struc claim
103
Firchow, introduction to Friedrich Schlegel’s Lucinde and the Fragments, 14-18.
104
Behler and Struc, introduction to Dialogue on Poetry, 41; from a letter to A.W.
Schlegel, 18 December 1797, no further reference given.
105
Behler and Struc, introduction to Dialogue on Poetry, 35.
106
Chaouli, 47.
107
Bishop, 2-3.
that the essence of Schlegel’s (and Novalis’s) aphorisms “lies in their

relationship to the infinite and in the despair of not being able to grasp
it in a more conventional way”108; on the other hand, the fragments
embody Schlegel’s ideal of a Romantic poetry that is “always
becoming.” The Absolute can never be reached; the attempt is
ultimately futile, but infinitely necessary and infinitely rich. From
fragments we constantly create and recreate new connections, draw
new conclusions: these “multiply in an endless succession of
mirrors.”109
In Lucinde, to avoid recording life in “the same systematic
and progressive way we experienced,”110 Julius recreated experience,
but any recollection is a recreation. Memories are fragments, ruins,
constantly deteriorating, constantly being recreated, but they are the
way we construct and reconstruct our past—whether personal or of a
nation, a people, a civilization. In her study of Jena Romanticism and
memory, Laurie Ruth Johnson argues that the fragment as a formal
structure “functions as a metaphor for the ambiguous capacities of
memory—like memory, the fragment both preserves something of the
past and yet also bears witness to the decay of that past, to loss.”111
But memory, like the fragment, also allows endless recreation in the
imagination—whether intentional or unacknowledged—endless new
forms arise.
Nineteenth-Century France
Nineteenth-century French authors could turn to a double

heritage: their own fragment writers—Rabelais, Diderot, Chamfort—
and the German Romantics who had, in turn, been influenced by them.
The extent and timing of the influences of German theory and
philosophy on the French, and specific points of contact (Madame de
Staël, Victor Cousin, and others), are beyond the scope of this study; a
108
Behler and Struc, introduction to Dialogue on Poetry, 36.
109
110
111
On the role of memory in the works of Schlegel, Laurie Ruth Johnson, The Art of
Recollection in Jena Romanticism: Memory, History, Fiction, and Fragmentation in
Texts by Friedrich Schlegel and Novalis (Tübingen: Max Niemeyer Verlag, 2002), 3.
few examples of French writers who clearly knew German theory and
the works of some German Romantics must suffice. Victor Hugo, in
the preface to his drama Cromwell, provides a manifesto of French
Romanticism that relies heavily on his understanding of German
Romantic thought; Baudelaire knew Jean-Paul (“le bon Jean-Paul,
toujours si angélique quoique si moqueur”112), he and Gautier (and
many others) knew Hoffmann,113 later in the century Laforgue lived in
Germany and knew Heine’s works well, and never must Wagner’s
influence be forgotten—on literature as much as music.
Though the extent of the spread of Schlegel’s specific theories
in France is unclear,114 his ideas have become instrumental in critical
studies of works of the French Romantics, perhaps in part because the
forms, the strategies, that Schlegel describes do appear, though they
are not linked directly to his philosophy.115
Lautréamont’s two sets of Poésies are fragment collections,
not poems but prose units of varying length, presented in the Oeuvres
complètes with line breaks at the ends of even the shortest, but without
spaces between the individual units.116 The second set parodies
well-known authors, including authors of fragment collections, as well
as those short phrases that have become aphorisms or maxims simply
by virtue of being repeated. Thus Pascal’s defense of fragments, a
sentiment shared by so many fragment writers:
J’écrirai ici mes pensées sans ordre, et non pas peut-être dans une
confusion sans dessein: c’est le véritable ordre, et qui marquera
toujours mon objet par le désordre même. Je ferais trop d’honneur
112
Blackall, 67, from Charles Baudelaire, “La Double Vie par Asselineau,” Oeuvres
complètes II, ed. Claude Pichois (Paris: Éditions Gallimard, 1976), 88.
113
Bishop, 85-86.
114
Furst, 120, for discussion of this topic.
115
Lloyd Bishop relies on Schlegel’s theories of arabesque narrative to trace forms of
Romantic irony in works in French from Diderot, who influenced Schlegel, through
Musset, Gautier, Baudelaire, and Flaubert, to Beckett. He notes the digressions and
the juxtaposition of opposites, the coexistence of contraries that the arabesque allows.
Rae Beth Gordon turns to Schlegel’s arabesque structure and its derivation from
ornament in her discussion of imagery and structuring devices in works of Gérard de
Nerval. (Rae Beth Gordon, Ornament, Fantasy, and Desire in Nineteenth-Century
French Literature (Princeton: Princeton University Press, 1992).
116
Lautréamont, Poésies II, in Lautréamont and Germain Nouveau, Oeuvres
complètes, ed. Pierre-Olivier Walzer (Paris: Éditions Gallimard, 1970), 273-92.
à mon sujet, si je le traitais avec ordre, puisque je veux montrer

qu’il en est incapable.
[I will write my thoughts here without order, and not, perhaps, in

a confusion without design: this is the true order, and this will
always mark my object with disorder itself. I would do too much
honor to my subject if I were to treat it with order, since I wish to
show that it is incapable of it.] 117
when reversed by Lautréamont, negates the entire fragment tradition:
J’écrirai mes pensées avec ordre, par un dessein sans confusion. Si

elles sont justes, la première venue sera la conséquence des autres.
C’est le véritable ordre. Il marque mon objet par le désordre
calligraphique. Je ferais trop de déshonneur à mon sujet, si je ne le
traitais pas avec ordre. Je veux montrer qu’il en est capable.
[I will write my thoughts with order, through a design without

confusion. If they are accurate, the first seen will be the
consequence of the others. This is the true order. It marks my
object through calligraphic disorder. I would do too much
dishonor to my subject, if I were not to treat it with order. I wish
to show that it is capable of it.] 118
Dante’s warning, “Lasciate ogne speranza, voi ch’intrate’,” becomes

an invitation to a party, “Vous qui entrez, laissez tout désespoir.”119
Gérard de Nerval was, in many ways, the quintessential
Romantic artist. He died young, a suicide, after suffering several
episodes of schizophrenia, perhaps induced by ill-fated love. His
works are filled with Romantic images: ruins, picturesque landscapes,
craggy mountains, and storms; he left fragments, unfinished works as
well as sketches, and he cultivated the forms of fragmentation in his
published works. He was familiar with the German Romantics and
traveled in Germany. He translated Faust, knew Heine and translated
his poetry, and, when he began writing fantastic tales, was much
influenced by Hoffmann. In his novel Aurélia, he employs a
117
Pascal quoted in: Walzer, notes to Poésies II, in Lautréamont, Poésie II, 1153.
André Breton and Paul Éluard will use the same approach in Notes sur la Poésie.
Walzer, in notes to Lautréamont, Poésie II, 1153-54.
118
Lautréamont, 275.
119
Lautréamont, 275.
structuring device similar to Hoffmann’s technique of interleaving, a

term used to describe narrative that alternates segments of two
separate stories, each intruding on the other. The most famous
example of this technique is Hoffmann’s Kater Murr, a bizarre
alternation of segments of the biography of Kapellmeister Johannes
Kreisler with segments of his cat’s autobiography, written—by the
cat—using Kreisler’s pages as blotting paper. The two were bound
together by chance; thus in the printed book, segments of the two
stories are run into each other without a break.120
In Aurélia Nerval alternates the narrative of reality with
memories of schizophrenic episodes, often without transition from one
to the other. Aurélia, part fiction, part autobiography, is an account of
a descent into madness and return to health. Nerval addresses his own
illness as well as contemporary fascination with connections of dream
to imagination and creativity. Aurélia opens with a doubling, “Le
Rêve est une seconde vie,”121 a sentence that is both introduction and
summary, foreshadowing images of doubling throughout the text. As
Tzvetan Todorov notes, the protagonist is doubled, split into two
personae: the sane narrator and the madman,122 a split made literal, as
Gordon notes, by the appearance of the Doppelgänger.123 The
narrator’s comments, explanations, and summaries—made from the
side of reason, in direct, clear statements—alternate with his
descriptions of past dreams and hallucinations. These descriptions are
vivid, profuse, colorful, some describing the very images of the
decorative arabesque and its potential for endless creativitiy:
Pendant la nuit … je m’étais cru transporté dans une planète

obscure où se débattaient les premiers germes de la création. Du
sein de l’argile encore molle s’élevaient des palmiers
120
John Daverio, Nineteenth-Century Music and the German Romantic Ideology
(New York: Schirmer Books, 1993), 61. Just as Starkie separated the digressions from
Don Quixote (see note 98 above), so Hans von Müller separated the Kreisler sections
from those written by Murr and published them individually as Das Kreislerbuch and
Lebens-Ansichten des Katers Murr. Steven Paul Scher, 309.
121
Gérard de Nerval, Aurélia, Oeuvres complètes, ed. Albert Béguin and Jean Richer
(Paris: Éditions Gallimard, 1966), 2:361.
122
Tzvetan Todorov, The Fantastic: A Structural Approach to a Literary Genre, trans.
Richard Howard (Ithaca, NY: Cornell University Press, 1975), 37.
123
Gordon, 32.
gigantesques, des euphorbes vénéneux et des acanthes tortillées

autour des cactus; les figures arides des rochers s’élançaient
comme des squelettes de cette ébauche de création, et de hideux
reptiles serpentaient, s’élargissaient ou s’arrondissaient au milieu
de l’inextricable réseau d’une végétation sauvage.
[During the night … I believed myself transported to a dark planet

where the first seedlings of the creation were struggling. From the
breast of still soft clay were rising gigantic palms, poisonous
spurge and acanthus twisted around cactus; the dry shapes of the
rocks rushed forward like skeletons from this sketch of creation,
and hideous reptiles were winding, stretching or swelling in the
midst of the inextricable web of a wild vegetation.]124
Dream and reality converge, one interrupting and overtaking

the other: as the protagonist walks down the street, the shapes of
reality suddenly begin to transform into monstrous beings and his
ability to interpret reality accurately fails. In two strands of narrative
that are never completely separate, just as the two personae of the
protagonist are never completely separate, narrator and madman
continually qualify the reality of the dream episodes—the madman
questioning whether they were truly real, the narrator whether they
were truly false.125
Digression creates the structural element of the arabesque in
narrative, but in order to identify digression (as Gordon emphasizes),
it is necessary to identify a principal narrative. In profuse ornament,
ambiguity arises when frame overpowers subject (as the original
grotesques seem to overpower the center portrait), or digression
overpowers primary narrative.126 In Aurélia, the narrating voice
documents the journey, the epic quest from madness to sanity, but just
as the dreams and hallucinations overpower the protagonist, so that
organizing narrative is overtaken and overpowered by the length and
vividness of the dreams and hallucinations—a profusion of ornament.
Does the rational voice provide a defining frame for the rich
arabesques of dream, or do the arabesques provide an overpowering
frame for the center, for reality? The identification of a principal
124
Nerval, Aurélia, I:376-77. My translation with thanks to Olivia Jamin and Anja
Pearre.
125
Todorov, 38-39.
126
Gordon, 34, 40.
narrative is compromised, but the two lines are too interconnected to

be co-equals. Nerval causes us to question the role of digression in his
story just as he causes us to question the role of dream—is it higher
reality or madness?
Aurélia is open-ended; the story could continue in a
never-ending cycle of illness and health, series after series of
alternating dream and sanity. The final sentences of the novel do little
to erase the possibility that the illness may recur, nor to answer the
questions raised by the entire structure. Is dream a second life, a
second level of reality? Is madness an illness or simply a higher
reality, a higher truth? The end is no end, the conclusion no
conclusion; there is no promise that the protagonist will not descend
into his hell—or perhaps, his heaven—again and again, returning to
“normal” again and again.
The found-manuscript ploy, though absent from the narrative
of Aurélia, is mirrored curiously in the circumstances surrounding the
book’s publication. Aurélia was intended to appear in two parts in the
Revue de Paris. The first appeared on 1 January 1855; Nerval
committed suicide by hanging on the night of 25 January; the second
part of the work was published posthumously on 15 February. Though
this second section was most probably in galley proofs at the time of
Nerval’s death, Alfred Dubruck reports that numerous rumors
circulated claiming that the final pages of the manuscript had been
discovered stuffed in the pockets of the hanged man’s clothing.127
Still, the intrigue of the “found manuscript” fragment must be
balanced by the need to complete that fragment: Béguin and Richer
report that letters found to Nerval from Jenny Colon, the real-life
model for Aurélia, were substituted by the editor for a missing section
in the novel—even though the narrative seemed to indicate that the
letters should have been to Aurélia from the protagonist.128 Rather
inconsistency than the unknown.
To come to terms with reality: this appears to be Nerval’s
quest, and surely, at least in part, the quest of all these previous
examples, chosen from the earliest stages of Western European
127
Alfred Dubruck, Gérard de Nerval and the German Heritage (The Hague: Mouton
& Co., 1965), 20.
128
Albert Béguin and Jean Richer, in notes to Nerval, Aurélia, 1:1296.
vernacular literature to the heart of Romantic fragment production and

criticism. The fragment does not belong to one period or philosophy,
but can be reinterpreted for any. Aurélia may lack the humor, satire,
and playful nature of some digressive narratives, but Nerval knew his
lineage as well as Frye—or at least well enough to steal it from Nodier
in his Histoire du roi de Bohême et de ses sept châteaux, who had in
turn lifted his title from Sterne’s Tristram Shandy. In Angélique,
Nerval writes:
“Et puis … ” (C’est ainsi que Diderot commençait un conte, me

dira-t-on.)
—Allez toujours!
—Vous avez imité Diderot lui-même.
—Qui avait imité Sterne …
—Lequel avait imité Swift.
—Qui avait imité Rabelais …
—Lequel avait imité Lucien. Et Lucien en avait imité bien
d’autres. … 129
[“And then … ” (That is how Diderot began a tale, they will tell
me.)
—Go on anyway!
—You have imitated Diderot himself.
—Who had imitated Sterne …
—Who had imitated Swift.
—Who had imitated Rabelais …
—Who had imitated Lucian. And Lucian had imitated many
others. …]130
Opening in mid-conversation, in the middle of a thought; a list

made of sentence fragments that trail off into dots of ellipsis (the holes
in the manuscript, in memory, in logical thought); interrupted by
asides; mesmerizing with word repetition: this excerpt, by using many
129
Gérard de Nerval. Angélique, Oeuvres complètes, 1:239. Nodier’s text reads “Et
vous voulez que moi, plagiaire des plagiaires de Sterne — Qui fut plagiaire de Swift
— Qui fut plagiaire de Wilkins — Qui fut plagiaire de Cyrano — Qui fut plagiaire de
Reboul — Qui fut plagiaire de Guillaume des Autels — Qui fut plagiaire de Rabelais
— Qui fut plagiaire de Morus — Qui fut plagiaire d’Érasme — Qui fut plagiaire de
Lucien — ou de Lucius de Patras — ou d’Apulée — car on ne sait lequel des trois a
été volé par les deux autres, et je ne me suis jamais soucié de le savoir …” Histoire du
roi de Bohême et de ses sept châteaux (Paris: Éditions Plasma, 1979), 26-27.
130
Trans. Gordon, 73-74.
of the strategies of fragmentation that mark the narratives it

enumerates, becomes a tiny mise en abyme of the broad subject it both
summarizes and embodies.
CHAPTER 2
BEGINNINGS AND ENDINGS
Others stem from a lack of parts … by reason

of cutting off, decisio, such as those who are
begotten without a hand or a head, whom the
Greeks call steresios; others are praenumeria,
when only a head or a leg is born. … It is
believed that the Blemmyae in Lybia are born
as headless trunks, that they have both eyes
and mouth in their chests; and that there are
others, born without necks, having eyes in
their shoulders.
—————Isidore of Seville
“Que signifie donc un discours parfait, achevé? Un texte qui

commence et finit? [What is the meaning of a perfect, complete
discourse? A text that begins and ends?]”1 Much may be forgiven in a
work if only the beginning presents, with certainty, even a few
concrete expository details, and the ending ties its bits and pieces
together convincingly. Beginnings and endings are powerful and
privileged edges; they function as both formal divisions and physical
boundaries. As frames (to paraphrase Calvino),2 they isolate the
work—separating it from what came before and what comes after;
simultaneously, through the very act of separation, they acknowledge
those surroundings. As thresholds, points of entry and exit, they usher
the listener from the world outside the work to the world within and
back, from silence to sound, from sound to silence. Beginnings raise
possibilities and set expectations, endings fulfill or frustrate them;
their effect on the work is crucial.3
Pierre Garrigues writes: “Tout fragment tient donc du miracle:
suspendu dans un espace sans espace. Ses arêtes, courbes ou
tranchantes, l’attirent en lui-même et hors de lui. [Every fragment is
1
Pierre Garrigues, Poétiques du fragment (Paris: Klincksieck, 1995), 11.
2
Italo Calvino, Note to Under the Jaguar Sun, trans. William Weaver (New York:
Harcourt Brace & Company, 1988), 86.
3
On Romantic composers’ tactics of beginning and ending, see Edward T. Cone,
Musical Form and Musical Performance (New York: W. W. Norton and Company,
1968); Leonard B. Meyer, Style and Music (Philadelphia: University of Pennsylvania
Press, 1989); and David Ferris, Schumann’s Eichendorff Liederkreis and the Genre of
the Romantic Cycle (Oxford: Oxford University Press, 2000).
like a miracle: suspended in a space without space. Its edges, curved

or sharp, pull it into itself and outside itself.]”4 When beginnings and
endings are fragmented, those powerful edges acquire the
multi-directional pull of the fragment, reaching both into and outside
the work in search of what the fragment lacks. The act of separation is
no longer clean, boundaries are no longer clear; the distinction
between the work and its surroundings blurs. Fragmentation makes
suspect the expectations and conclusions that beginnings and endings
present; it compromises their formal functions, sometimes provoking a
search for an alternate—a real—beginning or ending within the work,
somehow camouflaged by these deformed first and last sections.
Convention defines beginnings and endings: it was common
for the epic to begin in medias res, but in large part, the convention for
narrative rests on a chronological sequence, resolution of plot, and the
adequate identification of time, place, and character. Frank Kermode
quotes Simone de Beauvoir’s claim (in her autiobiography) that
L’Invitée is “a real novel, with a beginning, a middle, and an end”;
Kermode follows with this sentence, “And real novels do have these.”5
The convention is strong; deforming it can create a powerful effect.
Recall Rabelais’s manuscript, beginning and ending with the physical
fragment; its mangled and missing words ask what literature can
mean, acknowledge what time can do to the written document, to
interpretation, to memory. Later writers for these reasons and many
others have manipulated their narratives by mutilating both beginnings
and endings. Mallarmé gave Henri Cazalis “une recette” that he claims
to have invented and to practice, “‘Il faut toujours couper le
commencement et la fin de ce qu’on écrit. Pas d’introduction, pas de
finale.’ Tu me crois fou? Je t’expliquerai un jour que là n’est pas ma
folie.”6 [“‘Always omit the beginning and the end of what you write.
No introduction, no finale.’ You think I am mad? I’ll show you some
day that my madness lies elsewhere.”7]
4
Garrigues, 41.
5
Frank Kermode, The Sense of an Ending: Studies in the Theory of Fiction (Oxford:
Oxford University Press, 1966), 139.
6
Mallarmé, Letter to Henri Cazalis, 25 April 1864, Correspondance: 1862-1871, ed.
Henri Mondor (Paris: Gallimard, 1959), 117.
7
Mallarmé: Selected Prose Poems, Essays, and Letters, trans. Bradford Cook
(Baltimore: The Johns Hopkins Press, 1956), 82.
Chapter 2: Beginnings and Endings 65
In tonal music, the archetype of a complete work, or

movement, begins and ends in an established key, on the tonic chord,
though sometimes an introduction—a section set off from the main
part of the piece—may show a degree of tonal ambiguity that is
resolved as the introduction ends and the main part of the piece
begins.8 Works that do not conform to this convention are incomplete:
fragments.
Nineteenth-century composers sometimes began their pieces
in the tonic key but with a chord other than that built on the tonic note;
if that initial chord is long drawn out, as in Chopin’s Prelude in A
Minor, op. 28, no. 2, so that the listener’s expectation is misdirected,
or if the key is obscured through progressions difficult to place in a
key, as in Brahms’s Intermezzo in B-flat Minor, op. 118, no. 1, the
resulting tonal ambiguity tends to suggest the fragmentary.
Composers, with less frequency and far more caution, manipulated
and compromised endings as well; in such cases, of course, closure is
thwarted and the sense of incompletion all the stronger. Rosen cites
Chopin’s Prelude in F Major, op. 28, and Liszt’s 1844 setting of “S’il
est un charmant gazon” as two examples of the rare practice of ending
on a dissonance that clouds the final tonic, though without actually
negating it; Schenker’s examples of pieces with off-tonic endings
include Bach’s Prelude no. 3, BWV 999, and Chopin’s Mazurka op.
30, no. 2; the first song in Schumann’s Dichterliebe, “Im
wunderschönen Monat Mai” (also analyzed by Rosen), neither begins
nor ends on its tonic.9
8
A tonic is at once a goal and a frame of reference. In tonal music—music in a major
or minor key—notes and chords exist in a hierarchy organized around a tonic note; a
piece (or movement) is said to be in the key of that tonic. Notes and chords progress
from one to another according to certain conventions. The strongest of these
conventions are that the key be established at the beginning of the piece (in part by
beginning on the tonic chord); that, if the key is departed from, it be reestablished, and
that the piece end on the tonic chord. Western art music was tonal from some time in
the seventeenth century beyond the end of the nineteenth century. During this period,
pieces deviate from the first of these conventions very rarely, from the last, almost
never. Opera, as an extended, text-driven form, does not conform to these
conventions, though smaller units within the opera usually do.
9
Rosen, 79-80 and 96-98; Heinrich Schenker, Free Composition (Der freie Satz),
New Musical Theories and Fantasies, No. 3, trans. and ed. Ernst Oster, 2 vols. (New
York: Longman, 1979), 130-31 (par. 307) and Fig. 152/6 and 7.
The music of these composers was part of Debussy’s musical

inheritance. Just as writers have often developed their techniques and
strategies of fragmentation in response to earlier models, so did
composers—and Debussy was no exception. In this chapter I will
examine Debussy’s development of strategies of fragmenting
beginnings and endings in three works, the songs “Green” and
“Spleen” and the Prelude “Canope.”
Beginnings: “Green” and “Spleen”
“Begin at the beginning,” says Lewis Carroll’s King of

Hearts, but narrative is filled with opening strategies that obscure the
temporal order of events, the identity of characters, and what is
normally the very purpose of a beginning. They may begin in medias
res, where the protagonists’ situations are not specified, their identities
obscured by the use of pronouns without antecedents—such
beginnings had become so common that Genette can speak of a “topos
of initial ignorance”—or sometimes even in medias sententias
(Nutall’s phrase), in the middle of a thought, opinion, or conversation,
or literally mid-sentence.10 Such openings create fragments: some part
of the true beginning has been severed; the information necessary to
reconstruct it, if provided at all, has been repositioned later in the
work.
Hemingway’s For Whom the Bell Tolls, for instance, begins
without any clue to the identity of the characters speaking or their
circumstances:
He lay flat on the brown, pine-needled floor of the forest, his chin
on his folded arms, and high overhead the wind blew in the tops
of the pine trees. The mountainside sloped gently where he lay;
but below it was steep and he could see the dark of the oiled road
winding through the pass. There was a stream alongside the road
10
On the in medias res opening, a convention in the epic before being adopted by the
novel, see Gérard Genette, Narrative Discourse: An Essay in Method, trans. Jane E.
Lewin (Ithaca, NY: Cornell University Press, 1980), 36. On the in medias sententias
variant, A. D. Nuttall, Openings: Narrative Beginnings from the Epic to the Novel
(Oxford: Clarendon Press, 1992), 156; on the topos of initial ignorance, Genette, 191.
and far down the pass he saw a mill beside the stream and the
falling water of the dam, white in the summer sunlight.
“Is that the mill?” he asked.
“Yes.”
“I do not remember it.”11
Debussy knew several examples of the in medias res opening

well, and praised Pierre Louÿs in 1896 after having read his novel
Aphrodite, which opens,
Couchée sur la poitrine, les coudes en avant, les jambes écartées
et la joue dans la main, elle piquait de petits trous symétriques
dans un oreiller de lin vert, avec une longue épingle d’or.
[Lying on her breast, her elbows forward, her legs apart, and her
cheek resting in her hand, she pricked little symmetrical holes in a
green linen pillow with a long gold pin.] 12
Maeterlinck’s Pelléas et Mélisande opens (in a scene Debussy

did not set in his opera) with unidentified servants opening the doors
to an unidentified building for reasons that are not revealed. Even the
scene that Debussy used to open his opera offers more questions than
it answers: Golaud explains himself and his situation (he is a prince,
out hunting) but Mélisande, in Symbolist fashion, responds to his
queries with vague non-answers or with answers to questions that
were never asked. Her identity and background are never established;
the viewer is left to imagine as much history as individual curiosity
requires from a few clues: a crown, a land far away, great sadness.13
Joseph Conrad created multiple beginnings for his novel The
Secret Agent. The protagonist leaves his shop in Chapter 1, his journey
immediately interrupted by a digression that supplies background
information; Chapter 2 begins at the same spatial and temporal point
11
Ernest Hemingway, For Whom the Bell Tolls (New York: Charles Scribner’s Sons,
1940), 1.
12
Pierre Louÿs, Aphrodite: Moeurs antiques (Paris: Gallimard, 1992), 43; my trans.
Originally published in installments in the Mercure de France, in 1885-86, then under
separate cover in April, 1886.
13
See Richard Langham Smith, “The Play and its Playwright,” in Roger Nichols and
Richard Langham Smith, Claude Debussy: Pelléas et Mélisande, Cambridge Opera
Handbooks (Cambridge: Cambridge University Press, 1989), 18-19.
as Chapter 1.14 Debussy knew this novel and commented positively on

it to Jacques Durand in a letter dated 8 July 1910:
Avez-vous suivi, dans le journal Le Temps, le roman de J. Conrad,
intitulé L’Agent secret ? Il y a là-dedans, une collection de
crapules tout à fait réjouissante, et la fin atteint au sublime. C’est
décrit de la manière la plus tranquille, la plus détachée et ce n’est
qu’après avoir réfléchi qu’on se dit: “Mais tous ces gens-là sont
des monstres.”
[Have you been following Conrad’s novel The Secret Agent in Le

Temps? It’s full of the most splendid scoundrels, and the end is
magnificent. It’s expressed in an absolutely calm and detached
style and it’s only when you think about the story afterwards that
you say: “But all those people are monsters.”15]
These openings are fragments, as torn and mutilated as the

fictitious manuscript of Gargantua’s lineage. They provide a point of
entry but they do not usher the reader into the work with comfort and
assurance; instead they initiate a need to look both outside the work
(to outside sources, experience, or knowledge) and further into the
work in search of the missing information, the missing part of the
manuscript. In Garrigue’s evocative words, “Ses arêtes, courbes ou
tranchantes, l’attirent en lui-même et hors de lui.”16
The musical analogue of the fragmented opening is the
off-tonic beginning: the true beginning, the initial tonic, has been
severed and the information necessary to reconstruct it must come
from later within the work.17 The identity of that missing tonic is
revealed in retrospect (sometimes only at the end of the piece), just as
the details missing from the beginning of Aphrodite or For Whom the
Bell Tolls are supplied, or implied, as the text progresses.
Debussy uses such openings often and in many different
circumstances; following the example of other composers, he
frequently chooses to begin pieces with such progressions that
14
On The Secret Agent, Genette, Narrative Discourse, 37.
15
Debussy, Correspondance 1884-1918, 268; trans. Nichols, Debussy Letters, 221.
16
Garrigues, Poétiques du fragment, 41.
17
Schenker calls progressions that begin off the tonic but end on it auxiliary cadences;
he explains them as incomplete transferences of the fundamental structure I-V-I,
lacking the initial tonic of the bass arpeggiation. Schenker, 88-89 (par. 244-45).
function as more than just introductions:18 complete sections of a work

may be formed by progressions that begin off the tonic and reach it
only later (Brahms’s Intermezzo, op. 118, no. 1, and Debussy’s song
“Green”; see below); or an entire work may consist of a progression
that begins off the tonic and moves to it (Chopin’s Prelude in A minor,
op. 28, no. 2, and Debussy’s Prelude “La Terrace des audiences du
clair de lune”).19 The two works discussed below are examples of
Debussy’s response to this still suspect heritage of off-tonic
beginnings and a response to texts that purposefully play on
fragmentation.
Verlaine’s poems “Green” and “Spleen”—the first two
“Aquarelles” in his collection Romances sans paroles—are
companion poems: addressed to a silent and unresponsive beloved,
they present progressive stages in a love affair. In “Green”20 the poet
experiences physical fatigue (and probably emotional fatigue) as he
tries to please the beloved: unable to grasp the whole of this new love
affair, he (like Julius in Lucinde) tells his story out of order, changes
tone abruptly, and can speak of himself, the beloved, and nature only
in fragments—yet the poem unfolds within a conventional strophic
form. In “Spleen” the poet suffers the fatigue of boredom and of the
perpetual dread of loss; here, fragmentation of the poetic form mirrors
the undercurrent of disintegration in the poet’s mind.
Debussy’s settings of “Green” and “Spleen” are the final
songs (of six) in his Ariettes oubliées, published in 1888. Both are
18
See Matthew G. Brown, Debussy’s “Ibéria” (Oxford: Oxford University Press,
2003), 83-89, in particular.
19
For Schenker’s analyses of these works, see Schenker, 88-89 (par 244-45) and Fig.
110/a/3 and 110/d/3; for the Debussy Prelude and other examples, see Brown, 132-33.
20
Green is often associated with new life, spring, and new love; Schubert’s “Die liebe
Farbe” and “Die böse Farbe” from Schöne Müllerin provide two different—even
contradictory—perspectives on green as an emblem of love. Otto W. Johnston notes
that green symbolizes hope, joy, and youth, as well as their opposites, moral
degradation and folly. See Otto W. Johnston, “Chromatic Symbolism in Gottfried
Keller’s Romeo und Julia auf dem Dorfe” in Themes and Structures: Studies in
German Literature from Goethe to the Present, ed. Alexander Stephan (Columbia,
SC: Camden House, 1997) quoting Pierre Paul Frédéric de Portal’s Des couleurs
symboliques dans l’antiquité, le moyen-âge et les temps modernes (Paris: Treuttel and
Würtz, 1837). References to the green-eyed monster, jealousy, are common, and in
nineteenth-century decadence, green is a color of decay.
incomplete at their beginnings, each opening with a progression that

begins off the tonic, and they are disrupted within by harmonic
digressions or discontinuities and, in “Spleen,” by quotation.
Verlaine’s poem “Green”:

Voici des fruits, des fleurs, des feuilles et des branches
Et puis voici mon coeur qui ne bat que pour vous.
Ne le déchirez pas avec vos deux mains blanches
Et qu’à vos yeux si beaux l’humble présent soit doux.
J’arrive tout couvert encore de rosée

Que le vent du matin vient glacer à mon front.
Souffrez que ma fatigue à vos pieds reposée
Rêve des chers instants qui la délasseront.
Sur votre jeune sein laissez rouler ma tête

Toute sonore encor de vos derniers baisers;
Laissez-la s’apaiser de la bonne tempête,
Et que je dorme un peu puisque vous reposez.
[Here are fruits, flowers, leaves, and branches

And then here is my heart that beats only for you.
Don’t tear it to pieces with your two white hands
And to your eyes, so beautiful, may this humble present be sweet.
I arrive all covered still with dew,

That the morning wind ices over on my brow.
Let my fatigue, resting at your feet,
Dream of the cherished moments that will refresh it.
On your young breast let my head roll

Still ringing from your last kisses;
Let it be calm after the good tempest,
And I may sleep a little since you are resting.] 21
Given Verlaine’s reputation for formal innovation, “Green” is

remarkably conventional: three strophes of four alexandrines each, the
21
Verlaine, Oeuvres poétiques complètes, ed. Jacques Borel (Paris: Gallimard, 1962),
205; trans. Rita Benton, ed., Claude Debussy Songs 1880-1904 (New York: Dover
Publications, 1981), 167, minor changes. Influences on “Green” include Ophelia’s
song from Hamlet and “Roses de Saadi” by Marceline Desbordes-Valmore. Antoine
Adam, The Art of Paul Verlaine (New York: New York University Press, 1963), 96.
lines organized in couplets that correspond to complete clauses or

sentences and divide neatly into hemistichs, abab rhyme scheme, and
alternating paroxytonic and oxytonic line endings. Debussy’s setting
reflects this regularity (Fig. 2.3): the ABA1 musical form22 reflects the
strophic division of the poem, with the sections clearly delineated by
cadences23 and appropriate changes of key (G-flat major, D-flat major,
G-flat major, clearly indicated by the printed key signatures);24 phrase
structure preserves the integrity of couplet, line, and (usually)
hemistich; the text setting is syllabic, so that the text is clearly heard.
Yet within its formal regularity, Verlaine’s “Green” is made all of
fragments, its images and its internal structure pulling at its
conventional outer form; and Debussy’s setting reflects this
fragmentation.
The tiny narrative—the story of a poet/lover who returns,
tired, from a journey, bringing gifts to the beloved—begins with the
presentation of those gifts; the account of the journey and his resulting
fatigue appear only in the second strophe. Debussy’s setting also
begins in the middle—the middle of a very common chord
progression, I-ii-V-I.25 As Fig. 2.3 indicates, Debussy’s incomplete
progression ii-V-I lacks its initial tonic; the key of Gb can then be
confirmed only in retrospect, after the V resolves to I. Where Verlaine
entered into a situation already under way, Debussy begins with a
progression as it were already under way. For the following
discussion, the reader may refer to the poem above, and below to Fig.
2.1, which contains mm. (i. e. measures) 1-20 of the song, to Fig. 2.
which contains mm. 48-58 of the song, and to Fig. 2.3, a formal chart.
22
In musical analysis, sections in a work are often diagrammed using letters of the
alphabet: letters are repeated where music is repeated; if the music is similar but has
some significant differences, a superscript is added.
23
In music, the term refers to an arrival or a point of relative repose; a cadence
produces a degree of closure.
24
G-flat major, six flats; D-flat major, five.
25
In the analysis of tonal music, roman numerals indicate the scale degrees on which
the chords are built; upper and lower case are used to distinguish chord qualities—
“major”and “minor” respectively. The tonic chord (“I”) corresponds to the key of the
piece; the dominant chord (“V”) is the chord that most often leads to that tonic (and
often presages an impending resolution to it); “ii” belongs to a group of chords that
characteristically prepare the way for “V.”
Fig. 2.1, mm. 1-20 of “Green.”

