Formal Functions in Perspective: Essays On Musical Form From Haydn To Adorno

Formal Functions
in Perspective
Essays on Musical Form
from Haydn to Adorno
e d i t e d b y s t e v e n va n d e m o o rt e l e ,
j u l i e p e d n e au lt- d e s l au r i e r s , a n d
n at h a n j o h n m a rt i n
Formal Functions in Perspective
Vande Moortele.indd i 9/30/2015 7:50:27 PM

Eastman Studies in Music
Ralph P. Locke, Senior Editor
Eastman School of Music
Additional Titles of Interest
Analyzing Atonal Music: Pitch-Class Set Theory and Its Contexts

Michiel Schuijer
Analyzing Wagner’s Operas: Alfred Lorenz and German Nationalist Ideology

Stephen McClatchie
Aspects of Unity in J. S. Bach’s Partitas and Suites: An Analytical Study

David W. Beach
Bach to Brahms: Essays on Musical Design and Structure

Edited by David Beach and Yosef Goldenberg
Concert Music, Rock, and Jazz since 1945: Essays and Analytical Studies
Edited by Elizabeth West Marvin and Richard Hermann
A Dance of Polar Opposites:

The Continuing Transformation of Our Musical Language
George Rochberg
Edited by Jeremy Gill
The French Symphony at the Fin de Siècle:

Style, Culture, and the Symphonic Tradition
Andrew Deruchie
Music Theory in Concept and Practice

Edited by James M. Baker, David W. Beach, and Jonathan W. Bernard
Reviving Haydn: New Appreciations in the Twentieth Century

Bryan Proksch
A Theory of Music Analysis: On Segmentation and Associative Organization

Dora A. Hanninen
A complete list of titles in the Eastman Studies in Music series

may be found on our website, www.urpress.com.
Vande Moortele.indd ii 9/30/2015 7:50:47 PM

Formal Functions
in Perspective

Edited by Steven Vande Moortele,

Julie Pedneault-Deslauriers,
and Nathan John Martin
Vande Moortele.indd iii 9/30/2015 7:50:47 PM

Copyright © 2015 by the Editors and Contributors
All rights reserved. Except as permitted under current legislation, no

part of this work may be photocopied, stored in a retrieval system,
published, performed in public, adapted, broadcast, transmitted,
recorded, or reproduced in any form or by any means, without the
prior permission of the copyright owner.
First published 2015
University of Rochester Press

668 Mt. Hope Avenue, Rochester, NY 14620, USA
www.urpress.com
and Boydell & Brewer Limited
PO Box 9, Woodbridge, Suffolk IP12 3DF, UK
www.boydellandbrewer.com
ISBN-13: 978-1-58046-518-2
ISSN: 1071-9989
Library of Congress Cataloging-in-Publication Data
Formal functions in perspective : essays on musical form from Haydn to

Adorno / edited by Steven Vande Moortele, Julie Pedneault-Deslauriers, and
Nathan John Martin.
pages cm — (Eastman studies in music, ISSN 1071-9989 ; v. 127)
Includes bibliographical references and index.
ISBN 978-1-58046-518-2 (hardcover : alkaline paper) 1. Musical form. I. Vande
Moortele, Steven, editor. II. Pedneault-Deslauriers, Julie, editor. III. Martin,
Nathan John, editor. IV. Series: Eastman studies in music ; v. 127.
MT58.F68 2015
781.8—dc23 2015021591
A catalogue record for this title is available from the British Library.
This publication is printed on acid-free paper.

Printed in the United States of America.
Vande Moortele.indd iv 9/30/2015 7:50:47 PM

Contents
Acknowledgments vii
Introduction 1
Steven Vande Moortele, Julie Pedneault-Deslauriers,
and Nathan John Martin
Part One: Theoretical Studies in Haydn and Mozart

1 “Functial Formanality”: Twisted Formal Functions in
Joseph Haydn’s Symphonies 11
L. Poundie Burstein
2 Mozart’s Sonata-Form Arias 37

Nathan John Martin
Part Two: Nineteenth-Century Taxonomies

3 Formal Type and Formal Function in the Postclassical
Piano Concerto 77
Julian Horton
4 Saint-Saëns’s Cyclic Forms 123

Andrew Deruchie
Part Three: Schubert

5 Schubert’s “Deflected-Cadence” Transitions and the
Classical Style 165
Brian Black
6 “Heavenly Length” in Schubert’s Instrumental Music 198

François de Médicis
Vande Moortele.indd v 9/30/2015 7:50:47 PM

vi contents
Part Four: Text, Texture, and Form

7 Sentences in the Lieder of Robert Schumann:
The Relation to the Text 225
Harald Krebs
8 Parlante Talk: Texture and Formal Function in the

Operas of Verdi 252
Steven Huebner
Part Five: Analysis and Hermeneutics

9 Discipline and Punish among the Winds in the First Movement
of Beethoven’s First Symphony 295
Henry Klumpenhouwer
10 Laborious Homecomings: The “Ongoing Reprise” from

Clementi to Brahms 317
Giorgio Sanguinetti
Part Six: Schoenberg and Beyond

11 Dominant Tunnels, Form, and Program in Schoenberg’s
Verklärte Nacht 345
Julie Pedneault-Deslauriers
12 Form and Serial Function in Leibowitz’s

Trois poèmes de Pierre Reverdy 373
Christoph Neidhöfer and Peter Schubert
13 The Philosopher as Theorist: Adorno’s materiale Formenlehre 411

Steven Vande Moortele
Afterword 434
Janet Schmalfeldt
Notes on Contributors 441
Index 445
Vande Moortele.indd vi 9/30/2015 7:50:47 PM

Acknowledgments
The editors are grateful to the Society for Music Theory for helping to make
this project possible through a publication subvention, and to the Connaught
Fund of the University of Toronto for its financial support. We also thank
Ralph Locke, Sonia Kane, Julia Cook, Ryan Peterson, and Tracey Engel at the
University of Rochester Press for having shepherded the book to completion;
Massimo Guida and Dan Deutsch for preparing the final versions of the musi-
cal examples; the anonymous readers whose numerous suggestions we endeav-
ored to incorporate; and our respective families for having put up with us
during our crabbier moments. The book is dedicated to William E. Caplin in
humble recompense for the many things we learned under his tutelage.
Toronto; Ottawa; New Haven

January 2015
Vande Moortele.indd vii 9/30/2015 7:50:47 PM

Vande Moortele.indd viii 9/30/2015 7:50:47 PM
Introduction
Steven Vande Moortele, Julie Pedneault-
Deslauriers, and Nathan John Martin
Few writers have contributed as much to the revival of Formenlehre in current

English-language music theory as William E. Caplin. From his early articles
on the eight-measure sentence (1986) and the expanded cadential progres-
sion (1987) through to his current work on cadence typologies, Caplin has
consistently challenged the field to reengage with the conceptual resources
of German Formenlehre.1 The vocabularies introduced, reintroduced, reconfig-
ured, and refined in his writings—above all in his 1998 treatise Classical Form:
A Theory of Formal Functions for the Instrumental Music of Haydn, Mozart, and
Beethoven—have become an indispensable part of every working music theo-
rist’s conceptual armature.2 The essays in this volume engage with Caplin’s
theorizing in multiple ways and from diverse perspectives, testifying both to
its centrality and to its fecundity: it is difficult to imagine contemporary music
theorists writing seriously about musical form without orienting themselves, in
one way or another, toward the framework that Caplin has elaborated.
The central preoccupation of Caplin’s scholarship has unquestionably been
the formal processes on display in the musical monuments of high Viennese
classicism, that is, in the later instrumental music of Haydn and Mozart and
in the earlier works of Beethoven. Despite the eventual centrality of this rep-
ertoire to Caplin’s career, his path to it was indirect: a dissertation project on
theories of musical rhythm, undertaken at the University of Chicago under
the direction of Philip Gossett, led him to Carl Dahlhaus at the Technische
Universität Berlin from the fall of 1976 through the summer of 1978. Though
in Berlin ostensibly to study the history of music theory, Caplin discovered
something else: for his defining encounter with the Schoenberg–Ratz line—
an encounter whose influence on Caplin’s work is measured not least in the
central category of “formal function”—was mediated by Dahlhaus. Having
decamped to Montreal in the fall of 1978, Caplin began giving regular courses
on classical form, first to undergraduates and later to graduate students. What
began as an external imposition developed first into a research project and
Vande Moortele.indd 1 9/30/2015 7:50:47 PM

2 vande moortele, pedneault-deslauriers, and martin
then into a scholarly career. By the later 1980s, various drafts of what eventually
became Classical Form were in circulation as teaching texts, and certain char-
acteristic features of these pedagogical origins were carried over into the pub-
lished treatise: the gradual, systematic exposition of theoretical concepts, the
modesty of the authorial voice, the pellucid prose. An abiding concern with
pedagogy, indeed, has been a distinctive feature of Caplin’s scholarship, and it
was in that context that the present editors (two as graduate students, one as a
postdoctoral fellow) first began to engage intensely with his work.
Formal Functions
The key term in Caplin’s theorizing is formal function. In one sense, Caplin’s
notion of formal function is very general: as defined in the glossary to Classical
Form, it is “the specific role played by a particular musical passage in the formal
organization of a work.”3 At the same time, the term has acquired a far more
concrete meaning in the gradual unfolding of Caplin’s thinking. In this latter
sense, formal functions stand conceptually opposed to formal types. Types, which
are the traditional objects of theories of form, are the easier category to grasp.
They are conventionalized concatenations of musical units arranged into stan-
dard conglomerations—constructs like compound periods, small binaries, or
recapitulations. Functions, in contrast, are the musical building blocks out of
which types are formed: basic ideas, presentations, cadential ideas, and con-
tinuations or—up one level of magnification—main themes, transitions, and
developmental cores.
To describe them only thus, however, is to give formal functions far too
neutral a cast. For the interpretive richness of Caplin’s perspective lies in the
additional characterizations that such building blocks receive. One way to get
at this added value is to ask what formal functions do. And what they do is
this: formal functions impart a highly developed sense of temporal orienta-
tion within a work’s unfolding to those who know how to hear them. Listeners
attuned to formal function, simply put, know where (or rather “when”) they
are. Quite obviously, and even trivially, such orientation is in part contextual:
the piece has just begun, or has already been going on for some time now. Such
background contextual orientation, however, is complemented in Caplin’s the-
orizing by a meticulous account of the musical devices that are habitually asso-
ciated with particular locations in musical time.
Behind these associations—and indeed enabling their elaboration—is
the crucial distinction between intrinsic and contextual function. In its pur-
est form, contextual function is in effect positional function. It is extrinsic
in the sense that it depends upon the functions of surrounding passages, as
when one argues “this is the main theme, and that is the subordinate theme,
so this passage in the middle must be the transition.” Intrinsic function,

introduction 3
in contrast, depends on markers internal to the unit in question: “This is

a presentation because it is four measures long, features a basic idea and
its immediate repetition, and prolongs root-position tonic harmony.” The
outstanding achievement of Caplin’s thinking is, in the first place, to have
clearly articulated the idea of intrinsic formal function, and then to have
fleshed out that notion through a careful delineation of the musical fea-
tures that define such functions.
The stark binaries that the preceding paragraphs have heuristically pro-
posed must, of course, be probed: neither the sharp distinction between
intrinsic and contextual function nor indeed that between function and type
is entirely satisfactory if left as a simple opposition. Continuation function,
for instance, is intrinsically defined through four markers: acceleration of
the surface rhythm, acceleration of the harmonic rhythm, fragmentation
of the grouping structure, and sequential harmony. Of these, the first three
are clearly not exclusively internal: acceleration, for instance, can only be
acceleration in relation to some previously established norm. Similarly, at
the interthematic level, subordinate-theme function is intrinsically defined
by form-functional loosening; yet it is the main-theme function that estab-
lishes, in intraopus terms, what will be tight-knit. As for function versus type,
one sees upon reflection that the distinction is hardly absolute but rather
is relative and aspectual. If a “type,” to recall, is a concatenation of formal
functions, then the smallest, most atomic units—on both the inter- and intra-
thematic levels—are only functions; the highest, most macroscopic ones only
types. Thus, basic ideas (or main themes) are formal functions only; sen-
tences (or sonata forms) are only types.4 But the intermediate structures at
each level are functions in relation to their supervening types, types in rela-
tion to their constituent functions. A presentation, for instance, is a type if
viewed as a concatenation of basic ideas, but as the first half of a sentence, it
is a function. The same holds, mutatis mutandis, on the interthematic level.
Consider, for example, an exposition: it is a type consisting of a main theme,
transition, and subordinate theme, but is the first formal function in a sonata
form as a whole. Thus the identity of these intermediate units—the question
of whether they are functions or types—shifts in accordance with the aspect
under which they are considered.
Because of this interpenetration of function and type, functions themselves
come to be colored with implicit content. They are by no means the empty
vessels of Heinrich Schenker’s anti-Formenlehre polemicizing. Rather, Caplin’s
formal functions—and therefore also his types—bundle together distinct
material characteristics: a cadential idea, for instance, might present an imper-
fect authentic cadence, or a perfect one, or any of the various half-cadential
templates that appear in classical works. The same goes for the other small-
est functional units that the theory posits: basic ideas, contrasting ideas, and
continuation functions proper. All these functions are, in essence, families

comprising interrelated schemata (in something like Robert Gjerdingen’s

sense), from which the supervening functions and types are assembled
combinatorially.5
In Perspective
One consequence of its material determination is that Caplin’s theory of formal
functions is highly idiom-specific. The theory’s very richness—its fine-grained
delimitation of the classical style—entails a corresponding loss of generality.
The subtitle of Caplin’s treatise restricts its purview to “the instrumental music
of Haydn, Mozart, and Beethoven.” The range might have been further delim-
ited, as indeed Caplin acknowledges in the preface, to Haydn and Mozart after
1780 and Beethoven up to about 1810.6 Of course, Caplin makes occasional
incursions into later Beethoven or earlier Haydn. Yet he tends to do so in a
selective and ad hoc manner. While the general notion of formal function is
obviously relevant to a much wider range of music, the theoretical apparatus
that Caplin himself develops from his more specific understanding of the term
resists immediate application beyond the repertoire that stands at the front
and center of Classical Form.
One way in which the essays in the present collection put Caplin’s work “in
perspective,” then, is that they engage with aspects of formal functionality in
repertoires reaching beyond the instrumental music of Haydn, Mozart, and
Beethoven. They run the chronological gamut from Haydn and Clementi
to Schoenberg and Leibowitz; they discuss lieder and arias, symphonies and
concerti, opera and chamber music; they range from Vienna and Paris to
Milan and beyond. As such, they inevitably “loosen” some of Caplin’s stric-
tures; they aim to adapt and expand, to open up new analytical and theoreti-
cal vistas while continuing to engage with the basic themes and commitments
of Caplin’s work.
By no means, however, does our book offer a sustained, systematic—let
alone textbook-like—extension of Caplin’s theoretical apparatus. To be sure,
such extension is one aspect of the volume; it is even a central preoccupation
of some chapters. But the perspective we offer is both broader and more mul-
tifarious. Over and above the chronological, geographical, and generic open-
ings that they seek, the contributions in this volume are in no way confined
by the methodological and conceptual boundaries of Caplinian theory. This
becomes most obvious in the concern of many chapters—implicit or explicit—
to bring Caplin’s theorizing into dialogue with other music theories. Some
authors contextualize Caplin’s theory by focusing on the theoretical tradi-
tion on which Caplin himself draws: the Schoenberg–Ratz–Dahlhaus line of
Germanic Formenlehre ; others enrich the Caplinian perspective by drawing
upon the ideas of recent writers who have made considerable contributions

introduction 5
to the development of the theory of formal functions, first and foremost Janet
Schmalfeldt; others still confront Caplin’s approach with competing theoreti-
cal models, most notably that of James Hepokoski and Warren Darcy’s “Sonata
Theory.” Nor are the perspectives limited to music-theoretical and analytical
ones: several essays add a prominent historical, philosophical, or hermeneutic
aspect—modes of intellectual inquiry that complement the theory of formal
functions as practiced by Caplin himself.
In no aspect of this book does its diversity and eclecticism become more
apparent than in the authors’ use of terminology. Rather than impose the
usage of an orthodox Caplinian vocabulary, we have allowed individual authors
to adapt and modify Caplin’s terms and concepts freely—at times to the point
where a specifically Caplinian lexicon recedes into the background. Far from
seeing this diversity as a threat to the integrity of our collection, we embrace it
as a mark of its richness. Indeed, one measure of the success of a theory, in our
view, is the number of heresies it can inspire.
❧ ❧ ❧
The thirteen chapters in this volume are grouped thematically into five pairs
and one trio. Focusing on Haydn and Mozart, the first set of chapters (part 1)
deals with two of the composers who are central to Caplin’s own project. L.
Poundie Burstein discusses what he wittily calls “functial formanality” in the
symphonies of Joseph Haydn: instances where the composer skillfully twists
conventional formal functionality by expressing it in an unusual fashion.
Deliberately toying with the contextual and intrinsic qualities of a given for-
mal unit, Haydn may alter the standard sense of formal beginnings, middles,
and ends at any hierarchical level of a formal structure. Examining these pas-
sages in light of their unusual formal functions, Burstein argues, helps high-
light their expressive and narrative effect. In “Mozart’s Sonata-Form Arias,”
Nathan John Martin surveys the sonata-form schemes that appear in Mozart’s
operas from 1780 on. By applying and adapting Caplin’s formal functions to
this operatic repertoire, Martin is able to show how, from the comparatively
ample sonata templates of Idomeneo to the much more idiosyncratic ones of Die
Zauberflöte, Mozart progressively abridged his sonata-form arias and varied their
formal structure.
The next section (part 2) turns to large-scale nineteenth-century form.
Julian Horton offers an overview of the formal functions and types on display
in the first movements of postclassical piano concerti from Dussek’s opus 14
(1791) to Schumann’s opus 54 (1845). Rethinking form-functional theory
in view of this corpus, his chapter considers the formal types that underpin
first and second themes, the balance of tight-knit and loose organization that
obtains between them, and relationships between intrathematic levels and
large-scale forms. Throughout, Horton emphasizes the delayed reception

of Mozart’s piano concerti and highlights the implications of romantic con-

certo forms for nineteenth-century form in general. Andrew Deruchie surveys
the phenomenon of cyclic form in the music of Camille Saint-Saëns within a
broadly Caplinian framework. Combining analysis of the large-scale formal
organization of many of Saint-Saëns’s instrumental works with an interpreta-
tion of their form in the musicopolitical context of France in the first decades
of the Third Republic, Deruchie’s essay contributes to what may be called a
history—rather than a theory—of musical form.
Applying Caplinian analytical categories to the music of Franz Schubert, the
next two chapters (part 3)—by Brian Black and François de Médicis—take a
more traditional approach: rather than in expanding Caplin’s analytical tool-
box, their interest lies in demonstrating something essential about Schubert’s
music. Black examines a particular strategy that underlies a number of sonata-
form transitions across Schubert’s output, transitions he identifies as the
“deflected-cadence” type. These transitions exhibit a cadential redirection in
which a cadence is initiated in the home key, evaded, then reestablished only
to be diverted suddenly into a perfect authentic cadence in the subordinate
key, the actual modulation being accomplished exclusively by that final caden-
tial progression. De Médicis, for his part, brings out Schubert’s distinctiveness
by articulating one of its aspects in form-functional terms: in Schubert’s sym-
phonic sonata forms, the contrast between the largest-scale formal functions
of beginning (i.e., exposition) and end (recapitulation) on the one hand, and
of middle (development) on the other, is mitigated in that compositional tech-
niques that are used in the classical style to articulate one function, may in
Schubert occur in a unit that, as a whole, expresses another.
The two chapters filed under “Text, Texture, and Form” (part 4) investigate
aspects of formal functionality in two very different kinds of nineteenth-century
vocal music. Harald Krebs explores the use of sentences—the theme type most
commonly associated with Caplin’s theory—in songs by Robert Schumann.
Discussing both well-known and less familiar songs, Krebs not only recalibrates
Caplin’s sentence model in order to make it applicable to mid-nineteenth-cen-
tury syntax but also investigates the ways in which Schumann’s sentences relate
to the structure of the poems he is setting and how his sentences reflect the
meaning of those poems. Steven Huebner draws on earlier work on “lyric form”
to provide a detailed account of parlante texture in Verdi’s operas. Drawing his
examples from across Verdi’s oeuvre, he provides a careful description of both
the formal organization and the dramaturgical function of such textures, show-
ing how form-functional thinking can help define them more accurately.
The next section (part 5) turns to the hermeneutics of musical form.
Offering a close analytical and interpretive treatment of interactions between
form, instrumentation, and meaning, Henry Klumpenhouwer’s essay explores
an engaging feature of the subordinate theme in the first movement of
Beethoven’s First Symphony. The feature is dramatic in character, staging

introduction 7
the commission of an error and its subsequent correction in the woodwind

section of the orchestra—a drama that involves the subordinate theme and
the closing section in the exposition and extends to the images of those for-
mal elements in the recapitulation. In addition to taking a hermeneutic turn,
Klumpenhouwer’s analysis interacts suggestively with earlier writings about the
First Symphony’s subordinate theme, not only by Caplin but also by Hepokoski
and Darcy. Giorgio Sanguinetti turns his attention to the moment of recapitu-
lation, considering cases in which its onset is elided, disguised, or attenuated.
Through close analyses of examples from Clementi, Schubert, and Brahms, he
encourages us to regard “ongoing recapitulations” as gradual processes com-
parable to Odysseus’s arrival, at night and asleep, delivered unknowingly by his
Phaeacian hosts, on the shores of Ithaca.
The last three essays (part 6) in our collection concern themselves with the
Second Viennese School, broadly construed—with the music and thought of
those composers who established the theoretical tradition on which Caplin’s
work draws. Julie Pedneault-Deslauriers reads Schoenberg’s Verklärte Nacht
in conjunction with the Richard Dehmel poem it programmatically depicts.
The skeleton key unlocking her interpretation is the idea of a “dominant
tunnel”: a characteristic harmonic device that returns at key formal articula-
tions throughout the sextet, and which she likens to Caplin’s category of the
expanded cadential progression. The two remaining chapters in this section
deepen our understanding of the prehistory of Caplin’s theory by illuminat-
ing two little-known strands of the mid-twentieth-century Schoenbergian
Formenlehre tradition. Christoph Neidhöfer and Peter Schubert write about
René Leibowitz’s Traité de la composition avec douze sons, teasing out common-
alities between Leibowitz’s and Caplin’s understandings of musical form, and
demonstrating Leibowitz’s conceptual apparatus by means of an analytical
study of the composer’s own Trois poèmes de Pierre Reverdy for four voices and
piano. In the volume’s final essay, Steven Vande Moortele identifies unex-
pected parallels between current form-functional theory—in both its Caplinian
and Schmalfeldtian guises—and the “material Formenlehre” of Theodor W.
Adorno. Bringing together the various relevant fragments scattered across
Adorno’s writings of the 1960s, Vande Moortele reconstructs and contextual-
izes Adorno’s theory of form, and illustrates it by means of Adorno’s own ana-
lytical notes on the first movement of the Eroica symphony.
Notes
1. William E. Caplin, “Funktionale Komponenten im achttaktigen Satz,” Musiktheorie
1 (1986): 239–60; and “The ‘Expanded Cadential Progression’: A Category for the
Analysis of Classical Form,” Journal of Musicological Research 7 (1987): 215–57.
2. William E. Caplin, Classical Form: A Theory of Formal Functions for the Instrumental
Music of Haydn, Mozart, and Beethoven (New York: Oxford University Press, 1998).

3. Ibid., 254.
4. This relationship is somewhat differently treated in Steven Vande Moortele,
“Sentences, Sentence Chains, and Sentence Replication: Intra- and Interthematic
Formal Functions in Liszt’s Weimar Symphonic Poems,” Intégral 25 (2011), 129–30.
5. See Robert O. Gjerdingen, Music in the Galant Style (New York: Oxford University
Press, 2007), 10–16.
5. Caplin, Classical Form, 3.

Part One
Theoretical Studies in
Haydn and Mozart

Chapter One
“Functial Formanality”
Twisted Formal Functions in
Joseph Haydn’s Symphonies
L. Poundie Burstein
Introduction
Central to William E. Caplin’s concept of formal functionality is the notion
that “musical form directly engages our temporal experience of a work inas-
much as its constituent time-spans have the capacity to express their own loca-
tion within musical time.” As Caplin explains, “a composer’s ability to realize in
a convincing manner these kinds of temporal multiplicities accounts for expe-
rienced listeners (that is, those who are familiar with the host of compositional
conventions informing this style) being able to discern quickly just where a
particular passage lies within the overall temporal extent of a work.”1
Accordingly, whether a passage functions as a beginning, middle, or end
depends not merely on where it occurs within a composition but also on its
harmonic, melodic, rhythmic, textural, and rhetorical profile. For instance, in
most cases a subordinate theme functions as such not simply because it appears
after a main theme but also because it possesses qualities that are typical of a
subordinate theme. This can readily be tested by considering a normal sonata-
form exposition and imagining how it would sound if the main theme and
subordinate theme were swapped (making the appropriate transpositions). In
most instances the results would be strikingly unusual, to say the least, and the
same would be true if one were to imagine swapping segments on higher or
lower levels of structure.
But what can be established may also be disestablished, and crafty compos-
ers at times seem to seek out the strikingly unusual by twisting conventional
formal functions. In some cases the formal function of a passage is clear, but
nevertheless realized in a deformational manner.2 Put differently, in some
situations the formal function is strongly supported by certain features, but

12 l. poundie burstein
undermined by others—for instance, when a sense of beginning that is firmly

suggested by repetition structure and gesture nonetheless is counteracted by
harmonic and rhythmic elements. Other times the departure from the norm
is so drastic that the formal function itself becomes hazy, to the extent that
it may be difficult to discern whether a given passage acts as a beginning,
middle, or end.
Instances in which a conventional formal function is undermined do
not simply represent alternate formal possibilities. Rather, they are best
understood in relation to the standard functions from which they deviate.
Caplin notes that “if a given function is actually placed differently from
its expressed temporal position—if a medial function appears as a begin-
ning, for example—a kind of formal ‘dissonance’ will result.”3 As we shall
see, at times such a formal dissonance is “resolved” as the work proceeds,
and at other times it is left hanging, as it were. In all, the formal functions
that Caplin has elucidated are not intended to “represent standards of aes-
thetic judgment, such that passages deviating from the norm are devalued
in any respect.”4 On the contrary, such deviations often contribute vitally to
a composition’s expressive effect.
Appreciating the influence of such formal divergences is particularly impor-
tant when coming to grips with the works of Joseph Haydn. Many of his pieces
employ fascinating instances of form-functional conflict and may be regarded
as exemplars of this strategy. In this essay I shall discuss selected symphonic
movements by Haydn that exhibit some of the ways in which formal functions
may seem to be turned on their head, or at least called into question. In such
instances, features strongly associated with one type of location are found at
another—for example, when something more normally associated with a mid-
dle instead sounds at a beginning, or vice versa. In some cases this involves
aspects that appear on the phrase or subphrase level, and in others the form-
functional conflicts reach to the deepest levels of the structure. Through an
examination of some specific instances of such form-functional conflicts, along
with their bearing on the structural and expressive layouts of the compositions,
I hope to shed light not only on Haydn’s own works but also on the strategy of
playing with formal functions in general.
Deviating Middles and Ends

Let us begin by considering the main theme of Haydn’s Symphony no. 59 in A
Major (“Feuersymphonie”), which presents a rather straightforward example
of a form-functional conflict. This theme consists of a sentence whose presen-
tation begins in a boisterous manner (ex. 1.1). One would expect increased
momentum in the ensuing continuation⇒cadence, especially considering
that a main theme such as this usually has a tight-knit structure. As Caplin

“functial formanality” 13
Example 1.1. Haydn, Symphony no. 59 in A Major (“Feuersymphony,” 1768),

mvt. 1, mm. 1–9. NB: In this and other examples, the registers have been
slightly simplified.
has described at length, the continuation⇒cadence of a typical sentence

conventionally involves a shortening of melodic units into smaller segments,
an acceleration of harmonic rhythm, and an increase in rhythmic activity on
the surface of the music.5 Each of these features contributes to a building of
energy as the phrase approaches its conclusion.
But no such energy gain is found at the opening of this symphony, for the
continuation⇒cadence that follows the presentation (entering a half measure
early in m. 5) radically deviates from the norm. Instead of accelerating and
fragmenting, it involves a lengthening of the melodic units into bigger seg-
ments, a slowing down of the harmonic rhythm, and a decrease of rhythmic
activity on the surface of the music, along with softer dynamics and a thinning
of the orchestration.6
The oddness of this passage results not simply from its slackened energy
but also because this slackening takes place precisely at the moment where
one most expects an upsurge in energy. As such, the effect is greater than in
the main-theme group of the first movement of Haydn’s Symphony no. 26
in D Minor (“Lamentatione”), for instance, where the sudden slowing down
(in mm. 9–12, at the start of the second phrase of the main-theme group)
occurs within the presentation of a sentence. The formal manipulation at the
outset of Symphony no. 59 seems to encourage hermeneutic interpretation,

as the dynamic thrust of the opening measures is suddenly countered by

a continuation that suggests uncertainty and wavering. This wavering is
quickly shunted aside in the next passage (mm. 10–15), however, where the
continuation⇒cadence is particularly brusque. The remainder of the exposi-
tion likewise proceeds in an unhesitating manner, as though determined to
push aside the initial doubts.
Unconventional energy-decreasing passages may be found in many other
works by Haydn, not only within a continuation⇒cadence, but on larger
levels of the form as well. A prime example occurs in the developmen-
tal core of the first movement of Haydn’s Symphony no. 80 in D Minor.
Typically, the core of a sonata-form development section expresses a large-
scale medial function, and accordingly it usually possesses a characteristic
profile similar to that of a sentential continuation (which likewise serves
a medial function, albeit on a local level). As Caplin explains, “the core
of the development typically projects an emotional quality of instability,
restlessness, and dramatic conflict. The dynamic level is usually forte, and
the general character is often one of Sturm und Drang. The core normally
brings a marked increase in rhythmic activity projected by conventionalized
accompaniment patterns.”7
But this typical layout is not to be found in the first movement of Symphony
no. 80. The main theme of this movement opens in as fiery a manner as the
one from the “Feuersymphonie” cited above. Also as in the “Feuersymphonie,”
the continuation of the main theme in Symphony no. 80 slows down (although
here only momentarily) before plunging forward with an aggressive force that
continues for most of the exposition. The intensity abates only at the arrival of
the exposition’s codetta, which introduces a new theme that is entirely out of
character with what was heard in the movement thus far. This theme, in the
manner of a Ländler, is strangely lighthearted and comical, in great contrast to
the tempestuous themes that precede it.8
Of all the themes in the movement, the Ländler theme is the only one
that would be out of place in a development section. Ironically, it is pre-
cisely this theme that dominates the precore and core of the ensuing devel-
opment (ex. 1.2). Although within this section there are some instances
of fragmentation, sequential motion, and Sturm und Drang, in each case
these more typically developmental passages are interrupted by the reap-
pearance of the Ländler theme at a piano dynamic level. For the most part
the Ländler returns unchanged, except for its transposition to a series of
different keys, and before each entrance the momentum is further arrested
by a grand pause. To be sure, the extreme modulations here would be out
of place in any other section but a development. Still, the rhythmic pace
and textures here strongly seem to contradict their developmental setting,
creating a humorous effect.

After outlining harmonic motions up and down by thirds (see the annotations
under the voice-leading sketch of ex. 1.2), the developmental core concludes in
measures 120–24 by presenting the Ländler once again in the key of F major. This
is the key in which it made its first appearance at the end of the exposition in
measures 57–64. The notion that the Ländler’s appearance in measures 120–24
represents a return to its original state (of mm. 57–64; see the dotted slur in the
voice-leading sketch of ex. 1.2) is reinforced by the similar orchestration of these
passages (other statements of the tune each have slightly different instrumental
settings). As a result, there is a sense that the development section has not ulti-
mately progressed anywhere, but rather has mostly circled around the secondary
key of F. The underlying stasis is not fully challenged until the middle of the
recapitulation, where a forceful standing on the dominant in measures 136–46
leads to the large-scale return of the tonic in measure 147. The varied return of
the main theme in measures 128 and following, which is only slightly prepared
by a brief retransition based on a V7 in second inversion, seems to be swallowed
(in the manner of a parenthetical insertion) within this larger tonal motion.9
In all, despite its large-scale harmonic restlessness, the surface of this devel-
opmental core suggests stability and lack of conflict. That Caplin’s descrip-
tion, as cited above, of the standard features for a development section fits
ill with much of what happens in this symphony by no means indicates a flaw
in his approach, however. On the contrary, an understanding of conventional
formal functions helps contextualize unusual situations such as those seen
in Symphony no. 80 by underlining how they work in dialogue with stylistic
norms.10 Again, what is significant here is not just that an unusual theme
appears in this movement, but that it appears in such a way as to strongly dis-
rupt standard formal functionality. As a result, this development section seems
to express a type of assertion of human will against convention, as though
refusing to engage in the violent emotions expected of it.
Example 1.2. Symphony no. 80 in D Minor (c. 1784), mvt. 1: voice-leading sketch

Deviating Beginnings
The examples discussed above involve deformations that entail middles and
ends. In other instances, formal “dissonances” arise at sectional beginnings,
as in the category that Caplin describes as themes lacking an initiating func-
tion.11 This category usually involves subordinate themes. However, since
they have a large-scale medial function and typically are loosely knit anyway,
subordinate themes that lack an initiating function are quite common and
usually not so disruptive. Far less normal is a main theme that lacks an initiat-
ing function, since a main theme conventionally is tight-knit and has a large-
scale beginning function.
Main themes that lack an initiating function, or in which the initiating func-
tion is somehow problematic, nonetheless may be found in many of Haydn’s
works.12 For instance, consider the first movement of his Symphony no. 65 in
A Major, where the odd main theme opening seems to have repercussions that
extend deep into the movement. This symphony commences with a “noise-
killer” series of three chords that serve as a thematic introduction to the main
theme (ex. 1.3a). As Caplin explains, a thematic introduction “is generally
short, two to four measures at most. . . . The melodic-motivic component of
such an introduction is either weakly defined or entirely absent, so that the
expression of a genuine basic idea can be saved for the structural beginning
of the theme.”13 This description certainly could apply to what is found in
measures 1–2 here, except that the “genuine basic idea” that one expects to
follow is missing. Instead, the main theme proper starts immediately with a
continuation⇒cadence, as unstable harmonies and active rhythms suggest that
these measures begin as though in the middle of a larger gesture. Since it lacks
a clear presentation, this theme gives a somewhat nebulous impression, which
is abetted by the odd harmonic regressions in measures 4–5 and 8–9.
One might even wonder whether the gesture of measures 1–2 could instead
be retrospectively regarded as the presentation of a phrase that extends from
measures 1 to 6 (see the parenthetical annotation above the first measure in
example 1.3a). Yet if measures 1–2 were a presentation, they would be a most
unusual one, for their rhetoric strongly suggests a sense of “before-the-begin-
ning.” Furthermore, the gesture of measures 1–2 is texturally, dynamically,
and melodically so starkly separated from what follows in measures 3–6 that
if measures 1–6 are to be regarded as a unit, the resulting phrase would be an
extraordinarily disjointed one.
The odd main theme of this movement is counterbalanced by the subse-
quent themes, in which the formal functions are presented in a straightfor-
ward and unambiguous fashion. These subsequent themes may be regarded as
variants of the main theme, and as such they may be said to resolve the form-
functional dissonance of the movement’s opening. For instance, the theme of
measures 19 and following (ex. 1.3b) is a clear-cut sentence with an extended,

Example 1.3a. Symphony no. 65 in A Major (1769), mvt. 1, mm. 1–11

(EC = evaded cadence; PAC = perfect authentic cadence)
Example 1.3b. Symphony no. 65, mvt. 1, mm. 19–23: variant of main theme that
forms a clear-cut beginning of a sentence presentation
sequential continuation. A variant of measures 3–4 serves as the presentation of

this phrase, but now with its harmony and rhythm normalized. An allusion to the
opening three-note noise-killer gesture of measures 1–2 appears here as well,
now demoted to serving as part of the bass within the continuation⇒cadence
in measures 23 and following. The passage that leads to the conclusion
of the exposition (mm. 37ff., ex. 1.3c) provides further ironing out of the

Example 1.3c. Symphony no. 65, mvt. 1, mm. 37–40: further variant of main theme
forms part of extremely tight-knit phrase
main theme, as it presents yet another variant of this theme in the form of an
exceedingly tight-knit phrase. In all, the evolution of materials derived from
the main theme unfolds a type of narrative through the course of the exposi-
tion, in which seemingly unbridled thematic material is ultimately converted
into a theme whose formal functions are extremely orderly.
In other works by Haydn as well a form-functional dissonance presented
toward the outset of the movement is resolved as the movement proceeds.
A particularly celebrated example may be found in the first movement of
Haydn’s Quartet for Strings in G Major, op. 33, no. 5. The opening gesture of
this movement in measures 1–2 appears as a preamble to the main theme, and
it displays what Caplin describes as the content—but not the function—of a V–I
cadence.14 In subsequent passages, this gesture is placed in its more “proper”
position at the end of a phrase, where it has both cadential content and caden-
tial function (see, for instance, mm. 9–10 and 31–32).15
A somewhat similar but more extended example of this strategy arises in the
first movement of Haydn’s Symphony no. 90 in C Major. The exposition of this
movement opens literally with a passage that has a medial function, for the main
theme starts in the midst of the continuation of an oddly proportioned sentence
that had already begun within the Adagio introduction (ex. 1.4a).16 Such func-
tional conflicts appear continually throughout the movement: that is, almost
every time the main theme returns, it commences in the middle of a formal unit,
harmonic progression, or both. The only exception arises toward the end of the
movement, where the theme finally appears within its own phrase, although still
containing a harmonically unstable opening (mm. 218ff., ex. 1.4b). Not until the
very end of the movement is the gesture from the opening two measures put in
its “proper” formal place (see brackets in ex. 1.4b), as though to finally resolve
the tensions created by the form-functional disparities found throughout.17
Sometimes form-functional conflicts involve both beginning and endings, as
may be witnessed in the Trio from the third movement of Haydn’s Symphony
no. 64 in A Major (“Tempora mutantur”). This section begins with what
sounds reminiscent of a cadence, but which then turns out to function as the
basic idea at the start of a sentence (ex. 1.5). In other words, in the manner
discussed above, this opening has cadential content, not cadential function.
Matters become confusing when this theme returns in the last phrase of this

Example 1.4a. Symphony no. 90 in C Major (1788), mvt. 1, mm. 14–20: main
theme of exposition begins in the middle of a larger phrase (strings parts only)
Example 1.4b. Symphony no. 90, mm. 217–74: main theme appears within a
separate phrase; then opening motive appears as part of a cadence
small-ternary Trio, for it is somewhat ambiguous whether the outset of the A‫׳‬
section in measure 41 serves as a beginning, middle, or end. The contrasting
middle (mm. 33–40) that precedes the thematic return in the A‫ ׳‬section is
framed by a clear-cut sentence—or rather, what would be a clear-cut sentence
were it not derailed at its conclusion. At the end of this contrasting middle
there seems to be a half-cadential V decorated by appoggiaturas (m. 40).
Curiously, however, the appoggiaturas do not resolve until the start of the next
phrase, by which point the Trio’s main theme has already returned (now start-
ing directly with a V7 rather than with an anacrusis arpeggiation of the tonic
triad, cf. mm. 25 and 41).
Although measure 40 is probably best understood as a cadential endpoint,
the unresolved appoggiaturas in this measure nonetheless suggest that it lies in

Example 1.5. Symphony no. 64 in A major (“Tempora mutantur,” 1774), mvt. 3,
Trio, mm. 25–48

the middle of a larger harmonic and contrapuntal gesture. Likewise, although

the thematic return in measure 41 implies that this measure functions as a
formal beginning, there is a sense that it, too, appears in the middle of an over-
riding harmonic-contrapuntal motion. There is even a momentary hint that
measures 41–44, with their allusion to cadential content, provide an authentic
cadential end for the phrase that began in measure 33.18 The actual conclu-
sion of the Trio is itself somewhat muddied as well, for when the Menuetto
is repeated following the Trio section, its opening measures initially sound as
though they might serve as a codetta to the Trio.19 The various temporally dis-
combobulated moments in the Trio might well relate to the seeming preoc-
cupation with the disjointedness of time found throughout this symphony.20
The First Movement of Symphony

no. 81 and the Quiescenza
The first movement of Haydn’s Symphony no. 81 in G Major provides an
example of form-functional conflict on a yet grander scale. The main theme
of this movement unfolds in a fragmentary, almost impressionistic manner.
The sense of dislocation is heightened by the theme’s harmonic framework,
which is based on what at the time was a standard “after-the-ending” schema
involving a I–V7/IV–IV–V–I progression—typically placed over a static tonic
pedal point in the bass—that supports the implied melodic figure 8–♭7–
6–♮7–8 (ex. 1.6a). Robert Gjerdingen has dubbed this standard pattern a
“Quiescenza.” As Gjerdingen points out, the Quiescenza was commonly used
as an opening gesture in earlier eras, but by the middle of the eighteenth
century it had been established as a “stock schema employed for closing
rather than opening passages. . . . Just as a cadenza exploits a pause within
an important cadence . . . , so a Quiescenza exploits a moment of quiescence
following an important cadence.”21 To demonstrate its ubiquity, Gjerdingen
cites numerous examples where the Quiescenza appears at the ends of move-
ments or large sections of movements in works by C. P. E. Bach, François-
Joseph Gossec, Johann Baptist Wanhal, and other composers who flourished
in the mid-eighteenth century and later.
Example 1.6a. Quiescenza figure: abstract example

Haydn likewise often took advantage of the Quiescenza as an ending for-

mula, especially in his symphonies. Over a quarter of them contain at least one
movement in which a Quiescenza figure appears close to the end, including
the first movements of his Symphonies no. 8, 58, and 98, the second move-
ments of his Symphonies no. 36 and 39, the Menuetto of his Symphony no. 94,
and the Finales of his Symphonies no. 43, 73, and 101 (example 1.6b excerpts
one such ending).22
There are only three symphonic movements by Haydn that clearly employ
the Quiescenza within the stylistically less normative position at the very open-
ing of the main theme.23 One of these is Symphony no. 59, first movement
(cited in ex. 1.1 above), in which the typical sense of a Quiescenza figure is
powerfully undercut by a frenetic bass line that counteracts any sense of com-
ing to a close. Another is the second movement of Symphony no. 62 in D
Major, where the bass line is relatively pacific and in which the Quiescenza
opening sets the stage for an exceptionally serene movement, one that features
unusually long stretches of harmonic stability.24 The final one is Symphony no.
81, first movement, where a classic Quiescenza—appearing over a tonic pedal
point—opens not merely an internal movement, but indeed the entire compo-
sition (ex. 1.7a).25
To be sure, a Quiescenza figure does not necessarily always evoke a sense of
an ending, even when it appears in works from the later 1700s. This is especially
so when the use of the Quiescenza is strongly counterbalanced by rhythmic,
textural, and topical devices that strongly suggest a beginning, as is the case,
for instance, in the first movements of Haydn’s aforementioned Symphony no.
59 or the Piano Sonata no. 52 in E-flat Major.26 But no such counterbalancing
takes place in the opening of the first movement of Symphony no. 81; on the
contrary, as noted above, the rhythmic and textural features at the outset of
this symphony tend to work against a clear sense of beginning. As a result, not
Example 1.6b. Symphony no. 43 in E-flat Major (c. 1771), mvt. 4, mm. 196–202,
“after-the-end” passage based on Quiescenza schema

until its last few measures does it become apparent that this movement’s first
phrase serves as an antecedent, as a part of what turns out to be a period that
embraces the main theme. That the opening measures of this symphony could
serve as a stylistically suitable ending is supported by the return of the opening
material—only slightly modified—as the actual ending of the entire movement
(cf. exx. 1.7a and 1.7b). In these final measures, the Quiescenza serves effec-
tively in its standard role as an “after-the-end.”27
It is instructive to compare the beginning of this symphony with that of
Haydn’s Symphony no. 61 in D Major (ex. 1.8). The gestures used at the start
of Symphony no. 61 are practically clichéd signals of an opening, as the noise-
killer downbeat chord is followed by the arpeggiation of the tonic harmony
and then by the standard melodic figure that Gjerdingen labels as a “Prinner.”
The conventional schema found within this main theme—along with its more
self-assured melodic and textural profile—helps place the more unusual main
theme of Symphony no. 81 in relief: whereas Symphony no. 61 begins with
what is unmistakably well suited for an opening, the beginning of Symphony
no. 81 seems to send mixed signals.
The relative instability of the opening of Symphony no. 81 in turn poses
challenges for the recapitulation. Naturally, it is most typical for a main theme
to reenter at the start of a movement’s final large section that both begins and
Example 1.7a. Symphony no. 81 in G Major (1784), mvt. 1, mm. 1–12: beginning
of movement

Example 1.7b. Symphony no. 81, mm. 167–79: end of movement
Example 1.8. Symphony no. 61 in D Major (c. 1776), mvt. 1, mm. 1–8; cf. ex. 1.7a

ends in the tonic key.28 The return of the main theme at this juncture usually
marks the onset of the final rotation through the Anlage, that is, the basic the-
matic plan first presented within the exposition (see table 1.1).29 The simulta-
neous initiation of both the final large tonal section and the final presentation
of the Anlage usually is a dramatic highpoint of a movement, one that aptly may
be described with the term “point of recapitulation.”
Not all movements that otherwise conform to the sonata-form model are
constructed in this manner, however. In some cases, the final presentation of
the Anlage begins prior to the definitive return to the tonic key, so that the
onset of the final tonal section appears in the middle of the unfolding of the
Anlage (see table 1.2).30 This strategy is particularly amenable to those move-
ments in which a sense of beginning is undermined within the main theme:
after all, it is not easy to have a grand restart of a theme that even in its first
iteration started in a somewhat tentative manner.
Table 1.1. A standard major-key sonata-form framework
Exposition Development Recapitulation

1st Periode 2nd Periode, 1st half 2nd half
Begins in tonic key, Begins in key of V, often ends Begins and ends in
ends in key of V with half cadence, either in tonic key
the tonic key or in another
key (such as vi, iii, or ii)
followed by a recapitulation
First presentation Possible second presentation Final, slightly
of Anlage of Anlage, starting with main condensed
(movement’s basic theme, loose-knit and often presentation of the
thematic plan) incomplete Anlage, starting with
the main theme in the
tonic key
Table 1.2. An alternate sonata-form framework
First part Second part (no clear point of recapitulation)

1st Periode 2nd Periode, (retransition) 2nd half
1st half
First presentation ----- Second and final presentation of

of Anlage Anlage, beginning with main theme
(often starting in tonic key) and
continuing through the end of the
movement. This often embraces
“developmental” passages.

Among the works with tentative-sounding main themes whose layouts

conform to the pattern depicted in table 1.2 are Haydn’s Symphony no.
65 (whose unusual main theme was discussed above) and his Symphony
no. 81. Table 1.3 parses the first movement of Symphony 81 according to
its cadential and thematic structure. As is depicted here, the return of the
main theme that initiates the final statement of the Anlage (m. 73) as well
as the start of the final tonal section (m. 124) appears within the midst of
larger structures.31 Accordingly, the beginning of the final rotation through
the Anlage (in m. 73) is recognized only in retrospect, as it sneaks in within
the middle of a tonally unstable developmental passage (much as do those
moments that resume the Anlage within the retransition in mm. 94, 101,
and 111). Likewise, the final tonal section begins relatively unheralded, as
its entrance coincides with the return of the transition theme (in m. 124;
cf. mm. 24ff.), which clearly has a medial function and lies within the larger
unfolding of the Anlage.32
When confronting a situation such as this, any attempt to isolate one
moment as the official “point of recapitulation” arguably proves more of a
hindrance than an aid. Although one could certainly put forth reasonable
justifications for designating one moment or the other as the “official” onset
of the recapitulation, it is unclear what would be gained as a result.33 No mat-
ter where the recapitulation is located, it remains that any moment labeled
as the beginning of a recapitulation here lies in the middle of something and
thus would give rise to form-functional conflicts.
A double return of the opening of the main theme and main key is
delayed until the coda, where the main theme in the exposition now fol-
lows the transition and subordinate theme. To label the situation here as a
“reversed recapitulation” would be an incomplete description at best, how-
ever.34 Although the content of the main theme returns at the end of the
movement, its function does not. At the beginning of the movement, the
opening passage functions (however uncomfortably) as a main theme; at
the end of the movement, the analogous passage functions as a coda. As a
result, the formal function of this material has changed from behaving as
a beginning to serving in what would seem to be its more proper role as an
“after-the-ending.”35
In all, the appearances of the main theme throughout this movement
may be regarded to follow a type of narrative progression. At the start of
the movement, the main theme is forced to play the role of an opening ges-
ture, as it were. In the second half of the movement, the theme comes back
in the guise of a medial gesture, and thus still in conflict with its “natural”
role. Only at the end of the movement does the theme appear in its “proper”
place as a concluding gesture, as though to resolve the formal conflicts that
had persisted throughout.

Vande Moortele.indd 27
Table 1.3. Parsing of Symphony no. 81 in G Major, first movement
m. 1 m. 68 m. 73 m. 94 m. 124 m. 167
1st Periode 2nd Periode, 1st half (retransition) 2nd half appendix (coda) to
Starts in G, Eventually cadences on V of ii followed by Leads from V of ii Starts and end 3rd Periode
cadences standing on the dominant back to G major in G
in D
First Second and final presentation of Hint of main theme Unfolding “Content” of
presentation Anlage begins at nearly original starting in m. 94 and of Anlage main theme (but
of Anlage pitch level, but now within m. 101; latter part of continues with without main theme
motion to C; unfolding of Anlage main theme (cf. mm. the transition function)
interrupted by developmental 7ff.) hinted at in mm. theme (cf.
passages in mm. 79ff. 111ff. mm. 24ff.)
9/30/2015 7:50:56 PM
The Second Movement of Symphony no. 65:

The Unrepentant Formal Function
Many of the movements discussed above have been described according to a
similar narrative trajectory in which a problematic formal element presented
toward the beginning returns later in a normalized fashion, as if to resolve
the functional conflict presented earlier. But this narrative paradigm does
not fit in every case: in some works, a formal dissonance presented in the
opening measures of a piece remains unresolved at the movement’s end.
Such is the case in the final work to be examined in this essay, the Andante
second movement of Haydn’s Symphony no. 65 in A Major. Unlike in the
first movement of this symphony (discussed above; see ex. 1.3), the form-
functional conflicts in this Andante are never fully mollified; if anything, they
intensify as the piece progresses.
The main theme of this movement seems framed by a variant of a sen-
tence (ex. 1.9a), except that the presentation is missing (much as it was in
the main theme of the first movement; see ex. 1.3a above). As a result, the
theme begins with the strings playing a continuation⇒cadence in measures
1–3 (see top line of annotations in ex. 1.9a). The oboes and horns answer
this in measures 4–5 with a postcadential extension, which—as Caplin notes
in describing this formal function—“prolong[s] the final harmony of [the]
preceding cadence [while] the energy accumulated in the motion towards
the cadential goal is dissipated.”36
When the theme repeats in measures 6–8, however, a new interpretative
possibility arises: might the gesture in the oboe and horns retrospectively be
understood as a presentation for a phrase that extends from measures 4 to 8
(see italicized line of annotations in ex. 1.9a)? It would be an oddly propor-
tioned phrase to be sure, and the contrast between the presentation and the
continuation⇒cadence would be far greater than typical, yielding a rather frag-
mented structure. The passage in measures 13–16 presents a similar possibility
for reinterpretation: is the tutti outburst of these measures to be regarded as
the contrasting middle of a small ternary form, with the opening section of the
main-theme group returning in measure 17? Or are measures 13–16 retrospec-
tively to be understood as the presentation of a sentence that extends from
measures 13 to 19? Here, too, if these measures are indeed to be considered
a sentence, the resulting phrase would be extraordinarily disjointed. But then
again, so much else of this movement is so extraordinarily disjointed that the
possible presence of such an oddly balanced and fragmented phrase cannot
automatically be ruled out.
The main-theme group of the exposition concludes in measures 20–33 with
what likewise may be regarded as a sentence (see annotations above the staff
in ex. 1.9b). But what a bizarre sentence it is! Its proportions are peculiar, and
the segments that compose the phrase involve even greater contrasts than in

Example 1.9a. Symphony no. 65 in A Major (1769), mvt. 2, mm. 1–19: first parts of
main theme group
the preceding sections. Furthermore, the formal functions of the elements

heard at the outset of the movement are now switched around in their new
contexts: the flippant opening gesture from measure 1, played pianissimo, is
now wedded to a somber, forte unison passage to form the basic idea of this
phrase’s presentation. Consequently, the oboe and horn gesture that reap-
pears in measures 32–33 is changed from its previous role as a postcadential

Example 1.9b. Symphony no. 65, mm. 20–33: end of main theme group
extension (or, as noted above, was it a presentation?) to now function as a

continuation⇒cadence.
Or perhaps there is another way to understand measures 20–33. As sug-
gested in the italicized line of notations within the staff of example 1.9b, might
measures 20–31 be regarded as a weirdly expanded variant of measures 1–3?
If this is so, then these measures might be regarded in their previous manner
as a continuation⇒cadence, with what follows in measures 32–33 acting as a
postcadential extension.
Such ambiguities are not clarified as the movement proceeds. On the con-
trary, in the second half of the movement the functions are even more con-
fusing, as the seemingly erratic interruptions increase both in number and
in brazenness. The larger formal functions are caught up in the confusion
as well, for it is not completely clear where the recapitulation (or, for that
matter, the final run-through of the Anlage or the movement’s final tonal sec-
tion) begins. This is not to deny that one could make a good case for one
large-scale formal parsing or the other. Nevertheless, however this movement
is parsed, the formal divisions on all levels of structure surely are far murkier
than is typical.
The movement’s eccentric nature might well suggest a programmatic
underpinning. It calls to mind Haydn’s oft-cited remark that in one of his
early symphonic movements he attempted to portray an exchange between
God and a “foolish sinner,” in which “the dominant idea [is] how God speaks
to an unrepentant sinner, and pleads with him to reform, but the sinner in

his foolishness pays no heed to the exhortations.”37 Although Haydn did not
reveal which symphonic movement was based on this program, a number of
scholars have ventured reasonable guesses regarding the identity of the work,
including movements from his Symphonies no. 7, 22, 26, and 28.38 Haydn
was quite emphatic in stating that he used this specific program only once;
nevertheless, that there are so many viable candidates suggests that the gen-
eral narrative paradigm that underlies this program was not an uncommon
one for him.39
Its late dating argues against the second movement of Symphony no. 65 as
possibly being the unnamed early symphonic movement that Haydn specifi-
cally related to the God/sinner program. Still, this work does seem to loosely
follow the basic outline in which an authoritative and a flippant idea are in
dialogue with one another, and in which the flippant one remains steadfastly
“unreformed.” Throughout this movement the wayward opening alternates
with passages imbued with gravitas. At times these conflicting ideas seem to
combine to form phrases of sorts, although they never quite meld success-
fully. Unlike in many of the movements discussed above, the formal prob-
lems that result are not “solved” by the end of the movement: the unruly
opening gesture is never placed within its “proper” formal setting so as to
resolve the form-functional dissonances that it inspires. Rather, it seems to
spread further confusion as it continues, helping to muddy up the sense of
the large-scale formal design. At the end of both halves of the movement
(mm. 46–56 and 132–45), the initial part of the opening gesture appears to
celebrate its independence with a merry little dance. This dance concludes
by appropriating the unison texture and loud dynamics of the movement’s
more serious gestures (see mm. 54–56 and 143–45), as though to wrap things
up by thumbing its nose at authority.
Haydn himself may be regarded as a figure of authority, one who helped
establish the artistic standards of his era. As his early biographer Albert
Christoph Dies put it, “lucid arrangement, lucidus ordo, is not the least of
Haydn’s excellences.”40 Few equaled his skill at handling musical forms or his
ability—when he so desired—to compose convincing beginnings, middles, and
ends. But as the examples cited here suggest, Haydn did not always wish to do
so: sometimes, in the service of musical expressivity, he, too, seemed to enjoy
thumbing his nose at authority.
Naturally, the devices examined here may be found in works of other
composers as well, though rarely with the skill and powerful effect as wit-
nessed in the output of Haydn. As is suggested by this and the other exam-
ples discussed above, form-functional conflicts can take on various guises and
involve various parameters. For instance, sometimes it is a rhythmic feature
that seems at odds with its location within the formal design, and sometimes
it is a textural element, melodic figure, or harmonic element that appears to
be out of place.