Fig. 2.2. Mm. 48-58 of “Green.”

Fig. 2.3. Chart showing formal structures of the poetry and music of
“Green” (cont. on facing page).
Stanzas: I II
topic gifts plea gifts journey
Line #s 1 2 3 4 5 6
Musical form A transition B
Harmonic
model: Gb:ii ii----V---I interruption ii------V---I I V
ii-V-I (V/V)
Measures 1----4 5---------8 13--------16 17---------20 20-----23 24-----27
9--------12 28-----31
As the poem opens, the protagonist is eager but uncertain: he

has brought gifts for his beloved, but the urgency of desire has made
him unable to discriminate what is valuable, what is appropriate, from
what is not. He presents a string of fragments, bits and pieces,
anything he can find, everything he has: fruits and flowers, but even
leaves and branches, and a heart—his heart—offered almost in the
same breath with a few sticks and twigs.26 All are of equal value,
having no value except that of the beloved’s approval.27
Debussy’s opening is equally eager—the score indicates
joyeux et tendre—but in spite of the buoyancy of the rhythm and the
insistent harmonic and melodic repetition of the opening A-flat minor
chord, there are also hints of uncertainty: minor harmony is not
26
Borel describes “a banishment of the self” in “a world where man is no more than
an object, scarcely different from the houses and the roofs, the leaves and the flowers
that surround him.” Jacques Borel, Notes, in Verlaine, Oeuvres poétiques complètes,
184.
27
The hard c of the coeur (heart), however, distinguishes it from the labial phonetic
cognates v, f, b, p (votre, fruits, fleurs, feuilles, branches, puis). On such matters see
Kenneth Burke, “Musicality in Verse” in The Philosophy of Literary Form: Studies in
Symbolic Action, 3rd ed. (Berkeley: University of California Press, 1973), 369-78.
Burke sees one element of musicality in verse resulting from an increased use of a
type of alliteration created by phonetic cognates, families of similar consonant sounds
that create a “concealed alliteration.” The consonants n, d, t, d, and th (voiced and
unvoiced) form one family; m, b, p, v, and f form another.
Fig. 2.3 (cont.). Chart showing formal structures of the poetry and
music of “Green.”
III
plea/rest rest rest plea rest
7 8 9 10 11 12
retransition A1
V ii ii--------V----I interruption ii--V----I
32-------35 36---38--39 40----41 42---45 46--49 50--------53 54------58
usually the first choice for expressing joy; the texture stutters,
repeating itself every two measures, each repetition giving the
impression that it leaves a rising inner line (Eb-Fb-F in the tenor)
suspended, resolving back onto itself only to begin its ascent again;
the vocal line enters in m. 5 over the same material that began as an
introduction but now continues as accompaniment; and of course
A-flat minor is not the key indicated by the key signature.28 Listeners
would have no way of knowing this without a score, of course, and
would of necessity assume the key to be A-flat minor. When things do
change they change with astonishing abruptness: the eight measures of
A-flat-minor harmony are followed by a measure of D-flat major and
another of G-flat major—in other words, an authentic cadence (the
most common and decisive sort) finally confirming the key of G-flat
major; it is immediately repeated (mm. 9-12). Only at this point is the
listener able to interpret the long drawn out A-flat minor chord of the
opening as ii in G-flat major, ii being one of the chords normally used
to prepare for an authentic (V-I) cadence; and even here, the tonic
chord is compromised by the vocal line as it continues to outline an
A-flat-minor triad, by the comparative brevity of the V-I progression,
and especially by the presence of an unresolved Fb in an inner voice, a
note whose presence destroys the finality the chord would otherwise
have possessed.29
28
A-flat minor would require seven flats.
29
In precisely the same way as Chopin had compromised the F major triad at the end
of his Prelude No. 23 by adding an Eb to it.
In the third line, the poet’s tone changes: he interrupts the

presentation of the gifts, and the forward progression of the narrative,
with a desperate plea in the imperative voice and with a violent
image—one that turns the traditional metaphorical gift of the heart to
gory reality: “Ne le déchirez pas avec vos deux mains blanches.”30
The heart, an organ torn from the body, now torn into even smaller
fragments, not just with the hands as a unit, but with “deux mains
blanches,” fragments whose detachment from the body is somehow
emphasized through their being needlessly numbered.31 The title’s
green “watercolor wash”—David Scott’s evocative term32—is
momentarily obscured by white and stained blood red. This change of
tone tears not only at the poet’s heart, but also at the structure of the
strophe, fragmenting the quatrain into couplets, as Verlaine
exaggerates the natural tendency of the alternating rhyme scheme
(abab) to separate at the midpoint into two units.
Debussy parallels the poetic interruption of line 3 with a
thematic and harmonic interruption: the progression ii-V-I (A-flat
minor–D-flat major–G-flat major) clarified—with the reservation
noted above—in measures 9-12 is followed by a chord spelled as an A
dominant seventh, related only distantly to G-flat major and a tritone
(melodically and harmonically the least stable interval) away from the
root of the E-flat seventh chord that follows it. The initial rising
stepwise diatonic gesture (Ab–Bb–Cb) that began the first phrase of the
vocal line is now deformed, initially rising by whole tones to
encompass another tritone, G–C#. (See mm. 13-16). These measures, a
musical interpolation, interrupt the unfolding of the vocal line and any
predictable harmonic motion; indeed the E-flat seventh chord points
back to A-flat minor, just after the possibility of that key’s serving as
30
Arthur Wenk, Claude Debussy and the Poets (Berkeley: University of California
Press, 1976), 51, points out this reversal of metaphor and reality, noting the
metaphorical gift of the heart in the literature of courtly love. There is, however, in
this literature of courtly love, the tradition of the eaten heart—the heart of the lover
fed to the unfaithful lady by the jealous husband. Gregory Stone, The Death of the
Troubadour (Philadelphia: University of Pennsylvania Press, 1994), 101-8.
31
As des separated the members of the gift list, dental cognates d and n now
emphasize the words ne, déchirez, and deux (still surrounded by the labial p, v, m, b),
words that perform a harsher separation.
32
David Scott, Pictorialist Poetics: Poetry and the Visual Arts in Nineteenth-Century
France (Cambridge: Cambridge University Press, 1988), 105.
tonic has been denied by the D-flat seventh–G-flat major progression

of mm. 9-10, 11-12.33 As the lover interrupted the forward progression
of the narrative, Debussy has interrupted the forward progression of
the harmony with this tiny A-major seventh–E-flat major seventh
interpolation that reflects the brevity of the lover’s expression of
doubt, but also the depth of his concern.34
The fourth line of the poem pulls the strophe back together:
the alternating rhyme is a scheme of change, but also of recurrence.35
The repetition of the et that began line 2 marks a return to the
expression of hope, though less unbridled, presented in the first and
second lines;36 it also offers the first hint that the beloved is beautiful
(she has beautiful eyes) and that she might actually find the poet’s
strange conglomeration of gifts pleasing. Debussy sets this line with a
clear, uncompromised V/V (substituting for ii)-V-I (mm. 17-20).37
The “introductory material” now appears clearly in the tonic, as it
should have appeared at the beginning (“should have,” that is, in terms
of normal tonal practice); the section ends as it should have begun, by
establishing G-flat major without ambiguity. The functions of the
initial A-flat minor chords, the compromised tonic in measures 10 and
12, and the tonal interruption in measures 13-16 now become clear.
Verlaine continues his strategy of fragmentation in the second
and third strophes: the poet speaks of himself and the beloved
primarily in terms of body parts—brow, head, feet, breast; strophes 2
and 3 are both interrupted at the third line with the poet’s plea, but the
change in voice and subject matter is less dramatic. Debussy’s setting
of the second strophe (B section), in the key of D-flat major,38 differs
33
Brown, Debussy’s “Ibéria,” 82-83, has identified two types of musical insertion:
parenthetical episodes provide thematic diversion without destroying the unfolding
tonality; interpolations create discontinuities in the tonal motion.
34
Wenk’s analysis, though correct, obscures the disjunction of this interpolation, as
well as the one in mm. 50-52. Claude Debussy and the Poets, 52-54.
35
Clive Scott. French Verse-art: A Study (Cambridge: Cambridge University Press,
1980), 134.
36
Marked also by a return to the labial b, v, p—until a return to the dental d
emphasizes the final word doux.
37
V/V and ii have the same root; both normally function by preparing the way for a
subsequent V.
38
Precisely the key one would expect for the central section of an ABA form in G-flat
major.
from the outer sections: increased rhythmic motion in the setting of

the first two poetic lines reflects the journey; slower tempo and
decreased rhythmic activity reflect the change from motion to rest in
lines 3 and 4; the tritone root relationship reappears—this time G to Db
—at the opening of line 4, not line 3 as before, though it does not
function as a reflection of poetic disjunction as it did in the first
strophe.39
Except for its slower tempo and the omission of two measures
of piano introduction, Debussy’s reprise (for the third strophe) begins
as a literal repetition of the opening. The change of tempo—unusual in
the third section of an ABA form—is demanded by the dynamic of the
poem: Verlaine’s “Green”—for all that it does consist of three
strophes—clearly does not present statement-contrast-restatement, but
rather a progression toward repose and the alleviation of anxiety.
Debussy brings the musical motion to repose in several ways. The first
statement of the tonic in the reprise is compromised to a lesser degree
than in the opening: at m. 49 (corresponding to m. 12), though the Fb
still clouds the Gb chord, the melodic contour now conforms to the key
of G-flat major. Also, Debussy replaces the interruption from the A
section with one of a different character, compressed in length, and
used for a different purpose. As can be seen in Fig. 2.2, mm. 50-52
parallel the first-section insert in mm. 14-16, but whereas in the first
instance the roots of the two interpolated chords were distantly related
by the tritone, here they are less distantly related, by the semitone.40
They also connect more smoothly with what preceded (G-flat major
seventh–C-flat) and what will follow (B-flat seventh–E-flat minor–A-
flat minor–D-flat major–G-flat major). These differences serve to
decrease—rather than heighten—tension, and to reflect the movement
of the poem toward that state of repose, of motionlessness, that the
poet desires.
39
The G-major chord separates the dominant A-flat seventh from its tonic resolution,
D-flat major.
40
They are related as V7 (B-flat seventh) and VI (C-flat major), the chord to which V
most commonly resolves in the so-called “deceptive” cadence, where the V moves to
a chord other than the expected I. Debussy, however, minimizes this relationship by
making the C-flat chord move to the B-flat, reversing the normal procedure and
emphasizing this reversal through phrasing and dynamics.
How, in sum, does Debussy capture in song the doubts and

ambiguities of Verlaine’s poet/lover protagonist and the internal
tensions of the poem? First, he begins off the tonic on an A-flat-minor
chord, misleading listeners about what the key of the song will be and
disconcerting them in the process. Second, he destabilizes what ought
to be the most stable elements in the song, the opening and closing
“A” sections, by (in addition to beginning them off the tonic) clouding
the first cadences in the true key of G-flat major (mm. 9-12, 46-49),
then introducing the extraneous chord progressions of mm. 13-16 and
50-51 (which, in the first section, actually look back to the “false” A-
flat minor of the opening). Third, he extends the initial A-flat-minor
harmonies of both “A” sections disproportionately; lack of proper (or
expected) proportion is one of the factors through which fragments
create the tension and cause us to question reality. All these
ambiguities severely compromise the sense of form and the sense of
key (intimately bound together in tonal music), in effect calling those
conventions into question.
In “Spleen” the poet’s state of mind has changed dramatically.
Les roses étaient toutes rouges,
Et les lierres étaient tout noirs.
Chère, pour peu que tu te bouges

Renaissent tous mes désespoirs.
Le ciel était trop bleu, trop tendre,

La mer trop verte et l’air trop doux;
Je crains toujours,—ce qu’est d’attendre!—

Quelque fuite atroce de vous!
Du hous à la feuille vernie,

Et du luisant buis je suis las,
Et de la campagne infinie,
Et de tout, fors de vous, hélas!
[The roses were all red,

And the ivies were all black.
Dear, every little move you make

reawakens all my despair.
The sky was too blue, too tender,

The sea too green and the air too mild.
I fear always,—it is expected—!

Some agonizing flight of yours!
Of the holly with the varnished leaf,

And of the shiny boxwood I am weary,
And of the endless countryside,

And of everything except of you, alas!] 41
“Green” has become too green, the gifts of nature now seem
wearisome, even artificial (“varnished”); joy has turned to boredom,
anticipation to dread, the motionlessness of repose to agitation; and in
Debussy’s setting the Gb harmony that ended “Green” is reinterpreted
as bII of F minor42—another off-tonic beginning and a symbol of the
transformation of the poet’s state of mind.
In his study of French Symbolism, Houston notes Verlaine’s
reputation for the chanson grise, for “faint colors and insubstantial
outlines”; as he puts it, “All the imagery of bercement, fadeur,
langueur, pâleur, and monotonie in Verlaine is an attempt to conjure
away violent emotions with their unwanted consequences.”43 In
“Spleen” Verlaine acknowledges and expresses those violent
emotions; this poem is not a “conjuring away” of emotion, but rather a
longing for stasis that the poet has not yet achieved. If “Green” is a
poem of new love, of innocence, “Spleen” is a poem of the
degradation of that love, and a poem of decadence. The decadent
persona is jaded yet overly sensitive; he prefers the artificial to the
natural and the world of the inner mind to outer reality; he is ill in
mind and body, and proud of it; he seeks to vivify the inanimate and to
petrify the animate. In decadent writing, as Weir notes, description
serves to present attributes of character rather than to set a realistic
41
Verlaine, Oeuvres poétiques complètes, 205; trans. Benton, Debussy: Songs, 167,
minor changes.
42
The four flats in the key signature signal F minor.
43
John Porter Houston, French Symbolism and the Modernist Movement: A Study of
Poetic Structures (Baton Rouge, LA: Louisiana State University Press, 1980), 182.
background;44 the artificiality of the description in “Spleen”—”all

red,” “too blue,” “too tender,” “too green,” “varnished”—presents the
poet’s reaction to nature, or to an unnatural world of his own
creation.45 In either case, existence has become unbearable; he longs
for relief from boredom but also from the dread that the beloved will
leave—a release that can be bought only at the expense of the feared
outcome. Reversing the Pygmalion and Galatea story, the beloved
must become motionless, inanimate, a statue; the longed-for repose of
“Green” is now an obsession with control.
Where the poetic form of “Green” was conventional, that of
“Spleen” is not. In “Green” the fear that the poet’s heart will be torn
apart has now become the reality of the poem torn apart: “Spleen” is
presented on the page as six separate couplets, but they are not typical
rhyming couplets (aa bb cc dd, etc.); they rhyme ab ab cd cd ef ef. As
Wenk has pointed out, they are the halves of three four-line quatrains,
rhyming abab, that have been separated, split apart, at the midpoint. In
content, the fragmentation is more severe. Again, as Wenk points out,
couplets 1 and 3 are both descriptions in the imperfect tense; couplets
2 and 4 are present tense addresses to the beloved describing the
anxiety of the poet; couplets 5 and 6 form a unit that combines both
the descriptive elements and the anxiety46; however, Verlaine’s
destruction of the quatrain can be taken one step further: by content
and by verb tense (though not by rhyme) couplets 1 and 3 form a unit,
as do couplets 2 and 4.
“Spleen” is a variant of Schlegel’s digressive arabesque, a tiny
example of the device of interleaving familiar from Kater Murr and
Aurélia (See Chapter 1); in “Spleen” the conflicting emotions
interleaved are boredom on the one hand, fear and anxiety on the
other. Like Kater Murr and Aurélia, “Spleen” is a structural
representation, not just a telling, of the poet’s emotional state.
Debussy’s setting of “Spleen” is more fragmented than that of
“Green.” The auxiliary cadence progression that controls “Spleen”—
b
II-V-I—does not resolve to a clear, functional tonic until the final
44
David Weir, Decadence and the Making of Modernism (Amherst: University of
Massachusetts Press, 1995), 106.
45
Note Des Esseintes’s artificial environment in Huysmans’s À rebours.
46
Wenk, Claude Debussy and the Poets, 116-20.
cadence; thus the tonic is recognized, as in “Green,” only in

retrospect, but in “Spleen” only as the song ends.47 In “Spleen,” the
incomplete progression—bII-V—is not just prolonged, but repeated
distinctly four times (strophes 1, 2, 4, 6); as Fig. 2.4 (below) shows,
each repetition coincides with a statement of the main theme at the
pitch of its initial occurrence. This theme, perhaps an ironic borrowing
from the wedding song in Chabrier’s opera Gwendoline (Fig. 2.5),
always occurs in the accompaniment, but is never aligned with the text
of the respective couplets in the same manner; moreover, it is
fragmented in strophe 3 and transposed in strophe 5.48
Fig. 2.4. Chart showing formal structures of the poetry and music of
“Spleen.” (Continued on facing page)
Couplets & rhyme scheme 1 (ab) 2 (ab) 3 (cd)

Alternating emotions fatigue fear fatigue
Motivic statements a a1 extension
Musical form: 2 views
----1. through-composed
----2. two parts A
Harmonic model f: bII---V bII---V bII ---------------
Harmony f: bII---V bII---V --extension bII---
B: II--V-----bVII
Measures 1----------8 9-----13 14------------17
47
Noted by Wenk, Claude Debussy and the Poets, 119, and Roy Howat, Debussy in
Proportion (Cambridge: Cambridge University Press, 1983), 34-36. As mentioned
above, Chopin’s Prelude in A Minor is an example of such a technique: the dominant
is prolonged throughout that work, resolving to the tonic only as the piece closes.
Schenker gives this as an example; see 88 (par. 245) and Fig. 110/a3.
48
The quotation of the “Bénissez-nous” from the “Épithalame” in Chabrier’s opera
raises questions of extra-musical meaning. Perhaps Debussy saw a certain irony (or
sarcasm) in underlying the dread and boredom of “Spleen”—the relationship gone
sour—with wedding music, as the blessings Gwendolyn requests have, for the poet of
“Spleen,” become curses. Léon Vallas, Claude Debussy: His Life and Works, trans.
Maire O’Brien and Grace O’Brien (London: Oxford University Press, 1933; reprint,
New York: Dover Publications, Inc., 1973), 66.
Through these tactics—an almost insistent repetition of

harmonic and melodic material in an irregular and unpredictable
manner, the constant beginning anew without real resolution,49 and a
sustained harmonic tension created by the avoidance of the tonic—
Debussy maintains both the anxiety and boredom expressed by the
poet, though using a very different strategy from Verlaine’s
alternating emotional states. Debussy increases the harmonic tension
in “Spleen” by substituting bII for the ii of “Green”—increases the
tension in that bII is a non-diatonic chord and in that its root lies
distant from the root of the chord to which it moves (V) by a tritone,
the least stable and one of the most highly charged intervals in the
tonal system.
Fig. 2.4 (cont.) Chart showing formal structures of the poetry and
music of “Spleen.”
4 (cd) 5 (ef) 6 (ef)

fear fatigue fatigue/fear
a2 a3 (transposed) extension---a4
A1
bII----V---- -------------------------- ---------------bII--------V----I
bII----V---- IV# --------------V6/4 I+-------------bII--------V----I
18-----21--- 22-------------------25 26------------28-------34
49
Recalling the opening strategy of Conrad’s The Secret Agent, mentioned above.
Fig. 2.5. Comparison of measures 41-43 of the “Épithalame” from

Gwendoline with measures 1-4 of “Spleen.”
“Épithalame”
“Spleen”
Debussy’s use of motivic repetition and tonic avoidance in

“Spleen” does not take place within a conventional formal structure
that is as clearly defined as the ABA1 form of “Green.” Analyses that
rely too heavily on motivic repetition seem unsatisfactory at the level
of overall form and stretch the meaning of standard terminology.
Howat, relying on Golden Section proportion for his own formal
explanation of “Spleen,” acknowledges the motivic and harmonic
repetition, but admits that “to describe the form, truthfully enough, as
a tonally free rondo does nothing to explain what is crucial to its
expression”;50 nor, I would say, does Parks’s description of the form
of other works of Debussy that, like “Spleen,” consistently repeat and
extend motives as “a synthesis of strophic variations with …
continuous development interrupted by reprise.”51 Even the
appearance of the main motive at m. 18, the midpoint of the song (and
of the poem), unharmonized as at the opening, does little to suggest a
two-part form (see Fig. 2.4). Since there is no strong resolution to any
key area before the final tonic (though there is a weak one in mm.
25-6, see Fig. 2.4), and since the principal motivic repetition is
irregular, it is perhaps best simply to read the song as through
composed.
50
Howat, Debussy in Proportion, 34.
51
Richard S. Parks, The Music of Claude Debussy (New Haven: Yale University
Press, 1989), 224-25.
“Green” and “Spleen” are fragments; they exist within a tonal

harmonic framework, but for them, that frame has been broken. It has
not dissolved: recognition of the model whole, the convention, from
which they are derrived is crucial to understanding their structures.
Each of the songs is modeled on an incomplete chord progression—
”Green” on ii-V-I and “Spleen” on bII-V-I—just as other works are
modeled on complete chord progressions that begin and end on a
tonic. They are fragments, incomplete at their beginnings though
closed at their endings, and they are further fragmented by the internal
repetition of those progressions, or portions of them—ii-V-I for
“Green” and bII-V for “Spleen”—that govern their overall form. Those
incomplete progressions and their repetitions allow Debussy to
fragment the surfaces of the works and to build long-range tension,
but these fragments do not yet exist in the modern world of the collage
or of Warhol’s postmodern multiple, multi-colored Marilyns—a world
where the meaning of the part no longer depends on its relation to a
larger, overriding whole. It is in the nature of the Debussy’s
fragments—like those in the literary works that inspired them—not to
be ambiguous, but to appear ambiguous until enough information is
revealed, or discovered, to make an identification. In both “Green”
and “Spleen” the off-tonic beginnings create a certain level of
ambiguity, but in both cases strong final cadences—in section or
work—reveal the information necessary to place the preceding
material in its proper context. Debussy creates a sense of uncertainty
and expectancy by tearing at convention while ultimately maintaining,
even depending on, that convention.
Endings: “Canope”
In a photograph taken in 1910 in Debussy’s home in the Bois

de Boulogne, Debussy and Satie lean against a mantel; from the
mantel, just above Satie’s hand, a small Egyptian face stares out into a
world completely alien to its origins. It is one of a pair of lids to
canopic urns, funerary urns intended to hold certain internal organs in
the Egyptian mummification process, artifacts reduced to the status of
decorative objects. 52 These urn lids appear to be the source of the title
of Debussy’s tenth prelude from Book II, “Canope,” just as other
family possessions provided titles for Debussy’s works: the “Poissons
d’or” from Images, Book II, swim on a Japanese-lacquer wooden
panel, the “Puerta del vino” from the second book of Preludes appears
on a postcard of the Alhambra, while Jimbo and the Golliwog are only
two of the Children’s Corner and the Boîte à joujoux characters that
populated his daughter’s toy box.
The canopic urn lids are fragments in two senses: they are
remnants of the original urns—lost, perhaps broken—and they are
archeological artifacts, reminders of a distant and exotic culture. In his
classification of literary fragments, Roger Shattuck describes his
category labeled the linked fragment in terms of the archeological
artifact: from the pottery shard, the archaeologist reconstructs the
original vessel; from the realization of that vessel, with other
reconstructed artifacts, he recreates an entire civilization—its history,
beliefs, daily life, ultimate fate. This is the transcendental fragment of
Romantic and Symbolist literature, of Swedenborg, Novalis, Nerval,
Baudelaire, Mallarmé, and even Yeats and Stevens. The identity of the
literary fragment relies on the very convention it has broken, reaches
back to that convention, just as the identity of the artifact relies on its
origins in the past. From this perspective, as Shattuck claims, “nothing
stands alone, and the tiniest fragment of the universe breathes forth its
secret connections to everything else.”53
It is from this perspective of the archeological fragment that
many writers on Debussy’s prelude “Canope” view the referent of the
title—both as inspiration for Debussy and as interpretative guide for
the listener. They weave around the urns a web of connections, or
associations, centered on man’s ongoing fascination with ruins and the
past, regardless of the approach they will ultimately take in their
analyses. Robert Schmitz writes that Debussy “reaches far back to
archaic ceremonials, to the magnificent [city of] Canope, of which
52
This photograph is reproduced in François Lesure, Debussy: Iconographie musicale
(Paris: Éditions Minkoff & Lattès, 1980), 134. Lacking any evidence, I will ignore the
very real possibility that they are reproductions.
53
Roger Shattuck, “The Alphabet and the Junkyard,” in Fragments: Incompletion and
Discontinuity, ed. Kritzman, 36.
there remain only the ruins of temples and palaces.”54 James Baker
finds “little wonder, then, that the piece has an air of mystery, its
subdued harmonies redolent of ancient ruins.”55 David Lewin sums up
this attitude: “Debussy, to the extent he had the urn in mind, used the
word canope as a metonym for the city, the temple, the associated
rites, the dead civilization itself”; reflecting the significance for
personal inspiration that the artifact and the ruin have had at least
since Petrarch gazed over the ruins of Rome, he adds, “I am similarly
beset by fantasies about the person buried in the beautiful Egyptian
urn, the potter who cast that compelling art-work.”56
“Canope” is a tiny piece, only 33 measures long. (“Canope” is
given in its entirety in Fig. 2.6; motives labeled U-Z are indicated on
the score in large, bold type and circled.) A key signature of one flat is
in effect throughout; the chordal opening clearly emphasizes D minor,
as do all motives except X (m. 14) and Z (m. 17). The D chord at m. 7,
now with F#, tonicizes G,57 and this initiates a string of descending
fifths that lead, ultimately, to the dominant harmony at m. 18 and its
whole-tone extension. At the end of m. 25 a fermata allows the sound
to die away before a reprise of the opening motive in m. 26.
Subsequently motive W returns in mm. 30-33, harmonized differently,
but the final statement is incomplete: the closing melodic notes, Eb-D,
are not written, and there is no sense of melodic closure even though
D is already sounding (in the wrong voice) in the final chord. The
sense of incompletion, of the motive left unresolved, depends on aural
memory, the expectation created by the three previous statements of
the W motive, but also on the E-Eb-D descent in the first statements of
motive V (mm.7-8 and 20-21). The D tonal center is weakened by the
following: in m. 3, the descending fifth, G-C; in mm. 14-15, the
prominent C-major chords; and in the final measures, the C chord in
root position, even given the added ninth. Chromatic alterations in
mm. 4, 13, 23, and 28 also increase tonal uncertainty.
54
E. Robert Schmitz, The Piano Works of Claude Debussy (New York: Dover
Publications, 1966; originally published Duell, Sloan & Pearce, Inc., 1950), 182.
55
James M. Baker, “Post-Tonal Voice-Leading,” in Early Twentieth-Century Music,
ed. Jonathan Dunsby (London: Blackwell, 1993), 30.
56
David Lewin, “Parallel Voice-Leading in Debussy,” in Music at the Turn of the
Century, ed. Joseph Kerman (Berkeley: University of California Press, 1990), 64-65.
57
By functioning as V in reference to a G tonic.
Fig. 2.6. “Canope.”

Fig. 2.6. “Canope” (cont.).