In certain instances, the effect of the dislocation is rather subtle; this is par-
ticularly true when a single feature more characteristic of an ending seems to
have been placed as a beginning, or vice versa. At other times, the playing with
formal functions seems to create outright confusion, as in cases where several
elements more suited for a middle section appear at the outer edges of the
form. In such cases especially, an appreciation of the ways in which the music
departs from conventional formal functions can contribute vitally to a deep
understanding of the composition’s structural and hermeneutic implications.
Notes
1. William E. Caplin, “What Are Formal Functions?” in William E. Caplin, James
Hepokoski, and James Webster, Musical Form, Forms & Formenlehre: Three Methodological
Reflections, ed. Pieter Bergé (Leuven: Leuven University Press, 2009), 23, 25; see
also Michel Vallières, Daphne Tan, William E. Caplin, and Stephen McAdams,
“Perception of Intrinsic Formal Functionality: An Empirical Investigation of Mozart’s
Materials,” Journal of Interdisciplinary Music Studies 3, no. 1–2 (2009): 17–43.
2. The concept of deformation as an agent of musical meaning and significance is
most closely associated with the works of James Hepokoski and Warren Darcy, as
exemplified in their book Elements of Sonata Theory: Norms, Types, and Deformations
in the Late-Eighteenth-Century Sonata (New York: Oxford University Press, 2006). As I
shall argue here, this concept is compatible with the theory of formal functions as
espoused by Caplin as well.
Music of Haydn, Mozart, and Beethoven (New York: Oxford University Press, 1998),
111. Caplin describes a number of form-functional deviations and reinterpretations
throughout his writings.
4. Ibid., 4.
5. Ibid., 40–42. Regarding the relationship of the opening measures of this passage to
the Quiescenza schema, see the discussion of Haydn’s Symphony no. 81 below.
6. The deceleration is further attenuated when the main theme returns in the reca-
pitulation (mm. 79–89), where the continuation⇒cadence is drawn out to an
even greater extent. See also comments regarding the unusual nature of this main
theme in Elaine Sisman, “Haydn’s Theater Symphonies,” Journal of the American
Musicological Society 43, no. 2 (1990): 342–43; and A. Peter Brown, The Symphonic
Repertoire, vol. 2, The First Golden Age of the Viennese Symphony: Haydn, Mozart,
Beethoven, and Schubert (Bloomington: Indiana University Press, 2002), 113.
8. The striking use of the Ländler theme in this symphony has been insightfully dis-
cussed by various commentators, including James Webster, Haydn’s “Farewell”
Symphony and the Idea of Classical Style: Through-Composition and Cyclic Integration in his
Instrumental Music (Cambridge: Cambridge University Press, 1991), 167; and Elaine
Sisman, “Haydn, Shakespeare, and the Rules of Originality,” in Haydn and His World
(Princeton, NJ: Princeton University Press, 1997), 30–32.
9. As Sisman puts it, “the ostensible beginning of the recapitulation in fact turns that
material (m. 128) into a retransition with lengthy dominant pedal, and turns the

original F-major theme into the ‘real’ return [in m. 147], in D major”; see Sisman,
“Haydn, Shakespeare, and the Rules of Originality,” 31–32. See also comments
about this recapitulation in Webster, Haydn’s “Farewell” Symphony, 167; W. Dean
Sutcliffe, “Haydn Seek,” Musical Times 134, no. 1806 (1993): 447; and Brown, The
Symphonic Repertoire, 2:204. In Mozart, Haydn and Early Beethoven, 1781–1802 (New
York: Norton, 2009), 354, Daniel Heartz suggests that the recapitulation actually
begins in measure 147; depending on how one defines the term “recapitulation” (a
relatively modern term, unknown to eighteenth-century musicians), Heartz’s label-
ing might indeed be an appropriate one.
10. Caplin himself claims that his theoretical method is compatible with a dialogic
approach to form; see William E. Caplin, “Comments on James Hepokoski’s Essay
‘Sonata Theory and Dialogic Form,’” in Caplin, Hepokoski, and Webster, Musical
Form, Forms & Formenlehre, 90.
11. Caplin, Classical Form, 111–15.
12. Caplin discusses the category of main themes that lack an initiating function in
ibid., 199–201.
13. Ibid., 15.
14. William E. Caplin, “The Classical Cadence: Conceptions and Misconceptions,”
Journal of the American Musicological Society 57, no. 1 (2004): 81–85.
15. As a number of commentators have noted, the syntactic twisting of cadential clos-
ing gestures in Haydn’s works frequently has a witty or humorous effect; see, for
instance, Gretchen A. Wheelock, Haydn’s Ingenious Jesting with Art: Contexts of Musical
Wit and Humor (New York: Schirmer, 1992), 98–115; and Scott Burnham, “Haydn
and Humor,” in The Cambridge Companion to Haydn, ed. Caryl Clark (Cambridge:
Cambridge University Press, 2005), 71–72. For the use of closing gestures at open-
ings in general, see Norma Sherman-Ishayek, “Closing Gestures in Opening Ideas:
Strategies for Beginning and Ending in Classical Instrumental Music” (masters the-
sis, McGill University, 1991).
16. This sentence is itself a varied repetition of what appears in measures 1–8 of the
introduction. This main theme, along with its unusual formal function, is discussed
in Caplin, Classical Form, 199–200. I also discuss this movement and its formal con-
flicts in L. Poundie Burstein, “Comedy and Structure in Haydn’s Symphonies,”
in Schenker Studies, ed. Carl Schachter and Hedi Siegel (Cambridge: Cambridge
University Press, 1999), 2:78–81.
17. Caplin himself addresses the notion that a formal dissonance of an unusual
opening can be resolved by subsequent passages; see William E. Caplin, “Mozart,
Symphony No. 40 in G Minor, K. 550, II,” in A Composition as a Problem II, ed. Mart
Humal (Tallinn: Eesti Muusikaakadeemia, 1999), 155–62, where he notes (on p.
162) that in this work “Mozart sets up a conflict between an implied formal func-
tion and an actual formal placement—an ‘ending’ gesture occurring at a begin-
ning. . . . The conflict between function and placement is eventually resolved, not
surprisingly, in the recapitulation.”
18. Compare the formal twists in this symphony to those found in the Trio from the
third movement of Mozart’s Symphony in C (“Jupiter”), K. 551; regarding the
formal manipulations in the Mozart movement, see Jonathan Kramer, The Time of
Music: New Meanings, New Temporalities, New Listening Strategies (New York: Schirmer,
1988), 143–44; and Caplin, “The Classical Cadence,” 83–85.
19. This notion is bolstered by the return of the material from the first four measures of
the Menuetto as a codetta in the final measures of this section, following a perfect

authentic cadence in the tonic key (cf. mm. 1–4 and 21–24). Such form-functional
conflicts that span the divide between a Menuetto and Trio may be found in other
works by Haydn as well, as is noted in James Webster, “Haydn’s op. 9: A Critique of
the Ideology of the ‘Classical’ String Quartet,” in Essays in Honor of László Somfai on
his 70th Birthday: Studies in the Sources and the Interpretation of Music, ed. László Vikárius
and Vera Lampert (Lanham, MD: Scarecrow Press, 2005), 139–57 (esp. 144).
20. Regarding the possible programmatic implications of Symphony no. 64, see
Sisman, “Haydn’s Theater Symphonies,” 326–31; and Jonathan Foster, “The
Tempora Mutantur Symphony of Joseph Haydn,” Haydn Yearbook 9 (1975): 328–29.
21. Robert O. Gjerdingen, Music in the Galant Style (New York: Oxford University Press,
2007), 183.
22. Naturally, locating passages based on the Quiescenza schema relies on interpreta-
tion, and it is quite possible that others could reasonably come up with a slightly
different list of those symphonies by Haydn that conclude with this figure (which
is why I resist providing a complete list). Nonetheless, surely any fair accounting
would agree with the essential point put forth here: namely, that the Quiescenza is
a relatively common concluding figure in Haydn’s symphonic movements, but that
it is rarely found as an opening gesture in these works.
23. In a few cases, a Quiescenza is found in the second phrase of a main theme: see, for
instance, the first movements of Symphony “B” or Symphony no. 46, or the Finale
of Symphony no. 22.
24. Passages that are directly built either upon the Quiescenza figure or over a pedal
point in the bass take up over a third of the movement, including the entire core of
the development section.
25. The presence of the Quiescenza figure in Symphony no. 81 is discussed in
Markus Neuwirth, “‘Verschleierte’ Reprisen bei Joseph Haydn: Über einige
Fragwürdigkeiten eines anachronistischen Sonatenform-Paradigmas,” in Joseph
Haydn (1732–1809) [= Memoria, vol. 11], ed. Sebastian Urmoneit (Berlin: Weidler,
2009), 60–62. As H. C. Robbins Landon observes, one of the remarkable things
about this symphony is its “marvelous beginning, unique in the whole of Haydn’s
symphonic art”; see H. C. Robbins Landon, The Symphonies of Joseph Haydn (New
York: Macmillan, 1956), 393.
26. As Caplin claims regarding topoi, “if we can identify that a given topic is displaced
from its conventional formal position, yet the topic also displays musical charac-
teristics that are suitable for the formal position it actually occupies, there is little
reason to believe that the composer is toying with our expectations on the rela-
tion of topic to form, even if that relation is not as typical as some other one”;
see William E. Caplin, “On the Relation of Musical Topoi to Formal Function,”
Eighteenth-Century Music 2, no. 1 (2005): 121. The same surely would hold true for
the use of schemata: no single device can by itself establish a sense of beginning
or ending. I would add only that the sense of formal displacement or lack thereof
depends on the force of the formal associations of the topoi or the schemata as
well as on the distinctiveness of the “musical characteristics that are suitable for
the formal position [a passage] actually occupies.” Thus, for instance, the textures
and gestures in the openings of Haydn’s Piano Sonata in E-flat Major, Mozart’s K.
332, or Beethoven’s opus 1, no. 1, prevent them from sounding like endings, their
use of the Quiescenza figure notwithstanding. For a discussion of other works from
the later 1700s and afterward that begin with this figure, see Hepokoski and Darcy,
Elements of Sonata Theory, 91–92.

27. As Brown aptly puts it, the main theme “returns at the end of the movement,
because the brand of instability that it provided at the beginning underlines clo-
sure”; see The Symphonic Repertoire, 2:207.
28. This large final section in the tonic key corresponds to what Heinrich Christoph
Koch refers to either as a third Periode (if it follows a grand perfect authentic
cadence in a nontonic key) or as the second half of the second Periode (if there
is no preceding perfect cadence in a nontonic key, as is the case with Symphony
no. 81); see Koch, Versuch einer Anleitung zur Composition, 3 vols. (Rudolstadt and
Leipzig: Böhme, 1782–93), 3:304–11, 396–425; trans. by Nancy Kovaleff Baker as
Introductory Essay on Composition: The Mechanical Rules of Melody, Sections 3 and 4 (New
Haven, CT: Yale University Press, 1983), 199–201, 223–44. For reasons that should
become clear below, I deliberately avoid the term “recapitulation” to describe this
final, large tonal section.
29. Modern readers might note the similarity of this to the concept of “rotational form”
discussed at length in Hepokoski and Darcy, Elements of Sonata Theory.
30. Problems with applying the relatively modern concept of the “point of recapitula-
tion” to a number of works of the eighteenth century, including those that follow
the layout depicted in table 1.2, have been discussed at great length by Peter Hoyt;
see Peter Hoyt, “The Concept of développement in the Early Nineteenth Century,”
in Music Theory in the Age of Romanticism, ed. Ian Bent (Cambridge: Cambridge
University Press, 1996), 141–62; and Hoyt, “The ‘False Recapitulation’ and the
Conventions of Sonata Form” (PhD diss., University of Pennsylvania, 1998). I dis-
cuss this layout in greater length in L. Poundie Burstein, “Echt oder Falsch? Zur
Rolle der ‘falschen Reprise’ in Haydns Sinfonie Nr. 41,” trans. Felix Diergarten, in
Joseph Haydn (1732–1809), 97–129, as well as in L. Poundie Burstein, “True or False?
Reassessing the Voice-Leading Role of Haydn’s So-called ‘False Recapitulations,’”
Journal of Schenkerian Studies 5 (2011): 1–37.
31. Here, too, it is instructive to compare this movement with the first movement of
Symphony no. 61. Although the key structures of the second halves of these move-
ments share striking similarities, in Symphony no. 61 the last large tonal section
does begin with a clear return of its main theme, which—as noted above—in this
work is quite well suited for establishing a convincing sense of a beginning.
32. As Neuwirth notes, the solid manner in which the exposition’s transition begins—
which contrasts with the hazier nature of this movement’s main theme—helps
allow the material of the transition to serve as the start of a section within the move-
ment’s second half; see Neuwirth, “‘Verschleierte’ Reprisen,” 54–63. That the tran-
sition begins in the firm manner appropriate for an opening, however, does not
detract from that overall impression that it has a medial function, as becomes ever
clearer as the section continues.
33. Attempts to pinpoint the start of the recapitulation in the first movement of
Symphony no. 81 usually have been appropriately accompanied by an acknowl-
edgment of the problems involved; see, for instance, Charles Rosen, The Classical
Style: Haydn, Mozart, Beethoven (New York: Norton, 1998), 157–59; George Edwards,
“Papa Doc’s Recap Caper: Haydn and Temporal Dyslexia,” in Haydn Studies, ed. W.
Dean Sutcliffe (Cambridge: Cambridge University Press, 1998), 296; Ethan Haimo,
Haydn’s Symphonic Forms: Essays in Compositional Logic (Oxford: Clarendon Press,
1995), 145–60; and Brown, The Symphonic Repertoire, 2:207. The advantages and
disadvantages of the various proposed points of recapitulation for this movement
are discussed at length in Neuwirth‚ “‘Verschleierte’ Reprisen,” 54–63. Neuwirth

contends that the formal processes in this movement—with its multiple, ritornello-
like returns—are better understood as relating to those found in a typical concerto
movement rather than in terms of standard sonata-form paradigms.
34. A reading of this movement in terms of a “reversed recapitulation” is proposed
in Haimo, Haydn’s Symphonic Forms, 155; see also Brown, The Symphonic Repertoire,
2:207. Neuwirth discusses the various problems with reading a “reversed recapitula-
tion” in this movement in “‘Verschleierte’ Reprisen,” 58–59.
35. Another swapping of function involves measures 42ff. and 161ff. Although the the-
matic contents of these passages are similar, they have differing roles: the function
of measures 42 and following is that of a second part in a two-part transition (using
Koch’s terminology, a Grundabsatz in der Tonart der Quinte); the function of mea-
sures 161 and following, on the other hand, is that of a final part of the subordinate
theme (in Koch’s terminology, the Schlußsatz).
37. Haydn reported this to his friend and biographer Georg Griesinger, and he related
something similar to another of his biographers, Albert Christoph Dies; see Vernon
Gotwals, ed. and trans., Haydn: Two Contemporary Portraits (Madison: University of
Wisconsin Press, 1963), 62, 155; the translation used here is from Webster, Haydn’s
Farewell Symphony, 234.
38. See Webster, Haydn’s Farewell Symphony, 235; and Richard Will, “When God Met the
Sinner, and Other Dramatic Confrontations in Eighteenth-Century Instrumental
Music,” Music and Letters 78 (1997): 175–209.
39. Regarding Haydn’s remarks on the programmatic implications of his symphonies
with his biographers, see the illuminating discussion in Webster, Haydn’s Farewell
Symphony, 234–35; see also David Schroeder, “Orchestral Music: Symphonies and
Concertos,” in Clark (ed.), The Cambridge Companion to Haydn, 96. Will proposes
that this basic paradigm may be found in a number of works by other compos-
ers from the eighteenth century as well; see Will, “When God Met the Sinner,”
194–209.
40. Gotwals, Haydn, 199.

Chapter Two
Mozart’s Sonata-Form Arias

Nathan John Martin
In his influential 1972 study The Classical Style, Charles Rosen maintains that
Mozart’s mature operas make comprehensive use of sonata principles: “No
description of sonata form can be given,” he writes, “that will fit the Haydn
quartets but not the majority of forms in a Mozart opera.”1 At the opposite
extreme, James Webster contends in his 1991 article on Mozart’s arias that “a
catalogue of formal types in Mozart’s Da Ponte operas would include precisely
one aria in sonata form,” namely, Susanna’s “Venite, inginocchiatevi” (Figaro,
no. 12).2 Between these two poles—and accounting in part for the disparity
in their verdicts—lies a rich, and richly contested, scholarly terrain: some two
hundred years of competing agendas—political, historiographical, and ideo-
logical—whose tectonic shiftings erupted into the analysis of Mozart’s operas
in the early 1990s.3 In what follows, I adopt a calculated naiveté in undertaking
to bracket off these complexities so as to cast a fresh eye on those of Mozart’s
operatic arias that are, in some sense, “in” sonata form. My aim, in so doing,
is to reach a juste milieu between Webster and Rosen.4 For if Rosen’s commit-
ment to “sonata principles”—with all that phrase’s manifold accretions—leads
him to subsume even such prima facie unpromising candidates as Figaro’s “Se
vuol ballare” (Figaro, no. 3), Zerlina’s “Batti, batti” (Don Giovanni, no. 12), and
Donna Anna’s “Non mi dir” (Don Giovanni, no. 23) under the rubric of sonata
form,5 Webster’s polemicizing sins in the opposite direction: taken at face
value, his injunction would deny the obvious analogies that obtain between the
formal shapes of many Mozart arias and the more familiar templates operative
in his instrumental music, resemblances that Webster is elsewhere quite ready
to acknowledge.6
As a starting point, I tabulate some straightforward statistics, counting
in particular the proportion of arias in sonata form (in a sense to be clari-
fied momentarily) in generically matched works from the opposite ends of
Mozart’s later operatic career: first Idomeneo (Munich, 1780) versus La clemenza
di Tito (Prague, 1791), and then Die Entführung aus dem Serail (Vienna, 1782)
versus Die Zauberflöte (Vienna, 1791). What these comparisons suggest is that
sonata-form arias became significantly less prevalent in Mozart’s operas over

38 nathan john martin
the second half of his career.7 I then move to three more detailed analytical
case studies: of “Traurigkeit” (Die Entführung, no. 10), “Un’ aura amorosa”
(Così fan tutte, no. 17), and “Ach, ich fühl’s” (Die Zauberflöte, no. 17). A subtext
throughout is my attempt to prize apart any reflexive associations in the read-
er’s mind between sonata form, “tonal drama,” and onstage action. My princi-
pal concern, however, is to illustrate how, in his later arias, Mozart progressively
abridged the complete sonata structures that predominate in Idomeneo.
❧ ❧ ❧
In Idomeneo, the basic template for an aria can be schematically described

as follows: a standard number begins with a sonata exposition (preceded,
optionally, by an orchestral ritornello) and ends with a complete recapitula-
tion (sometimes with a concluding ritornello). Whatever text was sung to the
exposition’s music is repeated verbatim in the recapitulation, even when, in
minor-mode arias, the transposition of the subordinate theme(s) to the tonic
imparts a contrasting affective valence. In between the exposition and reca-
pitulation—where, in the corresponding instrumental form, the development
would go—there may or may not be an additional span of music about whose
typical formal properties I am as yet hesitant to generalize.8
Illia’s lovely third-act aria “Zeffiretti lusinghieri” (no. 19) can stand as an
example (ex. 2.1). The aria begins with an orchestral introduction (ritornello)
that synopsizes the exposition’s subsequent gestures. Measures 1–8 form an
eight-measure compound-basic-idea+continuation hybrid that leads to a tonic
half cadence (HC) at measure 8. (This same music will subsequently reappear,
in varied form, as the aria’s main theme.) To this initial tight-knit unit, Mozart
then appends a new four-measure continuation (beginning in m. 9) that is
supported exclusively by an expanded cadential progression. The projected
cadence is evaded at measure 13, however, and the entire continuation is
repeated, this time completing its arc and coming to rest with a perfect authen-
tic cadence (PAC) at measure 17. The entire introduction thus resembles a
kind of sixteen-measure hybrid in which a complete eight-measure theme type
is followed by a more expansive continuation phrase.9 (The second continua-
tion, incidentally, will reappear to close the exposition’s subordinate theme, so
that the entire ritornello is composed of materials subsequently heard in more
expanded form.)10
The entrance of the singer in measure 18 marks the beginning of the expo-
sition proper. The main theme (mm. 18–29) appears much as it was prefigured
in the introduction, except that its continuation phrase is now significantly
expanded (m. 5 becomes mm. 22–23, m. 6 becomes mm. 24–25, and the origi-
nal cadential function of mm. 7–8 is expanded to four full measures in mm.
26–29).11 After the main theme ends (I:HC arriving at m. 29), a subordinate
theme follows immediately in the dominant, beginning in measure 30. This

Example 2.1. Mozart, “Zeffiretti lusinghieri” (Idomeneo, no. 19), mm. 1–60
(continued)

Example 2.1.—(continued)
(continued)

mozart’s sonata-form arias 41
Example 2.1.—(concluded)
new function is again expressed through a tight-knit theme type, this time an
eight-measure sentence (compressed by a measure in its continuation), which
leads to an internal V:HC in measure 36.12 The subordinate theme (another
lightly expanded eight-measure sentence) begins at the pickup to measure 38
and reaches a partial close with an imperfect authentic cadence (IAC) in the key
of the dominant (V:IAC) in measure 48. As a rule, though, a subordinate theme
concludes with a PAC, and the theme’s final measures (mm. 45–48) are accord-
ingly repeated in measures 49–51. Finally, the PAC projected for the downbeat
of measure 51 is evaded when the languid continuation phrase first heard in
the opening ritornello (mm. 9–17) reappears in measures 52–60. That phrase,

itself interrupted (at m. 56) and then repeated, serves to bring the subordinate
theme, and with it the exposition as a whole, to a close in measure 60.
Following a brief postcadential passage (not shown), the aria launches into
a middle section, after which (from m. 82 on) the entire exposition is reca-
pitulated. The middle (ex. 2.2) is composed, in this instance, of two tight-knit
eight-measure theme types: an eight-measure sentence in E minor leading to a
i:HC in measure 72, followed by an eight-measure sentential hybrid that starts
in G major and modulates back to E minor at the beginning of its continua-
tion before reaching a second i:HC in measure 80 (reiterated in mm. 80–81).
The whole unit, it is worth underscoring, is organized along principles totally
different from an instrumental development section—a formal function that
almost never occurs in Mozart’s arias and ensembles. After this middle section,
the entire exposition is repeated, with all the normal tonal adjustments to the
subordinate theme occurring as expected.
The basic form just outlined is, again, the standard aria template in Idomeneo
(table 2.1). Of the opera’s thirty-two numbers, fourteen are arias, and all but
two of these conform to the type just exemplified. Six of these twelve “sonata-
form arias” have middle sections separating their full expositions from their
recapitulations, and in two cases—“No, la morte” (no. 27) and “Torna la pace
al core” (no. 31)—this interior section is in a contrasting tempo and meter,
a detail that of course suggests its filiation to the B section of the older da
capo form.13 The other six sonata-form arias answer instead to the so-called
sonatina or Type 1 sonata form consisting solely of an exposition followed by
a recapitulation (i.e., sonata without development).14 Of the two remaining
arias, Electra’s “D’Oreste, d’Ajace” (no. 29) is essentially a sonata form without
development, except that only the subordinate theme is recapitulated, while
Idomeneo’s “Vedrommi intorno” (no. 6) consists of a full sonata exposition in
a slow tempo (andantino) capped by a concluding allegro.15
If we turn now to Die Entführung aus dem Serail, written two years later, we
find a similar distribution of aria types (table 2.2). In this case, the arias make
up thirteen of the opera’s twenty-one numbers. The sonata-aria template pre-
dominant in Idomeneo is again the most common type, with seven of the arias
being cast in that form. The remaining numbers exhibit a variety of forms,
including the simple ternary (“Hier soll ich dich denn sehen,” no. 1), the
strophic lied (“In Mohrenland,” no. 18), the two-tempo rondò (“Wenn der
Freude Thränen fliessen,” no. 15), and the rondo (“O, wie will ich triumphie-
ren,” no. 19).16 One aria, Belmonte’s “O wie ängstlich” (no. 4) seems at once
to invoke and to deform the normal organization of a sonata exposition, and
ends with an expanded restatement of the main-theme material,17 whereas
Constanza’s “Martern aller Arten” (no. 11) opens with an extraordinary
extended concertante passage that also recalls the tutti exposition of a solo
concerto.18 Pedrillo’s second-act aria “Frisch zum Kampfe!” (no. 13), finally, is
a highly idiosyncratic construction.19

Example 2.2. “Zeffiretti lusinghieri,” mm. 65–81

Table 2.1. Aria types in Idomeneo
Number Title Formal type
1 “Padre, germani” (Ilia) Sonata (no development)

2 “Non ho colpa” (Idamante) Sonata (no dev.)
4 “Tutte nel cor vi sento” (Electra) Sonata (no dev.)
6 “Vedrommi intorno” (Idomeneo) Sonata exposition +
concluding allegro
7 “Il padre adorato” (Idamante) Sonata (no dev.)
10 “Se il tuo duol” (Arbace) Sonata with interior section
11 “Se il padre perdei” (Ilia) Sonata (no dev.)
12 “Fuor del mar” (Idomeneo) Sonata with interior section
13 “Idol mio” (Electra) Sonata (no dev.)
19 “Zeffiretti lusinghieri” (Ilia) Sonata with interior section
22 “Se colà ne’ fati” (Arabace) Sonata with interior section
27 “No, la morte” (Idomeneo) Sonata with interior section
29 “D’Oreste, d’Ajace” (Electra) Sonata (no dev.), incomplete
recapitulation
31 “Torna la pace al core” Sonata with interior section
(Idomeneo)
Table 2.2. Aria types in Die Entführung aus dem Serail
1 “Hier soll ich dich denn sehen” Ternary

(Belmonte)
3 “Solche hergelauf’ne Laffen” Sonata (no dev.)
(Osmin)
4 “O wie ängstlich” (Belmonte) Ternary (idiosyncratic)
6 “Ach ich liebte” (Constanza) Sonata (no dev.)
8 “Durch Zärtlichkeit” (Blonde) Sonata (no dev.)
10 “Traurigkeit” (Constanza) Sonata (no dev.)
11 “Martern aller Arten” Sonata (no dev.)
(Constanza)
12 “Welche Wonne, welche Lust” Sonata (no dev.)
(Blonde)
13 “Frisch zum Kampfe!” (Pedrillo) Idiosyncratic
15 “Wenn der Freude Thränen Rondò
fliessen” (Belmonte)
17 “Ich baue ganz” (Belmonte) Sonata with interior section
18 “In Mohrenland” (Pedrillo) Strophic lied
19 “O! Wie will ich triumphiren” Rondo
(Osmin)

The contrast between the distribution of aria forms in these earlier operas
and Mozart’s later practice is striking. In La clemenza di Tito, the sonata-form
design so prevalent in Idomeneo and Die Entführung appears exactly once; in Die
Zauberflöte, not at all. In the former (table 2.3), the standard form is instead a
hybrid sonata-ternary design in which a concise exposition is answered by the
recapitulation of the main theme only. Five of the opera’s eleven arias are in
this form (there are twenty-six numbers in all). Of the others, one (no. 21)
is a simple ternary, two are rondòs (nos. 19 and 23), two are sonata exposi-
tions with concluding allegros (nos. 2 and 9), and only Tito’s “Se all’impero”
(no. 20) resuscitates the main formal type of Idomeneo—indeed, in the by-then
archaizing version having a middle section in a contrasting tempo and meter.
The arias in Die Zauberflöte, on the other hand, seem to have in common only
the diversity of their formal designs (table 2.4). Nine of the opera’s twenty-one
numbers are arias, and of these, only one—Tamino’s “Dies Bildnis” (no. 3)—
corresponds even partially to the sonata-ternary template that prevails in Tito.
Of the others, three are strophic lieder (nos. 2, 10, and 15), one is a simple ter-
nary (no. 13), and one—the Queen of the Night’s first number (“Zum Leiden
bin ich auserkoren,” no. 4)—evokes the two-tempo rondò; Pamina’s “Ach, ich
fühl’s” (no. 17) and the Queen of the Night’s second aria (“Der Hölle Rache,”
no. 14) feature sonata expositions capped respectively by a compressed reca-
pitulation and an accompanied recitative; and Papageno’s “Ein Mädchen oder
Weibchen” (no. 20), finally, oscillates repeatedly between an F-major andante
theme and a C-major allegro.
Table 2.3. Aria types in La clemenza di Tito
2 “Deh se piacer mi vuoi” (Vitellia) Sonata exposition +

concluding allegro
6 “Del più sublime soglio” (Tito) Sonata-ternary
8 “Ah, se fosse intorno” (Tito) Sonata-ternary
9 “Parto, parto, ma tu ben mio” Sonata exposition +
(Sesto) concluding allegros
13 “Torno di Tito a lato” (Annio) Sonata-ternary
16 “Tardi s’avvede” (Publio) Sonata-ternary
17 “Tu fosti tradito” (Annio) Sonata-ternary
19 “Deh per questo istante” (Sesto) Rondò
20 “Se all’impero” (Tito) Sonata (no dev.) with
interior section
21 “S’altro che lacrime” (Servilia) Ternary
23 “Non più di fiori” (Vitellia) Rondò

Table 2.4. Aria types in Die Zauberflöte
2 “Der Vogelfänger bin ich ja” Strophic lied

(Papageno)
3 “Dies Bildnis” (Tamino) Sonata-ternary
4 “Zum Leiden bin ich auserkoren” Two-tempo rondò
(Die Königen der Nacht)
10 “O Isis und Osiris” (Sarastro) Strophic lied
13 “Alles fühlt der Liebe Freuden” Ternary
(Monostatos)
14 “Der Hölle Rache” (Die Königen der Sonata exposition
Nacht) +conclusion
15 “In diesen heil’gen Hallen” Strophic lied
(Sarastro)
17 “Ach, ich fühl’s” (Pamina) Sonata exposition
+compressed recap.
20 “Ein Mädchen oder Weibchen” 8-m period (andante) +
(Papageno) strophic lied (allegro)
What this brief survey suggests is that full sonata recapitulations became
progressively rarer in Mozart’s arias over the course of the 1780s. Sonata expo-
sitions, on the other hand, continue to figure prominently in his designs.
Of the eleven arias in La clemenza di Tito, for instance, all but three (the two
rondòs and Servilia’s “S’altro che lacrime”) begin thus. One expects, as a
result, to find a substantial repertory of techniques for abridging, compressing,
or otherwise condensing the complete recapitulations found in Idomeneo and
Die Entführung. The case studies that follow are designed to illustrate a selec-
tion of these techniques. My first example, Constanza’s moving aria d’affetto
“Traurigkeit” (Die Entführung, no. 10), establishes a kind of baseline in that it
corresponds very closely to the fully elaborated form that I illustrated above
with reference to “Zeffiretti lusinghieri.” Ferrando’s “Un’aura amoroso” (Così
fan tutte, no. 17) represents a compression of the fully worked form, and pro-
vides an instance of the sonata-ternary form so common in La clemenza di Tito.
Finally, with Pamina’s plangent aria d’affetto “Ah, ich fühl’s” (Die Zauberflöte, no.
17), Mozart compresses a near-complete sonata structure into what are very
nearly the bounds of a modest binary form.
“Traurigkeit” (Die Entführung aus dem Serail, no. 10)

“Immer noch traurig, geliebte Konstanze?” (Still sad, my dear Constanza?)
(emphasis added). Those words—the Pasha Selim’s (her husband, interlocu-
tor, and owner)—delimit Constanza when she first steps onstage toward the

end of the opera’s first act (1:7). The work’s psychological drama, its intrigue
over the first two acts, turns on whether the Pasha will “consummate” his mar-
riage—whether, that is, he will rape Constanza, as it seems he is lawfully enti-
tled to do within the fictitious legal context of the opera’s imagined Orient.20
Indeed, just before Constanza materializes, we have heard Pedrillo excite,
while ostensibly assuaging, his master’s anxieties on just this point: “Say, good
Pedrillo,” Belmonte asks uneasily when he comes upon his servant in act 1,
scene 4, “is my Constanza still alive?” “She lives,” comes the reply, “and I hope
still for you. . . . The Pasha is a renegade and has so much delicatesse that he
does not compel any of his wives to his bed, and so far as I know he still plays
the unrequited lover.”21
There are signs, however, that the Pasha’s self-mastery is imperfect, and if
Constanza’s aria directly apostrophizes Belmonte, it is nonetheless addressed
equally to the Pasha’s lightly veiled threats. Having welled up in the preceding
recitative, her Traurigkeit presses down on her with sudden, nauseating force.
Indeed, it leaves her momentarily unable to speak: the winds’ opening motto—
a stabbing G-minor triad that swells through an augmented-sixth chord to the
dominant—is an onomatopoetic gasp for breath (ex. 2.3). The abrupt pause
that follows serves to register the shock. For one brief moment, Constanza’s com-
posure has slipped; she has let fall the mask of dignified sorrow that she has worn
up to now: unbidden, unexpected, her grief overwhelms her; she stumbles, then
catches herself. The brief pause is enough, and she repeats the woodwinds’ ges-
ture (mm. 3–4), now naming it, and so beginning to master herself once more.
Traurigkeit ward mir zum Lose,

Weil ich dir entrissen bin.
Gleich der wurmzernagten Rose,
Gleich dem Gras im Wintermoose,
Welkt mein banges Leben hin.
Selbst der Luft darf ich nicht sagen
Meiner Seele bittern Schmerz:
Denn, unwillig ihn zu tragen,
Haucht sie alle meine Klagen
Wieder in mein armes Herz.
[Sorrow became my burden,

Because I was torn from you.
Like a worm-eaten rose,
Like the snow-burdened grass
My worried life wilts away.
Even to the air I can’t tell
My soul’s bitter pain:
For unwilling to bear them,
It breathes my cries
Back to my own poor heart.]

Example 2.3. “Traurigkeit” (Die Entführung aus dem Serail, no. 10), mm. 1–19
It is fitting, perhaps, that Constanza’s first sung gesture should stand apart
like a motto, an icon showing her with her dominant attribute. (That attri-
bute—her sorrow—is first sounded wordlessly, then labeled, like a figure in a
Byzantine mosaic.) Yet what follows in measure 5, where, propelled by the cir-
cling figures in the second violins, the aria seems first to get under way, is not
an initiation but a continuation: melodically, though not harmonically, these
measures are sequential, with corresponding accelerations in the surface and

in harmonic rhythm. Indeed, from measure 6 on, the harmonic progression

is cadential, complementing the soprano’s stepwise linear descent (from d2 in
m. 3 down to g1 in m. 8). The motion is interrupted, however, when the bass
lands on B♭ (not G) in measure 8. At the pickup to the following measure, a
second continuation phrase begins: from here to the downbeat of measure 10,
the third scale step is prolonged in the bass, with the initial i6 chord recolored
as the warmer diatonic mediant (III in m. 10), while Constanza’s thoughts turn
gradually to Belmonte (the “dir” of the text). But that brief idyll is shattered
when Constanza registers what she has just sung (“weil ich dir entrissen bin”);
the winds respond with a forte interjection (the unison D in m. 11), and she
leaps to her melody’s highest pitch (b♭2), under which the bass line begins its
cadential progression anew. Once more, however, the music’s forward impe-
tus is arrested: at the subito piano of measure 12, Constanza thinks again of
Belmonte, and the progression is deflected so as to land on the submediant at
the downbeat of measure 13. The entire cadential idea (mm. 11–13) is then
repeated in expanded form, with its implied emotional dynamics more explic-
itly spelled out (mm. 14–19): a new orchestral interjection (mm. 15–16) lingers
on Constanza’s reminiscence of Belmonte, adding a strong caesura to the line
at “dir,” and when she resumes (pickup to m. 17), it is on a note of resigned
pathos. The limping, stop-and-go quality of the entire theme (note the rhetori-
cal pauses in mm. 10, 13, 15, and 16) reflects, to my mind, at once Constanza’s
grief, and her disorientation.
After the closing cadence, the main theme’s angry wind interjections are
transformed into a throbbing B♭ pedal in the horns and bassoons, and these
cajoling sounds call forth a new, gentle figure on Constanza’s part (ex. 2.4,
mm. 20–22). Shadowed a sixth below by the cellos and basses, the idea is
repeated straightaway in measures 23–25, and the entire gesture comes thereby
to form the basic idea of a four-measure presentation inaugurating the aria’s
transition. Two sequentially related one-measure fragments mark the begin-
ning of the continuation, which reaches a half cadence to V/B-flat in measure
29. The abrupt shift to the relative major at the transition’s beginning, along
with the recessive dynamic and the gentle melodic lines, conveys a mood of
hushed expectancy, perhaps even hope, that sits incongruously with the mor-
dant textual imagery (“Gleich der wurmzernagten Rose, / gleich dem Gras im
Wintermoose”), even if a desolate note does creep in when d♭2 appears above
the concluding postcadential dominant pedal (mm. 29–32).
This dissociation—dissonance even—between the textual imagery and its
sonic instantiation continues in general through the first subordinate theme’s
initiating function (ex. 2.5, mm. 33–40), though the sense of groundlessness
that results when the cellos and basses drop out admittedly answers nicely to
the verses’ ethereal imagery. The theme’s basic idea appears first in the winds
(mm. 33–35), and its conclusion dovetails with Constanza’s answer: her phrase,
of course, is melodically identical, though this time unaccompanied; it now

Example 2.4. “Traurigkeit,” mm. 20–32
functions as a contrasting idea by virtue of the call-and-response effect.22 The

entire passage is repeated in measures 37–40 and so produces an eight-mea-
sure compound presentation, with the restatement of the basic idea likewise
eliding with its predecessor’s end. The last statement of the contrasting idea
breaks off, however, when the winds enter on a piercing diminished-seventh
chord (m. 41; on the word “Schmerz,” appropriately enough).23 From there,
the harmony continues through a cycle of applied diminished-seventh chords
to reach viiº56 of B-flat in measure 43 (and then on to the root-position tonic
at the half measure of 44). The basic voice leading is in descending parallel
tenths, which serve to carry the outer voices from d1–f2 (in m. 41) to B♭–d2 (in
m. 44). A short cadential idea (mm. 45–46) then brings the first subordinate
theme to a close.24

(continued)

At measures 47–48, fragmentation into one-measure groups and a circling

motion about the tonic initially suggest the beginning of a closing section, but
the music blossoms instead into a second, full-length subordinate theme.25
The presentation is of the “evolving” type, falling into two one-measure groups
that are followed by a two-measure unit (here internally expanded to fill four
full measures, mm. 49–52).26 The continuation begins in measure 53, with
a diminished seventh on “Herz” (m. 54), as a kind of aftershock recalling
mm. 41–43, and in measure 55, an expanded cadential progression begins to
unfold. The harmony resolves deceptively at measure 58, and again at measure
60, before being allowed to drop to the local tonic in measure 62. (Both times,
Mozart cannot resist a musical pun, the word “wieder” being set directly at the
point where cadential activity is reinitiated following the deceptive close.)
Following a short retransition (mm. 63–66) accomplished by means of reit-
erated 5–6 progressions, the aria’s recapitulation begins in measure 67 (not
shown). The main theme is restated unaltered in measures 67–85 (= mm. 1–19
of ex. 2.3), with the first departure from the exposition coming, as expected, at
the onset of the transition. That section is now appropriately compressed and
rewritten so as to remain in G minor: the original basic idea (mm. 20–25) is
replaced by a new one (mm. 86–87; ex. 2.6) that sounds at first like an appen-
dix to the preceding phrase, and the entire continuation is cut, so that the
passage reaches an extended standing on the dominant already in measure 89,

with that pedal point then being expanded by means of a dense new contra-
puntal elaboration in the winds.
After the transition, the exposition’s two subordinate themes are reprised
(beginning in m. 96; ex. 2.7), but with the original presentation of the first
(mm. 33–40 in ex. 2.5) replaced by a mournful new idea in the oboes and
basset horns (mm. 96–98) that receives a sepulchral reply from Constanza
(mm. 98–100). Still, despite the change in both aspect and affect, the phrase
structure of the original eight-measure presentation is preserved: the wood-
winds and the singer again answer one another in interwoven gestures, and
the second answering phrase is again cut off in mid-stride by a piercingly inter-
jected diminished-seventh chord (m. 104). Ingeniously, Mozart preserves the
soprano’s original chromatically filled-in descending third (f♮2–e♮2–e♭2–d2)
at pitch, and even harmonizes it with the same diminished seventh, though
now enharmonically reinterpreted so as to arrive at a G-minor rather than

B-flat-major triad at m. 107 (harmonizing d2).27 Following the first subordi-

nate theme’s cadential idea (mm. 107–9), the second subordinate theme reap-
pears essentially unaltered, except for the sudden move to the subdominant
in measure 114 and the elision of the original cadential extensions. After a
cadence in measure 121, Mozart repeats the entire theme (mm. 121–33 = mm.
109–21), this time introducing a new series of reiterated cadential ideas (mm.
133–43; ex. 2.8). As a result of these various expansions, the recapitulation is
not merely as long as the exposition; it even exceeds it in length, encompass-
ing as it does almost all of the exposition’s materials and introducing further
reiterations and repetitions.

“Un’ aura amorosa” (Così fan tutte, no. 17)

In the opening scene of Così fan tutte, Ferrando and Guglielmo make a Faustian
bargain: incensed at Don Alfonso’s having disparaged their lovers’ constancy,
they wager that Dorabella and Fiordiligi will be faithful to them come what
may (1:1). What comes is an elaborate ruse: at Don Alfonso’s behest, the men
feign having been called off to war, switch places, and, disguising themselves
as itinerant Armenians, pay court to each other’s partners. At their first pass,
both are coldly rebuffed. And Guglielmo, thinking he and Ferrando have won,
breaks character and laughs. “E voi ridete?” (And you’re laughing?), asks Don
Alfonso. Chastened, the two men promise henceforth to obey. Guglielmo
nonetheless allows himself to grumble about Don Alfonso’s strict regime: “Do
we get to eat today?” (Ed oggi non si mangia?). Not to worry, replies Ferrando:
the vision of our beloveds is food enough for us:

Un’ aura amorosa

del nostro tesoro
un dolce ristoro
al cor porgerà,
al cor che nudrito
da speme, da amore
d’un’esca migliore
bisogno non ha.
[A loving breath
From our beloved
Gives sweet refreshment
To the heart.
For hearts that feed
On hope, on love
No better fuel
Is needed.]
The formal organization of the aria to which these words are sung has been
the cause of some dispute. In a 1975 study, Sieghart Döhring took the aria to
be broadly ternary, a characterization that Mary Hunter disputed in her 1982
doctoral dissertation: for Hunter, the aria instead represents a modified sonata
form consisting of an exposition plus a “tonal return section.”28 More recently,
Webster, in commenting on the exchange, merely notes both analyses without
proposing to adjudicate between them, while Steven Rings takes the overall
form as self-evidently ternary.29
The aria begins with an extended small-binary theme (ex. 2.9).30 Following
a one-measure thematic introduction, measures 2–3 form a basic idea that is
answered by a dominant version in measures 4–5. The continuation brings
fragmentation into one-measure groups, as well as a far more active bass line
that moves in parallel sixths with the upper voice (E–D–C♯–G♯ supporting
c♯1–B–A–e1) before the pattern is broken on the last eighth note of measure
7. There, the seventh E–d1 between the bass and the vocal line ushers in a
two-measure cadential idea, and the harmony passes from vi through IV so as
to arrive at an HC on V at the downbeat of measure 9. The entire eight-mea-
sure unit thus forms a sentence, in the manner of a large antecedent. Rather
than continuing to a large consequent, however, this sentence gives way to a
contrasting middle (mm. 10–13, prolonging the V reached in m. 9), which
is followed in turn by a new continuation (mm. 14–17). From the fifth E–b1
on which the contrasting middle ends, the voice retakes e1 (on the last six-
teenth of m. 13). The resulting E–e1 octave is transformed into a dissonant sev-
enth when the bass steps up to F♯ in the following measure. For the moment,
though, that seventh is left unresolved because the upper line passes down
to c♯1 (m. 15) over a neighboring motion in the bass.31 The cadential idea’s

Example 2.9. “Un’ aura amorosa” (Così fan tutte, no. 17), mm. 1–23
onset is marked by a sudden subito forte B-minor chord in the orchestra on

the third beat of measure 15, and the passage reaches an IAC in measure 17.
The entire continuation is then repeated in exquisitely expanded form (mm.
18–23) in order to reach a PAC at measure 23. From the final tonic of the
IAC in measure 17, the bass passes down through G♯ while the upper voice
retakes e1, so that we arrive back (in m. 18) at the same F♯–e1 seventh that

appeared in measure 14. This second time around, the dissonant seventh is
treated as the first link in the incipient chain of 7–6 suspensions that blos-
soms over the next three measures: e1 resolves to d1 in measure 19 over a
chromatically inflected bass (F♯ becomes F♮), and the pattern begins anew
in measure 20 (d1 over E); this time, however, the bass steps upward as the
suspension resolves, so that we arrive back on the fifth F♯–c♯1. The entire
phrase (mm. 18–21) thus turns out to be an expanded version of measures
14–15: the upper voice moves down from e1 through d1 to c♯1, while F♯ is
prolonged in the bass, first through a chromatic inflection and then through
a lower neighbor. But if the “dolce ristoro” of which Ferrando sings is incom-
parably sweeter this second time round, it is also—through the unexpected
D-minor inflection in measure 19—more ambiguous in its valence: there is
perhaps a sense, in this brief moment, that the melancholy wisdom of the
opera’s conclusion peeks through the aria’s naive facade.
The aria’s transition begins in measure 24 with a new basic idea that deflects
the harmony toward F-sharp minor (ex. 2.10). Yet instead of reiterating this
idea, Mozart proceeds in measure 26 to a four-measure continuation in E major
consisting of two one-measure fragments followed by a two-measure cadential
idea. The entire four-measure span composes out an ascending fourth (f♯1–
g♯1–a♯1–b1 in the first violins, shadowed by d♯1–e1–e1–d♯1 in the voice and
second violins) over a dominant pedal embellished by an upper neighbor (C♯
in m. 28), which gives the effect of a half cadence arriving in measure 29. The
richly expressive melodic tritone A–d♯1 in the voice in measure 26, together
with the cumulative swell implied by the ascending line in the first violins, no
doubt provides a vivid depiction of the “hope” (speme) that Ferrando feels ris-
ing in his breast.
Immediately after the transition’s ending, the aria continues with a concise
subordinate theme (ex. 2.11). The presentation (mm. 30–33) here is also of
the “evolving” type (1 + 1 + 2),32 with the result that there is a sense of begin-
ning in medias res: despite the stable prolongation of the root-position tonic
beneath these measures, the fragmentation down to one-measure groups
perhaps gives this passage an initial air of continuation function. The actual
continuation, though, begins at measure 34 and brings a sharp increase in har-
monic rhythm together with a quick ascent up to the tenor’s highest note (a1).
Two measures later (m. 36), the theme makes a first attempt at cadential clo-
sure, but the cadence is evaded in measure 37, when the tenor leaps back up
to B and the harmony retakes I6 (through V24 on the last beat of m. 36). A sec-
ond attempt is likewise evaded (m. 39) before the theme is allowed to come to
rest with a PAC in E major at measure 41. The subordinate theme as a whole,
to summarize, is an eight-measure sentence, characteristically extended at its
cadential function through what Janet Schmalfeldt has evocatively called “one
more time technique.”33
Thus far, Mozart has crafted an entirely conventional, albeit highly com-
pressed, sonata-form aria. But the form’s first half, as it turns out, will be

Example 2.10. “Un’ aura amorosa,” mm. 23–29
balanced not by a full recapitulation but rather by an altered restatement of

the main theme only (mm. 42–73). Example 2.12 picks up this restatement
from measure 58. Up to now, the reprise has corresponded exactly to the origi-
nal version (i.e., mm. 42–57 = mm. 2–15), and this exact parallelism continues
until measure 62, where Mozart sidesteps a PAC that could have closed the
unit there (as it did in the corresponding place in the exposition). As in the
subordinate theme’s continuation (cf. mm. 36–37 of ex. 2.11), the bass slides
away from the dominant at the last possible moment (m. 62, third beat) to
bring the harmony back to I6, while the vocal line’s line descent is broken by
a leap back up to scale degree five. Mozart reiterates this progression twice
(mm. 63–65 and 65–67) before allowing the harmony to continue on to an
expansive subdominant in measures 68–70. That subdominant then passes

Example 2.11. “Un’ aura amorosa,” mm. 30–41
onward to a similarly elaborated dominant (embellished with a cadential 46)

in measures 71–72, and the arrival of the tonic in measure 73 completes the
expanded cadential progression begun back in measure 67. Thus, whereas no
material from the subordinate theme reappears in the reprise,34 Mozart none-
theless recomposes the main theme’s ending so as to incorporate those loos-
ening devices that suggest subordinate-theme function: namely, the cadential
extensions and expansions that make up the typical concluding rhetoric of a
subordinate theme. The overall design of the aria accordingly hovers some-
where between a sonata (without development) and a ternary form, though
it is worth reiterating that the resulting construct is not a “deformation” or
hybrid, but rather a standard option for operatic arias.35 Indeed, as we saw
above, this construct itself becomes the norm by La clemenza di Tito.
“Ach, ich fühl’s” (Die Zauberflöte, no. 17)

At some unspecified time before Die Zauberflöte begins, Pamina has been
abducted by Sarastro,36 and as the first act opens, is being held against her

Example 2.12. “Un’aura amorosa,” mm. 58–73
will in his Temple of Wisdom. Sarastro, it is hinted, intended at first to make

her his wife,37 and this veiled threat of sexualized violence finds explicit echo
in the actions of his servant Monostatos, who repeatedly tries to rape her. The
lingering resonance of Die Entführung—of the Pasha behind Sarastro, of Osmin
behind Monostatos, and of Belmonte and Constanza behind Tamino and
Pamina—is palpable. But Die Zauberflöte is a rescue opera that goes strangely
awry. No sooner does Tamino arrive (belatedly) on the scene, than he throws in
his lot with Sarastro’s priests: binding himself by a vow of silence, he submits to
the brotherhood’s arcane hazing. When Pamina—having endured yet another
assault (2:10), and then egregiously betrayed by her mother (2:8)—sees him
again (2:18), he waves her callously away. The constant violence—both physi-
cal and psychological—that Pamina endures is the practical correlate of the
outspoken misogyny of Sarastro and his priests.38 Is it any wonder, then, if she
is reduced to despair?
Her suffering coalesces in her poignant act 2 aria d’affetto “Ach, ich fühl’s.”39
The aria begins with an anguished cry—an inarticulate “Ach” (d2) that pierces
the orchestra’s G-minor vamp before two sobbing appoggiaturas lead d2 down-
ward to g1 (ex. 2.13). This first, involuntary cry elicits a second: an expressive
octave leap up to g2, and a second weeping descent that circles back to d2, with

Example 2.13. “Ach ich fühl’s” (Die Zauberflöte, no. 17), mm. 1–16
(continued)

Pamina’s voice lingering on that note’s upper neighbor (e♭2, on the downbeat
to m. 3). The revolving vocal line (beginning and ending on d2) is mirrored
in the orchestra’s obdurate accompaniment—the unyielding G pedals, both
superior and inferior, that at once frame and constrict the inner voices’ throb-
bing dissonances, all articulated to an obsessively trochaic rhythmic pulse. At
the beginning of the continuation (m. 3, half measure) the harmony dislodges
itself at last, with the bass dropping down to E♭ and then on to D (mirror-
ing Pamina’s e♭2–d2 sung to the word “verschwunden”) and so arriving pre-
maturely on the dominant at the downbeat of measure 4, where a mournful
descending line (again emphasizing e♭1–d1) intrudes in the bassoon.40
Ach, ich fühl’s, es ist verschwunden!

Ewig hin der Liebe Glück!
Nimmer kommt ihr Wonnestunden
Meinem Herzen mehr zurück!
Sieh Tamino! diese Tränen
Fließen, Trauter, dir allein.
Fühlst du nicht der Liebe Sehnen,
So wird Ruh im Tode sein!

[Ah, I feel it; it’s gone forever!

Forever lost, the joy of love!
You won’t come, you happy hours
Ever to my heart once more!
See Tamino! These tears
They flow for you alone.
Don’t you feel love’s longing?
Thus there will be peace in death!]
Beginning in measure 8, the aria’s main theme gives way to a subordinate

theme in the relative major. Having stood postcadentially on V7 of G for just
over two measures, the harmony expands outward to V6 of D at the half mea-
sure of measure 7. This latter is then wondrously transfigured into V7/B-flat
through a series of semitone displacements (c♯1→c♮1, e♮1→f1, e♮2→e♭2), and
we are lifted all at once onto a wholly new affective plane. With the relative
major reached, a sweeping new melody begins gradually to unfold. Indeed,
throughout this section, there is an abiding irony to the setting. Outwardly,
Pamina remains transfixed by sorrow: “Nimmer kommt ihr Wonnestunden/
Meinem Herzen mehr zurück.” The verb’s tense is present; its sense future.
But its valence is past: the musical tense, if there is such a thing, looks back-
ward; perhaps the downward contour of the new compound basic idea (mm.
8–10)—f2 arpeggiated down to d1—together with the deceptive bass motion
to vi at the gesture’s close (m. 10) still resonate with a note of residual tristesse
(the relentlessly trochaic rhythm, if submerged, is still there, a muted reminder
of mournful reality), but the continuation’s motion in ascending sixths (mm.
10–12), and even more so, the ascending arabesques capped with exquisite
coloratura that mark the resumption of continuation function following the
internal half cadence in measure 12—all these features tend instead to con-
jure up, perhaps even to revive, the “Wonnestunden” whose passing Pamina’s
explicit utterance laments; as she thinks back to happiness lost, the musical
setting suggests, some shard of it is reawakened in her, and the expansions
and coloratura come to seem, in this light, like a poignant attempt to prolong
the mirage for one more moment, before it gives way to the singer’s inevitable
pause for breath.41
All the more wrenching, then, is the shift back to G minor in the retran-
sition (ex. 2.14, mm. 16–20), a change made abruptly through two 5–6 pro-
gressions that lead back to V/G. Each sixth-phase in the cycle is punctuated,
moreover, by stabbing dissonant ninths in the flute and oboe, and the ascend-
ing scales leading to those ninths seem to stretch upward only to be shattered
on the piercing dissonances that they meet. With Pamina’s drooping answers,
“Sieh’, Tamino . . . diese Thränen,” the temporal frame is likewise wrenched
around abruptly to the present, with Pamina for the first time apostrophizing

Example 2.14. “Ach ich fühl’s,” mm. 17–38
(continued)

her absent lover directly. The whole unit (mm. 16–20) is cast in the orches-
tral-statement+vocal-response idiom common in the subordinate themes of
Mozart’s arias, but the harmony is sequential, with the ascending chromatic
line b♭1–b♮1–c2–c♯2–d2 being supported by descending-third plus rising-
fourth root motions that are spelled out explicitly in the bass (i.e., B♭–G–c1–A–
d1). Once the progression’s goal is reached (d2 is over V/G in m. 20), Pamina’s
vocal line slides quickly down a fourth to a1, a sonic analogue, no doubt, for
the paths traced on her cheeks by the tears she calls to the absent Tamino’s
attention. Here, for the first time in the aria, Pamina’s declamation breaks
away from the relentless trochees that have bound it thus far (unmistakably in
the main theme, more subtly in the subordinate theme): “fliessen, Trauter, dir

allein”—the line comes out in a jumble, its normative scansion reinstated only
with the repetition of its second hemistiche (“dir allein,” mm. 21–22). And
this prosodic effect—the artful simulation of artless immediacy—presages the
even more striking setting of the verse that follows: when Pamina finally speaks
aloud the question that has haunted her all along—“Fühlst du nicht der Liebe
Sehnen”—, the question’s force is registered both in the centrifugal added
sixth chord (C–E♭–G–A) that intervenes in measure 22, and in the repetition,
in measures 23–24, of “der Liebe Sehnen,” which leaves the line’s normative
declamation in shards (that it can, indeed will, be scanned into trochees is
clear from mm. 28–30).
Having reached its rhetorical highpoint, the aria continues to a highly com-
pressed recapitulation. The music that follows, which reprises only the subordi-
nate theme’s second half, now transposed to G minor, would seem to foreclose
on the aria’s projected end result, as the text it now sets (“so wird Ruhe im
Tode sein”) states with uncomfortable bluntness. But the passage lands, in
measure 27, on a deceptive rather than an authentic cadence, and Pamina reit-
erates, subito forte and in her highest tessitura, with the same urgent rhythms,
and the same chromatic assent from the retransition,42 the question on which
all her being is fixed. Yet already in the measures that immediately follow (mm.
28–30), where the crucial verse is repeated, its vital force has begun to cal-
cify. The subito piano and the newly ossified scansion suggest the answer that
Pamina infers, and with the reentrance of the subordinate theme’s continua-
tion phrase (beginning in m. 30), the aria begins to close down. The deceptive
resolution in m. 33 only delays the inevitable, and a mournful PAC is reached
three measures later, then twice reiterated, before Pamina lapses into silence.
❧ ❧ ❧
With “Ach, ich fühl’s,” Mozart arrives at a structure of extraordinary formal

economy and overwhelming affective force. The contrast between its concision
and the expansiveness of his earlier arias is striking. But I emphatically do not
want to say that “Traurigkeit” is too long. When I hear it done well, I want it
never to end. In its particular dramatic context, moreover, it is exceptionally
effective: its very expansiveness, which (along with other factors) marks it off
as an intrusion of seria style into the surrounding singspiel, sets off Constanza’s
soliloquy as a moment of particular gravitas. But the stroke can only work
thanks to the contrast between its dramatic register and the other numbers
that surround it, much as the gravediggers’ clowning heightens the pathos of
Hamlet’s apostrophe to Yorick. An opera composed exclusively of such arias—
as is Idomeneo—runs very long: one of the reasons, no doubt, that Idomeneo is so
extensively cut in many modern productions. Still, if we have learned anything
from the decades-long work undertaken by Daniel Heartz and his students, it
is that eighteenth-century operatic tastes are not our tastes, and vice versa.43

We should be wary, then, of supposing that Mozart progressively condensed his

arias so as to better conform to our preferences.
Nonetheless, the fact remains that his arias got shorter. To the langer
Geschmack of Idomeneo succeeded the concision of Die Zauberflöte. “Traurigkeit”
runs to 147 measures; “Ach, ich fühl’s” a mere 41. In the three case studies
presented here, I have tried to show some of the ways in which Mozart accom-
plishes this work of abbreviation. In so doing, I have suggested, pace Webster,
both the persistence and the prevalence of certain kinds of “sonata principles”
in Mozart’s operatic arias. But I would also insist, pace Rosen, that we try to
strip these formal observations insofar as possible of the various ideological
accretions with which they have come, over the years, to be encrusted. Sonata
and sonata-like forms are no more inherently dramatic than any others. Nor
should the seemingly innocuous observation that Mozart’s operas exhibit for-
mal procedures that recall those also on display in his instrumental works be
used as a criterion of valuation, still less as a cudgel to beat down the produc-
tions of the earlier Italian traditions of which Mozart was also in part the scion.
For the inherited prejudices that tell us otherwise, Mozart’s operas can serve
as a useful purgative: some of his most aesthetically satisfying and dramatically
effective numbers are in sonata form and some are not, and these qualities
have exactly nothing to do with these formal conventions.
Notes
1. Charles Rosen, The Classical Style: Haydn, Mozart, Beethoven (New York: Norton,
1972), 296.
2. James Webster, “Mozart’s Operas and the Myth of Musical Unity,” Cambridge Opera
Journal 2 (1990): 204.
3. For some of the issues involved, see Carolyn Abbate and Roger Parker,
“Dismembering Mozart,” Cambridge Opera Journal 2 (1990): 187–95; James Webster,
“To Understand Verdi and Wagner We Must Understand Mozart,” 19th-Century
Music 11 (1987): 175–93; John Platoff, “Myths and Realities about Tonal Planning
in Mozart’s Operas,” Cambridge Opera Journal 8 (1996): 3–15; Platoff, “Tonal
Organization in ‘Buffo’ Finales and the Act II Finale of Le nozze di Figaro,” Music and
Letters 72 (1991): 387–403.
4. I benefited, in framing the matter this way, from Katharina Clausius’s unpublished
paper “Texture and Syntax in Mozartean Dramaturgy,” and from many conversa-
tions with its author.
5. Rosen, Classical Style, 306, 308. For my analysis of “Non mi dir” as a two-tempo rondò,
see “Formenlehre Goes to the Opera: Examples from Don Giovanni,” in Mozart in Prague:
Proceedings of the International Conference of the Mozart Society of America and the Society for
Eighteenth-Century Music, 9–13 June 2009, Prague, ed. Kathryn Libin (Prague: Academy
of Sciences of the Czech Republic, forthcoming), examples 9–12.
6. See, for instance, his discussion of “Un’aura amorosa” in James Webster, “The
Analysis of Mozart’s Arias,” in Mozart Studies, ed. Cliff Eisen (Oxford: Clarendon
Press), 1:122.

7. I have not yet systematically categorized the arias in Mozart’s best-known operas, the
three great drammi giocosi written in collaboration with Lorenzo Da Ponte. I omit
them here for two reasons: first, if one is interested in the distance traversed in some
trajectory, it is best to consider the endpoints; second, I am being quietly polemical—
I feel strongly that Le nozze di Figaro, Don Giovanni, and (more recently) Così fan tutte,
great operas though they undoubtedly are, have received a disproportionate share of
our critical attention for reasons that I hope to discuss in a future article.
8. As Tim Carter shows in a recent article, there are two basic ways to adapt this model
form to a two-stanza aria text. If there is a contrasting section between the exposi-
tion and recapitulation, then this section sets the second stanza, whereas the first is
distributed across the exposition and reprised in the recapitulation. If there is no
middle section, then the main theme sets the first stanza, with the second being
divided between the transition and subordinate theme(s). See Tim Carter, “Two
into Three Won’t Go? Poetic Structure and Musical Forms in Mozart’s Idomeneo,”
Cambridge Opera Journal 24 (2012): 229–48.
9. See my “Formenlehre Goes to the Opera,” examples 7 and 9.
10. On this mirroring relationship between ritornello and aria, see Webster’s com-
ments on “Porgi amor” (“The Analysis of Mozart’s Arias,” 151–69).
11. That the expansions set in at the beginning of the continuation is probably not
incidental: fragmentation is comparatively unusual in sentences that are sung, as
opposed to those that are played. For another example, compare the orchestral
introduction of “Cinque . . . dieci” (Figaro, no. 1) to the form the theme takes once
Figaro enters.
12. One might ask whether the section beginning in measure 30 is not in fact a modu-
lating transition, with the subordinate theme beginning only at measure 38. The
crucial detail for me is that, having arrived at a half cadence on the dominant in
measure 29, the music then simply stays in the dominant for the ensuing presen-
tation (in the manner of Robert Winter’s “bifocal close”; see Robert S. Winter,
“The Bifocal Close and the Evolution of the Viennese Classical Style,” Journal of the
American Musicological Society 42, no. 2 [1989]: 275–337). This construction occurs
not infrequently in the Mozart arias (“Dies Bildnis,” Die Zauberflöte, no. 3; and “Ah
chi mi dice mai,” Don Giovanni, no. 3 provide further instances).
13. Compare Rosen’s discussion in Sonata Forms, rev. ed. (New York: Norton, 1988),
28–70.
14. Electra’s rage aria “Tutte nel cor vi sento” (no. 4), with its off-tonic reprise of
the main theme, is, of course, a Type 2 sonata in James Hepokoski and Warren
Darcy’s typology; see Elements of Sonata Theory: Norms, Types, and Deformations in the
Late-Eighteenth-Century Sonata (New York: Oxford University Press, 2006), 343–52,
353–87.
15. This form appears with relative frequency in Mozart’s operas. Other examples
include “Deh se piacer mi vuoi” (Clemenza, no. 2) and “Parto, parto, ma tu ben
mio” (Clemenza, no. 9). See table 2.3.
16. Throughout, I impose an orthographic distinction (only intermittently observed
in eighteenth-century sources) between the rondo—the familiar collection of
ABACA and related forms—and the rondò: an aria in two tempos, slow then fast,
for an aristocratic character (usually a woman), generally preceded by an exten-
sive accompanied recitative and placed at a dramatic highpoint. I elaborated on
the formal organization of the rondò in “Mozart’s Rondòs,” read at the Society for
Music Theory’s annual conference in New Orleans on November 4, 2012.

17. The aria starts to go seriously awry in measure 29, when E major (♯II) is replaced
by E minor, and the music is rerouted to an HC to V/B (m. 29). This last is followed
by an extended standing on the dominant (mm. 29–35), and after a conspicuous
pause, we plunge into D major for what might have been the beginning of a sub-
ordinate theme (mm. 36–48), except that we are in the global subdominant and
the unit’s eventual cadence (arriving in m. 55) is a half cadence to the home-key
dominant. The whole idiosyncratic construction ends up most closely resembling
an expansive ternary form, hence its label in table 2.2. It is tempting to see, in its
artfully constructed impression of formal indecision, an incisive psychological por-
trait of Belmonte.
18. On the relationship between Mozart’s concerto and aria forms more generally, see
Martha Feldman, “Staging the Virtuoso: Ritornello Procedure in Mozart, from Aria
to Concerto,” in Mozart’s Piano Concertos: Text, Context, Interpretation, ed. Neal Zaslaw
(Ann Arbor: University of Michigan, 1996), 149–86; and James Webster, “Are
Mozart’s Concertos ‘Dramatic’? Concerto Ritornellos versus Aria Introductions in
the 1780s,” in Zaslaw, Mozart’s Piano Concertos, 107–37.
19. The aria opens with a brief orchestral ritornello, followed by a main theme and
modulating transition (mm. 1–27). The ensuing section, however, measures 28–38,
is not a subordinate theme but merely stands on the dominant throughout. It is
followed by a restatement of the second half of the main theme. Up to this point,
the number seems to be unfolding as a ternary aria. But in measures 44–54, the
transition reappears and leads to a restatement of the music from measures 28–38
(= mm. 55–65) transposed to the tonic. In measures 66–71, the second half of the
main theme is restated once again (= mm. 39–45), and there follows an apparent
closing section (mm. 71–85). In measures 86–97, however, the main-theme’s tail is
again reiterated, before a new series of codettas and a closing ritornello bring the
aria to its end.
20. “Ich könnte befehlen, könnte grausam mit dir verfahren, dich zwingen.” (I could
command you, could act cruelly with you, compel you; 1:7. All translations are my
own.)
21. “Sag, guter Pedrillo, lebt meine Konstanze noch?”; “Lebt und noch, hoff’ich, für
Sie . . . Der Bassa ist ein Renegat und hat noch so viel Delikatesse, keine seiner
Weiber zu seiner Liebe zu zwingen. Und soviel ich weiß, spielt er noch immer den
unerhörten Liebhaber” (1:4). Pedrillo’s coy “soviel ich weiß” winks, naturally, at the
opposite—to Belmonte’s unease and the audience’s titillation. Belmonte’s anxiet-
ies over Constanza’s “faithfulness” (anxieties that are particularly ironic given her
name) will, of course, furnish the chief dramatic intrigue in the act 2 Finale.
22. This structure occurs commonly in subordinate themes in Mozart’s arias. See, for
example, “Ah chi mi dice mai” (Don Giovanni, no. 3) or “Porgi amor” (Figaro, no.
10).
23. On this way of joining a presentation and continuation, see Matthew BaileyShea,
“Beyond the Beethoven Model: Sentence Types and Limits,” Current Musicology 77
(2004): 11–12.
24. Some readers may be disturbed by the fact that I label this theme a sixteen-measure
sentence in example 2.5, despite the fact that it is only fourteen measures long. I
do so to indicate that I understand it as a compound sentence whose continuation
has been compressed rather than as a simple (eight-measure) sentence whose pre-
sentation has been extended. The qualification “sixteen-measure,” in other words,
refers to the ideal-typical model form being instantiated and not to its instantiation.

25. That is, the passage is an instance of what Caplin terms a “false closing section.” See
William E. Caplin, Classical Form: A Theory of Formal Functions for the Music of Haydn,
Mozart, and Beethoven (New York: Oxford University Press, 1998), 123, 129.
26. See Mart Humal, “Structural Variants of Sentence in Main Themes of Beethoven’s
Sonata Form,” in Composition as a Problem II, ed. Mart Humal (Tallinn: Eesti
Muusikaakadeemia, 1999), 34–48.
27. The trick works, of course, because the passage is transposed down a minor third,
and diminished-seventh chords divide the octave symmetrically into that same
interval.
28. Sieghart Döhring, Formgeschichte der Opernarie vom Ausang des 18. bis zur Mitte des 19.
Jahrhunderts (Itzehoe: George, 1975), 97–98; Mary Hunter, “Haydn’s Aria Forms:
A Study of the Arias in the Italian Operas Written for Esterhaza, 1755–1783” (PhD
diss., Cornell University, 1982), 45.
29. Webster, “Mozart’s Arias,” 1:117; Steven Rings, Tonality and Transformation (New
York: Oxford University Press, 2011), 171.
30. Rings, Tonality and Transformation, takes these measures as comprising two eight-
measure sentences, the second of which is extended (173). Measures 10–13 are
not, however, a presentation within the strict terms of Caplin’s theory, since they
prolong dominant rather than tonic harmony. Taking these measures instead as a
contrasting middle, and so the theme as a whole as a small binary, also seems to me
to provide a more satisfying global account of measures 1–23 in showing how they
cohere into a single main-theme function that is expressed across that entire span.
31. I pass over here, as tangential to the formal analysis, the harmonic complexities
that are elegantly probed in ibid., 175–81.
32. See Humal, “Structural Variants.”
33. Janet Schmalfeldt, “Cadential Processes: The Evaded Cadence and the ‘One More
Time’ Technique,” Journal of Musicological Research 12 (1992): 1–52.
34. Pace Webster, “The Analysis of Mozart’s Arias,” 1:122–23.
35. This is, of course, the “sonata-ternary” described above. I analyze another aria of
this type, Ubaldo’s “Dove sono” from Haydn’s Armida, in my article “Formenlehre
Goes to the Opera: Examples from Armida and Elsewhere,” Studia Musicologica
51 (2010): 399–404. On analogous forms in instrumental slow movements, see
James Hepokoski “Sonata Theory and Dialogic Form,” in William E. Caplin,
James Hepokoski, and James Webster, Musical Form, Forms & Formenlehre: Three
Methodological Reflections, ed. Pieter Bergé (Leuven: Leuven University Press, 2009),
86; and Caplin, Classical Form, 216.
36. The abduction is recounted twice, by the Three Ladies in act 1, scene 5, and then
again in the Queen of the Night’s first recitative and aria in the following scene (“O
zittre nicht/Zum Leiden bin ich auserkoren,” no. 4).
37. “Du liebest einen andern sehr, / Zur Liebe will ich dich nicht zwingen, / Doch geb
ich dir die Freiheit nicht” (1:19); later, though, he changes his tune: “Pamina, das
sanfte, tugendhafte Mädchen haben die Götter dem holden Jünglinge [Tamino]
bestimmt. Dies ist der Grundstein, warum ich sie der stolzen Mutter entriss” (2:1).
38. To Tamino’s avowal in the act 1 Finale that the Queen of the night has sent him,
the Priest replies incredulously, “Ein Weib hat also dich berückt?— / Ein Weib tut
wenig, plaudert viel; / Du, Jüngling, glaubst dem Zungenspiel” (1:15). Slightly
further on (1:18), Sarastro patronizingly councils Pamina: “Ein Mann muss eure
Herzen leiten, / Denn ohne ihn pflegt jedes Weib, / Aus ihrem Wirkungskreis
zu schreiten.” Were that not enough, the priests’ duet (no. 11) in act 2, scene 3

stipulates: “Bewahret euch vor Weiber Tücken, / Dies ist des Bundes erste Pflicht.”
And these examples could be readily multiplied. By act 2, scene 5, Tamino has
imbibed their attitudes as his own: “Sie [Die Königin der Nacht] ist ein Weib,
hat Weibersinn!” Attendant on these attitudes is the restrictive conception of
masculinity evident in the repeated injunctions—Tamino’s (2:3) and the priests’
(2:6)—“Papageno! Sei ein Mann!”
39. Compare Webster’s analysis in “Cone’s ‘Personae’ and the Analysis of Opera,”
College Music Symposium 29 (1989): 45–50.
40. The ending of the main theme (m. 4) poses a theoretical problem, since there
is no half cadence there: the bass has arrived on scale degree 5 already in mea-
sure 3 and stays there, with descending linear motions (57–46–35) over top. There
are half-cadential progressions in measures 5–7 (It6–V in both cases), but these
sound postcadential, in part because of the shift in register, in part because the
soprano’s descent d2–c2–b♭1–a1 spans measures 1–4 and has reached its terminus
before these half-cadential progressions begin. So one might wish to argue that the
cadence in fact arrives on the downbeat of measure 5. In favor of this admittedly
contentious analysis, the following detail might be raised: while the chord on the
downbeat of measure 4 is quite obviously V7, it is far less obvious whether its succes-
sor (on the downbeat of m. 5) is a triad or a seventh chord. In the literal sense, obvi-
ously the former, since no seventh is sounded. But is a seventh implied? If so, we
have another reason to discount the idea of a half cadence here, since the terminal
harmony of that configuration cannot (for Caplin) bear a dissonant seventh. If not,
however, then we must explain what happened to the seventh sounded in measure
3. Where did it resolve? I would suggest: to the sixth (B♭) above the sounding bass
(D) at the half measure of measure 4. But if B♭ resolves the dominant’s seventh,
then the harmony of which it is a part—the apparent second-inversion G-minor
chord in the second half of measure 4—cannot represent the same harmony that
supported the seventh, since the dominant’s seventh is an essential, rather than
an incidental, dissonance. And if the G-minor chord is not a dominant chord, it
must be the only other candidate: namely, the initiating tonic of a half-cadential
progression. On this pattern, see also Nathan John Martin and Julie Pedneault-
Deslauriers, “The Mozartean Half Cadence,” in What Is a Cadence? Theoretical and
Analytical Perspectives on Cadences in the Classical Repertoire, ed. Markus Neuwirth and
Pieter Bergé (Leuven: Leuven University Press, 2015), 185–213.
41. Though if so, it is now marked by the pathos of its irreality, and that irreality is
captured beautifully in the transient quality that B-flat major here assumes, hav-
ing been reached, after all, by a chromatic sleight of hand of the kind more often
found in Schubert’s transitions.
42. The identical upper line (b♭1)–b♮1–c2–c♯2–d2 is now supported by a bass moving
E♭–D/F–e♮1–e♭1–d1, and harmonized VI–viiº56(34)/iv–IV6–Ger6–V. In view of this
recomposition and the subsequent return of the subordinate theme’s second half,
as well as the wholesale repetition of the aria’s text (from “Fühlst du nicht” to the
end) that begins here, some readers might wonder whether measures 27–38 are
not a formal repeat of measures 17–27, with the resultant form taking on the binary
aspect alluded to above. This is a reading I would dispute. First, the alleged repeti-
tion is clearly far from exact: measures 17–20 are omitted, and measures 20–24 are
significantly recomposed. Second, the “second part” is not closed at measure 27
(as the binary reading would require), since the music reaches only a deceptive
cadence there. I see the handling of measures 17–38 instead as strictly analogous

to what Mozart does in the recapitulations of “Or sai chi l’onore” (Don Giovanni,
no. 10) and “Ah taci, ingiusto core” (Don Giovanni, no. 15). In the former, at mea-
sure 47 of the aria, Mozart interrupts the reprise of the main theme by a deceptive
cadence; three measures of interpolated music lead to a tonic HC (m. 49); and
then the second half of the main theme is repeated (mm. 50–56) to complete the
form. In the latter, Mozart allows the recapitulation of the main theme to close with
a PAC (m. 67); the interpolated passage that follows (mm. 67–73) is in this case a
recomposition of the exposition’s nonmodulating transition (mm. 14–19), but the
tonic HC projected for measure 73 is undone when the orchestra adds a sforzando
seventh to the E-major chord on that downbeat. The subsequent measures bring
the repetition of the main theme’s second half that marks the number’s formal
end with the arrival of a tonic PAC in measure 79. All of these passages are, I think,
good illustrations of the way in which motivic and thematic material can migrate
across interthematic functions.
43. See, for instance, Daniel Heartz, From Garrick to Gluck: Essays on Opera in the Age of
Enlightenment, ed. John A. Rice (Hillsdale, NY: Pendragon Press, 2004).

Part Two
Nineteenth-Century
Taxonomies

Chapter Three
Formal Type and Formal

Function in the Postclassical
Piano Concerto
Julian Horton
William E. Caplin’s theory of formal functions offers one of the most substan-
tial accounts available of the formal strategies of Viennese-classical instrumen-
tal music. Elaborating and modernizing the Formenlehren of Arnold Schoenberg
and Erwin Ratz, Caplin formalizes the conventions bridging the gap between
the whole-movement forms that are the habitual starting point for architec-
tonic theories and the motivic processes underpinning Schoenbergian con-
cepts of developing variation.1 In so doing, he mediates skillfully the atomistic
and global perspectives characterizing what Mark Evan Bonds calls the “gen-
erative” and “conformational” attitudes into which much nineteenth- and
twentieth-century formal thinking divides, supplying a syntax (to borrow a lin-
guistic metaphor that Caplin himself adopts) arising from the vocabulary and
grammar of tonal melody and harmony, from which whole-movement forms
are constructed.2
Caplin’s theory however raises two obvious questions. First, it compels spec-
ulation about the ubiquity of his Viennese syntax: would the theory look the
same if it were grounded evidentially in (for example) the music of Boyce,
Clementi, and Dussek rather than Haydn, Mozart, and Beethoven? Second, we
might wonder about his theory’s historical reach: is a comparable theory of
nineteenth-century syntax possible, and if so, then to what extent does it main-
tain Viennese-classical conventions?
This chapter furnishes a tentative, generically restricted response to the
second of these questions, and a somewhat historically tangential response
to the first, by examining the functional syntax of a geographically dispa-
rate but formally consistent repertoire, namely, the postclassical piano con-
certo, paying special attention to first-movement form. My remit is both
theoretical, in that I aim to devise a model for understanding syntax in an

78 julian horton
as yet sparsely investigated body of music, and historical, in that I hope to

use this model to build a coherent picture of concerted practice in a par-
ticular phase of its development.
My historical purview spans two revolutionary watersheds, beginning with
Dussek’s concerti composed after the French Revolution of 1789 and ending
with Liszt’s two piano concerti, conceived in 1839–40 but brought to their final
form between 1848 and 1855—that is, in the aftermath of the 1848 revolutions.
This is not to pursue any direct analogy between the genre’s development, its
cultural-political circumstances, and concomitant theoretical issues, although
I recognize the musicological value of such a project. Rather, the revolution-
ary years are useful dividers because they demarcate changes in Dussek’s and
Liszt’s circumstances, which in both cases initiated creative phases during which
important generic practices were consolidated. Dussek’s move to London in
1789 in order to escape the revolution in Paris initiated an eleven-year stay in
England, in which time he completed eight concerti, from opus 15 of 1789 to
opus 40 of 1798, thereby consolidating a concerted style, form, and syntax that
came to dominate the concerto repertoire in London and elsewhere. By the
end of the first decade of the nineteenth century, the model of first-movement
form advocated in London had been widely disseminated, comprising a basic
component of the practice of the so-called virtuoso concerto, thanks to the
international reach of composers entering the London sphere, from Field,
Cramer, and Steibelt between 1790 and 1810 to Kalkbrenner, Moscheles, and
Sterndale Bennett after 1820.
Liszt’s turn to concerto composition after his move to Weimar in 1848 is,
in contrast, defined by an interplay of innovation and consolidation, through
which virtuoso characteristics submit to an agenda of (seemingly) radical for-
mal experimentation. The motivation behind this project reflects, in part, a
shift of compositional and aesthetic values arising in the wake of the early-
century critique of virtuosity, defined by Carl Dahlhaus as a transition from
improvisation to thematic logic as “modes of cognition,” and more recently
explained by Jim Samson as the replacement of “the perfect musical perfor-
mance” with “the perfect performance of music.”3 Liszt’s dual program of
formal reform and thematic integration—evinced in the merging of first-move-
ment form and the multimovement cycle, and in overlaid strategies of cycli-
cal thematic transformation—is, however, a synoptic response to his concerti’s
virtuoso genealogy as much as it stimulates a new phase of concerted composi-
tion. Not only are the idioms of Liszt’s concerti apparent piecemeal in much of
the earlier repertoire, but crucially, his formal experiments are also prefigured
in various concerti, which either tend toward the conflation of form and cycle
or at least collapse larger-scale formal distinctions.
The development of concerto first-movement form between these two
historical poles has been characterized in a number of ways. Several com-
mentators have drawn attention to the tendency for the eighteenth-century

formal type and formal function 79
sonata-ritornello hybrid to give way to a unitary sonata form, a process nascent

in various works of the 1820s and 1830s that truncate parts of the form (nota-
bly the first ritornello), and realized fully in the concerti of Mendelssohn and
Schumann, which abandon almost completely the discourse of ritornello and
solo as a large-scale formal determinant. Stephan Lindeman identifies this
development as one of two dominant postclassical trends: the “symphonic”
concerto integrates soloist and orchestra within a unitary sonata-form first
movement; the “virtuoso” concerto subordinates the orchestra to a dominant
soloist.4 James Hepokoski and Warren Darcy present a more focused view,
characterizing the symphonic variant as a post-Mozartian phenomenon:
Mozart’s concerto-sonata syntheses were continued by Beethoven and others.

Eventually, with Mendelssohn especially, the initial ritornello of the Type 5
concerto came to seem redundant, old-fashioned, something that had out-
worn its original raison d’être. With its excision, what had been the favored
format for concerto first movements . . . collapsed into the Type 3 pattern
[the standard “textbook” sonata form]. . . . The history of the concerto in the
eighteenth century and beyond, developing alongside the symphony, is that
of gradually being attracted to the latter’s principles, finding ways of adapt-
ing itself to them while retaining important features of its own identity, but
eventually (around the fourth decade of the nineteenth century) succumb-
ing rather totally to them.5
Significant though this trend is, a historical narrative that makes the con-
vergence of concerted and symphonic sonata forms its central feature is
overly reductive. In the first place, composers embracing the “Type 3” sonata
form are atypical in this period: against the two solo concerti of Mendelssohn
and the single concerto of Schumann, we might contrast dozens of works
maintaining the sonata-ritornello variant, from Dussek, Field, Steibelt, and
Cramer through Kalkbrenner, Moscheles, Hummel, and Weber to Ries,
Chopin, Sterndale Bennett, Herz, Litolff, and Henselt. Second, the Austro-
German focus of such a narrative belies the repertoire’s internationalism: to
narrate the genre’s history as a progression from Mozart to Mendelssohn via
Beethoven is to neglect its striking cosmopolitanism in this time. A detailed
picture would acknowledge a bewildering diversity of “centers,” from London
and Paris to Saint Petersburg, populated by composers who were themselves
highly mobile. Dussek, for example, was born in Bohemia and studied in
Prague, made his way to Paris via posts in Russia and Lithuania, moved to
London to escape the 1789 Revolution, and eventually finished his career in
Prussia and Paris; and Dussek is by no means exceptional.6 As a result, the
concept of a dominant Austrocentric, and particularly Mozartian, model of
concerto first-movement form is hard to sustain. This problem is exacerbated
by the protracted reception of Mozart’s concerti in this period: they have lit-
tle purchase on the performing canon before 1820 at the earliest; and their

80 julian horton
dissemination by other means is notably uneven.7 Consequently, any theory

designed to account for early nineteenth-century practice taking its bear-
ings from Mozart has to deal with the fact that the conventions established
in much of this music were consolidated and disseminated before Mozart’s
forms were fully absorbed.8
Caplin’s approach is advantageous in this context precisely because of its
syntactic rather than architectonic orientation: a theory of concerto first-move-
ment form that proceeds from syntax toward large-scale design can operate
independently of a formal paradigm derived from one source in the reper-
toire, while a study of this music’s syntactic distinctiveness may go a long way
toward accounting for formal practices that the Mozartian paradigm struggles
to explain. Underpinning this attitude is a conviction that the characteristics
distinguishing postclassical from classical first-movement forms are the large-
scale effects of accommodating novel syntactic patterns and their attendant
tonal and topical relationships. At its most extreme, this negotiation is mani-
fest in Liszt’s conflations of form and cycle, which I will argue result from the
proliferation of interthematic presentational and transitional structures. It also
produces an array of less radical designs, which alter the balance of the form as
Mozart conceived it.
Theoretical Preliminaries
Capturing these syntactic shifts, however, requires reconsideration of aspects
of Caplin’s theory. In particular, I want to reconceptualize elements of his
critical threefold distinction between grouping, function, and type. In Classical
Form, grouping and function are construed as frequently congruent but
essentially separate categories: grouping is defined as “the variety of discrete
time spans organized hierarchically in a work,” whereas function connotes
the “role that [a grouping structure] plays in the formal organization of a
work.”9 Although much formal analysis assumes the synonymy of the former
with the latter—as Caplin has it, “a musical group . . . is assigned a single
functional label, and, conversely, a given function is understood to take place
within the confines of a single group”—Caplin insists on the two categories’
theoretical separation, since “a group may express more than one function
simultaneously,” or else “several consecutive groups may express the same
formal function.”10
The point is well-taken; yet, because the identification of “discrete time
spans” inevitably depends on functional criteria, grouping and function con-
verge in the common ground of segmentation. The “beginning–middle–end”
paradigm that Caplin has sketched as a basic context for classical formal
function—which maps onto the division into initiation, continuation, and
cadence—is, for instance, only perceptible at all to the extent that it marks

out a grouping structure.11 The issue, in this respect, turns on the question of
how the interaction of grouping and function informs conceptions of musical
hierarchy. Although Caplin is of course right to assert that “several consecutive
groups may express the same formal function,” the concatenation of groups
under a common function invariably produces a larger grouping, the boundar-
ies of which are functionally defined.
Mindful of the grouping problems to which nineteenth-century syntax gives
rise, I want therefore to revisit some of Caplin’s terminology. Specifically, I
will argue for the flexible application of terms that Caplin applies in a more
restricted sense. In Classical Form, for instance, functions fall into two broad
classes: the “formal properties of the various phrases or sections associated
with a single thematic unit” constitute intrathematic functions; the “higher-level
formal syntax” obtaining in the functional relationship between theme and
form is housed under the concept of interthematic function.12 These two terms
delineate a grouping hierarchy: the grouping of intrathematic functions into
themes produces interthematic functions. In order to differentiate functions
occurring at a particular grouping level from the grouping level itself, I dis-
tinguish between inter- and intrathematic groupings, as groupings within the
overall hierarchy, and inter- and intrathematic functions, as functions occuring
within those hierarchical levels. I want, further, to situate these levels within
the hierarchy outlined in table 3.1.
Altogether, the table posits a fourfold hierarchy for whole-movement forms:
total form constitutes a movement’s overall design; large-scale groupings comprise
the major divisions of a total form; interthematic groupings comprise the groups
that make up the large-scale level; intrathematic groupings comprise the functional
Table 3.1. Grouping hierarchy
Grouping level Content
Genre Concerto, symphony, sonata, quartet, and so forth
Cycle First movement, Adagio, Scherzo, Finale, and so forth
Total form Sonata form, rondo form, and so forth
Large-scale Exposition, development, recapitulation, ritornello 1,

solo 1, and so forth
Interthematic First theme, second theme, transition, closing section,

and so forth
Intrathematic Presentation, continuation, cadence (sentence);

antecedent, consequent (period), and so forth
Motivic “a,” “b,” and so forth

82 julian horton
divisions of a thematic group. Moving up the hierarchy, we might situate total

forms within movement cycles, and movement cycles within genres; moving
down the hierarchy, we can identify the motivic level as the smallest meaning-
ful formal subdivision.13 Genres might be termed interopus groupings; move-
ment cycles operate at the level of the work (opus); and large-scale, intra- and
interthematic groups are all intraopus categories.
Whereas this hierarchy is, for the present purposes, a fixed construct, func-
tions may exhibit hierarchical mobility, the capture of which leads me to mod-
ify further Caplin’s usage. To the extent that, for example, an expositional first
theme and a sentential presentation phrase both introduce material that begs
continuation, they are both presentational in function, but at different hier-
achical levels. Similarly, “continuation” might apply intrathematically within
a sentence, or interthematically to a transition, which is altogether continu-
ational beyond the first-theme group, indicating a distinction between intrath-
ematic and interthematic continuation. My broadening of usage also operates
between thematic types. A presentation phrase initiates a sentence, and an
antecedent phrase initiates a period, but both facilitate the process of interthe-
matic presentation, that is, the presentation of a first theme.
We can formalize these observations by differentiating between general-
ized and level-specific functions, and adapting Caplin’s usage accordingly.
Presentation, in these terms, is a generalized function because it can serve
intrathematic, interthematic, or large-scale ends: statements are presentational
within a sentence; sentences can be presentational within an exposition; and
expositions can be presentational within a sonata form.14 Other functions are
more hierarchically localized. Cadential functions, although structurally vital,
are restricted to the intrathematic level: it would mean very little to describe
a second-theme group altogether as cadential.15 Table 3.2 elaborates this
perception, differentiating level-specific and generalized functions within a
hypothetical sonata form, and nominating four of Caplin’s categories (presen-
tation, continuation, cadence, framing function) and two additional categories
(re-presentation, linkage).
In fact, this perspective can be useful in the analysis of larger Viennese-
classical designs. The sentential period is a concise example; an instance taken
from the last movement of Mozart’s Piano Sonata K. 309 is shown in example
3.1. Altogether, this theme forms a single thematic group. Its interior, how-
ever, expresses a proliferation of lower-level groupings, which replicate func-
tions. At the first level of segmentation, the periodic type predominates: in this
context, the antecedent has to be regarded as presentational, the consequent
as an end function, and both are demarcated respectively by medial and con-
cluding cadences. At the lower grouping levels, presentation, continuation,
and cadence are all present within the antecedent and consequent, since each
consists of a sentence, the latter modified by repetition and the installment
of a perfect authentic cadence. The functions comprise groupings at every

Table 3.2. Generalized and level-specific functions within a hypothetical

sonata form
Total form: sonata form
Function Grouping level
Presentation Intrathematic level: statement; basic idea
Interthematic level: first theme; second theme
Large-scale function: exposition
Re-presentation Intrathematic level: repetition
Interthematic level: reprise
Large-scale function: recapitulation
Continuation Intrathematic level: continuation; contrasting idea;

model–sequence
Interthematic level: transition
Large-scale function: development
Cadence Intrathematic level: cadential function; contrasting idea
Linkage Intrathematic level: caesura fill
Interthematic level: transition; retransition
Large-scale function: development
Framing Intrathematic level: anacrusis; codetta
Interthematic level: Eingang or thematic introduction;

closing section
Large-scale function: introduction; coda
level; but the groupings are ramified at a higher level into a single functional
span—the main theme—which performs one function (it is interthematically
presentational). In brief, the concatenation of functions always defines group-
ing, even though groupings do not always demand “a single functional label.”
Broadly speaking, the difference between function and type reduces to the
differentiation of the formal task that material performs and the organiza-
tional conventions it adopts in so doing. As Caplin has noted:
I see classical form arising out of a common set of formal functions, which
are deployed in different ways to create multiple full-movement types. The

84 julian horton
Example 3.1. Mozart, Piano Sonata K. 309, mvt. 3
common element is not [for instance] sonata form per se, but rather the
functions that make up the various forms. Thus we can recognize the appear-
ance of subordinate-theme function . . . in a short minuet form, in a moder-
ately sized rondo form, in a large-scale concerto form, and, of course, in a
sonata form.16
In other words, while types are defined by a particular arrangement of func-

tions, functions are expressed in a plurality of possible types.
As a general rule, classical types are level-specific concatenations of func-
tions. The thematic type “period” is formed from lower-level intrathematic
functions, and does not normally migrate upward through the hierarchy (we
would never encounter a “periodic” exposition). Similarly, the total-formal
type “sonata form” arises from a specific arrangement of large-scale functions,
which tolerates only a certain amount of variation before the typical label
becomes invalid (the extent of its elasticity is of course open to debate). In

brief, formal typology involves the classification of a set of functional concat-

enations across the gamut of intramovement grouping levels, which exhibit
sufficient consistency as to suggest membership of a single total-formal class.
As we shall see, however, mid-century developments in formal practice—most
obviously the conflations of form and cycle in Liszt’s so-called double-func-
tion forms, recently theorized by Steven Vande Moortele as “two-dimensional
form”—have ramifications for the level-specificity of types: where sonata form
and four-movement cycle “identify” (to use Vande Moortele’s term), a formal
and a cyclical type have come to coexist at the same grouping level.17
This chapter adds analytical substance to these observations by investigat-
ing five central aspects of concerto first-movement form in this repertoire:
ritornello first-theme syntax (R1 A); solo first-theme syntax (S1 A); B-theme
syntax (R1 and S1); disposition of cadential functions; and whole-movement
form and the conflation of grouping levels (“two-dimensionality”). A glossary
of terms used is given in table 3.3. Of particular concern is the question of
how lower-level syntax conditions form, fundamental to which is the tendency
toward intrathematic and interthematic expansionism, a practice that can
seem radically promiscuous in Viennese-classical terms.
Form, Function, and Syntax in the

Postclassical Piano Concerto
Thematic Syntax and Intrathematic Proliferation
As we move into the nineteenth century, a number of syntactic habits arise,
which, despite owing some allegiance to classical precedent, often evade ade-
quate description in terms predicated purely on the music of Haydn, Mozart,
and Beethoven. One important such characteristic is the tendency of nine-
teenth-century themes toward a degree of syntactic looseness, which exceeds
classical precedent in two ways: first, through the novel arrangement of func-
tions within a recognizable classical design; and second, through the propa-
gation of functions within a broader grouping level. I call such functional
promiscuity proliferation, since it projects the impression that the core mate-
rial is generating a plurality of continuations within the remit of a given inter-
thematic or large-scale context. This habit is manifest in all thematic regions,
mindful of Caplin’s observation that “loose” design is generally a subordinate-
theme property in Viennese classicism.18
Generally, such designs arise because composers marshal aspects of clas-
sical syntax in novel formations, augmenting them by cadential expansion
or the insertion of consecutive continuation functions to produce expansive
hybrid or compound forms. A brief survey of expositional thematic syntax
in postclassical piano concerti serves to illustrate this. The most compact

86 julian horton
Table 3.3. Glossary of terms and symbols

Term Meaning
A First theme
A1, B1, and so forth Reprise
A1, A2, A3; B1, B2, B3, and so forth New material under the same function
a1, a2, a3; b1, b2, b3, and so forth Motives within the same function
B Second theme
C Closing section
DE Display episode
HC Half cadence
IAC Imperfect authentic cadence
MC Medial caesura
PAC Perfect authentic cadence
RT Retransition
R1, R2, R3, and so forth Ritornello succession
S1, S2, S3, and so forth Solo succession
TR Transition
⇒ Functional transformation
(“becoming”)
 Elision
→ Modulation
themes in this repertoire often seem closest to Mozartian precedent. The R1

A of Field’s Concerto no. 1, first movement, of 1799, shown in example 3.2,
for instance, comprises a sentential period. Field’s theme consists of a sixteen-
measure period,19 in which both antecedent and consequent are fully senten-
tial. Each presentation phrase comprises a thematic statement and response;
the continuation phrases deploy a model-sequence technique accelerating the
presentation’s harmonic rhythm; and the antecedent’s half close is balanced by
a perfect cadence at the end of the consequent, which is elided with the start
of the transition.
Evolving in London in the 1790s, Field’s early concerted style owes much
to that of Dussek, although Dussek’s thematic designs can seem somewhat less
orthodox. The R1 A of Dussek’s opus 14, composed in 1791 and quoted in

Example 3.2. Field, Piano Concerto no. 1, mvt. 1: R1 A
example 3.3, also consists of a compound periodic design, but one in which
both antecedent and consequent phrases divide into statement-response, and
a single grouping functioning both as continuation and cadence.20
The bifocal historical perspective of much of this early postclassical reper-
toire, as a body of music that both absorbs and exceeds the purview of late
eighteenth-century style, is very clear here. In Dussek’s opus 14, the arrange-
ment of functions is consistent with designs apparent in much earlier music,
but the result is a theme that appears comparatively unbalanced because the
medial and concluding cadences seem premature in relation to the dimen-
sions of the groupings they succeed.

88 julian horton
Example 3.3. Dussek, Piano Concerto, op. 14, mvt. 1: R1 A
Hummel’s six published concerti, written between 1811 and 1833, furnish
more complex examples.21 The R1 A of his Concerto, op. 113, first movement
(1827) is given in example 3.4. The theme is distinctive for its displacement of
functional characteristics. Measures 1–4.3 function as a compound basic idea,
from which measures 4.3–8.1 flow as a continuation. This continuation, however,
has the hallmarks of a Schoenbergian liquidation, because Hummel here iso-
lates a single motive from the statement (marked “a3” in ex. 3.4) and diminutes
it toward the phrase’s apex in measure 7. Measures 8–11.2 occupy the position
of a cadential function; yet any suggestion that the V7 chord in measure 8 will

resolve cadentially is dispelled by the deceptive move toward vi, and the tonic
is attained in measure 11 via a linear bass ascent. As a result, A♭ is confirmed
as tonic by the voice leading, but not by a cadential bass progression. Hummel
seems to dissociate function and rhetoric (the expected cadential function fails
to materialize) and also to relocate associated thematic processes (liquidation
occurs in the continuation, not the cadence). At the next grouping level, we
see that this whole unit acts as a presentation phrase, initiating a four-measure
continuation in measures 12–16.1, a deceptive cadence in measures 16–19, and
finally an expanded cadential progression (ECP) leading to a PAC in measures
20–23.1. Altogether, the theme resembles an expanded sentence, in which the
presentation phrase has the grouping structure, if not the precisely ordered
harmonic or thematic characteristics, of an antecedent–continuation⇒cadence
hybrid. But the interior of Hummel’s theme seems to challenge some of the
basic harmonic-rhetorical associations of the high-classical style.
Example 3.4. Hummel, Piano Concerto, op. 113, mvt. 1: R1 A

90 julian horton
The R1 A of Hummel’s opus 89 (ex. 3.5), composed in 1819, is more intri-

cate again. After a two-measure introduction, Hummel launches a thematic
design that, at its most complex, yields no fewer than five intrathematic layers,
two of which conflate competing functions into the same grouping. On the
largest scale, the theme has the character of an expanded sentence. The pre-
sentation phrase is, however, a multilayered compound design, consisting of a
statement and response, which can also be viewed respectively as the antecedent
and consequent of a sixteen-measure period (the vertical square brackets indi-
cate this conflation). The basic idea is moreover in each case itself a compound,
devolving into the statement and response of a critical Hauptmotiv (marked “a”).
Hummel’s continuation falls into three leisurely stages, described as continua-
tion phrases 1, 2, and 3 in example 3.5, which express the characteristics of a
sentential continuation, in that they progress from the statement and repetition
of a model derived from the presentation phrase (“a”), which is then subject to
fragmentation as the continuation proceeds, eventually reducing to the second
half of “a” alone. Here, however, the process is considerably expanded, unfold-
ing over a tonic–dominant vamp in measures 17–27.1 and a chromatic octave
bass ascent in measures 27–40 before the group is rounded off by a four-measure
PAC elided with the first measure of TR.
Whereas R1 A themes tend to expand thematic types from within, S1 A
themes betray a consistent multipart design, which is quite different from
Mozartian precedents.22 This design involves the presentation of a (frequently
new) solo-entry first theme in two topically distinct intrathematic groupings: a
bravura preface, which is commonly periodic, or employs an expanded state-
ment-response design; then a nocturne-like “cantabile episode” (my term),
which can exhibit anything from a periodic to a small-ternary form. Solo-entry
themes are in consequence distinguished by a rather different kind of prolifera-
tion than R1 A themes and tend toward a bipartite form that very often cannot
be described under an expanded unitary thematic type. This functional plurality
reflects a topical discourse that is basic to postclassical virtuosity: the preface and
cantabile episode express an essential contrast between bravura pianism and an
imitation of song reflecting operatic idioms, especially early nineteenth-century
bel canto opera. Taken together with the brillant transition that normally ensues,
this functional division of labor embodies in nuce the three topical attitudes of
the virtuoso concerto: bravura, brillant display, and song.
Examples 3.6–3.8 show the S1 A and start of TR in three concerti by Dussek:
opus 29, published in 1795; opus 49 of 1801, his first to be composed after
leaving London; and his last concerto, opus 70 of 1810. In opus 29, the preface
consists of a lavish sentence, split into a chordal statement and response, and a
continuation and cadence constructed from a series of arpeggiated and scalar
flourishes. A brief cantabile episode follows, which appears periodic, although
the consequent’s final cadence is undercut by the tonic pedal persisting in
measures 112–13.1. In opus 49, the topical succession is more succinct, but

Example 3.5. Hummel, Piano Concerto, op. 89, mvt. 1: R1 A

Example 3.6. Dussek, Piano Concerto, op. 29, mvt. 1: S1 A (solo only)


94 julian horton
nonetheless apparent; the harmonic details and thematic rhetoric, how-

ever, suggest competing functional readings. The balance of phrases implies
two periods, since both preface and episode devolve into binary designs of
near-equal length (two seven-measure phrases in the preface; two eight-
measure phrases in the episode). Yet the phrase endings in the preface
are not cadential, a feature that lends them the character of the statement
and response of a sentence, albeit one from which no orthodox continua-
tion ensues. Altogether, we can read the group as two types in a compound

design (period–period), or as one highly irregular hybrid (statement–

response–antecedent–consequent), in which the statement is formed as a
compound basic idea. More ingeniously, the preface in opus 70 is built from
a figure introduced at the end of R1, which initiates a compound basic idea–
continuation⇒cadence design forming the antecedent to a large period, the
consequent of which is supplied by the cantabile episode. The display epi-
sode follows from measure 130 after a PAC.
Other composers adopting this model sometimes truncated it. Many
of Johann Baptist Cramer’s S1 A themes, for example, move straight from
preface to display episode; this practice was subsequently taken up by
Sterndale Bennett (for instance in the Concerto no. 3 of 1836). Later solo-
entry first themes often betray a heightened sense of continuity, or of inte-
gration with R1, features that perhaps respond to the intervening spread
of Viennese examples, as well as to an emergent sense of the significance
of “thematic logic,” which comes to the fore in the concerti of Schumann
and Liszt. The S1 A of Moscheles’s opus 58 (ex. 3.9), completed in 1820 and
performed twice in London in 1822, evinces both Dussek’s first-theme rheto-
ric and post-Beethovenian strategies of integration. Moscheles begins with
a statement–response preface presenting new material; thereafter, cantabile
and display elements alternate within three phrase groups—a presentation
and two continuation phrases, each delineated by mediant half cadences.
The persistent harmonic instability of the episode, and especially its rapid
turn toward III, however, undermines any clear differentiation of first-theme
and transitional functions; as the open-headed arrow in example 3.9 indi-
cates, it is perhaps more appropriate to state that A1 “becomes” TR, in the
sense of functional transformation theorized by Janet Schmalfeldt.23 The
association of transition and display is not properly consolidated until the
medial caesura, reached at measure 167. Moscheles is also more motivically
minded than Dussek. The cantabile episode presents a variant of R1 A, yield-
ing two strands of motivic action: measures 141–43 extract and liquidate a
motive labeled as “a1”; measures 147–52 invert the whole of the cantabile-
episode statement. In fact, the movement is monothematic, deriving R1 and
S1 A, B, and C in from the same material.
A more rhetorically straightforward example, which continues Moscheles’s
trend toward R1–S1 integration, is found in Chopin’s opus 11, completed in
Warsaw in 1830 (ex. 3.10).The preface–cantabile episode–display episode suc-
cession is retained and sharply articulated. The preface is periodic and invokes
Dussek’s prefatory rhetoric; the cantabile episode comprises a small-ternary
form. Syntax, topic, and cadence are precisely aligned: the preface closes with
an imperfect authentic cadence; the cantabile episode with a PAC. At the same
time, the integration of S1 and R1 resembles Moscheles’s opus 58, because
both S1 A elements have an R1 precedent: the preface is based on R1 A; the
cantabile episode on R1 TR.

Example 3.9. Moscheles, Piano Concerto, op. 58, mvt. 1: S1 A (solo only)
(continued)


Example 3.10. Chopin, Piano Concerto, op. 11, mvt. 1: S1 A (solo only)
(continued)

Whereas A-theme syntax in this repertoire is marked by expansionism, B

themes are, at least in the early postclassical repertoire, often more compact.
They are, moreover, frequently shared between R1 and S1, contrasting the ten-
dency to differentiate orchestral and solo A themes, and fundamentally depart-
ing from Mozart’s practice, which usually allots the soloist distinctive B-theme
rather than A-theme material. Nevertheless, the repertoire yields some strik-
ingly postclassical designs. Dussek’s opus 14, movement 1 and Cramer’s opus

100 julian horton
10, movement 1, summarized in Tables 3.4 and 3.5, exhibit an alternation of

B-theme and display materials that has no Mozartian parallel. The B groups
of both first movements comprise a double thematic presentation framing
a central display episode. Dussek’s initial B presentation (mm. 147–54) is a
period, in which the antecedent concludes with a half cadence in V and the
consequent with a V/V PAC. There follows a twenty-four measure display epi-
sode having the character, if not the postcadential function, of a closing sec-
tion. This culminates in a caesura over V36/V, after which a three-measure fill
leads to the second B-theme presentation at measure 179, which reprises the
antecedent from measures 147–50, but replaces the original consequent with
a four-measure continuation and PAC. The closing section proper ensues from
measure 186.
Cramer’s opus 10 is a more difficult case. The B theme itself is more func-
tionally complex than Dussek’s; and the fact that both B-theme presenta-
tions close with dominant PACs generates problems of interthematic analysis,
because Cramer seems to have duplicated the definitive dominant cadence
(sonata theory’s “essential expositional closure”) demarcating the end of B
and the attainment of the exposition’s tonal goal.24 As table 3.5 explains, the
first B-theme presentation, measures 173–85.1, begins with an antecedent–
continuation⇒cadence design concluding with a half close. Cramer follows
this with a consequent beginning on the supertonic in measure 181. Yet the
response to measures 173–80 is unbalanced; rather than answering the ante-
cedent hybrid with a modified variant to produce a compound period, the
music moves swiftly to a PAC, generating an unorthodox hybrid–consequent
form. The second B presentation, measures 199–212.1, is a variant of this
design, as table 3.5 also shows, elongated by the addition of a two-measure fill
between the half cadence and the consequent, which then reproduces exactly
the PAC from measures 184–85.1.
The situation gains complexity when we realize that the two display episodes
begin with the material that initiates S1 TR in measure 145. This means that
the B-theme presentations are in effect framed and punctuated by three mate-
rially similar display episodes. The result is the ambiguity explained in table
3.5: either there is an interthematic division at the start of B, such that it alto-
gether comprises a ternary design with a brillant contrasting middle; or both
B presentations are episodes within a five-part design beginning with TR. The
tonal scheme emphasizes the former, since TR modulates, whereas the B–DE–
B1–C(DE1) scheme prolongs V; the material rhetoric however favors the latter,
cutting across the tonal structure and imbuing the B presentations with the
character of contrasting episodes.
All of this is quite different from Mozart’s habit of contrasting a unitary S1 A
theme derived from R1 A with a succession of novel S1 B themes. Dussek’s and
Cramer’s subordinate-theme looseness stresses correspondence between ritor-
nello and solo by playing on the expected affiliations of topic and function.

Table 3.4. Dussek opus 14, movement 1: S1 B and C
Measures 147–54 155–78 179–861 186–2151
Large-scale S1
function 1
Large-scale Exposition
function 2
Interthematic B TR (DE1) B1 C (DE2)

function
Intrathematic period expanded period 1. 186–981 11. 198–2151

function/type cadential→ postcadential→ postcadential→
standing on V/V cadence cadence
Tonal plot V → V/V PAC→V → PAC → PAC → PAC
9/30/2015 7:51:37 PM
Table 3.5. Cramer opus 10, movement 1: S1 A, B, and C
Mm.: 122–44 145–72 173–851 185–98 199–2121 212–331
Large-scale S1
function 1
Large-scale Exposition
function 2
Interthematic A1 (bravura TR (DE) B DE1 (TR-based) B1 C (DE2: TR-based)
function preface)
Intrathematic expanded model– hybrid postcadential hybrid postcadential →
function/type period sequence → ⇒ contrasting middle standing on V/V →
standing on cadence
V/V
Tonal plot I → V:HC MC V → PAC → PAC → PAC
competing
segmentations:
9/30/2015 7:51:37 PM
Display episodes implying expositional closure are diverted back into B-theme
presentations, in Cramer’s case producing a multiplicity of structural perfect
cadences, which problematizes the differentiation of B and C.
Cadence, Prolongation, and Functional Articulation

The disposition of cadences in Cramer’s opus 10 underscores the need to
take account of the evolving functional role of cadence and prolongation in
this repertoire, in addition to scrutinizing its predominant thematic types.
Generally speaking, the syntactic function of the cadence follows two opposed
developmental paths: either cadential functional markers are increasingly
sharply articulated, or else they are dissolved into prolongation or displaced to
classically unorthodox locations.
The first trend is clarified in example 3.11, which appraises the distribu-
tion of cadential functions in three solo expositions by Cramer—from opus
26, 38, and 51, respectively—composed between 1801 and 1811. It is clear that
the relationship between interthematic subdivision, topical discourse, and
authentic cadence in these movements is invariant; notwithstanding changes
in intrathematic organization, interthematic functional divisions, and shifts of
topic remain precisely coordinated with cadential closure. As already observed,
the S1 cantabile episode is rare in Cramer’s music; rather, an S1 A character-
ized by homophonic, chordal rhetoric leads, via a perfect authentic cadence,
into a brillant display episode signaling the transition. This in each case pro-
duces a dominant half-close caesura preparing a cantabile B theme, which
itself elides with a display episode closing section via a PAC. Cramer repro-
duces these practices precisely, without any of the diversity of TR and B-group
rhetoric and organization explored in Mozart’s mature concerti.
Example 3.12, by contrast, shows R1 A and TR in the first movement of
Chopin’s Concerto op. 21 of 1829. The A theme amounts to an antecedent–
continuation⇒cadence hybrid, beginning with an antecedent phrase that
moves to a medial half close, but acquiring the properties of a sentence from
the middle of measure 4, which initiates a continuation built from single-mea-
sure units and a closing gesture rounding the phrase off with the tonic.
The final segment is, however, not properly speaking cadential at all, instead
comprising a V36–i progression briefly extended by means of a voice exchange
in the outer parts. By Caplin’s terms, this phrase is incomplete, because it has
no concluding cadential function. Yet no reasonable analysis would describe
what happens next as an intrathematic continuation. Chopin supplies new
material from measure 9, which in terms of texture and rhetoric strongly
resembles a transition. In fact, no clear cadential progression occurs until
the tonic half-close medial caesura ending the transition in measures 35–36;
and no authentic-cadential closure occurs until measure 38, at the end of the
first phrase of the B theme and in the relative major. Strictly speaking, all the

Example 3.11. Cramer, Concerti, op. 26, mvt. 1, op. 38, mvt. 1, and op. 51, mvt. 1:
disposition of structural cadences in S1 (solo only)

Example 3.12. Chopin, Piano Concerto, op. 21, mvt. 1: R1 A, TR, and start of B
(continued)

106 julian horton
material of measures 5–35 is in some sense prolongational; if we cleave to

Caplin’s Viennese syntactic criteria, then we arrive at the odd conclusion that
the entire thirty-six-measure span comprises one giant compound thematic
type generated by the accretion of multiple lower-level groupings, which are
consistently denied cadential articulation. This would mean that there is no
TR, but a large loose-knit A group proceeding directly into the B theme. In
brief, if—as Caplin has noted in his engagement with the analytically notorious
first theme of Beethoven’s “Tempest” Sonata—classical main themes capital-
ize on a tension between syntactic stability and textural instability, then the

analytical problem to which Chopin’s A and TR give rise is that the syntax has
in key respects become as unstable as the texture.25
Chopin’s music might be served better by a revised set of syntactic priori-
ties. Above all, here we have to abandon the necessary correlation of cadence
and intrathematic ending, and simultaneously admit concepts of thematic liq-
uidation and noncadential tonic projection as meaningful alternatives. Thus,
although Chopin’s measures 6–8 are not cadential, they are plainly liquida-
tory in Schoenberg’s sense, being constructed almost entirely from the ante-
cedent’s anacrusic Hauptmotiv, which is inverted toward the end of measure
6, divested of its initial sixteenth note on the second beat of measure 7, and
stated twice in its prime form in the bass and in inversion in the violins. This is,
moreover, a compression of the motivic action of the continuation, which com-
prises a twofold assertion of the Hauptmotiv punctuating a dominant prolonga-
tion. We thus have to distinguish here between ending and closure: the theme’s
ending is signified by motivic fragmentation and tonic assertion, but it is not
synonymous with cadential closure. Like many nineteenth-century composers,
Chopin exploits the structural and expressive potential of cadential deferral,
but this does not bring about the collapse of interthematic categories. Instead,
we have at once to grasp the articulation of interthematic functions by motive
and rhetoric and their subsumption into large continuous bass progressions,
within which perfect authentic cadences are few and carefully located.
One consequence of these practices is an increased propensity for struc-
tural elision: both interthematic and large-scale functions often begin before
the bass has moved to articulate closure. Example 3.13 shows the end of
TR, the start of B, and aspects of its continuation in Mendelssohn’s Piano
Concerto, op. 25, movement 1 of 1831. The transition arrives on V/III at
measure 71, but this arrival occasions no bass motion onto III for the start of
the B theme. Rather, the bass V remains active beneath the initial B-theme
statement and response in measures 73–76.26 More radically, Mendelssohn
does not then proceed to a clear continuation phrase in this key; instead, he
maintains the F pedal, but engineers a mode switch in measures 77–78, pro-
voking a modulation to D-flat, in which key the music cadences with a PAC in
measures 81–83.1. A second presentation then sets off in D-flat, spawning a
series of continuation phrases in measures 88–108, culminating in measures
105–8, where a single motivic residue is liquidated over vii°7/V in D-flat. No
D-flat PAC however ensues; Mendelssohn pivots to B-flat minor via a recall of
A-theme material, before reprising B over III36 in measure 117 above display
episode rhetoric in the piano.
In classical circumstances, Caplin would perhaps argue here for a loose B
group with an internal harmonic digression, awaiting a PAC that will demarcate
the true closing section. Yet no such PAC arrives. There is a perfect cadence
in III in measures 124–25, rounding off the presentation of B begun in mea-
sure 117; but this is rhetorically subsumed into the grouping that measure 117

Example 3.13. Mendelssohn, Piano Concerto, op. 25, mvt. 1: summary of B,
showing relationship between phrase endings and structural cadences
(continued)


110 julian horton
initiates, since measure 125 sets off a continuation returning to the material
and texture of measure 117. The trill over V/III reached in measure 152 seems
to herald the exposition’s end, presaging a PAC and postcadential ritornello.
Neither, however, appears: V/III is deflected to ii56/iv in measure 155, after
which the soloist carries the music into a brief development section without
cadential punctuation.
Mendelssohn’s design plays havoc with classical functionality because the
intimate relationship between cadence and thematic rhetoric is dislocated.
Theme B enters before the medial standing on V has completed; this V is not
resolved, and the theme’s presentation phrase cadences in ♭V. The second
attempt at stable presentation contrastingly begins over a local root-position
tonic, but is denied cadential closure, leading at length back to III. The closing
section then functions as a stable mediant presentation of B, which is elided
with the development through cadential evasion. Neither putative subordinate
tonality is allowed fully to stabilize, because Mendelssohn never permits pre-
sentation and continuation to lead on to a structural cadence in the same key
and be confirmed by postcadential material.
Mendelssohn’s and Chopin’s practices have profound implications for
sonata-theoretical analysis. The problems engendered in Mendelssohn’s opus
25 fundamentally reorient the debate between William Rothstein, Caplin, and
Hepokoski and Darcy about how to locate an expositional closing section, cen-
tral to which is the question of whether subordinate-theme ending is defined
by the “first satisfactory perfect authentic cadence that proceeds onward to dif-
fering material,” as Hepokoski and Darcy put it, or whether, as Caplin con-
tends, subordinate themes close with the final PAC of the exposition.27 Such
arguments rely at base on the decisive presence of perfect cadences. But the
examples considered here marshal recognizable expositional intrathematic
syntax and interthematic rhetoric, while radically undermining the crucial role
that cadence plays in this context.
Formal Consequences of Proliferation: Normative Schemes,

Formal Conflation, and “Two-Dimensional” Form
The syntactic formations appraised here evolve in the 1790s in response to
a complex of systemic, stylistic, and technological factors. In part, they are
bound up with the topical and stylistic changes compelled by developments
in instrumental technology. The concerted styles of Dussek and others both
responded to and provoked changes in piano design, particularly as pioneered
by Broadwood in London in that decade, and from 1800 by Erard in Paris.
Two developments are especially noteworthy: first, the instrument’s compass
expanded from five to six octaves between 1790 and 1792; second, advances in
sustaining-pedal design enabled a greater variety of left-hand accompanimen-
tal patterns, since the registral separation of bass and internal parts becomes

progressively easier.28 These changes in turn consolidated the topical attitudes

described above: bravura and display idioms capitalized on the instrument’s
expanded compass; changes of pedaling capability enabled a cantabile idiom,
which eventually becomes the nocturne style. The solo syntax of the virtuoso
concerto evolved as a formal-functional vehicle for these idioms, installing a
system of syntactic conventions that work closely with topical discourse.
Attempts to accommodate these priorities at the level of whole-movement
forms produce a concept of first-movement form that is quite different from its
Mozartian forebear.29 Contrasting Mozart’s tendency to favor monotonal ritor-
nelli, the modulating R1 overwhelmingly predominates, often cast as a loose
ternary design, in which R1 B is framed by two A-based tonic regions. This pro-
motes the sense of a double exposition that remains weak in Mozart’s concerti,
because S1’s tonal dialectic is anticipated in R1. The form is also distinctive in
the relationship between R1 and S1 material. Whereas Mozart tended to allo-
cate common A material and contrasting B material to soloist and orchestra,
virtuoso composers after Dussek favored the opposite strategy: R1 B and S1 B
correspond; R1 A and S1 A differ. This is a direct result of the incorporation
of keyboard idioms and their associated topics: material appropriate to an S1
bravura preface or cantabile episode is hard to integrate as R1 A material; but
it becomes a first principle of the form that the soloist should expose the style’s
basic topical attitudes immediately. Movements exhibiting R1 A/S1 A corre-
spondence usually impose the topical characteristics of the bravura preface or
cantabile episode on material that had been differently treated in R1. Chopin’s
opus 11 offers a clear instance: R1’s martial A theme becomes infused with
bravura figuration in S1’s preface; and R1 TR is reinterpreted by the soloist as a
cantabile episode preceding S1 TR.
This interaction of syntax and topic accounts for postclassical variation
across the form. S1 A and TR express in nuce idioms that inform the entire
design, consolidating an alignment of syntax, topic, and formal function that
is far less fluid than in Mozart’s concerted style. The brillant manner apparent
in S1 TR is standardized as the basis of S1 C, S3 C, and the development core.
The nocturne style recurs in S1 B and frequently in the development precore,
where it is sometimes conjoined with a variant of the S1 preface. Three vari-
ants predominate at the point of recapitulation. The most direct strategy here
treats R3 as the first-theme recapitulation, and commences S3 with a reprise of
the cantabile episode or TR (Chopin’s opus 11, mvt. 1 and Field’s no. 7, mvt.
1, for instance). Alternatively, R3 can also function entirely as a retransition, so
that S3 is aligned with the start of the recapitulation (as in Chopin’s opus 21,
mvt. 1, although note here the drastic truncation of S1 A and TR that follows).
As a middle ground between these two extremes, R3 can sometimes encom-
pass both RT and the start of the recapitulation (Field’s Concerto no. 3, for
example). The cadenza in R4, a mandatory characteristic of Mozart’s forms, is
eliminated by Dussek in the 1790s and becomes a rarity thereafter. Composers

112 julian horton
restoring the cadenza often relocate it (as in Moscheles’s opus 58, in which
the cadenza appears in S3 C, or Ries’s opus 132 of 1823, where it forms the
substance of TR in the recapitulation). Beethoven’s habit of enlarging the
cadenza in its Mozartian location is certainly anomalous in its time. Later
attempts to revisit Beethoven’s practice, for instance, the first movement of
Schumann’s opus 54, tend to integrate the cadenza in a decidedly postclassical
fashion. Thus Schumann begins his first-movement cadenza over a German-
^
sixth chord, and leaves it via a sustained trill on 5 , which fails to cadence before
the orchestra enters.
The key factor in determining the dimensions of this form is proliferation:
the essentially unclassical proportions of the virtuoso first movement are a con-
sequence of its tendency to enlarge interthematic groupings to accommodate
the style’s generative topical discourse. By the 1830s and 1840s, this tendency
had (paradoxically) spawned higher-level processes of conflation, whereby pre-
viously distinct functional levels collapse into each other. This is manifest ini-
tially in movements dissolving the hybrid form’s two-tier large-scale functional
division (R–S succession and exposition–development–recapitulation) into
one sonata scheme, notably in Mendelssohn’s two concerti of 1831 and 1837,
respectively. More radical still is the emergent habit of collapsing the differen-
tiation of form and cycle, described by William S. Newman as “double-function
form” and theorized more extensively by Vande Moortele as “two-dimensional”
sonata form; the locus romanticus, as Vande Moortele puts it, is of course Liszt’s
B-minor Piano Sonata.30 This is consistently a consequence of proliferation.
In Liszt’s sonata and other similar works, the functional “identification” of the
large-scale divisions of the sonata form and the cyclical division of movements
is enabled by the tendency to compose extended compound designs at the
interthematic level.
Prior to Liszt’s Concerto no. 2 of 1848–53, conflation is a rare phenom-
enon at higher functional levels. Two related, but rather weakly projected,
instances can be found in the first movements of John Field’s Concerto no. 7
and Schumann’s Concerto op. 54.31 Field capitalizes on the widespread ten-
dency in contemporaneous piano concerti to begin the development with a
solo nocturne episode, expanding the first stage of the development precore
into an entirely self-contained nocturne, which he later extracted and pub-
lished separately as his Nocturne no. 12. Retrospectively, we see that this noc-
turne is also the slow movement: the concerto is putatively a three-movement
cycle, but the slow movement is displaced into the first movement, where it
functions as an interpolation between the end of R2 and the development
precore. Field’s movement is, however, a long way from Liszt’s practice:
first and slow movements are conflated by interpolation alone; there is no
broader identification of higher functional levels, nor is there the marked
proliferation of interthematic functions that is the lower-level correlative of
this strategy in Liszt’s music.