In numerous attempts to explain the organization and structure

of this work, analysts have employed topics ranging from the
harmonics of the Javanese gamelan through the rules of medieval
musica ficta to modern set theory.58 Tonal analyses of one sort or
another all acknowledge a conflict between D minor and C major (or
between D dorian and C mixolydian), but are divided in their opinions
on the ultimate tonality of the piece. Arthur Wenk, Robert Schmitz,
Serge Gut,59 and Amy Dommel-Diény 60 conclude that the tonality is D
minor; David Lewin and James Baker settle on C major, Thomas
Warburton claims bitonality.
Given the final C major chord and the incomplete statement of
motive W in the final measures, those who claim D minor (or dorian)
tonality obviously must stress the inconclusiveness of the ending, the
acceptance of an open-ended fragment; however, neither Wenk nor
Schmitz attempts an explanation. As Schmitz puts it, “The final chord
does not assuage our thirst for a final cadence. We are left with a
choice of three tonal centers and yet none”; 61 his three tonal centers
are C, G, and D, none of which can be reconciled with the others.
Gut claims D is the central axis on which the piece turns: the
final chord, he asserts, should not be analyzed as a triad with added
ninth, but as a summary of the fifths Debussy uses through the piece.62
Dommel-Diény calls D “an immutable tonic” throughout, and a
symbol of the eternity represented by the funerary urns. She provides
support for her reading through a detailed account of voice-leading,
and describes the final chord as an appoggiatura to an imaginary D-
minor chord (which she supplies in one graph).63 Though her
harmonic plan is unconvincing in that it indicates no dominant to tonic
progression, she is the only one of those four to provide such
analytical support of a compositional strategy.
58
Roberts, 164-5; Lewin, 68; Parks, 148-50.
59
Serge Gut, “‘Canope’ de Debussy, analyse formelle et structure fondamentale,”
Revue musicale de la Suisse romande, 33/2 (May 1980), 60-65.
60
A. Dommel-Diény, L’Analyse harmonique en examples de J. S. Bach à Debussy
(Neuchâtel: Éditions Delachaux et Niestlé, 1967), 57-63.
61
Schmitz, 183.
62
Gut, 60-65.
63
Dommel-Diény, 57-63.
Those who analyse the piece in C major (or C mixolydian)

must stress a progression toward closure, and all do so with more
current approaches to harmonic analysis. Baker concludes that “There
is a rightness—even an inevitability—to the final cadence.”64 Lewin
says of the ending, “The priority of D as a melodic final [has] receded
into a vague and fragrant memory of things past.”65 Lewin and Baker
thus imply that the piece is a fragment, but open at the beginning
rather than the end. Even Warburton’s bitonal analysis notes that “the
tone C gains in prominence” through the piece,66 though his analysis
weakens the notion of the tonally open fragment by having the two
tonal areas function separately.
Many pieces by Debussy—like “Green” and “Spleen”—do
begin away from the tonic, but end rather more decisively on it,
clarifying the preceding material; this is clearly not the case with
“Canope,” whose ending intentionally obscures or compromises what
preceded it. There are models, precursors, for the fragment ending in
the music of Chopin and Schumann. Among musical fragments that
end off the tonic or on a tonic that is severely compromised, Charles
Rosen includes Chopin’s Prelude in F Major. He describes the
accented Eb added to the final F-major sonority as an unresolved detail
that can compromise the conventions of the form without quite
destroying them. Also connecting the Romantic fragment to the
infinite, Rosen notes that in this context the Eb does not so much
weaken the final cadence as make it more mysterious, prolonging the
final chord beyond the confines of the form.67 Kallberg writes that
Chopin may have intended an ironic gesture by including the Eb, thus
allowing the reinterpretation of the tonic chord as a dominant-seventh,
and introducing the possibility that the Prelude could serve as a
preparation for a following work in Bb—the original function of the
prelude as a genre before Chopin transformed it into the independent
concert work (see Chapter 6).68 This Prelude becomes even more a
64
Baker, 35.
65
Lewin, 70.
66
Warburton, “Bitonal Miniatures by Debussy from 1913,” Cahiers Debussy, 6
(1982): 11.
67
Rosen, 96-97.
68
Kallberg, Chopin at the Boundaries: Sex, History, and Musical Genre (Cambridge:
Harvard University Press, 1996), 156.
fragment when deprived of both a clear final tonic and resolution in a

following work. Rosen also mentions the Romantic form of Chopin’s
Mazurka op. 17, #4, ending with a broken fragment of its melody,
completed in the correct register but the wrong voice, as a part of the
accompaniment rather than in the melodic line. As Rosen notes, the
last bars, which return to the opening phrase, shake the firm plagal
cadence in A minor but not the sense of tonality. They turn the piece
into an ideal Romantic fragment: complete and provocative,
well-rounded and yet open.69
The ending of “Canope” recalls these techniques.70 Like the
Mazurka (op. 17, #4), Debussy’s melody ends, an unresolved
fragment, even though the resolution is sounding in a different voice
(and in a different register as well). The final chord is more
problematic since the tonic chord does not sound in the last measures,
only the tonic pitch D. Nevertheless, the return of the opening material
at pitch just before the closing, in conjunction with the unresolved
melody (motive W) allows the unresolved dissonance of the final
measures to prolong the D, in Rosen’s words, beyond the confines of
the works itself; as a fragment, “Canope” turns both back into itself
and outside itself in search of closure. In this form, the piece expresses
what the urns represent: the certainty of death and uncertainty of what
comes after.
Though often without this level of final insecurity, the
repetitive motive (an echo, an afterthought?), static harmony, and
ever-decreasing level of sound and speed that mark the final measures
of “Canope” became in Debussy’s output almost as much a topos as
the in medias res beginnings identified by Genette in nineteenth- and
twentieth-century literature. Many of Debussy’s endings are
reminiscent of Berthold Hoeckner’s representations of Romantic
distance in music (see Chapter 1). He explains the ending of
Schumann’s Papillons as such—sound gradually fading, dying away
on the dominant-seventh chord whose notes vanish one by one, the
final tonic touched only briefly.71 At least once, Debussy specifically
69
Rosen, 419.
70
Recall Debussy’s familiarity with Chopin’s music, especially Chopin’s influence on
Debussy’s music documented by Howat.
71
Hoeckner, 55-132.
links such an ending to two Romantic images: horn calls and sound
fading away in the distance—that distance which allowed both words
and paintings to be transformed into music. The score of “Les Sons et
les parfums tournent dans l’air du soir” from Book I of the Preludes
has at the fourth measure from the end the indication Comme une
lointaine sonnerie de cors [like the distant sound of horns] and for the
last two measures, which are horn calls transposed to the piano,
Encore plus lointain et plus retenu [yet further off and slower].
Similarly, the fading final chord of “Canope” reflects this aesthetic of
distance inherited from Romanticism—the distant civilization, the
distance of exoticism, and the distance of death—become more
pronounced as the Symbolists inherit from the Romantics and from
Baudelaire in particular a heightened sense of alienation from society.
Both the assurance that Dommel-Diény assigned to the
conclusion by predicting the resolution of the final dissonances to a
specific D-minor chord, as well as the “rightness” and “inevitability”
Baker assigned to the final cadence he reads as C, rob “Canope” of the
character of the Romantic and Symbolist fragment—and point to a
certainty about the afterlife that is out of character with the subject. At
the very core of Romanticism, the fragment’s broken state is seen as a
parallel to the incomplete art work: the unfinished is a reflection of the
ongoing process of creation just as the broken is a product of the
ongoing process of destruction; the Symbolist fragment is only a
refinement of that aesthetic doctrine: the whole, the ideal, becomes
less attainable, the degree of incompletion more pronounced, the role
of the audience more involved in the process of recreation.
Fig. 2.7 presents a graph that shows how the D minor tonality
is maintained with a fragmented ending. The complete progression in
D minor from mm.1-26 supports a series of nested scales descending a
fifth; this descent includes both the natural and flat second degrees of
the scale. The descending pattern is repeated in the final measures, but
not completed. The flat second scale degree introduces the
chromaticism that colors much of the work. Note in Fig. 2.6 that
motives W, X, and Y are all derived from this descent. Fig. 2.6 also
shows that motives Y, and Z are related to the descending fifth
introduced in motive U. In this graph the prelude is rounded tonally
(“complete” and “well-rounded,” to quote Rosen, above), but open at
its end (again in Rosen’s words, “provocative,” and “open”), D minor
is compromised but not negated, the lack of resolution of the final

notes prolonging it far beyond the confines of the form.72
Fig. 2.7: Counterpoint graph of “Canope”
ctpt 5-------------8 5-----8 5 5-----------8

harmony i v/vi iv fifths seq. V WT i
motives U V W X Y Z V U W
mm. #s 1 7 11 14 15 17 18 20 26 30
Commenting on a Novalis fragment from Pollen Dust, Rosen

has written, “This fragment itself is not only provisional, tentative, but
glories in its refusal to reach the definitive; and it is given to us with a
premonition of its own death, imminent or deferred, with the prospect
of its ruin.”73 Rosen eloquently and aptly characterizes the Romantic
fragment—in words that could as easily have been conceived in
reference to “Canope.”
72
Rosen, 418-19.
73
Rosen, 94.
CHAPTER 3
ARCADIAS AND ARABESQUES
Satyrs are little mannikins with hooked noses,

horns in their foreheads and feet like those of
goats such as the one Saint Antony saw in the
desert, who also, questioned by this servant of
God, is said to have replied, saying ... “I am
mortal—one of those inhabitants of the desert
whom the pagans, deceived by various errors,
worship as Fauns and Satyrs.”
The legacy of Classical Greece and Rome, left to us in stone

ruins and manuscript fragments, its stories told through the composite
creatures that people myth and legend, has been a recurring source of
inspiration for Western makers of fragments at least since Petrarch
pondered the ruins of Rome. The nineteenth century was no exception,
despite the myriad influences and interests that competed for the
attention of artists: nationalism, naturalism, exoticism, among many
others. In France, the intermingled and interdependent groups labeled
Parnassians, Decadents, and Symbolists used aspects of the Classical
tradition for their own purposes. Gilbert Highet, in his extended
discussion of this reappropriation and reinterpretation of Classical
ideals, highlights several areas: the Parnassian appeal to the
universality of emotional restraint and formal control, visible most
certainly in the poetry of Leconte de Lisle and the paintings of Puvis
de Chavannes, but also in aspects of the works of Mallarmé and the
Symbolists; the exaggerated notion of personal freedom and sexual
license in the ancient world (contrasted with the repressive control of
Christianity) seen in the works of Pierre Louÿs and Gustave Moreau,
among others (writers and painters often labeled Decadents); and the
Symbolist trove of mythological figures, from the faun to Narcissus,
laden with centuries of meaning, capable of functioning as either
subject or frame, center or decoration.1 The ancient world made a
perfect destination for escapist literature, more concrete than the lands
of pure imagination and recalling more from communal knowledge
and expectation, but at the same time still shrouded in the mist of the
past, hiding in uncertainty, opening doors to many interpretations and
1
Highet, 437-65 and 501-10.
misinterpretations.2 The nineteenth century made many of its

fragments on this rich base. Debussy’s interaction with these makers
of new classical fragments is represented in this chapter by his tone
poem based on Mallarmé’s L’Après-midi d’un faune, which premiered
in December 1894, and his settings of poems from the Chansons de
Bilitis, poems first published by Pierre Louÿs that same month.
Faune and Arabesque
Debussy invited Mallarmé to the first performance of Prélude

à “l’Après-midi d’un faune” with the following note:
Cher Maître, Ai-je besoin de vous dire la joie que j’aurai si vous
voulez bien encourager de votre présence, les arabesques qu’un
peut-être coupable orgueil m’a fait croire être dictées par la Flûte
de votre Faune.
[Dear Master, Need I say how happy I should be if you were kind
enough to honour with your presence the arabesques which, by an
excess of pride perhaps, I believe to have been dictated by the
flute of your faun.]3
Debussy’s use of the term arabesque—in his note to

Mallarmé, as the title of two works for piano, and in other instances—
has been the subject of several studies.4 Arabesque was an essential
term in aesthetic theories in both literature and the visual arts in
France in the late nineteenth century, particularly in the circles
Debussy frequented: Mallarmé and the Symbolists drew on concepts
related to the arabesque from Gautier, Baudelaire, and Edgar Poe; in
the visual arts, Maurice Denis and art nouveau movement as well as
2
For more on classical themes in the escapist literature of decadence, Jennifer Birkett,
The Sins of the Fathers: Decadence in France 1870-1914 (London: Quartet Books,
1986), 225-35.
3
Jean-Jacques Eigeldinger, “Debussy et l’idée d’arabesque musicale,” Cahiers
Debussy 12-13 (1988-89): 67, transcription and facsimile of original note; trans.
Nichols, Debussy Letters, 75, with minor changes.
4
In particular, see Eigeldinger and Françoise Gervaise, “La notion d’arabesque chez
Debussy,” La Revue musicale 241 (1958): 3-23.
Chapter 3: Arcadias and Arabesques 97
the Pre-Raphaelites drew on Ruskin5; Gauguin and Moreau also used

the term arabesque in their writings on art.
Debussy could have used the term arabesque in his note as a
(no doubt somewhat insincere) gesture of self-deprecation, the novice
daring to approach Cher Maître with his meager offering. In the
Dialogue on Poetry, Schlegel wrote of Diderot’s Jacques, “To be sure,
it is not a work of high rank, but only an arabesque,” and the term had
not lost its association with its origins in ornament—bauble, trifle—
even in Debussy’s time. Schlegel continued, “But for that reason it has
in my eyes no small merit; for I consider the arabesque a very definite
and essential form or mode of expression of poetry,”6 and Debussy
echoed that sentiment, prevalent in his time as well, when he later
wrote of the “‘arabesque musicale’ ou plutôt ce principe de
‘l’ornement’ qui est la base de tous les modes d’art” [the musical
arabesque or rather this principle of ornament which is the basis of all
the varieties of art.]7
Given the importance of the term arabesque in Mallarmé’s
theories, Debussy might have used it here, as Eigeldinger says, to give
Mallarmé what he wanted to hear.8 For Mallarmé, as Bradford Cook
points out, every art form is defined by its own arabesque, but each is
a multiple of one great art, the total arabesque.9 This Mallarméan
concept invites an interpretation of Debussy’s work as musical
arabesque intertwining with the poetic arabesque of Mallarmé’s poem
(absent but present in memory), two very real facets of the ideal total
arabesque. The origin of arabesque as frame also meshes with the
interpretation of Debussy’s Prelude as musical gloss of Mallarmé’s
poem; and fits well with the story that Debussy was originally to have
composed incidental music for a dramatic reading of the poem. At any
5
Ruskin’s ideas were transmitted in part through Sizeranne, Ruskin et la religion de la
beauté. See Richard Langham Smith, notes to Debussy on Music: The Critical
Writings of the Great French Composer, collected by François Lesure, trans. and ed.
by Richard Langham Smith (Ithaca: Cornell University Press, 1977), 31.
6
Friedrich Schlegel, Dialogue on Poetry and Literary Aphorisms, 96.
7
La Revue blanche, 1 May 1901, quoted in Eigeldinger, 10. Debussy further qualifies
his use of the term ornament, saying that it has nothing to do with musical
ornamentation.
8
Eigeldinger, 10.
9
Bradford Cook, introduction to Mallarmé: Selected Prose Poems, Essays, and
Letters, xvi.
rate, from Schlegel’s time forward, though varying in precise meaning

from group to group and context to context, arabesque retains the
double sense of both mere ornament and the basis of all creativity:
these are two strands of one common and prevalent notion. Both
Mallarmé’s poem and Debussy’s tone poem are examples of
arabesque structures, digressive narratives, deformed and interrupted,
but based, nevertheless, on recognizable conventions.
L’Après-midi d’un faune, subtitled “eclogue” after Virgil’s
designation for pastoral poems, is Mallarmé’s most famous homage to
antique myth and its continued power to inspire.10 The faun is a
grotesque, a creature made of fragments, part man, part goat;
Mallarmé has taken the interleaving technique of Hoffmann’s
rambunctious Kater Murr—who so boldly crossed the line between
the animal world and the human—and with subtlety and delicacy has
transformed it into the inner monologue of the timid faun, his
vacillation between sleep and wakefulness, dream and reality, on a
sultry afternoon in an Arcadia that never existed except in
imagination.
The poem is written in alexandrines, in rhyming couplets; on
this model he makes his arabesques. The first line of the faun’s
monologue begins in medias res (with a reference to something that
has happened in the past and about which the reader is ignorant) and
ends with a sentence fragment.
Ces nymphes, je les veux perpétuer.
Si clair,
[Those nymphs, I want to make them permanent. / So bright,]11
10
Robert Greer Cohn, Toward the Poems of Mallarmé (Berkeley: University of
California Press, 1965), 13. The following analysis draws on the work of Cohn as well
as the following: Rosaline Crowley, “Toward the Poetics of Juxtaposition: L’Après-
midi d’un faune,” Yale French Studies 54 (1977), 33-44; Houston, 56-58; Guy
Michaud, Mallarmé (New York: New York University Press, 1965), 37-38 and 89-92;
Roger Pearson, Unfolding Mallarmé: The Development of a Poetic Art (Oxford:
Clarendon Press, 1996), 108-43; Maurice Z. Shroder, “The Satyr and the Faun: A
Definition of Romanticism,” Symposium: A Quarterly Journal in Modern Foreign
Literature 23 (fall—winter 1969): 346-53.
11
Mallarmé, L’Après-midi d’un faune, lines 1-2, Oeuvres complètes, ed. Henri
Mondor and G. Jean-Aubry (Paris: Éditions Gallimard, 1945), 50; trans. Austin, in
Debussy, Prelude to “The Afternoon of a Faun,” 23.
By placing the caesura at points where it is not expected,

Mallarmé fragments the alexandrine, as in this first line. Here the
twelve syllables are broken, both by syntax and by placement on the
page, into groups of three, seven, and two syllables rather than the
accepted groups of six, or even four. The rest of the poem proceeds
through sentence fragments indicated by dots of ellipsis, dashes, blank
space,12 and other signs of omission and interruption—physical signs
earlier used by Rabelais, Cervantes, Sterne, and others who depicted
incomplete manuscripts in their texts.13 The faun, like the narrator in
Aurélia, tells his own story, attempting to separate dream from reality:
he saw some nymphs, or maybe swans; he captured two of them; they
escaped before he could make love to them. The story is told as a
factual account, told as if it had happened in the past, in three separate
sections that are marked with quotation marks, and printed in italic
type. In Mallarmé’s letters to his editor Edmond Deman, he described
the use of italics as the printed equivalent of handwriting, indicating,
according to Graham Robb, a draft, a fiction, rather than reality.14 The
first of these sections is introduced by the word “contez” in capital
letters; the second by the word “souvenirs,” again in capital letters:
“tell” and “memories.” Like Petrarch and Montaigne, the faun
questions telling and memory, for telling is always reconstructed from
memory, and memory is always imperfect and always itself a
reconstruction. Surrounding this narrative is the faun’s gloss—his
musings, his emotions, his questions—told in the fictional present.
These sections frame the narrative as the grotesque frames the center
portrait it surrounds. Embedded in the final section of the gloss is a
tiny mise en abyme, a fragment story that sums up the faun’s
experience. In this short dream, he dares to hold Venus, but waking is
his punishment as the dream trails off in dots of ellipsis. The line
continues, after a break on the page, with the word “non.” This was a
dream; perhaps the nymphs were also a dream. There is no resolution
12
In the original edition of L’Après-midi d’un faune the intervals of blank space on
the page take almost as much space as the sections that are italicized. Crowley, 43.
13
Bowie claims many of Mallarmé’s poems could be profitably studied for their gaps,
elisions, and discrepancies. Malcolm Bowie, Mallarmé and the Art of Being Difficult
(Cambridge: Cambridge University Press, 1978), 56.
14
Graham Robb, Unlocking Mallarmé (New Haven: Yale University Press, 1996),
213.
of any section of the poem, or of the poem itself, but, in fact, it hardly
matters. As the last line reminds us, “Couple, adieu; je vais voir
l’ombre que tu devins” [Couple, goodbye; I go to see the shadow you
became]: 15 whether real or dream, everything becomes memory.
Debussy’s tone poem also bears the marks of the arabesque,
of fragmentation. As Matthew Brown shows in his analysis of Prélude
à “L’Après-midi d’un faune,” Debussy’s work may be interpreted as a
standard ternary form—ABA1 followed by a coda—but one where
fragmentation on many levels has obscured the structure to such an
extent that it has opened the door to many differing analyses. 16
Brown’s formal divisions are summarized in Fig. 3.1 below.
Fig. 3.1. Prélude à “L’Après-midi d’un faune,” Brown’s formal

divisions
mm. 1-30 31-36 37-54 55-78 79-106 106-110

A digression transition B A1 coda
E whole tone to Db Db E E
opening with material composing
incomplete extends V of out of mm
progression in E, delays 20-30, with
mm. 1-13 transition to allusion to
vii7/V-V9-I Db B section
The work opens with what Brown labels a transformation of

an incomplete progression: vii/V-V9-I in E major. Thus Debussy
begins in medias res, in the middle of a chord progression, just as
Mallarmé began the faun’s tale in the middle of his musings. The
tonic, E major, is reached only in m. 13 with a clear V9-I, but that
point of arrival and clarification is masked by its brevity and by
misalignment with the end of the melodic phrase above it.17 The
progression has been interrupted, not harmonically but aurally, by a
15
Mallarmé, L’Après-midi d’un faune, final line, Oeuvres complètes, 53; my
translation.
16
Matthew Brown, “Tonality and Form in Debussy’s Prélude à “L’Après-midi d’un
faune,” Music Theory Spectrum 15 (1993): 127-43.
17
The opening progression VII7 of V-V9-I in mm. 113 actually bridges the first two
statements of the flute theme. See Brown, “Tonality and Form,” 134.
sort of hole in the manuscript—m 6 is completely silent; m. 7 begins a

repeat of material from mm. 4-5, doubling back on itself just as the
visual arabesque may double back before going on. The effect also
mimics memory retrieval, recollection. Memories are often fragments,
incomplete, inexact, brought to the forefront of thought in stages, first
dimly then more clearly. Rosen stresses Schumann’s ability to mimic
the imperfect process of memory by introducing first a short fragment
of a known melody, followed by a longer version, still incomplete.18
Debussy often uses a similar technique of repetition and expansion to
lead the listener into his works: the opening of “Danseuses de
Delphes” and “Les Sons et les parfums tournent dans l’air du soir,”
both from Book I of the Preludes, are examples, as is the opening of
Prélude à “L’Après-midi d’un faune.” In these works, Debussy is not
recreating from recollection, but literally creating the impression of
recollection. The faun’s memories are the source of his creativity,
though they may not be memories, or even dreams, but rather
constructs of the imagination, mimics of memory.
Mallarmé questions dream, memory, and reality with
digressions, interruptions of the narrative; Nerval did the same when
he intertwined reality and psychotic episodes in Aurélia; Cervantes
used his interpolated stories to comment on reality; Verlaine’s tiny
interpolations reveal the uncertainty of his protagonist, his questioning
of the attitude of the beloved. Just as Mallarmé delayed the faun’s
narrative with his gloss, so Debussy waylayed his musical progression
with the musical equivalent of asides or parentheses (such as the brief
two-measure interruptions in “Green” that echo the change of tone in
the poem). One such harmonic digression or interpolation is found in
mm. 30-39, where whole-tone material functions to extend the
dominant chord of E major, as Brown shows, and to delay the
modulation to D-flat major, but does not advance the primary tonal
progression of the work. In his article, Brown includes a passage from
Schenker’s Free Composition describing such delays in the harmony
and compares them to life:
In the art of music, as in life, motion toward the goal encounters

obstacles, reverses, disappointments, and involves great distances,
18
Rosen, 98-100.
detours, expansions, interpolations, and, in short, retardations of

all kinds. Therein lies the source of all artistic delaying, from
which the creative mind can derive content that is ever new.19
Both Debussy and Mallarmé derived “content ever new” from the
“reverses, detours, interpolations” they made to accepted and
recognizable models.
Debussy also uses a compositional technique that parallels,
though only loosely, the use of the mise en abyme—often expressed as
an embedded or interpolated text whose purpose is either to reflect in
miniature, or through similarity to shed light on, the principal work
that contains it. The faun’s decision to dream/imagine that he goes to
Mt. Etna and rapes Venus, in spite of “sûr châtiment” [sure
punishment], captures his frustration, his dreams of conquest, his
impotence, and ends with dots of ellipsis as uncertain of its ending as
the poem itself.20 To form his A1 section, Debussy did not simply
repeat the A material literally or with some alteration; instead, as
Brown has shown, Debussy chose a small section, mm. 20-30 of the A
section, which he expanded (composed-out), including in the process
“subtle reminiscences” of motivic material and the tonality (C#/Db) of
the B section. 21 This fragment of A then becomes a tiny reflection, as
well as the basis of, the A1 section. It functions much as a mise en
abyme through its explanatory power, its representation in miniature
of the final A section, though it is not an interpolation or interruption.
Mallarmé wrote of the arabesque:
A l’égal de créer: la notion d’un objet, échappant, qui fait défaut.

Semblable occupation suffit, comparer les aspects et leur nombre
tel qu’il frôle notre négligence: y éveillant, pour décor,
l’ambiguïté de quelques figures belles, aux intersections. La totale
arabesque, qui les relie, a de vertigineuses sautes en un effroi que
reconnue; et d’anxieux accords. Avertissant par tel écart, au lieu
de déconcerter, ou que sa similitude avec elle-même, la soustraie
en la confondant.22
19
Brown, “Tonality and Form,” 136, quoting Schenker, 5 (sec. 3).
20
The Mad Tryst, the story the narrator reads to Roderick in The Fall of the House of
Usher, a work Debussy knew intimately, functions as another example of mise en
abyme.
21
Brown, “Tonality and Form,” 140.
22
Mallarmé, La Musique et les lettres, Oeuvres complètes, 647-48.
To create is to conceive an object in its fleeting moment, in its

absence. To do this, we simply compare its facets and dwell
lightly, negligently upon their multiplicity. We conjure up a scene
of lovely, evanescent, intersecting forms. We recognize the entire
and binding arabesque thus formed as it leaps dizzily in terror
play disquieting chords; or, through a sudden digression (by no
means disconcerting), we are warned of its likeness unto itself
even as it hides.23
We recognize the arabesque in all its myriad forms only against

expectation, its facets as convention deformed, its power as the
tension between the fragment and the whole.
Chansons de Bilitis: Louÿs and the Tradition of the Hoax Poem
The past speaks with the voice of authority and confers a

legitimacy on its ruins and relics that contemporary works must earn.
A false claim to this authority, coupled with contemporary tolerance
of direct descriptions of sexuality based on the notion that
pre-Christian morality permitted greater sexual freedom, allowed
Pierre Louÿs’s literary forgery, the Chansons de Bilitis, to enjoy great
success both before and after his hoax was revealed. Louÿs’s alleged
translations of the fake poems of an imaginary Greek poet, and
Debussy’s responses to them, are of particular interest for several
reasons: these prose poems are related directly to the great
eighteenth-century literary hoaxes, an important element in the climate
that allowed the Romantic fragment to develop; they are examples of a
reinterpretation of the themes of classical mythology and a
representation of views of life in the ancient world written by a
Symbolist/Decadent poet who was also Debussy’s close friend;
specifically, the text of “Le Tombeau des naïades,” the third and last
of Debussy’s settings of prose poems from this collection, illustrates
another of his musical interpretations of the theme of fragmentation.
The eighteenth century was mad for ruins; two of the more
bizarre manifestations of that mania were sham ruins and hoax poems.
Neither of these phenomena was new—the earliest report of a fake
23
Mallarmé: Selected Prose Poems, Essays, and Letters, 48-49.
ruin occurs in the sixteenth century 24 and the literary forgery was
certainly older25—but the eighteenth century experienced a frenzy in
this construction of decay. The creators of architectural and literary
fakes worked with skill, care, and great seriousness, as witnessed by
Joseph Heely’s description of a Gothic castle built in 1747-48:
And to keep the whole design in its purity—to wipe away any
suspicion of its being any otherwise than a real ruin, the large and
mossy stones, which have seemingly tumbled from the tottering
and ruinous walls, are suffered to lie about the different parts of
the building, in utmost confusion. This greatly preserves its
intention, and confirms the common opinion of every stranger, of
its early date; while, to throw a deeper solemnity over it, and to
make it carry a stronger face of antiquity, ivy is encouraged to
climb about the walls and turrets.26
The great literary forgers were no less intent on maintaining

this deception. Many hoax poems were presented in a fragmented
condition to ape the genuine found manuscript;27 hoax authors,
including the greatest of the eighteenth century, Macpherson (the
Ossian poet) and Chatterton (the Rowley poet), included prefaces that
detailed the circumstances of the recovery of the manuscripts they
claimed to have found, as well as biographical data that situated the
supposed original authors in time and place. Rabelais and Cervantes
provided similar (though not so complete) descriptions for the
imaginary found manuscripts they used as props for their fictions, but
these eighteenth and nineteenth-century forgeries were not props: they
were presented to the public as literary works themselves.
Once these hoaxes were uncovered, they presented critical
problems that placed them in a category quite different from both the
legitimate literary fragments of the ancients (known most often
through published translations in such popular collections as the Greek
24
Harries, 62.
25
Roger A. Pack, “Two Classical Forgeries,” American Journal of Philology 110
(1989): 479-80.
26
Harries, 65-66, quoting Joseph Heely, Letters on the Beauties of Hagley, Envil, and
the Leasowes, with Critical Remarks and Observations on the Modern Taste in
Gardening, 2 vols. (London: R. Baldwin, 1777), 175-76.
27
Levinson, 34.
Anthology) and the planned fragments of the seventeenth and

eighteenth centuries. Fragments of ancient Greek and Latin poems
provided the reader in the eighteenth century with a form whose
incompletion was a mark of its antiquity, comparable to the missing
nose on an antique bust, and interpreted only as the result of the ruin
of time, not as the result of artistic intent. Such fragmentation did not
need to be questioned, since it was not part of the initial artistic plan.28
Many fragments of the seventeenth and eighteenth centuries (those of
Bacon, Hamann, Diderot, and Sterne among many others—see
Chapter 1) were made quite deliberately, were proclaimed fragments
by their authors who openly used fragmentation and incompletion to
confront any number of concerns. On the other hand, the hoaxes, once
discovered, created interpretative and critical problems, particularly if
the hoax translations had been judged excellent before they were
uncovered. Legitimate translation admits a level of imperfection: the
impossibility of knowing the past—ancient languages, scripts,
customs—and the mutilated state of the original text all contribute to
accepted shortcomings in the translation, to holes the translator could
not be expected to fill. In these modern hoaxes, such problems—such
holes—were less easily forgiven. Either their imperfections, including
a certain lack of originality, had to be attributed to shortcomings on
the part of the writer, or had to be justified as integral to the form of
the work; they could not be blamed on the difficulties of dealing with
the past. The legitimate translation asked the reader to accept
incompletion on the basis of unavailability of information; the hoax
writer asked the reader to accept incompletion as necessary to the
telling of his tale, necessary to his plan. At the same time, questions of
creation and origin relevant to the meaning of the true antique had no
meaning for the hoax; background information supplied by the hoax
author, normally accepted at face value, had to be interpreted along
with the main text. According to Levinson, this confrontation prepared
general audiences and critics for the many nineteenth-century
fragments whose incompletion was an integral part of both the work
itself and the aesthetic behind it, the aesthetic of longing and
28
Levinson, 20-21. As sources of translations of such fragments, Levinson notes in
particular the eighteenth and nineteenth-century editions of The Greek Anthology and
The Palatine Anthology as well as collections of writings by individual authors.
incompletion. As Levinson observed, “Without the hoax poems, the