These characteristics are more tangibly nascent in Schumann’s A-minor

Phantasie of 1841, although they are somewhat blunted in the music’s final ver-
sion as the first movement of opus 54, which is of course a three-movement
concerto. Schumann voiced the aspiration to create a concerted work min-
gling aspects of form and cycle in 1836, when he observed the need for “a
type of one-movement composition in moderate tempo in which an introduc-
tory or preparatory part would take the place of a first allegro, the cantabile
section that of the adagio and a brilliant conclusion that of the rondo.”32
These remarks have been taken as the basis for a “double-functional” read-
ing by Claudia Macdonald, who posits a sonata form that mimics the “allegro-
adagio-rondo sequence,” giving the design outlined in table 3.6.33 In fact,
only the nocturne conveys any sense of self-containment, exhibiting a closed
phrase design, a unity of topic, and prolonging a single key (♭I). The result
is that Schumann’s nocturne has the character of an interpolated movement,
whereas the other cyclical elements “identify” with their corresponding large-
scale formal regions.34
The extent of this identification is, however, limited. Aside from the obvi-
ous fact that the movement is a unitary sonata rather than a sonata-ritornello
hybrid, the main feature distancing Schumann’s exposition from earlier prac-
tice is the extent of interthematic proliferation, which generates the (albeit
relatively weak) illusion that a single exposition and an entire movement
form share the same formal span. The A theme itself (mm. 4–19.2) is unam-
biguous, comprising a period in which the antecedent is orchestral and the
consequent is taken by the soloist, the latter closing with a PAC. Thereafter,
however, the distinction between transition and subordinate theme is consis-
tently blurred, thanks to the tendency for apparent thematic presentations to
collapse into transitions. In other words, the music exhibits interthematic prolif-
eration: Schumann composes neither a single subordinate theme nor a more
loose-knit multipart group, but a succession of subordinate-theme candidates
linked by transitions. The first instance of this occurs in measures 36–58 (ana-
crusis m. 35). Measure 36 seems to initiate a B theme, a sense reinforced by the
arrival on III, concomitant resolution of the preceding G harmony and pre-
sentation of fresh material (admittedly adumbrated in the closing stages of the
Table 3.6. Schumann opus 54, movement 1: two-dimensional reading
Mm.: 1–155 156–84 185–258 259–401 402–57 458–544
Cycle: first slow finale

movement movement
Large-scale exposition develop- recapitu- cadenza coda

function: ment lation

114 julian horton
transition, beginning in m. 25.4). Yet this music struggles to maintain its pre-
sentational function, and by measure 48 has decisively yielded to a new transi-
tional phase, the goal of which is the return of theme A at measure 59, at which
point the impression of a monothematic exposition starts to emerge. But
again, the sense of arrival here is premature; although measures 59–66 appear
loosely sentential (presentation mm. 59–62; continuation 62.4–64; authentic
cadence mm. 65–67.1), they begin over a 46 chord and only find their point of
cadential focus going into the section beginning at measure 67.
Measures 67–102 have a stronger claim to presentational stability, supply-
ing a cadentially open-ended small-ternary design (statement–response–con-
tinuation in mm. 67–76; contrasting middle in mm. 77–94; reprise in mm.
95–102) and confirming an A-based B theme grounded in a tonicized III.
Yet if the prior material projects a consistent tendency toward adumbra-
tion, then here the music’s B-theme status is challenged because it is more
strongly redolent of a closing section, an impression reinforced by its display-
episode character; the proliferation of transitional functions seems, in short,
to have engulfed the B theme’s interthematic space, causing it instead to be
projected onto the closing section.
Two features prevent us from relaxing into formal certainty here. First, even
though the exposition is rhetorically closed with the tutti beginning at mea-
sure 134, the span of music from measure 67 to this point exceeds the normal
formal remit of a closing section, comprising a tripartite design, which in itself
projects the rhetoric of a main theme–subordinate theme–closing section suc-
cession (A theme in mm. 67–102; B theme in mm. 103–11; C theme in mm.
112–34.1). And because the B-theme candidate here recovers that attempted
at measure 36, the entire exposition subdivides into two large phases (mm.
4–66 and 67–134; “rotations” in Hepokoski’s and Darcy’s terms or “parallel
forms” in Linda Correll Roesner’s usage) founded on the same basic ordering
of material.35 Second, Schumann more or less completely abandons the caden-
tial markers of a classical exposition after the end of the A theme. In particular,
there is no perfect-cadential confirmation of III, but rather a series of authen-
tic cadences, which persist even as we move into the closing tutti.
In sum, Schumann’s decision to start the development with an interpolated
slow movement responds to the fact that measures 1–155 have the rhetorical,
if not the tonal, characteristics of a double rather than a single exposition,
and, as such, project elements of a whole-movement form (the sonata without
development) housed within one broad expositional span. Put another way,
we can argue that the urge to collapse the cycle into the form develops out
of the latter’s tendency to expand interthematically: the super-abundance of
chained interthematic groupings inflates the large-scale functions from within
to the point where maintenance of a clearly delineated movement cycle seems
redundant. (Schumann’s decision to compose a relatively slight intermezzo as
the slow movement of opus 54, which is elided with the Finale, displays an

ongoing consciousness of the cyclical weakening that the first movement’s two-
dimensionality threatens.)
The urge toward conflation of form and cycle is fully realized in Liszt’s
Concerto no. 2, which comprehensively conflates sonata form and sym-
phonic movement cycle in the manner of his near-contemporaneous
B-minor Sonata. Again, the key issue for the present purposes is the rela-
tionship between the work’s “two-dimensionality” and the promiscuity of
functions at lower levels. On the broadest scale, the outline of a sonata
form is apparent (exposition mm. 1–289; development mm. 290–420; reca-
pitulation mm. 421–512; coda mm. 513–89); simultaneously, the markers
of a four-movement cycle are discernible, if not precisely aligned with the
sonata scheme (first movement mm. 1–147, Scherzo mm. 148–212, slow
movement mm. 213–89, Finale, mm. 421–589).
The veneer of clarity that this reading projects, however, rapidly dissolves
under close analysis. The exposition is especially problematic. The concerto
begins with a seventy-two measure section resembling a slow but fairly clear-
cut A-theme presentation (orchestral antecedent, mm. 1–12; solo/tutti con-
sequent, mm. 13–28.1, concluding with a perfect cadence), followed by a
two-part continuation (mm. 28.2–46 and 47–72), which modulates and increas-
ingly resembles a transition. This in turn prepares the martial D-minor subordi-
nate theme commencing at measure 73, which is broadly sentential (statement
mm. 73–82.1; response mm. 82–90.1; continuation mm. 90–96; half cadence
mm. 97–98.1). A basic ambiguity arises in the relationship between these two
themes: although presentational of primary material that is unambiguously
reinforced at the start of the recapitulation, the opening has the character of
a slow introduction; and although the D-minor march is in one sense an inter-
thematic response to the opening, its character identifies more closely with the
topical features common to concerted first themes. In brief, the trajectories “A
theme–B theme” and “introduction–A theme” are simultaneously in play and
establish a dual functional perspective. This is reinforced by tonality: the open-
ing establishes the global tonic A major, which comes to sound like a dominant
upbeat in relation to the D-minor march.
Subsequent events compound these ambiguities, taking up the parallax of tran-
sition and presentation already encountered in Schumann’s opus 54. The section
initiated at measure 98 appears transitional, leading as it does into the new 86 music
entering at measure 116; however, this music also quickly becomes prefatory to
the tutti entering at measure 148, which introduces an important new idea in the
key of B-flat minor, and which initiates the Scherzo at the cyclical level. And this
again yields to a transitional process with the piano entry from measure 181, which
moves toward the D-flat-major nocturne variant of the first theme commencing at
measure 224, already identified as a putative slow movement.
There are several ways of reading all of this. Mindful of the work’s concerted
genealogy, measure 116 could be regarded as initiating a closing section, an

116 julian horton
impression reinforced with the tutti entry at measure 148, which has the char-
acter of a ritornello closing a solo exposition; this yields the interpretation
given in table 3.7a. If measures 1–72 are introductory, then this reading col-
lapses, since no subordinate theme is present between A and the ritornello.
In order to preserve a sonata perspective, we are obliged instead to take the
view outlined in table 3.7b and regard measures 72–98.1 as A, measures 98–223
as a multipart transition and measures 224–89 as a B group, which however
also closes the exposition. A further alternative is to posit two expositions, the
second of which is the rhetorical mirror of the first (see table 3.7c). Thus the
thematic contrast adagio–march underpinned by the subdominant motion
I–iv in A proceeds to the thematic contrast scherzo–nocturne articulating a
localized i–III relation in B-flat minor. The structure is framed by the opening
theme, which by measure 224 has exchanged its primary for a subordinate-
theme function. By these terms, measures 1–147 function as a first movement,
while the music between measures 148 and 289 is both expositional at the level
of the sonata form and functions as two movements within the cyclical scheme.
The insufficiency of these readings taken in isolation is not only a product of
their codependence but also arises from the sense of thematic and formal “becom-
ing” that the music projects. Thus the initial material appears to be in transit to
a thematic presentation, even as it projects a clear syntactic identity. After the
D-minor march’s presentation, the sense of functional “becoming” intensifies: at
each stage before the recovery of the first theme at measure 224, the introduc-
tion of new material is rendered provisional by the music’s tendency to become
retrospectively transitional. The design becomes less ambiguous from the start of
the development. Measures 290–394 are entirely developmental of earlier mate-
rial, focusing on the combination and reworking of the march, the tutti, and the
solo material that follows it in measures 181–206. Measures 395–420 are altogether
retransitional, being poised over a dominant pedal, and rehearse the material of
Table 3.7a. Liszt, Piano Concerto no. 2, exposition, reading 1
Mm. 1–28 29–72 73–115 116–47 148–81 182–212
Cycle: first movement scherzo
Large-scale exposition
function 1:
Large-scale S1 R1 S2
function 2:
Interthematic A TR B C tutti solo

function 1:
Tonality: I → iv → (♭IV) → ♭ii →♯iii

Table 3.7b. Liszt, Piano Concerto no. 2, exposition, reading 2

Mm. 1–72 73–981 98–115 116–47 148–212 213–23 224–89
Cycle: first scherzo slow movement

movement
Large-scale function 1: introduction exposition
Large-scale function 2: S1 R1 S2
Interthematic function 1: A TR B
Tonality: I → iv → (♭IV) → ♭ii→♯iii ♭IV
Table 3.7c. Liszt, Piano Concerto no. 2, exposition, reading 3
Mm. 1–28 29–72 73–981 98–115 116–47 148–81 182–212 213–23 224–89
Cycle: first movement scherzo slow movement
Large-scale function 1: exposition 1 exposition 2
Large-scale function 2: S1 R1 S2
Interthematic A TR B C A TR B
function 1:
Tonality: I → iv → (♭IV)→ ♭ii→ ♯iii → ♭IV
9/30/2015 7:51:47 PM
118 julian horton
measures 115–47. And measures 421–512 serve as a condensed recapitulation,

which eliminates the ambiguities surrounding the Adagio theme in the expo-
sition (as introduction, A theme, and B theme) by making it the substance
of a monothematic reprise (A theme mm. 421–39.1; transition mm. 439–61;
A as tonic subordinate theme mm. 462–512). Finally, measures 513–89 func-
tion unambiguously as a coda. Altogether (and to borrow Vande Moortele’s
terminology), there is identification between the Finale of the cycle and the
recapitulation of the form, but the development is “exocyclic,” existing only at
the large-scale formal level.36
The paradox of a conflation of functions arising at one formal level being
a consequence of their proliferation at another is sharply articulated here.
The expository material in this work appears to burst its large-scale formal
constraints: where classical composers would content themselves with the pre-
sentation of a twofold material contrast related by a single transition, Liszt’s
presentations, even more so than Schumann’s, are manifold and collapse con-
tinuously into transitions.
Conclusions
This study’s aim is by no means to offer a comprehensive syntax for this rep-
ertoire, let alone for early nineteenth-century concerted forms in general.
What I have attempted is rather a sketch of this music’s syntactic character, an
explanation of how this syntax conditions changes in large-scale design, and
a (brief) consideration of its close interaction with early nineteenth-century
pianistic topics and rhetoric.
Such a project uncovers a basic lacuna in the literature on nineteenth-century
sonata forms in particular, and on nineteenth-century forms in general: although
much has been written on this repertoire, few authors start from the premise that
a model of syntax needs to be established before formal analysis can proceed.37
In short, the project on which Caplin embarks for Viennese classicism currently
has no nineteenth-century counterpart. More commonly, postclassical sonatas
are approached with the underlying conviction that they stand in relation to a
reified classical scheme, which thereby furnishes (by presence or absence) the
theoretical terms of reference. Such an approach is useful for revealing how early
nineteenth-century forms depart from (“deform”) an extrapolated repertoire of
classical norms, but it also has manifest disadvantages. First, it forces all postclas-
sical forms into dialogue with a model derived from a relatively restricted (and
usually canonically predetermined) sample of works. Second, such models are
often applied with little thought for generic specificity. Any theory of postclassi-
cal symphonic sonata form would be remiss if it ignored the sovereign example
of Beethoven, even if it recognized a plurality of precedents beyond Beethoven’s

oeuvre. But a theory of postclassical concerto first-movement form cannot be so

monocular in its search for precedents; the installation of Mozart or Beethoven
as dominant precursors is supported neither by the growth of formal practice in
this time nor by the reception history of Viennese examples.38
The syntactic-historical approach I advocate therefore unearths a verdant
territory for Formenlehre. Nineteenth-century instrumental forms yield a wealth
of syntactic formations, which demand explanation in terms of both the formal
functions they perform and the role they play in conveying topical discourse.
As an arbiter of function, this syntax acts, as it does in earlier music, as a means
of mediating form and system: the distribution of functions responds to the
problem of how to articulate tonal relations as relations between material
grouping structures. Yet because tonality as it develops toward the mid-century
embraces a chromaticism operating at a greater structural depth than in its
high-classical counterpart, the set of relations that nineteenth-century syntax
mediates is concomitantly expanded. Moreover, since the development of this
system is tied to a style that makes expressive capital out of cadential deferral,
the syntax evolves to convey all this; the result is a revision or expansion of clas-
sical categories, such that their rhetoric is preserved while their clear affiliation
of cadence and end function is undermined.
Substantial engagement with nineteenth-century sonata forms from this
perspective would require the taxonomy of functions and types in all generic
contexts, together with assessment of how their disposition conditions form
in each case. Given sufficient space, I would prosecute a parallel historical
agenda, the motivating conviction of which is that we should reconstruct as far
as possible the circumstances in reception history that condition the develop-
ment of a form in any given genre before we attempt any higher-level observa-
tions about norm and deviation. In other words, I would urge that suppositions
about the relationship between a repertoire’s formal practices and their pre-
cursor models be severely constrained by considerations of generic-historical
specificity. Too often, Formenlehre is happy to reinforce the practice, initiated
in the nineteenth century, of deriving norms from a canon of works selected
on aesthetic grounds or to reinforce cultural-political agendas. The orienta-
tion of A. B. Marx’s and Carl Czerny’s notions of concerto first-movement form
toward Beethoven, for example, served the twin convictions that Beethoven’s
music is aesthetically superior and historically central.39 As a basis for com-
positional pedagogy, this attitude facilitated the carrying forward of such
aesthetic-political convictions in the production of new music. But the true cir-
cumstances of the development of classical forms into the nineteenth century
are plural, geographically disparate, and generically variable. So long as we
allow this nineteenth-century canonical mentality to condition, from afar, the
way we ground theoretical models, this plurality will remain beyond the grasp
of formal theory.

120 julian horton
Notes
1. See William E. Caplin, Classical Form: A Theory of Formal Functions for the Instrumental
2. See Mark Evan Bonds, Wordless Rhetoric: Musical Form and the Metaphor of Oration
(Cambridge, MA: Harvard University Press, 1991), esp. 13–52.
3. See Carl Dahlhaus, Nineteenth-Century Music, trans. J. Bradford Robinson (Berkeley:
University of California Press, 1989), 134–42; and Jim Samson, Virtuosity and the
Musical Work: The Transcendental Studies of Liszt (Cambridge: Cambridge University
Press, 2003), 134.
4. Stephan D. Lindeman, Structural Novelty and Tradition in the Early Romantic Piano
Concerto (Hillsdale, NY: Pendragon Press, 1999), 25–26.
5. See James Hepokoski and Warren Darcy, Elements of Sonata Theory: Norms, Types, and
Deformations in the Late-Eighteenth-Century Sonata (New York: Oxford University Press,
2006), 434–35.
6. For an account of Dussek’s career, see Howard Allen Craw, “A Biography and
Thematic Catalogue of the Works of J. L. Dussek” (PhD diss., University of Southern
California, 1964).
7. On the early reception of Mozart’s piano concerti, see Claudia Macdonald,
“Mozart’s Piano Concertos and the Romantic Generation,” in Historical Musicology:
Sources, Methods, Interpretations, ed. Stephen A. Crist and Roberta Montemorra
Marvin (Rochester, NY: University of Rochester Press, 2004), 302–29.
8. I have dealt with this matter as it affects the London repertoire in some detail in
Julian Horton, “John Field and the Alternative History of Concerto First-movement
Form,” Music and Letters 92 (2011): 43–83.
9. See Caplin, Classical Form, 4.
10. Ibid., 4.
11. See William E. Caplin, “What Are Formal Functions?” in Caplin, James Hepokoski,
and James Webster, Musical Form, Forms & Formenlehre: Three Methodological Reflections,
ed. Pieter Bergé (Leuven: Leuven University Press, 2009), 21–39.
13. This claim raises the matter of whether motives should be regarded as formally gen-
erative in classical syntax, an issue that fundamentally differentiates Caplin’s and
Schoenberg’s positions. See Caplin, Classical Form, 4, responding to Schoenberg,
Fundamentals of Musical Composition (London: Faber, 1967), which theorizes the
motive as a basis for understanding phrase types.
14. In this respect, the notion of large-scale presentation comes close to what
Hepokoski and Darcy intend by rotation: the exposition “presents” a body of mate-
rial, which later large-scale functions might adopt as a template and vary. See their
Elements of Sonata Theory, 23 and 611–14.
15. This claim is of course open to challenge from a Schenkerian perspective, which
would regard all local cadential events as prolongational of a deep-structural
cadence. The question of how the Schenkerian concept of hierarchy differs from
the functional conception adopted here are unfortunately too complex to be
addressed here. Incidentally, I think Riemann, too, might describe an entire sec-
ond-theme group as cadential.
16. Caplin, “What Are Formal Functions?” 32.
17. Steven Vande Moortele, Two-Dimensional Sonata Form: Form and Cycle in Single-
movement Instrumental Works by Liszt, Strauss, Schoenberg, and Zemlinsky (Leuven:
Leuven University Press, 2009), 24–26.

18. On the distinction between tight-knit and loose designs, see Caplin, Classical Form,
17 and also 97–99.
19. See ibid., 65–69.
20. On such “irregular” sentential formations lacking a continuation in Haydn’s music,
see Matthew Riley, “Haydn’s Missing Middles,” Music Analysis 30 (2011): 37–57.
21. Hummel’s music represents one of the most tangible links between this reper-
toire and Mozart. Hummel studied with Mozart in Vienna between 1786 and 1788,
and Mozart’s K. 466 was in Hummel’s repertoire as a pianist from the early 1790s
onward. The disparities between Hummel’s syntax and Mozart’s are therefore all
the more striking.
22. On the solo-entry first theme in Mozart’s concerti, see Joel Galand, “The Large-
Scale Formal Role of the Solo Entry Theme in the Eighteenth-Century Concerto,”
Journal of Music Theory 44 (2000): 381–450; and also Hepokoski and Darcy, Elements
of Sonata Theory, 498–512.
23. See in particular Janet Schmalfeldt, “Form as the Process of Becoming: The Beethoven-
Hegelian Tradition and the Tempest Sonata,” Beethoven Forum 4 (1995): 37–71; and more
recently In the Process of Becoming: Analytical and Philosophical Perspectives on Form in Early
Nineteenth-Century Music (New York: Oxford University Press, 2011).
24. On the subordinate theme and the essential expositional closure, see Hepokoski
and Darcy, Elements of Sonata Theory, 117–49.
25. See William E. Caplin, “Beethoven’s Tempest Exposition: A Springboard for Form-
functional Considerations,” in Beethoven’s Tempest Sonata: Perspectives of Analysis
and Performance, ed. Pieter Bergé, Jeroen D’Hoe, and William E. Caplin (Leuven:
Peeters, 2009), 87–126, esp. 91–93.
26. James Hepokoski might regard this as a “zero-module”; see, for instance, Hepokoski,
“Approaching the First Movement of Beethoven’s Tempest Sonata through Sonata
Theory,” in Bergé et al. (ed.), Beethoven’s Tempest Sonata, 181–212, esp. 187.
27. See William Rothstein, Phrase Rhythm in Tonal Music (New York: Schirmer, 1989),
116; Hepokoski and Darcy, Elements of Sonata Theory, 120; Caplin, Classical Form, 122,
which states that “in movements containing a single subordinate theme, a postca-
dential closing section almost always follows the perfect authentic cadence ending
that theme. In movements containing a subordinate-theme group, the closing sec-
tion follows the cadence ending the last theme of the group.”
28. On these matters, see David Rowland, “Pianos and Pianists, c. 1770–c. 1825,” in
The Cambridge Companion to the Piano, ed. David Rowland (Cambridge: Cambridge
University Press, 1998), 22–39; on the relationship between pedal technology
and the development of the nocturne style, see also Rowland, “The Nocturne:
Development of a New Style,” in The Cambridge Companion to Chopin, ed. Jim Samson
(Cambridge: Cambridge University Press, 1992), 32–49, esp. 36–42.
29. For overviews of first-movement form in the virtuoso concerto, see John Rink,
Chopin: The Piano Concertos (Cambridge: Cambridge University Press, 1997),
2–6; and Claudia Macdonald, Robert Schumann and the Piano Concerto (New York:
Routledge, 2005), 19–35.
30. See William S. Newman, The Sonata since Beethoven (Chapel Hill: University of
North Carolina Press, 1969), 134; Vande Moortele, Two-Dimensional Sonata Form,
35–58 and “Beyond Sonata Deformation: Liszt’s Symphonic Poem Tasso and the
Concept of Two-Dimensional Sonata Form,” Current Musicology 86 (2008): 41–62.
On Liszt’s sonata forms, see also Richard Kaplan, “Sonata Form in the Orchestral
Works of Liszt: The Revolutionary Reconsidered,” 19th-Century Music 8 (1984):
142–52.

122 julian horton
31. For a more detailed consideration of Field’s Concerto no. 7 and its relationship to
Schumann’s opus 54, see Horton, “Field and the Alternative History.”
32. See Schumann, “Pianoforte. Concerte. John Field, 7tes Concert mit Begl. des
Orch . . . I. Moscheles, 5tes Concert mit Orchest.,” Neue Zeitschrift für Musik 4/29
(1836): 122–24, trans. by Henry Pleasants as “Ignaz Moscheles: Piano Concertos
nos. 5 and 6,” in Schumann on Music: A Selection from the Writings (New York: Dover,
1965), 107–9; quote from p. 108).
33. Claudia Macdonald, “‘Mit einer eignen außerordentlichen Composition’: The
Genesis of Schumann’s Phantasie in A minor,” Journal of Musicology 13 (1995):
240–59, esp. 253–54). On this matter, see also August Gerstmeier, Robert Schumann.
Klavierkonzert a-moll, Op. 54 (Munich: Fink, 1986), 25–26.
34. As I have argued elsewhere, Schumann surely borrows directly from Field in
this respect. The relationship is borne out not only by the formal similarities at
this point but also by Schumann’s reception of Field’s concerto: see Schumann,
“Pianoforte. Concerte. John Field, 7tes Concert mit Begl. des Orch . . . I.” trans. by
Henry Pleasants as “John Field: Piano Concerto no. 7” in Schumann on Music, 106;
and Horton, “Field and the Alternative History.”
35. See Hepokoski and Darcy, Elements of Sonata Theory, 611–14; and Linda Correll
Roesner, “Schumann’s Parallel Forms,” 19th-Century Music 14 (1991): 265–78.
36. Vande Moortele, Two-Dimensional Sonata Form, 27.
37. Schmalfeldt’s In the Process of Becoming is a recent honorable exception; for
other, less wide-ranging addresses on the subject, see Matthew BaileyShea, “The
Wagnerian Satz: The Rhetoric of the Sentence in Wagner’s Post-Lohengrin Operas”
(PhD diss., Yale University, 2003); and James Hepokoski, “Framing Till Eulenspiegel,”
19th-Century Music 30 (2006): 4–43.
38. The literature on the postclassical concerted repertoire is more substantial for some
centers than for others. See, for example, Thomas B. Milligan, The Concerto and
London’s Musical Culture in the Late Eighteenth Century (Ann Arbor, MI: UMI Research
Press, 1983); Therese Ellsworth, “The Piano Concerto in London Concert Life
between 1801 and 1850” (PhD diss., University of Cincinnati, 1991); Macdonald,
Robert Schumann and the Piano Concerto; Jeffrey Cooper, The Rise of Instrumental Music
and Concert Series in Paris, 1828–71 (Ann Arbor, MI: UMI Research Press, 1983); and
Charles David Lehrer, “The Nineteenth-Century Parisian Concerto” (PhD diss.,
University of California, Los Angeles, 1990). All these studies, however, note the
striking dearth of Mozart in the repertoire before 1820 at the earliest.
39. A. B. Marx’s description of concerto first-movement form is found in Die Lehre von
der musikalischen Komposition (Leipzig: Breitkopf und Härtel, 1847), 4:437–40, esp.
439; Czerny’s appears in School of Practical Composition, trans. John Bishop (London:
Cocks, 1848), 1:159–64. Marx cites Mozart, Beethoven, and after them Moscheles,
Mendelssohn, Chopin, and Hummel; Czerny takes Beethoven’s opus 37 as an
exemplum, but additionally acknowledges a range of virtuoso composers.

Chapter Four
Saint-Saëns’s Cyclic Forms

Andrew Deruchie
Music critics and historians have long recognized the importance of cyclic form
to late nineteenth- and early twentieth-century French music. Most composers
of instrumental music active in France during this period adopted the prac-
tice of effecting large-scale relationships in their symphonies, concerti, sona-
tas, chamber compositions, and other works by reintroducing or developing
themes or motives from earlier movements in later ones. Nevertheless, writers
addressing this repertoire tend to associate such procedures above all with the
music of César Franck from the late 1870s until his death in 1890 and with that
of some of his students, including Guy Ropartz, Ernest Chausson, Guillaume
Lekeu, and above all Vincent d’Indy. Benedict Taylor’s recent monograph on
cyclic form in the romantic era, one of the most sophisticated studies on the
topic to date, points to Franck as the late-century apogee; the same composer
revealingly stands as the only French member of the cyclic-form pantheon as
defined by New Grove and the Oxford Dictionary of Music.1 To be sure, Franck and
d’Indy warrant emphasis. Both cultivated cyclicism with remarkable complex-
ity and finesse, and d’Indy, purposefully taking up his mentor’s legacy, champi-
oned “la forme cyclique”—he numbered among the earliest writers to use the
term—with particular vigor, employing it in almost all of his multimovement
works and promoting it to the status of a historically determined canon in his
pedagogical and theoretical writings.2 Moreover, appearing to legitimize a priv-
ileged status for these composers, contemporary critical discourses—a topic we
shall revisit—often considered cyclic form a hallmark of Franckiste aesthetics
and even a proprietary compositional technique.
Camille Saint-Saëns, on the other hand, has attracted relatively little atten-
tion as a practitioner of cyclicism. One of the nation’s leading composers of
instrumental music in the late nineteenth century and a continuing, if fading,
presence in the first two decades of the twentieth, Saint-Saëns in fact employed
the procedures enumerated above extensively and did so from the beginning of
his long career. Early examples include the A-Minor Piano Quintet (1855), the
first two violin concerti (1858 and 1859), and the Second Symphony (1859).
He would follow these up with a long string of other cyclic works in the next

124 andrew der uch ie
three decades, including the Second, Third, and Fourth Piano Concerti (1868,
1869, and 1875), the First Violin Sonata (1885), and the Third (“Organ”)
Symphony (1886), all among his most respected and frequently performed
compositions. Contrary to d’Indy, Saint-Saëns did not consider cyclic form an
aesthetic sine qua non. A self-declared eclectic, he resisted tying his reputation
to specific compositional procedures or historical pedigrees. Plenty of his com-
positions make no use of intermovement thematic or motivic connections, and
he made no special pleas for these in his writings. Nevertheless, such proce-
dures clearly constituted important components of his technical toolbox, and
he plainly valued the kinds of large-scale relationships and formal effects they
could bestow upon his compositions. Writers who attach special importance to
“firsts” have often observed that Franck’s Trio in F-sharp Minor of 1841 stands
as the earliest foray into cyclic form by any French composer active in the late
nineteenth century, but this work would remain for decades the lone example
in its composer’s oeuvre. On the other hand, one could make a compelling
case, on the basis of the above-named compositions alone, that Saint-Saëns had
emerged as a major proponent of cyclic form long before Franck made it one
of his compositional signatures, and before d’Indy’s career even began.
To this end, this chapter seeks to draw attention to the cyclic procedures
Saint-Saëns employed and to the particular quality of formal integration he
effected with them. It offers panoptic analyses of five representative cyclic
works—the First Violin Concerto (1859), the First Cello Concerto (1872), the
Fourth Piano Concerto (1875), the First Violin Sonata (1885), and the Third
Symphony (1886)—which provide varying perspectives on the composer’s
craft and offer an account of how his approach evolved. What emerges, I hope,
will not only contribute to a fuller historical account of cyclic form in late nine-
teenth-century French music, but will also help define Saint-Saëns’s position
among the aesthetic factions operating during that period in the nation’s musi-
cal culture. Recent work by Michael Puri, Marianne Wheeldon, Brian Hart,
Serge Gut, and others has cumulatively made plain what the fuzzy definition
offered in this essay’s opening paragraphs implies: far from a singular practice,
cyclic form stands as an umbrella category that admits an array of composi-
tional procedures.3 A theme from one movement may return in another in
whole or in part, literally or transformed; a motive from one theme may serve
as a germ that produces other, distinct themes; cyclic form may involve a single
thematic recurrence or many, and in principle it may involve any theme return-
ing at any point in any subsequent movement; a theme may return smoothly or
with a violently disruptive effect; and, no less than in a single-movement form,
the return or development of a theme in a multimovement context may estab-
lish any manner of ongoing or overarching formal process. Thus, if on one
level cyclicism can be viewed as a predilection Saint-Saëns shared with such
aesthetically disparate composers as d’Indy, Fauré, Edouard Lalo, and Debussy
(to say nothing of Bruckner, Brahms, Tchaikovsky, Dvořák, Elgar, Mahler, and

saint-saëns’s cyclic forms 125
Schoenberg), it can also serve as an analytical touchstone to underscore dif-

ference, not only in matters of musical form in the abstract but also in the
spheres of cultural values, ideology, and so on. Although neither a compara-
tive account nor a hermeneutic one is possible here, the pages that follow may
serve as groundwork for future research focusing on these issues.
❧ ❧ ❧
Important studies by Taylor (on Mendelssohn) and Puri (on Ravel) have set out
new and sophisticated theoretical approaches emphasizing the phenomenologi-
cal qualities cyclic procedures may engender. By recalling material from a pre-
vious movement, cyclic forms possess a special capacity to juxtapose a musical
past with a musical present. The delineation and intermingling of a “then” and a
“now” may produce complex modes of temporality, not generally available to sin-
gle-movement forms, and thereby aestheticize memory, a keystone of romantic
and decadent subjectivity and of historical consciousness.4 These methods seem
promisingly applicable to a wider range of repertoire. Nonetheless, what follows
will largely dispense with their specialized theoretical formulations and instead
frame questions of formal organization (and implicitly of musical time) in the
more familiar vocabulary and conceptual apparatus of the Formenlehre tradition
as William E. Caplin has revived it and especially as Steven Vande Moortele has
recently extended it. For contrary to the practice of Mendelssohn and Ravel,
where thematic-motivic connections bestow relationships on otherwise discrete
and more or less autonomous movements, it seems more fruitful to conceptual-
ize Saint-Saëns’s approach to cyclicism as one that embeds the formal functions
of a single-movement sonata form—main theme, exposition, coda, and so on—
among the sundry movements of his cycles.
Although Saint-Saëns could have drawn on well-known precedents by
Schumann and Schubert, and on more obscure ones by Moscheles and the
Belgian violinist-composer Henri Vieuxtemps, the music of Liszt appears to
have served as his most immediate source for this strategy. The two compos-
ers first met around 1852, and a warm and enduring friendship developed as
did a professionally supportive relationship: Liszt arranged the 1877 premiere
of Samson et Dalila in Weimar, and Saint-Saëns tirelessly championed his col-
league’s music in France. Liszt’s oeuvre, as is well known, includes a number of
works that group multiple movements into a single continuous span, the Piano
Sonata in B Minor and the symphonic poems Tasso, Die Ideale, and Les Préludes
among them. As Vande Moortele has demonstrated, these “two-dimensional”
forms interleave a cycle of four movements (that is, Allegro–slow movement–
Scherzo–Finale, or some variant thereof) with a sonata form, dissolving the
normative, hierarchical relationship between form and cycle and placing both
on the same plane. In this scheme, sections of the overarching sonata form,
or portions of contiguous sections, may coincide with movements in the cycle

(the exposition, or perhaps the exposition and the beginning of the develop-
ment, may express the first movement; the recapitulation and/or the coda
may do double duty as the cycle’s Finale; and so on). Movements may also be
interpolated between or within sections of the overarching form, temporar-
ily deactivating it, and portions of the form may play no role in the embed-
ded cycle. Such works therefore demand that listeners attentively following the
progress of the form continuously ask themselves whether what they are hear-
ing unfolds in the dimension of the overarching sonata scheme, in that of the
multimovement cycle, or in both.5
Some of Saint-Saëns’s cyclic works, especially those dating from the
early part of his career, adhere closely to this pattern. Both the First Violin
Concerto and the First Cello Concerto interweave three-movement cycles
with an overarching sonata form. Table 4.1 gives an overview of the former.
The top portion of the diagram shows the overarching sonata form; the bot-
tom portion shows how this form interacts with the cycle. Like all two-dimen-
sional sonata forms, this one begins as a normative, single-dimension sonata
form. The exposition lays out two main themes; a transition, a subordinate
theme, and a development section follow. The A-major return of main theme
1 at measure 109 suggests the onset of the recapitulation. This function, how-
ever, immediately fizzles: the German augmented-sixth chord built on B♭ in
measures 102 and 104 instantly recolors the tonic as V of IV, undercutting
the effect of tonal return, and a cadenza dissipates main theme 1 at measure
117. If a cadenza at this point in the first movement of a nineteenth-century
concerto would seem unremarkable, the formal unit that follows at measure
118 is difficult to reconcile with a sonata form.6 A new theme in D major
(the tonality implied by the augmented sixths in the abortive recapitulation)
emerges at measure 128, preceded by a little introduction. This theme initi-
ates a miniature ternary form, comprising also a tonally unstable, develop-
ment-like middle section (which references the little introduction) and a
D-major reprise of the theme. Bracketed off from what precedes it not only
by the new key but also—and especially—by changes in tempo and meter,
this whole section (mm. 118–72), with its fresh thematic material and rela-
tively elaborate form, registers as a more or less autonomous slow movement
embedded within the overarching sonata form and not as a continuation of
that form. In Vande Moortele’s theoretical framework, the dimension of the
overarching form here goes dormant, and listeners become aware that a new
dimension—that of the cycle—has emerged.
When the embedded slow movement concludes, the original tempo and
meter return, and the overarching sonata form resumes with the arrival of the
recapitulation. The subordinate theme appears first, off tonic in D minor. Now
it is the turn of the cycle to go dormant, as this formal unit plays no role in
that dimension. The recapitulation continues with the restatement of the two
main themes in A major. The return of the tonic and the energetic, fanfare-like

Table 4.1. First Violin Concerto: two-dimensional sonata form
Over- Exposition Development Recap.? Recapitulation

arching
form Main Main Sub. Main th. Sub. Main Trans- Coda
theme theme theme 1 and theme themes ition
1 2 cadenza 1 and 2
A+ A+ C#– A = V/D D+ V D– A+ A+ A+
Cycle First movement Slow movement Finale
Intro A B A
D+ D+
9/30/2015 7:51:48 PM
quality of the first inject the character of a Finale (which listeners will expect
given that the concerto has included a slow movement) into the recapitulation,
reactivating the dimension of the cycle, such that both dimensions are now
present. Like the embedded slow movement, the Finale expresses its own com-
pact ternary design, for which main themes 1 and 2 serve as the main theme
group. As in the exposition of the overarching sonata form, the open-ended
main theme 2 shades into the transition, which in the context of the Finale’s
local form serves as a developmental middle section; an emphatic retransition
(not shown in table 4.1) follows at measure 250. Here Saint-Saëns faced a com-
positional challenge: how to round off the Finale’s local form? Recalling one
or both of the main themes would risk redundancy, since these same materi-
als had already served the formal function of recapitulation (in the context of
the overarching sonata form) some fifty measures earlier. He solved this prob-
lem by providing formal closure to the Finale with the coda of the overarching
design. The latter section begins with a fresh theme, based on the continuation
phrase—but dispensing with the characteristic, opening two-measure idea—of
main theme 2, now adorned with a waltz-like accompaniment (ex. 4.1). This
theme retains enough of main theme 2 to satisfy the Finale’s need for reca-
pitulation, yet differs sufficiently, notably in its deft recasting in an initiating
role of material that had previously served a medial (continuation) function,
to eschew a sense of superfluous repetition.
The First Cello Concerto (1872) pursues a broadly similar two-dimensional
design, likewise projecting a three-movement cycle (featuring a minuet instead
Example 4.1a. Saint-Saëns, First Violin Concerto, main theme 2, reh. A to A+6

Example 4.1b. First Violin Concerto, coda of Finale, reh. N to N+4
of a slow movement) and a sonata form onto one another (see table 4.2). It
does, however, evince some important differences. For one, the cycle’s three
movements assume more elaborate internal forms, which in scope and com-
plexity approach what one might expect of a concerto comprising separate
and fundamentally independent movements. Paralleling the earlier work, the
opening allegro coincides with the exposition and development sections of
the overarching form. Three themes serve as its basis: a main theme firmly
in the tonic A minor, a subordinate theme sitting on the dominant of F, and
a vigorously cadential closing theme (see exx. 4.2, 4.3, and 4.4a below). The
development of the overarching form extraordinarily juxtaposes extended,
at-pitch quotations of the main and subordinate themes. Preserving even his
scoring, Saint-Saëns simply transplants A+7 to A+15 (from the main theme)
and B+16 to B+24 (from the subordinate theme) to the development, stitch-
ing these together to produce measures E–8 to E+12, altering only the bass of
the subordinate-theme citation to now emphasize the dominant of B-flat. This
strikingly literal recall of exposition material functions as an off-tonic recapitu-
lation of the cycle’s first movement, bringing a measure of conclusion to the
local sonata form while leaving the overarching form’s tonal process incom-
plete and deferring its resolution.

Table 4.2. First Cello Concerto: two-dimensional sonata form
Over- Exposition Development Retrans- Recap. Coda

arching ition
form
Main Subor- Closing Main Main Main
theme dinate theme and sub. theme theme,
theme themes closing
theme
A– V/F F+ V/B♭ B♭+ V/A A– F+ A– A– A+
Cycle Exposition Develop- Recap. A A B A′ Exposition Develop- Recap.

ment ment
First movement Minuet Finale
9/30/2015 7:51:49 PM
Table 4.3. First Cello Concerto: form of the exposition
m. 1 Reh. A B+16 C–16 C+20
Main theme Transition Sub. theme Transition Closing

resumes theme
Antecedent Conse- Presentation Continuation,

quent? only Cadential
A– HC HC V46 F+ 6
4 F+:PAC
As in the Violin Concerto’s abortive recapitulation, Saint-Saëns employs

this local recapitulation as a tonal pivot to the interpolated middle movement,
which follows in B-flat major, the Neapolitan, an unusual relationship even by
mid-century standards, though one the composer employed often.7 The ear-
lier Violin Concerto’s slow movement is, at fifty-five measures, concise, and it
assumes a highly compressed form: the initial “A” section consists simply of a
twelve-measure sentence arriving at a half cadence (HC); the contrasting mid-
dle prolongs the dominant for eight measures; and the reprise expands the
continuation phrase of “A.” The Cello Concerto’s middle movement, on the
other hand, is a full dance form, comprising a repeat of the opening minuet—
itself a rounded binary design—with obbligato solo cello, a waltz-like trio in
G minor, and a written-out da capo with the solo cello. With some rejigging
of its beginning and conclusion, this minuet could stand on its own in a con-
ventional cycle of discrete movements.8 Saint-Saëns similarly fleshed out the
work’s Finale. The earlier concerto’s third movement expresses a simple, com-
pact ternary form, but in the later work the Finale pursues a more complex
sonata form featuring two contrasting themes (bridged by a lengthy transi-
tion), a brief development (mostly over a dominant pedal), and a recapitula-
tion of the main theme, though not the subordinate theme.
The Cello Concerto also differs from the First Violin Concerto in that the
two dimensions interact in more complex ways. As in the earlier work, the mid-
dle movement stands as an interpolation into the form, playing no role in it,
and it is likewise followed by a segment, in this case the retransition, which
functions exclusively in the dimension of the overarching form. But where
the Violin Concerto’s Finale coincides with most of the recapitulation and the
coda of the overarching form, in the Cello Concerto these all occur succes-
sively: Saint-Saëns interpolates the Finale, deactivating the dimension of the
cycle for its duration, and the recapitulation and coda of the overarching form
play no role in the dimension of the cycle.
Saint-Saëns also lays for this concerto’s large-scale design a type of ground-
work not present in the earlier opus 20. As noted above, two-dimensional
forms confound the hierarchical relationship that normally inheres between

a work as a multimovement whole and the large-scale forms (sonata, rondo,

ternary, and so on) of the movements that compose it. This hierarchy can be
extended further downward: full-movement forms fall into sections (exposi-
tion, development, recapitulation, minuet, trio, and so on); sections, to adopt
Caplin’s vocabulary, comprise interthematic functions (main theme, transition,
subordinate theme); interthematic functions comprise sets of intrathematic
functions (antecedent, consequent, presentation, continuation, cadential).9
Saint-Saëns’s A-Minor Cello Concerto destabilizes the normative relationships
between intra- and interthematic functions right from the opening exposition.
Here boundaries between main theme, transition, and subordinate-theme
group are fluid and ambiguous (table 4.3). The main theme (m. 1) initially
appears to express a compound period: a sentence produces a half cadence
(HC) at measure 16, and its basic idea returns at letter A to initiate a conse-
quent. A sudden thickening of the orchestral fabric, a shift to submediant har-
mony, and a crescendo (A+7) suggest the dissolution of consequent function
and the onset of the transition, but this sentence veers back to A minor to arrive
at another HC. Chromaticism and sequential spinning-out of main-theme
material (B+6 to B+11) follow a brief standing on the dominant; on account
of the main theme’s tonal openness and duplicate antecedent functions, these
measures probably register as a developmental middle section, now suggest-
ing that the theme will pursue a ternary form, an expectation reinforced by
the 46 chord built above E at B+12. However, parsimonious voice leading con-
verts this last sonority into V7/F, and the subordinate theme (ex. 4.2), topically
characteristic in its legato lyricism and attenuated rhythmic activity, unexpect-
edly emerges, leaving listeners no choice but to retrospectively interpret B+6
to B+15 as the transition. In its melodic-motivic design, the subordinate theme
assumes a sentential design, though it sounds over a pedal C, which undercuts
the continuation function affected by fragmentation in the melody, entirely
negates any sense of cadence, and gives the entire theme a presentational
shade. When activity in the bass does begin, the theme ceases, and the transi-
tion resumes, picking up on the same 46 chord and with the same main-theme
material as when it broke off; it leads to the closing theme shown in example
4.3, which emphatically secures F major as the secondary key. Like the sub-
ordinate theme, this segment lacks a full complement of formal functions,
eschewing presentation and expressing only continuation and cadence. To
summarize, the subordinate theme occurs, without the main theme having
clearly concluded, embedded in a transition that has yet to properly depart
from the home key; when the transition does definitively arrive at F major,
it yields to a fragmentary closing theme that belatedly supplies formal func-
tions missing from the largely presentational subordinate theme. Put differ-
ently, the hierarchy that would group ensembles of intrathematic functions
into discrete interthematic functions dissolves. By confounding hierarchy at

Example 4.2. First Cello Concerto, reh. B+13 to B+33

Example 4.3. First Cello Concerto: closing theme, reh. C+20 to C+28
this relatively low formal level, the concerto’s exposition prepares its large-
scale design: the dynamic relationship between sonata form and sonata cycle
springs forth from fluid relationships between main theme, transition, and
subordinate-theme group.
In Liszt’s two-dimensional forms, transformations of themes from the over-
arching sonata form typically serve as themes for interpolated movements,
integrating the two formal dimensions. This does not occur in Saint-Saëns’s
First Violin Concerto; in that work, the slow movement’s theme and introduc-
tion are entirely new. The First Cello Concerto, on the other hand, evinces a
degree of motivic continuity between its dimensions: as example 4.4 shows, the
main theme of the overarching form (and of the first movement) emphasizes
neighbor motion between scale degrees 5^ and 6 ^
. This motive returns in the
main theme of the interpolated minuet (ex. 4.4b) and again in the subordi-
nate theme of the Finale (ex. 4.4c). The theme initiating the coda (ex. 4.4d)
additionally quotes a fragment of the Finale’s subordinate theme, which, as
noted above, does not receive a recapitulation in the movement’s local sonata
form. By this omission, Saint-Saëns again sidestepped the pitfall of redundant
closure that might have resulted from the juxtaposition of two full recapitula-
tions (i.e., that of the Finale’s local form and that of the overarching form).
Example 4.4d thus ties up a loose end and reconciles the competing demands
of the two formal dimensions: it offers enough of an echo of example 4.4c to
stand as the missing tonic-key reprise of the Finale’s subordinate theme, while
striking the ear as sufficiently new to mark the beginning of the overarching
form’s coda.
The formal complexity of the Cello Concerto’s three embedded movements
and the work’s more sophisticated interleaving of sonata form and sonata

Example 4.4a. First Cello Concerto: main theme, mm. 1–9
Example 4.4b. First Cello Concerto: minuet theme, reh. F to F+8

Example 4.4c. First Cello Concerto, reh. O to O+7
Example 4.4d. First Concerto, reh. R to R+8

cycle carry two important results. First, these factors make for added bulk.
The Violin Concerto is extraordinarily compact, lasting less than twelve min-
utes in most performances. The Cello Concerto usually clocks in at around
twenty minutes, still brief by contemporary standards but far more substantial
than its predecessor. These developments also recalibrate the balance between
single-movement form and multimovement cycle. Most listeners will hear
the Violin Concerto as a single movement that makes reference to the three
movements of a conventional concerto; for this reason, it is sometimes called a
“Konzertstück.”10 The Cello Concerto, on the other hand, places overarching
form and embedded cycle on the same plane. That is, listeners will perceive, in
equal measure, a single-movement form and a three-movement design.
❧ ❧ ❧
Let us turn now to three of Saint-Saëns’s later cyclic forms, all among his best-
known instrumental compositions. The Fourth Piano Concerto (1875), the
First Violin Sonata (1885), and the Third Symphony (1886) swell in scale rel-
ative to the earlier concerti, assuming dimensions typical of their genres by
late nineteenth-century standards. These works forsake two-dimensional form,
which sees all movements unfold within a single continuous sonata form,
exchanging it for a four-movement design whereby the opening Allegro is
fused to the Adagio, and the Scherzo to the Finale, to make two large, con-
tinuous parts (see table 4.4). Despite this bipartite layout, listeners will rec-
ognize in each work four discrete, full-scale movements, clearly demarcated
by changes in tempo and (usually) meter, much as they will in Schumann’s
Fourth Symphony. Where the First Violin Concerto registers as a single-move-
ment form that references the three-movement concerto cycle, and where the
First Cello Concerto poises cycle and form in equilibrium, in these works the
balance tips decidedly toward the cycle. Nonetheless, they retain the formal
premise of the First Violin and Cello Concerti. That is, in each, Saint-Saëns dis-
tributes the formal functions of a single-movement sonata form across a multi-
movement span.
Table 4.4. Large-scale layout of First Violin Sonata, Fourth Piano Concerto,
and Third Symphony
Part I Part II
Allegro Adagio Scherzo Finale
1 2 3 4

As table 4.5 shows, the Violin Sonata’s first movement remains syntacti-
cally incomplete. The exposition lays out two themes (exx. 4.5a and 4.5b),
the second, in F major, an arresting tune that supposedly inspired the famous
“petite phrase” by the fictional composer Vinteuil in Proust’s À la recherche du
temps perdu, and which was perhaps reverberating in the back of Stravinsky’s
mind when he composed the Finale to the Firebird.11 The development section
includes a fugato (ex. 4.5c) that incorporates elements of both themes, assum-
ing the rhythm of the first and the pitch contour, transposed into the minor, of
the second. In the recapitulation, the main theme and the fugato return in the
tonic, but the subordinate theme appears in E-flat, the Neapolitan. Off-tonic
recapitulations, of course, occur commonly enough in nineteenth-century
practice, adding drama to sonata forms by postponing large-scale resolution
to the coda. But this does not happen here. Indeed, there is no coda, and the
tonic never returns: the music remains in E-flat, and the slow movement fol-
lows in that key, leaving the subordinate theme, and with it the allegro move-
ment’s tonal process, unresolved.
The slow movement pursues a ternary design, with the E-flat “A” section
flanking a contrasting middle that settles in G major and minor after start-
ing out in G-flat. The third movement continues to shift between these same
tonalities, passing from G minor in the Scherzo to E-flat major in the Trio,
which repeats in the same key after the reprise of the Scherzo. At the move-
ment’s conclusion, Saint-Saëns reinterprets E-flat as the Neapolitan to prepare
the home-key dominant, which arrives at the tempo change to allegro molto
that initiates the D-major Finale. Given the lack of tonal resolution in the first
movement and the emphasis placed on E-flat in the Adagio and the Scherzo,
one might posit something approaching a single-movement Ursatz spanning
the sonata’s four movements, with the first progressing from D to E-flat, the
second and third prolonging the Neapolitan by way of parsimonious voice
leading highlighting G, the outset of the Finale arriving at the dominant and
discharging into the tonic, and the Finale’s structural cadence clinching the
tonic and completing the progression (see ex. 4.6). The sonata-form Finale
resumes and completes the thematic process abandoned in the opening move-
ment. As Timothy Jones observes, the unresolved first-movement subordinate
theme returns in the final coda, first in B-flat and then in D major, and so
finally receives its missing home-key recapitulation.12 Jones might also have
noted that the Finale’s main theme quotes a fragment of this same theme (see
exx. 4.5b and 4.7), preparing its belated return and lubricating its entry into
the Finale’s form.
Where the Violin Sonata suspends its overarching thematic process at the
conclusion of the opening movement and resumes it near the end of the
Finale, the Fourth Piano Concerto and the Third Symphony elaborate a tonal
and thematic process across their entire spans, with individual movements
assuming the functions of sonata-form sections. The Piano Concerto’s opening

Table 4.5. First Violin Sonata: large-scale form
PART I
1 Allegro 2 Slow movement
Exposition Development Recapitulation A B A
Main theme Sub. Main theme Sub. theme

theme
D– F+ D– E♭+ E♭+ G♭ G D E♭
PART II
3 Scherzo 4 Finale
Scherzo Trio Scherzo Trio Exposition Development Recapitulation Coda
Main Sub. Main theme Sub. Sub.

theme theme theme theme
of I
G– (D) G– E♭+ G– E♭ = ♭II/D D+ A+ D+ D+ B♭+

D+
9/30/2015 7:51:54 PM
Example 4.5a. First Violin Sonata, mm. 1–11
Example 4.5b. First Violin Sonata: subordinate theme, mm. 76–83

Example 4.5c. First Violin Sonata: development section fugue, mm. 152–65
Example 4.6. First Violin Sonata: overarching tonal design

Example 4.7. First Violin Sonata, Finale, mm. 1–6
movement, unusually, unfolds a set of variations, in all of which each section

of the binary theme repeats with decoration before the next section follows,
nesting variation within variation (see table 4.6). Just as unusually, the entire
movement remains in the tonic C minor; no modulation occurs until the tran-
sition to the A-flat-major slow movement. Thus while this concerto’s opening
movement, contrary to the other works we have examined, does not remain
tonally unresolved, it nonetheless registers as incomplete since it entirely lacks
harmonic adventure. The second movement will therefore strike listeners as a
continuation of the first rather than a fresh and autonomous musical structure.
Given that the Adagio effects the work’s first modulation and introduces, for
the first time, new thematic material, the first two movements together come
to function as an exposition on a higher formal level, respectively assuming the
roles of main and subordinate theme areas (see table 4.7).
With the Allegro–Adagio movement pair expressing a large-scale exposi-
tion, the Scherzo–Finale pair must elaborate the form they initiate and bring
it to conclusion. Given these roles, the concerto’s two large, continuous parts
(see again table 4.4) suggest the two sections of a sonata form conceived as a

Table 4.6. Fourth Piano Concerto: first-movement form

Theme Variation 1 Variation 2
a a′ b b′ a a′ b b′ a a′ b b′
Table 4.7. Fourth Piano Concerto: large-scale form
Cycle 1 First movement 2 Second movement 3 Third movement 4 Fourth movement (rondo)
Theme Varia- Varia- Theme Theme Develop- Th. 1 Th. 2 Scherzo Trio Scherzo Intro. = Main theme /\/\/\/ Coda
tion tion 1 2 ment Theme+ Popular Themes = Theme 1
coda of I theme of II of II
C– A♭+ A♭- C- A♭+ A♭– C+ A♭– C– E♭+G- E♭+ C– E♭+G– C– A♭+ C+ C+
Over- Main theme Subordinate theme area Development Recapitulation Coda

arching
Form
9/30/2015 7:51:58 PM
binary plan, perhaps Saint-Saëns’s motivation for this distinctive layout. The
third movement engages some procedures characteristic of a development sec-
tion. As shown in table 4.7, this Scherzo juxtaposes melodic-motivic fragments
from the preceding exposition: Saint-Saëns bases measures 1–33 on the transi-
tion from the first movement to the second, and measures 34–74 reintroduce
the first movement’s theme.13 These materials flit rapidly between keys that
remain only momentarily stable, with the first theme, firmly ensconced in the
tonic throughout the opening movement, now bouncing between G minor
and E-flat major. The Trio, on the other hand, remains in a stable E-flat major
and introduces a new theme in a popular style (see ex. 4.8), attributes largely
incommensurate with a development section. To apply Vande Moortele’s con-
ceptual apparatus, the dimension of the form here goes dormant, with the
passage functioning exclusively in the dimension of the cycle, and so both sat-
isfying the generic requirement that a scherzo include a trio and providing
needed contrast to the turbulence that otherwise characterizes the movement.
The Finale provides formal completion to the overarching sonata form by
recapitulating the themes of the second movement—which in the concerto’s
overarching form functions as a subordinate-theme group—in the tonic key. A
slow introduction first recasts the Adagio’s second theme as a C-minor fugato
and then restates its first theme in its original chorale texture and in its original
A-flat-major tonality. A transformation of this same theme (see ex. 4.9) materi-
alizes in C major at the tempo change to serve as the rondo-form Finale’s main
theme. The only other theme presented off tonic anywhere in the concerto is
the E-flat-major tune from the Trio. Although this melody never returns, the
Finale’s main theme does recall its popular style, similarly unfolding as a single,
sparse melodic line—which the pianist can practically play with a single fin-
ger—and falling into four symmetrical phrases. It additionally assumes a penta-
tonic profile, not an attribute of example 4.8 but a mode of pitch organization
recognized as a characteristic of popular music by the burgeoning discipline of
folk-music scholarship in fin-de-siècle France.14 The Finale, then, recapitulates
the style of the Trio theme, resolving it, so to speak, into the tonic.
The Third Symphony stands as Saint-Saëns’s most famously cyclic composi-
tion. As is well known, the main theme of the opening movement (ex. 4.10a)
returns, transformed in the manner of Liszt, the symphony’s dedicatee, in each
subsequent movement. Also suggesting Liszt, these transformations etch a
Example 4.8. Fourth Piano Concerto: popular-style theme in Scherzo, reh. K+8
to K+13

Example 4.9a. Adagio’s second theme returns in Finale’s introduction
Example 4.9b. Adagio’s second theme returns transformed as Finale’s main theme
trajectory from C-minor strife at the outset to a tonic-major apotheosis that

sees the theme transformed into a magnificent chorale (ex. 4.10b). However,
Lisztian thematic transformation constitutes only one element of a rich and
sophisticated thematic design. This principal, cyclic theme comprises two
important motives, x and y (to adopt Daniel Fallon’s convenient nomencla-
ture), which first sound separately in the slow introduction (ex. 4.11).15 As
example 4.12 reveals, Saint-Saëns derives virtually all of the symphony’s other
themes from one or both of these motives, often treating them to complex and
subtle manipulations, including retrograde and inversion, and sometimes trun-
cating y. The opening-movement subordinate theme (ex. 4.12a), for example,
largely derives from y: minus its first pitch, this motive initiates the melody, and
a retrograde version forms the second half of the first measure. In the third
measure, an entire y sounds in retrograde, with the third inverted to a sixth.
Motive x also factors in the melody’s fourth and fifth measures in addition
to providing the bass line for measures 2 through 5. Some especially intricate
motivic work arises in a trio-like subsection of the third movement’s middle
section, comprising a pair of melodic ideas, shown in examples 4.12f and
4.12g. In the first, an x and a y interlock in a four-measure model that repeats
sequentially three times; in the final iteration, an elaborated x extends the line
up to C, where the second “trio” melody begins. This tune, likewise sequential,
seamlessly strings together iterations of y and x, alternating between them.
The per aspera ad astra narrative traced by the transformations of the cyclic
theme interlocks with an overarching sonata process spanning the symphony’s
four movements (table 4.8). The opening Allegro, as in the two early concerti and
the First Violin Sonata, remains syntactically incomplete. In the recapitulation,

Example 4.10a. Third Symphony, mvt. 1: main theme, mm. 12–14
Example 4.10b. Third Symphony, mvt. 4, reh. S
(continued)

Example 4.10b.—(concluded)
Example 4.11. Third Symphony, slow introduction, mm. 1-4

Example 4.12a. Third Symphony, mvt. 1, reh. F to F+5
Example 4.12b. Mvt. 2, reh. Q–14 to Q–3
Example 4.12c. Mvt. 3, mm. 1–5
Example 4.12d. Mvt. 3, reh. C+12 to C+15
Example 4.12e. Mvt. 3, reh. D–2 to D

Example 4.12f. Mvt. 3, reh. F+8 to F+14
Example 4.12g. Mvt. 3, reh. G
Example 4.12h. Mvt. 3, reh. Q+6
Example 4.12i. Mvt. 4, reh. V
the main theme returns in the home key, but the subordinate theme returns
in F major; as in the First Violin Sonata, the movement lacks a coda and
the recapitulation never returns to the tonic, modulating instead to D-flat,
the tonality of the second movement. Interrelating the two movements, D♭
occurs at several points in the first, including in the opening measure (see
ex. 4.11 above) and at the outset of the development. Significantly, D-flat is
also the tonality, distinctive given the C-minor context, of the subordinate
theme and therefore the principal rival to the tonic. Consequently, not only
does the Allegro remain unresolved and formally fragmentary, but its central
tonal premise also becomes written across the entirety of Part I (the Allegro–
Adagio pair) and thus reinscribed upon a higher formal order: the conflict
between C and D-flat, initially articulated by theme areas, now becomes reart-
iculated by entire movements. As in the Fourth Piano Concerto, Saint-Saëns
confounds the normative, hierarchical relationship between the cycle and

the forms of the movements that comprise it by transferring the function

of exposition to a higher level of structure (see fig. 4.1): the first two move-
ments respectively come to function as the main and subordinate theme
areas of the overarching form.
Also as in the Fourth Concerto, the symphony’s third movement assumes
the function of a development. Its contrasting middle section (C+12 to K+13)
employs some characteristic compositional techniques. In the first of two
alternating subsections (occurring at C+12 and H), the music quickly passes
through a series of momentarily stable keys, with brief snatches of thematic
material (examples 4.12d and 4.12e, and the rapid-fire scales on the piano at
D–6) quickly succeeding one another. The other subsection (at F+8 and I+4)
more closely resembles a conventional trio, although Saint-Saëns presents its
materials (exx. 4.12f and 4.12g) as model-sequences—traditional develop-
ment-section workhorses—rather than as conventional theme types.
The Scherzo portion of the movement functions as a development primar-
ily by effecting a rapprochement between the principal tonal centers of part
1 and thereby working out the main formal problem posed in exposition of
the overarching form. As table 4.9 shows, D-flat major factors conspicuously
here. It first appears as the Neapolitan in the ternary Scherzo’s contrasting
(“b”) subsection (at A+7), where tonicization via its dominant seventh adds
emphasis. This local harmonic move becomes composed out upon the reprise
of the Scherzo’s “a” subsection, the opening sentence of which now modulates
to arrive at an HC in D-flat instead of in C (ex. 4.13). This D-flat, which ini-
tiates the much-expanded sentence (A+18 to C+12) leading to the Scherzo’s
final cadence, also comes to function as a large-scale Neapolitan. Saint-Saëns
stretches this sentence out, allowing for yet more emphasis on D-flat, by care-
fully manipulating phrase rhythm. From a purely harmonic-syntactic point of
view, the cadential dominant appears to arrive at B+12 and resolve two mea-
sures later. The movement, however, has thus far fallen almost exclusively into

Overarching form: Exposition

First movement Second movement
(Allegro) C- (Adagio) D‫ڷ‬+

Allegro: Exposition

Main theme Subordinate theme

C- D‫ڷ‬+

Figure 4.1. Third Symphony: exposition transferred to higher formal level

Table 4.8. Third Symphony: large-scale form

1.
Intro. Exposition Development Recapitulation
m. 1 11 Reh. F H+9 G M O+7
Main theme = Subordinate Cyclic theme Main theme Subordinate theme

Cyclic theme theme
C– V–♭VI D♭+ F+ E+ C– F+ V of D♭
=V/D♭ no tonic, no cadence!
2.
A B A
Reh. P+35 S+3 U V
Main theme Variation Cyclic theme Main theme

(organ enters) (organ)
D♭+ D♭+ Minor keys D♭+
9/30/2015 7:52:04 PM
3.