fragment might have remained the province of the antiquarian, the
metaphysician, and the connoisseur of sensation and sensibility.”29
In the tradition of the Ossian and Rowley poets, Pierre Louÿs
contributed his own manuscript hoax to a world in part still too naive,
or still too anxious to uncover fragments of the past, to recognize
immediately the joke he played. The Chansons de Bilitis were
supposedly translations of the works of Bilitis, a sixth century BCE
Greek-Phoenician poet and contemporary of Sappho, who
documented her life through her poetry. She records three stages: her
childhood and sexual initiation in Pamphylia in Asia Minor, her
lesbian life with Mnasidika in Mytilene in the company of Sappho,
and her later life as a courtesan on Cyprus. Louÿs published the
Chansons de Bilitis in 1895, with all the accoutrements needed to give
them the authority and legitimacy of the ruin.30 As a younger man,
impressed by Leconte de Lisle’s translations of the Iliad and the
Odyssey, he had taught himself Greek. Before he wrote the Chansons
de Bilitis, he published the first French translations of the epigrams of
the Syrian-Greek poet Meleager31 and Lucian’s Courtesans’
Conversations.32 The complete title of the Bilitis poems, Chansons de
Bilitis traduites du grec pour la première fois par P. L., ostensibly
presented the reader with a translation made by a reputable
translator.33 Macpherson had recorded his hoax epic in rhythmical but
29
Levinson, 35. Complete discussion, 34-48.
30
The first edition is dated 1895, but the actual publication date was 12 December
1894. See H. P. Clive, Pierre Louÿs (Oxford: Clarendon Press, 1978), 110.
31
In 1893, Louÿs published first a collection of ten epigrams with a life of Meleager
in the Mercure de France, followed by the larger Les Poésies de Méléagre, published
by the Librairie de l’art indépendent. Highet, 458, and Clive, 98. Many of Meleager’s
epigrams were found in the Greek Anthology. They were also imitated by French
Renaissance poets, especially Ronsard, whom Louÿs admired. Clive, 69.
32
Highet, 457-58.
33
Louÿs at one point intended to publish scholarly notes for his fake translation. He
imitated some authentic ancient Greek poems, and quoted from others, in order to
claim that Bilitis was the original author and that others of her time and after had
copied her. Lawrence Venuti, “The Scandal of Translation,” French Literature Series
22 (1995): 26-28.
non-rhyming prose; 34 Louÿs made his fake translations into French

prose poems, each paragraph a verse unit of the supposed original
Greek.35 Following Macpherson and Chatterton, Louÿs provided a
preface that summarized the life of Bilitis and explained how he had
come to find and translate her poems. He added legitimacy to Bilitis
and her work by claiming that her tomb had been discovered by a
German, Professor G. Heim, who was also the first editor of her
poems (Bilitis’ saemmtliche Lieder zum ersten Male herausgegeben
und mit einem Woerterbuche versehen, von G. Heim, Leipzig, 1894);
Louÿs described the circumstances of the preservation and discovery
of the tomb and the poems written on its walls, here functioning as his
found manuscript. They are not described as crumbling, defaced, or in
any way marred, but Louÿs, perhaps as a gesture to the condition of
ruin, lists several poems in the index as untranslated. He gives the
impression either that he had found an incomplete work, or that those
poems were, for some reason, untranslatable. David Grayson points
out that Louÿs may have been implying that they were too risqué to
translate.36 This also gave him the opportunity to publish more Bilitis
poems at a later time, which, in fact, he did.
Why Louÿs chose to perpetrate this hoax is unclear.
Macpherson’s and Chatterton’s motivations to forgery may have
arisen as a direct response to the value the eighteenth century placed
on genius and originality. The works of the great masters of the past
presented standards impossible to meet, much less surpass; the
eighteenth-century writer had to ask—whether consciously or
unconsciously—how to become a great poet in his own age.
Macpherson’s answer, and subsequently Chatterton’s, was an appeal
to the ready-made authority of the past—an authority that relieved
them of the greatest burden, the burden of originality, and allowed
flaws to be attributed to the physical condition of the original and to
difficulties of translation.37 The nineteenth century saw no decline in
34
Robert Folkenflik, “Macpherson, Chatterton, Blake and the Great Age of Literary
Forgery,” The Centennial Review 18 (1974): 380-81.
35
Pack, 482.
36
David Grayson, “Bilitis and Tanagra: Afternoons with Nude Women,” in Debussy
and his World, ed. Jane Fulcher, 118.
37
Folkenflik, 378-82.
the problems facing the eighteenth century poet. Issues of originality

and genius were certainly no less important, and the specter of the past
loomed perhaps even larger. Whether Louÿs wrote the Chansons de
Bilitis to tease the professors, whether he intended it to be uncovered,
believing the publicity of a literary scandal was worth any negative
response, or whether he approached his hoax as a legitimate response
to the same problems Macpherson and Chatterton had faced, his
strategy is remarkably similar to theirs.38
In addition to the hoax format, Louÿs built on other fragment
forms and traditions. He created a lyric cycle: even his prose poem
paragraphs are separated by blank space on the page like stanzas of
poetry—in the form of Hellenistic epigrams, one of Louÿs’s critics
wrote.39 The sequence of isolated lyric moments bound loosely
together, separated from a narrative they only hint at, opens the
door—as it has since Petrarch defined the genre by omitting the prose
narrative from his Rime sparse—to multiple interpretations, to active
imaginative participation on the part of the reader. Louÿs is not as
subtle as Petrarch or others who write in this genre: he is torn between
the inherent uncertainty of the genre and the need to justify Bilitis in
order to maintain his hoax. As a result, his prefatory Vie de Bilitis
must answer questions his prose poems leave unanswered.
Louÿs’s work, like Isidore’s monsters, inhabits the margins,
the edges, the barely acceptable: the hoax uncovered remains a hybrid,
always appearing to be something it is not; prose poems and lyric
cycles made of prose poems are hybrids; Bilitis’s sexual exploits are
presentable only in the context of their fake origin; Louÿs writes to his
brother that the work is a souvenir of Algeria,40 a country both foreign
and exotic, like Meriem ben Atala, the North African woman who
served in part as a model for Bilitis.41 The hoax itself, not so cleverly
concealed and soon revealed, caused great controversy, not only
38
Folkenflik, 378-82 and Pack, 483 for more information on this point. As Clive, 110,
reports, “Louÿs pointed out to a correspndent in1896, anyone with a knowledge of
German should have put on his guard by the professors name, since ‘G/Heim =
Geheim = Le mystérieux’.” Perhaps “secret” is a better English translation.
39
Venuti, 32.
40
Venuti, 31.
41
Venuti, 30, and Clive 10-46. Louÿs dedicated the first edition to Gide, referencing
Meriem ben Atala by initials only. She was romantically involved with both men.
because it inhabited the land of the unclassifiable, and not only

because any forgery questions the distinctions between translation,
authorship, and scholarship (as Venuti notes), but also because Louÿs,
whether advertently or inadvertently, triggered nationalistic and
political responses. Louÿs promoted Sappho’s homosexuality—
harshly denied in some circles, especially German, but more
acceptable in France—when he depicted Bilitis in Sappho’s company,
and he promoted and accepted Jewish culture when he gave Bilitis the
Syrian name for Aphrodite. On both accounts he was taken to task by
defenders of Sappho and by antisemites.42
Debussy and Bilitis
Debussy responded three times to the Chansons de Bilitis. He

first set three of the poems, on his own initiative, in 1897-98: “La
Flûte de Pan” (#20 in the first edition), “La Chevelure” (not included
in the original edition but sent to Debussy and later published in
Mercure de France in August of 1897), and “Le Tombeau des
naïades” (#31 in the original). Debussy’s setting of “La Chevelure”
was first published in L’Image under the title “Chanson de Bilitis,”
with decoration by Kees van Dongen.43 The songs were published as a
set in 1899 and first performed as a group in 1900 by Blanche Marot.44
42
Venuti, 34-35, gives more information. The poems were first published before the
height of the Dreyfus Affair. Dreyfus was arrested in 1894 but Zola’s article
“J’accuse” was not published until 1898; afterwards, Louÿs sided with the anti-
Dreyfus camp. See Nichols, The Life of Debussy (Cambridge: Cambridge University
Press, 1998), 92.
43
Information on the publication of the poems and songs comes from Clive, 100-12,
141-42, and 170-71, and Wenk, Claude Debussy and the Poets, 171-96.
44
Regardless of her status at the time of Debussy’s famous interrogation of Blanche
Marot’s mother concerning the state of the young woman’s sexual awareness, by the
time of the first performance, Blanche Marot was the mistress of Debussy’s
benefactor Georges Hartmann, as Debussy knew. Nichols, The Life of Debussy, 96.
Debussy’s supposed conversation with the mother is evidently reported first by
Blanche Marot herself in Charles Oulmont, Noces d’or avec mon passé (Paris, 1964),
70. Roger Nichols, Debussy Remembered (Portland, Oregon: Amadeus Press, 1992),
59. Jeanne Raunay refused to perform the songs for moral reasons. Lesure, Claude
Debussy, 228.
Debussy’s second response to Bilitis took the form of

incidental music for a private reading in 1901 of eleven of the prose
poems accompanying tableaux vivants and mime. None of the three
prose poems Debussy had previously set were included. This
incidental music, scored for two harps, two flutes, and celesta is
delicate and atmospheric, as Orledge says, stressing modal and whole
tone lines.45 Debussy’s music blurs any supposed musical reference to
ancient Greece into the typical Romantic formula for the exotic, no
more authentic than Louÿs’s attribution of oriental sexual practices
and North African landscapes to ancient Greece46—the elusive Other,
so easy to bring to mind, so difficult to differentiate and define. In a
final response to Bilitis, Debussy reworked and expanded the
incidental music as the Épigraphes antiques for piano duet.47
Louÿs and Debussy were such close friends during this time
that it seems likely Debussy would have been one of those who knew
about the hoax even before publication, though this is not made clear
in the letters. Debussy’s song cycle presents only the early life of
Bilitis.48 It is less direct than Louÿs’s original, not simply because of
the limited number of poems he set, but because of his choice.
Debussy omits poems in which Bilitis speaks directly of emotion or
action: her tactics to evade her mother in order to visit Lykas, the
undesirable shepherd she loves, or the account of her rape, or her
child. The first two settings are texts of sensuality and desire, without
narrative detail; the final poem appears, on the surface, to be a
statement of unexplained, but presumably inevitable, disillusionment,
but in the following discussion, I advance a less desolate
interpretation, one that suggests a new response to fragmentation as a
source of renewal and creativity.
45
Robert Orledge, Debussy and the Theater (Cambridge: Cambridge University
Press, 1982), 245-49.
46
Venuti, 33-35 and Highet, 458.
47
Orledge, 245-49, gives more information on this expansion. On these works, see
also Grayson, “Bilitis and Tanagra,” 121-123.
48
Susan Youens, “Debussy’s Song Cycles,” The NATS Bulletin (Sept./Oct., 1986): 13,
lists the Chansons de Bilitis as one of only three of Debussy’s song collections that
qualify as song cycles; all contain minimal narrative.
“Le Tombeau des naïades”

Le long du bois couvert de givre, je marchais; mes cheveux
devant ma bouche se fleurissaient de petits glaçons, et mes
sandales étaient lourdes de neige fangeuse et tassée.
Il me dit: “Que cherches-tu!” “Je suis la trace du satyre. Ses petits

pas fourchus alternent comme des trous dans un manteau blanc.”
Il me dit: “Les satyres sont morts.
“Les satyres et les nymphes aussi. Depuis trente ans il n’a pas fait
un hiver aussi terrible. La trace que tu vois est celle d’un bouc.
Mais restons ici, où est leur tombeau.”
Et avec le fer de sa houe il cassa la glace de la source où jadis

riaient les Naïades. Il prenait de grands morceaux froids, et les
soulevant vers le ciel pâle, il regardait au travers.
[Along the wood covered with frost, I walked; my hair in front of

my mouth was decorated with little icicles, and my sandals were
heavy with muddy, packed snow.
He said to me: “What are you seeking?” I am following the trail

of the satyr. His little cloven hoofs alternate like holes in a white
mantle.” He said to me: “The satyrs are dead.
“The satyrs and the nymphs also. For thirty years there has not
been a winter so terrible. The trail that you see is that of a buck.
But let us remain here, where their tomb is.”
And with the iron of his hoe he broke the ice of the spring where
once the naiads laughed. He took large cold pieces and, raising
49
them toward the pale sky, he looked through them.]
In “Le Tombeau des naïades,” Bilitis searches a winter

landscape for the elusive nymphs and satyrs she has previously
pursued through the sunny meadows and deep forests of her youth.
She follows false tracks, perhaps as false as the other near-sightings
mentioned in her earlier poems, but these tracks finally lead to her
goal—too late. An unidentified man tells her that the nymphs and
satyrs have all perished in the cold; he shows her the tomb of the
49
Pierre Louÿs, Les Chansons de Bilitis (Paris: Albin Michel, 1965), 76; trans.
Benton, Debussy: Songs, 174.
naïads. The poem is, of course, a not very subtle marker of the end of
youth and youthful dreams: after this, Bilitis leaves her home and
writes no more of her childhood, her first love, or her infant.
The poem is a fragment, beginning in the middle of a journey,
or simply a trek through the snow; the text offers no explanation, nor
do the previous poems. Only Louÿs’s preface informs us that Bilitis’s
love affair ended and that she abandoned her child at the end of the
first stage of her life, marked by this poem. Stephen Rumph’s
interpretation draws on parallels between “Le Tombeau des naïades”
and the opening of Pelléas et Mélisande: the young woman, seated by
a spring, discovered in the forest by an older man, questioned by
him.50 Rumph reads into this poem, and Debussy’s setting, a very
negative masculine/feminine conflict: the unidentified man is
ultimately responsible for usurping Bilitis’s poetic voice, responsible
for her loss of self and the death of her creativity. When he breaks the
ice covering the spring of the naïads, he reenacts the rape of Bilitis
reported in a previous poem.51 There is much room for interpretation
in this poem,52 and while recalling Pelléas et Mélisande is logical—
Debussy had finished the bulk of the opera before he started these
songs—any study that considers similarities in character and situation
should perhaps also consider similarities in the music.
In his interpretation of Pelléas et Mélisande, Richard
Langham Smith discusses Debussy’s symbolic use of tonality: he
claims that C major (more accurately, it is often C as the dominant of
F major) accompanies darkness; that F-sharp major and its dominant
accompany both the appearance of light and the striving for the ideal,
which in the opera, according to Langham Smith, is Mélisande herself.
50
Stephen Rumph, “Debussy’s Trois Chansons de Bilitis: Song, Opera, and the Death
of the Subject,” The Journal of Musicology 12 (1994): 477-83.
51
Rumph fails to point out the hoe as an instrument used in agriculture, especially
during growing season. It is unusual in the winter scene, and evocative when seen as
an instrument associated with fertility that the man uses to break the ice/hymen.
52
Rumph’s interpretation is based on several assumptions that I would hesitate to
present as fact: we do not know that the man is older: we know only that he is male,
that he carries a hoe, and that he knows that they are near the tomb of the naïads.
Anyone could have known it was the coldest winter in thirty years; he could be Lykas
or a stranger; they could have come upon each other in the forest or they could have
been walking together.
In act 2, scene 3, G-flat (F-sharp) major tonality accompanies the

moonlight flooding the entrance of the grotto and Pelléas’s line, “voici
la clarté”; also, Mélisande’s “Je t’aime” in act 4, scene 4 is followed
by a move to F-sharp major and includes a reference to ice breaking:
Pelléas sings “On a brisé la glace avec des fers rougis!” [The ice has
been broken with red-hot irons.]53
Does “Le Tombeau des naïades” reflect some of the same
tonal significance? Debussy’s setting is a fragment, beginning with a
key signature of three sharps (F-sharp minor) on a G-sharp half
diminished seventh chord (i.e. ii7 in that key) but ending quite
emphatically and undeniably, after the appropriate key signature
change in m. 25, with a dominant-seventh to tonic resolution in and on
F-sharp major. The only real indication of a tonic key area before m.
25 occurs in m. 9, where an F-sharp minor chord is arpeggiated, but
the inflection is modal, with no leading tone, and is unconvincing. In
addition, Debussy continues the negative association of C major by
arpeggiating a C major chord in mm. 11-14, cancelling the C# and G#
with accidentals. Here, Bilitis explains her quest; a root position C
major chord accompanies the man’s response, “Les satyres sont
morts.” Only through the final clarity of the F-sharp major section
does the initial key signature seem relevant. Both the repeated motion
of the inner voices and the weak chord progressions leading to the
final section in F-sharp major may reflect Bilitis’s wandering, her
unfruitful search for the nymphs and satyrs; the final section in F-
sharp major begins at the climax of the piece—the highest pitch in the
voice and the loudest dynamic marking—just as the man breaks the
ice and raises pieces to the sky, to light, to the ideal, as Wenk notes.54
If any vestige of Langham Smith’s tonalities of darkness and light
remains in this song, I question Rumph’s negative reading. If F-sharp
minor represents the search for the old myths, the search of the poet
for inspiration (just as Louÿs was inspired by ancient myth and
legend), the F-sharp major represents the light, the source of that
53
Richard Langham Smith, “Tonalities of Darkness and Light,” in Claude Debussy:
Pelléas et Mélisande, 107-39. My translation.
54
Wenk, Claude Debussy and the Poets, 193-94. Cohn notes the similarity to
L’Après-midi d’un faune, when the faun raises the grape skins to the sky and looks
through them.
inspiration. F-sharp major, accompanying the breaking of the ice and

raising of the pieces to the light of the sky, represents both the freeing
of the old myths and their reinterpretation in a new poetry, not their
deaths and the loss of their ability to inspire. The poem then ends with
the triumph of both the old symbols and Bilitis’s ability to use them in
a new way as she grows from a child into a mature poet.
Symbolist poets and writers use ice and glass as symbols of
sterility as well as the gateway toward the ideal. Since both Debussy
and Louÿs were part of the Symbolist milieu and were well acquainted
with Mallarmé’s works,55 two of his poems may serve as illustrations.
In “Le Vierge, le vivace et le bel aujourd’hui” the swan/poet/sign
(cygne/signe) is trapped in the ice of a frozen lake; he cannot or will
not free himself to reach the sky, the creative act. The promise of
fertility inherent in the word vierge cannot be realized unless the ice is
broken.56 In “Les Fenêtres,” an old man, dying in a white hospital
room, looks out the window at sunset. He sees/hallucinates beautiful
images, infinity, through the glass; then the transparent glass becomes
translucent, his own face is reflected in the glass, and he sees himself
as an angel. He wants to break the glass, to escape, but again, he, like
the swan, cannot.57
Bilitis walks through the cold snow, unable to find the
mythological creatures she has sought all her life, unable to find
happiness with the man she loved or the child she abandoned. All her
hopes are now frozen under the ice/window pane. The man, agent of
action for Louÿs’s (in modern terms) sexist view, frees the old
symbols and points Bilitis toward the ideal—new symbols drawn from
fragments of old myth, or, if we accept John Porter Houston’s reading
of the frozen lake in “Le Vierge, le vivace et le bel aujourd’hui” as a
state of mind, haunted by memories of what never happened, then the
ice breaking could free Bilitis from the myths that never were real (she
55
Louÿs started attending the mardis in 1891 according to Wallace Fowlie, Mallarmé
(Chicago: The University of Chicago Press, 1953), 16.
56
Fowlie, 96-101; Bernard Weinberg, The Limits of Symbolism: Studies of Five
Modern French Poets (Chicago: The University of Chicago Press, 1966), 170-186.
Many interpretations are similar.
57
Fowlie, 31-32.
had never been able to see the nymphs or satyrs).58 Perhaps the man
does reenact the rape that occurs earlier in the cycle, but to Louÿs, it
was the rape that freed the woman to move forward, that freed Bilitis
to leave her unhappy situation and to become a poet. In his novel
Aphrodite, the heroine—having just left her homeland and no longer
having her mother’s wrath to fear—rides into Tyre proudly showing
on her legs the blood that proves she is no longer a virgin. Perhaps in
Louÿs’s rather unfortunate sexist view, rather than taking away
Bilitis’s voice, as Rumph suggests, the man offers her another chance
to find it. Debussy’s F-sharp major setting of this section, tonally so
clear compared to the rest of the piece, surely presents the end of the
poem in a positive light. Perhaps the tonal clarity, now able to
incorporate the continuous ostinato patterns into a tonally stable
environment, is itself a fragment of the old system, resolving the tonal
inconclusiveness of the rest of the song, providing the key to the
puzzle just as the myths, in bits and pieces, offer their key to creativity
to so many poets of this time.
58
Houston, 114-15.
CHAPTER 4
THE SKETCH
A monster was born of Alexander’s wife

which had human upper parts, which were
dead, and the lower parts of various animals,
which were living, indicating the king’s
sudden death, for the worse parts outlived the
better. But these monsters which are sent as
signs do not live very long, dying as soon as
they are born.
The white-hot flash of inspiration frozen before intellect and

craft overlay genius; a glimpse, perhaps voyeuristic, into the
realization of the idea; the face-to-face encounter with failure, with the
artist’s inability to merge idea with form—whatever we choose to see
in the sketch, it always fascinates. We ponder it, analyze it, frame it to
exhibit in museums, and, as with any fragment, are compelled, at
some level, to complete it. The very word “sketch” in the title of a
work, even one the artist or writer presents for exhibition or
publication, is enough to place the work in a special category for the
viewer, who has now been given permission to “finish” it in his
imagination, and for the critic, who may assess it from a different
perspective than a supposedly completed composition.
As the fragment tradition of the eighteenth century fed
Romanticism, the sketch—mirror image of ruin—achieved a level of
acceptance and appeal that challenged the status of the finished work,
that questioned how “finished” could even be determined or defined.
Debussy’s direct response to the sketch, and to the role of the sketch
in his milieu, is the subject of this chapter. I will concentrate on two
compositions: D’un cahier d’esquisses, a piano work that, by title,
involves itself in this tradition, and Morceau de concours, a piece that
Debussy created by literally patching together two short, unaltered
sections from the only known sketches for his unfinished opera Le
Diable dans le beffroi.
The sketch and the ruin: both are incomplete, both have the
power to provoke a sense of sadness or nostalgia, both inspire the
imagination of the observer. Pliny the Elder’s observations echo
sentiments familiar today:
Another most curious fact and worthy of record is that the latest
works of artists and the pictures left unfinished at their death are
valued more than any of their finished paintings. … The reason is
that in these we see traces of the design and the original
conception of the artists, while sorrow for the hand that perished
at its work beguiles us into the bestowal of praise.1
The Renaissance knew the aesthetic value of the incomplete in

both the sketch and the ruin, but made a distinction between the two:
though related by surface resemblance—by lack of detail and finish—
only the sketch possessed the inherent sense of spontaneity and
immediacy that offered a glimpse into the moment of artistic
inspiration.2 The sketch, however, was still only the preliminary step.
The Renaissance recognized, but did not purposely make, fragments.
The rough, broken surfaces of the mutilated Belvedere torso might
resemble the rough, uncut stone of Michelangelo’s unfinished Slaves,
but the Slaves were never meant to be exhibited in their incomplete
state, and they found their first home not in a public place or even on
display in the gallery or home of a wealthy patron, but rather in a
grotto. There, among the collected artifacts—the shells, stones, and
other bits and pieces so often decorating these underground
fantasies—the Slaves joined the collage that so resembles a
three-dimensional grotesque.3
The eighteenth century continued to parallel the sketch with
the ruin, but did not limit the effect of spontaneity to the sketch. Wind
illustrates this change in attitude by summarizing a section of the
Reverend William Gilpin’s Essay on Picturesque Beauty: “Works of
art decomposed by Nature resemble those left unfinished by the hand
of man: in both cases irregular and accidental shapes convey a sense
of spontaneity.”4 Further, the literary sketch was equated with the
1
David Rosand, “Composition/Decomposition/Recomposition: Notes on the
Fragmentary and Artistic Process,” in Fragments: Incompletion and Discontinuity, ed.
Kritzman, 22, citing Pliny, Nat. Hist. 35.145.
2
Rosand, 20-22.
3
A photograph of the grotto with the Slaves is reproduced in Naomi Miller, Heavenly
Caves: Reflections on the Garden Grotto (New York: George Braziller, 1982), 8.
4
Wind, 40.
Chapter 4: The Sketch 119
visual sketch, as seen in Samuel Johnson’s remarks on the unfinished

manuscripts of the poet Edmond Smith:
… if some of them were to come abroad, they might be as highly

valued as the sketches of Julio and Titian are by the painters;
though there is nothing in them but a few outlines, as to the design
and proportion.5
Harries reports Diderot’s summary of many

eighteenth-century associations with the sketch:
Le mouvement, l’action, la passion même sont indiqué par

quelques traits caractéristiques, et mon imagination fait le reste. Je
suis inspiré par le souffle divin de l’artiste, Agnosco veteris
vestigia flammae [An IV, 23]; c’est un mot qui réveille en moi une
grande pensée. Dans les transport violens de la passion, l’homme
supprime les liaisons, commence une phrase sans la finir, laisse
échapper un mot, pousse un cri et se tait; cependant j’ai tout
entendu; c’est l’esquisse d’un discours. La passion ne fait que des
esquisses. Que fait donc un poëte qui finit tout? Il turne le dos à la
nature.
[Movement, action, passion itself are indicated by a few

characteristic lines, and my imagination does the rest. I am
inspired by the divine breath of the artist. … In the violent throes
of passion, one suppresses connections, begins a phrase without
finishing it, lets a word escape, cries out and falls silent; and yet I
have understood everything; it is the sketch of a discourse.
Passion creates only sketches. What then does a poet do who
finishes everything? He turns his back on nature.]6
These are the attitudes toward the sketch that the nineteenth century
inherited: the sketch stimulates the viewer’s imagination and allows
the viewer to identify with the creative process; it communicates as
much as or more than the finished work; it finds its form in language
as well as the visual arts; and it allies its maker with the ongoing
process of creation in nature.
The meaning and description of finish in a work changed from
period to period. “Unfinished” can only be defined as less than the
5
Harries, 58.
6
French and English translation, Harries, 106-107, from Diderot, Salons, ed. Jean
Seznec and Jean Adhémar (Oxford: Clarendon 1963), 3:248.
acceptable level of “finished,” and artists frequently incorporated into

their works some element of what was considered, in their time, an
unfinished texture.7 An appreciation for the unfinished increased in the
eighteenth century, but the majority of works labeled sketches were
still truly preliminary, meant to serve the creation of a finished work,
whether or not that work was ever brought to completion. Sketches
were a private matter: they retained the most personal elements of the
artist’s working method—his preliminary gestures, unrefined
compositional design, rough brush strokes; they resided most properly
in the artist’s sketchbooks and in his studio, viewed by only a select
few friends and connoisseurs, and were not intended for the general
public.
Sketching was also the working method of the trained
professional. In France, sketches were part of the training of artists in
the Academy. They were considered necessary preliminary steps,
classified and labeled according to their phase in the sequence toward
the properly finished work: esquisse, esquisse peinte, étude, ébauche
among others. The Academy was concerned with both appropriateness
of subject matter and the finish that erased rough edges, broken lines,
brush strokes—all evidence of the sketch. Finish guaranteed the
amount and quality of work that had gone into the painting.8
In the nineteenth century, as the romantic ideal of the
incomplete spread, the sketch moved from these private spaces to the
public arena. Measured against the Academy’s officially defined
finish, artists began to present to the public works that could be
defined only as sketches, not just in the area of surface finish but also
composition (balance, proportion) and subject matter. Constable and
others of his time began making sketches that were not intended for
any particular finished work, but were ends in themselves. Sketches
were bought and sold, moved from the studio to the walls of homes,
and eventually to the walls of galleries.9 Once the training tool of the
Academy, they became a protest against its rigor and its control over
7
Rosand, 21, lists Leonardo, Bosch, Giorgione, and the Housebook Master as
Renaissance examples.
8
Rosen and Zerner, 226-7.
9
Wendelin Guentner, “The Inscription of the Sketch in the 19th-Century French
Journal,” Nineteenth-Century French Studies 27 (1999): 281-283.
the success of artists in Paris, and also a protest against the bourgeois
demand for this Academy-defined precision in finish as proof of value
for the dollar.10
The process of sketching itself moved from the domain of the
trained artist to that of the well-to-do amateur. Rather than buying
paintings to document their travels, the wealthy now sketched a
record—a predecessor of the photo album; their sketches were visible
evidence that they had enough money and leisure time not only to
travel but also to learn to draw. The tool of the professional became
the pastime of the wealthy. Sketches by professionals were also in
demand, but not those preliminary sketches valued for the glimpse of
creativity and genius they were thought to offer; instead, the buying
public wanted the same quick studies of scenes that they were
themselves attempting—the “found” sketch rather than the sketch
“made” by the inspiration and imagination of the artist.11 This type of
sketch became a commodity, and its dependence on the taste of the
public lowered its prestige. It was never judged by the same criteria as
the completed work because it was never placed in the same category.
Partly as a result of this shift, the sketch-artist, especially the
sketch-artist trying to make a living while establishing himself as a
reputable serious artist, began to present himself as a sort of dilettante,
a flâneur, strolling about wherever chance happened to lead him,
sketching whatever sight happened to please him, selling the work to
whomever it happened to please. His sketches were works that he
tossed off; they were not examples of what he could do if he painted
as a professional. They showed his skill, but not his imagination and
genius: they were fragmentary.12
10
Rosen and Zerner, 226-7.
11
Richard Shiff, Cézanne and the End of Impressionism: A Study of the Theory,
Technique, and Critical Evaluation of Modern Art (Chicago: University of Chicago
Press, 1984), 74. Shiff treats the theories of “making” and “finding,” and explains the
distinction between two types of sketch identified by the Academy: the esquisse was a
compositional sketch, showing the arrangement of the forms, and grew out of
inspiration and imagination; the étude was quick study that grew directly from nature,
was more passive, less consciously creative. Chapter 7, especially 74-75.
12
Alison Byerly, “Effortless Art: The Sketch in Nineteenth-Century Painting and
Literature,” in Criticism: A Quarterly for Literature and the Arts 41 (1999): 350-52.
As journals and magazines grew in popularity and sales,

writers (again, especially those attempting to build a reputation) were
often forced to write short articles for publication in these venues in
order to make a living. They relied on the interest of the buying public
just like the sketch-artists.13 Writers (Irving, Dickens, Thackeray are
examples given by Alison Byerly) often labeled such pieces
“sketches,” and often included many of the techniques used by earlier
fragment writers: the mutilated document as a prop, the midsentence
ending, dots of ellipsis, and parenthetical expressions. Also like the
sketch-artists, these writers masked a reliance on the taste of the public
behind the impression of a work tossed off, the situation happened
upon and reported, the “found” rather than the planned. The title
“sketch” alleviated the burden of being judged by the same standards
as the finished work; it excused an informal style and a lack of what
would have been considered professional quality and finish.14
Debussy’s Sketches
Debussy left important sketches for some major completed

works (Ibéria, La Mer, Pelléas et Mélisande), for works he planned
but never finished (Le Diable dans le beffroi, La Chute de la maison
Usher, Rodrigue et Chimène), as well as other bits and pieces in
assorted sketchbooks. For the most part, during his life, his sketches
remained in his possession or among his private circle of friends and
supporters, just as had the sketches of artists from the Renaissance
through the eighteenth century. Sometimes, at small gatherings, he
would improvise (a practice often equated with sketching) or play a
work in progress. He also gave away autograph copies of his scores
(Gabrielle Dupont, his first long-term mistress, received the autograph
of Prélude à “l’Après-midi d’un faune”). His sketches were, for the
most part, very much a private matter.
To what extent Debussy bought into the aesthetic of the sketch
is uncertain. Howat notes that Debussy published each prelude of
13
The sketch entered the public world again as lithographs began to be published in
newspapers and books. Guentner, 283.
14
Byerly, 352-53.
Book II, either in part or completely, on three staves, and that in the
manuscript of the first three of these preludes, he included four staves
per system, though he never used more than three at a time. He
suggests that Debussy may have intended to “give the impression of
an orchestral sketch in short scores” rather than finished works.15 Such
sketches were common intermediate steps for composers working out
melodic and harmonic material before writing out a complete
orchestration.
The word “sketch” appears in the titles of two of Debussy’s
published compositions, D’un cahier d’esquisses and La Mer: trois
esquisses symphoniques;16 moreover, the tiny piano work Page
d’album could refer to a sketch book, though “album” more often
means a book used to collect autographs and other memorabilia.17
Debussy’s most truly sketch-like works, D’un cahier
d’esquisses and the tiny Morceau de concours, are among the few
pieces Debussy published in magazines. Both retain close ties (they
show similar or identical material) to Debussy’s sketches for other
works and may call to mind certain connections to the changing role
of the sketch in nineteenth century.
15
Howat, foreword to Préludes, Livre I, Livre II, ed. Roy Howat with the
collaboration of Claude Helffer, Oeuvres Complètes de Claude Debussy, Série I, vol.
5 (Paris: Durand-Costallat, 1995), 12m(French), 16 (English). The four-staff
manuscript layout of Prelude no. 1, “Brouillards,” can be seen in facsimile on page
182.
16
Debussy may have borrowed the subtitle for La Mer from Paul Gilson, who wrote
an earlier symphony also titled La Mer and subtitled “ esquisses symphoniques.”
Simon Trezise, Debussy: “La Mer” (Cambridge: Cambridge University Press, 1994),
33-35.
17
Artists and musicians would add a drawing or a composition to the album of a
friend or someone who was wealthy and influential. Though many such entries are
sketches, some are complete works. Maria Szymanowska (1789-1831), Polish pianist
and composer, owned several albums containing sketches, short writings, and
compositions of many artists of her time. Album Musical: Maria Szymanowska, trans.
Renata Suchowiejko (Krakow: Musica Iagellonica, 1999). Debussy wrote several little
musical notes to his second wife, Emma Bardac, but according to Robert Orledge,
none of these appear in published works. There is also no evidence to support
Debussy’s claim that the music for his children’s ballet, La Boîte à joujoux, was based
on music he had written for Christmas and New Year albums for children. Robert
Orledge, Debussy and the Theatre, 182; Robert Orledge, “Debussy’s Musical Gifts to
Emma Bardac” Musical Quarterly, 62/4 (1976): 536-553.
D’un cahier d’esquisses
In 1904 Debussy published the piano piece D’un Cahier

d’esquisses in Paris illustré. That same year, Roland-Manuel quoted
Ricardo Viñes during a conversation stimulated by Viñes’s having
either played, or listened to Debussy play, D’un cahier d’esquisses:
Debussy had declared his dream of composing music whose form

was so free as to seem improvised; to produce works that sounded
as if torn from the pages of a sketchbook [arrachées aux pages
d’un cahier d’esquisses].18
The circumstances of the conversation are questionable. As