Scherzo Contrasting Middle Scherzo
m. 1 Reh. C+12 K+14 O+2
Main theme and cyclic theme (Piano enters) Main theme + cyclic theme Transition to finale (Fugato;
quotes main theme of II)
C– C–
D♭ becomes functional ♭II of C D♭ = ♭II of C V–♭VI -----------------------------------------------V
4.
Introduction Exposition Development Recap. (thematic) Coda

(tonal
recap.)
Reh. S–8 S T V Y AA+11 BB+6 FF+9
Cyclic theme Cyclic Subordinate Cyclic theme Cyclic theme as chorale Subordinate
as theme as theme theme
Chorale fugue
Fugato (Organ and (with organ) Organ

piano)
C+ C+ B+/D+ G+: V–I Minor mode /\/\/\/\/\/\/\/\/\/\/\/ E+/G+ C+: V–I C+
9/30/2015 7:52:04 PM
Table 4.9. D-flat in scherzo section of third movement

3.
a b a
Reh.A A+7 A+10 A+17 B+16 C+6
C–: I–V I–V I V/♭II–♭II–V I I–V C–: V–I ♭II--- V♯!– I
of ♭II------
HC HC HC HC No PAC
cad.!
four-measure phrases, engendering an insistent hypermeter. The tonic at B+14

falls in a weak metrical position, so no sense of cadence obtains, and the phrase
continues unbroken through to the next hypermetric downbeat at B+16. The
harmony there is once again D-flat major, which after lingering for four mea-
sures progresses to the true cadential dominant at C+6. Clearly, the harmonic
role of D-flat has evolved. In the Allegro and the Adagio (insofar as the latter
registers as a secondary key in a higher-order formal scheme), D-flat acts as a
destabilizing, disruptive entity, a rival to the tonic’s claim to sovereignty. In the
Scherzo, on the other hand, it functions—invariably, repeatedly, and at mul-
tiple structural levels—as the Neapolitan and thereby affirms the tonic by pre-
paring the dominant. Put differently, D-flat has become tonally “domesticated”
through its absorption into the syntactic fold of C.
This whole process of domestication repeats in the reprise of the Scherzo,
after which begins a similar one involving D-flat’s dominant. A-flat first appears
in a structural capacity in the opening movement’s exposition, where it arises
as the deceptive resolution of C minor’s dominant to initiate the transition
(letter E; see again table 4.8) and prepare the D-flat tonality of the subordinate
theme. A-flat functions in the same capacity at the corresponding place in the
recapitulation, and there too the music winds up in D-flat for the Adagio. At
the conclusion of the Scherzo (O+2), A-flat arises once more as a deceptive
resolution of C minor’s dominant, leaving the movement, like the first, tonally
open, and it threatens to deflect the music away from the home key yet again.
(Although a C-minor triad materializes at letter O, it holds a weak hypermetric
position and consequently does not register as a resolution of the dominant,
which surges through to the metrically strong A-flat chord at O+2.) The eighty-
measure transition to the Finale, comprising a fugato on its own thematic
idea (ex. 4.12h), lingers on A-flat. This time, however, A-flat winds up neither
leading to D-flat nor truly destabilizing the tonic, as the passage concludes by
settling back onto V of C to launch the Finale. Just as D-flat becomes domesti-
cated in the Scherzo, the once harmonically disruptive A-flat here assumes its
normative syntactic and contrapuntal role as a neighbor to the dominant.

Example 4.13. Third Symphony, mvt. 3: scherzo section, modulation to D-flat,
reh. A+14

The Scherzo having played a development-like role, the beginning of the

Finale fulfills the function of recapitulation abandoned in the first movement.
It does so in several ways. Most emphatically, at letter S an apotheosis-like
return of the cyclic theme occurs, now transformed into the chorale shown
in example 4.10b and resolved into the tonic major, which marks the telos of
the work’s per aspera ad astra narrative. The introduction also includes a sub-
tler deed of recapitulation. As Ralph Locke has observed, the fugato subject
in the transition from the Scherzo (ex. 4.14b) quotes at pitch the beginning
of the slow movement’s main theme (ex. 4.14a), now reharmonized in A-flat.16
Following a blast from the organ that initiates the Finale, a rhythmically ener-
gized version of this theme (ex. 4.14c) appears in C major: thematic mate-
rial introduced in the D-flat Adagio, which in the symphony’s overarching,
four-movement form functions like a subordinate theme area, here becomes
grounded in the tonic. Saint-Saëns even brings timbre into the process. The
symphony famously includes two unconventional instruments, organ and
piano. The former is introduced in the D-flat Adagio movement and the lat-
ter in the modulating, development-like trio subsection of the third move-
ment. At the outset of the Finale, both play in C for the first time, their timbres
“resolved” (like the popular style in the Fourth Concerto) into the tonic.
Example 4.14a. Third Symphony, mvt. 2: main theme, reh. P+37ff.
Example 4.14b. Transition to Finale, reh. Q+6ff.
Example 4.14c. Beginning of Finale, reh. S–7ff.
❧ ❧ ❧
Although his career would span a further thirty-five years and include more
works in Viennese classical genres, after 1886 Saint-Saëns largely renounced

thematically integrated multimovement designs like the examples we have

addressed here. He perhaps felt he had pursued cyclic form as far as he could
in the Third Symphony, as some of his correspondence suggests.17 Cultural-
political factors likely weighed most heavily. Starting in the mid-1880s, Franck
composed the majority of his most influential cyclic works, including the
Prélude, choral et fugue (1884); the Violin Sonata (1886); the Prélude, aria et final
(1887); the Symphony (1888); and the String Quartet (1889). This remarkable
cluster followed the Piano Quintet of 1879. D’Indy also made his first mature
foray into cyclic form around the same time with the Symphonie sur un chant
montagnard français (1886), which scored a major public success and helped
nudge him to the front rank of French composers. A piano trio and a pair of
quartets, all ostentatiously cyclic, would follow from d’Indy’s pen before the
century’s end. Some of Franck’s other students with public profiles similarly
adopted cyclic procedures in the 1890s, notably Ropartz (two symphonies, a
quartet, and others), Chausson (a symphony, a piano quartet, and the Concert),
and Lekeu (a piano trio and a violin sonata).
These developments might have caused Saint-Saëns little concern. However,
as d’Indy and some of Franck’s other disciples gained stature and banded
together as an increasingly influential aesthetic “camp” united under the ban-
ner of their teacher’s legacy, his relationship with them became fractious. He
held a measure of respect for Franck’s music, but hardly considered it wor-
thy of widespread emulation; and although Saint-Saëns admired Wagner,
the Franckistes in his estimation drew excessively on the style and ideas of
Der Meister, to the detriment of French music.18 Contributing greatly to the
rift, d’Indy and some of Franck’s other students began to seize control of the
Société nationale de musique in the early 1880s.19 Saint-Saëns had cofounded
this composers’ league and concertizing organization to revitalize the nation’s
musical culture in the aftermath of the Franco-Prussian war and the bloody
Paris Commune, and in the ensuing decade it had become a vitally important
platform for French orchestral and chamber music. Yet he found himself with
little choice but to resign when d’Indy’s party passed a motion in November
1886 to include foreign works on the society’s previously French-only programs,
a development that flatly contradicted the society’s mandate as Saint-Saëns had
originally formulated it. Although Saint-Saëns maintained a civilized profes-
sional relationship with d’Indy and his Franckiste colleagues, he would remain
bitter for years.20 He even came to resent the benevolent Franck himself, who
kept distant from the Société nationale’s politics, and who rarely expressed
anything but admiration for Saint-Saëns. Given these factors, it seems easy to
imagine the composer coming to feel ambivalent about cyclic form.
The same factors make a revealing context for Saint-Saëns’s First String
Quartet, written in 1899. This work appears to parody the approaches to cycli-
cism in the quartets of d’Indy, especially the well-received second of 1897 (in
which a four-note cell suffuses most of the themes) and Franck (in which

themes from previous movements reemerge juxtaposed and superimposed in

the final coda). Saint-Saëns’s quartet abounds with quasi-tangible connections
between themes involving miniscule motivic details, especially accented minor
seconds, which will catch many listeners’ ears but seem too fleeting and insub-
stantial to register as binding elements in the manner of d’Indy’s pellucid cell.
Such relationships seem purposefully liminal and calculated to straddle the
threshold of apprehensibility, and they result in a distinctive affective quality:
where d’Indy’s procedure fosters perceptions of genetic relationships between
themes, Saint-Saëns’s elicits only a sensation of déjà entendu. The final coda
similarly teases by reintroducing readily recognizable fragments from the slow
introduction (rehearsal 48), encouraging listeners—especially those having in
their ears the monumental peroration at the corresponding spot in Franck’s
Quartet—to anticipate a more substantial, climactic return. What actually fol-
lows is a mirage of one: the texture, timbre, and rhythm at rehearsal 49 unmis-
takably recall the Scherzo, but the melodic line itself derives from the Finale’s
main theme; Saint-Saëns then offers a flash of the quartet’s opening (compare
the second violin at 50+4ff. to the first violin at m. 13ff.), but the Finale’s main
theme quickly asserts itself, dissolving any sense of return. Where the coda to
Franck’s Quartet effects a full-blooded synthesis of themes, Saint-Saëns’s feints
at the same but follows through only with trompe l’oreille and phantoms.21
After the E-Minor Quartet, Saint-Saëns did on one occasion revisit a com-
positional strategy similar to those outlined above: the Second Cello Concerto
of 1902 assumes a two-dimensional-sonata design much like that of the First
Cello Concerto. But this would be the last time he would do so. This work’s
appearance coincided with d’Indy’s launching of a vigorous campaign to
claim for Franck the singular provenance of cyclic form and thereby to claim
for himself and his Franckiste colleagues an exclusive pedigree. In his com-
position classes of the late 1890s at the Schola cantorum (the Parisian music
academy he helped found), d’Indy elaborated a history of the procedure,
which maintained that Franck had inherited it directly from the venerable
Beethoven. (He acknowledged that Berlioz, Schumann, Mendelssohn, Liszt,
and others had attempted cyclic forms, but he dismissed their efforts as deriva-
tive or misguided.) These lectures were later codified as the Cours de composi-
tion musicale, but even before the relevant volumes were published in 1909 and
1911, d’Indy disseminated this same narrative in other, more public, forums,
including a multipart article appearing in 1904 in the weekly Revue musicale de
Lyon, edited by his friend, colleague, and eventual biographer Léon Vallas, and
his influential 1906 biography of Franck.22 He probably also did so in public
lectures given under the auspices of the Schola.23 Members of d’Indy’s circle
also participated in the campaign. Vallas, for example, proposed this same evo-
lutionary lineage in articles appearing in 1904 on Georges-Martin Witkoski’s
E-Minor String Quartet and D-Minor Symphony and would repeat it often in
his columns in the ensuing decade.24 The publication of volume 2 of the Cours

elicited a number of press reviews, with some recapitulating d’Indy’s history of

cyclic form, authored by critics both within his circle and outside of it, notably
Adolphe Jullien.25
Whatever writers might have thought of d’Indy’s highly problematic his-
tory of cyclic form—with its specious, ad hoc dismissals of Schumann, Berlioz,
Mendelssohn, and Liszt—enough of them proved perfectly willing to accept
Franck as its leading proponent in late nineteenth-century France and d’Indy
as his heir. These included not only members of d’Indy’s circle and those sympa-
thetic to his camp (notably Gaston Carraud) but also the likes of the young critic
Jean Marnold, an emerging champion of Debussy and Ravel.26 Indeed, by late in
the decade, it became routine for writers to associate cyclic form with Franck and
his followers. To cite just one improbable example, Charles Chambellan, a critic
for the sophisticated Mercure musical, likened Strauss’s treatment of themes in the
Finale of the Symphonia domestica to Franck’s practice.27
Saint-Saëns may well have recognized all of this, and he surely recognized that
his own cyclic works rivaled in sophistication those of his countrymen. But that
would have mattered little. Between the sheer numbers of cyclic works mem-
bers of Franck’s circle produced and the effectiveness of their discursive cam-
paign, cyclic form became firmly and, it would seem, indelibly associated with
the Franck–d’Indy axis. In a forty-page review of the Cours de composition, Saint-
Saëns took issue with many facets of d’Indy’s text, notably the position in music
history it accorded to Franck.28 But he did not even bother to dispute d’Indy’s
account of cyclic form or his implicit claim that he and his Franckiste colleagues
were its sole legitimate present-day progenitors. Indeed, so strong became the
association of cyclic form with Franck and d’Indy that it appears to have made
an impact on one of the period’s most creative and independent musicians. As
Wheeldon has observed, the youthful Debussy plainly emulated the cyclic design
of Franck’s String Quartet in his own quartet of 1893, possibly with an eye to gain-
ing a performance at the Société nationale.29 But in his in his three late sona-
tas, for cello (1915), flute, viola, and harp (1915), and violin (1917), Debussy
engaged cyclic procedures with altogether more reserve, perhaps weary of being
“identified with a musical heritage that began with Beethoven and culminated
with d’Indy and his students at the Schola cantorum.”30 Saint-Saëns would have
had little choice but to react even more strongly. His rivals had appropriated and
come to hold a firm lock on a category of compositional procedures he had for
decades employed regularly and fruitfully: Saint-Saëns may well have found him-
self forced out of the business of cyclic form by the competition.
Notes
1. Benedict Taylor, Mendelssohn, Time, and Memory: The Romantic Conception of Cyclic Form
(Cambridge: Cambridge University Press, 2011), 1–2; Hugh Macdonald, “Cyclic
Form,” Grove Music Online, http://www.oxfordmusiconline.com/subscriber/article/

grove/music/07001 (accessed April 16, 2014); “Cyclic Form,” The Oxford Dictionary
of Music, 2nd rev. ed., Oxford Music Online, http://www.oxfordmusiconline.com/
subscriber/article/opr/t237/e2652 accessed April 16, 2014).
2. D’Indy’s writings on cyclic form include his Cours de composition musicale, vol. 2, bk. 1
(Paris: Durand, 1909), 375–422, and vol. 2, bk. 2 (Paris: Durand, 1911), 121–78.
3. Michael Puri, Ravel the Decadent: Memory, Sublimation, and Desire (New York: Oxford
University Press, 2011), 32–51; Marianne Wheeldon, “Debussy and La Sonate
cyclique,” Journal of Musicology 22 (2005): 645–59; Brian Hart, “Vincent d’Indy and
the Development of the French Symphony,” Music and Letters 87 (2006): 237–61 and
“The French Symphony after Berlioz: From the Second Empire to the First World
War,” in A. Peter Brown, The Symphonic Repertoire, vol. 3, The European Symphony
from ca. 1800 to ca. 1930: Great Britain, Russia, and France (Bloomington: Indiana
University Press, 2008), 527–722; Serge Gut, “Y a-t-il un modèle beethovénien pour
la symphonie de Franck?” Revue européenne d’études musicales 1 (1990): 59–79. See
also Andrew Deruchie, The French Symphony at the Fin de Siècle: Style, Culture, and
the Symphonic Tradition (Rochester, NY: University of Rochester Press, 2013); and
Taylor, Mendelssohn, Time, and Memory, 9–16.
4. Taylor, Mendelssohn, Time, and Memory, and “Cyclic Form, Time, and Memory in
Mendelssohn’s A-Minor Quartet, Op. 13,” Music and Letters 93 (2010): 45–89; Puri,
Ravel the Decadent, 32–51. See also the chapter on d’Indy’s Symphonie sur un chant
montagnard français in Deruchie, The French Symphony, 152–84.
Movement Instrumental Works by Liszt, Strauss, Schoenberg, and Zemlinsky (Leuven:
Leuven University Press, 2009). On Liszt’s B-Minor Sonata, see 23–24 and 35–57, on
Tasso see 59–71, and on Die Ideale see 71–78. See also Steven Vande Moortele, “Two-
Dimensional Symphonic Forms: Schoenberg’s Chamber Symphony, Before, and
After,” in The Cambridge Companion to the Symphony, ed. Julian Horton (Cambridge:
Cambridge University Press, 2013), 268–84; and “Beyond Sonata Deformation:
Liszt’s Symphonic Poem Tasso and the Concept of Two-Dimensional Sonata Form,”
Current Musicology 86 (2008): 41–62.
6. See, for example, the first movement of Mendelssohn’s Violin Concerto, in which
the cadenza prefaces the recapitulation; Tchaikovsky would similarly situate the
cadenza in the first movement of his Violin Concerto.
7. On the composer’s usage of the Neapolitan, see Sabina Teller Ratner, “The Piano
Works of Camille Saint-Saëns” (PhD diss., University of Michigan, 1972), 208.
8. According to Saint-Saëns’s early English-language biographer, Watson Lyle, in the
composer’s day this minuet was sometimes performed as a stand-alone piece. See
Lyle, Camille Saint-Saëns: His Life and Art (London: K. Paul, Trench, Trubner, 1923),
101.
10. See, for example, the editions published by Fischer (New York, 1915) and Thomi-
Berg (Munich, 2003).
11. On the relationship between Saint-Saëns’s theme and Proust’s fictional phrase, see
Jean Gallois, Camille Saint-Saëns (Sprimont: Mardaga, 2004), 250–52, and James
Harding, Saint-Saëns and His Circle (London: Chapman and Hall, 1965), 201.
12. Timothy Jones, “Nineteenth-Century Chamber and Orchestral Music,” in French
Music Since Berlioz, ed. Richard Langham Smith and Caroline Potter (Aldershot:
Ashgate, 2006), 63–64.

13. As Puri has observed, the descending chromatic scale segments that figure in the
transition from the first movement to the second and again at the opening of the
Scherzo themselves develop the first movement’s theme in that they compress the
pitch-class collection of its first five measures into the two tetrachords [A♭–G–
F♯–F] and [E♭–D–D♭–C]. Puri, “Saint-Saëns, Ravel, and Their Piano Concertos:
Sounding Out a Legacy,” in Saint-Saëns and His World, ed. Jann Pasler (Princeton,
NJ: Princeton University Press, 2012), 345–46.
14. See, for example, d’Indy’s introduction to his first collection of folk songs from
his ancestral Vivarais region, Chansons populaires du Vivarais (Paris: Durand, 1900).
See also Jann Pasler, “Race and Nation: Musical Acclimatization and the Chansons
populaires in Third Republic France,” in Western Music and Race, ed. Julie Brown
(Cambridge: Cambridge University Press, 2007), 147–67; and Stéphane Giocanti,
“De l’histoire à l’esthétique régionalistes,” Les Cahiers de mémoire d’Ardèche et temps
présent 53 (1997): 31.
15. Daniel Fallon, “The Symphonies and Symphonic Poems of Camille Saint-Saëns”
(PhD diss., Yale University, 1973), 359–430.
16. Ralph P. Locke, “The French Symphony: David, Gounod, and Bizet to Saint-Saëns,
Franck, and Their Followers,” in The Nineteenth-Century Symphony, ed. D. Kern
Holoman (New York: Schirmer, 1997), 176.
17. “I have given all that I have to give,” he famously wrote to his publisher Durand,
“what I have done I shall never do again.” This pronouncement is usually taken to
express the composer’s disinclination to write another symphony, which he did not,
but perhaps it applied (or also applied) to compositional technique. Quoted in
Stephen Studd, Saint-Saëns: A Critical Biography (London: Cygnus Arts, 1999), 155.
18. See the Introduction to Saint-Saëns’s Harmonie et mélodie, a compilation of his criti-
cism published in 1885, in Camille Saint-Saëns on Music and Musicians, trans. and ed.
Roger Nichols (New York: Oxford University Press, 2008), 3–11.
19. See Michael Strasser, “Ars Gallica: The Société nationale de musique and Its Role in
French Musical Life, 1871–1891” (PhD diss., University of Illinois, 1998), 369–443.
20. On the composer’s relationship with d’Indy, see Jann Pasler, “Saint-Saëns and
d’Indy in Dialogue,” in Pasler, Saint-Saëns and His World, 287–303.
21. D’Indy himself heard much of this in the quartet, though if he perceived any irony
he did not let on. His analysis identifies the melodic minor second as the “cyclic
bond for the entire work,” but frowns upon this cell for “making an insufficiently
distinct and lasting impression on the listener.” He likewise concludes that the
“excessively weak character” this interval bestows on materials from the slow intro-
duction “detracts greatly” from the effectiveness of their recall in the final coda.
D’Indy, Cours de composition, vol. 2, bk. 2, 266–67. Georges Servières reacted simi-
larly; see Servières, Saint-Saëns (Paris: Alcan, 1923), 113–15.
22. D’Indy, “De Bach à Beethoven,” Revue musicale de Lyon, November 13, 1904, 37–39;
November 20, 1904, 49–51; and November 27, 1904, 61–65; and d’Indy, César Franck
(Paris: Alcan, 1906).
23. See Jane Fulcher, French Cultural Politics and Music: From the Dreyfus Affair to the First
World War (New York: Oxford University Press, 1999), 28.
24. Léon Vallas, “Le Quatuor en mi de G. M. Witkowski,” Revue musicale de Lyon, March
2, 1904; and “La Symphonie en ré mineur de G-M Witkowski,” Revue musicale de
Lyon, May 22, 1904, 341–43.
25. Adolphe Jullien, “Revue musicale,” Journal des débats politiques et littéraires, December
12, 1909.

26. Gaston Carraud, “La Musique pure dans l’école française contemporaine,” La
Revue musicale S.I.M., July 15, 1910, 483–505; see esp. 486, 496, 497, and 503; and
Jean Marnold, “Musique,” Mercure de France, March 1, 1908, 158.
27. Charles Chambellan, “Revue de la quinzaine,” Le Mercure musical, June 15, 1906, 359.
28. Camille Saint-Saëns, Les Idées de M. Vincent d’Indy (Paris: Pierre Lafitte, 1919).
29. Marianne Wheeldon, “Debussy and La Sonata cyclique.”
30. Ibid., 688.

Part Three
Schubert

Chapter Five
Schubert’s “Deflected-Cadence”
Transitions and the Classical Style
Brian Black
Schubert has long been famous for the striking character of his modula-
tions. In fact, they are considered a hallmark of his style, particularly in
his sonata-form transitions. Here, the precise moment that the subordinate
key enters is often highlighted as an extraordinary event, capable of infus-
ing a new and intense atmosphere into the subordinate theme.1 To achieve
this effect, Schubert uses a variety of schemes. One of the most prominent
of these consists of a “deflected-cadence” strategy involving two successive
cadential progressions. The first, which occurs in the home key, may either
achieve closure or be thwarted by an evaded or deceptive cadence. It is fol-
lowed immediately by the second cadence, which begins the same way, only
to be diverted at the last moment into a perfect authentic cadence (PAC)
in the subordinate key. The modulation is thus accomplished exclusively
by the second cadence, which both ends the transition and ushers in the
subordinate-key region.
Schubert’s deflected-cadence strategy constitutes a small, yet significant,
departure from the classical transition, which usually targets the dominant,
rather than the tonic, of the new key. The use of a concluding PAC and
the unusual way it is set up create a new effect that distinguishes this type
of transition from those of Haydn, Mozart, and Beethoven, as well as many
of Schubert’s own generation. What is more, the scheme itself has serious
ramifications for other sections of the form, especially when elevated to a
broader structural level, as shall be seen later in this chapter. Thus in this
one detail of Schubert’s sonata forms we see a subtle transformation of the
classical style indicative of a new approach to sonata form and its aesthetics
in the early nineteenth century. This situation emerges more clearly in a
further comparison of Schubert’s deflected-cadence strategy with conven-
tional classical practice.

166 brian black
The Classical Sonata-Form Transition

in the Theory of Formal Functions
We will begin with a discussion of the characteristics of the classical transi-
tion as defined by William E. Caplin’s theory of formal functions. In Caplin’s
theory, the transition (what concerns us in this case is the modulating type)
destabilizes the home key by a modulation that targets the dominant of the
new key, either through a half cadence (HC) or a simple dominant arrival (a
noncadential move to the new dominant). The subsequent subordinate theme
confirms the new key by a PAC.2 Due to this differentiation in function, a PAC
is not accepted as a viable ending for a classical transition.3
Usually the transition is separated from the beginning of the subordi-
nate theme by a clear break in the texture (what James Hepokoski and
Warren Darcy refer to as the “medial caesura”). At times, however, there
is no break and it is not possible to locate the initiating phrase of a sub-
ordinate theme, so that the transition seems to run on, eventually end-
ing with a PAC in the subordinate key. Caplin considers this situation to
exhibit a fusion of transition and subordinate-theme functions: “Some
expositions witness the fusion of transition and subordinate theme func-
tions into a single thematic unit. This unit not only modulates to the sub-
ordinate key—a prime constituent of transition function—but also closes
with a perfect authentic cadence in that key—a fundamental requirement
of the subordinate theme’s function. In such cases, it is not possible to
find an appropriate initiating function for the subordinate theme, even
in retrospect.”4 Here, Caplin insists on the maintenance of the distinction
between transition and subordinate-theme function according to the role
of the PAC as the principal means of tonal confirmation:
Many cases of transition/subordinate-theme fusion . . . seem on the surface

to be transitions that close with a perfect authentic cadence in the subordi-
nate key, instead of the more normal half cadence. To speak in this manner,
however, is to recognize authentic cadential closure as a legitimate deviation
from the normal ending of a transition. But sanctioning this cadential pos-
sibility blurs a theoretical distinction fundamental to this study, namely, that
the authentic cadential confirmation of a subordinate key is an essential cri-
terion of the subordinate theme’s function. It is thus theoretically more con-
sistent to interpret such passages as cases of fusion than to risk confusing the
fundamental characteristics of interthematic formal functions.5
It is transition/subordinate-theme fusion that provides the closest point of

comparison with Schubert’s strategy, above all in the use of the PAC as a goal.
Thus we will begin by looking at a classical example of this structure as found
in the first movement of Mozart’s String Quartet in D Minor, K. 421/417b
(ex. 5.1).6 The passage in question begins in measure 9 after the concluding

schubert’s “deflected-c adence” transitions 167
PAC of the main theme. It consists of an initial dominant prolongation in D

minor (mm. 9–14) followed by a dramatic shift to the dominant seventh of
F major, the subordinate key (latter half of m. 14). From this point on the
music remains in F, moving eventually to a PAC in the new key at measure 24.
Nowhere in this process is the beginning of a subordinate theme clearly artic-
ulated, even though the whole passage presents a series of harmonic events
that usually occur across the transition and the subordinate theme: the move
to an HC in the subordinate key F major (m. 18) and the ensuing dominant
prolongation (mm. 18–20) usually mark the end of a transition while the sub-
sequent PAC in F (m. 24) conventionally serves as the end of a subordinate
theme. The section thus clearly exhibits a telescoping or fusion of a transition
and subordinate theme. What is more, no functionally complete subordinate
theme emerges after the concluding PAC. There is a feeling of a new begin-
ning in m. 25 with different melodic material and a new accompaniment, but
the underlying harmonic structure here (mm. 25–28; not shown) and continu-
ing up to the end of the exposition consists of repetitions of an expanded per-
fect authentic cadential progression, an ending not an initiating gesture. It is
as if the “new theme” that enters in measure 25 were actually part of something
already in progress. This further reinforces the feeling that the end of the pre-
ceding section is involved in some way with subordinate-theme, rather than
exclusively transition, function.
Of particular importance for our discussion of Schubert’s transitional pro-
cess is the substantial weight and time given to the new key in the present
example. Over half of the passage resides in F major (from the latter half of m.
14 to m. 24). Moreover, that key is carefully prepared by its dominant, which is
emphasized through the HC and subsequent dominant prolongation in mea-
sures 18–20. The PAC at measure 24 thus confirms a key whose presence has
been strongly felt since measure 14. As shall be seen, this is quite a different
situation from what is found in Schubert’s deflected-cadence transitions.
The Deflected-Cadence
Transition in Schubert’s Sonata Forms
Schubert’s modulating transitions may be divided into two general categories.
The first maintains the traditional focus on the dominant of the subordinate
key with a concluding HC or dominant arrival and subsequent prolongation.
The second represents a stylistic change from classical practice as defined by
Caplin—one in which the focus shifts to the new tonic through a conclud-
ing PAC. The deflected-cadence type forms an important subgroup of this
category.7 Such transitions achieve their special character through a carefully
constructed “surprise” modulation and thus are related in effect to some of
Schubert’s famous tonal feints.8

Example 5.1. Mozart, String Quartet in D Minor, K. 421/417b, mvt. 1, mm. 5–25
(continued)

As mentioned at the outset of this chapter, Schubert’s deflected-cadence

strategy consists of two cadences in close proximity to each other. The first, in
the home key, provides the model of expectation for the second, which begins
the same way but is suddenly redirected to a PAC in the subordinate key. This
scheme occupies the very end of the transition. In fact, its goal tonic often
elides with the beginning of the subordinate theme. The modulation employs
a conventional pivot chord, but the feeling of tonal movement is heightened
by the element of surprise at the precise point where the second cadence
diverges from the first. This moment creates a marked shift from one tonal
plane to the next. A straightforward example is found in the first movement
of the early Violin Sonatina in G Minor, D. 408, from 1816 (see ex. 5.2). Here,
the transition begins with a restatement of the main theme’s basic idea (mm.
12–14), leading to a cadential progression whose arrival on the tonic is dramat-
ically interrupted at measure 16. The cadence’s initial V of IV–IV6 is taken up

170 brian black
Example 5.2. Schubert, Violin Sonatina in G Minor, D. 408, mvt. 1: transition,

mm. 11–18
again (mm. 16–17), but the C-minor 36 chord, acting as the pivot, now prepares
the dominant of B-flat major rather than that of the home key. The new domi-
nant’s unexpected entry creates the modulation’s striking effect of a gentle lift
up to the new key at the cadential arrival in the next measure, which helps to
give the ensuing subordinate theme its sweet, intimate character.
Schubert also uses this strategy for modulations by a fifth, as can be seen in
the first movement of the Piano Quintet in A Major, D. 667 (“The Trout”). The
modulation of the previous example involved a change of mode from minor to
major that enhanced its effect. In the present case the mode remains the same,
but the unusual character of the key change creates a similar lift onto a new
tonal level. The transition begins in measure 37 as a restatement of the second
main theme. The theme’s projected cadence, however, is interrupted by a sud-
den turn through the parallel minor to its relative major, C. Example 5.3 begins
at the point of arrival of C major’s dominant seventh in measure 56. The music
immediately moves back to A major and a new cadential progression (mm.

Example 5.3. Piano Quintet in A Major, D. 667 (“The Trout”), mvt. 1, mm. 56–65

172 brian black
57–60). The cadential arrival is evaded in measure 60 by another intrusion of

the dominant seventh of C major and by a repetition of the preceding run-up
to the cadence (mm. 56–57, repeated in mm. 60–61).9 The cadence, reinsti-
tuted in measure 62, is then deflected from A major to E major by a pivot on
the A-major chord, which now functions as the pre-dominant IV in E. As in
the previous example, the entry of the new dominant in measure 63 creates a
palpable shift to the subordinate key, an effect quite distinct from a typical clas-
sical modulation. The cadential arrival in measure 64 elides with the beginning
of the first of a series of subordinate themes in E major.
Schubert’s Deflected-Cadence
Transitions as a Stylistic Shift
The preceding examples of Schubert’s deflected-cadence strategy differ radi-
cally from the fusion of transition and subordinate-theme functions outlined
in Caplin’s theory. Above all there is little time spent in the new key. In fact,
the key’s arrival is reserved as a deliberate surprise for the very end of the pro-
cess, usually occurring on the dominant of the second cadence. As a result,
the function of tonal confirmation that Caplin identifies with the PAC in clas-
sical practice is no longer operational: the concluding cadence of Schubert’s
scheme can hardly confirm something that has not even been hinted at up
to that point. Furthermore, while the PAC still acts as a closing gesture with
regard to the phrase structure, it has become an initiating gesture with regard
to the tonal structure, since it alone launches the music into the subordinate
key and in such an abrupt manner. Consequently, we cannot talk of a fusion
of functions in these passages: they are exclusively transitions, with full con-
firmation of their concluding key left to the ensuing subordinate theme.10 In
fact one rarely finds Schubert’s subordinate themes beginning in medias res
as in Mozart’s D-minor String Quartet. Rather, they are clearly articulated by a
strong feeling of a new beginning, both melodically and harmonically.11
By this proposal I am not supporting Hepokoski’s and Darcy’s acceptance
of the PAC as a valid ending for a transition in classical practice. Quite the
contrary, I find Caplin’s distinction between transition and subordinate-theme
function convincing for the music of Schubert’s predecessors. I am arguing
instead that what is occurring in Schubert’s deflected-cadence strategy repre-
sents a subtle shift in style.12 And this shift can be seen not only in the unusual
character of his modulations but also in the ramifications that his new modula-
tory technique holds for the form as a whole.
Oddly enough, many of the features on which the strategy depends for
its effectiveness have been cited as weaknesses in Schubert’s sonata forms.
Repetition, for instance, has drawn a great deal of fire from Schubert’s crit-
ics, yet here, it is vital for the success of the scheme.13 Two parallel cadences

are needed, since the surprise of the modulation is created by the unexpected
divergence of the second cadence from the established pattern of the first. As
we have seen, not just the cadence, but large stretches in the run-up to the
cadence are also usually repeated specifically to make the surprise more effec-
tive. Furthermore, the home key must be maintained until the deflection of
the second cadence in order to create the heightened sense of a tonal shift
between the home and subordinate key. Consequently, Schubert’s deflected-
cadence transitions either dwell in the tonic for most of their length or travel
in circles away and back again before moving on to the subordinate tonality.
Classical transitions or transition/subordinate-theme fusions usually follow a
more focused path to the new key, announcing their tonal goal in advance by
approaching the new dominant through its applied dominant or diminished
seventh, or through an augmented-sixth chord. In fact, this is such a standard
practice that it is considered by many commentators as an essential component
of a successful modulation. As Charles Rosen states, “merely going to the domi-
nant and staying there will not work. . . . What follows must still return to V of
V and almost always to V of V of V as well—at least if the music has any ambi-
tion.”14 However, the careful preparation of the new dominant is not possible
in Schubert’s deflected-cadence transitions. Such a clear signal of intent would
give away the tonal surprise beforehand, thus destroying the effect of the sec-
ond cadence’s modulation.15
Finally, the success of the plan clearly depends on different tonal beginning
and end points, since otherwise the whole poetic effect would be lost in a rather
prosaic repetition of a cadential progression in the same key. Thus in the reca-
pitulation, the two-key scheme must be transposed in some way to achieve a
proper realignment toward the tonic. For those schemes moving from tonic to
dominant in the exposition, a possible solution lies in restating the scheme in
the subdominant, either by a complete subdominant recapitulation, as in the
first movement of the “Trout” Quintet (ex. 5.3), or in a subdominant readjust-
ment of the transition at some point before the deflected-cadence modulation.
Those deflected-cadence strategies involving other tonal relationships call
upon individual solutions, but they, too, are often transposed intact in the reca-
pitulation. In the case of the Violin Sonatina in G Minor (ex. 5.2), Schubert
uses a subdominant transposition shortly after the beginning of the transition.
This results in an interrupted cadence in C minor followed by a successful PAC
in E-flat major, the key of the first subordinate theme. The second subordinate
theme is cast in B-flat major, which then yields to the tonic in its final cadences.
The resulting tonal plan of this movement is quite unusual by classical stan-
dards (see table 5.1).
According to Rosen, “even if we were to call this a sonata form for lack of a
better term, its distance from classical procedure is evident and so is its loose-
ness.”16 There is still a definite logic in the form, however: the succession of
keys (E-flat–B-flat–G) in the latter half of the recapitulation reverses that of the

174 brian black
Table 5.1. Violin Sonatina in G Minor, D. 408, mvt. 1: tonal plan,

exposition vs. recapitulation
Exposition
Main Theme Transition Sub. theme 1 Transition Sub. theme 2 Retransition
mm. 1–12 12–18 18–32 32–35 35–50 51–53
G– B♭+ E♭+
Recapitulation
Main theme Transition Sub. theme 1 Transition Sub. theme 2 Coda
86–98 98–106 106–21 121–24 124–40 140–45
G– E♭+ B♭→G– G–
exposition (G–B-flat–E-flat), so that the music cycles back to the tonic through
the same key to which it was first deflected in the exposition. Furthermore, the
tonic is reentered through a PAC in G minor, the very gesture withheld in the
original modulation by deflection (ex. 5.4).
It is interesting to compare Schubert’s practice with that of one of his con-
temporaries, Ludwig Spohr. Spohr’s use of the deflected-cadence strategy
exhibits a more classical handling of the form in retaining the strategy’s special
character—as, for example, in the first movement of his String Quartet in G
Minor, op. 4, no. 2 (1804). In the exposition (ex. 5.5a), the modulation from
G minor to the subordinate key of B-flat major is accomplished by a deflected-
cadence strategy very similar to that of Schubert’s G-minor Sonatina. The
model cadence in G minor is set up in measures 18–20. It is reiterated in mea-
sures 22–23, but ends in a deceptive cadence (m. 24), which leads directly to
the PAC in B-flat major that ushers in the subordinate key at measure 26. As in
the Schubert Sonatina, the arrival of the new dominant in measure 25 coupled
with the transposition of the cadential melody to B-flat creates a gentle lift into
the new key.
For the recapitulation (ex. 5.5b), the main theme returns in G major. The
modal change allows the recapitulation to begin in the tonic and still employ
the minor-to-relative-major relationship of the transition’s deflected-cadence
strategy. Spohr accomplishes this by setting up a brief digression to G’s rela-
tive minor, E (mm. 123–27), then returning to G major through a deflected-
cadence strategy (mm. 126–29) involving a deceptive cadence in E minor at
measures 125–27, answered by a PAC in G major at measure 129.
Spohr’s alterations to the recapitulation thus reassert the tonic key while
preserving the mechanics of the exposition’s modulation intact. The modula-
tion’s effect, however, is seriously weakened by having the transition’s events so
strongly enfolded within G major. In fact, the E-minor digression lends only a
fleeting coloration to the main key. Consequently, when G reemerges in mea-
sure 129, it does not possess the same force as if it had entered as an entirely

Example 5.4. Violin Sonatina D. 408, mvt. 1, recapitulation: return to G minor,
mm. 132–36
Example 5.5a. Spohr, String Quartet in G Minor, op. 4, no. 2, mvt. 1: exposition,
transition, mm. 19–26

176 brian black
Example 5.5b. String Quartet, op. 4, no. 2, mvt. 1: recapitulation, transition,

mm. 122–29
new tonal region. Furthermore, this return is accomplished through the very
same cadence in G (mm. 128–29) that occurred an octave lower only six mea-
sures before (mm. 122–23). The repetition here borders on the redundant
when compared to the parallel passage in the exposition, where the first cadence
in G minor (mm. 19–20) is subtly transformed by its transposition to B-flat major
at the end of the transition (mm. 25–26). Thus Spohr has sacrificed the special
character of the transition’s modulation to the traditional requirements of the
recapitulation—something that Schubert is reluctant to do.

The Deflected-Cadence Strategy in

Schubert’s Early Sonata Forms
Schubert’s first concentrated attempt at mastering sonata form is found in a
series of string quartets he wrote for his family quartet between 1811 and 1816.
Oddly enough for a composer so famous for his modulations, a fully devel-
oped transitional process is not present in the expositions of the earliest sonata
forms.17 By 1814, however, his sonata-form movements all exhibit effective
transitions—including the earliest examples of his use of the deflected-cadence
strategy, as can be seen in the opening movement of the String Quartet in
B-flat Major, D. 112, written in early September 1814.
The transition of this sonata form is extensive and complex, featuring not
one, but three surprise tonal shifts (table 5.2). It is initiated by a sudden turn to
G minor (ex. 5.6a, mm. 35ff.) involving a series of three interlocking cadences
in that key: deceptive (mm. 35–37), deceptive again (mm. 39–41), and finally
a PAC (mm. 41–45). At the transition’s midpoint the music returns unexpect-
edly to B-flat major (mm. 72–73), but subsequently moves back to G minor. It
then equally unexpectedly cadences in F major (mm. 95–103) to usher in the
subordinate theme.
Both the internal return to B-flat major and the final arrival in F employ
deflected-cadence schemes. In the first instance (ex. 5.6b, mm. 61–73), an
evaded cadence in G minor (mm. 61–67) is answered by a PAC in B-flat major
(mm. 67–73). The passage that precedes each cadence adds to the modula-
tion’s sense of mystery, suspending all momentum in the long held D of
the first violin.18 In the second instance (ex. 5.6c, mm. 95–103), an evaded
cadence in G minor interlocks with the final PAC in F major, the common link
being the G36 chord that signals the failure of the first cadence while serving as
the pre-dominant of the second.
The beautiful effect of the F-major cadence depends not only on local paral-
lels but also on a broader recall across the whole transitional process. The forte
outbursts and sudden silences in the cadences at the transition’s end echo those
of the cadences at its beginning. In fact, the primary tonal motion—the move
from B-flat major and the final arrival in F major—is concentrated in these two
points. The overall modulation consists essentially of a conventional pivot on
the G-minor chord as the submediant of B-flat, with the pivot expanded into
a key area of its own (ex. 5.6d).19 Through this expansion, G minor serves
as a blind to the transition’s ultimate tonal destination: most of the transition
proper is taken up with the dramatic struggle between G minor and B-flat
major, but the decision falls at the last moment to F major, an unexpected third
option. The F-major cadence seems to relax all of the previous tensions by dis-
solving the unsettling eighth-note triplet motion and sudden interruptions

Table 5.2. String Quartet in B-flat Major, D. 112, mvt. 1: tonal plan
of exposition
Main theme Transition Sub. theme
mm. 1–34 35–103 103–56
B♭+ G– → B♭+ G–→ F+
35–73 73–92 93–100
X X X
cadential deflected cadence deflected
complex (62–73) cadence
(35–45) (95–103)
↑ ↑
Parallel cadential complexes
Example 5.6a. Schubert, String Quartet D. 112, mvt. 1: cadential complex at

beginning of transition, mm. 33–49

Example 5.6b. String Quartet D. 112, mvt. 1: deflected cadence scheme at mid-
point of transition, mm. 60–73
Example 5.6c. String Quartet D. 112, mvt. 1: deflected cadence scheme at end of
transition, mm. 94–103

180 brian black
Example 5.6d. Basic strategy for modulation from B-flat major to F major
that preceded it into its steady whole-note rhythm and hushed dynamics. As in
previous examples, the arrival of the subordinate key thus constitutes a striking
event with a special emotional color.
Certainly the transition in D. 112 is lengthy and, in many respects, convo-
luted. Due to its great extent, its unusual path, and its “over effortful empha-
sis on G minor,” Susan Wollenberg considers it an indication of what James
Webster refers to as Schubert’s “reluctance” to leave the tonic.20 These features,
however, can equally be seen as an attempt to reinvigorate the modulation to
the dominant, which by this time threatened to become a mere routine.21 Not
only does the deflected-cadence modulation imbue the dominant with a com-
pletely new atmosphere at its point of arrival, but the preceding insistence on
G minor also creates a feeling of distance between B-flat and F. Together, these
effects lift the move to the dominant out of the ordinary and infuse it with new
power. What is impressive here at such an early stage in Schubert’s career is
the effective execution of this plan over a substantial span of music. The next
example, the first movement of the String Quartet in G Minor, D. 173 (from
March 1815), shows similar long-range planning.
Here, the transition’s modulation is itself a fairly straightforward and
concise version of the deflected-cadence strategy: a deceptive cadence in
G minor (ex. 5.7, mm. 41–43) interlocks with a PAC in B-flat major (mm.
43–45). As in D. 112, however, the scheme’s references have been broadened:
the cadential deflection recalls striking elements of the concluding PAC of
the main-theme group (mm. 21–25), specifically the melodic slip from F♯
to F♮ and the cello’s ascending line from B♮ to E♭. Thus the modulation to

Example 5.7. String Quartet in G Minor, D. 173, mvt. 1: end of main theme to end
of transition, mm. 13–45
(continued)

182 brian black
B-flat major involves three cadences, with the final deflected one redirecting
the bass line of the PAC in measure 22 up to F as the new dominant, rather
than returning to D as in measure 23.
In the recapitulation, Schubert increases the span of music involved in the
cadential deflection while radically altering the scheme itself (ex. 5.8). The
main-theme group returns in the relative major, B-flat. Its concluding cadence
begins in B-flat, reaching a lengthy prolongation of the cadential 46 (mm. 145–
50). At measure 151 the cadential 46 veers to the dominant of G minor, which
in turn leads to a restatement of the concluding cadential progression in the
home key (161–71), culminating with a PAC in measure 171 that elides with
the subordinate theme. The final cadence of the main-theme group thus serves
as a transition, and it does so through an elaborate variation of the deflected-
cadence strategy.22

Example 5.8. String Quartet D. 173, mvt. 1: cadential deflection in the
recapitulation, mm. 139–71
(continued)

184 brian black
Later Variations on the Deflected-Cadence Strategy

Schubert’s subsequent use of cadential deflection in his transitions shows an
inventive flexibility that produces a number of interesting variations on the
basic scheme. A telling example can be seen in the Finale of Schubert’s Piano
Sonata in A Major, D. 664 (ex. 5.9).23 In this instance, a weak cadence in the
tonic, consisting simply of a V7–I motion, is initiated halfway through the tran-
sition (mm. 24–25), but is evaded in measure 26 with the repetition of the tran-
sition to that point (mm. 26–31, restating mm. 19–24). This leads to a more
fully developed cadential progression that shifts immediately to the dominant
key (mm. 31–33). Unlike the previous examples, however, the resolution of
the penultimate dominant is left hanging in midair by a measure of silence,
followed by the arrival of the new tonic, displaced down an octave. The effect
of the cadential evasion is quite witty and its delightful rhythmic dislocation is

Example 5.9. Piano Sonata in A Major, D. 664, mvt. 3: transition to beginning of
subordinate theme, mm. 19–40

186 brian black
immediately echoed in the new theme by a sudden stop on the submediant in

measures 36–37.24
In this movement Schubert relies on a subdominant recapitulation to pre-
serve the modulation’s effect. Yet there is no question of a mechanical transpo-
sition here, as is usually assumed in such recapitulations. The strategy is subtly
reworked to express the significance and emotional character of the return to
the home key (ex. 5.10a). In the first cadence of the scheme, the music is sud-
denly diverted to F major by the enharmonic resolution of C♯ as D♭ to the C of
the ensuing cadential 46 (mm. 145–46).25 As a result, the second cadence (mm.
152–54) is now approached through F major rather than D major. This new
key relationship has a decided effect on the character of the second cadence’s
deflection to A major. Rather than creating the exposition’s bright outward
movement by an ascending fifth, as an unaltered subdominant transposition
would, the shift from F to A allows the music to emerge from the flat into the
sharp region as if it were returning home. This feeling of homecoming seems
to be summed up in the focus on A in the melody (mm. 153–54). Example
5.10b presents the transition’s final cadence as it would have unfolded in an
unaltered subdominant transposition to highlight the completely different
character of the shift from D major to A major.
Schubert’s deflected cadences can also have a highly unsettling effect, as
seen in the first movement of the Piano Trio in E-flat Major, D. 929 (see ex.
5.11). Here the transition (mm. 12–48) suggests the nonmodulating type. It
moves very quickly to the dominant (mm. 15–23) but then circles back sequen-
tially through ♭III and ♭VI to an HC in the tonic (mm. 33–35). The half-
cadential gesture is immediately repeated (mm. 36–41), now colored by modal
mixture with the cadential dominant preceded first by ♭III, then by the minor
tonic. When the same passage is repeated yet again, however, the flattened-
third degree becomes the root of V7 within an abrupt PAC in the key of the
minor ♭vi (C-flat minor, spelled as B minor, mm. 44–48).26 This modulation
violently wrenches the music into a completely unexpected and remote tonal
region. What follows (mm. 48–84) suggests a subordinate theme with its new
melodic and rhythmic material. However, the disturbance of the sudden mod-
ulation echoes in the harmonic instability of the whole passage, which passes
sequentially from B minor back through E-flat major to cadence in B-flat major
(m. 84).27 B-flat major is then sustained by a series of more conventional sub-
ordinate themes in that key.
As we have seen, many of the instances of Schubert’s deflected-cadence
strategy are found in the composer’s early works. In fact, Schubert experi-
mented with the strategy most intensively between 1814 and 1820. There are
some later examples, as in the first movement of the Piano Trio in E-flat Major,
D. 929, discussed above, or the Impromptu in C minor, op. 90, no. 1, which
is not a sonata form. A number of late masterpieces, however, elevate caden-
tial deflection from a local means of modulation to a higher formal level. In

Example 5.10a. Piano Sonata D. 664, mvt. 3: recapitulation up to subordinate
theme, mm. 138–58
Example 5.10b. Piano Sonata D. 664, mvt. 3: hypothetical transition from D major
to A major, approach to final cadence

Example 5.11. Piano Trio in E-flat Major, D. 929, mvt. 1: transition to beginning of
the subordinate theme, mm. 12–48
(continued)

(continued)

190 brian black
such cases, the unusual features of Schubert’s transitions that arise from his
deflected-cadence strategy exert a more powerful influence on the form. This
is true especially when such a strategy serves as the controlling framework of
his three-key expositions. Take, for example, the first movement of the Piano
Sonata in B-flat Major, D. 960 (ex. 5.12a). In this instance, the transitional pro-
cess is launched by a cadential deflection at the very end of the main theme,
where a projected PAC in B-flat major is interrupted by the move from the
dominant seventh to its common-tone diminished seventh (mm. 44–45). The
diminished seventh in turn reroutes the music to a PAC in F-sharp minor by
resolving to the latter’s dominant in measure 47 (a situation reminiscent of
that in the E-flat-major Piano Trio). There then follows a long, tonally unsta-
ble section, which circles back from F-sharp minor to B-flat major, where the
arrival and prolongation of the tonic 46 suggests the resumption of the original

Example 5.12a. Piano Sonata in B-flat Major, D. 960, mvt. 1: expansion of

deflected cadence strategy in the exposition: beginning of transitional process,
mm. 43–50
cadence in the home key (ex. 5.12b, mm. 70ff.). Once again the eruption of
the common-tone diminished seventh deflects the cadence (mm. 71–72), but
this time to the cadential 46 of a successful PAC in F major, which ushers in the
subordinate theme in measure 80.28
When taken together, the first interrupted cadence in B-flat major (mm.
44–45) and the final PAC in F major (mm. 73–80) also outline a deflected-
cadence strategy on a larger scale—one that provides the mechanism for the
overarching modulation of the exposition (summarized in ex. 5.12c). Both
cadences are linked by the shared common-tone diminished seventh and
its originating F-major chord.29 The surprise progression from the F-major
dominant to its common-tone diminished seventh, which interrupts the first
cadence, is answered by the common-tone resolution of that same diminished
seventh back to the F-major chord, now serving as the cadential 46 in the last
cadence.30 Thus the intervening tonal region is framed by the two closely
related members of a deflected-cadence strategy.
That Schubert circles back through the home key after a remote shift away
arises from his projection of the deflected-cadence strategy onto a higher
structural level. This tonal configuration, though, does not result from some
deep-seated inhibition, but rather arises from the projection of the deflected-
cadence strategy onto a higher structural level. As in its local incarnation, the
scheme is based on the close relationship of the two cadences involved, spe-
cifically their starting from the same tonality and their use of a shared “hinge”
chord (the diminished seventh), which is consequently heard to function as

192 brian black
Example 5.12b. Piano Sonata D. 960: end of transitional process, mm. 69–80
Example 5.12c. Basic modulation to F major
the plan’s revolving doorway between two different keys. Thus the music must
return through the tonic in order to initiate the second cadence in the same
key as the original cadence. This leads to an interesting departure from the
accepted theoretical view of the key relations within a classical three-key expo-
sition. Many writers, including Rosen and Caplin, consider that the middle key
of this plan lies on the road to the dominant, invariably the last tonality in a
classical thee-key exposition.31 But in the first movement of D. 960, the middle
and final keys, F-sharp minor and F major, respectively, do not follow one upon
the other, but are generated each time through the tonic. In other words,

rather than forming a straight line of successively related keys, Schubert’s plan
describes a sort of circular motion that relates each subsequent tonality directly
back to the tonic.32 This is typical of many of the composer’s three-key exposi-
tions and it is decidedly un-classical.33
Conclusion
Schubert’s development of the deflected-cadence strategy reveals much about
his art, both on a personal level and a broader historical one. As far as his per-
sonal style is concerned, we can perceive in this scheme an acute sensitivity to
the expressive potential of tonal relationships as well as great craftsmanship
in making these relationships such a compelling force in his music. His han-
dling of the scheme never degenerates into formulaic mannerism, but reveals
instead a flexibility and ingenuity that create a wide range of effects suited to
the expressive meaning of each individual work. To this day such modulations
still seem fresh and spontaneous. Yet they are not merely local moments of
inspiration, but carefully worked-out constructions that demand an absolute
control of the composer’s craft as well as an understanding of the psychology
of implication and expectation.34 This is most evident in how Schubert deals
with the problems that the scheme poses in the recapitulation. Here, rather
than what was once considered a mechanistic approach, we find a creative
reworking of the scheme that integrates it into the larger tonal plan. And this
is true of the composer even at an early age.
On the broader historical level, the developments we have seen in Schubert’s
sonata forms provide an excellent illustration of how a significant transforma-
tion in style arises not necessarily from a radical sea change, but rather from
a slight reorientation out of which emerge significant ramifications for the
aesthetics and structural logic of the form. Schubert’s deflected-cadence strat-
egy is an unassuming innovation, yet it has far-reaching consequences. First,
it employs the PAC in a drastically new role. Rather than serving as an ending
gesture to ground or confirm a tonality, the PAC has become an initiating ges-
ture—the very springboard of the modulation. This simple reassignment of
function not only creates some of the most exquisite local effects with regard to
the entrance of the subordinate key, but the modulation’s unusual mechanism
also leads to fundamental changes in the tonal structure of the form, especially
when it operates on a higher level, as in the Piano Sonata in B-flat Major, D.
960. In such cases, direct linear modulation is replaced by a circular movement
dependent upon harmonic and tonal cross-references and the subtle use of
repetition. Taken together then, the novel changes flowing from Schubert’s
deflected-cadence strategy create a form that, although it reproduces many of
the surface elements of the classical style, has already departed significantly
from that style as far as its harmonic workings and its tonal plan are concerned.

194 brian black
Notes
1. Klaus Rönnau was one of the first to propose that Schubert’s unusual preparation
of the subordinate key gives that key its special character and thus becomes quasi-
motivic in nature. See his “Zur Tonarten-Disposition in Schuberts Reprisen,” in
Festschrift Heinz Becker, ed. Jürgen Schläder and Reinhold Quandt (Laaber: Laaber
Verlag, 1982), 439. For a more in-depth discussion of the special character of the
modulations in Schubert’s sonata-form transitions, see Hans Joachim Hinrichsen,
Untersuchungen zur Entwicklung der Sonatenform in der Instrumentalmusik Franz
Schuberts (Tutzing: Schneider, 1994), 51–59; Susan Wollenberg, “Schubert’s
Transitions,” in Schubert Studies, ed. Brian Newbould (Aldershot: Ashgate, 1998),
16–61; as well as her more recent Schubert’s Fingerprints: Studies in the Instrumental
Works (Farnham: Ashgate, 2011), 47–98. Neither Hinrichsen nor Wollenberg
deals with the actual cadential mechanics of Schubert’s transitions discussed in
this chapter.
2. For transition function and the role of the HC and dominant arrival, see William E.
Caplin, Classical Form: A Theory of Formal Functions for the Instrumental Music of Haydn,
Mozart, and Beethoven (New York: Oxford University Press, 1998), 131–35. For sub-
ordinate-theme function and the importance of perfect authentic cadential clo-
sure, see 97–98. In the present discussion I am focusing exclusively on the specific
harmonic role Caplin assigns to these parts of the form. I am not discussing the
broader question of temporality associated with Caplin’s beginning, middle, and
end functions. For the debate between Caplin and James Hepokoski over the latter,
see William E. Caplin, James Hepokoski, and James Webster, Musical Forms, Forms
& Formenlehre: Three Methodological Reflections, ed. Pieter Bergé (Leuven: Leuven
University Press, 2009).
3. This is one of the more controversial tenets of Caplin’s theory. An alternative view
is advanced by James Hepokoski and Warren Darcy, who do accept the PAC as a
possible ending for a transition, a construction they term a “third-level default
option” in their Elements of Sonata Form Theory: Norms, Types, and Deformations in the
Late-Eighteenth-Century Sonata (New York: Oxford University Press, 2006), 27–28.
The essential difference in the two treatments of the transition stems from each
theory’s respective grounding—in melody and rhetoric for Hepokoski and Darcy,
and in harmony and function for Caplin.
5. Ibid., 203.
6. Caplin lists this movement as exhibiting transition/subordinate-theme fusion in its
exposition (ibid., 280n30).
7. The other subgroup simply cadences in the new key without a cadential deflection.
See, for example, the first movement of the Symphony in C Minor, D. 417 (“The
Tragic”), measures 63–67.
8. Transitions involving a tonal feint move to the “wrong” dominant to prepare the
subordinate key, as in the first movement of the Piano Trio in B-flat Major, D. 898,
measures 49–59, where the transition ends on the V of D minor (III) and the sub-
ordinate theme begins on the tonic of the subordinate key, F major. See also the
first movement of the String Quartet in G Major, D.887, measures 58–65, for a simi-
lar plan. For a discussion of the latter case, see Wollenberg, “Poetic Transitions,”
57–61.

9. This is an example of what Janet Schmalfeldt refers to as the “one more time tech-
nique.” See her “Cadential Processes: The Evaded Cadence and the ‘One More
Time’ Technique,” Journal of Musicological Research 12 (1992): 1–51.
10. I would maintain that those Schubert transitions ending with PACs but not involv-
ing a deflected-cadence strategy are also purely transitions. In such cases, it is their
dynamic character, motivic homogeneity, and harmonic focus that express their
intended function. This is a subject, however, for a separate study.
11. The one exception to a clear-cut subordinate theme following upon the deflected-
cadence transition may be found in those three-key expositions in which the initial
subordinate key area is unstable and eventually yields to the second subordinate
key, as in the first movement of the Piano Trio in E-flat Major discussed below.
Here, and in other instances, the transitional process continues, despite the projec-
tion of a new thematic beginning. See also the first movement of the Piano Sonata
in B-flat Major, D. 960, also discussed below.
12. Caplin’s emphasis on dominant harmony as the goal of the classical transi-
tion allows Schubert’s departure from this norm to stand out quite noticeably.
A more inclusive treatment of the classical transition with respect to its possible
cadence types, as in Hepokoski and Darcy’s theory, tends to obscure the novelty of
Schubert’s practice.
13. For one of the most detailed and influential criticisms of Schubert’s penchant for
repeating musical material at all levels of his sonata forms, see Felix Salzer’s “Die
Sonatenform bei Franz Schubert,” Studien zur Musikwissenschaft 15 (1928): 86–125.
Here Salzer argues that repetition is an essential element of Schubert’s lyricism, but
considers it inimical to the very principles of sonata form, which he believes flows
from the dynamic impulse of improvisation with its constant forward motion. Many
of Salzer’s arguments are brilliantly countered by Carl Dahlhaus in his “Sonata Form
in Schubert: The First Movement of the G-Major String Quartet,” in Schubert: Critical
and Analytical Studies, ed. Walter Frisch (Lincoln: University of Nebraska Press, 1986),
1–12. For a discussion of attitudes toward repetition in Schubert’s Finales and the
issues involved, see Thomas Denny, “Too Long? Too Loose? And Too Light? Critical
Thoughts about Schubert’s Mature Finales,” Studies in Music 23 (1988): 25–52.
14. Charles Rosen, Sonata Forms, rev. ed. (New York: Norton, 1988), 236.
15. The famous modulation to G major in the first movement of the “Unfinished”
Symphony in B Minor, D. 759, constitutes an ironic twist with regard to Schubert’s
deliberate avoidance of dominant preparations in these cases. The modulatory
scheme is based on the deflected-cadence strategy, although the two cadences
are not parallel. Here the final emphatic PAC of the main theme (mm. 36–38)
is answered immediately in the winds with a progression that suggests a PAC in G
major (mm. 38–42). (The concluding root of the tonic chord is supplied outside
the winds by the pizzicato G in the basses.) The link between the two cadences is
the sustained D, which indeed is the dominant of the new key, but as the common-
tone thread it is not perceived as a dominant until the last measure before the
cadential arrival in G. Due to its functional ambiguity, it serves to create a myste-
rious floating feeling before the entrance of the subordinate theme, rather than
focusing on and thus preparing the new key.
16. Rosen, Sonata Forms, 357.
17. See in particular the first movements of the String Quartets in G Minor/B-flat Major,
D. 18; in D Major, D. 94; in C Major, D. 32; and in C Major, D. 46. For a discussion

196 brian black
of this development, see Brian Black, “Schubert’s Apprenticeship in Sonata Form:

The Early String Quartets” (PhD diss., McGill University, 1996). Carl Dahlhaus
also deals with the unusual character of the early quartets in “Formprobleme in
Schuberts frühen Streichquartetten,” in Schubert Kongreß Wien 1978: Bericht, ed.
Otto Brusati (Graz: Akademische Druck- und Verlagsanstalt, 1979), 191–97.
18. The common launching point in the bass is C, which supports the pre-dominant
in both cadences. Unlike later schemes of this sort, however, it supports a different
chord in each cadence: A diminished for the G-minor cadence and C minor for the
modulation to B-flat.
19. The length and character of this transition is related to the unusual, block-like
construction of the exposition, in which the main theme, transition, and subor-
dinate theme each occupy a distinct and substantial tonal and motivic space. This
leads to a certain form-functional ambiguity. Webster considers this movement an
early example of the three-key exposition, thus implying that the G-minor area is
a subordinate theme of some sort. Dahlhaus, on the other hand, refers to this sec-
tion as a “transition with the character of a development section.” This is much
closer to the mark: the passage’s rhythmic drive and intensity, its wavering between
the tonic and submediant keys, and its final modulation to the dominant all argue
for its status and function as a transition. See James Webster, “Schubert’s Sonata
Form and Brahms’s First Maturity, I,” 19th-Century Music 2, no. 1 (1978): 26n18; and
Dahlhaus, “Formprobleme,” 196.
20. See Wollenberg, “Schubert’s Transitions,” 33. For Webster’s original comment, see
“Schubert’s Sonata Form,” 30–31.
21. The dominant’s loss of power for Schubert was first proposed by Webster in
“Schubert’s Sonata Form,” 24.
22. To a large extent, the recapitulation recalls that of the first movement of the
Sonatina in G Minor, D. 408, for it reverses the key scheme of the exposition (G
minor to B-flat major), regaining the tonic from B-flat through the PAC in G minor
that was denied in the exposition’s original cadential deflection.
23. The sonata’s date of composition is uncertain and could be either the summer of
1819 or 1825.
24. This feature has been prepared by the complete halt on the leading-tone seventh
chord preceding each cadence (m. 24 and m. 33). It is intensified in the second
cadence where the music seems to hang between the two tonalities, the balance
being tipped by the enharmonic reinterpretation of F♮ as E♯. Such enharmonic
play becomes important in the recapitulation’s transition.
25. The progression involves the conversion of the half-diminished seventh into the
fully diminished seventh of D minor, which essentially resolves by semitone dis-
placement to the dominant seventh of F.
26. Admittedly the cadential progression here consists of merely a V7–I progression in
the new key. But the gesture is heightened rhetorically by the long prolongation of
the V7 and the climactic sweep down to the tonic in the treble of the piano part.
27. Webster thus considers this passage to be part of an expanded transition; see
“Schubert’s Sonata Form,” 29. Hinrichsen, however, accepts it as a subordinate
theme; see Untersuchungen zur Entwicklung, 113–14.
28. The actual deflection of the cadence occurs at the resolution of the B♭ 46 to the
dominant minor on the last beat of measure 73. Richard Cohn discusses the whole
passage in some detail in “As Wonderful as Star Clusters: Instruments for Gazing
at Tonality in Schubert,” 19th-Century Music 22, no. 3 (1999): 213–32. His article,

based on Neo-Riemannian theory, deals with issues different from those addressed
in the present chapter. However, Cohn does recognize a parallel between the
aborted cadence in B-flat major and the final cadence in F. He locates this parallel
in the implied D-minor and A-minor harmonies before the respective cadential 46s.
Other discussions of this famous transitional passage include Charles Fisk, Returning
Cycles: Contexts for the Interpretation of Schubert’s Impromptus and Last Sonatas (Berkeley:
University of California Press, 2001), 242–43; Richard Taruskin, The Oxford History
of Western Music, vol. 3, The Nineteenth Century (New York: Oxford University Press,
2005), 96–100; and Charles Rosen’s seminal discussion in Sonata Forms, 258–61.
29. Rosen draws attention to the return of the diminished-seventh chord at this point
and characterizes the intervening passage as a “magnificent detour”; see Sonata
Forms, 261. My example 5.12c builds on his example (ibid.) to illustrate the control-
ling deflected-cadence strategy of the broad transitional process.
30. My analysis of this particular moment in the form differs in a small, but important,
detail from Cohn’s. Cohn proposes that an implied resolution from a V24 to a i6 in A
minor intervenes before the arrival of the cadential 46 in F major (see “Star Clusters,”
220 and 222, ex. 2.) While this interpretation is logical, it does not account for the
brightening effect of the progression across the measure between measures 73 and
74. This is best viewed, without the A-minor coloring here, as a direct move by com-
mon tone from the diminished seventh on B♮ to the cadential 46 in F major. The
G♯-to-A♮ motion is not a leading-tone resolution at this point, but instead a modal
change on the third of the F chord. The previous move to the cadential 46 of B-flat
(mm. 69–70) has the same bright quality from the very beginning of measure 70,
which similarly argues against an implied D-minor harmony before the arrival of
the B♭46 on the third beat of the measure. Instead, the effect at the beginning of
measure 70 is that of a deceptive resolution of the V24 of D minor to the B♭46 chord.
31. Rosen, Sonata Forms, 246–47; Caplin, Classical Form, 119–21.
32. This observation complements Suzannah Clark’s assertion that in certain minor-
mode sonata forms, Schubert’s tonics can act as an axis around which the keys
are related, rather than a pole. However, in this instance Clark is referring to the
position that the tonic occupies within the boundaries of the form’s tonal space,
rather than to the trajectory of the modulations in a three-key exposition. See her
Analyzing Schubert (Cambridge: Cambridge University Press, 2011), 234–39.
33. For a discussion of Schubert’s circular transitional strategies, in which the tonic is
referenced during the transition to the second subordinate key in a three-key expo-
sition, see Webster, “Schubert’s Sonata Form,” 30.
34. The myth of Schubert as a naive natural genius and its effects on the composer’s
reception history is discussed in detail by Clark in Analyzing Schubert, particularly
in chapter 1, “Singing Schubert’s Praises: The Voice of Vogel in Schubert’s Early
History,” 6–55.

Chapter Six
“Heavenly Length” in
Schubert’s Instrumental Music
François de Médicis
For romantic composers, the creation of large-scale instrumental works pre-

sented a particular challenge: they sought to emulate Beethoven’s powerful
achievements all the while avoiding mere epigonism. Robert Schumann was
well aware of this problem when he undertook his own first symphony, as is
obvious from his review article of Schubert’s Symphony in C Major (“The
Great”), D. 944.1 As is well known, Schumann praises Schubert’s work as a dis-
tinctly successful realization in the grand-scale genre, and considers him to be
an important alternative to Beethoven. The composer-critic points repeatedly
to the sense of monumentality and expansiveness in this work. For instance,
he describes it as a “fat novel in four volumes by Jean Paul, never-ending” and
marvels at its “refreshing sense of inexhaustible wealth.” But a very special
sense of temporality emerges from Schumann’s most often quoted comment
about Schubert’s symphony: the attribution to the work of divine or “heavenly
length.”2 A deeply evocative and penetrating statement about Schubert’s style,
this comment does not refer simply to the music’s temporal span (in the physi-
cal, clockwork sense), and there is of course no suggestion of boredom, as the
qualifier “heavenly” makes clear enough. More profoundly, the description
alludes to a very special psychological experience of temporality, the feeling of
a broadening and stretching of time, phenomenological in nature.
The analytical literature provides us with various attempts to describe the
special sense of time flow in Schubert’s instrumental music in terms of the
organization of musical parameters, such as harmony, motivic organization,
phrase structures, and formal design.3 This chapter pursues this line of inves-
tigation and reexamines parameters already familiar: first, the use of sequence
in sonata form (more specifically the presence of “thematic sequence” in the
development, and of “developmental sequence” in the exposition); and sec-
ond, the use of uniform grouping structures and accentuation patterns in
the exposition’s themes. My contribution lies in a closer examination of these

“heavenly length” in schubert’s instr umental music 199
parameters and of how their interaction weakens the differentiation of the

main sections of the form, thus creating a sense of large-scale coherence that is
distinct from the classics and that helps to project a typically Schubertian expe-
rience of temporality. This study refers to numerous instrumental works by
Schubert but centers on two representative large-scale pieces, the Finale from
the Symphony in C Minor (“Tragic”), D. 417, an early work that neatly encap-
sulates and clearly illustrates the main Schubertian procedures that form the
focus of this chapter; and the Finale from the Symphony in C Major, D. 944,
which exemplifies a mature and masterful use of those features. Schubert’s
unique formal approach will be emphasized through a comparison with
selected classical works.
Thematic and Developmental Sequences

Composed in 1816 when Schubert was only nineteen years old, the C-minor
Symphony is already fully Schubertian. This becomes clear from several dis-
tinctive features of the Finale (see the formal layout in table 6.1). First, it is a
movement of ample proportions, covering nearly five hundred measures, and
lasting about fifteen minutes. Also, the exposition adopts a special tonal plan,
which replaces the classical binary opposition of main and subordinate keys
with a looser chain of three tonal areas, commonly referred to as a three-key
exposition.4 Schubert’s fondness for the use of thematic, closed forms in the
sonata exposition is also illustrated in the main theme, which adopts a ternary
design. As James Webster notes, this typically Schubertian procedure is unusual
in classical works, and would later influence Brahms, among other compos-
ers.5 Moreover, the end of the transition produces an unusual closure, with a
cadence that elides with the beginning of subordinate theme 1. Generally in
classical works (and, not infrequently, also in early romantic ones), the transi-
tion ends with a prolongation of the subordinate key’s dominant, which pre-
pares the arrival of the first subordinate theme in the new tonic. Brian Black
has identified the omission of this very striking formal marker, resulting in a
linkage of the transition and the subordinate theme, as a gesture typical of
Schubert.6 Finally, a comparison between the exposition and the recapitula-
tion shows the characteristic interplay between parallel major and minor
modes. The main theme, originally cast in the minor mode during the exposi-
tion, is brought back in the major for the recapitulation.
These features, specifically Schubertian but in some instances precursors to
common nineteenth-century procedures, alter in no fundamental way the pro-
jection of the movement’s tripartite organization, nor the means to sustain and
project large-scale form. They appear rather as extensions of the classical prac-
tice, and demonstrate the flexibility of sonata form as well as its adaptability to
the new emerging romantic style. But some unusual elements in the internal

200 françois de médicis
Table 6.1. Schubert, Symphony in C Minor, D. 417 (“Tragic”), mvt. 4:

formal layout
Three-key exposition (mm. 1–194)
Intro (mm. 1–5) i

Main theme (mm. 5–63) i ABA′
Transition (mm. 63–85) i → VI
Subordinate theme 1 (mm. 85−129) VI → III sentence (expanded)
Subordinate theme 2 (mm. 129–63) III sentence (expanded)
Closing section (mm. 164–94) III
Development (mm. 195–292) Modulating
Recapitulation (mm. 293–486)
Main theme (mm. 293−351) I–vi ABA′

Transition (mm. 351–72) vi → IV
Subordinate theme 1 (mm. 373–417) IV → I sentence (expanded)
Subordinate theme 2 (mm. 417–51) I sentence (expanded)
Closing section (mm. 451–86) I
organization of the different sections weaken the sense of differentiation

between the main sections of the sonata form, namely, between the exposi-
tion and recapitulation on the one hand, and the development on the other.
To study these special features, we need first to review the formal organi-
zation of the development. My discussion of this formal section is based
on the framework developed by William E. Caplin, which synthesizes the
main dimensions of a development usually encountered in other accounts
of sonata form and describes particularly well the temporal ordering of its
constitutive elements.7 According to this model, the development generally
divides in three sections: a precore, a core, and a standing on the dominant.8
As its name implies, the core is the main section of the development. It com-
prises one or more sequences organized as follows: a model is first intro-
duced, then replicated sequentially, and followed by fragmentation, leading
to new shorter segments that may or may not relate sequentially among
themselves. The expression “precore” offers a neutral designation referring
to whatever materials take place between the end of the exposition and the
core. At the end of the development, a standing on the dominant prolongs
the dominant of the main key and prepares the return of the main theme for
the onset of the recapitulation.
At first glance, the development of Schubert’s Finale conforms perfectly to
this model, and a comparison with a standard development, such as that of the
first movement of Beethoven’s Piano Sonata in C Major, op. 53 (“Waldstein”)

shows no major discrepancies, as we can observe by comparing figures 6.1a

and 6.1b. Both developments follow the normative tripartite design outlined
above. In both works, the core is subdivided into two parts, labeled core 1 and
core 2 in figure 6.1. Schubert’s symphony features a standing on dominant
after each core, the first one in the key of the subdominant, the second in the
home key. In Beethoven, only the second core is followed by a standing on
the dominant (in the main key). Three of the four cores produce fragmenta-
tion. In Beethoven’s core 1, the four-measure segments of the initial sequence
give way to a new two-measure sequence; in core 2, the fragmentation of four-
measure segments produces nonsequential two-measure units. In Schubert,
only core 1 produces fragmentation; the eight-measure segments of the initial
sequence are followed by two-measure sequential segments.
There is, however, a striking difference between the two developments that
is not apparent from figure 6.1. It resides in the internal organization of the
sequences, shown in examples 6.1a and 6.1b. Schubert’s core 1 features what
I will call a “thematic sequence,” as opposed to the “nonthematic sequence”
illustrated by Beethoven’s core 1. As its name implies, the expression “thematic
sequence” refers to a sequence that, like the one found in Schubert’s core 1, is
based on the transposition of a thematic structure. In example 6.1a, the model
from measures 223–30 appears as a theme most obviously because it adopts a
conventional form: it corresponds to the archetypical eight-measure period, as
codified by Schoenberg and further discussed by Erwin Ratz and by Caplin.9
It consists of two four-measure phrases, an antecedent and a consequent,
that lead to an imperfect authentic cadence (IAC) and to a perfect authentic
cadence (PAC), respectively.
Pre-core (mm. 195222)
Core 1 (mm. 223–54):
model (8) seq. (8) seq. (8) frag.
mod. (2) seq. (2) seq. (2) seq. (2) link (2)
Standing on the dominant (mm. 257–64) V/iv
Core 2 (mm. 265–76):
model (4) seq. (4) seq. (4) link (4)
Standing on the dominant (mm. 281–93) V
Figure 6.1a. Schubert, Symphony in C Minor, D. 417, mvt. 4: overview of the

development, mm. 195−292

Pre-core (mm. 86–95)
Core 1 (mm. 96111)
model (4) seq. (4) frag.
new model (2) seq. (2) seq. (2) frag.
(1) (1)
Core 2 (mm. 112–35)
model (4) seq. (4) seq. (4) frag.
(2) (2) (2) (2) (2) frag
(1) (1)
Standing on the dominant (mm. 281–93) V
Figure 6.1b. Beethoven, Piano Sonata in C Major, op. 53 (“Waldstein”), mvt. 1:

overview of the development, mm. 86−155
At the same time, a passage need not adopt a conventional formal pattern
to convey a sense of thematic structure; indeed, sonata themes occasionally
present unique, unconventional designs. Hence, the model in Schubert’s core
1 also projects a thematic character in a broader sense, because it starts with
characteristic material and an individualized and well-defined profile that first
clearly establishes the tonic and subsequently leads to a cadence. The distinc-
tive lyrical tone of the segment, which contributes to its thematic nature, con-
trasts with the turbulent, Sturm und Drang character often associated with the
onset of the first core. Moreover, a special effect of transfiguration is created in
the particular way in which the sequence is carried out. The theme concludes
in its opening key and is transposed as a block for the next segment, a major
third down. This generates a chromatic (common-tone) modulation, in which
a melodic link is provided at the juncture of every segment pair. For instance,
the pitch A both concludes the initial statement of the model in measure 230,
as 1^ in A major, and initiates the transposition of the next segment, as 3 ^
in
F major.10 (The second core of Schubert’s symphony features more conven-
tional developmental characteristics: it is based on a nonthematic model, and
expresses a Sturm und Drang character.)
In contrast to the unusual thematic sequence found in Schubert’s first core,
Beethoven’s core 1 illustrates the more normative nonthematic sequence (ex.
6.1b). If the latter differs sharply from the former, it is nonetheless difficult
to describe its essential features in positive terms. It uses more fleeting mate-
rial that is of a flowing, transitory nature. We can safely say that the four-mea-
sure segment starting at measure 96 does not adopt a conventional thematic

Example 6.1a. Schubert, Symphony in C minor, D. 417, mvt. 4, mm. 223−38

Example 6.1b. Beethoven, Piano Sonata in C Major, op. 53 (“Waldstein”), mvt. 1,

mm. 96−103
pattern, but this does not take us very far: I have already mentioned that not all
classical themes adopt a stereotypical design. In this case, however, the segment
lacks certain essential traits of thematic organization, whether it is conven-
tional or not. Only four measures long, it is too brief and brings no cadential
closure. The passage is fleeting and the x and y motives are too redundant
to allow for formal differentiation—features common in sequences, but not
in themes. The section’s transitory character is apparent in the shape of its
bass line: as can be seen from the boxed notes on the score, the bass’s gradual
ascending motion (interrupted by a voice exchange at the beginning of m. 97,
but recaptured afterward) is conceived so as to continue until the onset of the
subsequent sequence at the upper fourth, and seamlessly merge with it. Finally,
in contrast to the lyrical tone of Schubert’s core, Beethoven’s conveys an agi-
tated, Sturm und Drang character.
The use of thematic sequence in Schubert’s symphony has an important
impact on the characterization of formal functions: on a local level, it produces
a lyricization of the development; on a larger scale, it undermines the contrast
between the main sections of the form as a whole, to the extent that the devel-
opment integrates the kinds of thematic structures that are usually restricted to
the exposition and the recapitulation. Moreover, Schubert’s symphony further
weakens the differentiation of the sonata form’s tripartite organization because

the second subordinate theme in the exposition features a long, modulating

sequence that readily evokes a developmental core. This theme adopts a sen-
tence-like design, a form that theoretically can feature short sequences. Here,
however, the sequences are developed to such an extent that their weight and
developmental character is comparable to those found in the second core of
the development: in both cases, the sequential segments are four measures
long, and both sequences express a Sturm und Drang quality.
The use of thematic sequences in the development and of core-like sequences
in the exposition is not specific to this symphony but rather constitutes a fre-
quent procedure in Schubert’s instrumental music. To cite only examples from
well-known instrumental works, thematic sequences can be found in the devel-
opments of the first movements of the Piano Quintet in A Major, D. 667 (“Die
Forelle,”), the String Quintet in C Major, D. 956, and the Piano Trio in B-flat
Major, D. 898, as well as in the fourth movement of the Symphony in C Major,
D. 944.11 Similarly, core-like sequences can be found in the first-movement
expositions of the String Quartet in A Minor, D. 804 (“Rosamunde”), the String
Quartet in D Minor, D. 810 (“Der Tod und das Mädchen”), the String Quartet in
G Major, D. 887, and the Piano Sonata in A Major, D. 959.
The weakening of contrast between thematic and developmental sections
is especially prominent in Schubert’s late works, including the Finale of the
C-major Symphony. In this movement, the development section presents a
thematic sequence, and although the subordinate theme does not feature any
developmental sequence, it contains a close equivalent in the form of a devel-
opmental digression. This lengthy passage (mm. 257–337) avoids sequential
activity, but develops previous motives in a very loose, nonthematic structure
and in an unstable, chromatic setting. In the development, two sections (mm.
385–466 and 467–559) successively work out materials from the closing theme
(mm. 337–53) and from the subordinate theme.12 The first section corre-
sponds to core 1 (see fig. 6.2) and features a thematic sequence that starts with
a lyrical tone and produces a steady increase in tension. The second section
combines sequential writing characteristic of a second core with a dominant
prolongation suggesting a standing on the dominant, prolonging first V of V
(mm. 449–511), and later on, after a brief link (mm. 512–15), the proper dom-
inant of the main key (mm. 515–98).13 As shown in example 6.2, core 1 starts
with a long model organized as a modulating hybrid that generates a modula-
tion a fifth down each time it is replicated. It comprises a twelve-measure ante-
cedent and a thirteen-measure continuation,14 whose last measure elides with
the onset of the sequence.15 Fragmentation subsequently produces eight-mea-
sure segments that adopt the form of a modulating antecedent (it divides into
two four-measure units that are derived, respectively, from the first and the last
four measures of the antecedent initiating core 1), generating a transposition
up a minor third. Core 2 presents the conventional nonthematic sequence,
associated with the usual Sturm und Drang character.

Exposition (1−380)
Main theme (1−106)

Transition (107–64)
Subordinate theme (165−256)
Developmental digression (257−337)
Closing section (337–80)
Development (381–598)
Link (381-84)
Core 1 (385-448)
Standing on V/V (449–511):
S.o.V/V (449–66) Core 2 (467–89) Core 2 var. (489–511)
Model (8) seq. (8) frag. (3) (2) (2)
Link (512−15)
Standing on V (515–98)—pedal of G, end deflected toward ♭III6 (591–98)
Core 2 (515−37) Core 2 var. (537–59)
Model (8) seq. (8) frag.
(3) (2) (2)
Figure 6.2. Schubert, Symphony in C Major, D. 944 (“The Great”), mvt. 1: overview
of exposition and development
Schubert’s tendency to indulge in discursive development in the subordi-

nate theme’s continuation has not gone unnoticed in the scholarly literature,
and Martin Chusid’s interesting observations on the matter are worth citing
here. Chusid relates the developmental portions of the subordinate theme
to Czerny’s theorization of sonata form.16 According to Czerny, the subordi-
nate theme (or middle subject to follow the theorist’s own terminology) starts
with “a new and more beautiful and pleasing melody than all which precedes.”
This phrase is followed by an amplification (“a new continuation . . . which
generally consists either of moving or brilliant figures, and is terminated by a
cadence”).17 Chusid relates this last quotation to the developmental portions
of Schubert’s subordinate themes, but a closer look at Czerny’s text suggests
significant differences.
Czerny’s theoretical description of the intensified continuation in the sub-
ordinate-theme group is rather general and its practical applications seem
unclear. Nonetheless, possibilities emerge when one compares it to a similar
notion discussed by Adolf Bernhard Marx and extrapolates from the musi-
cal illustrations provided by the two authors. According to Marx, the struc-
ture of the exposition comprises a “main Satz” (Hauptsatz), a “subsidiary Satz”
(Seitensatz), a “Gang,” and a “closing Satz” (Schlusssatz). Satz and Gang refer to
closed thematic utterance and open-ended transitional passage, respectively.18

Example 6.2. Symphony in C major, D. 944, mvt. 4, mm. 385-414
(continued)

Czerny’s and Marx’s theories each emphasize developmental traits in the sec-
ond part of the subordinate-theme group, albeit different ones: Czerny refers
to musical activity and intensity, Marx to structural instability and openess.
The musical illustrations Marx and Czerny provide help to interpret their
ideas more precisely. The examples, all drawn from piano solo and four-hand
literature, illustrate a procedure that helps to mark the end of the subordinate-
theme group and was first developed in the classical concerto genre but later
frequently adopted in fast sonata forms in general. It consists of an intensifica-
tion occuring toward the end of the subordinate-theme group, which often

features an expanded cadential progression leading to a climax.19 The first

movements of Mozart’s Sonata for Piano Four-Hands in D Major, K. 123a/381
quoted by Czerny, and of Beethoven’s Piano Sonatas in F Minor, op. 2, no. 1
and in E-flat Major, op. 31, no. 3 analyzed by Marx all feature some form of
intensification, for example through rhythmic acceleration, melodic progres-
sion to a higher register, or dynamic increase—an intensification that culmi-
nates in an expanded cadential progression.20 In two instances (Mozart’s K.
123a/381 and Beethoven’s opus 31, no. 3), the parallel with the concerto is
even reinforced by the trill appearing at the climax of the progression—a ges-
ture that typically marks the end of the solo and the imminent entry of the
orchestral ritornello in the concerto genre.
The procedures described by Czerny and Marx and exemplified by Mozart
and Beethoven are quite different from Schubert’s intensifications discussed
above, but they can nonetheless help clarify what Schubert does. Whereas the
intensification in Mozart’s and Beethoven’s sonatas occurs within looser but
still thematic structures, in Schubert’s works it appears in developmental sec-
tions characterized by sequential writing and Sturm und Drang character, or at
least in unthematic digressions (with a more or less pronounced developmen-
tal character, depending on how much the intensification elaborates on previ-
ous motives, features chromatic harmonic writing, and modulates to remote
tonal areas). What are we to conclude from all of this? Schubert might have
been trying to imitate a gesture found in classical sonata form, that of an inten-
sification of the discourse at the end of the exposition—a model that he emu-
lated, in his own way, with new means, deliberately or not.
Phraseology
In addition to the special use of sequential and thematic structures, another
element affects the differentiation of main sections in Schubert’s practice of
sonata form, a parameter I call “phraseology,” which combines traits of group-
ing structure and accentuation patterns.21 Grouping structure is hardly a new
concern in formal analysis, and the study of phraseology has some precedents
on which I will comment later. Nonetheless, its great potential for highlight-
ing form-functional characterization has not yet been fully exploited. Example
6.3a shows an analysis of the grouping structure in core 1 of the first move-
ment of Schubert’s C-minor Symphony according to Lerdahl and Jackendoff’s
grouping preference rules.22 The structure shows one striking feature: the two-
measure level tends to repeat and perpetuate itself with remarkable uniformity
and perseverance. This is partly due to the obsessive prevalence of the initial
motive, labeled “x.” All four-measure groupings begin with this motive, and
from the fourth system onward (i.e., after m. 247), motive x appears at every
measure. Since the motive starts with a durational accent (due to the longer

value of the half note tied to eighth note in comparison with its surround-
ings), it creates a great uniformity of accentuation as well, resulting in a quasi-
hypnotic effect.
Given the passage’s formal context, these features are hardly surprising:
they can be expected in a core, because the development of one or more
motives through sequencing and fragmentation necessarily produces great
motivic redundancy, and this in turn will produce strong uniformity of surface
groupings and accentuation. This uniformity is also a desired characteristic in
a development section, for it creates unrest, ongoing motion. What is more
peculiar is to find the same type of phraseology in a thematic section, as is
the case in the main theme of the C-minor Symphony (shown in ex. 6.3b).
Comparing example 6.3b with example 6.3a, we recognize the same general
features: the prevalence of two-measure groupings, often associated with
a recurring motive, labeled “y” here. This motive initiates all four-measure
groups, and for some substantial parts, two-measure groups as well (see mm.
5–17, 21–29, and 45–55). Furthermore, motive y is closely related to x : it has a
different melodic shape but adopts the same rhythm, consequently producing
the same accentuation pattern.
Such similarity in phraseology between thematic and developmental sec-
tions is found throughout Schubert’s output, and is also illustrated in the
Finale of the C-major Symphony (I will return to this in the last part of my
chapter). This uniformity is unusual when measured against classical practice,
which tends to use different phraseologies to contrast thematic and develop-
mental sections. The comparison of themes from piano sonatas by Mozart
and Schubert in examples 6.4a and 6.4b illustrates the difference. The over-
all grouping structure is similar in both cases, with eight-measure units that
divide into four-measure segments, and four-measure segments that subdivide
roughly in turn into two-measure ones. The differences in phraseology reside
in significant details of grouping structure and accentuation pattern.
In Schubert, a one-measure rhythmic cell x and its variant x‫ ׳‬produce a
uniform one-measure surface grouping that generates the whole theme. In
contrast, Mozart’s motivic organization is more varied, and slight asymmetries
occur at the two-measure level. The opening two-measure group is really two
and a half measures long because it includes an upbeat, while the next group
is truly one measure and a half, because it starts at the beginning of measure 3
and reaches only as far as the middle of measure 4.
The motivic organization influences the accentuation in different ways.
In Schubert, the shape of motive x creates a durational accent on the first
beat, and its repeated use reproduces the accent at the beginning of each
measure. In contrast, the variety of motivic material in Mozart favors a non-
uniform accentuation. The theme’s quasi-continuous eighth-note rhythm gen-
erates durational accents when the motion stops at measures 1, 4, 5, and 8.
Consequently, a displacement of the accentuation affects the grouping at the

Example 6.3a. Symphony in C minor, D. 417, mvt. 4, mm. 223-57: core 1

Example 6.3b. Symphony in C minor, D. 417, mvt. 4, mm. 5-63: main theme
(continued)

Example 6.3b.—(continued)
(continued)

Example 6.3b.—(continued)
(continued)

Example 6.4a. Piano Sonata in D major, D. 850, mvt. 2, mm. 1-8
two-measure level: in the first and third groups, the stress is placed on the first
measure, and in the second and fourth groups, on the second measure.
The tendency of early romantic composers such as Schubert or Schumann
to use uniform, redundant grouping structures has been discussed by numer-
ous theorists. Edward T. Cone has famously referred to this phenomenon by
the terms hypermeasure or Viertaktigkeit, or by the expression “four-bar phrase
tyranny”; this last expression is also found in William Rothstein, while Rosen
and others use Vierhebigkeit.23 Previous authors, however, described this type

Example 6.4b. Mozart, Piano Sonata in D major, K. 284, mvt. 1, mm. 1-8
of phraseology exclusively with respect to thematic sections, and failed

to recognize its common use in development sections (even in classical
works); furthermore, they overlooked the distinctive feature that sets early
romantic composers apart, namely, the attenuation of phraseological con-
trast among sections.
Aesthetic Issues
Let us review the different techniques discussed in this chapter to evaluate
their combined impact on the perception of the overall form, and discuss
briefly the aesthetic appeal of one controversial technique. All these pro-
cedures—the thematic sequence used in the development, the core-like
sequence or digression in the exposition, and the uniform, developmental
phraseology in the exposition—affect one’s experience of overall form in
that they undermine the contrast between the development section and the
outer thematic sections. But at the same time, their effect is not so strong as
to destroy every sense of large-scale organization and lead to chaos and anar-
chy. I believe that the weakening of formal contrast resulting from these pro-
cedures, combined with the expansion of form, helps convey a very special
experience of time felt in many of Schubert’s works—a temporal experience
that might have inspired Schumann’s evocation of “heavenly length.” These

procedures typify Schubert’s style and occur throughout his instrumen-

tal oeuvre, including such prominent late works as the C-major Symphony.
Although some of these techniques have sometimes been denigrated in the
literature, it appears that Schubert’s style developed over the years not by
stripping these idiosyncratic features from his music, but by using them in
aesthetically more convincing ways.
One of these parameters has elicited particularly negative comments that
need to be addressed at some length, partly because they clash with the very
conclusions of this study. An early derogatory comment concerning uniform
phraseology comes from Schumann—of all people—and seems to question
whether this parameter could even contribute to the powerful experience of
“heavenly length.” Indeed, while Schumann had high praise for the C-major
Symphony, he was more critical of Schubert’s last three piano sonatas (D. 958
in C Minor, D. 959 in A Major, and D. 960 in B-flat Major).24 In these pieces,
the composer-critic denounces a special way of developing ideas, the descrip-
tion of which seems to match the notion of uniform phraseology. For him, the
expansiveness of these works results from “the spinning out of certain general
musical ideas instead of adding new threads to them from phrase to phrase. . . .
It is as though there could be no ending, nor any embarrassment about what
should come next. Even musically and melodically it ripples along from page
to page, interrupted here and there by single more abrupt impulses—which
quickly subside.”25 Such a negative account of Schubert’s typical uniformity of
phraseology seems to rule out this parameter as a legitimate component of the
divine temporal stretching, or “heavenly length,” that Schumann reserves for
the C-major Symphony. But on the other hand, that symphony also indulges
immoderately in uniform phraseology, while, surprisingly, this does not seem
to bother Schumann.26 To be sure, the redundancy did not escape Donald
Francis Tovey, for whom the subordinate theme arises “so inevitably and so
astonishingly out of the four premonitory repeated notes of the horn, and
stretching itself ad infinitum while the violins madly turn somersaults with the
persistent [triplet figure first introduced in m. 2].” Tovey reports that when
Mendelssohn rehearsed the Finale with the London Philharmonic, the begin-
ning of that passage “caused the players to giggle and behave so badly that he
had to withdraw the work.”27
One could infer that, for some reason, Schumann did not notice the strong
motivic redundancy in the symphony. But it could also be argued that for him,
it did not impair the aesthetic result of this piece. After all, we witnessed a simi-
lar attitude with Tovey, although with regard to a different procedure. For the
British music critic, the use of developmental sections in the exposition is dam-
aging to the balance of the piece in principle, but it is excused when employed
in a convincing way (as in the “Unfinished” Symphony). Schumann’s condem-
nation of uniform phraseology in Schubert can probably be explained par-
tially by the influence of the romantic cult of originality. Schoenberg would

later criticize the procedure on the grounds of modernist valuation of con-

stant renewal in his essay “Brahms the Progressive,” where he proposes a prin-
ciple of historical development based on the concept of “musical prose” that
involves grouping structures and motivic development. Schoenberg defines
musical prose as “a direct and straightforward presentation of ideas, without
any patchwork, without mere padding and empty repetitions.” His historical
survey from Bach up to himself (!) praises above all composers who avoid regu-
lar grouping structures and motivic redundancy, like Brahms and especially
Mozart. He scorns lesser composers who indulge in sterile symmetry and rep-
etition, including the eighteenth-century practitioners of the galant style such
as Keiser, Telemann, and Mattheson, who “live only because of the musicolo-
gists’ interest in dead, decayed matter.”28 In order to rescue important canoni-
cal composers that remained aloof from the development of tighter and more
concise presentation of ideas, like Schubert, Verdi, and Johann Strauss, or to
salvage works like Die Zauberflöte, Schoenberg is obliged to revert to criteria
different from musical prose: the “power of expression,” which legitimizes the
“real, lasting popularity” of these composers.29
The sensitive writings of Cone and Rosen have paved the way for more
nuanced reevaluations of this Schubertian trait in the past decades.30 Still,
direct assessments of the positive effect of this procedure remain scarce. A
rare exception is Alfred Brendel’s insightful commentary on the three last
piano sonatas.31 In this text, Brendel takes issue with Schumann’s judgment
of these sonatas, considering the insistent repetition of motives as deliberate
and successful in its effect. He maintains that “Schubert’s ‘length’ only appears
obsessive where the music is intended to express an obsessive state of mind.”32
Repetitiveness contributes to the character of a work, a quality that Brendel lik-
ens in the C-minor Sonata to “demons descend[ing] to strangle or mercilessly
to chase,” concluding that “it is precisely the obsessional quality of the C-minor
Finale which makes this movement convincing—in a good performance.”33
Here, as we have witnessed with Schoenberg, Tovey, and Cone, analytical obser-
vations provide arguments for aesthetic evaluation, even though (or precisely
because) theoretical and aesthetic discourses evolve on parallel, independent
planes. And if Schumann was receptive to the heavenly dimension in Schubert,
for all his awareness of the repetitive nature of the late piano sonatas, he was
oblivious to their dark side, a dimension long obliterated that has only begun
to be acknowledged recently.
Notes
1. “Die C-dur Sinfonie von Franz Schubert,” published initially in the Neue Zeitschrift
für Musik in 1839, reprinted in Robert Schumann, Gesammelte Schriften über Musik
und Musiker von Robert Schumann, ed. Martin Kreisig (Leipzig, 1914), 459–64.

2. Schumann, Gesammelte Schriften, 463. It is interesting to note that a similar reference

to a stretching of temporality is found in a text written by Schubert himself, the
famous Mein Traum: “And I felt as though eternal bliss were gathered together into
a single moment.” Reproduced in Otto Erich Deutsch, The Schubert Reader: A Life of
Franz Schubert in Letters and Documents, trans. Eric Blom (New York: Norton, 1947),
298–99.
3. Susan Wollenberg analyzes instrumental works that include long sections of loose
materials with abundant repetitions and uniform rhythms, and likens the proce-
dures Schubert uses there to his lieder writing. She further studies both the inter-
nal organization of these sections as well as the expressive relationships of the
various sections, and impact of cyclical recalls across the sections. See her Schubert’s
Fingerprints: Studies in the Instrumental Works (Aldershot: Ashgate, 2011), esp. chap.
9, “Heavenly Length,” 245–86. Ryan McClelland calls attention to hypermetrical
irregularities and ambiguities in a Schubert Klavierstück. See his “Hypermeter,
Phrase Length, and Temporal Disjuncture in Schubert’s Klavierstück No. 3 (D.
946),” in The Unknown Schubert, ed. Barbara M. Reul and Lorraine Byrne Bodley
(Aldershot: Ashgate, 2008), 157–75. Su Yin Mak shows how Schubert’s late instru-
mental style may be viewed as an expression of Friedrich Schiller’s concept of the
elegiac, grounding her analytical observations in Schenkerian reduction, ryth-
mic structure, musical topics, and formal markers. See her “Schubert as Schiller’s
Sentimental Poet,” Eighteenth-Century Music 4, no. 2 (2007): 251–63. According to
Barbara R. Barry, Schubert’s instrumental music projects a special sense of form
(digressive and nonteleological, with deflections to remote keys and sections of
inward reflection and lyricism). This the author contrasts to Beethoven and inter-
prets as a transposition of expressive techniques and gestures borrowed from lie-
der writing. See “The Matrix Revisited: A Reconsideration of Schubert’s Sonata
Form Movements,” Journal of Music and Meaning 6 (2008), n.p. Older studies by
Carl Dahlhaus, Charles Rosen, Alfred Brendel, and John Daverio that continue to
inform more recent literature will be cited below.
4. On this Schubertian procedure see David W. Beach, “Harmony and Linear
Progression in Schubert’s Music,” Journal of Music Theory 38 (1994): 1–20; and
James Webster, “Schubert’s Sonata Form and Brahms’s First Maturity,” 19th-Century
Music 2 (1978): 18–35.
5. Webster, “Schubert’s Sonata Form,” 19. See also James Webster, “Schubert’s Sonata
Form and Brahms’s First Maturity (II),” 19th-Century Music 3 (1979): 18–35.
6. Brian Black, “Schubert’s Apprenticeship in Sonata Form: The Early String
Quartets” (PhD diss., McGill University, 1996), 254–57; see also his contribution to
the present volume.
139–59. This model was developed for the classical repertoire, but it proves to be
useful for most sonata forms from the first half of the nineteenth century.
8. In Caplin, a standing on the dominant generally appears at the end of the retransi-
tion (Classical Form, 157).
9. Arnold Schoenberg, Fundamentals of Musical Composition, ed. Gerald Strang with the
collaboration of Leonard Stein (London: Faber and Faber, 1967), 20–24; Erwin
Ratz, Einführung in die musikalische Formenlehre: über Formprinzipien in den Inventionen
J. S. Bachs und ihre Bedeutung für die Kompositionstechnik Beethovens, 3rd ed. (Vienna:
Universal Edition, 1973), 23–28; Caplin, Classical Form, 9–12; 35–48.

10. The motion from I to ♭VI, which results from the modulation a major third down,
is related to important motivic relations that pervade the whole symphony. Already
the key plan of the different movements shows a departure from the tonic C minor
in the first movement toward the key of the submediant A-flat in the slow move-
ment, followed by a return to the tonic in the last movement through the relative
in the Minuet. The key relationships within individual movements or sections also
feature prominent modulation to or tonicization of the submediant: in the first
Allegro, the exposition goes from I to ♭VI, and the subordinate theme features a
symmetrical division of the octave in descending major thirds; the contrasting epi-
sode of the slow movement starts in the key of VI (m. 53); the Trio of the Scherzo
modulates to ♭VI at the beginning of B; and the three-key exposition of the Finale
goes from I to ♭VI to III. Finally, the move to the submediant also appears in
numerous surface details: the C–A♭ leap at the beginning of the slow introduction
to the first movement, the gradual melodic ascent from C to A♭ in the main theme
from the first movement, and so forth.
11. One finds a somewhat similar procedure in Beethoven’s Symphony no. 3 (Eroica).
The development of the first movement is famous for introducing what is com-
monly called a new theme. From a formal standpoint, things are a bit more com-
plex (which makes them also more interesting): the exact nature of the material
is ambiguous and suggests alternatively thematic and core-like structure. It first
sounds as a theme based on new material, which in itself is not that unusual in a
development (as noted by Rosen in Sonata Forms, 2; familiar examples of this proce-
dure occur at the beginning of the developments of the first movement of Mozart’s
Piano Sonata in F Major, K. 332, or in the Finale of Beethoven’s Piano Sonata in F
Minor, op. 2, no. 1). What is more special is the four-measure dominant prepara-
tion that precedes it and calls attention to it (in a similar way, standings on the
dominant are used to introduce important themes before the beginning of the sub-
ordinate theme group or the recapitulation). But the “quasi” theme evades caden-
tial closure, and the beginning of its transposed recurrence suggests sequential
treatment typical of a developmental core. In the end, Schubert’s and Beethoven’s
procedures present some similarity in their mix of thematic and sequential fea-
tures, but the former’s straightforwardness contrasts with the latter’s ambivalence.
12. The closing theme itself derives from a cadential idea in the subordinate theme
(mm. 249–53).
13. The combination of a twelve-measure antecedent and a thirteen-measure conse-
quent adds up to a period of twenty-five measures. Each period creates a one-mea-
sure elision with the beginning of the next segment, and since this is difficult to
represent on a chart without contradicting the overall proportions, for the sake of
convenience, figure 6.2 shows the length of the period’s grouping as nonoverlap-
ping segments of twenty-four measures.
14. The thirteen-measure segment actually features some structural ambiguity because
it begins with a restatement of the basic idea of the previous twelve-measure seg-
ment, and its structure could be interpreted as a consequent phrase. The con-
tinuation function is suggested by the faster pace in harmonic rhythm and the
fragmentation resulting from a series of melodic imitations entering at a two-mea-
sure distance.
15. In spite of the final prolongation of the conventional dominant of the main key, an
unexpected twist brings back the beginning of the main theme in the key of E-flat
major, or ♭III. The tonic harmony of the home key is not established for the entire

main theme of the recapitulation, and its confirmation is postponed until the
first cadence of the subordinate theme (m. 785). This results in a form of elision
between the end of development and beginning of recapitulation, a fairly common
procedure in Schubert’s instrumental music, discussed in Daniel Coren “Ambiguity
in Schubert’s Recapitulations,” Musical Quarterly 60 (1974): 568–82.
16. Martin Chusid, ed., Franz Schubert, Symphony in B Minor (“Unfinished”): An
Authoritative Score (New York: Norton, 1971), 76.
17. Carl Czerny, School of Practical Composition (1848[?]; repr., New York: Da Capo
Press, 1979), 35. The reference to the original German text is Schule der praktischen
Tonsetzkunst (Bonn: Simrock, ?1849–50).
18. Adolf Bernhard Marx, Musical Form in the Age of Beethoven: Selected Writings on Theory
and Method, ed. and trans. Scott Burnham (Cambridge: Cambridge University
Press, 1997), 91.
20. The same holds true for the pastiche that Czerny composed after Mozart’s model.
See Carl Czerny, School of Practical Composition, 37–40 for the Mozart musical illus-
tration, 41–42 for the theoretical discussion, and 43–46 for Czerny’s pastiche. Cf.
Marx, Musical Form in the Age of Beethoven, 134–36. For the specific passages contain-
ing intensification, see Mozart’s K. 123a/381, measures 18–21 (and possibly up to
the beginning of m. 27); Beethoven’s opus 2, no. 1, measures 26–41, and opus 31,
no. 3, measures 64–82.
21. See François de Médicis, “La spécificité des thèmes à retour dans l’œuvre instru-
mentale de Brahms” (PhD diss., McGill University, 1998).
22. Fred Lerdahl and Ray Jackendoff, A Generative Theory of Tonal Music (Cambridge,
MA: MIT Press 1983), 43–67.
23. Edward T. Cone, Musical Form and Musical Performance (New York: Norton, 1968),
79. At first sight, Cone’s definition of hypermeasure in terms of regular grouping of
measures seems more general than my notion of phraseology. But his commentary
adds other dimensions that make both look very similar: “[Hypermeasure] is espe-
cially likely to occur whenever several measures in succession exhibit similarity of
motivic, harmonic, and rhythmic construction.” See also William Rothstein, Phrase
Rhythm in Tonal Music (New York: Schirmer, 1989); Charles Rosen, “Ritmi di tre bat-
tute in Schubert’s Sonata in C Minor,” in Convention in Eighteenth- and Nineteenth-
Century Music: Essays in Honor of Leonard G. Ratner, ed. Wye J. Allanbrook, Janet M.
Levy, and William P. Mahrt (Stuyvesant, NY: Pendragon Press, 1992), 131–221; Ira
Braus, “An Unwritten Metrical Modulation in Brahms’s Intermezzo in E Minor, Op.
119, No. 2,” in Brahms Studies, ed. David Brodbeck (Lincoln: University of Nebraska
Press, 1994), 1:161–70. This notion goes back to Wagner’s idea of “Quadratur
des Tonsatzes” (or “period” in English). See, for example, his “Zukunftsmusik,”
in Richard Wagner, Judaism in Music and Other Essays, trans. William Ashton Ellis
(Lincoln: Nebraska University Press, 1995), 316–17.
24. See Schumann, “Franz Schuberts letzte Kompositionen,” Gesammelte Schriften,
327–31.
25. Robert Schumann, Schumann on Music: A Selection from the Writings, trans. and ed.
Henry Pleasants (New York: Dover, 1988), 143.
26. Schumann himself is known to have adopted a highly uniform phraseology. See
Cone, Musical Form, 79, and Rosen, Sonata Forms, 393. The latter mentions as exam-
ples Schumann’s Piano Sonata no. 1 in F-sharp Minor, op. 11, and some of the
symphonies (which he does not identify).

27. Donald Francis Tovey, Essays in Musical Analysis (London: Oxford University Press,
1948), 1:210. Brendel tells also that when Mendelssohn “conducted the C-major
Symphony in Leipzig, he felt obliged to make cuts” (“Schubert’s Last Sonatas,”
79)—possibly to avoid too much repetition.
28. Arnold Schoenberg, “Brahms the Progressive,” in Style and Idea (Berkeley:
University of California Press, 1984), 409, 415.
29. Ibid., 415.
30. Cone, Musical Form, 79–82; Rosen, “Ritmi di tre battute” and Sonata Forms, 392–93. In
spite of their more favorable discussion of the procedure, these authors are quick
to justify successful use of uniform phraseology by the interference of unsettling
means, such as metrical ambiguity or hypermetrical irregularity.
31. Brendel, “Schubert’s Last Sonatas.”
32. Ibid., 84. This quote refers to Schubert’s sketching methods, which show that
Schubert was very self-critical, and that his “obsessiveness” was intentional.
33. Ibid., 125.

Part Four
Text, Texture, and Form

Chapter Seven
Sentences in the Lieder

of Robert Schumann
The Relation to the Text
Harald Krebs
My chapter extends William E. Caplin’s theory of formal functions in two

ways: it applies the theory to vocal rather than instrumental works, and to
romantic rather than classical music.1 Neither of these extensions is new.
There have been several earlier explorations of the applicability of the theory
to vocal works. Matthew BaileyShea has studied Wagner’s use of the sentence
in his post-Lohengrin music dramas;2 Nathan John Martin has investigated the
use of Caplin’s thematic types in Mozart’s and Haydn’s operas, and has dis-
cussed an example of a sentence in a song from Schumann’s Dichterliebe;3 and
Stephen Rodgers has studied Schubert’s use of sentence form in Die schöne
Müllerin.4 As for applications to the nineteenth-century repertoire, Caplin
himself has emphasized the continuing relevance of his theme types to nine-
teenth-century music, and much of the work on vocal music cited above deals
with music of that century.5
In this chapter I follow in the above authors’ footsteps as I focus on
Schumann’s use of the sentence in his lieder. My aims are twofold: to investi-
gate (1) how Schumann’s sentences relate to the structure of the poetic texts,
and (2) how his sentences reflect the meaning of the texts.
Preliminary Matters
Before embarking on this investigation, I consider two preliminary issues: the
nature of the nineteenth-century sentence, and the role played by a theme type
that we associate with classical instrumental music within a short vocal genre

226 harald krebs
of the romantic period. The authors cited above acknowledge that numerous
deviations from classical norms must be expected as one ventures beyond the
original purview of form-functional theory. The many deviations from normal
tight-knit theme types that Caplin has discovered within the classical reper-
toire are encountered much more frequently in nineteenth-century music. In
his discussion of the sentence, Caplin notes that even in tight-knit sentences,
there may occur expansions beyond the normative two-measure duration of
the basic idea, extensions of continuation function and expansions of caden-
tial function, and compressions of the continuation phrase.6 He demonstrates
that loosely knit sentences, typical of subordinate themes, deviate even more
drastically from the basic tight-knit format. For the presentation phrase, he
describes and illustrates the following possibilities: (1) more than two repeti-
tions of the basic idea; (2) repetition of the entire phrase; and (3) weakening
of tonic prolongation within the phrase.7 There are even more possibilities of
loosening techniques for the continuation-cadential portion of a sentence: (1)
extension of the continuation function by “excessive” repetition of fragments;
(2) initiation of the continuation phrase with units of the same duration as the
basic idea, and fragmentation of these units only thereafter; (3) extension of
cadential function by various techniques of avoidance of the expected perfect
authentic cadence (PAC); (4) repetition of the cadence; and (5) expansion of
one or more of the harmonic components of the cadence.8
But aside from an increase in instances of these classically sanctioned
sentence types in nineteenth-century music, we can, according to Martin
and Rodgers, expect sentences in “novel constellations.”9 Martin describes
a sentence whose presentation eschews prolongation of the tonic harmony
(“Florestan” from Schumann’s Carnaval),10 whereas Rodgers finds sentences
in Schubert’s Die schöne Müllerin in which a series of presentations of differ-
ent basic ideas is followed at long last by a continuation.11 BaileyShea similarly
mentions sentences in Wagner’s music dramas that postpone the continuation
(or avoid it completely). He also cites sentences that do not follow the typical
teleological path of the classical theme type, that avoid cadential closure, or
that do not exhibit the typical proportions of a classical sentence.12
Given the flexibility of the sentence theme type, especially in nineteenth-
century repertoires, some delimiting of the category is necessary if it is not to
become meaningless. For the purposes of this chapter, I shall impose the fol-
lowing minimum requirements for a sentence: (1) it must begin with a presen-
tation phrase comprising at least two statements of a basic idea;13 and (2) the
presentation phrase must be followed by a unit that is more unstable and active
(the continuation). This minimal definition allows for all of the sentence types
briefly mentioned above; it does not impose a specific duration on the basic
idea (b.i.), a specific number of repetitions of the b.i., a particular harmonic
structure on the presentation phrase, or a specific series of events on the con-
tinuation portion.14

sentences in the lieder of robert schumann 227
It is clear from the work of the aforementioned authors that not only the
nature but also the placement of sentences must be expected to be more flex-
ible in nineteenth-century music than in music of the classical period. The
types of loosely knit sentences that Caplin outlines as well as the novel types
that BaileyShea, Martin, and Rodgers describe do not occur in nineteenth-cen-
tury works only in those contexts in which loosely knit sentences are typically
found in the classical style (that is, after a tight-knit initial theme, within sub-
ordinate themes and transitions). As a result, the progression from tight-knit
to more loosely knit themes characteristic of classical sonata expositions does
not necessarily govern those nineteenth-century lieder (or other small forms)
that employ classical theme types. If a given short nineteenth-century piece
contains more than one sentence, for example, it is just as likely to begin with
some loosely knit or novel sentence type and to progress to a tight-knit type as
the other way around. Since a lied can be very short indeed, it is also possible
for an entire lied to consist of a single sentence—an option that does not exist
for classical works.
It should not surprise us that Schumann, who was very familiar with instru-
mental music of the classical period, should have employed a theme type that
is central to that music. What might be more surprising is his use of this theme
type in his vocal music. But as Martin has shown, the sentence is not exclusively
an instrumental theme type; it is also prominent in classical operas.15 More
relevant to Schumann is the frequent presence of sentences in a class of songs
that were models for him and for most early nineteenth-century German song
composers, namely, folk songs. My browsing through a collection of German
folk songs uncovered numerous examples of sentences, one of which I repro-
duce here (ex. 7.1).16
Sentences in Schumann’s Lieder

In order to demonstrate the prevalence of the sentence theme type in
Schumann’s lieder, I offer a few introductory examples, mostly from famil-
iar Liederjahr songs. I begin with some examples that are quite straight-
forward, then show some that, albeit less similar to classical models, still
qualify as sentences.
The vocal line of the first song of Frauenliebe und Leben (ex. 7.2) begins
with a three-measure b.i. that is immediately repeated a step higher (with
some modification). The following eight measures constitute a typical con-
tinuation-to-cadential phrase, featuring fragmentation; a two-measure unit is
repeated almost without change (mm. 7–9 and 9–11), whereupon a one-mea-
sure unit is sequenced downward by step (precisely retrograding the slower
upward sequence of the presentation phrase—mm. 11–13). A PAC is postponed
by a deceptive resolution to vi in measure 15, but the expected cadence is

228 harald krebs
Example 7.1. An example of a sentential German folk song
Example 7.2. Schumann, “Seit ich ihn gesehen,” op. 42, no. 1, mm. 2−17
completed in the piano part in measure 17. The beginning of the next stro-
phe, with its melodic scale degree 5, overlaps with the scale degree 1 of the
cadential tonic.
The first song of the opus 39 Liederkreis, “In der Fremde” (ex. 7.3), is mod-
eled in its entirety on a sixteen-measure sentence. It begins with a twofold
statement of a four-measure b.i. that clearly prolongs tonic harmony (mm.
1–9). The continuation immediately departs from the tonic and becomes more

Example 7.3. “In der Fremde,” op. 39, no. 1, mm. 1−25
active in various ways. The piano part (not shown here) twice states a one-
measure idea based on a rising and falling perfect fifth. The harmonic rhythm
increases from a predominantly whole-note pace to a consistently half-note
one.17 The melody contributes to the greater sense of activity by its increasing
disjunctness and by its exceeding of the very limited ambitus of the presenta-
tion phrase. An allusion to the b.i. in the subdominant key (mm. 15–19) leads
into a PAC in the tonic, which closes the sentence at measure 21. The remain-
der of the song is a wistful, tonic-prolonging coda.18
Another example of an almost “classical” sentence comes from a slightly
less familiar song—one of Schumann’s Kerner settings, opus 35. The vocal
line of “Wanderlied” (ex. 7.4) begins with two statements of a fanfare-like,
tonic-prolonging b.i. The second statement moves the fanfare upward a notch
within the prolonged tonic, and at the last moment breaks out of that harmony
by moving to V of ii, and by replacing the f2 of the first statement with an

230 harald krebs
Example 7.4. “Wanderlied,” op. 35, no. 3, mm. 3−11
intensifying g2 (m. 6). In the continuation phrase the harmonic rhythm speeds
up drastically; the static tonic that occupied almost the entire presentation
phrase is succeeded by a quarter-note, then an eighth-note surface-level har-
monic rhythm (although on a deeper level, as shown in the example, ii har-
mony is prolonged for one and a half measures—mm. 7–8). An emphatic PAC
concludes the sentence, which is then repeated to set the following two stanzas
of the poem.
Additional examples of straightforward sentences from Schumann’s famil-
iar Liederjahr songs include the opening of “Auf einer Burg,” op. 39, no. 7
(mm. 1–18—the second statement of the four-measure b.i. is partially trans-
posed down a third, and the continuation features sequencing of two-measure
fragments) and the opening of “An meinem Herzen,” op. 42, no. 7 (mm.
2–9—see ex. 7.5). (The sentence, with a two-measure b.i., is repeated for each
stanza of the poem, a closing cadence being postponed until the final state-
ment of the sentence. The postlude of the song is also a sentence, based on
a different b.i.) Another source of obvious sentences is the Lieder-Album für
die Jugend, op. 79—the first opus from Schumann’s second outpouring of lie-
der (1849). Schumann was clearly aiming for a folk-like air in many of these
songs, his agenda being to draw young people from the familiar folk-song rep-
ertoire toward the more ambitious art song.19 A number of the songs begin
with sentential four-measure phrases (where the b.i. is only one measure long);
examples are the first four measures of the two “Zigeunerliedchen,” of “Des
Knaben Berglied,” and of “Die wandelnde Glocke.” The first two phrases of
“Lied Lynceus des Türmers” are also sentential (mm. 1–4 and 5–8).