Howat has shown, dates and entries in the Viñes diaries offer no
corroboration and, in fact, indicate that the comment, if it was made,
might have been made about another work.19 Even so, both the
comment itself and Debussy’s title echo the nineteenth-century
fascination with the fragment in the form of the unfinished sketch.
Debussy, speaking through M. Croche, does make one documented
statement that indicates he did associate the musical sketch with
freedom from traditional formal structures, with imagination, and with
rapid changes of idea:
Certainly Chopin’s nervous disposition let him down when it

came to the endurance required in composing a sonata. But he did
make some finely wrought “sketches,” and it is at least agreed that
he invented his own way of handling the form, not to mention the
marvelous music he achieved in doing so. He was a man with
abundant imagination, and would flit from one idea to another
without demanding a one hundred percent commission on the
transaction—which is what some of our more celebrated masters
do.20
Howat believes that D’un cahier d’esquisses was originally

written as the second of a set of three piano works titled “Masques,”
18
Roy Howat, “En route for L’île joyeuse: The Restoration of a Triptych,” Cahiers
Debussy 19 (1995): 42. Howat quotes Roland-Manuel, Á la gloire de Ravel (Paris,
1938), 65.
19
Howat, “En route,” 42.
20
Debussy on Music, 47.
“Deuxième sarabande,” and “L’Île joyeuse”21 intended to be published

as Suite bergamasque. This suite never materialized: Masques and
L’Isle joyeuse were each published separately, the “Deuxième
sarabande” disappeared, and Debussy used the title Suite bergamasque
for a set for piano written in 1890 and published in 1905.
Although Debussy’s use of self-quotation and borrowing is
well documented (see Chapter 5), D’un cahier d’esquisses is perhaps
the most extreme example. It was written during 1903-4, when
Debussy was also working on major works such as Estampes, La Mer,
the Images for piano (first series), and the unfinished Le Diable dans
le beffroi, as well as Masques and L’Isle joyeuse.22 As Howat has
shown, D’un cahier d’esquisses borrows heavily from many of these
pieces. Fig. 4.1 lists the quotations identified by Howat, with two
additions (d, f). Fig. 4.2 places those themes in possible formal
structures.
Fig. 4.1. Borrowings in D’un cahiers d’esquisses.
a = Masques, m. 3 e = L’Isle joyeuse, m. 67

b = La Mer, mvt. 1, after rehearsal 9 f = Reflets dans l’eau, m. 1
c = L’Isle joyeuse, m. 79 g = L’Isle joyeuse, m. 219
d = La Mer, mvt. 3, Chorale h = Masques, coda, 362
Fig. 4.2. Possible formal structures of D’un cahier d’esquisses.
Sections I II III IV
measures 1-10 11-28. 29-44 45-53
Form
Howat Intro A A1 Coda
Alternate A B B1 A coda
Themes a b c d trans e f— c+ d+ g-cad a h e g
measures 1 6 11 13 20 29 38 43 45 47 50
Harmony ii Ger6 I ii vi—vi/wtV-wtV I ii vi Ger6 ii V I IV I
I
21
Howat writes L’île [sic] joyeuse, but notes later that Debussy changed the spelling
to L’Isle, alluding to his stay on the English island with Emma. Howat, “En route,”
38.
22
Howat, “En route,” 48, notes a reference to a now lost autograph dated January
1904.
Almost every measure of the piece contains borrowed material

of one sort or another. Howat uses the copious borrowings from
Masques and L’Isle joyeuse as evidence that D’un cahier d’esquisses
is the missing middle work that, situated between Masques and L’Isle
joyeuse , would form a triptych. Certainly this is a provocative idea,
though Debussy’s handling of motives in these three pieces is not
typical of that found in his other cyclic works and Howat offers no
explanation for the separate publication of the three pieces. Another
explanation is that the piece was, as its title suggests, literally lifted
from the pages of a sketchbook. Roland-Manuel’s account (given
above) would support this view, as would the make-up of a
well-known sketchbook from the years just after 1903-4 that contains
a conglomeration of jottings from many pieces Debussy was writing at
the time. The so-called Images sketchbook23 from 1906-9 includes
sketches for Gigues, Rondes de printemps, and all three movements of
Ibéria; “Voiles,” “La Cathédrale engloutie,” and “La Fille aux
cheveux de lin” from Book I of the Preludes; an entry marked
“Angelus” and one marked “Bouddha”; as well as many unidentified
scribbles. With its collection of shared material, D’un cahier
d’esquisses suggests a parallel, fictional (so far as we know)
sketchbook, and its title is an accurate description of the content of the
work.
Debussy’s title raises issues that should be of interest to
analysts. By referring to the sketchbook, to a common source for ideas
found in more than one work, Debussy acknowledged the conscious
use of his technique of borrowing and reworking. This is a clear
indication that his reuse of material in other instances is deliberate and
has musical and extramusical implications that should not be ignored
in any study of his work. In D’un cahier d’esquisses, Debussy’s use of
fragments that appear in so many of his other pieces suggests the
sketchbook of the professional artist. At least since the Renaissance,
these sketchbooks contained much more than a series of scenes; they
included reproductions of famous works, bodies drawn in different
poses, objects drawn from different angles, composition sketches,
drawing exercises—perhaps ears, or hands, or feet—as well as the
23
Now in the Lehman Deposit at Pierpont Morgan Library in New York.
results of sudden inspiration or the quick study of the beautiful scene

that was happened upon.
Accepting D’un cahier d’esquisses as a piece made from the
fragments of a sketchbook may also offer a clue to the unusual
circumstances of its publication. According to Howat, in the eight
years prior, Debussy had published only two pieces in magazines—
two versions of his Sarabande, both meant to advertise the
forthcoming publication of the suites for which they formed the
middle movements.24 Howat believes the publication of D’un cahier
d’esquisses had a similar purpose, even though no advertisement
appeared with the piece. He suggests that Debussy was playing a little
joke on his followers by publishing the piece mysteriously; perhaps its
appearance under these circumstances would give the message that he
had a “work in progress” or that readers should “watch this space,” as
Howat puts it. The fact remains that no advertisement accompanied
D’un cahier d’esquisses and the piece never appeared as part of any
larger work—an unusual occurrence for Debussy’s piano works of this
time.25 Perhaps Debussy’s choice of title was related to his decision to
publish the work separately in a periodical, but not as an
advertisement for a set that would include it. Perhaps he chose this
title for the same reasons that the sketch-artists and journalists used
the word “sketch” in their titles: to present it as something tossed off,
to excuse it from being judged by the same standards as the other
works in sets, to justify its solo position by presenting it as a “found”
and not a planned work.
If Roland-Manuel and Viñes reported Debussy’s quotation
(“music whose form was so free as to seem improvised”) correctly,
and if it can be assumed to refer to this piece, then clearly Debussy
wanted to capture that sense of spontaneity that links the sketch and
24
The first suite for piano, titled Images, never appeared. It was published in 1977 by
Presser as Images (oubliées). The second was Pour le piano. Howat, “En route,” 49.
25
Howat, “En route,” especially 49.
the improvisation.26 Two facts of publication support this view. The

title as it appears in the original publication is preceded by three dots
of ellipsis, indicating something incomplete, and Henri Vanhulst
reports that this is the first work for piano that Debussy wrote (perhaps
more correctly, published) on three staves.27 “Howat has suggested
that Debussy published the Preludes, Book II, on three staves to
suggest orchestral score. (See further discussion in Chapter 6.) Yet
Debussy’s title suggests not an improvisation, for D’un cahier
d’esquisses implies that the bits and pieces had been jotted down in a
sketchbook beforehand, but a grotesque, a creature made of parts that
either do not necessarily belong together, or do not appear in their
correct relationship to each other.
The grotesque has always been closely linked to the sketch, to
the work that arises from imagination and inspiration—from the
working of the mind without interference of convention or reality.
Montaigne compared his Essais to the grotesque (see Chapter 1), those
fanciful decorative borders where vegetation merges with human and
animal parts to create composite creatures from the world of the
imagination—creatures made of fragments. The sixteenth century
allied the grotesque with imagination divorced from reason. Kayser
quotes Dürer: “If a person wants to create the stuff that dreams are
made of, let him freely mix all sorts of creatures.”28 Here Debussy has
freely mixed all sorts of parts, all sorts of creatures; they are
recognizable to anyone familiar with his music. Rosen says that
quotation is the memory made public; these quotations are both
Debussy’s memories and the listeners’ memories, and they constantly
26
Jeffrey Kallberg (“Chopin and the Aesthetic of the Sketch: A New Prelude in E flat
Minor?,” Early Music 29 [2001]: 409) offers this quotation from Eugène Delacroix’s
diary: “We spoke of Chopin. He told me that his improvisations were much bolder
than his finished compositions. It is the same, no doubt, with a sketch for a painting
compared to a finished painting … Perhaps there is less scope for the imagination in a
finished work than in a sketch.” Eugène Delacroix, Journal 1822-1863, ed. André
Joubin (Paris, 1931-32; rev. ed. 1980), 330; translation adapted by Kallberg from
Michele Hannoosh, Painting and the “Journal” of Eugène Delacroix (Princeton:
Princeton University Press, 1995), 72.
27
Henri Vanhulst, “L’éditeur bruxellois Schott frères et D’un cahier d’esquisses de
Claude Debussy,” Cahiers Debussy no. 26 (2002): 6; facsimile of first publication,
11-13.
28
Quote attributed to Dürer by Kayser, 22.
pull us out of D’un cahier d’esquisses and into the memory of those
quoted works. Its very fabric, the quotations, functions as the parts of
the grotesque, never allowing it to coalesce into a perfect whole.
The formal structure of D’un cahier d’esquisses also creates a
grotesque. This freedom to accept formal distortion and genre mix, to
confuse the essential with the decorative, the center with the frame,
creates the formal grotesque or arabesque. In D’un cahier d’esquisses,
the concept of the grotesque again sheds light on Debussy’s
compositional practice. Fig. 4.2 shows the piece divided into four
main sections: mm. 1-10, 11-28, 29-44, and 45-63. According to
Howat these sections can be labeled as introduction, A, A1, and coda.
Under this scheme the main motive of the piece (m. 11) is taken from
L’Isle joyeuse. The opening motive from Masques is given a
subordinate role. My alternative reading (fig. 4.2, “alternate”) suggests
that the situation is not quite so clear-cut. First, Howat’s introduction
and coda together make up more than half as much material (18
measures) as the main part of the piece (32 measures). Second,
Howat’s designation coda devalues the formal function of the opening
motive and its return near the end. Third, although the tonic harmony
appears at the beginning of the last section (m. 45), the decisive
cadence in D-flat major, with its leading tone resolution, does not
occur until m. 49. Fourth, the final allusion to L’Isle joyeuse and the
strong movement to the subdominant in m. 50 are perhaps classic
signs for the start of a coda.
What this alternate reading underscores is that the formal
function of the four sections is anything but obvious. As we might
expect from a grotesque, Debussy seems to blur the distinction
between the essential and the decorative aspects of a composition. It is
simply not obvious whether Howat’s introduction and coda frame the
movement or whether they are an integral part of it, as in my
alternative. The elaborate network of tonal and thematic
cross-references between the sections only serves to blur their
functions even more.
D’un cahier d’esquisses satisfies two essential criteria of the
grotesque: its form clearly distorts expectations of formal structure;
and thematically, rhythmically, and even harmonically it is a creature
made of parts. Whether they were torn from an existing, now lost,
sketchbook remains a mystery, but for the listener familiar with
Debussy’s music today, D’un cahier d’esquisses must bring to mind

all those pieces it either quotes or that grew from common material. If
this grotesque is subtle, then, as Hugo says of the grotesques of the
primitives and ancients, some were “barely deformed.”29
Morceau de concours
In January of 1905, Debussy allowed another work to be

published in a magazine: the Paris periodical Musica included a very
short work for solo piano, now titled Morceau de concours, as one of
six anonymous pieces, given without title or other information that
might reveal any of the six composers’ identities, used in a “name that
composer” competition.30 As many other Debussy scholars have
noted, mm. 1-7 and 19-22 of that work duplicate two segments of the
only known surviving sketch for Debussy’s unfinished opera Le
Diable dans le beffroi (based on Edgar Allan Poe’s short story “The
Devil in the Belfry”) dated on the first page of the sketchbook as
August, 1903.31 A comparison of Morceau de concours with the
sketch for “Le Diable dans le beffroi” shows that Debussy copied two
passages from the sketchbook exactly and in sequence, the last fact
obscured by Lockspeiser’s publication of the pages out of order.32
Separating these two sections are ten measures that consist of a single
bass line and two snippets of the opening theme (suggesting a repeat
29
Hugo, Cromwell, 71.
30
The competition pieces were published in the January 1905 issue, #28. The winner
and statistics were published in the April, 1905 issue, #31. The six composers whose
works were included were Gaston Serpette, Camille Saint-Saëns, Cécile Chaminade,
Jules Massenet, Rudolphe Berger, and Debussy. Of 530 entrants in the competition,
158 did not correctly identify the composer of any work, 194 identified one composer,
103 identified 2, 57 identified 3, 9 identified 4. Only Madame Maillard, professor of
piano and solfège (48, boulevard Rochechouart, Paris) identified all six correctly,
winning a piano. Massenet received 304 correct identifications, Debussy 218,
Chaminade 150, Saint-Saëns 146, Berger 80, and Serpette 6. See Musica, no. 31
(April 1905): 63.
31
The sketchbook, now in the Bibliotèque nationale, Paris, (B.n. MS 20634), contains
Debussy’s score and scenario.
32
Between the two, there is one minor difference in m. 2, but the sketch is difficult to
read.
and expansion of the opening) followed by a sequence consisting of

parallel major triads moving by semitones, tones, and augmented
seconds. For publication, Debussy replaced this passage with a
different sequence constructed of newly-composed materials drawn
first from one, then from the other whole-tone scale, and added a coda.
The sketch and Morceau de concours are compared in Fig. 4.3;
identical passages are highlighted in grey.
Fig. 4.3. Chart comparing measures in Morceau de concours with “Le

Diable dans le beffroi” sketch.
Morceau 1-8 9-10 11-18: 19-22: 23-27:

(m. 8 new material based on sequence sequence coda
duplicates opening motive statement repetition
m. 7)
I V I—WT I I
Diable [1-7] 8-17: 18-25: 26-29: no more
p 1, last repetition of opening sequence sequence material
system theme in bass clef with statement repetition
added accompaniment
I subdom I I
Morceau de concours then consists simply of two sections of

Le Diable dans le beffroi extracted separately and reattached via a
whole-tone sequence. Unlike the identifiable fragments of the
decorative grotesque that become part of a fabric made of many parts,
or the fragment that intrudes on a work in process, these fragments
neither interrupt nor become absorbed into another texture. Rather, the
fragments themselves become the subject of the work.33
Given the subject matter of “The Devil in the Belfry,” some
connection to the grotesque, both as compositional technique and
33
Although the effect would be lost to listeners, since the music of the sketch would
be unknown to them, the process is reminiscent of one of Géricault’s studies of
severed limbs. In some of these, Géricault arranged mutilated body parts with great
care into a balanced composition, painted slightly larger than life on a large canvas,
with careful attention given to lighting and surface finish. With the exception of
subject matter, they meet the requirements of the finished work. Rosen and Zerner
refer to them as gruesome puns on the idea of the still life. See Rosen and Zerner,
Plate 4.1 for a reproduction, and 46-48 for more information on Géricault.
subject matter, is not unwarranted. Debussy’s description in the

scenario he prepared for the opera, which is also contained in the
sketchbook with the music, describes the devil: “Il est vêtu d’un strict
habit noir et ne ressemble au Diable que par l’éclat de son regard et
par la structure des mains.” [He is dressed in a severe black suit of
clothes and he resembles the devil only through the flash of his eyes
and the structure (shape) of his hands.] He begins his disruption of the
daily routine of a well-ordered Dutch village by making the clock
strike thirteen at noon.
Debussy’s scenario is a rather bizarre juxtaposition of two
different scenes. The first scene takes place in Poe’s controlled little
Dutch village, Vondervattimitis, and it closes with the devil playing a
violon poche, leading the inhabitants away in a “gigue fantastique”
towards the canal—rather like a mix of the “Pied Piper of Hamelin”
and the Totentanz.34 The second scene, however, has nothing to do
with Poe’s story. The curtain rises to a “décor … complètement
changé,” as Debussy writes in his sketchbook; the village by the canal
in Holland now resembles a village on a river in Italy. Everyone
dresses and behaves in a manner quite different from their former lives
(the men wear their hats sideways and the women their blouses open).
Debussy then manipulates an ending that completely disarms the
perversity of Poe’s tale: through the power of prayer and true love, the
devil is undone.
Debussy had worked on his Poe operas for many years; it is
likely that he would not have spent undue time and effort on the
project for Musica, and it seems quite sensible that he would have
pulled out any sketch that could have been made ready for publication
quickly. Was his choice of material from the sketchbook, or his choice
of the new sequence material added to glue these sketchbook sections
together, influenced by any desire to hide or to reveal his identity
through this music? Did Debussy create what Gérard Genette terms a
self-pastiche, an intentional imitation of his own style, as Verlaine did
in his own “À la manière de Paul Verlaine,” which imitates both his
34
Similarities to Stravinsky’s L’Histoire du soldat are interesting. As mentioned
above, the earliest sketches for Le Diable date from 1903, well before Debussy met
Stravinsky, but Debussy’s letters show that he continued to work on both Poe projects
well after he did know Stravinsky.
style and the themes that appear frequently in his work?35 While this
seems the perfect opportunity for Debussy to have created a similar
work, Morceau de concours does not have the atmospheric, subtle
quality Debussy may have been best known for. (The work is more in
the style of “Minstrels” from Preludes, Book I, which was published
in 1910.) The opening measures certainly stress the tritone, perhaps
identified with Debussy’s music, but the new sequence, added for the
magazine publication, seems the most obvious clue since it is made of
whole-tone material. Was Debussy, contrary to the idea of
self-pastiche, trying out what he considered a new style, as a letter to
André Messager from September 1903 indicates?
Quant aux personnes qui me font l’amitié d’espérer que je ne

pourrai jamais sortir de Pelléas, elles se bouchent l’ oeil avec soin
… Il est probable du reste, que les mêmes personnes trouveront
scandaleux d’avoir abandonné l’ombre de Mélisande pour
l’ironique pirouette du Diable …
[Those people who are kind enough to expect me never to

abandon the style of Pelléas are well and truly sticking their finger
in their eye. … Quite likely, the same people will find it
scandalous that I should have abandoned Mélisande’s shadows for
the Devil’s ironical pirouette … ]36
Whatever Debussy’s intentions, considerably fewer than half (218 of

the 530 entrants) correctly identified the composer of Pelléas et
Mélisande.37
35
Gérard Genette, Palimpsests: Literature in the Second Degree, trans. Channa
Newman and Claude Doubinsky (Lincoln: University of Nebraska Press, 1997), 124.
36
Debussy, Correspondance 1884-1918, 186; Debussy Letters, 141.
37
See n. 329, above.
CHAPTER 5
AUTO-QUOTATION
They have made the Chimaera out to be a

three-formed beast: a lion’s features, the
hinder parts of a dragon, and the middle parts
of a goat. Certain naturalists say that this was
not an animal, but a mountain in Cilicia,
nursing in some places lions and goats,
burning in some and in other places full of
serpents. Bellerophontes made this place
inhabitable, whence he is said to have killed
the Chimaera.
The use of quotation and related procedures—citation,

allusion, borrowing—is common in Western European art forms, even
in the nineteenth century when theories of genius and originality
moved to the forefront of criticism and artistic endeavor. The creators
of fragmented works often make blatant use of quotation: Rabelais’s
examples range from the learned texts of the ancients to street songs;
Montaigne’s Essais are filled with quotations used to lend credibility
to a point he wishes to make or to demonstrate the breadth of his
knowledge; Petrarch quotes other poets (Arnaut Daniel, Cavalcanti,
Dante, among others) as well as proverbs. Gordon notes that “Nerval’s
intertextual borrowings from the Renaissance and the seventeenth and
eighteenth centuries are so excessive that they create an utterly novel
approach to textuality.”1 From the voice of authority to the voice of
the ridiculous, quotation is used for its power to create relationships
between the world within the work at hand and the world outside.
With only a minimum of material, quotation forces comparisons and
builds tension as the reader questions the degree to which the
borrowed or quoted material can or should be assimilated.2 Quotation
has the power to disintegrate, to pull the audience out of the quoting
work and into the quoted work, and thus to make fragments of both
texts, each seemingly incomplete without information transferred from
the other, each seemingly richer with it.
1
Gordon, 20.
2
Herman Meyer, The Poetics of Quotation in the European Novel, trans. Theodore
Ziolkowski and Yetta Ziolkowski (Princeton: Princeton University Press, 1968), 6,
discusses the tension of assimilation.
Debussy was a master of quotation and allusion. His letters

and writings are filled with quotations—from poems he has read or
set, from the text of Pelléas et Mélisande, from his own previous
writings. Laforgue and Baudelaire are among his favorites, but no one
he has read or heard is safe from his pen as he quotes and (purposely)
misquotes, consistently, across the pages and through the years.
Debussy’s music is no different. From some of the earliest studies of
his compositions to the latest, scholars have believed that the
identification of borrowed tunes—even borrowed chords and chord
progressions—is integral to understanding certain aspects of his
compositional procedure as well as to ascertaining extra-musical
meaning.3 There is no question that Debussy was a magpie, lining his
nest with all the shiny objects that caught his eye or entertained his
ear, using them to his own advantage, constructing for them new
surroundings, new contexts, new functions.
Both Daverio and Rosen include the use of quotation in their
discussions of music and the romantic fragment, particularly with
regard to the compositions of Robert Schumann. Few composers, as
Daverio points out, have so systematically and successfully employed
allusion in music.4 Among the various categories of quotation,
Daverio includes examples that add a historicizing dimension: if the
listener associates a quotation with a particular historical event or era,
the juxtaposition of past and present increases the awareness of the
distance between the two. Examples include the “Marseillaise” in
Faschingsschwank aus Wien and Schubert’s “Gretchen am Spinnrade”
in No. 2 of the Intermezzi. But Debussy’s collection of quotations
must surely compete with Schumann’s, if not surpass it. Daverio’s list
calls to mind Debussy’s quotation of the “Marseillaise” at the end of
“Feux d’artifice” and the subtle quotation of the alto line of the
Brahms Waltz Op. 39, #15, in “Les Fées sont d’exquises danseuses”
(both in Book II of Debussy’s Preludes for piano). Debussy is no snob
when it comes to musical quotations: popular songs and children’s
3
Vallas is an early writer who identifies many of Debussy’s quotations and
self-quotations. There are several studies claiming influence on Debussy’s style based
on the accumulated number of quotations or references from other composers, for
example, Robin Holloway, Debussy and Wagner (London: Eulenburg Books, 1979).
4
Daverio, 59.
Chapter 5: Auto-Quotation 137
song (“Nous n’irons plus au bois,” “Do l’enfant, do,” “Camptown

Races”) find a setting as easily as Brahms or Wagner.
Quotation often plays a role in parody, satire, and caricature.
Debussy’s pieces inspired by toys provide outlets for humor—whether
sarcastic, simply light-hearted, or tinged with nostalgia. If the
grotesque is made of the juxtaposition of incongruous parts, then the
ballet for children, La Boîte à joujoux, certainly qualifies. It is filled
with quotations and references, from Gounod’s Faust to moments that
allude to Stravinsky, from quotations of Debussy’s own published
works including “Jimbo’s Lullaby” and “The Little Nigar,” to
previously unused material from his sketch books.5 “Golliwog’s
Cakewalk” (from the Children’s Corner) forms a double grotesque—a
caricature of a caricature—since the cakewalk Debussy spoofed
originated as an African-American parody of white dancers. The
juxtaposition of the Golliwog’s dance, forming the outer sections of
the work, with the satirical quotation of the famous Tristan chord in
the style of a “music-hall ballad,”6 in the middle section, is
purposefully awkward, and that awkwardness is heightened and
reflected by the abrupt “cakewalk” interruptions in the Tristan ballad.
The audience recognizes a grotesque combination of styles, even
before it recognizes just how out of place Tristan is when paired with
the Golliwog or placed in the new harmonic setting Debussy provided.
Even the cover of the Children’s Corner, the work in which
“Golliwog’s Cakewalk” appears and which Debussy designed, is a
visual quotation and a bizarre little grotesque. The disembodied head
of the Golliwog, drawn by Debussy as a balloon tethered by a string,
floating over the elephant Jimbo, resembles several of Odilon Redon’s
disembodied heads—drawings of flowers with faces, many tied to the
ground by a thin stem that seems incapable of supporting their
weight.7 Hugo said of the grotesque, “Il y est partout; d’une part, il
crée le difforme et l’horrible; de l’autre, le comique et le bouffon.”8 [It
5
Orledge, Debussy and the Theatre, 182-85.
6
Roberts, 216.
7
Linda P. Cummins, “Debussy and the Grotesque,” Annual Meeting, American
Musicology Society, Southern Chapter, University of Alabama, February 1997;
Margaret Cobb, “Debussy and Le Roman de Rosette,” Cahiers Debussy 22 (1998): 75-
87. A striking example, A Flower with a Child’s Face, is reproduced in Dorra, 50.
8
Hugo, preface to Cromwell, 71.
is everywhere; on the one hand it creates the deformed and the

horrible; on the other the comic and the buffoon.]
Auto-quotation or auto-citation can be a particularly effective
form of quotation and may serve as a more powerful way of signing
than a simple signature in the corner or name on the title page, for it
places the artist in the work itself, not peripheral to it. It can also
function in much the same way as the authorial intrusion: through a
reference to another of the artist’s own works, thus to the artist as
creator/maker, the fiction of the work at hand is destroyed. (This is a
story being told, not something happening now; this is a painting, not
a view out a window or a peek into a currently existing situation.)
Visual artists placed themselves in paintings in several ways. These
include not only the self portraits or examples of the “artist in his
studio” genre that make the artist the focal point, or the numerous
cases where the artist gives his face to a character in a painting, but
especially works in which the artist inserts his presence as an artist
but not as principal subject into the painting as a way of stating his
role in the act of its creation. Often this was achieved through the use
of mirrors or other reflecting devices, as in the tradition that Jan van
Eyck followed when he painted The Arnolfini Double Portrait.9 In
other instances, the artist makes his presence as creator known by
placing a copy of his own work in the painting. Matisse’s Large Red
Interior includes his own Pineapple hanging on the wall over the
table; Manet’s portrait of Émile Zola clearly shows his Olympia, or a
sketch for it, hanging in the upper right-hand corner.10 These paintings
within paintings serve to remind us, more clearly than any signature,
that we are viewing something made by the artist. In music the
recognition of a quotation from another of the composer’s works can
also make the listener aware that he is hearing something that has been
created, crafted, by the composer.
9
Victor I. Stoichita, The Self-Aware Image: An Insight into Early Modern
Meta-Painting, trans. Anne-Marie Glasheen (Cambridge: Cambridge University
Press, 1997), 192-95, 247-55. Velázquez turned this tradition on its head in Las
Meninas, where he appears as a central figure and the mirror reflects what is most
likely the subject of the painting itself.
10
The Matisse paintings are reproduced in Yves-Alain Bois, Matisse and Picasso
(Paris: Flammarion, 1998), 208-9, Manet’s portrait of Zola in Shiff, 31.
Chapter 4 investigated Debussy’s use of auto-quotation in the

context of the value the late nineteenth-century inherited and
continued to place on the sketch. This chapter will present examples
of Debussy’s use of auto-quotation to create a musical web of
allusions similar to that attributed to Schumann by Daverio—a web
that is literary in purpose and character, creating musical ideas that
appear in multiple compositions, much as some nineteenth-century
writers created characters who appeared as either major or minor
figures in multiple stories. Each work is complete within itself and yet
not complete since each quotation, each reappearing character, directs
the attention of the reader outside the work at hand.11 The effect is one
of fragmentation at the level of the individual work, but also of
connection, binding multiple works in a larger context. Debussy’s “La
Sérénade interrompue” from Preludes, Book I, with its borrowing
from “Le Matin d’un jour de fête” (the last movement of Ibéria) and
“Nuages,” the first of the orchestral Nocturnes, with its connection to
act 2, scene 3, of Pelléas are two of many possible examples.
Ibéria and “La Sérénade interrompue”
As mentioned in Chapter 3, Rosen stresses Schumann’s ability

to make a quotation sound like a memory, to mimic the imperfect
process of recollection by introducing first a short fragment followed
by a version longer, though also incomplete. His example is the
quotation from Papillons that interrupts “Florestan” in Carnaval.12
Papillons fragments its host—not simply by being included, but by
literally interrupting the piece in progress twice, introducing a new
tempo and sonority (Daverio also notes a change of key area, from G
minor to B-flat major), before it is absorbed into its new context.
Papillons fragments “Florestan,” but it also connects—not only two of
Schumann’s individual works, but two of his collections of
miniatures: Papillons and Carnaval.
In “La Sérénade interrompue” from the first book of Preludes,
Debussy also uses blatant auto-quotation, binding his serenade to one
11
Daverio, 59-61.
12
Rosen, 98-100.
of his most famous Spanish works, Ibéria.13 The serenade begins on

the dominant (of B-flat minor), “comme en préludant,” in the manner
of a guitarist warming up—typical gestures of the preludes as a genre:
they are short and improvisatory, and end in the key of the work that
follows, though they do not necessarily begin in that key. Eventually a
slow, plaintive melody begins, breaks off with a piano return to the
preluding, which is then suddenly and violently interrupted by a brief
two-and-one-half measures, très vif, forte, with a different key
signature (A minor). Soft preluding and a continuation of the serenade
dissolve in a melodic cadenza (beginning in m. 76), when suddenly,
softly, out of nowhere, lointain (m. 80 in Fig. 5.1), appear the first
measures of the third movement of Debussy’s Ibéria, “Le Matin d’un
jour de fête,” (marked “Dans un rythme de marche lointaine, alerte et
joyeuse” in the Ibéria score). The key of the interruption is D major,
not the E-flat of the original. The “guitar” then interrupts the Ibéria
fragment twice (m. 85-86 and 98-90 in Fig. 5.1), the second time
leading back to the serenade in the tonic.
Debussy, like Schumann, has interrupted himself, interrupted
his own serenade with another of his own works. Unlike Schumann’s
Papillons reference, neither this interruption from Ibéria nor the
earlier and more violent interruption (which does not seem to be a
quotation) is actually incorporated into the texture of the serenade:
they remain outside. Also unlike Schumann, who inserts the word
“(Papillon?)” in his score at the second appearance of the quotation,
Debussy does not sign his interruption with words. The connoisseur
will know; the cultivated audience such as Mallarmé desired, the
audience that is willing to work with the artist to decipher his
meaning, will discover this clue easily.
13
Both Ibéria and the first book of Preludes were first published in 1910. See Brown,
Ibéria, for complete sketch history of that work. Brown, 144, notes that, although the
autograph for “La Sérénade interrompue” is undated, Howat (introduction to
Preludes, Book I: The Autograph Score, iv) has speculated late 1909 or early 1910,
the time when Debussy was preparing the premiere of Ibéria.
Fig. 5.1 Mm 78-91 of La Sérénade interrompue.
Fig. 5.2. Reduction of last measures of the second movement and first
measures of the third movement of Ibéria
Ibéria, end of second movement

Fig. 5.2, cont.

Ibéria, beginning of third movement; return of second-
movement theme in m. 5
Debussy’s interruption quotes not only musical material, but

also compositional technique—the technique of interruption. As “Le
Matin d’un jour de fête,” the third movement of Ibéria, opens, the first
theme—the theme that interrupts “La Sérénade interrompue”—is itself
interrupted by material recalled from the end of the second movement,
“Les Parfums de la nuit.” (See Fig. 5.2) The enchaînez indication
between the movements and the alternation of leftover
second-movement thematic material with the new third-movement
material blurs the line between the two. With this subtle interweaving
of themes, Debussy creates the effect of a march, heard at a distance,
invading and interrupting the previous movement as the day of the
festival interrupts the previous night (though, formally, it is the second
movement that interrupts the third). In “La Sérénade interrompue,”
unlike Ibéria, the borrowed theme from the third movement is foiled:
it does not take over or even become absorbed; it simply disappears.
Its effect, however, is far from inconsequential, for through it Debussy
not only recalls his own motive, and the effect as well as the
compositional technique of interruption, but also his fascination with
Spanish music and the subcategory of his compositional styles that
fascination has created.14 Through this interpretation, we see Debussy,

like van Eyck in the mirror, at work.
Pelléas and Nuages
Debussy wove a subtle web of allusion, of meaning and extra-

musical representation—one that ultimately extended beyond his own
compositional output—when he reworked material from his opera
Pelléas et Mélisande as the opening of Nuages from the orchestral
Nocturnes.15 Fig. 5.3 shows the motive from act 2, scene 3, of the
opera (which I will call the “grotto motive”) followed by the theme in
the first measures of Nuages. Maeterlinck also wove such a web when
he transposed the brief grotto scene in Pelléas et Mélisande (act 2,
scene 3, in both play and opera) to the setting of the climax of his next
play Alladine et Palomides. Both artists took a brief, though key,
element from an earlier work, and highlighted it in a later one; in each
case, the expanded reappearance creates a sense of the fragmentary, a
sense that not all has been revealed in either work, that the role in the
one must surely inform the role in the other. In the Symbolist world of
both these artists, often a world of understatement and subtlety, an
inherent sense of incompletion both magnifies and is magnified by the
attraction of such reused fragments.
Scholars once believed that Debussy composed Nuages before
Pelléas. This view was based on two assumptions: that the orchestral
Nocturnes are derived from two aborted scores, the Trois Scènes au
Crépuscule from 1892 and the three Nocturnes for violin and
orchestra from 1894-6; and that having started Pelléas in 1893,
14
Arthur B. Wenk, Claude Debussy and Twentieth-Century Music (Boston: Twayne
Publishers, 1983), 105-9, describes Debussy’s Spanish style.
15
Pelléas and Nuages share many of Debussy’s compositional strategies, including
thematic transformation and techniques of fragmentation such as interpolation
(discussed in Chapter 3 in connection with Prélude à “l’Après-midi d’un faune”).
Major interpolations occur in Pelléas in act 2, scene 3, beginning at 469, and in
Nuages at measures 84 and 86.
Debussy completely revised it in 1898.16 A study of the extant

sketches gives reason for challenging this view. No musical
connections have ever been found between Nuages and the two earlier
works; on the contrary, the only extant sketch for the Trois Scènes au
Crépuscule (Paris, BN, Ms. 20632) has nothing to do with the
orchestral score,17 and though no sketches exist for the Nocturnes for
violin and orchestra, Debussy’s reference to completing “three other
Nocturnes” suggests that these were different pieces rather than
reworkings of the original Nocturnes.18 As for Pelléas et Mélisande,
the earliest known sketches for act 2, scene 3, from 1895 (Paris, BN,
Ms. 20631) contain the passage given in Figure 5.3. These sketches
are so advanced and written with such confidence that they may even
have been copied from prior sketches for the same piece. Based on
information available today, it seems that Debussy worked out the
theme in the opera, then reworked it for Nuages.
With this adjusted chronology, the motive in Pelléas may now
be seen to have musical and extra-musical implications for Nuages,
while the prominence given to the opening motive of Nuages may
encourage a closer look at its development and use in the opera. In the
same way, the importance Maeterlinck gives the grotto scene in the
later play Alladine and Palomides may lead the reader to return to the
grotto scene in Pelléas, to reinvestigate its importance in that story.
The fragment pulls apart, and at the same time seeks connections.
16
David Grayson, The Genesis of Debussy’s “Pelléas et Mélisande,” UMI Research
Press Studies in Musicology (Ann Arbor, Michigan: UMI Research Press, 1986)
includes dating of sketches for Pelléas et Mélisande.
17
Denis Herlin, “Trois Scènes au Crépuscule (1892-1893): Un Premier Projet des
Nocturnes?” Cahiers Debussy No. 21 (1997), 11-14 shows facsimiles of sketches.
18
Though Debussy has mentioned Nocturnes for violin and orchestra to his publisher
Georges Hartmann in letters of 31 December 1897, 14 July 1898, 16 September 1898,
1 January 1899, and 3 April 1899 (describing various states of completion), in a letter
dated 3 July 1899, he writes to Hartmann: “Je terminerai La Saulaie, puis trois autres
[emphasis mine] Nocturnes et les Nuits blanches.” [I will finish La Saulaie, then three
other Nocturnes and the Nuits blanches.] Debussy, Correspondance 1884-1918, 128,
134, 138, 141, 150-151.
Fig. 5.3. Comparison of act 2, scene 3, of Pelléas and the opening of

Nuages.
Pélleas et Mélisande, act 2, scene 3
Nuages, mm. 1-2
Alladine and Palomides, both promised to others, have fallen

in love and professed that love; as punishment, they have been bound,
blindfolded, and sealed in a grotto (act 4). Freeing themselves of their
bonds and blindfolds, they “see” in the darkness (with the “sight” of
the blind that Maeterlinck so values) a world they create within their
minds. The grotto is beautiful, with clear water and blue roses, and as
Palomides exclaims, “It is as if the sky had flowed all the way down
here.” Standing dangerously near the edge of a cliff, they embrace,
even as their rescuers arrive. Light streams in as the rocks at the
entrance are dislodged, but the light reveals reality—an ugly, dirty
cave, not a beautiful paradise; the lovers deliberately plunge into the
water. Perhaps they did not understand that they would be freed,
perhaps they chose the life of the mind over reality. They do not
drown, but never recover and shortly die of unknown causes.19
Pelléas and Mélisande come to their grotto simply so that she
will be able to describe it to Golaud, to further substantiate the lie she
has told in the previous scene: she lost her ring there. The lie arises
from guilt, from an awareness of her love for Pelléas. The scene is
short, little more than a tableau. At the entrance, they also wait in
darkness, wait for light so that they can see to enter. Foreshadowing
Palomides, Pelléas describes the grotto, “Quand on y allume une petite
lumière, on dirait que la voûte est couverte d’étoiles, comme le ciel.”
[If you light a candle inside the cave, you would say that the vault is
covered with stars, like the sky.] When the moon emerges from behind
a cloud and light floods the grotto (to a G-flat major arpeggio, one of
the tonalities of light in the opera), they see three starving beggars
inside, leaning against each other, asleep. A famine in the outside
world is as severe as the parallel spiritual famine of the characters in
the story. The light has again revealed the truth—the three beggars are
not three fates, or the future, but the present, the three main characters,
weak and leaning on each other in a precarious balancing act, their
destinies so intertwined that even a subtle change can spell disaster.
For Maeterlinck, the grotto represents the interior life, the life
of the mind. As Patrick McGuinnes states,
descents into subterranean worlds function as determined

“symbols” of a descent into the unconscious and as “correlatives,”
projections, and extensions, part of an overarching structural
system in which distinctions between inner and outer worlds are
continually blurred.20
Both Palomides and Pelléas have described the grotto of the

imagination; the light reveals the grotto of reality. The four characters
react differently to the distinction between mind and reality: Alladine
and Palomides turn away from the light, rejecting reality (even though
19
The similarities to Pelléas et Mélisande are clear; Bettina Knapp, Maurice
Maeterlinck (Boston: Twayne Publishers, 1975), 78, reports that Maeterlinck once
called Alladine et Palomides a “decoction of Pelléas.”
20
Patrick McGuinnes, Maurice Maeterlinck and the Making of Modern Theatre
(Oxford: Oxford University Press, 2000), 144.
it means rescue), choosing their inner dream even as death; Mélisande

and Pelléas turn back toward the light, choosing to face the reality that
the moonlit grotto revealed. Mélisande seems actively aware of this
decision: as they leave, she turns away from Pelléas and away from
their growing attachment, “Laissez-moi; je préfère marcher seule”
[Leave me alone, I prefer to walk alone]. If the destiny of both couples
is ultimately death, then Alladine and Palomides embraced that fate in
the grotto; Pélleas (perhaps passively) and Mélisande (actively)
rejected it.
Just as Maeterlinck’s grotto scene has implication for the rest
of Pelléas, Debussy’s “grotto” motive also appears in variant forms in
other parts of the opera; he uses it to create a subcycle within the
overall progression of the play. It is prominent in three scenes; all
occur outside at night and progress through increasingly more certain
premonitions of death. In act 1, scene 3, Pelléas and Mélisande look
out over the sea, contemplating the possibility that the ship leaving the
harbor will be wrecked by a storm. In act 2, scene 3, Pelléas and
Mélisande enter the grotto at night to find the three starving beggars
whose deaths are as real a possibility as they are a premonition of
future events. In act 4, scene 3, little Yniold hears sheep “pleurer”
[weeping] and then, after noting “Il n’y a plus de soleil” [There is no
more sunlight], he watches the herd being led past him; the shepherd
tells him the sheep are not on their way to the fold: the assumption is
slaughter. Allusions to the deaths of Pelléas and of Mélisande are
unmistakable. Mélisande believes the ship that may be wrecked is the
ship that brought her to Allemonde; the three starving paupers are
surely Pelléas, Mélisande, and Golaud; the sheep are led to slaughter
as fate leads the characters to the inevitable conclusion.21
Embedded in this progression of scenes is a series of delays.
Yniold’s scene, opening with his futile attempts to move a heavy rock
and recover his gold ball and ending with the sheep on their way to
slaughter, actually retards the progression of the drama itself. It is a
digression serving merely to emphasize the futility of man’s actions
21
There are other brief variants of the fragment motive in all the night scenes, but the
motive is conspicuously absent from the night scene where Golaud actually kills
Pelléas (with the exception of a brief variant in act 4, scene 4, just after Pelléas asks
Mélisande to come out of the light and into the shade), also from any indoor night
scenes, and from any day scenes, including the scene in the dark vaults.
and the inevitability of death. In the first two scenes, though, it is

Mélisande who delays, hesitating to initiate or acknowledge her
relationship with Pelléas, thus postponing the inevitable outcome. In
act 1, she simply refuses his hand offered in assistance; in act 2, she
turns away from him in the grotto. These hesitations forestall their
final meeting in the forest, act 4, scene 4, where she makes excuses for
being late—excuses that tell us that even at the end, she hesitated.
These hesitations seem on one level simply in keeping with
the passive Symbolist heroine, but this cycle Debussy creates with his
use of the “grotto” motive may show that he understood Mélisande as
a more active character, not one who is unable to defy destiny, but one
who acts to do so in the only ways available. Her hesitations are
actions, the act of turning away from Pelléas, and like the sheep scene,
they delay the final outcome. Debussy noted in a letter to Henri
Lerolle22 that act 2, scene 2, the scene immediately preceding the
grotto scene, marked a turning point in the opera. Mélisande lied to
Golaud about where and how she lost the ring he gave her. The lie is
blatant and unnecessary. A modified version of the truth—the ring
slipped off as she reached toward the water in the fountain in the
forest—would have sufficed, but her guilt fabricates a response that
places her physically far from the place where she willingly and
willfully played her dangerous game of toss with the ring. It now lies
beneath the water like her crown at the beginning of the opera—the
crown the unknown “he” had given her and that she no longer wanted.
The lie is an active, if still unrealized, acknowledgement of how little
she cares for Golaud and how guilty she feels about her feelings
toward Pelléas. Golaud has insisted she return to the grotto, even in
the dark, to look for the ring, and that she take Pelléas with her. The
grotto scene shows Mélisande now aware—if not completely aware of
her feelings, then completely aware of her situation. At that moment,
she is no longer the passive Symbolist heroine; she is active and her
actions defy destiny.
The adjusted chronology discussed at the beginning of this
chapter, placing the composition of act 2, scene 3, of Pelléas before
that of Nuages, directly effects the extramusical associations of the
grotto motive and its reworking in Nuages, as well as of other
22
Debussy, Correspondance 1884-1918, 17 August 1895, 111-112.
connected borrowings. In a published explanation of the titles used for

the Nocturnes, Vallas reports that Debussy wrote, “‘Nuages’ renders
the immutable aspect of the sky and the slow, solemn motion of the
clouds, fading away in grey tones lightly tinged with white.”23 In
Pelléas, the motive always appears with the night sky. It is not present
in any indoor scenes or day scenes, and it is not always connected to
clouds. One of the few moments when some variant of it is not present
in act 2, scene 3, occurs when Pelléas says “No stars are out tonight on
this side; Let us wait til the moon has broken through that large
cloud.” When Pelléas and Mélisande watch her ship leave the harbor,
there are not clouds, only low-lying mist; they do not fear for the ship
because a storm approaches (in fact, the sea is still and there is no
mention of clouds), but because there have been storms the previous
nights. There is no mention of clouds in act 4, scene 3, when Yniold
sees the sheep. Perhaps it is a subtle notion, but it seems clear that the
opening motive of Nuages was, for Debussy, the immutable aspect of
the sky, the dark background against which the clouds may pass, just
as the clouds, the ship, the sheep, all pass across the immutable aspect
of fate.
Debussy’s borrowing extends in both chronological
directions. Vallas (and others) have accused Debussy of taking “every
note” of the opening of Nuages from Mussorgsky’s song The Noisy
Day Has Sped its Flight from the Sunless cycle.24 Lockspeiser,
defending Debussy’s originality and being perhaps sensitive to other
than the top voice, responds, “As for Vallas’s contention, repeated
unceasingly in programme notes, that the indefinite, flowing theme of
Debussy’s Nuages derives note for note from Mussorgsky’s song, …
this, if still likely to be upheld, can only be put down to a peculiar
insensitiveness to concepts of melody that were utterly opposed.”25
Debussy openly admired Mussorgsky (especially Boris Godonov) and
clearly could have known the song—and there are similarities, though
there are also many differences. The matter should be opened to
23
Vallas, 112; French not available; source unattributed.
24
Vallas, 113.
25
Edward Lockspeiser, Debussy: His Life and Mind, 2nd ed. (Cambridge: Cambridge
University Press, 1978), I:53. The notion is still upheld, among others, by J. Peter
Burkholder, Donald J. Grout, and Claude V. Palisca, A History of Western Music, 7th
ed. (New York: W.W. Norton and Company, 2006), 665.
further examination by the possibility that Pelléas was at least

sketched before Nuages, and that there are fewer similarities between
the Mussorgsky song and the grotto motive than between the song and
Nuages. Note, though, that the title of the cycle is Sunless, and the
song concerns the end of day, not clouds.
A striking case of borrowing, this time from Debussy, is
Stravinsky’s reworking of the main theme from Nuages at the
beginning of his opera Le Rossignol. The similarity between the two
scores was noticed as early as 1933 by Vallas, and is so close that
Glenn Watkins has accused Stravinsky of “near plagiarism.”26
Watkins’ charge is not so clear-cut, however, since evidence supports
the idea that Stravinsky knew the scores of both Pelléas and Nuages
before he began Le Rossignol, and his opera actually opens at the pitch
of the motive in act 2, scene 3, of Pelléas. Borrowing, quotation, even
auto-quotation often raise complicated issues of timing, influence, and
sheer coincidence. All these connections, these “memories made
public”27 in bits and pieces, invite further investigation.
26
Watkins, Pyramids at the Louvre, 39.
27
Rosen, 111.
CHAPTER 6
PRELUDES: A POSTLUDE
Portents, portentum, are named from

portending, portendere, that is from “showing
in advance,” praeostendere; . . Monstrosities,
monstrum, are named from an “admonition,”
monitus, because they point out something by
signaling, or because they indicate what may
immediately appear.
Expectations determine our perception of the fragment work;

the paratext—that accumulation of material defined by Genette as the
productions that accompany the text: author, title, preface,
illustrations, epigraphs, dedications, genre indications1—determines
many of those expectations. The title often provides both the first
clues to the fragment text and the genre or other convention against
which the fragment will work: Petrarch’s Latin title for the
Canzoniere, Rerum vulgarium fragmenta, alerts the reader to expect
more than a simple lyric collection; Cervantes’ introduction to Don
Quixote prepares the reader for the satire of the chivalric romance that
follows; and Sterne’s The Life and Opinions of Tristram Shandy,
Gentleman, announces a narrative that the author then consistently
evades. Yet, as discussed in Chapter I, conceptions of a convention, of
a genre, can change over time, causing modern audiences to see
Petrarch’s fragments as less startling and Cervantes as more disjointed
than either might have been viewed in its own time. Extensions of
paratextual elements are also subject to transformation over time; in
particular the author’s name carries with it reputation, sex, and
historical period, and to the extent that these are known by the reader,
also determine expectations.
When Debussy published his Preludes for piano, with that title
and in groups of twelve (“The fateful number!” Vallas writes 2), he
drew consciously on the tradition of such collections published in key
order: Bach’s forty-eight Preludes and Fugues,3 but more directly on a
1
Gérard Genette, Paratexts: Thresholds of Interpretation, trans. Jane E. Lewin
(Cambridge: Cambridge University Press, 1997), passim.
2
Vallas, 210.
3
2x24 proceeding chromatically, with each major key followed by its parallel minor.
work that was also much influenced by Bach, Chopin’s twenty-four

Preludes, op. 28,4 a work Rosen indentifies as one of the musical
responses to the romantic fragment.5 These works are closed and
ordered collections: ordered because they follow a predetermined,
systematic succession of keys (a framing device); closed because they
present all the possible keys in that succession; collections because
key order was in no way teleological, serving merely as a sort of
directory for presenting and locating a prelude in a particular key, and
in no way affecting presentation in any performance venue.6
The most direct referent of Chopin’s paratext—his title as
well as number and order of individual members of the collection—
was the tradition of published preludes, some in collections in key
order, meant to teach or to aid those less adept at preluding (at
improvising an introduction to a longer work as a means of easing
both the performer and the listener into the major work that would
follow). In a very real sense, works of this genre were collections of
fragments, of works that were not intended to stand alone as complete,
but only intended to be completed by another work outside the
collection. Preludes were sometimes short, improvisatory (the written
preludes arose from a tradition of improvisation), sometimes
technically difficult (meant to warm up the fingers, try out the
keyboard, and show off a little), often open tonally (to lead more
effectively into the work that followed). Their fragmentation was an
element of their function, but was not necessarily, if ever, part of an
aesthetic statement of fragmentation on the part of the composer. Key
order was significant only in that it offered, in a systematic
presentation, all the choices necessary to precede a work in any key.
These collections were closed, and ordered by practicality.
4
These progress through the circle of fifths in the direction of increasing sharps, each
prelude in a major key followed by one in its relative minor.
5
Rosen, 82-83.
6
Rosen, 83, notes that such methods of organization that transcend performance were
not new to musicians and gives as an example the third part of Bach’s Clavierübung.
(chorale preludes arranged in order of the ordinary of the mass, opening with the St.
Anne’s Prelude and closing with that Fugue). Liszt’s Transcendental Etudes (first
version 1826, second 1837, final 1851) are another example of the many collections
ordered by key but never intended to be played as a cycle.
Chapter 6: Preludes: A Postlude 153
Chopin’s Preludes did not meet the expectations signaled by

his paratext: not only the expectations of genre and function
announced by the title, number, and order, but also the expectations of
his own reputation as well. Schumann evidently imagined that Chopin
would have written longer, more imposing preludes that would move
that genre from subsidiary to independent status, much as Chopin’s
Etudes had moved the exercise to the stage:
I confess I imagined them differently, and designed in the

grandest style, like his Etudes. Almost the opposite: they are
sketches, beginnings of Etudes, or so to speak, ruins, individual
eagle pinions, all disorder and wild confusion.7
Through their brevity, their sketchiness, their ruin, and their

confusion—exaggerated characteristics of the prelude—Chopin
created works that, as Kallberg notes, actually did transform the genre.
Kallberg offers evidence that Chopin did use them in performance as
preludes to other works, in the original function of the genre, but that
he often performed the Preludes as individual works or in small
groups as well.8 When Chopin altered their function, when he chose
not to use them simply to lead into another work, the preludes became
intentional fragments—intended to stand on their own, in their
incompleteness, hedgehogs, not just pieces of a puzzle waiting to be
put together—and the act of separation a conscious aesthetic statement
on Chopin’s part. Just as Petrarch omitted the connecting narrative of
the prosimetrum, creating the holes, the blank spaces between his
lyrics, through which a narrative is suggested by its absence but never
stated, so Chopin omitted the possibility of conclusion, leaving instead
the silence through which the fragment echoes its incompletion, never
resolved.
Today the Preludes suffer from the same problem that often
affects the interpretation of earlier fragment forms: the loss of
7
Schumann, Neue Zeitschrift für Musik 7 (22 December 1837), 200; reference and
translation from Kallberg, Chopin at the Boundaries, 146.
8
Kallberg, Chopin at the Boundaries, 150-152. Kallberg gives evidence of at least
one occasion when Chopin used Prelude no. 8 (F-sharp minor) as an introduction to
the F-sharp major Impromptu, op. 36. Chopin’s last work published under the title
Prelude is a substantial work (op. 45) and establishes the use of the title for what is,
basically, a new genre.
historical perspective. They are reinterpreted by a world that has lost

the practice of preluding in piano recitals, that has embraced the
performance of complete collections, regardless of original intent, that
has lost the distinction between publication format and performance
practice, and that has come to misinterpret the closed collection as a
cycle whose meaning depends on the inclusion of all members in
order. As David Ferris points out, in traditional modern analysis,
defining traits of the musical cycle include “coherent narrative and
immutable order.”9 A collection, on the other hand, has no such
requirements, even though individual members of the collection may
relate to one another, and to the collection as a whole, in a variety of
ways. Complete performances of Chopin’s opus, though they would
not have been an option in Chopin’s lifetime, are now considered
appropriate, desirable, even necessary, perhaps fueled by the notion
that the key sequence functions as a substitute narrative or requires a
cyclic performance. Such performances, as Kallberg has eloquently
and persuasively argued, deny the fragment quality of the individual
preludes, are contrary to Chopin’s own performance practice, and rely
for their justification on anachronistic notions of genre and prejudiced
methods of analysis.10
Debussy’s work also contradicted expectations raised by his
paratext. Though his Preludes are numbered, he abandoned any
discernible key order, thus creating a closed but unordered collection,
and one that is closed simply through association with Chopin and
other prelude collections—the number twelve having meaning only in
memory. Debussy also added titles, atypical of the prelude. These
titles he placed unusually at the ends of the works, preceded by three
dots of ellipsis and enclosed in parentheses.11 Though Debussy’s
Preludes, in their use of interpolations and quotations, off-tonic
beginnings, and incomplete endings, are fragmented—similar in this
9
Ferris, 3. Ferris questions the validity of this distinction in the nineteenth century.
10
Kallberg, Chopin at the Boundaries, 149.
11
Three dots of ellipsis follow the title “Soirée dans Grenade” from the Estampes and
precede D’un cahier d’esquisses (see Chapter 4).
respect to Chopin’s 12—the titles, coupled with the greater length and
more extensive formal structures, blurred the distinction between the
prelude and the character piece—a category into which other Debussy
piano works, including the Estampes and Images I and II, fall. Thus
Debussy’s paratext sends mixed signals. The Preludes are a grotesque,
a joining of genres.
Debussy’s abandonment of key order has triggered searches
for a substitute plan, one that would explain the order in which he
chose to present his Preludes for publication and perhaps reveal that
the collection is closed by more than mere reference to the number
twelve—that would, in fact, provide evidence to consider the Preludes
an ordered collection or even a cycle. Roy Howat has reported that
Debussy stressed certain “pivot notes”(Bb in the first part of Book I
and Db in the last part of Book II), and he has observed a shift of
emphasis from C to Db through Book II, but he is careful to say that
these are not key progressions.13 In addition, as has been the case with
12
Roy Howat has identified similarities between certain Debussy Preludes and those
of Chopin, and of other Chopin works; musical influence is evident. (Howat,
“Chopin’s Influence on the Fin de Siècle and Beyond,” 262-269.) Debussy scholars
have often used language that makes clear reference to Chopin and to the prelude as a
genre prior to its reinterpretation as character piece. Oscar Thompson stresses the
improvisatory form of the preludes, describing them as “more like sketches than
paintings” (Oscar Thompson, Debussy, Man and Artist [New York: Tudor Publishing
Company, 1940], 263-264.) Schmitz labels them “perfect miniatures” and compares
them favorably to the sketch: “a few telling scratches of the pen—the picture is
complete.” (Schmitz, 129) Vallas excuses what he perceives as unevenness of quality
by writing, “It is probable that in some cases Debussy utilized old sketches.” (Vallas,
210.) Vallas also claims that “Debussy regarded these little pieces less as works
intended to be played by themselves than as real preludes, short introductions to more
important pieces in the same keys. This so-called revolutionary showed many traits
characteristic of the conservative upholder of tradition.” He also stated that “Debussy
… gives them the character of an improvisation or fantaisie which is best suited to this
very free type of composition.” (Vallas, 208 and 210.) Roy Howat attributes lack of
proportional structure in many of the Preludes to their brevity: since they are short,
they do not need “more complex hidden unifying devices”; he proposes another
explanation stemming from the title Préludes, and suggests “that a sense of
incompleteness at the end of such pieces is apt.” He notes that incomplete harmonies
at the end of “Brouillard” may correspond to its “unresolved proportional series.”
(Howat, Debussy in Proportion, 158.)
13
Howat, “Chopin’s Influence on the Fin de Siècle and Beyond,” 266-67; foreword to
Préludes, Livre I, Livre II, 12 (French), 16 (English).
Chopin’s works, some scholars have noted tonal and motivic

connections between certain preludes in their published order and
among the preludes as a group. This evidence, even if more
compelling, would not necessarily dictate performance order, but it
has been misinterpreted as grounds for an “ideal” performance of the
complete set, in order—essentially turning the collection into a cycle
driven, in Debussy’s case, primarily by order in the publication (since
no convincing substitute for Chopin’s key order has been
uncovered).14 Many defend the practice despite the fact that Debussy’s
own performances in public and private gatherings, as well as those
given by his favored pianist, Ricardo Viñes, and others in his lifetime,
followed Chopin’s lead: the Preludes were played alone or in small
groups. Vallas has confirmed that performance tradition, stating that
Debussy “did not like the complete series to be played as a whole,
according to the present custom.”15 Debussy’s preludes depart from
Chopin’s model only in ways that make them even more a collection:
the omission of a controlling sequence, and the addition of titles that
in no way form a consistent narrative, dissolve any convincing frame.
It is just as difficult to imagine Debussy’s intention to thwart the
fragment as it is to imagine Chopin’s, for both arose from a literary
and artistic milieu of fragmentation—for Chopin, Schlegel and the
Romantics; for Debussy, late nineteenth-century decadence, a subject
to which I will return shortly.
Kallberg has laid at least part of the blame for anachronistic
performance practice in Chopin’s Preludes at the feet of late
nineteenth- and early twentieth-century analysts and critics (and the
performers who follow them) who “locate the highest artistic
achievement not so much in the individual preludes themselves … as
in the nuanced motivic relationships that may be teased out of all
twenty-four of the preludes working together.”16 This also explains, in
part, why those who advocate complete performances of Debussy’s
Preludes do so. Any approach to analysis that prejudices an over-
14
Roberts, 239-42, advocates cyclic performances of the individual books of Preludes
and offers reasons supporting cyclic performances of both books in one performance,
advising an intermission between.
15
Vallas, 210.
16
Kallberg, Chopin at the Boundaries, 149.
arching unity, though valuable in its own right for the insights it
offers, can be used to distort the relationship of part to whole, to
devalue the part. The results can affect not only practice in the
performative arts but also approaches to interpretation and criticism in
all arts. Kallberg’s concerns are shared by other scholars. David Ferris
has expressed reservations about modern expectations of unity in
Romantic song cycles, cautioning against “organicist conceptions of
musical structure,” with an emphasis on the search for a unified
musical whole, and against using notions of unity and coherence to
distinguish Schumann’s “true cycles” from “mere collections.”17
Steven Rendall laments anachronistic expectations of unity in his
discussion of Montaigne’s Essais. He notes that the Renaissance
reader (who was often read to) “did not assume that the meaning of
any given segment of discourse was dependent upon its place in the
overall structure of the text in which it occurred.” Such an approach to
a text encourages attention to the individual segment, and discourages
placing its primary value in its connections to the other segments, to
the whole.18 (Other examples of misinterpretation of notions of unity
and fragmentation were mentioned in Chapter 1.)
The problem of maintaining an historically appropriate view
of the relationship of part to whole, and of placing appropriate value
on unity and fragmentation, is exacerbated by the fragment work. The
collection of fragments may easily be mistaken for a cycle (and I use
the term cycle to mean a work in which order of performance is
significant) simply because the fragment invites the search for
relationships such as those Kallberg mentions, motivic and otherwise;
the fragment always reaches toward completion, invites the audience
to search within the individual member, out to other members of the
collection, and even outside the collection itself. To return to
Garrigues’s description in Chapter 2, the edges of every fragment pull
it both back into itself and outside, “en lui-même et hors de lui.”19 The
creator of the fragment collection may include connections, may
purposely set up a tension between fragment and whole, between unity
and disorder, between the lack of resolution and the possibility of
17
Ferris, 4-5.
18
Rendall, 72-75.
19
Garrigues, 41.
resolution—but resolution is not the goal. As Rosen puts it for