Example 7.5. “An meinem Herzen,” op. 42, no. 7, mm. 2–9
Some of the sentences in the Lieder-Album für die Jugend are slightly less obvi-
ous exemplars of the theme type. In “Er ist’s” (ex. 7.6), the two-measure b.i.
is sequenced, but its rhythm is significantly altered in the second statement,
the shorter note values suggesting the “fluttering” that the poet mentions. The
piano then offers a third statement of the b.i. at the initial pitch level. This
statement could be heard as an unorthodox addition to the normative presen-
tation, but it also suggests the beginning of the continuation (it is this ambigu-
ity of function that renders the sentence unorthodox). The example shows the
latter interpretation; in the varied statement of the b.i. in measure 7, a latent
downward sequence within that idea is actualized, and this immediate down-
ward transposition of a one-measure segment establishes the fragmentation
(and therefore acceleration) that is expected in a continuation. The piano
provides the final cadence without the participation of the voice. In “Frühlings
Ankunft” (ex. 7.7), the end of the second statement of the b.i. is significantly
altered; up until the last note, this statement sequences the first down by step—
but then, in an appropriately optimistic gesture, the melody bursts out of the
downward sequence and presents a note that is higher than any within the first
statement. The piano part (shown in small notes) states a third rendition of
the b.i. as the voice begins the continuation, so that the boundary between the
presentation and the continuation is again blurred.
Similar obfuscations occur in other lied opera by Schumann. The middle
section of “Widmung” (ex. 7.8), for instance, begins with two four-measure
segments. Their initial three measures are virtually identical in terms of mel-
ody, but the first is supported throughout by a tonic pedal (in the key of ♭VI),
while the second leads to a half cadence (HC); the end of the melody of the
second statement is altered to accommodate this significant change in har-
mony. The passage that follows these nearly identical four-measure segments is

232 harald krebs
Example 7.6. “Er ist’s,” op. 79, no. 23, mm. 3−14
Example 7.7. “Frühlings Ankunft,” op. 79, no. 19, mm. 1–8
considerably more active: the bass begins to participate in the throbbing triplet
chord pattern that was established by the right hand in measure 14, and the
melody not only uses shorter note values but also subdivides into two-measure
rather than four-measure segments (mm. 21–23, 23–25, 25–27, 27–29). The
first two of these segments are closely related; in measures 23–25, the preced-
ing segment is sequenced up a fourth (with rhythmic modifications). The
second pair of segments is not a pair in the same sense; there is no melodic
similarity, and the two segments are gathered together by the dominant pedal
(in the home key) that underlies them.

Example 7.8. “Widmung,” op. 25, no. 1, mm. 14−29
Two features of this section are unusual with respect to the classical sen-
tence: the significant melodic and harmonic difference of the initial four-
measure statements, and the lack of a cadence at the end. But the opening
segments are similar enough to qualify as statements of the same b.i. and
hence, to form a presentation phrase; and the undeniable increase in activity
in the following passage results in a reasonable continuation segment. The
continuation, in fact, exhibits the 1+1+2 proportions found in many classical
sentences. In spite of its unusual features, I regard the excerpt as a convinc-
ing sentence.20
The Sentence and Poetic Form in Schumann’s Lieder

After this introductory glance at Schumann’s use of the sentence in his lieder,
I move on to a consideration of the relationship of the sentences to the poetic
text—first, the relationship between sentence structure and poetic form.
Rodgers, who has investigated this issue in Schubert’s Die schöne Müllerin, finds
a correlation between the use of sentences and rhyme schemes in that cycle:

234 harald krebs
Schubert often sets an initial pair of rhyming lines to the presentation phrase
of a sentence.21 The same is true of many of Schumann’s sentences; most of
the sentences shown in the above examples illustrate this point. In examples
7.2 and 7.4, the two statements of the b.i. set couplets exhibiting the rhyme
scheme ab and cb, respectively; the “b” rhymes match the musical rhyme of
the two basic ideas. In example 7.3, the two b.i. statements both set couplets
exhibiting the scheme ab. Each of the two b.i. statements in example 7.8 sets a
single rhyming line. The opening passages of “Auf einer Burg” and “An mei-
nem Herzen” (ex. 7.5) are further illustrations of correspondences between
presentation phrases and the initial rhymes of the poems.22
A pair of rhyming lines, or a pair of rhyming couplets, does map neatly onto
the twofold b.i. of the presentation phrase of a sentence—but what of the con-
tinuation? The shape of a typical continuation is not as obvious a fit to com-
mon rhyme schemes of German lyric poetry. Repeating, or nonrepeating but
bipartite rhyme schemes such as a/a, ab/ab, or ab/ba perfectly match the form
of a presentation phrase, but not that of a continuation phrase; the continua-
tion and cadential portions of the latter are not commonly separated by a clear
caesura, and it is certainly not typical of a continuation to be divided into two
similar units. Because of this imperfect mapping between musical and poetic
form, there exist numerous sentences in lieder where only the presentation
phrase corresponds to the rhyme scheme.23 Schumann’s “Auf einer Burg” illus-
trates. The two-measure segments of the continuation (mm. 9–18), although
they do match the lineation of the poem, exhibit no relationship to its rhyme
scheme; the latter (as in the presentation) is abba, but the musical content of
the two-measure segments follows the incongruent scheme aaab.
In many of the continuation portions of Schumann’s lied sentences, there
is nonetheless some correspondence to the poem’s rhymes. Schumann cre-
ates this correspondence in two ways: either he manipulates the poem so that
it fits into a continuational framework, or he deviates from the norms for the
continuation so that the poem’s rhyme scheme is matched. Example 7.2 illus-
trates the first strategy. As was mentioned, the initial b.i. statements match
the rhyme scheme of the initial couplets; the first b.i. statement sets the cou-
plet “Seit ich ihn gesehen, / glaub’ ich blind zu sein”; and the second state-
ment sets the couplet “wo ich hin nur blicke/ seh’ ich ihn allein.” The rhyme
“sein/allein” is perfectly matched in terms of contour and rhythm by the
corresponding endings of the two b.i. statements. The following four lines,
set to the continuation portion of the sentence, continue the same rhyme
scheme—but Schumann avoids a similar highlighting of the rhyming words
that end the two couplets (“vor,” “empor”). The musical settings of these
words do not correspond at all; in fact, Schumann does not even pause after
the first couplet (at “vor”), but links the two couplets by beginning a voice
exchange during the final word of the first couplet and completing it as the
second couplet begins (m. 11). The concealing of the rhyme “vor/empor”

and hence of the joint between the couplets prevents an atypical hiatus in the
middle of the continuation.
Example 7.3 involves a similar manipulation of the poem during the con-
tinuation. The relevant lines are “Wie bald, ach wie bald kommt die stille Zeit,
/ da ruhe ich auch und über mir / rauscht die schöne Waldeinsamkeit/ und
Keiner kennt mich mehr hier”; the rhymes “Zeit/-keit” and “mir/hier” main-
tain the abab scheme established in the first stanza.24 Schumann, however, ren-
ders these rhymes inaudible by avoiding a pause after “mir,” the end of the first
couplet. He positions this word, in fact, in the midst of the aforementioned
allusion to the initial b.i., thus depriving the word of any articulative function.
The continuation thereby gains a completely different shape from the presen-
tation (which did reflect the rhyme scheme and the couplet form), and the
overall sentential structure of the song becomes clearly perceptible.
Another example of manipulation of a poem to mold it to continuational
norms is found in the final song of Frauenliebe und Leben. In the presentation
phrase, each of the repeated b.i. statements sets one of the initial couplets (with
ab rhymes—“getan/traf, Mann/-schlaf”). The even rhymes, “traf” and “-schlaf,”
are clearly audible because they occur at the ends of the almost identical b.i.
statements, which are articulated by arrivals on V and by rests—but even the
odd rhymes “getan” and “Mann” are perceptible, because both are associated
with relatively long f1s (mm. 3 and 6). The situation changes in the continua-
tion of the sentence. The poem continues the same rhyme scheme (“hin/leer,
bin/mehr; zurück/fällt, Glück/Welt”), but Schumann avoids the earlier clear
correlation between repeated music and rhymes. His vocal rhythm causes the
words “ich bin” to be attached to the following line (“nicht lebend mehr”);
he thereby musically renders Chamisso’s enjambment, but entirely obscures
the rhyme “hin/bin.” The rhyme “leer/mehr” is slightly more perceptible
because these words are followed by rests and because the corresponding lines
are set to the same rhythm (cf. mm. 9–11 and 13–15); the pitches, however,
are completely different. During the setting of the final four lines, the odd
rhyme “zurück/Glück” is again imperceptible (there is no pause on “-rück”
and the pitches are different). The even rhyme “fällt/Welt” is partially ren-
dered in the music by the use of the same pitches for both words (c♯1–d1–e1),
but the rhythms at the ends of the corresponding even-numbered lines do not
match. In short, whereas Schumann regularly creates musical correspondences
to repeating and bipartite rhyme schemes during presentations, he sometimes
weakens or even conceals such correspondences during continuations, thereby
preventing nonnormative articulations and repetitions during those units.
Of course a continuation phrase may exhibit articulations and repetitions: it
may begin with fragmentation—with repeated (usually sequenced) units half
the duration of the b.i. Example 7.8 demonstrates a manipulation of a poem
to create a correlation with a continuation phrase of this form. As was men-
tioned, the continuation begins with two-measure units (half the duration of

236 harald krebs
the four-measure b.i.), the second unit sequencing the first. Schumann speeds
up the declamatory rhythm at this point, so that the rhyming lines occupy two
measures as opposed to four; the sequential two-measure units perfectly match
the rhyming couplets.
Example 7.4 shows the opposite strategy: here, Schumann deviates from
the typical form of a continuation to accommodate the rhyme scheme of the
poem. Both the presentation and the continuation set a rhyming couplet.
During the continuation, Schumann makes no attempt to sweep the first mem-
ber of the rhyming pair (“Haus/hinaus”) under the rug; on the contrary, he
pauses dramatically on a high note at “Haus” (m. 8), and leads into the pause
with a ritardando. This continuation, then, has an uncharacteristic caesura in
the middle.
Schumann employs the sentence theme type to match aspects of poetic
structure other than the rhyme scheme. In example 7.8, the twofold b.i. state-
ments match not only the rhyme “Frieden/beschieden” but also the repeated
line-incipit “Du bist.” In fact, the latter parallelism is even more relevant to
the musical structure than the rhyme, since the strongest musical parallelism
occurs at the beginning of the b.i. statements rather than at the end. Example
7.9 shows a sentence that matches a somewhat less obvious poetic correspon-
dence. The initial line of the poem subdivides into two phrases that are par-
allel in structure and syntax, though contrary in meaning: “Es flüstert’s der
Himmel, / es murrt es die Hölle” (’Twas whispered in Heaven, ’twas muttered
in Hell—“it” being the letter H, as Schumann wittily announces by the whole-
note B♮ in m. 1).25 The following line does not subdivide in the same manner
(“Nur schwach klingt’s nach in des Echo’s Welle” [only weakly does it resound
in the echo’s wave]). These two lines perfectly match the structure of a typical
sentence, and Schumann does set them to this theme type; the two syntacti-
cally corresponding phrases of the first line are set to two statements of a one-
measure b.i. (with just one changed pitch at the end), and the continuation
matches the unbroken flow of the second line.
Near the end of the same song, the opening material returns, the pre-
sentation now setting a line that lists two geographical locations (“In
Griechenland klein, an der Tiber Borden” [In Greece, it is small; on the
banks of the Tiber]). Again, the continuation sets a line that is not bisected
in this manner (“ist’s größer, am größten in Deutschland geworden” [it is
larger, but it has grown largest in Germany]). A similar example of the cor-
relation of the presentation with a short list is shown in example 7.10. The
first line enumerates two signs of spring, young green shoots and fresh grass,
each of which Schumann sets to a triadic b.i. (in the vocal line, the inversion
of the triad changes in the second statement, but the piano part, not shown
in the example, imitates the voice’s first statement precisely). The following
unbroken line, “Many a heart has returned to health through you,” is set to a
similarly unbroken continuation.

Example 7.9. “Räthsel,” op. 25, no. 16, mm. 1–5
Example 7.10. “Erstes Grün,” op. 35, no. 4, mm. 1–5
I demonstrated that subdivisions within the continuation section may

reflect poetic rhymes (see the analysis of “Widmung” above)—but these sub-
divisions may also relate to other kinds of poetic parallelisms. In example 7.5,
where the presentation’s two-measure segments correspond to rhyming lines,
the repeated one-measure segments that initiate the continuation (mm. 5–7)
reflect the corresponding, albeit retrograded statements—“joy is love” and
“love is joy”—into which the given line subdivides.
The Sentence and Poetic Meaning

We have seen that numerous sentences in Schumann’s songs map in some
fashion onto the poetic structure. In the final section of this chapter, I con-
sider correspondences between sentential structure and the meaning of the
poem. BaileyShea, Rodgers, and Martin have done important work in this area.
BaileyShea argues that a sentence as an abstract category has no “specific dra-
matic quality,” but that within a specific sentence, the manner in which the
basic idea is manipulated might well have expressive meaning. BaileyShea also

238 harald krebs
makes the important point that dramatic expression can result from the way
the composer’s choices with respect to the various components of the sentence
“interact with the larger, discursive background of sentences in general”; that
is, he finds that unusual sentences gain meaning against the background of
the normative sentence schema. He describes and illustrates three “rhetorical
categories of sentence expression” in Wagner’s operas.26 Rodgers convincingly
relates Schubert’s presentations with multiple pairs of basic ideas (he calls
them “manic sentences”) to a sense of “inner tension, . . . expectancy and pent-
up emotion.” The eventual relinquishment of the basic idea in the continua-
tion, on the other hand, provides a sensation of release of tension. Rodgers
demonstrates that these sensations align with the emotions of the protagonist,
as conveyed by the poet.27 Martin provides one example of the text-expressive
use of a sentence in Schumann’s lieder: he shows that the sentence in mea-
sures 5–17 of the first song of Dichterliebe “circles around to ‘end’ with its own
beginning” (i.e., the cadence of m. 17 is at the same time the beginning of
another statement of the basic idea), and that this unusual sentence contrib-
utes to Schumann’s expression of the poetic idea of unresolved yearning.28
I offer some additional examples of text-expressive sentences in Schumann’s
lieder. Like Rodgers’s and Martin’s, most of my examples involve nonnorma-
tive sentences. My first example is a “manic” sentence, much like those that
Rodgers discovers in Schubert’s Die schöne Müllerin. “Warte, warte, wilder
Schiffsmann,” from the Liederkreis, op. 24 (ex. 7.11), is certainly a “manic” song.
The protagonist refers to the boatman whom he addresses as “wild,” but it is he
himself who best fits that description. His language is intense and unbridled;
as he takes farewell from his beloved and from his native continent, he asks
his eyes and his body to spew forth streams of blood with which he will write
of his pain, and he revels in wild accusations of the woman from whom he is
fleeing. Appropriately, Schumann does not set these desperate lines to coher-
ent musical sentences. He does, however, evoke sentential structure from the
outset by repeating several basic ideas. The initial four-measure b.i. is repeated
twice (mm. 1–4, 5–8, 9–12). There is already an element of “wildness” in the
treatment of this idea: neither of the two repetitions is exact. The first repeti-
tion is exact in the voice part, but the first two measures of the piano part (mm.
1–2 and 5–6) are altered. The second repetition, on the other hand, is exact
as far as the piano is concerned (cf. mm. 5–8 and 9–12), but the voice part is
simplified to a skeletal 1^–5^–1
^
.
Measures 12–16 bring two statements of a softer, contrasting idea; this pas-
sage does not sound like a continuation following from the initial presentation,
but rather like a new presentation. And more presentations follow: in measures
17–24, a four-measure idea is sequenced a step downward, and in measures
25–32, a different four-measure idea (itself dividing into two repeated state-
ments) is sequenced a third upward. After a restatement of some of the open-
ing material in the piano, we hear a passage whose form can be interpreted in

Example 7.11a. “Warte, warte, wilder Schiffsmann,” op. 24, no. 6, mm. 1–54
(continued)

Example 7.11a.—(continued)
(continued)

Example 7.11a.—(concluded)
two ways. In measures 37–44, a two-measure segment, consisting mostly of a

tritonal ululation, is repeated, then followed by a “wild” four-measure synco-
pated passage; this eight-measure unit forms a coherent sentence. The entire
sentential unit, however, is immediately sequenced a major third upward (a
sixth downward in the voice). The mini-sentence notwithstanding, I prefer to
interpret measures 37–44 as an eight-measure b.i., and the slightly expanded
repetition of this b.i. in measures 45–54 as completing another member of the
ongoing parade of presentations.29
A wailing descending semitone over the expanded final dominant har-
mony of the latter presentation (the word “Oh” is Schumann’s addition to the

242 harald krebs
Example 7.11b. “Warte, warte, wilder Schiffsmann,” mm. 96−111
poem!) leads to a return of the opening, and indeed of the three initial pre-
sentations; measures 55–66 correspond to measures 1–12, measures 66–70
to measures 12–16, measures 71–78 to measures 17–24, and measures 79–92
to measures 25–32. Some of these restatements intensify the earlier versions.
The downward sequence of measures 17 and following is replaced by a rising
sequence. To the upward-third sequence first stated in measures 25 and fol-
lowing, Schumann adds a third segment (mm. 87–92), and then, after a “wild”
piano intrusion, a modified fourth segment (which, if the tenor can manage
it, culminates on the highest vocal pitch in the song—a notated a2). With this
shriek (on the word “death”!), the vocal portion of the song ends.
The piano, however, launches into a substantial postlude, based on the intro-
ductory material. The first half of the postlude—measures 99–107—is the first
music that can qualify as a continuation-cadential section on the same level as
the many presentations that have been offered. This passage perfectly accords
with expectations for such a section. After a precipitous downward sequence
based on a one-measure fragment of the introductory material (mm. 99–104),
and a repetition of the same fragment (mm. 104–5), the continuation is brought
to a close by a PAC. The remainder of the postlude consists of codettas.

The long succession of presentations in this song, like similar successions in

songs studied by Rodgers, creates an effect of increasing tension, as expectations
for a continuation are repeatedly defeated. This effect is enhanced by the afore-
mentioned intensifications in the latter portion of the song. The tension gener-
ated by the manic sentence is, of course, perfectly appropriate for this poem.
As BaileyShea has noted in Wagner’s music, many of Schumann’s nonnorma-
tive sentences generate meaning against the background of normative ones—
although in Schumann’s case, the background is usually not a “general discursive”
one, but one or more articulated normative sentences. A good illustration is “Ich
wandelte unter den Bäumen” (ex. 7.12), from the same cycle as the preceding
example. The piano introduction is a beautiful, nearly normative sentence; the
second statement of the one-measure b.i. sequences the first a third higher so that
a melodic peak is attained. In the continuation, a downward chromatic upper
voice gradually releases the tension; in an inner voice, the second motive of the b.i.
(half a measure in duration) is heard three times, and a liquidating eighth-note
descent concludes the sentence. The only unusual feature of this sentence is the
lack of a cadence. The vocal line begins with a longer normative sentence: a two-
measure b.i., prolonging tonic harmony, is repeated with minor rhythmic altera-
tions. The five-measure continuation (mm. 8–13) begins with a new two-measure
unit, then presents three repetitions of a one-measure fragment, introduced by the
voice and taken over by the piano. The sentence ends with a V7 harmony, includ-
ing a poignant sustained upper neighbor. The lack of a true cadence is, again, the
only unusual feature. The sentence is immediately repeated, this time ending even
more poignantly with a minor V9 chord.
In measure 23, the ninth (G) magically becomes the local tonic (♭VI in
terms of the overall B major). In this new key, a new basic idea, four measures
long and prolonging the new tonic, is stated twice. Given the earlier music, one
expects a continuation to follow this presentation—but instead, after a pause,
the initial sentence in B major returns and is restated in full, this time end-
ing with a PAC (the piano provides scale degree 1 above the final vocal D♯).
Thereafter, the song is closed by a restatement of the introductory sentence.
The sentences in this song do not, with one exception, exhibit correspon-
dences to the structure of the poetry. The rhyme scheme is abab. During the
B-major sentences, the two statements of the b.i. set the first ab pair of the
given stanza, and the continuation sets the second pair. Only the G-major
presentation matches the rhyme scheme. Since the b.i. during this passage
is four measures in length, it accommodates two lines of poetry, each state-
ment setting an ab pair; the twofold statement of the b.i. thus corresponds to
the b rhymes. But it is clear that Schumann’s main concern in this song is not
to map his sentences onto the poetic form. In cases like this, one might ask
if the meaning rather than the structure of the poem motivates the senten-
tial structure—and indeed, this is a fruitful line of inquiry. The first two com-
plete B-major sentences set the words of the protagonist—his description of his

Example 7.12a. “Ich wandelte unter den Bäumen,” op. 24, no. 3, mm. 1−13
Example 7.12b. “Ich wandelte unter den Bäumen,” mm. 14−30

Example 7.12c. “Ich wandelte unter den Bäumen,” mm. 31–45
perambulations beneath the trees and of his pain, and his plea to the birds,
who are singing some little word that opens old wounds, to be silent. The fol-
lowing G-major section—a presentation without a continuation—sets the birds’
response (a maiden walked here before the man did—and it was from her that
they learned the “beautiful golden word” that so distresses the protagonist). The
man’s enigmatic final statement, expressing denial and distrust, is again set to a
complete sentence. Schumann’s strategy is clear: human statements are set to
complete, coherent sentences, whereas the statement of the birds is “othered” by
being set to an incomplete formal unit. The unusual clarity of the surrounding
sentential forms renders the avian nonsentence all the more striking.30
In “Stille Liebe,” op. 35, no. 8 (ex. 7.13), Schumann uses the contrast between
normative and unusual sentences in a different way. The poem deals with the dif-
ficulty of expressing love; the poet wishes to sing his beloved’s praises, and does
indeed sing, but he bewails his failure to express all that he feels—his failure to

Example 7.13a. “Stille Liebe,” op. 35, no. 8, mm. 1−16
Example 7.13b. “Stille Liebe,” mm. 17−28

Example 7.13c. “Stille Liebe,” mm. 29–49
celebrate adequately the beloved in song. The hesitant pauses in the piano
introduction already suggest repeated unsuccessful attempts to communi-
cate.31 The form of the introduction, and of the two subsequent piano inter-
ludes, ties in with this interpretation. During the introduction, a two-measure
b.i. is stated twice to create a normative presentation. The final two measures of
the introduction function as a syntactically correct continuation-cadential seg-
ment; a twofold statement of a two-sixteenth-note rising fragment is followed
by a melodic descent that leads to a PAC in the tonic key. This continuation is,
however, surprisingly brief in relation to the presentation; it occupies only two
measures rather than the expected four. Against the background of the norma-
tive sentence, then, the introduction could be regarded as a failed utterance.
Two interludes (mm. 14–18 and 26–31), related to the introduction and
inevitably heard in relation to it, depart even further from the sentential norm.
In the first interlude, a minor-mode version of the initial b.i. is followed by a
curtailed sequential repetition (the last note is not extended by a fermata).

248 harald krebs
This second statement of the b.i. leads directly into an IAC in the key of ♭III,
which overlaps with the next strophe; there is no equivalent of the continua-
tion rhetoric (the fragmentation) briefly heard in measure 5. In the second
interlude, the b.i. (still in the key of ♭III) is again sequenced downward by
thirds, as during the introduction—but a third, slightly simplified statement,
another third lower, appears in measures 29–30. The chromatically descend-
ing sixteenth notes in measures 30–31 allude to the continuation of measure 5
(which is based on a longer sixteenth-note descent), but this is a mere vestige
of a continuation. Furthermore, the inverted dominant in measure 31 weakens
the cadential effect that was present in the preceding interlude.
The vocal strophes, all three of which are almost normative sentences,
actualize the discursive background against which we can interpret the intro-
duction and interludes as failed, unsuccessful utterances. In each strophe, a
two-measure b.i. is sequenced upward by step to form a presentation, and a
four-measure continuation, concluding with a PAC, follows. (The only miss-
ing element is a sense of acceleration during the continuation.) These nor-
mally proportioned sentences suggest the successful, eloquent and expressive
“Lieder” that the lyric I so urgently desires to produce.
The postlude, interestingly, “corrects” the nonnormative aspects of the
piano’s earlier sentences. The b.i. statements remain unchanged, but the con-
tinuation portion is expanded; Schumann augments the sixteenth notes of
measure 5 to eighth notes, and extends the sequenced fragments from two
notes to four. He also renders the cadence substantially longer and stronger
than in the earlier piano sentences; the postlude ends with a PAC whose dominant
and tonic harmonies both occupy two measures—a striking contrast with the curt
V–I at the end of the introduction, with the IAC of the first interlude, and with the
complete lack of cadence in the second interlude. As a result of the expansions in
the continuation-cadential portion of the postlude, the typical proportions of the
sentence theme type are approached much more closely here than in any earlier
piano solo passage; a four-measure presentation is followed by a six-measure, as
opposed to a two-measure or an even shorter continuation-cadential portion. It
is also noteworthy that in the latter portion of the postlude, the hesitant gestures
of the b.i. yield to flowing, uninterrupted gestures. Both the gestural quality of
the postlude and its near-normative sentence form might refer, as do the vocal
sentences, to the successful communication of love for which the lyric I strives. By
ending the song with such a sentence, rather than with a broken one of the sort
that was stated in previous solo piano passages, Schumann suggests there is some
hope that the desired message of love might in the end be composed.
I have shown in earlier studies that Schumann mobilizes many aspects of
musical structure in his effort to reflect the meaning of his poetic texts.32 From
the analyses presented here, it is apparent that Schumann’s compositional
choices with respect to theme type, too, are determined by his understanding
of, and sensitivity to, the poetry that he chose to set to music.

Notes
2. Matthew BaileyShea, “The Wagnerian ‘Satz’: The Rhetoric of the Sentence in
Wagner’s Post-Lohengrin Operas” (PhD diss., Yale University, 2003); and “Wagner’s
Loosely Knit Sentences and the Drama of Musical Form,” Intégral 16/17 (2002/3):
1–34.
3. Nathan John Martin, “Formenlehre Goes to the Opera: Examples from Armida
and Elsewhere,” Studia musicologica 51 (2010): 387–404; Martin, “Schumann’s
Fragment,” Indiana Theory Review 28 (2010): 85–109; Martin, Formenlehre Goes to
the Opera: Examples from Don Giovanni,” in Mozart in Prague: Proceedings of the
International Conference of the Mozart Society of America and the Society for Eighteenth-
Century Music, 9–13 June 2009, Prague, ed. Kathryn Libin (Prague: Academy of
Sciences of the Czech Republic, forthcoming).
4. Stephen Rodgers, “Sentences with Words: Phrase Structure and Poetic Structure in
Schubert’s Die schöne Müllerin,” Music Theory Spectrum 36 (2014): 58–85.
5. William E. Caplin, “Classifying Harmonic Progressions,” keynote presentation at
the West Coast Conference of Music Theory and Analysis, University of Victoria,
2002. For a recent application of Caplin’s theories to nineteenth-century music, see
Steven Vande Moortele, “Sentences, Sentence Chains, and Sentence Replication:
Intra- and Interthematic Formal Functions in Liszt’s Weimar Symphonic Poems,”
Intégral 25 (2011): 121–58.
6. Caplin, Classical Form, 40, 47, 48.
7. Ibid., 99.
8. Ibid., 101–11. Caplin mentions additional loosening techniques, too numerous to
list in full here, on pp. 111–21.
9. Martin, “Schumann’s Fragment,” 90.
10. Ibid., 94–97.
11. Rodgers, “Sentences with Words,” 71–83.
12. BaileyShea, “Wagner’s Loosely Knit Sentences,” 11–14.
13. This condition excludes the “monofold sentence” (a sentence in which the presen-
tation consists of a single statement of a basic idea), proposed in Mark Richards,
“Viennese Classicism and the Sentence Idea: Broadening the Sentence Paradigm,”
Theory and Practice 36 (2011): 189–92. Like Rodgers, I regard the repetition of the
basic idea as an essential feature of the sentence concept (see Rodgers, “Sentences
with Words,” 59n).
14. The requirement for instability during the continuation excludes forms of the type
AABA (where the final A is a complete and unaltered return of the opening section).
Such forms are very common in nineteenth-century lieder. A final return of the basic
idea (A), however, lends such units a sense of stability and of being “rounded off”
that seems to me to set them apart from the sentence category. William Rothstein
refers to such units as “quatrains”; see Phrase Rhythm in Tonal Music (New York:
Schirmer, 1989), 107. Both Rodgers and BaileyShea accept quatrains as sentences;
see Rodgers, “Sentences with Words,” 69, and BaileyShea, “Beyond the Beethoven
Model: Sentence Types and Limits,” Current Musicology 77 (2004): 16–17.
15. See Martin, “Formenlehre Goes to the Opera: Examples from Armida and Elsewhere”;
Martin, “Schumann’s Fragment”; and Martin, “Formenlehre Goes to the Opera:
Examples from Don Giovanni.”

250 harald krebs
16. Engelbert Humperdinck incorporated the melody shown in example 7.1 into
Hänsel und Gretel; it is assumed, however, that it was a preexisting folk melody.
See Das große Hausbuch der Volkslieder, ed. Walter Hansen (Munich: Mosaik Verlag,
1978), 113. I found the following additional instances of sentential folk songs in
the above anthology: “Walpurgisnacht” (p. 19), “Zwei Königskinder” (p. 94), “Jetzt
gang i ans Brünnele” (p. 97), “Ein Männlein steht im Walde” (p. 127; this melody
was also incorporated into Hänsel und Gretel), “Nun ade, du mein lieb Heimatland”
(p. 136), “Auf de schwäb’sche Eisebahne” (p. 234), “Lustig ist es im grünen Wald”
(p. 242), and “Blaublümelein” (p. 282). I did not count quatrains (AABA forms—
of which there are many!) as sentences. Nor did I count melodies that are deemed
folk songs in Germany but for which the composer is known (e.g., Carl Zelter and
Johann Reichardt), although there are numerous sentences in this category. The
latter composed melodies, in a deliberately simple style, provide additional evi-
dence that for German song composers there is an affinity between the sentence
and folk song. We shall encounter some relevant examples of folk-like melodies by
Robert Schumann later on.
17. The accelerated harmonic rhythm is not evident from the example. I have not shown
every chord in my examples; chords with embellishing functions are omitted.
18. The allusion to the basic idea in measures 16–17 is just that—a brief allusion, in a
nontonic key. I therefore do not regard it as a return of the opening, and do not
consider the form of the song as being AABA—a form that I have rejected as a pos-
sibility for sentences.
19. See Jon Finson, “Schumann’s Mature Style and the ‘Album of Songs for the
Young,’” Journal of Musicology 8 (1990): 231–32.
20. The HC at the end of the first phrase of the middle section raises the possibil-
ity of an analysis of this phrase as an antecedent, and of the entire section as a
hybrid of the form “antecedent–continuation” rather than a sentence. An ante-
cedent, however, normally consists of a basic idea followed by a contrasting idea.
Since the first phrase of this section begins with two statements of the same b.i., it
evokes, to my ears, the quality of a presentation phrase rather than of an anteced-
ent. The opening section of “Widmung” (mm. 2–13), incidentally, also exhibits
some sentential features. Measures 5–13 certainly conform to the expected char-
acteristics of a continuation; in fact, these measures begin like the unit encom-
passed by measures 21–29, which I have already termed a continuation. I do not
find, however, that the two initial vocal segments (mm. 2–3 and 3–5) are similar
enough to be regarded as the same b.i.; I therefore reject an analysis of the open-
ing section as a sentence.
22. Example 7.7 illustrates that Schumann’s presentation phrases do not always
set rhyming lines. The first statement of the b.i. sets an a rhyme and the second
statement a b rhyme. The presentations of the short sentences from opus 79 that
were mentioned above (in “Zigeunerliedchen,” “Des Knaben Berglied,” “Die wan-
delnde Glocke,” and “Lied Lynceus des Türmers”) do not correspond to the rhyme
schemes of the given poems. The basic ideas in these songs are so brief that they
cannot accommodate entire poetic lines.
23. Many of the examples from Die schöne Müllerin listed in Rodgers, “Sentences with
Words,” fall into this category.
24. Schumann slightly modified the last two lines of Eichendorff’s poem, but the
changes do not affect my points. For information about the modifications, see the

LiederNet Archive, http://www.recmusic.org/lieder/get_text.html?TextId=5174,

accessed April 23, 2015.
25. The poem is not by Byron, as older scores would have it; it is an adaptation by Karl
Friedrich Kannegießer of an English poem by Catherine Fanshawe.
26. BaileyShea, “Wagner’s Loosely Knit Sentences,” 20–21. BaileyShea’s examples do
fulfill the dramatic functions to which he alludes, but in some of them, it is less
the sentential structure than its particular harmonic and melodic building blocks
that creates the effect; for example, in a sentence from the end of the second act
of Siegfried, it is the slow tempo and the augmented triads that suggest Siegfried’s
exhaustion rather than the sentence theme type per se. Wagner’s choices of tempo
and harmony could have achieved the given dramatic effect even if he had used a
different theme type.
28. Martin, “Schumann’s Fragment,” 104–5.
29. Caplin uses the term “mini-sentence” in Analyzing Classical Form: An Approach for the
Classroom (New York: Oxford University Press, 2013), 88–89.
30. The “otherness” of the latter passage is enhanced by its remote key, its lower
dynamic level, its higher register, and its nonsyntactic harmony (in mm. 25–26
and 29–30, ii7 moves to I6 instead of to V, and at the ends of both b.i. statements,
Schumann leaves the harmony hanging on a I46).
31. I heard the pianist Hartmut Höll suggest this meaning for the introduction during
a master class at the Musikhochschule in Karlsruhe many years ago.
32. See Harald Krebs, Fantasy Pieces: Metrical Dissonance in the Music of Robert Schumann
(New York: Oxford University Press, 1999), 156–71; Krebs, “The Expressive Role
of Rhythm and Meter in Schumann’s Late Lieder,” Gamut 2, no. 1 (2009), Special
Feature—A Music-Theoretical Matrix: Essays in Honor of Allen Forte (Part I),
267–98, http://trace.tennessee.edu/gamut/vol2/iss1/9/; Krebs, “Meter and
Expression in Robert Schumann’s Op. 90,” in Rethinking Schumann, ed. Roe Min
Kok and Laura Tunbridge (New York: Oxford University Press, 2010), 183–205;
and Krebs, “Fancy Footwork: Distortions of Poetic Rhythm in Robert Schumann’s
Late Songs,” Indiana Theory Review 28 (2010): 67–84.

Chapter Eight
Parlante Talk
Texture and Formal Function
in the Operas of Verdi
Steven Huebner
Thanks to important work by Julian Budden, Scott Balthazar, Alessandro

Roccatagliati, and, especially, Harold Powers in the 1980s, Abramo Basevi’s
Studio sulle opere di Giuseppe Verdi (1859) has assumed a major place in reflec-
tions on ottocento musical dramaturgy.1 Basevi, a critic and composer himself,
established the basis for a Formenlehre applicable to the repertory from Rossini
to Puccini by providing vocabulary for analysis and outlining parameters for
the organization of set pieces. While he did not set out to write a rigorous
treatise, his book supplied enough guideposts to future students of the reper-
tory to develop his observations, especially as related to the solita forma (usual
form) of ensemble pieces in the sequence of tempo d’attacco, adagio, tempo di
mezzo, and cabaletta.2 His study gave authentic credentials to later and more
systematic analytical methods, an authenticity that some have understood as
mere veneer because Basevi himself privileged descriptive accounts and ad
hoc observations.3 In this view, the urge to explicit exposition of composi-
tional reflexes or to the plotting of variants against an implicit conventional
grid emanates more from the twentieth-century academy than from the nine-
teenth-century theatrical world. Yet the same might be said of most critical
routines informed by primary sources applied by scholars today: authenticity
is inherently a fragile concept.4
Basevi did, however, give extended analytical attention to one stylistic fea-
ture of Verdi’s operas: parlante texture. It receives the most sustained treatment
of any technical term in his book, perhaps objectively not very long (about
seven hundred words in two places) but certainly elaborate in the context of
his usual approach. And, as Basevi’s present-day editor tells us, he used the
term sixty-seven times in the course of the Studio.5 Despite its relative weight

PARLANTE talk 253
and ostensible authenticity, Basevi’s discussion of parlante had just as small an

immediate impact on methods of understanding the syntax of ottocento opera as
did his other categories of analysis. The first twentieth-century Anglo-American
scholar to take up the term appears to have been Joseph Kerman in Opera as
Drama (1956), which, given the importance and fame of the book, made for an
auspicious resuscitation. Kerman positioned parlante in a discussion about the
evolution toward greater continuity in nineteenth-century Italian opera. Verdi,
wrote Kerman, “gradually reduced the extent of the recitatives and blended
them into a more continuous texture,” something accomplished not only
through “Gluckian methods” in recitative but chiefly by a “technique known
at the time as parlante.”6 Kerman continued by defining parlante as a texture
that “hold[s] together a considerable passage of advancing dialogue by means
of a systematic motivic ground-plan in the orchestra; the voices chime in with
low-grade counterpoint as best suits the verbal phrase or sentiment.” In short,
the orchestra “provides the musical coherence” in a texture that opened the
way for Italian composers to achieve continuity because it elided easily with
both recitative and voice-dominated, lyrical music. Roots of the practice lay in
eighteenth-century opera buffa, with examples ready to hand for many listen-
ers in the chattering figure heard in the second Finale of Le nozze di Figaro after
Susanna has emerged from the adjacent room (“Susanna son morta”) or the
moment in the act 1 Finale of Don Giovanni when the eponymous hero offers
refreshments to his guests (“Ehi, caffè! Cioccolate!”).
Having set the parameters, Kerman used parlante as a tool in his larger criti-
cal project. He took early ottocento composers such as Rossini and Bellini to task
for mechanistic spinning out of a kind of parlante bedeviled by an excess of
“standard regular patterns” that produce a merely stagnant orchestral frame-
work while the dialogue moves forward. Verdi made the technique “genuinely
dramatic,” by which Kerman implicitly meant responsive to changing nuances
in the verbal text or, if uniform, attentive to the dramatic purpose of setting a
consistent mood. He singled out the passage that culminates in Violetta’s out-
burst “Amami, Alfredo” in La traviata as well as the texture of the duel between
Cassio and Montano in Otello as exemplary instances of the technique, at least
with respect to their dramatic efficacy.
For David Kimbell, parlante is “a quasi-naturalistic declamation, commonly
a dialogue [that] is set against statements, restatements, interludes, variations,
and developments of an orchestral theme.”7 Harold Powers, for his part, pro-
posed the pithy formulation of “where the orchestra provides the continuity
and the voices are fitted in” in a discussion about Puccini that understands
this composer’s cultivation of “instrumental continuity with voice-over” as
an outgrowth of parlante texture instead of derivative of “some sort of endlose
Melodie.”8 Indeed, the idea of “fitting voices in” to an orchestra that assumes the
lead is inherently inimical to organic Wagnerian thinking, just as is Kerman’s
idea of vocal “low-grade counterpoint.” Rather, it suggests an independent,

254 steven huebner
form-defining role for the orchestra instead of symbiotic entwining with the
voice. Such chronological extensions of the practice to Puccini provide incen-
tive to come to a fuller understanding of parlante texture because they touch
upon wider stylistic issues related to the impact of Wagnerian music drama
in Italy. That impact is undeniable, but the long tradition of parlante writing
reminds us that Italian opera evolved independently as well.
I will argue that concepts such as Kerman’s “systematic motivic ground-
plan” and orchestral “coherence” call for further exegesis that draws on
considerations of phrase structure and formal function. Phrase structure in
parlante texture, as I understand it, often operates on a premise related to what
has been called, with reference to another texture, the “lyric prototype”: lyrical
voice-dominated melodic writing.9 This prototype has clear parallels to phrase
structure in Viennese classical instrumental music, so effectively discussed by
William E. Caplin; an analytical methodology pertinent to the “lyric proto-
type” involves consideration of the strength of cadences, the presentation and
fragmentation of materials, and the phenomenon of return. Caplin’s work is a
vital stimulus to this project—indeed to all studies of phrase structure in tonal
music—because of its systematic approach to cadential progression and har-
mony more generally. Yet writing for voice in nineteenth-century Italian opera
necessarily also calls for somewhat different analytic criteria than writing for
instruments in eighteenth-century Vienna, particularly in assessing cadential
strength (for example, the sounding of the third scale degree in the voice in
the final tonic chord of a perfect authentic cadence undermines a sense of
closure in the “lyric prototype”).10 In my understanding, the extension of “lyric
prototype” elements to the orchestra in parlante texture is an important crite-
rion for identifying the latter—and one of the goals of this chapter is precisely
to develop a method for identification of this texture. This means imbuing the
concept of parlante with the same kind of form-functional thinking that I have
applied to the “lyrical prototype” elsewhere:11 one important characteristic of
parlante texture is that the formal functions articulating the larger structure
are projected mainly by the orchestra rather than by the vocal line. But phrase
structure is not the only distinguishing criterion: factors such as the amount of
voice-orchestra doubling and the length of the motivo also play a role. By con-
sidering these various criteria, we shall see that it is not even entirely clear that
the passage before “Amami, Alfredo” in La traviata that Kerman cites exempli-
fies the texture.
❧ ❧ ❧
Basevi’s main discussion of parlante occurs during lengthy asides in his begin-
ning-to-end descriptions of I Lombardi alla prima crociata (1843), an early opera
that followed quickly upon the heels of Verdi’s smash hit Nabucco (1842), and
La traviata (1853).12 In the second of these interventions Basevi describes four

PARLANTE talk 255
kinds of parlante. The leading orchestral figure might (1) capture the essence
of the dramatic situation, (2) express the emotions of a character, (3) graze the
ear (“solleticare l’orecchio”), or (4) assume the role of filler (“quelli che ser-
vono di ripieno”). The first two possibilities relate to projecting the drama, the
third to a more general aesthetic quality, and the fourth vaguely to form—or so
it would seem. Moving through these in reverse order: Basevi does not explain
what he means by “filler.” He does, however, use the expression parlante di ripi-
eno later in his study for an episode in the introductory double chorus to Les
vêpres siciliennes (1855) where French soldiers are set off against Sicilian peas-
ants: to a sprightly leading passage in the strings, some of the minor French
characters jovially gloat over how they intend to exploit the local population
(“Aussi dans ce pays”).13 Ripieno has a distinctly pejorative edge here, as Basevi
disapprovingly observes that this parlante does nothing to prepare the return
of the double chorus, which consequently reemerges completely by surprise
and unwanted (“che giunge inaspettata, e nulla desiderata”). “Filler” seems to
mean unnecessary and expendable, but says little about texture and syntax.
Basevi might have made more compelling use of the concept with reference
to highly conventional moments in the pre-Verdian repertory. I am thinking
here particularly of the so-called ritornello between statements (often between
solo and duet renditions) of a cabaletta melody, where the orchestra some-
times takes the lead as characters repeat text from the tempo di mezzo (as in
the cabaletta of Semiramide–Assur in the second act of Rossini’s Semiramide) or
introduce new dialogue (as in the cabaletta of Adalgisa–Pollione in the first act
of Bellini’s Norma). With respect to parlante that “grazes the ear,” Basevi writes
of passages in opera buffa that allow the spectator to concentrate on the words
and acting of the character to a pleasing accompaniment, an explanation that
either significantly detracts from the principle of the leading orchestra or per-
haps speaks to a delicate balancing act between voice and instrument. Fleshing
this out, one imagines that the attraction of this kind of parlante was its capacity
to provide a lively, effervescent and energetic background to comic high jinks
onstage—that is, to provide orchestral music not necessarily specific to the
action at hand but more generally applicable to the comic mode (for example,
the rapid-fire sixteenth notes with thirty-second note turn on the violins and
winds at “È Rosina, or son contento” in the first-act Finale of Rossini’s Il barbiere
di Siviglia). Basevi’s first two categories (projection of atmosphere and commu-
nication of a character’s inner feelings) are more tailored to specific context,
but a useful refinement for the former are situations where the leading orches-
tra belongs to the diegesis of the opera—for example, the dance music at the
beginning of La traviata or Rigoletto or the final scene of Un ballo in maschera—
in addition to functioning as a way to enhance the mood.
Basevi’s other discussion of parlante in the Studio focuses more closely on
texture.14 In the third act of I Lombardi, the lovers Giselda and Oronte resolve
to elope in an extended four-section duet. Basevi notes the existence of a short

256 steven huebner
parlante with “lively, spirited” (briosa, vivace) orchestration in the first section,
similar (he says) to the parlante in the duet between Abigaille and Nabucco
in the third act of Nabucco. In parlante, writes Basevi, “the motivo occurs in the
instrumental part instead of in the voice” (il motivo sta nella parte strumentale,
anzichè nella vocale) but, despite this, the orchestra should not detract from the
singer but rather increase his or her appeal much like “the rich royal vestments
that add new brilliance to the authority of the prince” (le ricche vesti reali, che
aggiungono nuovo splendore alla autorità del principe). In light of Basevi’s negative
comments about Wagner elsewhere in the book, one suspects that a defense
of italianità lurks behind his cautionary note about the relative importance of
voice and orchestra. Taking up a theme to which he returns often, and which
came to be viewed as a hallmark of Verdi’s style, Basevi praises Verdi’s conci-
sion (the parlante passage here fills twenty-three measures in a tempo d’attacco
that is sixty-seven measures long) while also casting aspersions on Donizetti’s
tendency to indulge too copiously in the texture.
Basevi continues by identifying different kinds of parlante. In par-
lante melodico, “while the orchestra unfolds the dominant motivo, the vocal
part . . . follows the said motivo at the unison or at the third or sixth” (mentre
nell’istrumentazione si svolge completamente il motivo dominante, la parte vocale . . .
seguita il detto motivo o all’unisono o di terza o di sesta). Sometimes it can be
difficult to distinguish between this kind of texture and “song accompa-
nied by instruments” (canto accompagnato dagl’istrumenti)—or what I will call
(lyrical) voice-dominated texture as distinct from declaimed recitative or
parlante—but, Basevi assures his reader, with careful reflection it is usually
possible to determine whether the motivo belongs principally to the voice
or the orchestra. Basevi coins the term parlante armonico for another kind
of parlante where “the vocal part does not have its own significant melody
but produces a kind of counterpoint to the motivo of the accompaniment”
(la parte vocale non ha melodia propria rilevante, ma fa quasi un contrappunto al
motivo dell’accompagnamento). As with parlante melodico, Basevi hints at certain
problems with the term around the edges: parlante armonico is difficult to dis-
tinguish from both parlante melodico, on the lyrical side of the continuum,
and recitativo obbligato, on the declaimed side. (The latter term, more appro-
priate to operatic repertories that have the contrasting category of recitativo
secco, implies an active orchestra that underpins declamation as well as an
avoidance of foursquare phrases and melodic repetition). In relatively lyrical
manifestations, the vocal part might follow the orchestra a fair amount, and
in more declaimed examples the “orchestra does not contain a motivo that is
very clear and developed” (non porta un motivo molto chiaro e sviluppato). In
light of the fuzzy parameters around parlante melodico and parlante armonico, it
is a little surprising that Basevi proposes a third category of parlante misto as a
combination of the first two, but he has nothing additional to say about it (di
cui non ha luogo intrattenerci maggiormente).

PARLANTE talk 257
As an illustrative example of parlante melodico, Basevi directs his readers

to the duettino between the thugs Rustighello and Astolfo in the first act of
Donizetti’s Lucrezia Borgia (1833), and for parlante armonico, he cites the tempo
d’attacco of the Giselda–Oronte duet in I Lombardi that provided the initial
trigger for his discussion. Example 8.1 shows the first half of Donizetti’s piece
and example 8.2 a Verdi duet similar to Giselda–Oronte according to Basevi
(Abigaille–Nabucco in Nabucco), but which for my purposes will provide a com-
pact illustration of some additional points.
Example 8.1. Donizetti, Duet Rustighello–Astolfo in Lucrezia Borgia, act 1
(continued)


PARLANTE talk 259
Donizetti’s entire parlante passage extends thirty-two measures. In addition to

its satisfying Basevi’s main parameter for parlante (that the main motivo be given
to the orchestra), it is noteworthy that the passage follows the “lyric prototype” of
ottocento opera—except, of course, for the fact that the orchestra leads. The six-
teen-measure type, normally subdivided into four-measure phrases, gets doubled
in length to a sequence of eight-measure phrases (ex. 8.1 shows the first half of
the melody). Taken on its own, the orchestral part nearly exactly corresponds to
Caplin’s category of the compound period. The text follows the normative pat-
tern as well—that is, two versi lirici quatrains (of ottonari) here increased to four,
each of which stretches out over eight measures.15 The voices follow the orches-
tral motive (which we might roughly describe as a dotted quarter followed by a
dotted sixteenth, usually after a leap upward), but only some of the time, actu-
ally for less than half the passage. For the most part, this motivo, scored darkly
for clarinet and bassoon with a pizzicato accompaniment, would be impossible
to detect were the sung dialogue to be heard by itself. Sounded at the begin-
ning in the voices (mm. 1–4), it disappears until the last measure of example
8.1. The limited presence of the main motive in the voice does not suggest the
equality of parlante melodico. It is striking that Basevi’s observation about the thin
line between parlante armonico and parlante melodico should apply even to a pas-
sage that he cites as a clear example of the latter. Considerations of formal func-
tion lurk behind this categorization, for I suspect that the orchestra’s lyric-form
design encouraged Basevi to the category of parlante melodico. A more consistent
pattern of doubling—here and in other examples of both parlante melodico and
armonico—would have been difficult for Donizetti (and later Verdi) to create in
view of another principle (not noted by Basevi) that plays out in a great many
parlante passages: melodic repetition in the orchestra tends to be accompanied
by new vocal rhythms. Whereas in the first measures the voices adhere to the
orchestra, the repetition (after m. 9) spawns vocal independence.
In the Nabucco chapter, Basevi calls the Abigaille–Nabucco parlante (ex.
8.2a) “short and lively” (breve e brioso) and claims that it is “fitted very stylishly
with the recitativo” (incastrato, con bel garbo, nel recitativo). The problem with
this observation is that Verdi positioned the parlante armonico to which Basevi
refers (m. 14 of ex. 8.2a) not after a recitative but after another parlante that
initiates the number, one that is coordinated with a shift to versi lirici (here
settenari). Perhaps Basevi felt that the number began with a recitativo obbligato:
patches of recitative do occur within the versi lirici borders of set pieces in the
repertory, and we will recall that he felt that obbligato could be difficult to
distinguish from parlante proper. Nevertheless, the beginning—strings pre-
dominate here—provides a symmetrical orchestral framework that strongly
indicates parlante as the prevailing texture.
At this point in the opera, Nabucco has gone mad. The slave Abigaille, pre-
sumed to be his daughter, informs him that she has usurped the crown and he
responds with incredulity:

Example 8.2a. Verdi, Duet Abigaille–Nabucco in Nabucco, act 3, “Donna, chi sei?”
(continued)

Example 8.2a.—(continued)
(continued)

262 steven huebner
Nabucco: Donna, chi sei?

Abigaille: Custode
Del seggio tuo qui venni! . . .
Nabucco: Tu? Del mio seggio? Oh, frode!
Da me ne avesti cenni?
(Nabucco: Woman who are you?

Abigaille: I came here as custodian of your throne!
Nabucco: You? Of my throne? Oh, fraud! Did I command you to do so?)
Verdi sets up the beginning of the duet as an eight-measure period (a

theme type heard often as the first half of a sixteen-measure “lyric prototype”)
based on a nervous turning figure that suggests Nabucco’s trepidation in his
encounter with Abigaille. The first four measures end with V/vi harmony and
measures 5–8 move to a perfect authentic cadence on the dominant. The
presence of dialogue here is typical of parlante in Verdi’s work. (The more
natural habitat for solo parlante was comic opera of the previous generation,
and indeed one of Basevi’s examples of parlante armonico is the famous move-
ment “Numero quindici” in the first act of Rossini’s Il barbiere di Siviglia where
Figaro repeats a single pitch seventy-three times above the leading orchestra.)
Also characteristic is the way Verdi follows the sense of the text in distribut-
ing it over the symmetrical phrase structure in the orchestra: he negotiates
the relatively soft line ending after “Custode” by pushing through to the next

Example 8.2b. Duet Abigaille–Nabucco in Nabucco, act 3, “Esci! invan mi
chiedi pace”

264 steven huebner
poetic line on a dotted rhythm, thereby working independently of the form-

functional properties of the orchestra by bridging over from the end of the
antecedent to the beginning of the consequent with the vocal phrase. The
measures right after the cadence on B♭ in example 8.2a (m. 8) illustrate just
how easily parlante can elide with music that takes on characteristics of voice-
dominated texture. Nabucco cannot contain his disbelief: over a B♭ pedal he
twice repeats his confused question (“Da me ne avesti cenni?”) and returns
to his protestation against Abigaille’s fraud, repeating “Oh frode!” The text
repetition and dactylic orchestral rhythm that build to the end of his phrase
are more typical of voice-dominated lyrical music. Abigaille will exert complete
control over the hapless Babylonian king, and it is she who steers the music
back to E-flat in order finally to crush her “father” with a sonorous and vocally
elaborate cadence later at measures 32–33.
On the way (at m. 14), she appropriates Nabucco’s bass B♭ to launch a par-
lante moment of her own to a figure that percolates with her sense of impend-
ing triumph. This is the music referred to by Basevi as parlante armonico,
analogous to his Giselda–Oronte example from I Lombardi. As with the initial
parlante, the passage is only eight measures long but now, instead of a period,
it consists of two identical phrases in the winds without a terminal cadence (an
open-ended form-functional situation in the “lyric prototype” that I have called
“balanced phrases”).16 Differences in formal function may of course be enlisted
for hermeneutic purposes. Here these distinctions match the rhetorical char-
acter of the poetry: Nabucco’s declaration of fraud (“O frode”) is an emphatic
reaction to Abigaille (hence, local-level closure), whereas Abigaille’s remark
that the Babylonian people have lost patience with the rebellious Hebrews
seems merely a preparatory salvo for her ensuing demand that Nabucco sign
the decree of their death (hence, the more open-ended “balanced phrases”).
Although Abigaille follows the winds quite closely in measures 18–19 (and
sings alone), the passage sounds like parlante because the orchestra initiates
it so conspicuously and the voice does not duplicate two of its most salient
motives (the syncopated leap in m. 19 and the turning figure in the next mea-
sure). This new parlante merges with voice-dominated writing based on an
orchestral motive derived from the first parlante, the kind of fluid interchange
of different parlante motives that extends back to Mozart. The entire passage
builds impressively to a high B♭ and cadenza-like flourish further to oppress
the prostrate king as Abigaille demands that he sign the Hebrews’ warrant.
The E-flat cadence that I mentioned earlier marks an emphatic close to the
first part of the tempo d’attacco. Example 8.2a skips over the subsequent mea-
sures and continues by showing the reprise of Abigaille’s parlante slightly fur-
ther on in order to demonstrate how vocal lines may be differently disposed
over repeated music in the orchestra.
The duet is idiosyncratic, however, inasmuch as the parlante figure sur-
faces yet again as Abigaille’s response to Nabucco in the cabaletta, a moment

PARLANTE talk 265
predictably praised by Basevi for economy of means but also one that illustrates
differences between parlante and voice-dominated music (ex. 8.2b). Abigaille
launches from a strong upbeat figure (“Esci! invan mi chiedi pace”) and then
is largely doubled by the orchestra throughout. Even though the listener has
absorbed this motivo as belonging to the orchestra, the fundamental principle
that I suggest is that in cases of consistent doubling the default approach in
Italian opera is to hear the texture as voice-dominated—not to mention that
the expert listener is programmed to expect both slow sections and cabalettas
to act in this way.
❧ ❧ ❧
Having filled out Basevi’s definitions of parlante, I continue by looking at fur-

ther examples of ambiguous edges around the texture, both as it abuts voice-
dominated writing and as it may resemble Basevi’s recitativo obbligato. Example
8.3 shows a part of the episode after the long Violetta–Germont duet in the
second act of La traviata: Violetta has just sealed the letter informing Alfredo
of her intention to return to Paris when he unexpectedly comes upon her
(“Che fai?”).
The dialogue builds to Violetta’s passionate outburst “Amami Alfredo.”
According to Basevi, Verdi “conceived the whole passage as a parlante” (è tutto
immaginato a guisa di parlante) and it will be recalled that Kerman also identi-
fied the passage before “Amami Alfredo” as parlante, perhaps taking his cue
from Basevi. Yet the case for application of this textural category to Alfredo’s
appearance seems weak. An agitato orchestral figure accompanies the rapid-
fire dialogue between Violetta and Alfredo, first above a decorated domi-
nant of B-flat major, then over a dominant of A-flat major/minor, and finally
through an enharmonic reinterpretation of a dominant-tonic progression in E
major. Speaking for the parlante side, one might recall that dialogue character-
izes the vast majority of parlante passages and that, in addition, here the orches-
tra unfolds a sharply profiled motive beneath the conversation. In Powers’s
formulation, the orchestra supplies “continuity” and the voices are “fitted in.”
But one might challenge with the question: fitted into what? For the figure is
also a cliché of agitato accompaniments, and really only a step away from a suc-
cession of block chords; it remains uniform throughout, as it might in a voice-
dominated aria. The music is tonally open-ended and does not unfold the kind
of symmetrical phrase structure we observed in the Nabucco example.
Notwithstanding Basevi’s description, then, it would seem most compelling
to describe the texture as allied to recitativo obbligato, for instead of truly leading
with an independent articulation of phrases, the orchestra merely provides an
accompaniment to this exchange. We might therefore formulate the following
guidelines: in passages where neither the voice nor the orchestra is organized
into foursquare phrases (or compressions and extensions thereof), where the

266 steven huebner
Example 8.3a. Scena, Violetta–Alfredo, La traviata, act 2, “‘Che fai?’ ‘Nulla . . .’”
orchestra remains fixated on a single motive of a measure’s length or less, and

where chord-to-chord voice leading seems more critical than melodic ductus in
the orchestra’s measure-to-measure evolution, parlante is generally less opera-
tive as a textural category than recitativo obbligato.
A related, if nonetheless distinct, example to which these guidelines might
be usefully applied occurs a little later in the gambling scene just after the play-
ers withdraw for supper, leaving Violetta alone with Alfredo (“Mi chiamaste?”;
ex. 8.4). The orchestral strings garner attention immediately by launching
a tense and agitated motivo that captures the essence of Alfredo’s distress at
Violetta’s recidivism, and a quick dialogue between the characters ensues.
Parlante figure or accompaniment? Unlike my previous example, in form-
functional terms the exchange between Violetta and Alfredo spools out in

PARLANTE talk 267
Example 8.3b. Scena, Violetta–Alfredo, La traviata, act 2, “Di lagrime avea d’uopo”
foursquare phrases and rhymed couplets for almost its entire length.
Symmetrical “balanced phrases” mark the beginning: Alfredo starts the first
four-measure phrase with a question. Violetta completes the phrase and starts
the second phrase with the same music heard earlier from Alfredo; this time
it is he who completes the phrase. For its part, the orchestra restricts itself to
a repeating one-measure figure. Although the orchestra is symmetrically orga-
nized (mm. 3–6 recur in 7–10), this pattern results from adhering to voices
that command attention because of their more elaborate melodic lines. Had
the orchestra presented material that extended beyond a single measure (as in
our Nabucco example), then the case for parlante texture here might have been
more robust. In this context, therefore, the first two measures do not sound
like an orchestra assuming phrase-structural leadership in parlante as much as
vamping to set up a voice-dominated dialogue.