Chopin’s op. 28, “the opposing demands of the opus as a whole and of
each individual prelude are intended to coexist without being
resolved.”20 Ferris notes that the fragment cycle “implies structural
connections and hints at larger meanings, but it never makes them
explicit or definitive.” Fascination with the ruin, the sketch, the
fragment lies in the invitation to completion, but also in the
knowledge that there can be no definitive completion, only endless
possibilities. Firchow noted that the reader of Schlegel’s Athenaeum
Fragments was invited to recreate a theory of literature that actually
never existed, that even Schlegel could not communicate as a logical,
ordered whole. The creators of fragments—Debussy and Chopin
included—could and often did make wholes, not fragments; we must
remember that the intentional fragment is just that—intentional—and,
paraphrasing Bishop, allows the author to transcend what cannot be
resolved or contained in the forms and means available.21 We should
not accept Don Quixote without the interpolated stories, or Kater Murr
separated into its two narratives, or Schlegel’s fragments organized by
topic, or Petrarch’s Canzoniere with added prose narrative—or one
definitive performance order for Chopin’s or Debussy’s Preludes.
Much speculation has surrounded Debussy’s misplaced titles,
centering on two areas: the search for their sources and for an
explanation for their unusual placement. Believable, if not completely
verifiable, references for all but four22 have been tracked down, of
which the following are representative. Three titles refer to poems he
had previously set: “Les Sons et les parfums tournent dans l’air du
soir” is from Baudelaire’s “Harmonie du soir”; “La Fille aux cheveux
de lin” is the fourth of Leconte de Lisle’s “Chansons écossaises”; “Le
Vent dans la plaine” is the epigraph to Verlaine’s “C’est l’extase
langoureuse.” Some are possessions: canopic urns on the mantel
(“Canope”), a postcard he received from de Falla or Viñes (“La Puerta
del Vino”), illustrations in Chouchou’s books (Rackham’s “Les Fées
sont d’exquises danseuses” from an edition of Barrie’s Peter Pan in
20
Rosen, 83.
21
Bishop, 2-3.
22
Those four are “Des Pas sur la neige,” “La Sérénade interrompue,” “Brouillards,”
and “Bruyères.”
Kensington Garden. Many are recollections: a Greek sculpture on

display at the Louvre (“Danseuses de Delphes”), Loïe Fuller
dancing—or sailboats?—(“Voiles”), musical clowns who performed
where Debussy spent the summer of 1905 (“Minstrels”), an article in
Le Temps (“La Terrasse des audiences du clair de lune”), the Bastille
Day celebration of 1912 (“Feux d’artifice”). Howat reports an
unconfirmed story of a wine label that inspired “Les Collines
d’Anacapri” (or perhaps it was stories of Anacapri by Axel Munthe).
“Les Tierces alternées” may simply describe the musical figuration.23
The sources of this mélange are all part of Debussy’s personal life, his
possessions, bibelots, souvenirs, memories.
The history of collecting leads not only from the King’s
Kunstkammer to the state or private museum, but also from the
cabinets of curiosities to grandmother’s curio cabinet. Unicorn horns,
monster’s skulls, saint’s bone, and precious jewels may have been
replaced by sea shells from last year’s vacation, a picture of the latest
grandchild, and a commemorative spoon from the St. Louis World’s
Fair, but much of the instinct behind both is the same: the desire to
acquire possessions, whether to show material wealth or simply to
provide the comfort of being able to own; the notion of defining
oneself to others through possessions—this is where I’ve been, what
I’ve done, who I am; the need to control time, to reclaim the past by
retelling history (whether a private history or that of a civilization)
through artifacts—souvenirs, fragments, ruins.
Debussy was a collector. Madame Gérard de Romilly reports
that he spent many hours in an antique shop near her family’s house,
often spending his fee for her lessons there. He also evidently acquired
a number of his possessions as gifts: Madame de Romilly recounts
Debussy’s admiration of a medieval box and an engraved bronze
platter that belonged to her father. After Debussy’s comment that
“they’d look good in my apartment,” her mother sent them to him.24
Arthur Hartmann comments on Debussy’s collection of “priceless
Japanese and Chinese oddities,” but also reports that after receiving,
23
Howat, foreword to Préludes, Livre I, Livre II, 12-13 (French), 16-17 (English),
supplies a complete list.
24
Roger Nichols, Debussy Remembered, 54. Madame de Romilly studied singing and
piano with Debussy at the turn of the century.
with little enthusiasm, the gift of a set of heirloom buttons Hartmann

had collected in Norway, Debussy announced that there was
something in the Hartmann house that he wanted. When questioned,
Debussy identified an odd little pebble on Hartmann’s desk. From
then on Debussy kept it on his own writing desk, perhaps in company
with Arkel, the figurine he took with him on his travels. Clearly his
tendency to collect extended beyond the quotation and allusion found
in his music and letters.25
From his possessions and his memories, Debussy selected not
the most valuable objects or memories of the most important events
but the quotidian; not his collection of “Japanese oddities” but a
postcard, perhaps a wine label, favorite books, a night out. The
Preludes are musical representations of this scrapbook of the
everyday.26 There are, through Western history, genres of art and
literature that depict or describe collections. Notable examples include
the Renaissance “Cabinets of Curiosities” paintings. These sometimes
depicted real collections, and even functioned as catalogs for those
diverse juxtapositions of valuable works of art with sea shells and
telescopes; they exhibited wealth, and could include the owner
(depicted visiting his cabinet, or through a portrait hanging in the
cabinet). Often the collection is imaginary, put together for allegorical
purposes.27 In a similar category are paintings of artists in their
studios, with all their works in various states of completion around
them.
Though the Cabinet paintings have frozen the order of the art
works, objects, and people represented, there is every indication in
many of them that that order is flexible. Many objects stand on the
25
Nichols, Debussy Remembered, 208-09. Hartmann was a Hungarian violinist.
26
Other writers have described the Preludes as catalogs or summaries of Debussy’s
compositional style. Heinrich Strobel, Claude Debussy (Paris: Éditions Balzac, 1940),
222, claims that in the Preludes Debussy “a résumé toute sa création” and Oscar
Thompson, 263, writes that they sum up the technical and harmonic devices which
Debussy had been developing since 1880. Arthur Wenk, Claude Debussy and
Twentieth-Century Music, 93-113, lists several Preludes as examples of what he calls
Debussy’s masks, subsets of Debussy’s styles that he has identified. The two books
also mirror each other, each having a Spanish piece (“La Sérénade interrompue” in
Book I, “La Puerta del Vino” in Book II), a Ragtime (“Minstrels” in Book I,
“‘General Lavine’—excentric—” in Book II), etc., though not in the same order.
27
Stoichita discusses such paintings in detail.
floor, leaning against something, others are being moved, others being
held for examination. Though an interpretation of the Cabinet painting
might be based on the positions of the canvases in that painting, there
is no indication that the cabinet itself was static or fixed. The
impression is often one of movement captured mid-action and frozen
in time, like a snapshot.
“La Sérénade interrompue” takes on a special role when
viewed from the angle of the Preludes as a musical equivalent of the
“Cabinets of Curiosities” paintings, especially since the author
(Debussy) is describing his own collection. Stoichita analyzes the
cabinet of Cornelis van der Geest painted by Willem van Haecht, in
which the artist has signed a small canvas, a replica of one of his own
works, placed in the foreground of the larger work. Not only the
inserted painting, but also the signature (and therefore the author as a
“name”) “becomes an intertext.”28 Similarly, Debussy’s Ibéria
quotation becomes not simply a quotation, but a musical signature; a
reference not simply to another work, but to Debussy himself and, as
mentioned in Chapter V, to one of his compositional techniques.
Various nineteenth-century literary subgenres reflected the
increasing French, especially Parisian, fascination with acquisition and
collection that arose in part as a result of a rising economy. The
bourgeois apartment became a sort of giant curio cabinet, filled with
bibelots; Emily Apter describes it as “increasingly like a museum in
which curios, antiques, and personal memorabilia were lovingly
displayed.”29 The description of homes, whether grand or small, and
of the objects that filled them, became a literary topos. Baudelaire
wanted to combine three Poe texts—”The Domain of Arnheim,”
“Landor’s Cottage,” and “Philosophy of Furniture”—under one title,
Habitations imaginaires; Balzac’s description of the boarding house
in Le Père Goriot is famous; Edmond de Goncourt even worried about
the fate of his objects after his death:
Pour les objets que j’ai possédés, je ne veux pas, après moi, de
l’enterrement dans un musée, dans cet endroit où passent des gens
ennuyés de regarder ce qu’ils ont sous les yeux; je veux que
28
Stoichita, 229.
29
Emily Apter, “Cabinet Secrets: Fetishism, Prostitution, and the Fin de Siècle
Interior,” Assemblage, no. 9 (June 1989), 6-19.
chacun de mes objets apporte à un acquérant, à un être bien

personnel, la petite joie que j’ai eue en l’achetant.
For the objects that I possessed, I do not wish after my death

interment in a museum, in that place where men pass, bored at
looking at what they have before their eyes; I wish that each of
my objects bring to a buyer, to a very private person, the bit of joy
that I have had in buying them. 30
Debussy also worried about his objects, evidently personified

them, and perhaps projected his own feelings on them. Dolly Bardac
claims to have found a note written by Debussy as he prepared for a
trip, “Do not put Arkel in the trunk; he doesn’t like that.”31 Judging
from his letters and other reports, as an adult Debussy never enjoyed
traveling abroad. Few, if any, of his souvenirs or other prized
possessions appear to have been collected on his trips abroad, made
primarily to conduct and promote his works. The Amsterdam lawyer
and banker Richard van Rees claims to have entertained a very
homesick Debussy during a conducting trip to Holland. Jacques
Durand notes that he was a “stay-at-home” and both his stepchildren,
Dolly and Raoul Bardac, claim that (by the time they knew him)
Debussy rarely went out if he could avoid it, except, as both note, to
the bookseller’s or to buy Chinese antiques.32 His collection of
preludes includes no reference to travel of the sort that Louÿs made
when he wrote to his brother that the Chansons de Bilitis were a
souvenir of Algeria.33 Wallace Fowlie writes that Mallarmé “lived
more in his apartment than in the world.” His objects—”his fans,
vases, books, and bibelots”—made his parlor a refuge that reflected
his elitist attitude toward poetry and his rejection of popular
literature.34 These objects also found their way into his poetry, and
hold clues to deciphering his works. Debussy’s apartment was also his
30
Séverine Jouve, Obsessions and Perversions dans la littérature et les demeures à la
fin du dix-neuvième siecle (Paris: Hermann Éditeurs des sciences et des arts, 1996),
170, excerpted from Goncourt’s Journal, 1887; my translation.
31
Nichols, Debussy Remembered, 198.
32
Nichols, Debussy Remembered, 231, 195, 196, 198. It is difficult to know at what
point Debussy’s illness began to affect his desire to remain at home.
33
Venuti, 31.
34
Wallace Fowlie, Poem and Symbol: A Brief History of French Symbolism
(University Park, PA: The Pennsylvania State University Press, 1990), 52.
refuge and his objects, like Mallarmé’s, became involved to some

extent in his works.
The fictional extreme of this reclusiveness and obsession with
possessions is, of course, Des Esseintes, the anti-hero of Huysmans’s
À rebours. That work is perhaps the most famous prose representative
of the late nineteenth-century French decadent movement—a
movement that might be summed up as both a structural and topical
obsession with the object over its context, the part over the whole, the
decorative over the essential.35 Huysmans’s 1903 retrospective Preface
to À rebours, a novel begun in 1881 and published in 1884, echoes, by
accident or plan, Schlegel’s outline for the Romantic novel:
… de secouer les préjugés, de briser les limites du roman, d’y
faire entrer l’art, la science, l’histoire, de ne plus se servir, en un
mot, de cette forme que comme d’un cadre pour y insérer de plus
sérieux travaux …
… to shake off prejudices, to break apart the limits of the novel,

to introduce art, science, and history into it; in a word, no longer
to use that form except as a frame into which more important
undertakings might be inserted.36
Huysmans has taken to the extreme Schlegel’s intention (as presented

by Julius in Lucinde) to destroy order, to remove it, and to recreate
experience in the imagination,37 for in À rebours there is little if any
order, and the only experiences are those of the imagination and
memory. It is a novel in which the roles of narrative and of description
have been reversed or turned inside out, a digressive narrative,
categorized by Laurence Porter as an example of the genre that Frye
labels anatomy (see Chapter 1).38 Huysmans first provides an
introductory account of the circumstances of Des Esseintes. He is the
model decadent hero—or anti-hero: passive, inactive, perverted,
succumbing to illness rather than overcoming it. He is the sick last
son, divorced from society, the end of a family, paralleling the sick
35
David Weir, Decadence and the Making of Modernism, especially 1-21 and 82-97.
36
Text and translation in Laurence M. Porter, “Decadence and the Fin-de-Siècle
Novel” in The Cambridge Companion to the French Novel from 1800 to the Present,
edited by Timothy Unwin (Cambridge: Cambridge University Press, 1997), 96.
37
Schlegel, Lucinde, trans. Firchow, 45.
38
Porter, “Decadence and the Fin-de-Siècle Novel”, 95.
society divorced from nature, the end of a civilization. He has

retreated with all his possessions into a sort of live-in cabinet of
curiosities—with his Moreau Salomé, his liquor organ, his jewel
encrusted live turtle, and his first edition of Mallarmé’s L’Après-midi
d’un faune. Near the end of the book, a doctor advises him to return to
Paris and the society of man. Between this opening and lack of
closure, cluster a collection of chapters that, as Weir and Porter both
note, could be presented in any order since there is no plot
development, no real chronological progression; Weir even points out
that “chronological progression ceases to be of any value to an
understanding of the work at hand.”39 Chapters on literature, interior
decoration, perfumery, and so on are structurally interchangeable;
individual chapters—with the exception of a few in which Des
Esseintes recounts remembered experiences—are equally devoid of
narrative progression. Each chapter, like a guided tour through his
various cabinets—cabinets of reality, cabinets of memory—consists of
description and recollection. In À rebours the digressive novel has
disintegrated into little more than a collection, its external structure a
mirror of the content of its chapters.
Debussy’s placing of descriptive titles of the individual
Preludes at the end of each is quite unusual; so unusual, in fact, that
Genette claims it is the only example he has discovered.40 Genette
notes that these are titles of parts, not of the whole, but fails to note
that the individual Preludes actually do have titles, of a sort that he
acknowledges in his discussion, in the normal position: their
numbers—and, as mentioned above, the number of Preludes is
significant. In Genette’s categorization, rhematic titles are those which
designate genre, number, opus number in music, etc.; they refer to the
work as an object, to its classification. Thematic titles, on the other
hand, in some way indicate or allude to the subject matter of the text.
For Genette, the choice of rhematic or thematic titles indicates one of
two antithetical positions that the author may take in regard to his text:
rhematic titles indicate a “restrained stance” on the part of author, one
that recalls a certain classical dignity and later realistic seriousness;
thematic titles imply a “demonstrative—indeed insistent—stance on
39
Porter, “Decadence and the Fin-de-Siècle Novel,” 95-96; Weir, 95-96.
40
Genette, Paratexts: Thresholds of Interpretation, 65.
the part of the author toward his work.”41 Certain titles mix rhematic
and thematic elements, particularly genre and subject: Genette lists
Hume’s Treatise on Human Nature.42
Had Debussy placed both number (rhematic) and descriptive
(thematic) titles together on the first page of each Prelude, he might
have weakened the notion of the works as members of the Prelude
genre, or he might simply have emphasized his grotesque joining of
old and new types of prelude (of traditional prelude and character
piece). Instead, by separating the two types of the title, and by
positioning the descriptive title at the end of each Prelude, he lends
importance to the number, which does in fact serve as the section (or
Prelude) title, and calls into question the proper function of the
descriptive title. The distance he creates between the two is not simply
a spatial separation, but a distance that questions the real genre and
function of the work, and that highlights a certain ambivalence on
Debussy’s part toward the two different authorial stances that Genette
equates with the two types of title. How restrained, or how
“demonstrative—indeed insistent” is Debussy’s stance toward this
work?
Explanations of and attitudes toward the descriptive titles
vary. Marguerite Long calls them postscripts, and recalls Debussy and
Emma together choosing names for some of them, presumably after
the compositions were complete.43 Most explanations tend to
downplay the significance of the titles as though Debussy felt he had
to include them against his wishes or better judgment, and so hid
them. Wenk believes that the titles at the end reflect an introspection
created by the reclusion forced on Debussy by cancer. Schmitz claims
that this practice places the music first in importance; he also notes
that it stimulates enjoyment as an afterthought, offers a helping hand
to those who need it, or a confirmation (for those who evidently came
up with the right answer), and are of no use to those who have found
their own thoughts and need no suggestions.44 Perhaps they reflect the
Symbolist reticence to name: Mallarmé declared, “To name an object
41
Genette, Paratexts: Thresholds of Interpretation, 78-79, 315.
42
43
Marguerite Long, Au Piano avec Claude Debussy (Paris: Gérard Billaudot, 1960),
102.
44
Wenk, Claude Debussy and Twentieth-Century Music, 17; Schmitz, 130.
is largely to destroy poetic enjoyment, which comes from gradual

divination. The ideal is to suggest the object”;45 Schmitz must have
found that a frustrating practice, and claims that Debussy may have
put them at the end “to avoid the coyness of those who have a subject
and won’t say what it is.”46 In the same vein as Mallarmé, Redon
wrote, “Designating my drawings by giving them a title is sometimes
too much, so to speak. A title is justified only when it is vague,
indeterminate, and even confusedly aiming at the equivocal.”47 Howat
suggests a stage between the descriptive titles of the Images and the
Estampes and the strictly pedagogical headings of the Etudes:
Never one to labor a point, he [Debussy] may have felt that the
picturesque titles heading his earlier Estampes and Images had
achieved their object of encouraging extra-musical evocation, and
that the music should now be allowed to tell its own story first.48
These misplaced “titles” are a part of the paratext, and if

Debussy placed them at the ends of the Preludes because he did not
want them to function as titles, then perhaps they function as
something else; a few possibilities follow. Genette notes that chapter
titles are already accessible only to those who either read the book, or
perhaps flip through pages; in theory, if not practice, Debussy’s titles
would be limited to the performer. Schumann inserted texts in his
scores, private messages to the performer: “Papillon?” identifies the
thematic quotation in Carnaval (See Chapter 4); the final dance of
each book of the Davidsbündlertänze is prefaced by a sentence: “Here
Florestan stopped and his lips quivered in pain”; “Superfluously,
Eusebius added the following and his eyes filled with tears of
happiness.” Perhaps the unusual placement of Debussy’s descriptive
titles puts them in the category of such moments of intimacy between
45
Stephane Mallarmé, Enquête avec Jules Huret, quoted in translation in Claude
Debussy, Prelude to “The Afternoon of a Faun”: An Authoritative Score, Mallarmé’s
Poem, Backgrounds and Sources, Criticism and Analysis, edited by William W.
Austin (New York: W. W. Norton and Company, 1970), 109-110
46
Schmitz, 130.
47
Odilon Redon, À soi-même (Paris: Librairie José Corti, 1961), 26-27; quoted in
Claude Debussy, Prelude to “The Afternoon of a Faun”: An Authoritative Score,
Mallarmé’s Poem, Backgrounds and Sources, Criticism and Analysis, 125-26.
48
Howat, foreword to Préludes, Livre I, Livre II, 12-13 (French), 16-17 (English).
performer and composer, moments not to be shared with the audience

except as the performer wishes to transmit them through the music,
moments when the composer will be more open to the subject of his
work.
Given Debussy’s debt to Chopin, perhaps the misplaced titles
were a nod to the tradition of assigning inauthentic titles to Chopin’s
Preludes. This tradition is more complex than the simple assignment
of programmatic short titles such as “Raindrop” or “Winter Wind,”
and recalls a common nineteenth-century practice denoted by the term
melodeclamation, a recitation of poetry against the background of an
instrumental performance. The most widely disseminated poems for
melodeclamation accompanied by Chopin’s works were those from
Kornel Ujejski’s cycle of poems titled Tlumaczenia Szopena
[Translations of Chopin], first published in 1866, and circulating
widely for decades in Polish as well as French, English, and German
translations. Individual Preludes were even published with these titles,
and in some instances the title of the poem or its first line came to
function as a title for the Prelude, as in the case of Ujejski’s
“Ascension-dream” for Prelude no. 7.49
By placement and format (the dots of ellipsis)—and
ultimately by their function or end result—these descriptive titles
resemble what Genette labels terminal epigraphs. Epigraphs are
usually placed at the beginning of a work or a section; they are usually
quotations—the epigraph is almost always a text, but can be a
reproduction such as a drawing (Sterne uses one in Tristram Shandy)
or a musical score (three measures from Stravinsky’s Le Sacre du
printemps introduce The Consecration of Spring by Alejo
Carpentier.50 Epigraphs may comment on or elucidate both the text
and the title; they provide connections between two texts that the
reader “puts together” during the reading of the epigraphed text. This
process changes when the epigraph appears after the text, when the
reader is already familiar with the epigraphed text. As Genette notes,
49
I am indebted for this information to Halina Goldberg. The topic is briefly treated in
Mieczyslaw Tomaszewski, “Wechselbeziehungen zwischen Literatur und Musik in
der polnischen Frühromantik,” Welttheater: Die Künste im 19. Jahrhundert, edited by
Peter Andraschke and Edelgard Spaude (Freiburg im Breisgau: Rombach, 1992), 202-
10.
50
“for the reader, the relationship between introductory epigraph and

text is still prospective, but the significance of the terminal epigraph,
appearing after the text, is more obvious and more conclusive.”51 That
has been the effect of Debussy’s descriptive indications, even if it was
not his intention, for almost every scholar who defends downplaying
the titles because Debussy put them at the end also goes to great
length to explain what the titles are and where they come from—and
to then describe how the music reflects the title. In truth, we are
convinced that these mysterious “titles,” in the wrong place, are at
least as significant as those he placed at the beginning. In a statement
that recalls Mallarmé’s attitude, Genette notes that some epigraphs, if
given before a work, “would give the game away too soon.”52
Epigraphs are usually the words of a writer other than the
author of the work at hand. Many of Debussy’s titles, while they
identify his possessions, are not his own words. They are quotations—
from a poet, from a postcard, from a newspaper. As Genette notes of
the terminal epigraph, the author has left the last word to someone else
and often the author of the epigraph is more important than the
epigraph itself because in this way the author of the work at hand
chooses his peers and thus his place in the pantheon.53 If so, Debussy
has placed himself in a mixed pantheon: Baudelaire, Dickens, the
daily newspaper, a postcard. Epigraphs may also be autographic,
chosen from works of the author himself, which is also a possibility
given that, either directly, assuming the Preludes name Debussy’s own
works and possessions, or indirectly, assuming they are the words or
images of others that then lead us back to Debussy’s own possessions
and works (as the Baudelaire poem leads to Debussy’s song and to his
piano Prelude), Debussy always returns to Debussy, to his cabinet, his
souvenirs.
Debussy, in the aesthetic act of recreating memory, has
created a collection of fragments—fragments that can be arranged and
rearranged, but never achieve a perfect form, memories that defined
his life in fragments and mirrored the fragmentation of his world,
inside his apartment and out. Debussy’s collection, his response to
51
52
53
Genette, Paratexts: Thresholds of Interpretation, 158-59.
Chopin, is an act of preservation, but in fragments—remnants of the

Prelude genre juxtaposed with the character piece, remnants of his
own existence. Yet Debussy, as surely as Petrarch, realized that
preservation is ultimately impossible. Like Mallarmé’s faun, he must
relate (CONTEZ) and remember (SOUVENIR) among the dots of
ellipsis, the holes in the manuscript, the fragments of old forms, both
proof of loss and of creative potential, the futility and necessity of
reconstructing the ruins of memory, memory in ruins.
Dante wrote:
In that part of the book of my memory before which little could be

read, a rubric is found that says: Incipit vita nova. Beneath this
rubric I find written the words that it is my intention to transcribe
into this little book: if not all of them, at least their substance.54
That substance, from the book of memory, from the book of the mind,
Petrarch could only transcribe in lyric moments, Montaigne in
disordered attempts, Schlegel in “flashes of intuition,”55 and Debussy
in musical souvenirs.
54
Dante Alighieri, Vita Nuova, trans. Dino S. Cervigni and Edward Vasta (Notre
Dame: University of Notre Dame Press, 1995), 47.
55
Eichner’s phrase, 46.
BIBLIOGRAPHY
Adam, Antoine. The Art of Paul Verlaine. New York: New York
University Press, 1963.
Adler, Guido. Handbuch der Musikgeschichte. Vol. 2. Tutzing: Hans

Schneider, 1961.
Apter, Emily. “Cabinet Secrets: Fetishism, Prostitution, and the Fin de

Siècle Interior.” Assemblage, no. 9 (June 1989): 6-19.
Austin, William. Music in the Twentieth Century from Debussy

through Stravinsky. New York: W. W. Norton and Sons,
1966.
Bacon, Francis. Works. Vols. 1, 6 and 8. Edited by Spedding, Ellis,

and Heath. Boston, 1861-64. Quoted in Kenshur, Open Form
and the Shape of Ideas.
Bagehot, Walter. Literary Studies. Vol. 2. Edited by Richard Holt

Hutton. London: Longmans, Green, and Co., 1879. Reprint,
New York: AMS Press, Inc. 1973.
Baker, James M. “Post-Tonal Voice-Leading.” Early Twentieth-

Century Music, edited by Jonathan Dunsby, 20-41. London:
Blackwell, 1993.
Bakhtin, Mikhail. Rabelais and His World. Translated by Hélène

Iswolsky. Bloomington: Indiana University Press, 1984.
Baudelaire, Charles. Oeuvres complètes. 2 vols. Edited by Claude

Pichois. Paris: Éditions Gallimard, 1975-76.
Béguin, Albert. L’Âme romantique et le rêve. Paris: Corti, 1937.

Quoted in Furst, Counterparts.
Bellaigue, Camille. Revue des Deux Mondes, Paris, May 15, 1902.
Quoted in Slonimsky, Lexicon of Musical Invective.
Birkett, Jennifer. The Sins of the Fathers: Decadence in France 1870-

1914. London: Quartet Books, 1986.
Bishop, Lloyd. Romantic Irony in French Literature From Diderot to

Beckett. Nashville, TN: Vanderbilt University Press, 1989.
Blackall, Eric A. The Novels of the German Romantics. Ithaca:

Cornell University Press, 1983.
Bois, Yves-Alain. Matisse and Picasso. Paris: Flammarion, 1998.
Boulez, Pierre. Stocktakings of an Apprenticeship. Translated by

Stephen Walsh. Collected and presented by Paule Thévenin.
Oxford: Clarendon Press, 1991.
Bowie, Malcolm. Mallarmé and the Art of Being Difficult. Cambridge:

Cambridge University Press, 1978.
Briscoe, James R. Debussy: A Guide to Research. New York: Garland

Publishing, Inc., 1990.
Brown, Matthew. Debussy’s “Ibéria.” Oxford: Oxford University

Press, 2003.
__________. “Tonality and Form in Debussy’s Prélude à ‘L’Après-

midi d’un faune’.” Music Theory Spectrum 15 (1993): 127-43.
__________, and Linda Cummins. “Pelléas et Mélisande, Nuages, Le

Rossignol: Critical Concepts and Cognitive Consequences.”
Paper read at annual meeting of American Musicological
Society, Baton Rouge, February 1995.
Bibliography 173
Burke, Kenneth. “Musicality in Verse.” In The Philosophy of Literary

Form: Studies in Symbolic Action. 3rd edition. Berkeley:
University of California Press, 1973.
Byerly, Alison. “Effortless Art: The Sketch in Nineteenth-Century

Painting and Literature.” Criticism: A Quarterly for Literature
and the Arts 41 (1999): 349-64.
Calvino, Italo. Under the Jaguar Sun. Translated by William Weaver.

New York: Harcourt Brace & Company, 1988.
Cervantes, Miguel de. Don Quixote of La Mancha. Translated and

edited by Walter Starkie. New York: The New American
Library, 1957.
Chaouli, Michel. The Laboratory of Poetry: Chemistry and Poetics in

the Work of Friedrich Schlegel. Baltimore: The Johns
Hopkins University Press, 2002.
Clevenger, John R. “Achille at the Conservatoire.” Cahiers Debussy

19 (1995): 3-35.
__________. “Debussy’s Paris Conservatoire Training.” In Fulcher,

Debussy and his World, 299-361.
Clive, H. P. Pierre Louÿs. Oxford: Clarendon Press, 1978.
Cobb, Margaret. “Debussy and Le Roman de Rosette.” Cahiers

Debussy 22 (1998): 75-87.
Cohn, Robert Greer. Toward the Poems of Mallarmé. Berkeley:

Cone, Edward T. Musical Form and Musical Performance. New

York: W. W. Norton and Company, 1968.
Crowley, Roseline. “Toward the Poetics of Juxtaposition: L’Après-

midi d’un faune.” Yale French Studies 54 (1977): 33-44.
Cummins, Linda P. “Debussy and the Grotesque.” Annual Meeting,

American Musicological Society, Southern Chapter,
University of Alabama, February 1997.
Dante Alighieri. Vita Nuova. Translated by Dino S. Cervigni and

Edward Vasta. Notre Dame: University of Notre Dame Press,
1995.
Daverio, John. Nineteenth-Century Music and the German Romantic

Ideology. New York: Schirmer Books, 1993.
Debussy, Claude. Correspondance 1884-1918. Edited by François

Lesure. Paris: Hermann éditeurs des sciences et des arts, 1993.
__________. Debussy Letters. Selected and edited by François Lesure

and Roger Nichols. Translated by Roger Nichols. London:
Faber and Faber, 1987.
__________. Debussy on Music: The Critical Writings of the Great

French Composer. Collected by François Lesure. Translated
and edited by Richard Langham Smith. Ithaca: Cornell
__________. Préludes, Livre I, Livre II. Edited by Roy Howat with

the collaboration of Claude Helffer. Oeuvres Complètes de
Claude Debussy, Série I, vol. 5. Paris: Durand-Costallat,
1995.
__________. Prelude to “The Afternoon of a Faun”: An Authoritative

Score, Mallarmé’s Poem, Backgrounds and Sources,
Criticism and Analysis. Edited by William W. Austin. New
York: W. W. Norton and Company, 1970.
Bibliography 175
__________. Songs 1880-1904. Edited and translated by Rita Benton.

New York: Dover Publications, 1981.
Delacroix, Eugène. Journal 1822-1863. Edited by André Joubin.

Paris, 1931-32; rev. ed. 1980. Quoted in Dorra, Symbolist Art
Theories and Kallberg, “Chopin and the Aesthetic of the
Sketch.”
__________. “Réalisme et idéalisme.” Oeuvres littéraires. Vol. 1.

Paris: Crès, 1923.
Diderot, Denis. Observations sur la Sculpture et sur Bouchardon. In

Oeuvres complètes, Vol. 8. Edited by J. Assézat and Maurice
Tourneux. Paris: Garnier, 1875-77. Quoted by Weinshenker,
“Diderot’s Use of the Ruin Image.”
__________. Salons. Edited by Jean Seznec and Jean Adhémar.

Oxford: Clarendon 1963.
Dommel-Diény, A. L’Analyse harmonique en examples de J. S. Bach

à Debussy. Neuchâtel: Éditions Delachaux et Niestlé, 1967.
Dorra, Henri, ed. Symbolist Art Theories: A Critical Anthology.

Berkeley: University of California Press, 1994.
Dronke, Peter. Verse with Prose from Petronius to Dante: The Art and
Scope of the Mixed Form. Cambridge: Harvard University
Press, 1994.
Dubruck, Alfred. Gérard de Nerval and the German Heritage. The

Hague: Mouton & Co., 1965.
Duval, Edwin M. The Design of Rabelais’s Pantagruel. New Haven

and London: Yale University Press, 1991.
__________. “Rabelais and Textual Architecture.” A New History of

French Literature, 154-59. Edited by Denis Hollier.
Cambridge: Harvard University Press, 1994.
Eichner, Hans. Friedrich Schlegel. New York: Twayne Publishers,

Inc., 1970.
Eigeldinger, Jean-Jacques. “Debussy et l’idée d’arabesque musicale,”

Cahiers Debussy (1988-89): 5-14.
Eliot, T. S. The Wasteland. T. S. Eliot, the Complete Poems and Plays,

1909-1950. Harcourt, Brace and Company, 1958.
Elson, Louis. Boston Daily Advertiser, March 4, 1907. Quoted in

Slonimsky, Lexicon of Musical Invective.
Engster, Dan. “The Montaignian Moment.” Journal of the History of

Ideas 59/4 (Oct 1998): 625-50.
Ferris, David. Schumann’s Eichendorff Liederkreis and the Genre of

the Romantic Cycle. Oxford: Oxford University Press, 2000.
Folkenflik, Robert. “Macpherson, Chatterton, Blake and the Great

Age of Literary Forgery.” The Centennial Review 18 (1974):
378-91.
Fowlie, Wallace. Mallarmé. Chicago: The University of Chicago

Press, 1953.
__________. Poem and Symbol: A Brief History of French

Symbolism. University Park, PA: The Pennsylvania State
Friedmann, Michael L. “Approaching Debussy’s ‘Ondine’.” Cahiers

Debussy 6 (1982): 14-22.
Bibliography 177
Frye, Northrop. Anatomy of Criticism: Four Essays. Princeton:

Princeton University Press, 1990.
Fulcher, Jane, ed. Debussy and his World. Princeton: Princeton

Furst, Lilian. Counterparts: The Dynamics of Franco-German

Literary Relationships 1770-1895. Detroit: Wayne State
University Press: 1977.
Garrigues, Pierre. Poétiques du fragment. Paris: Klincksieck, 1995.
Genette, Gérard. Narrative Discourse: An Essay in Method.

Translated by Jane E. Lewin. Ithaca, NY: Cornell University
Press, 1980.
__________. Palimpsests: Literature in the Second Degree.

Translated by Channa Newman and Claude Doubinsky.
Lincoln: University of Nebraska Press, 1997.
__________. Paratexts: Thresholds of Interpretation. Translated by

Jane E. Lewin. Cambridge: Cambridge University Press,
1997.
Gervaise, Françoise. “La notion d’arabesque chez Debussy.” La Revue

musicale 241 (1958): 3-23.
Gordon, Rae Beth. Ornament, Fantasy, and Desire in Nineteenth-

Century French Literature. Princeton: Princeton University
Press, 1992.
Görres, Joseph. Mythengeschichte der asiatischen Welt. Vol. 2

Heidelberg: Mohr and Zimmer, 1810. Quoted in Dorra,
Symbolist Art Theories: A Critical Anthology.
Grafton, Anthony. The Footnote: A Curious History. Cambridge:

Harvard University Press, 1998.
Grayson, David. “Bilitis and Tanagra: Afternoons with Nude Women”

In Fulcher, Debussy and his World, 117-140.
__________. The Genesis of Debussy’s “Pelléas et Mélisande.” UMI

Research Press Studies in Musicology. Ann Arbor, Michigan:
UMI Research Press, 1986.
Grout, Donald J. and Hermine Weigel Williams. A Short History of

Opera. 4th ed. New York: Columbia University Press, 2003.
__________ and Claude V. Palisca. A History of Western Music. 6th

ed. New York: W.W. Norton and Company, 2001.
Guentner, Wendelin. “The Inscription of the Sketch in the 19th-

Century French Journal.” Nineteenth-Century French Studies
27 (1999): 276-89.
Gut, Serge. “‘Canope’ de Debussy, analyse formelle et structure

fondamentale.” Revue musicale de la Suisse romande 33
(1980): 60-65.
Hannoosh, Michele. Painting and the “Journal” of Eugène Delacroix.

Princeton: Princeton University Press, 1995. Quoted in
Kallberg, “Chopin and the Aesthetic of the Sketch.”
Harpham, Geoffrey Galt. On The Grotesque: Strategies of

Contradiction in Art and Literature. Princeton: Princeton
Harries, Elizabeth. The Unfinished Manner: Essays on the Fragment

in the Later Eighteenth Century. Charlottesville: University
Press of Virginia, 1994.
Bibliography 179
Heely, Joseph. Letters on the Beauties of Hagley, Envil, and the

Leasowes, with Critical Remarks and Observations on the
Modern Taste in Gardening. 2 vols. London: R. Baldwin.
1777.
Hemingway, Ernest. For Whom the Bell Tolls. New York: Charles
Scribner’s Sons, 1940.
Herlin, Denis. “Trois Scènes au Crépuscule (1892-1893): Un Premier

Projet des Nocturnes?” Cahiers Debussy No. 21 (1997): 3-24.
Highet, Gilbert. The Classical Tradition: Greek and Roman Influences

on Western Literature. Oxford: Oxford University Press,
1985.
Hoeckner, Berthold. “Schumann and Romantic Distance.” Journal of

the American Musicological Society 50 (1997): 55-132.
Holloway, Robin. Debussy and Wagner. London: Eulenburg Books,

1979.
Homer, The Iliad. Translated by Robert Fagles. New York: Penguin

Books, 1991.
Houston, John Porter. French Symbolism and the Modernist

Movement: A Study of Poetic Structures. Baton Rouge:
Louisiana State University Press, 1980.
Howat, Roy. “Chopin’s Influence on the Fin de Siècle and Beyond.”

In The Cambridge Companion to Chopin, edited by Jim
Samson, 246-83. Cambridge: Cambridge University Press,
1992.
__________. “En route for L’Isle joyeuse: the Restoration of a

Triptych.” Cahiers Debussy 19 (1995): 37-52.
__________. Debussy in Proportion. Cambridge: Cambridge

Hugo, Victor. Cromwell. Paris: Garnier-Flammarion, 1968.
Immerwahr, Raymond. “Romantic Irony and Romantic Arabesque

Prior to Romanticism.” The German Quarterly 42 (1969):
665-85.
Isidore of Seville, Etymologiarum Sive Originum Liber XI. Translated

by William D. Sharpe in “Isidore of Seville: The Medical
Writings.” Transactions of the American Philosophical
Society, new series, 54 (1963): 3-70.
Jarocinski, Stefan J. Debussy: impressionnisme et symbolisme. Paris:

Éditions du Seuil, 1970.
Johnson, Carroll B. Don Quixote: The Quest for Modern Fiction.

Boston: Twayne Publishers, 1990.
Johnson, Laurie Ruth. The Art of Recollection in Jena Romanticism:

Memory, History, Fiction, and Fragmentation in Texts by
Friedrich Schlegel and Novalis. Tübingen: Max Niemeyer
Verlag, 2002.
Johnston, Otto W. “Chromatic Symbolism in Gottfried Keller’s

Romeo und Julia auf dem Dorfe.” Themes and Structures:
Studies in German Literature from Goethe to the Present.
Edited by Alexander Stephan. Columbia, SC: Camden House,
1997.
Jouve, Séverine. Obsessions and Perversions dans la littérature et les

demeures à la fin du dix-neuvième siècle. Paris: Hermann
Éditeurs des sciences et des arts, 1996.
Bibliography 181
Kallberg, Jeffrey. “Chopin and the Aesthetic of the Sketch: A New

Prelude in E flat Minor?” Early Music 29 (2001): 408-22.
__________. Chopin at the Boundaries: Sex, History, and Musical

Genre. Cambridge: Harvard University Press, 1996.
Kayser, Wolfgang. The Grotesque in Art and Literature. Translated by

Ulrich Weisstein. Bloomington: Indiana University Press,
1963.
Kenshur, Oscar L. Open Form and the Shape of Ideas: Literary

Structures as Representations of Philosophical Concepts in
the Seventeenth and Eighteenth Centuries. Lewisburg:
Bucknell University Press; London and Toronto: Associated
University Presses, 1986.
Kermode, Frank. The Sense of an Ending: Studies in the Theory of

Fiction. Oxford: Oxford University Press, 1966.
Knapp, Bettina. Maurice Maeterlinck. Boston: Twayne Publishers,

1975.
Kritzman, Lawrence D., ed. Fragments: Incompletion and

Discontinuity. New York: New York Literary Forum, 1981.
__________. “Rabelais’s Splintered Voyage.” In Kritzman,

Fragments: Incompletion and Discontinuity, 53-70. New
York: New York Literary Forum, 1981.
Lacoue-Labarthe, Philippe and Jean-Luc Nancy. The Literary

Absolute: The Theory of Literature in German Romanticism.
Translated by Philip Barnard and Cheryl Lester. New York:
State University of New York Press, 1988. Originally
published as L’Absolu Littéraire. Paris: Éditions du Seuil,
1978.
Lautréamont. Poésies II. Lautréamont and Germain Nouveau. Oeuvres

complètes. Edited by Pierre-Olivier Walzer. Paris: Éditions
Gallimard, 1970.
Lesure, François. Claude Debussy. Paris: Klincksieck, 1994.
__________. Debussy: Iconographie musicale. Paris: Édition Minkoff

& Lattès, 1980.
Levinson, Marjorie. The Romantic Fragment Poem: A Critique of a

Form. Chapel Hill: The University of North Carolina Press,
1986.
Lewin, David. “Parallel Voice-Leading in Debussy.” In Music at the

Turn of the Century, edited by Joseph Kerman. Berkeley:
List of winners of Musica’s composer identification contest. Musica,

no. 31 (April 1905): 63.
Lockspeiser, Edward. Debussy: His Life and Mind. 2nd ed. 2 vols.
Cambridge: Cambridge University Press, 1978.
Long, Marguerite. Au Piano avec Claude Debussy. Paris: Gérard

Billaudot, 1960.
Louÿs, Pierre. Les Chansons de Bilitis. Paris: Albin Michel, 1965.
__________. Aphrodite: Moeurs antiques. Paris: Gallimard, 1992.
MacPhail, Eric. “Antiquities and Antiquaries.” In A New History of

French Literature, edited by Denis Hollier, 209-13.
Mallarmé, Stéphane. Correspondance: 1862-1871. Edited by Henri

Mondor. Paris: Gallimard, 1959.
Bibliography 183
__________. La Musique et les lettres. Oeuvres complètes. Edited by

Henri Mondor and G. Jean-Aubry. Paris: Éditions Gallimard,
1945.
__________. L’Après-midi d’un faune. Oeuvres complètes. Edited by

Henri Mondor and G. Jean-Aubry. Paris: Éditions Gallimard,
1945.
__________. Selected Prose Poems, Essays, and Letters. Translated

by Bradford Cook. Baltimore: The Johns Hopkins Press,
1956.
Martínez -Bonati, Félix. “Don Quixote” and the Poetics of the Novel.
Translated by Dian Fox. Ithaca: Cornell University Press,
1992.
Mazzotta, Giuseppe. The Worlds of Petrarch. Durham: Duke

McFarland, Thomas. Romanticism and the Forms of Ruin:

Wordsworth, Coleridge, and Modalities of Fragmentation.
Princeton: Princeton University Press, 1981.
McGuinnes, Patrick. Maurice Maeterlinck and the Making of Modern

Theatre. Oxford: Oxford University Press, 2000.
Meyer, Herman. The Poetics of Quotation in the European Novel.

Translated by Theodore and Yetta Ziolkowski. Princeton:
Princeton University Press, 1968.
Meyer, Leonard B. Style and Music. Philadelphia: University of

Pennsylvania Press, 1989.
Michaud, Guy. Mallarmé. New York: New York University Press,

1965.
Miller, Naomi. Heavenly Caves: Reflections on the Garden Grotto.

New York: George Braziller, 1982.
Montaigne, Michel de. The Complete Essays of Montaigne. Translated

by Donald M. Frame. Stanford: Stanford University Press,
1965.
__________. The Complete Works of Montaigne: Essays, Travel

Journal, Letters. Translated by Donald M. Frame. Stanford,
California: Stanford University Press, 1957.
__________. Essais. Vol. 1. Paris: Garnier, 1962. Quoted in Kenshur,

Open Form and the Shape of Ideas.
Nerval, Gérard de. Aurélia, Oeuvres complètes. Vol 2. Edited by

Albert Béguin and Jean Richer. Paris: Éditions Gallimard,
1966.
Nichols, Roger. The Life of Debussy. Cambridge: Cambridge

__________. Debussy Remembered. Portland, OR: Amadeus Press,

1992.
__________ and Richard Langham Smith. Claude Debussy: Pelléas et

Mélisande. Cambridge Opera Handbooks. Cambridge:
Cambridge University Press, 1989.
Nodier, Charles. Histoire du roi de Bohême et de ses sept châteaux.

Paris: Éditions Plasma, 1979.
Nuttall, A. D. Openings: Narrative Beginnings from the Epic to the

Novel. Oxford: Clarendon Press, 1992.
Orledge, Robert. Debussy and the Theater. Cambridge: Cambridge

Bibliography 185
__________. “Debussy’s Musical Gifts to Emma Bardac.” Musical

Quarterly 62/4 (1976): 536-553.
Oulmont, Charles. Noces d’or avec mon passé. Paris, 1964.
Pack, Roger A. “Two Classical Forgeries.” American Journal of

Philology 110 (1989): 479-480.
Parks, Richard S. The Music of Claude Debussy. New Haven: Yale

Pearson, Roger. Unfolding Mallarmé: The Development of a Poetic

Art. Oxford: Clarendon Press, 1996.
Petrarch. Petrarch’s Lyric Poems: The “Rime Sparse” and Other

Lyrics. Translated and edited by Robert M. Durling.
Peyre, Henri. What is Symbolism? Translated by Emmett Parker.

Tuscaloosa: The University of Alabama Press, 1974.
Originally published as Qu’est-ce le symbolisme? Paris:
Presses Universitaires de France, 1974.
Portal, Pierre Paul Frédéric de. Des couleurs symboliques dans

l’antiquité, le moyen-âge et les temps modernes. Paris:
Treuttel et Würtz, 1837.
Porter, Laurence M. The Crisis of French Symbolism. Ithaca: Cornell

University Press, 1990 .
__________. “Decadence and the Fin-de-Siècle Novel.” In The

Cambridge Companion to the French Novel from 1800 to the
Present. Edited by Timothy Unwin. Cambridge: Cambridge
Rabelais. The Complete Works of François Rabelais. Translated by

Donald M. Frame. Berkeley: University of California Press,
1991.
__________. Gargantua. Oeuvres complètes. Edited by Mireille

Huchon. Paris: Gallimard, 1994.
Rabkin, Eric S. “Spatial Form and Plot.” Spatial Form in Narrative.

Edited by Jeffrey R. Smitten. Ithaca: Cornell University Press,
1981.
Raffel, Burton. Gargantua and Pantagruel. New York: W. W. Norton

& Company, 1990.
Rajan, Balachandra. The Form of the Unfinished: English Poetics

from Spenser to Pound. Princeton: Princeton University Press,
1985.
Redon, Odilon. À soi-même. Paris: Librairie José Corti, 1961.
Regosin, Richard L. “Montaigne and His Readers.” A New History of

French Literature. Edited by Denis Hollier, 248-53.
Rendall, Steven. “In Disjointed Parts/Par articles décousus.” In

Kritzman, Fragments: Incompletion and Discontinuity, 71-83.
Robb, Graham. Unlocking Mallarmé. New Haven: Yale University

Press, 1996.
Roberts, Paul. Images: The Piano Music of Claude Debussy. Portland,

Oregon: Amadeus Press, 1996.
Roland-Manuel. À la gloire de Ravel. Paris: Nouvelle Revue critique,

1938.
Bibliography 187
Rosand, David. “Composition/Decomposition/Recomposition: Notes

on the Fragmentary and Artistic Process.” In Kritzman,
Fragments: Incompletion and Discontinuity, 17-30.
Rosen, Charles. The Romantic Generation. Cambridge: Harvard

__________ and Henri Zerner. Romanticism and Realism: The

Mythology of Nineteenth-Century Art. New York: Viking,
1984.
Rumph, Stephen. “Debussy’s Trois Chansons de Bilitis: Song, Opera,

and the Death of the Subject.” The Journal of Musicology 12
(1994): 477-83.
Said, Edward W. Beginnings: Intention and Method. New York:

Columbia University Press, 1985.
Sayendra, Ramon. “Liszt’s Open Structures and the Romantic

Fragment.” Music Theory Spectrum 19 (Fall 1997): 184-205.
Schenker, Heinrich. Free Composition (Der freie Satz), New Musical

Theories and Fantasies, No. 3. Translated and edited by Ernst
Oster. 2 vols. New York: Longman, 1979.
Scher, Steven Paul. “Hoffmann and Sterne: Unmediated Parallels in

Narrative Method.” Comparative Literature 28 (1976): 309-
25.
Schlegel, Friedrich. Charakteristiken und Kritiken I (1796-1801).

Edited by Hans Eichner. In Kritische Friedrich-Schlegel-
Ausgabe. Vol. 2. Munich: Schöningh, 1967. Quoted in
Harries, The Unfinished Manner.
__________. Dialogue on Poetry and Literary Aphorisms. Edited by

Ernst Behler and Roman Struc. University Park, PA: The
Pennsylvania State University Press, 1968.
__________. Friedrich Schlegel’s “Lucinde” and the Fragments.

Translated and edited by Peter Firchow. Minneapolis:
University of Minnesota Press, 1971.
__________. Literary Notebooks, 1797-1801. Edited by Hans Eichner.

Toronto: University of Toronto Press, 1957.
Schmitz, E. Robert. The Piano Works of Claude Debussy. New York:

Dover Publications, 1966. Originally published Duell, Sloan
& Pearce, Inc., 1950.
Schumann, Robert. Neue Zeitschrift für Musik. 22 December 1837.

Quoted in Kallberg, Chopin at the Boundaries.
Scott, Clive. French Verse-art: A Study. Cambridge: Cambridge

Scott, David. Pictorialist Poetics: Poetry and the Visual Arts in

Nineteenth-Century France. Cambridge: Cambridge
Shattuck, Roger. “The Alphabet and the Junkyard.” In Kritzman,

Fragments: Incompletion and Discontinuity, 31-38.
Shiff, Richard. Cézanne and the End of Impressionism: A Study of the

Theory, Technique, and Critical Evaluation of Modern Art.
Chicago: University of Chicago Press, 1984.
Shroder, Maurice Z. “The Satyr and the Faun: A Definition of

Romanticism.” Symposium: A Quarterly Journal in Modern
Foreign Literature 23 (Fall-Winter 1969): 346-53.
Bibliography 189
Slonimsky, Nicholas. Lexicon of Musical Invective: Critical Assaults

on Composers since Beethoven’s Time. 2nd edition. Seattle:
University of Washington Press, 1969.
Smith, Richard Langham. “The Play and its Playwright.” In Nichols

and Smith, Claude Debussy: Pelléas et Mélisande, 1-29.
__________. “Tonalities of Darkness and Light.” In Nichols and

Smith, Claude Debussy: Pelléas et Mélisande,107-39.
Stoichita,Victor I. The Self-Aware Image: An Insight into Early

Modern Meta-Painting. Translated by Anne-Marie Glasheen.
Stone, Gregory. The Death of the Troubadour. Philadelphia:

University of Pennsylvania Press, 1994.
Strobel, Heinrich. Claude Debussy. Paris: Éditions Balzac, 1940.
Szymanowska, Maria. Album Musical: Maria Szymanowska.

Translated by Renata Suchowiejko. Krakow: Musica
Iagellonica, 1999.
Thompson, Oscar. Debussy, Man and Artist. New York: Tudor

Publishing Company, 1940.
Thorel, Jean. “Les Romantiques allemands et les symbolistes

français.” Entretiens politiques et littéraires 3 (1891): 85-109.
Quoted in Peyre, What is Symbolism and Furst, Counterparts.
Todorov, Tzvetan. The Fantastic: A Structural Approach to a Literary

Genre. Translated by Richard Howard. Ithaca, NY: Cornell
Tomaszewski, Mieczyslaw. “Wechselbeziehungen zwischen

Literature und Musik in der polnischen Frühromantik.” In
Welttheater: Die Kunste im 19. Jahrhundert, edited by Peter
Andraschke and Edelgard Spaude, 202-210. Freiburg im
Breisgau: Rombach, 1992.
Trezise, Simon. Debussy: “La Mer.” Cambridge Music Guides.

Vallas, Léon. Claude Debussy: His Life and Works. Translated by

Maire O’Brien and Grace O’Brien. London: Oxford
University Press, 1933. Reprint. New York: Dover
Publications, Inc., 1973.
Vanhulst, Henri. “L’éditeur bruxellois Schott frères et D’un cahier

d’esquisses de Claude Debussy.” Cahiers Debussy no. 26
(2002): 3-13.
Venuti, Lawrence. “The Scandal of Translation.” French Literature

Series 22 (1995): 25-38.
Verlaine, Paul. Oeuvres poétiques complètes. Edited by Jacques Borel.

Paris: Gallimard, 1962.
Vickers, Nancy J. “Diana Described: Scattered Woman and Scattered

Rhyme.” Critical Inquiry 8 (1981): 85-97. Quoted in Harries,
The Unfinished Manner.
Villey, P. Les Sources et l’évolution des essais de Montaigne. Paris,

1908. Quoted in Highet, The Classical Tradition.
Vordtriede, Werner. Novalis und die französischen Symbolisten.

Stuttgart: Kohlhammer, 1963.
Warburton, Thomas. “Bitonal Miniatures by Debussy from 1913.”

Cahiers Debussy, 6 (1982): 5-15.
Bibliography 191
Watkins, Glenn. Pyramids at the Louvre: Music, Culture, and Collage

from Stravinsky to the Postmodernists. Cambridge: The
Belknap Press of Harvard University Press, 1994.
__________. Soundings: Music in the Twentieth Century. New York:

Schirmer books, A Division of Macmillan, Inc., 1988.
Weinberg, Bernard. The Limits of Symbolism: Studies of Five Modern

French Poets. Chicago: The University of Chicago Press,
1966.
Weinshenker, Anne Betty. “Diderot’s Use of the Ruin-Image.” In

Diderot Studies 16 (1973): 309-329.
Weir, David. Decadence and the Making of Modernism. Amherst:

University of Massachusetts Press, 1995.
Weissmann, Adolf. The Problems of Modern Music. Translated by M.

M. Bozman. London: J. M. Dent and Sons Ltd., 1925.
Originally published as Die Musik in der Weltkrise, 1922.
Wenk, Arthur B. Claude Debussy and the Poets. Berkeley: University

of California Press, 1976.
__________. Claude Debussy and Twentieth-Century Music. Boston:

Twayne Publishers, 1983.
Wilkins, E. H. Life of Petrarch. Chicago: University of Chicago Press,

1961. Quoted in Durling, ed., Petrarch’s Lyric Poems.
Williams, David. Deformed Discourse: The Function of the Monster

in Mediaeval Thought and Literature. Montreal: McGill-
Queen’s University Press, 1996.
Wind, Edgar. Art and Anarchy. 3rd edition. Chicago: Northwestern

Youens, Susan. “Debussy’s Song Cycles.” The NATS Bulletin

(Sept./Oct., 1986): 13-15.

Pub - Debussy and The Fragment Chiasma 18 PDF

Uploaded by

Copyright:

Available Formats

Pub - Debussy and The Fragment Chiasma 18 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pub - Debussy and The Fragment Chiasma 18 PDF

Uploaded by

Copyright:

Available Formats

Debussy and the Fragment

Amsterdam - New York, NY 2006

The paper on which this book is printed meets the requirements of

Chiasma seeks to foster urgent critical assessments focussing upon joinings

I would like to thank Professors Matthew G. Brown, now of the

“Gifted as he was, there was nothing cosmic or inspired about

notre art … et la ruine de notre être” [the decomposition of our art …

perception of unity as the ultimate goal of art—a unity analysts and

[Have you played the Images … ? Without false vanity, I think

In 1904 Laloy introduced a concert with a talk intended to show

loved the music of Schumann and understood its capricious

Ultimately, Rosen observes that the classical hierarchy of

had been shaken to the ground by Schumann and Chopin, shored

Rosen’s statements parallel literary comparisons between

With this conversion to mythology and symbolism, Schlegel

Lilian Furst expands this comparison in an extended work, noting

language, religious feeling.”22 A comparison of texts by German

The arts should be brought together again, and bridges sought

foreshadows Baudelaire’s notion of “Correspondances” in the first

Comme de longs échos qui de loin se confondent

Vaste comme la nuit et comme la clarté,

[Nature is a temple where living pillars

Like extended echoes which mingle far away

Henri Dorra speculates that the source for Baudelaire’s image of

distance captures the imagination of the musician: distant philosophy

The appropriate way to determine whether a painting is melodious

Dorra points out similarities to Delacroix’s aesthetic theories:

There is an impression that results from a certain arrangement of

In the distance, everything becomes music.

our position in a historical continuum.”32 The purpose of this book is

examine strategies of fragmentation in individual works written either

A portent therefore, does not arise contrary to

Man’s fascination with the fragment is ancient, his uses of it

story, Cervantes opened Don Quixote with endorsements as fake and

part of the image and to the artificiality of the convention of portrait

that fragment writers build on the work of previous creators of

“Voi ch’ ascoltate in rime sparse il suono / di quei sospiri …”

by a prose narrative explaining the poem’s structure and the

Petrarch presents a chronology, writing 365 poems plus a final

For Petrarch the term [fragment] expresses the intensely

As a reader of classical manuscripts, Petrarch knew both the

Like Petrarch’s Canzoniere, Montaigne’s Essais also establish

particularly the annotated Adages of Erasmus.28 These collections do

essay “Of Friendship” begins with a quotation comparing his writing

If I had written to seek the world’s favor, I should have bedecked

In a vivid description of their structure, Montaigne described

As I was considering the way a painter I employ went

A lovely woman tapers off into a fish.

I do indeed go along with my painter in this second

Montaigne’s preface announced his Essais as autobiography.

Countering the expected chronology is a deformed

series of disconnected wanderings of the mind. As the works

Our actions are nothing but a patchwork. … We are all

Anthony Grafton writes, “The introspective, wide-ranging Essays of

Given the sheer numbers of their imitators and the acceptance

In 1534 and again in 1535, Rabelais was in Rome as private

a i ? enu le grand dompteur des Cimbres

Alcofribas performs the job of the audience for any

grotesques. Gargantua’s mother, for instance, eleven months pregnant,