268 steven huebner
Example 8.4. Finale, La traviata, act 2 (encounter between Violetta and Alfredo)
It is worth recalling that, from the start of his career, Verdi cultivated elabo-
rate and evocative accompaniment figures, often in closed slow sections and
cabalettas with vocal lines that might be highly declaimed in the manner of,
say, Rigoletto’s “Cortiggiani vil razza dannata” or Lady Macbeth’s sleepwalk-
ing scene “Una macchia.” The latter piece has little outright melodic repeti-
tion, but the voice still “leads” because if imagined on its own, it would strongly
articulate a succession of four-measure phrases. With a different layout of the
voice, say one that might have drawn attention away from its phrase-structural
regularities, could the orchestral figure for Lady Macbeth (characterized by a
one-measure module of shuddering strings and poignant ♭6–5 ostinato in the
English horn) have served as the leading strain in parlante? The same question
might be asked of the Violetta–Alfredo passage “Mi chiamaste” in example 8.4.
Perhaps, but, as I have intimated, most parlante passages in Verdi’s early and
middle-period work are driven by repeating orchestral material that extends
beyond a single measure to two and four measures, whereas repeating single-
measure figuration generally appears in the role of accompaniment. More
extended material for the orchestra has the very effect of drawing attention to
its leading status in the texture. In short, for textural nomenclature as it relates

PARLANTE talk 269
to music that operates in symmetrical phrases (or extensions and compressions

thereof), phrase-structural self-sufficiency of the voice or orchestra together
with consideration of the length of the orchestral motive works well as a crite-
rion for distinguishing voice-dominated passages from parlante ones.
In cases where the voice and orchestra both articulate phrases with equal
clarity, the default position should be to privilege the voice-as-leader in tex-
tural analysis. But not all cases are clear, especially in music that features a
succession of textures—common enough in the kinetic passages of early and
middle-period Verdi but even more thoroughgoing in Simon Boccanegra, and
then even more so much later in Otello and Falstaff. As part of his extended dia-
tribe against the first version of Simon Boccanegra, Basevi called the music with
which the Fiesco–Simone duet in the prologue begins (at “Qual cieco fato”)
“a kind of parlante melodico” (una specie di parlante melodico). Following this,
Simone gives voice to “a real cantabile” (un vero cantabile) of only eight mea-
sures (“Sublimarmi a lei”) followed by another “parlante” (“Io fea plauso”).17
Example 8.5 shows the beginning of this confrontation duet between the
upstart Corsair and the Genoese patrician, but leaves out the eight measures of
voice-dominated cantabile (“Sublimarmi a lei”) where Basevi’s textural catego-
rization is unproblematic.
At “Qual cieco fato” and again at “Sul tuo capo,” bassoons and cellos double
the voice for four measures. Or is it that the voice doubles the leading bassoon
and cello line? In an extended analysis of the conventional shapes that lurk
behind the unconventional surface of this duet, Powers accepts Basevi’s cat-
egorization of parlante melodico for these measures.18 According to the guide-
lines I suggest, however, the vocal part trumps the instrumental one in the
assignment of the default “lead” in cases of consistent doubling. Basevi pre-
sumably called these voice-dominated measures parlante by taking a bird’s-eye
view of the context. The orchestra does indeed launch both the antecedent
and consequent of an idiosyncratic but nonetheless symmetrical period. Given
the strength of the orchestral lead, our perception might be that the orches-
tra continues in this role as the voice gets added on. But counterbalancing
this consideration is the fact that the orchestral figuration and motive change
completely at the introduction of the voice in both the antecedent and conse-
quent, suggesting that the most compelling way to hear the orchestral explo-
sion is as a kind of instrumental refrain between passages of voice-dominated
texture, while also recalling that stretches of any scena or tempo di mezzo may be
governed by shifting textures and fluid transitions between these textures. The
real parlante here occurs at “Padre mio, pietà t’imploro” (against the leading
strain on the oboe, violin, viola), but it too slips into voice-dominated texture
after four measures (at “il perdono a me concedi”), and then a few measures
of recitative. Following the voice-dominated cantabile (not shown), the figu-
ration changes once again, but the rhythmic motive from the cantabile car-
ries on, and the texture still operates for six more measures, with the voice

Example 8.5. Duet Fiesco–Simone, Simon Boccanegra, prologue
(continued)


272 steven huebner
predominating by means of strong downbeats (at “Io fea plauso al tuo valore”).
As Powers observed, this is “hardly the simple parlante of Basevi’s descrip-
tion.”19 After a threefold succession of tonicizations (C♯ to E to G), Fiesco
settles on staunchly reiterated pitches as he continues to resist Simon’s offer to
suspend their hostility; continued downbeat articulation of the voice and lack
of motivic profile in the orchestra makes the texture sound like declamation
with orchestral accompaniment.
A similar analytical issue arises in another study where Powers considers the
Alzira–Zamoro duet in the second act of Alzira (1845).20 A clear parlante passage
driven by a two-measure agitato module in the orchestra (“Ah! l’ombra sua”)
leads to a voice-dominated a due cadence, much as in the tempo d’attacco of the
Abigaille–Nabucco duet that we examined. Notwithstanding the texture at the
cadence, for Powers this entire passage is parlante armonico. And in his view, the
subsequent music (“Qual mai prodigio”) continues with parlante melodico. Yet
not only is “Qual mai prodigio” set off from the voice-dominated a due cadence
by the performance indication pausa lunga, but the orchestra also doubles both
voices throughout this passage, suggesting the default position of voice-domi-
nated texture. One suspects that Powers chose the parlante designation because
the two characters remain in conversation, but I would suggest that the distinc-
tion between solo delivery and dialogue (whether declaimed or lyrical) not serve
as the sole criterion for separating parlante texture from other textures.
In a variant of the Alzira example, Amelia and Riccardo launch their love
duet in the second act of Un ballo in maschera with a breathless parlante that
merges with voice-dominated writing (and not parlante melodico as Powers per-
haps might have had it). Example 8.6a shows the parlante and the beginning
of the voice-dominated music at “Conte, abbiatemi pietà.” Unlike our Alzira
example, the parlante here does not build to an a due cadence: for the moment,
the two characters remain in conflict as Amelia spurns Riccardo and he insis-
tently pledges his love. In this parlante, the vocal parts double the leading mel-
ody on the flute and violin quite closely, and certainly could be imagined as
falling into symmetrical phrases on their own. Verdi thus comes close to voice-
dominated writing. But the orchestra still remains more continuous than the
vocal parts, particularly in the third and seventh measures of the excerpt, and
after the eighth. It also articulates the rhythmic motive, which incorporates a
weak-strong repeated note, more conspicuously than the vocal parts and devel-
ops the melody beyond a single measure, in fact into repeated four-measure
balanced phrases. Melodico seems the apposite qualifier, and more convincing
than Basevi’s Lucrezia Borgia example of this texture.
Later, in the tempo di mezzo of the same duet, Amelia breaks down to confess
her love explicitly to Riccardo in a justly famous passage where the cellos whis-
per a lyrical tune beneath her revelation (ex. 8.6b). After a beginning where
the strings (together with, momentarily, all winds) double the vocal parts, the
instruments issue the most complete rendition of the melody. Here one might

Example 8.6a. Duet Amelia–Riccardo, Un ballo in maschera, act 2, “Son la vittima
che geme”
(continued)


Example 8.6b. Duet, Amelia–Riccardo, Un ballo in maschera, act 2, “La mia vita . . .
l’universo”
(continued)


PARLANTE talk 277
plausibly argue that because a single motive runs through the entire pas-
sage, the parlante actually starts at the beginning of the example with a few
measures of melodico texture (“La mia vita”). Such an argument reads “back-
ward” from later in the passage, and cannot a priori be invalidated as an
analytical approach, but my own preference as an analytical premise is to
adopt an experiential orientation to trace textures as they unfold in time
in order to identify parlante. The merging of voice-dominated lyrical music
with parlante based on the same motivic material recalls a related procedure
that Verdi used increasingly in his later career, one heard just before the
duet in Amelia’s great aria “Ma dall’ arido stello divulsa,” where a com-
plete melody sung in one strophe gets taken up in the next strophe by the
orchestra (English Horn in Amelia’s piece) while the voice declaims above.
Example 8.6b continues with more parlante melodico: the very beautiful
cello melody that accompanies Amelia’s declaration (“Ebben si t’amo!”).
In a tour de force analysis of the second act of Un ballo in maschera, Powers
once noted that the cellos begin at “Ebben si t’amo” with parlante armonico
that becomes melodico at “Ma tu nobile.”21 I am more reluctant to micro-
track parlante to this extent and prefer instead to understand the whole
passage after “Ciel pietoso” (including both “Ebben si t’amo” and “Ma tu
nobile”) as parlante melodico, but my difference with Powers does illustrate
how the temporal perspective that one chooses to apply can affect distinc-
tions between parlante armonico and parlante melodico. At any event, what the
music at “Ma tu nobile” does not constitute is a return to voice-dominated
writing, because the continuity demonstrated by the leading cello strain
leaves the impression that it is the voice that grafts onto the instruments at
this point, and not vice versa.
❧ ❧ ❧
The close relationship of parlante with other textures in early and mid-
dle-period Verdi will be apparent from my brief survey and, perhaps not
surprisingly, this relationship becomes more pronounced for Otello and
Falstaff. One reason is that the general loosening of quadratic phrase struc-
ture in these late works means that an important form-functional param-
eter for distinguishing among textures is not as readily available: a grid of
phrases established principally by the orchestra that begin and end in the
same key and often have a connection, however loose, to the sixteen-mea-
sure “lyric prototype.” Another reason is that with the fading away (though,
it should be stressed, not the complete disappearance) of conventional
slow sections and cabalettas/strette, wider stretches of both operas became
available for the mixing and blending of textures, with attendant fuzzy

278 steven huebner
boundaries between them. Nevertheless, conventional moments of parlante

still do surface. In Otello, one of these occurs in the act 3 trio as Jago sidles
up to Cassio in order to coax Desdemona’s name to his lips, while Otello
bears hidden witness (“Vieni l’aula è deserta”). The violins clearly lead in
this eight-measure passage, but now the music sounds more like a classi-
cal sentence (with a two-plus-two presentation phrase followed by continu-
ation leading to a cadence) than the lyric-form prototype I have discussed
so far. The famous bacio theme near the end of the act one Desdemona–
Otello duet also unfolds in a parlante texture that suggests sentence form,
but now more distantly because the initial two-plus-two presentation occurs
over a chromatically descending bass (to settle on the dominant of the
supertonic). This example is related to the parlante in Amelia’s “Ma dall’
arido stelo divulsa” as an instance where the leading orchestral strain is
slow and lyrical, unusual in middle-period Verdi but much more common
in the late works, and then especially prevalent in the music of Puccini.
In Falstaff a relatively conventional moment of parlante sounds at the end
of the first part to act 2. A complete sixteen-measure lyric-form prototype
in the strings here mirrors the stiff, and somewhat comical, formality that
Falstaff and Ford adopt toward one another as they leave the inn. Other
parlante moments may be readily identified: the reprise of the Nanetta–
Fenton duet in the first act (“Torno all’assalto”: a lyrical orchestral melody
heard initially in voice-dominated texture), the lyrical melody spun out by
the English horn on Ford’s first appearance (“Signore, V’assista il cielo!”: a
continuously developed line with few internal symmetries), or the minuet
for the wedding scene at the end (“Già s’avanza il corteggio nuziale”: adher-
ing to the older tradition of dance music as the foundation of parlante, as in
the first act of La traviata).
But a great many other passages in Otello and Falstaff are open to debate.
In the scene for Quickly and Falstaff at the beginning of act 2, Kimbell has
argued that the motive associated with “Dalle due alle tre”—the hour of
Mistress Ford’s availability—provides the material for an extended parlante.22
It is true that the motive sneaks in and out of the texture, and recurs later in
the act, but to my ears it is integrated into the accompaniment or appears as
orchestral punctuation to relatively declamatory lines. There is little impres-
sion of the “Dalle due alle tre” motive as continuously spun out by an orches-
tra that firmly takes the lead and that might be profitably discussed with
respect to formal function.
For his part, Kerman also folded a discussion of late Verdian music domi-
nated by a single motive into his early pathbreaking discussion of parlante.
Example 8.7a shows the beginning of his example from Otello. To set the
scene: Jago has just engineered Cassio’s inebriation by leading a collective
drinking song, and Cassio now begins to scuffle with Roderigo. The poetry

PARLANTE talk 279
spools out in versi sciolti until a set of lines in doppio senari that begin at
Cassio’s “Nessun più ti salva!”:
Cassio (barcollando): Andiam!

Montano: Che vedo?
Jago (a Montano): (Ogni notte in tal guisa
Cassio preludia al sonno.
Montano: Otello il sappia.)
Cassio (come sopra): Andiamo ai baluardi . . .
Roderigo, poi Tutti: Ah! Ah!
Cassio: Chi ride?
Roderigo (provocandolo): Rido d’un ebro . . .
Cassio (scagliandosi contro Roderigo): Bada alle tue spalle!
Furfante!
Roderigo (difendendosi): Briaco ribaldo!
Cassio: Marrano!
Nessun più ti salva!
Montano: Frenate la mano,
[Cassio (staggering): Let’s go then!

Montano: What’s this I see?
Jago (to Montano, approaching him closely): Every night in like manner
does Cassio prelude sleep.
Montano: Othello shall know about it!
Cassio: Let’s go to the platform.
Roderigo, then all: Ah! ah! . . .
Cassio: Who laughs?
Roderigo (provoking him): I laugh at a drunkard!
Cassio: Defend yourself! (flinging himself at Roderigo) Scoundrel!
Roderigo (defending himself): Drunken rogue!
Cassio: Knave! No one can save you now!
Montano (separating them by force and turning to Cassio): Hold your
hand, sir, I pray you.]
Here Kerman celebrated “irregular parlante” that “shows how Verdi could
blend the texture into other elements of a scene.” He continued by observing
that the section is “developmental, organized by a clear harmonic movement
and by rather insignificant motivic material—which has however excellent dra-
matic interest in that it is derived from the Drinking Song.”23 Example 8.7b
illustrates one of the drinking-song source passages that Kerman seems to have
had in mind, taken from the beginning of the third verse, a moment when
Cassio, now inebriated, interrupts Jago before he can complete his portion

Example 8.7a. Otello, act 1: beginning of brawl that follows the drinking song,
“Ogni notte in tal guise”
(continued)

PARLANTE talk 281
of the music. The orchestral figuration in example 8.7a is similar. Imagined

without the triplet flourishes in alternate measures (and the woodwind trills
in every second measure, also derived from the drinking song but not shown
in the vocal-score reduction), it might function as a conventional chordal
accompaniment figure, albeit one with a grotesque touch provided by thick
downbeat trombone chords. This invites a question: were Verdi to have indeed
composed the music for the beginning of the brawl with the same orchestral
figure as in the third strophe of the drinking song, would that passage be heard
as parlante? Perhaps not—the motivo would not be “chiaro,” to employ Basevi’s
descriptor—but there are other parameters to consider. Although the passage
is developmental throughout (tracing a descending sequence from G-sharp
minor to F-sharp major to E major and then to D), the first eight measures of
the orchestra get repeated at lower pitch with slight variation (at “Ah! Ah!”),
followed by four measures of dominant preparation for D. A key marker of
parlante is that the declamation here maps differently onto each half of the

282 steven huebner
Example 8.7b. Otello, act 1: beginning of brawl that follows the drinking song,
“Fuggan dal vivido nappo i cordadi”
symmetrically organized orchestra in the first sixteen measures. The conversa-

tion even begins in the middle of a poetic line, and the second eight-measure
unit starts in the middle of one as well. Kerman’s epithet of “irregular” parlante,
then, applies only to the vocal part (and bearing in mind that, strictly speaking,
“irregular” parlante is rare because in most instances it is precisely “regularity”
at the level of the orchestra that distinguishes parlante). These factors combine
with the presence of a characteristic instrumental motive to produce parlante in
this excerpt, even recognizing that the distinction between “characteristic” and
“conventional” will sometimes be open to debate, as Kerman’s remark about
the “rather insignificant motivic material” of this parlante attests. The continu-
ous orchestra energizes the rapid-fire dialogue that culminates in Cassio draw-
ing his sword, a fine illustration of Basevi’s (and Kerman’s) principle of how
some parlante episodes can truly play to the heart of a dramatic situation.
Verdi often deployed parlante across long time spans in his middle-period
and late operas: all told, the parlante that occurs during Cassio’s brawl lasts
sixty-seven measures as it moves beyond the motive shown in example 8.7b
to new, although related, orchestral material. Other long parlante episodes in
Verdi’s oeuvre are the trio in the first act of Il trovatore (which includes an excit-
ing upward stepwise sequence of its main phrase), the gambling scene in La

PARLANTE talk 283
traviata (a large ternary form with nested lyric prototypes), the Amelia–Simon
duet in the first act of Simon Boccanegra (a complete lyric form parlante in the
tempo d’attacco, the motivic material of which is developed in a renewed tempo
di mezzo parlante), and the act 1 trio in Aida (which also features genuine devel-
opment of motivic material and where an extended agitato parlante animates
a parte parallel stanzas for all the characters, music that more conventionally
would have been set as a voice-dominated ensemble).
Despite the presence of other textures—including a voice-dominated
buildup to the final cadence, just as we observed in our Nabucco example—
parlante predominates in the last-mentioned example from Aida to the extent
that we might call it a parlante set piece. The locus classicus in Verdi’s oeuvre is
the Rigoletto–Sparafucile duet in the first act of Rigoletto, noteworthy as par-
lante armonico over a relatively lyrical strain in the muted low strings that pro-
duces the sinister complexion of the moment. Another example of a parlante
set piece is the act 3 trio in Otello where, over scherzando writing for the instru-
ments, Jago playfully persuades Cassio to reveal the handkerchief before the
concealed Otello (“Essa t’avvince coi vaghi rai”).
These pieces certainly deserve more extended reflection in the context of this
study, but to draw a contrast to the relatively small scope of my first (and paradig-
matic) example from Nabucco, let us turn instead to what must surely be the most
elaborate parlante passage in Verdi, the opening number of Falstaff (ex. 8.8 shows
a succession of excerpts from the piece). The spectator plunges into a dramatic
situation of total confusion as Dr. Cajus storms on with strident accusations of
skullduggery against Falstaff and his pitiful henchmen Bardolfo and Pistola.
Arrigo Boito wrote doppio settenari here, the longest line length Verdi
employed, and his poetry provided flexibility not only by that fact alone but
also in the numerous enjambments introduced to bridge over the mid-verse
caesurae and line endings. For example, two irregularities near the beginning
are the bridging-over of the caesura after the seventh syllable in the very first
line (the seventh syllable falls on the first syllable of “Falstaff” in Cajus’s “Sir
John Falstaff!”) and the very soft syntactical break after “bottiglia” (which goes
with “Di Xeres” of the next line). Yet Boito also adheres to the “opposite” pole
of lyric construction by using an absolutely regular succession of rhyming cou-
plets—that is, a scheme with close-to-the-surface audibility:
Dr. Cajus: Falstaff!

Falstaff: Olà!
Dr. Cajus Sir John Falstaff!!
Bardolfo: Oh! che vi piglia?!
Dr. Cajus: Hai battuto i miei servi! . . .
Falstaff: Oste! Un’altra bottiglia
Di Xeres.
Dr. Cajus: Hai fiaccata la mia giumenta baia,
Sforzata la mia casa.

284 steven huebner
Falstaff: Ma non la tua massaia.

Dr. Cajus: Troppa grazia! Una vecchia cisposa. Ampio Messere,
Se foste venti volte John Falstaff Cavaliere
Vi sforzerò a rispondermi.
Falstaff: Ecco la mia risposta:
Ho fatto ciò che hai detto.
Dr. Cajus: E poi?
Falstaff: L’ho fatto apposta.
[Dr. Cajus: Falstaff!

Falstaff: Ola!
Dr. Cajus: Sir John Falstaff!!
Bardolfo: What’s the matter?
Dr. Cajus: You beat my servants!
Falstaff: Innkeeper, another bottle of Xeres
Dr. Cajus: You have beaten my mare and broken into my house.
Falstaff: But not your housekeeper
Dr. Cajus: Thank you very much. An ugly hag. Ample Sir, if you were Sir
John Falstaff twenty times over I would force you to answer me.
Falstaff: Here is my answer. I have done as you have described.
Dr. Cajus: And so?
Falstaff: I did it all intentionally.]
The music at the beginning of Falstaff gives the impression of anything

but regularity, as a strong initial weak-beat accent in the first measure catches
the listener off guard and undermines the sense of meter (ex. 8.8a). James
Hepokoski hears an initial three-measure phrase at the local level;24 the next
cadence occurs on the downbeat of the seventh measure. A quarter-century
before, Verdi might have put a quadratic orchestral grid in place here for par-
lante. Taking into account this phrase-structural convention, we might adjust
our perception to imagine a chord on the first beat of the first measure (elimi-
nating the second-beat accent) together with an elision of the first phrase (now
four measures, including the downbeat of the fourth) with the second four-
measure phrase (that is, using the last chord of the first phrase as the launch-
ing point for the second). One might think of what Verdi actually wrote as a
collapsed accordion of eight measures reduced to seven, a beginning that, in
a phrase-structural sense, gets “tripped” and knocked immediately off stride,
appropriate to the flustered and confused ranting of Dr. Cajus. Indeed, when
the figure returns at the original pitch level later in the number (“Vi smen-
tisco!”), the music parses as a two-plus-two presentation phrase followed by
eight measures of motivically derived continuation (see ex. 8.8c). Following
the irregularities of the brief orchestral prelude, the voices bring a more sym-
metrical two-plus-two instrumental sequence, but now it is the poetry that is
unevenly divided (with the bridging of the caesura in the first line on “Sir

Example 8.8a. Falstaff, act 1: opening scene, Vivace, “Falstaff!”

Example 8.8b. Falstaff, act 1: opening scene, “Ho fatto ciò ch’hai detto”

Example 8.8c. Falstaff, act 1: opening scene, “Vi smentisco!”
(continued)

Example 8.8c.—(continued)
(continued)

PARLANTE talk 289
Example 8.8c.—(concluded)
John Falstaff!” that I mentioned earlier). Further phrase-structural ambigu-

ity ensues at measure 11 of example 8.8a and results in a metrical obfusca-
tion of the rhyme between “piglia” (downbeat) and “bottiglia” (mid-measure
and mid-pattern). A new orchestral pattern springs up in measure 14, based
on a slight extension of measure 11 (that is, the four sixteenth-note figure
derived from the first measure followed by a two eighth-note tag). One might
expect the material of measure 14 to work its way to a point of articulation
four measures later, but instead it gets stopped in its tracks with a prolonged
rest on the last beat of measure 16. In short, the initial motive gets reworked
in a variety of ways, and in this respect the first scene of Falstaff seems like a
culmination of developmental strategies already heard in earlier pieces such
as the Amelia–Simon duet, the Aida trio, or the drinking song and brawl of
Otello. But now quadratic phrase structure has faded much more into the
background, a mere point of reference. Yet the leading status of the orches-
tra is never in doubt, and the conversation occurs at a rapid pace, just as it so
often does in earlier parlante armonico.
To fill out the number, Verdi adds a second parlante figure in E major at mea-
sure 26 (shown in ex. 8.8b). In contrast to the first motivo, its initial presenta-
tion is utterly foursquare and even domesticates the sixteenth-note motive heard
since the beginning as an ancillary contrapuntal flourish on the fourth beat.

290 steven huebner
Julian Budden aptly describes the passage as an illustration of Falstaff’s corporeal

immensity, and in the music that follows, this motivo stands for the old knight as
much as the quirky first figure seems to represent Dr. Cajus. Budden also moves
up a form-functional level by arguing that Falstaff’s material acts as a second
theme in an implied sonata-form movement, the procedure evoking the absent
convention of the operatic overture.25 For Roger Parker the number is a “kind
of mock sonata form.”26 The argument for a sonata-form design rests on the
presence of two motivi, subsequent extensive development of both of them, and
a return to the initial material in the tonic (ex. 8.8c), as well as a reprise of the
second parlante motivo in C major. But the case for sonata form is unconvinc-
ing. We have already observed the presence of multiple parlante figures within
single numbers in the tradition. As far back as our Abigaille–Nabucco duet
example, each of the figures may be linked to a character. And a recapitulation
of an E-major “second theme” in the C-major tonic seems a dubious proposi-
tion. Example 8.8c illustrates the continuation phrase to the first subject after its
“recapitulation” to the point where it joins the second parlante motivo. The music
actually heads in the direction of E major, so that the C-major setting of the sec-
ond motivo emerges as something of surprise when Falstaff calls for Bardolfo
(“Bardolfo! Chi ha vuotate le tasche”). The move to this key is actually a stepwise
jacking-up from the last unadorned statement of the second parlante motivo in B
major (not shown in exx. 8.8a–c), where Falstaff addressed his other henchman
Pistola. That is, a temporally disjunct sequencing of the same motivo links two
similar dramatic actions (Falstaff calling out for his sidekicks). The C major at
“Bardolfo!” seems a result of this local-level process, and following this moment,
the second parlante figure does indeed turn to the initially prepared E major
(also not shown in ex. 8.8c), now in augmented note values.
As a parlante set piece, the number is unusual in that it adroitly combines
two parlante figures exposed in different keys, but conventional inasmuch as
by the end of the number, each figure returns to its own tonic (C major, E
major) in the manner of earlier (single motivo) parlante set pieces such as the
Rigoletto–Sparafucile duet or the Aida trio. Certainly, much of the center of
the piece is taken up by development of the second parlante, but the manner
here is very much as in the Otello brawl, with quadratic phrases and modulatory
sequences along a circle of fifths. Verdi’s combination of the two parlante fig-
ures leaves every impression of a playful comedic spirit, as does his engineering
of Dr. Cajus’s remark “Non è finita” as the textual trigger for continued devel-
opment, an emblem of comic excess. In a kind of coda to the piece, after the
recapitulatory moments I have just discussed, C-major and E-major chords con-
tinue to jostle one another, the latter functioning as the dominant of A minor.
Playfully, once again, Verdi brings his parlante number to a close with Pistola
and Bardolfo indulging in a bit of dissonant counterpoint on “Amen”: Pistola
starts off from C to outline V/V in that key, but Bardolfo enters in imitation at
the ninth with V/V of D (that is, a linear outline of an E7 harmony).

PARLANTE talk 291
The use of two keys, then, results from a sophisticated and ludic approach
to parlante rather than from the integration of a model borrowed from instru-
mental music; it is eminently operatic inasmuch as each of the keys attaches
to one of the feuding characters on the stage. Emanuele Senici has written
evocatively about the historicism of Falstaff, to which the third-act minuet and
final fugue bear ample witness.27 The pertinence of historicism to the initial
number would relate not to an “academic” form derived from instrumental
music but rather to a way of writing with deep roots in the Italian operatic tra-
dition itself. Those roots include an approach where the orchestra consistently
carries the most complete versions of leading motives and where symmetrical
phrase structure operates either explicitly or as a background model.
Notes
1. See the recent critical edition Abramo Basevi, Studio sulle opera di Giuseppe Verdi,
ed. Ugo Piovano (Milan: Rugginenti, 2001). All translations from Basevi are my
own. For prominent references to Basevi’s book, see Julian Budden, The Operas of
Verdi, 3 vols. (London: Cassel, 1973–81); Scott Balthazar, “Evolving Conventions in
Italian Serious Opera: Scene Structure in the Works of Rossini, Bellini, Donizetti,
and Verdi, 1810–1850” (PhD diss., University of Pennsylvania, 1985); Balthazar,
“The Forms of Set Pieces,” in The Cambridge Companion to Verdi, ed. Scott Balthazar
(Cambridge: Cambridge University Press, 2004), 49–68; Harold S. Powers, “‘La
solita forma’ and ‘The Uses of Convention,’” Acta musicologica 59 (1987): 65–90;
Powers, “Basevi, Conati, and La traviata: The Uses of Convention,” in Una piacente
estate di San Martino: Studi e ricerche di Marcello Conati, ed. Marco Capra (Lucca: LIM,
2000), 215–35; Alessandro Roccatagliati, “Le forme dell’opera ottocentesca: il caso
Basevi,” in Le Parole della musica, vol. 1 of Studi sulla lingua della letteratura musicale in
onore di Gianfranco Folena, ed. F. Nicolodi and P. Trovato (Florence: Leo S. Olschki,
1994), 311–34.
2. Piovano’s excellent edition italicizes all technical vocabulary, allowing for quick
identification of Basevi’s usage.
3. For this perspective see Roger Parker, Leonora’s Last Act: Essays in Verdian Discourse
(Princeton, NJ: Princeton University Press, 1997), 42–60.
4. For a development of this point, see Steven Huebner, “Structural Coherence,”
in Balthazar, The Cambridge Companion to Verdi, 139–53. See also Peter Schubert,
“Authentic Analysis,” Journal of Musicology 12 (1994): 3–18.
5. Basevi, Studio, 125.
6. Joseph Kerman, Opera as Drama, rev. ed. (Berkeley: University of California Press,
1988), 115.
7. David Kimbell, “Instrumental Music in Verdi’s Operas,” in Balthazar, The Cambridge
Companion to Verdi, 165.
8. Harold S. Powers, “Form and Formula,” Studi Pucciniani 3 (2004): 21.
9. Pertinent readings include Joseph Kerman, “Lyric Form and Flexibility in Simon
Boccanegra,” Studi verdiani 1 (1982): 47–62; Scott Balthazar, “Rossini and the
Development of Mid-Century Lyric Form,” Journal of the American Musicological
Society 41 (1988): 102–25; Steven Huebner, “Lyric Form in Ottocento Opera,” Journal

292 steven huebner
of the Royal Musical Association 117 (1992): 123–47; Giorgio Pagannone, “Mobilità
strutturale della lyric form: Sintassi verbale et sintassi musicale nel melodramma ital-
iano del primo Ottocento,” Analisi 7/20 (1997): 2–17.
10. In “Lyric Form in Ottocento Opera,” I broach such analytical criteria. Whereas most
other approaches to the prototype deploy an alphanumeric system to label phrases
largely on the basis of melodic content, my approach considers cadence and har-
mony as well.
11. Huebner, “Lyric Form.”
12. The passage in the La traviata discussion appears only in the first version of the
Studio published in the journal L’Armonia in 1858. See Basevi, Studio, 291–92.
14. Ibid., 136.
15. Versi lirici refers to Italian poetry organized into consistent rhyme schemes and stan-
zas, as opposed to the more irregularly organized versi sciolti. For a short discus-
sion of Italian verse forms, see Paolo Fabbri, “Metrical and Formal Organization,”
in Opera in Theory and Practice, Image and Myth, ed. Lorenzo Bianconi and Giorgio
Pestelli (Chicago: University of Chicago Press, 2003), 200–209. For more detail, see
Anselm Gerhard and Uwe Scheikert, Verdi Handbuch, 2nd ed. (Stuttgart: Metzler,
2013), 201–22.
16. See Huebner, “Lyric Form,” 127.
18. Powers, “Form and Formula,” 17.
19. Ibid., 18.
20. Harold S. Powers, “Verdi’s Monometric Cabaletta-Driven Duets: A Study in Rhythmic
Texture and Generic Design,” Il Saggiatore musicale 8 (2000): 281–323.
21. Harold S. Powers. “‘La dama velata’: Act II of Un ballo in maschera,” in Verdi’s Middle
Period 1849–1859: Source Studies, Analysis, and Performance Practice, ed. Martin Chusid
(Chicago: University of Chicago Press, 1997), 309.
22. Kimbell, “Instrumental Music in Verdi’s Operas,” 165–67.
23. Kerman, Opera as Drama, 116.
24. James Hepokoski, Giuseppe Verdi: Falstaff (Cambridge: Cambridge University Press,
1983), 1.
25. Julian Budden, The Operas of Verdi, rev. ed. (Oxford: Clarendon Press, 1992), 3:447.
26. Roger Parker, Leonora’s Last Act, 112.
27. Emanuele Senici, “Verdi’s Falstaff at Italy’s Fin de Siècle,” Musical Quarterly 85
(2001): 274–310.

Part Five
Analysis and Hermeneutics

Chapter Nine
Discipline and Punish among

the Winds in the First Movement
of Beethoven’s First Symphony
Henry Klumpenhouwer
Example 9.1 provides a condensed score for measures 53–88 of the first move-
ment of Beethoven’s First Symphony. The passage represents the second theme,
which Donald Francis Tovey considers a “small and comic sonata for orches-
tra.”1 In 1838, Berlioz wrote that the symphony as a whole, “by its form, by its
melodic style, and by its sobriety of harmony and instrumentation, is altogether
distinct from the other compositions of Beethoven by which it was succeeded.
The composer evidently remained in course of writing it [sic], under the influ-
ence of Mozart’s ideas; which he sometimes enlarges, and everywhere imitates
with ingenuity.”2 He writes of the second theme in particular: “by means of a
half-cadence repeated three or four times, we arrive at an instrumental design
in imitations at a fourth; our astonishment at finding which in such a place
is increased by the fact that the same design has been often employed in the
overtures of several French operas.”3 Setting the agenda for a number of more
recent studies of the first movement of opus 21, Hermann Kretschmar, in his
Guide to the Concert Hall, writes that the “second theme is pure Mozart. The
jubilant Nachgesang that follows appears verbatim in the ‘Jupiter’ and other
symphonies of the Master of Salzburg.”4 Berlioz’s and Kretschmar’s linkage to
Mozart in general and Kretschmar’s connection to Mozart’s Symphony no. 41
in C Major, K. 551 (the “Jupiter”) in particular have been repeated by Elaine
Sisman, who also argues that the First Symphony recalls elements not only of
the “Jupiter” Symphony but also of Haydn’s Symphony no. 97. Carl Schachter’s
comparison of the modulatory plan in the development sections of the First
Symphony and the “Jupiter” Symphony clearly belongs to this tradition, as
does Daniel Heartz’s discussion of the movement.5

Example 9.1. Beethoven, Symphony No. 1 in C Major, op. 21, mvt. 1, mm. 53–88
(continued)

(continued)

(continued)

discipline and punish among the winds 299
The question of Beethoven’s psychological dependence on Haydn and

Mozart is certainly an interesting one, but we shall set it aside for now to focus
on certain structural and formal features of the theme. The passage is discussed
by James Hepokoski and Warren Darcy in connection with the general prob-
lem of distinguishing the second thematic section from the closing section.6
In their view, second themes end with the first satisfactory perfect authentic
cadence in the second theme’s key (an “essential expositional closure” in their
terminology). In the example at hand, they argue that the cadence in measure
77 does not qualify as the essential expositional closure because the G-major
tonality is immediately unseated on the second quarter of measure 77 by G
minor, and because the character of the music that begins there has the effect
of undoing the closing power of the cadence. In Hepokoski and Darcy’s read-
ing, expositional closure is prepared, carried out, but then denied, a process
that represents the last in a series of denials of formal convention present ear-
lier in the second theme. They read the theme itself in Satz or sentence form,
whose presentation cycle features a period. The continuation module (which
Erwin Ratz sensibly calls Entwicklung)7 begins in measure 69 and is made pos-
sible or, better still, necessary, by the half cadence in measure 68, which denies
a conventional consequent in measures 61–68.
In the analysis by Hepokoski and Darcy, the theme begins with a conven-
tional antecedent of eight measures, which they parse into Satz structure: pre-
sentation cycle (mm. 53–56), continuation (mm. 57–58), and cadential gesture
(mm. 59–60). It is worth observing that the antecedent is itself characterized
by a strong antecedent–consequent sensation within its own presentation cycle,
between the basic idea (mm. 53–54) and its transformation (mm. 55–56).
Accordingly, we can extend the hyper- or metaphrase structuring observed
by Hepokoski and Darcy—namely, that the second theme forms a Satz, whose

300 henry klumpenhouwer
presentation cycle instantiates a periodic structure—to the structure of the

Satz-shaped antecedent and consequent of the hyper-presentation cycle itself.
Table 9.1 helps to clarify this structure. Along the top, the figure parses the
music from measures 53 to 88 according to Satz structure. The components of
the Satz individually also follow the Satz schema, and the figure numbers the
distinct component Sätze for reference: Satz 1 extends from measures 53 to 60;
Satz 2, from measures 61 to 68; Satz 3, from measures 69 to 77; and Satz 4, from
measures 77 to 88.
We have already noted that according to Hepokoski and Darcy, the first
two Sätze constitute the antecedent and the consequent of a conventional
period and, taken together, form the presentation cycle of the large Satz that
constitutes the second theme. Satz 3 (Kretschmar’s “jubilant Nachgesang . . .
that appears verbatim in the ‘Jupiter’”) functions as its continuation module.
Satz 4 represents a formally unconventional element, designed, in Hepokoski
and Darcy’s account, to deny the closure of the second theme in measure 77.
The structure presented in table 9.1 is immensely complex, because it fea-
tures Sätze and periods at alternating levels of organization. The complexity
emerges not only from the structural depth this pattern creates, but more
fundamentally from the profound differences between the two theme-scripts
themselves. Periods are symmetrical in design: they provide the sense of an
object and its dual, of a query and its reply. (However utterly conventional they
seem to us, periods are among the most interesting formal developments in
Western music: the script allows one to hear a stretch of previously unheard
music and then predict a second stretch of music of equal length that formally
closes off both stretches as a single unit.) The Satz, however, is entirely differ-
ent in design and orientation. It is not symmetrical. And whereas the period is
directed principally toward its midpoint at the end of the antecedent, toward
the moment when one has (more or less) formed an image of the entire conse-
quent, the Satz is directed toward its end. Generally speaking, one never devel-
ops an image of a particular Satz until it is over. Therefore, the period and the
Satz are not simply two distinct formal scripts for thematic structure. They form
two poles of a structural dynamic that governs something we might think of as
a thematic “process,” at least among “School-of-Haydn” composers.8
Hepokoski and Darcy’s analysis of the second theme must be read against
William E. Caplin’s earlier analysis of the passage. In “Structural Expansion in
Beethoven’s Symphonic Forms,” Caplin argues that the second theme (more
precisely, the “subordinate-theme group”) of the First Symphony instantiates
“the additive technique of stringing together a series of self-contained subor-
dinate themes, a procedure especially favored by Mozart” and contrasts this
technique with “more integrated” subordinate themes in the Third and Ninth
Symphonies.9 Caplin parses measures 53–88 into two distinct subordinate
themes: the first subordinate theme corresponds to the first three Sätze in table
9.1; the second corresponds to Satz 4. In Caplin’s analysis, measures 88–100

Table 9.1. Hepokoski and Darcy’s parsing of the second theme
Satz
Presentation cycle Durchführung-Cadence
Satz 1 (mm. 53-60) Satz 2 (mm. 61-68) Satz 3 (mm. 69-77) Satz 4 (mm. 77-88)
Period
Antecedent Consequent
Pres. cycle Durchf.-Cad. Pres. cycle Durchf.-Cad. Pres. cycle Durchf.-Cad.
Period Period
Ant. Cons. Ant. Cons.
9/30/2015 7:53:34 PM
constitute a third subordinate theme, one that Hepokoski and Darcy label
as a closing theme (or more precisely, “a generically standard P-based
C1-theme,” which is to say, a closing theme based on fragments of the first
theme).10
There are many ways in which Hepokoski and Darcy’s analysis differs quite
dramatically from Caplin’s: the relation of measures 53–77 to period and Satz
schemata; the structural meaning of the cadence in measure 68; the point at
which the second theme or subordinate-theme group concludes. Yet there are
also important ways in which the two analyses generally agree: both regard
the music as one large Satz whose components exhibit period and Satz form,
although they present different accounts of how the parts constitute the whole.
And both analyses regard the second theme as a narrative of denial, although
each presents its own account of the formal mechanics in play.
Setting these particular questions aside for the moment, I wish to focus on
the meaning of the music in measures 77–88. This is precisely the focus of
Caplin’s more recent comments on the second theme of the First Symphony
in the essay “What Are Formal Functions?” In the context of contrasting his
theory of formal functions with systems based on formal types, Caplin elabo-
rates upon certain features of his first analysis. He writes:
Though these bars [mm. 77–80] appear to “begin” the second subordinate
theme, they actually sound more medial in function, for they feature con-
tinuational characteristics such as sequential harmonies and repeated one-
bar units. What follows in mm. 81–83 brings cadential harmonies, but in the
wrong key. The theme finally ends with a genuine cadential function in mm.
84–88, culminating in a perfect authentic cadence. Thus while this theme
contains two of the three functions of the sentence form—continuation and
cadential—a clear functional beginning is actually missing, and so the theme
seems to start, in some sense, already in its middle.11
Before continuing our discussion, I wish to invoke in addition to the score

a particular performance of the symphony, because it interacts fruitfully with
the problematic at hand. I have in mind Willem Mengelberg’s performance
on April 14, 1940, with the Concertgebouworkest.12 There are a number of
interesting features in the performance, but I want to focus on Mengelberg’s
remarkable tempo choices at the end of Satz 3 and in Satz 4 in Hepokoski
and Darcy’s account—or, in Caplin’s account, at the end of the first subordi-
nate theme and in the second subordinate theme. Rather than recounting
Mengelberg’s execution of the passage myself, I invoke Richard Taruskin’s
description from Text and Act:
Both in the exposition and in the recapitulation, Mengelberg brusquely reins

the four staccato fortissimo chords preceding the passage [our Satz 4] so that
it starts way under tempo (in the low 90s by the metronome) and gradually

regains both the original tempo and loud volume over its whole eleven-mea-
sure expanse, played as a steadily mounting accelerando cum-crescendo.
(Beethoven marks the crescendo from pp to f over the last two-and-a-half bars
only.)13
Mengelberg’s performance certainly brings about a much stronger sense of

closure in measure 77 than Hepokoski and Darcy observe in the score.14 In
this way, his performance helps us appreciate Caplin’s efforts to argue a more
profound break with the previous music. Yet it would not be very interesting
to pose the analytical problematic purely as a matter of how best to parse the
formal divisions, as simply a question about where the second theme ends. A
more engaging way to begin is to reframe the question so that it addresses the
nature of the relationship between the music beginning in measure 53 and
that beginning in measure 77. However one parses the music here, one must
agree that Satz 4—Caplin’s second subordinate theme—is a personal rather
than a conventional element of the movement. This quality, I think, is what
motivates Mengelberg’s performance.15 He is not simply telling us where he
believes the second theme ends (though in my view this is one of the effects
of his articulations). There are two other effects I wish to mention. First, by
performing a grand ritardando into measure 77, Mengelberg creates time
for the tragic character of Satz 4 to emerge more convincingly than it would
through the sudden attenuation of volume alone. Without this extra time, the
brief turn to G minor is in danger of projecting a maudlin rather than tragic
character. The accelerando that follows—and which is made necessary by the
slowing down at measure 77—allows Beethoven’s crescendo to explode more
forcefully into the forte at measure 88.
So, whatever else it does, Mengelberg’s performance intensifies the
moment’s pathos. Yet it also affects the moment’s logos, because it gives the
sense that the function of the music at measure 77 is to pick up a loose end,
correct an error, or restore something lost in the preceding passage. From
this perspective, the slower tempo at measure 77 brackets it out from the
previous music.16 The accelerando (which Mengelberg initiates in small
increments around m. 80 but does not intensify until m. 83) suggests that
a successful correction has taken place, allowing the music to return to its
schedule at measure 88.
Let us pursue this angle a bit further. If there is a sense of correction or
restoration in the fourth Satz (which does not necessarily contradict the sense
of denial in Hepokoski and Darcy’s account), then we need to isolate a cor-
responding sense of error earlier in the music and describe its nature in terms
that relate strongly to the effect we have in mind with respect to Satz 4. Before
we do so, however, we ought to acknowledge an existing critical tradition that
singles out Satz 4 for special comment. Kretschmar, in his discussion cited at
the beginning of this chapter, writes of measures 77–88:

Immediately after that [i.e., after the Nachgesang, Satz 3, or the extended continua-
tion] however, Beethoven himself steps into the orchestra. It is the moment where
pianissimo takes over (m. 77), where the basses quietly ponder the initial motive of
the second theme, and the other instruments state dark and restless harmonies.
The oboe finds the exit from the eerie spell [unheimliche Verzauberung]. This is the
first time Beethoven brings the demonic element into the symphony.17
Sisman discusses measures 77–88 along very similar lines, analyzing the music
as an instance of “the ombra topic of supernatural operatic scenes.” Channeling
Kretschmar, she writes that in measure 77, “we enter a mysterious shadow
world in the low strings.”18
Generally speaking, both of these accounts regard the passage as Caplin and
as Hepokoski and Darcy do: as a denial of the previous measures. But unlike
these two accounts, they view the denial along topical rather than formal lines.
As a result, they provide no particular interpretive means to relate the excerpt to
the preceding music. In fact, both Kretschmar and Sisman regard the meaning
of the excerpt as directed primarily outside its local context: in Kretschmar’s case,
to other demonic elements in other symphonies by Beethoven; and in Sisman’s,
to the appearance of similar events in Mozart’s “Jupiter” Symphony, and in par-
ticular to what she calls “a swerve to the minor” in the “Jupiter” (eerily, also in
m. 77). So, while Kretschmar sees in the passage a breakthrough of Beethoven’s
own (demonic) personality after what he (Kretschmar) regards as a theme that
conforms itself to Mozart’s personality (this is what he has in mind when he
writes that “Beethoven himself steps into the orchestra”), Sisman regards the
passage as one element in Beethoven’s homage to Mozart’s “Jupiter” (and as we
discover, to Haydn’s Symphony no. 97 as well). What interests us here, however,
is to explore in the passage the sense of restoration or correction we detected,
which will allow us to integrate the passage with the second theme itself.
Example 9.2 initiates the exploration by focusing on the opening four mea-
sures of Satz 1, beginning in measure 53. This is the passage Berlioz had in
mind when he described an “instrumental design in imitations at a fourth,”
adding that “our astonishment at finding this in such a place is increased by
the fact that the same design has been often employed in the overtures of
several French operas.”19 The passage also brings to mind Hugo Riemann’s
discussion in his Catechism of Orchestration of what he describes as the “orches-
trational ideal of classic composers.” He writes:
The classics have gone a step farther [than previous composers], and besides
dividing up the tutti into the three groups of strings, woodwinds and brass,
have also made the single voices stand out from the tutti in variegated alter-
nation, not with pretentious soli, but with single motives fitting into the whole
with a most effective variety of expression. . . . The breaking up of the compact
massiveness of orchestral writing by means of an alternate use of instruments
is one of the fairest fruits of the classic art of instrumentation.20

Example 9.2. Satz 1, presentation cycle: antecedent, mm. 53–57
The notion that the presentation cycle of Satz 1 fragments the compact mas-
siveness of the music that precedes it is an interesting audible feature of the
second theme, especially because the mass coalesces again in the continuation
and cadence of Satz 1. This is not simply a matter of noting woodwind solos
and orchestral tuttis: what interests us in Riemann’s remarks is that he sees
the solos and tuttis as bound in a single dynamic of fragmentation and coales-
cence. The strings have the capability of presenting the entire orchestral mass
on their own, he says, because the timbres of the various instruments involved
are relatively similar. Solo woodwinds, on the other hand, have distinctive, indi-
vidual properties, which causes us to objectivize them (in contrast to the string
orchestra, which we easily “subjectivate”). But since solo winds produce sound
as we do, by means of air columns, their objectivity appears to us as human in
nature, so that we tend to regard them as characters or personalities—an idea
that will become central in our investigation.21
Example 9.2 isolates the oboe and flute parts and indicates the correspond-
ing harmonies underneath the staves with both function labels and Roman
numerals. We observed earlier in connection with table 9.1 that in this pre-
sentation cycle there is a strong sense of an antecedent/consequent relation-
ship between the idea in measures 53–54 and its transformation in measures
55–56.22 The feeling of period structure emerges from both harmonic and
melodic relations. The harmonic progression of the basic idea is T–D; that of
the transformed idea, D–T. The melodic structure of the transformed idea is,
roughly speaking, a transposition up a diatonic second of the opening idea’s
melodic structure. Accordingly, the passage relates strongly to the Meyer–
Gjerdingen “changing-note” schema, which in turn brings about the feeling of
periodic structuring we observed earlier.23 The continuation modulates (tem-
porarily) to D major.
Example 9.3 reproduces the corresponding elements in Satz 2, where
Hepokoski and Darcy observe the second theme’s first denial of closure,
brought about by concluding with a half cadence in G major at the end of
its continuation (though without modulating to D major, as Satz 1 does). In
other words, the second theme promises conventional period structure, but by

leaving the period harmonically open with a half cadence, leaves that promise
largely unfulfilled. In Caplin’s reading, the cadence in measure 68 is evaded
altogether, so that the music that follows functions as an “expanded cadential
progression” closing the grand period formed by Satz 1 and Satz 2 rather than
as the beginning of a new, relatively independent Satz.24 We might additionally
observe that the three cadences we have heard so far in connection with the
second theme—the medial caesura at measure 52, and the cadences of the
antecedent and the consequent at measures 60 and 68, respectively—produce
an interesting series of closes, each very different in character, but all avoid-
ing a full cadence on G, either by underreaching or by overreaching: a half
cadence in C major (G as Dominant) in measure 52, followed immediately by
G presented as Tonic (m. 53); a full cadence in D major (D as local Tonic in m.
60), and a half cadence or evaded cadence (depending on the reading) in G
major (D as Dominant in m. 68).
Let us turn our focus to the means (aside from cadences) by which Satz 2
carries out a denial of closure. Examining example 9.3, we see that the open-
ing idea in the consequent repeats the melodic structure of the corresponding
measures in the antecedent in example 9.2. This is of course purely conven-
tional. The orchestration, however, is different, which, in its own way, also
emerges from purely conventional thinking. As we will see, the imaginative or
personal elements in the passage have to do with the interaction between the
orchestration and the harmonic-melodic structure, not with any element on
its own. In example 9.2, the oboe initiates the melody and is answered by the
flute. In example 9.3, first and second violins present the melody, doubled at
the octave, and are answered by the flute and oboe, also doubled at the octave.
But the flute-and-oboe response at measure 62 is supported by the Parallelklang
of the subdominant, and not by the dominant, as it is in the corresponding
measure in example 9.2. The appearance of subdominant-functioning har-
mony, a function withheld altogether in the antecedent (Satz 1), reconfigures
the harmonic characteristics of the basic idea entirely, especially in connection
with the harmonies of the transformed idea in measures 63 and 64. In example
9.2, the symmetrical harmonic reversal between the idea and its transforma-
tion is polar or binary in design. In example 9.3, the corresponding symmetri-
cal harmonic reversal orbits around a center rather than alternates between
two (unequal) poles, Stufe I and Stufe V. The progression from the subdomi-
nant representative at the end of the basic idea in measure 62 (A minor) to
the dominant at the beginning of its transformation in measure 63 is a much
more powerful articulation of G as tonic than are the corresponding measures
in example 9.2, since by articulating G-major tonality’s harmonic boundaries,
the subdominant and dominant bring about the idea of G-as-tonic as the medi-
ating element. (Subdominant is to the tonic as the tonic is to the dominant.)
Pace Schenker, subdominant and dominant functions are more powerful
agents in creating a sense of tonic than are Stufen I and V.25

Example 9.3. Satz 2, presentation cycle: consequent, mm. 61–65
Yet G-functioning-as-tonic does not appear at the end of the transformed

idea in measure 64. Instead, Beethoven supports it with G-functioning-as-
dominant rather than G-functioning-as-tonic, as he did in the corresponding
measure in example 9.2. The moment in measure 64 also involves a prominent
melodic disturbance. In example 9.2, the flute’s reaction in measure 56 to the
oboe’s statement of the basic idea’s animating motive in measure 55 contin-
ues the melodic pattern proposed in measures 53–54; in measure 64, the flute
reacts to the violins’ statement in measure 63 by denying the pattern, by begin-
ning its statement with f♮3 rather than d3 as it did in measure 56. The flute’s
actions change the dynamics of the situation considerably. The changing-note
character of the antecedent’s presentation cycle is seriously weakened in the
presentation cycle of Satz 2. The flute’s new melody in measure 64 and the
particular harmonic function it brings about effectively negates rather than
echoes the violins’ statement in measure 63. By overturning the violin’s state-
ment of the motive beginning on A, replacing it with a statement of the motive
on F♮, the flute creates a new design that extends rather than balances and
consolidates the pattern of rising fourths from g2 (in m. 61) to c3 (m. 62) to
f♮3 (m. 64). In this account, the violins’ statement of the motive beginning
on A is heard as “crossed-out.” Example 9.4 collates these effects. The exam-
ple brackets the four forms of the animating idea: a G-form in measure 61, a
C-form in measure 62, an A-form in measure 63, and an F-form in measure 64.
The example indicates with an arrow that the first pitch of the C-form in mea-
sure 62 is determined by the final pitch of the G-form. By negating the A-form
of the motive, the F-form is able to continue two patterns begun by the G-form
and the C-form in measures 61 and 62. First, it takes its initial pitch (though
chromatically altered from F♯ to F♮) from the end of the C-form in measure
62. Second, the F-form completes a pattern of three rising fourths, beamed
together in the example, from one form to the next. We will have more to say
about the structure provided in example 9.4 later on.
In addition to the A-form of the motive in measure 63, something else has
been crossed out. Instead of joining the flute at the octave in measure 64 as

Example 9.4. Melody, Satz 2, presentation cycle: consequent, mm. 61–64
it should, the oboe goes missing altogether. The analytical question is: why?
Drawing on what we might call the dramatic content of the music at this point,
and bearing in mind Riemann’s idea that solo winds tend to represent person-
alities, the disappearance of the oboe in measure 64 takes on the quality of a
refusal. The oboe will simply not join the flute in disrupting the logic estab-
lished in the antecedent, and in negating the violins’ statement in the previous
measure. Perhaps it is because the music the violins play here is just the music
the oboe played in the presentation cycle of Satz 1. And, looking back to exam-
ple 9.2, we also recall that it was the oboe who established the motivic design
itself. In Satz 1, it taught the animating motive of the presentation cycle to the
flute and by extension to the violins in Satz 2. Now, in measure 64, its first pupil
has a different idea.
There is nothing in this account so far that cannot be coordinated with
Caplin’s or Hepokoski and Darcy’s assertion that the consequent contains a
structural denial, except to say that they limit the effect to the meaning of the
cadence of Satz 2. It is Satz 4, the music after the proposed essential exposi-
tional closure in measure 77, that provides a second cadential denial by replac-
ing G major with G minor, and from the perspective of pitch class, by replacing
B♮ with B♭. We observed earlier, in connection with Mengelberg’s perfor-
mance, that Satz 4 also projected a sense of correction or restoration. Let us see
if we can bring the dramatic content we have elicited earlier to bear on Satz 4.
Looking back to measure 77 in example 9.1, we notice that the bassoon begins
Satz 4 by restarting Satz 1 (or is it Satz 2?), this time in G minor. The antiphonal
character of Satz 1 and Satz 2 seems absent at first, but it emerges if we attend
to the bassoon in measure 77 and to the low strings in measure 78. In measure
79, the low strings begin the animating motive on F♮, which triggers a change
in direction. The overall result is that by the beginning of measure 79, the bas-
soon and low strings have recreated (though without the octave transfers) the
beamed pattern depicted in example 9.4, the logic of which we argued was
brought about by the flute’s actions in measure 64: we hear a G-form followed
by a C-form and an F-form of the animating motive. Here, the third statement
of the animating motive, its A-form in measure 63, which we asserted was effec-
tively crossed-out by the flute, is entirely absent. And in measure 79, just when
the low strings play F♮ (completing the summary of events isolated in ex. 9.4),
the missing oboe reappears (as a locally dominant instrument), playing f♮2,

the pitch it refused to play in measure 64. Moreover, the oboe plays f♮2 for a
very long time, more than twice as long as the time demanded by the structure
of the animating motive itself. The descending motive is further transformed,
from a descent through a fifth to a descent through a fourth.26 The elongation
of the first pitch has the effect of disturbing the Vierhebigkeit of the moment. In
measure 82, the oboe plays d2, the pitch negated by the flute’s F in measure
64. It is here that the sense of correction or restoration in Satz 4 takes a defi-
nite shape. The eleven unconventionally parsed measures (2+3+3+3) contain
four events pertinent to the idea of correction or restoration: first, there is a
return to the context in which the flute’s mistake (its expression of individual-
ity) takes place; second, we are presented with a synopsis of the effect on Satz
2 of the flute’s actions; third, the oboe articulates the precise nature of the
mistake (namely, playing an F-form of the animating motive); and fourth, the
oboe articulates, supported by the bassoon, the correct form of the motive, its
D-form. It is also worth observing that the role of the bassoon in Satz 4, gener-
ally speaking, is the role performed by the oboe in Satz 1, whereas the role of
the oboe in Satz 4 is the role played in Sätze 1 and 2 by the flute. In other words,
the oboe and the bassoon together reeducate the flute by modeling for it its
incorrect action and then indicating the correct one, a D-form of the motive.
The quiet dynamics (supplemented in Mengelberg’s performance by the slow
tempo) have the effect of bracketing Satz 4 out from the surrounding music, as
does the hypermetric shift to three-measure units. The dynamics also make a
connection to the presentation cycles of Sätze 1 and 2: all the other music in the
second theme is played loudly and is organized in conventional vierhebig units.
Once the correction has been made, the music returns to the locally normative
dynamic (and in Mengelberg’s performance, the locally normative tempo). The
disappearance of distinct thematic-melodic content in Satz 4’s continuation and
cadence modules after the oboe’s second statement in measures 82–84 (which,
formally speaking, completes the presentation cycle of Satz 4) certainly helps
project this idea. In a way, then, the second theme has a structural happy ending.
Whatever logic the flute disturbed has been reestablished.
Now our attention naturally turns to the second theme’s appearance in
the recapitulation, especially since it is conventional to regard the recapitu-
lation as the moment of restoration, correction or fulfillment of disrupted,
flawed, or promised features found in the exposition. In any case, according
to Anton Reicha, it is certainly conventional to do so for Beethoven’s sonata
forms. Linking Beethoven’s forms to the laws of classical poetics, Reicha
argues that the exposition represents the “tying of the knot,” while the reca-
pitulation, functioning as a denouement, represents its unraveling.27 In the
case at hand, however, the second theme in the exposition contains both the
mistake and its own restoration. As a result, any sense of correction or res-
titution in the recapitulation’s second theme would focus principally on its

Example 9.5. Satz 1, presentation cycle: antecedent, recapitulation, mm. 206–10
appearance in the tonic key. And that would be true whether or not the reca-
pitulated second theme still contained all the music related to the dramatic
content just discussed.
Example 9.5 explores the matter further. The music given is the presen-
tation cycle of Satz 1 as it appears in the recapitulation, shown along the
lines of example 9.2. It begins conventionally enough, by transposing the
presentation cycle of example 9.2 up a fourth. The harmonic progression
has changed from T–D//D–T to T–Sp//D–T, the progression associated
with the consequent in the exposition. The antiphonal character persists in
example 9.5, where the oboe’s music, suitably transposed, is played by the
reeducated flute doubled by the clarinet. The flute’s part in the exposition
is played here by the oboe and bassoon, which together carried out the
flute’s reeducation in the exposition. In measure 208, the clarinet drops
out of its partnership with the flute. This leaves the flute to play the form
of the motive—the D-form—it did not play in measure 64, where it trans-
formed the supporting G chord into the dominant of C major rather than
the tonic of G major. In this way, the clarinet’s missing music in measure
208 has an entirely different dramatic effect than did the oboe’s missing
music in the exposition. For one thing, the phenomenon occurs here in
Satz 1, the antecedent, and not in Satz 2, as it had in the exposition. For
another, it is not refusing to double the flute, because there is no reason
to do so: the flute is only playing what is conventional for it to play at the
moment. So the clarinet is withholding its music. After all, it did not com-
mit the error in the exposition: the flute did that alone. As a result, mea-
sure 208 serves as the last, humiliating moment of the flute’s reeducation.
Example 9.6, which parallels example 9.3, provides the presentation cycle
of Satz 2 in the recapitulation, the (transposed) site of the original action
in the exposition. The instruments involved here are just those involved in
Satz 1, except that the first violins replace the clarinet in pairing with the
flute. Furthermore, the violins, unlike the clarinet, are happy to support
the flute through the entire presentation cycle.

Example 9.6. Satz 2, presentation cycle: consequent, recapitulation, mm. 214–18
Given the dramatic context we have established so far, measure 217 is

quite remarkable. The flute’s error in the exposition appears again in the
recapitulation, suitably transposed. But this time, it is the oboe who commits it.
And this time, it is the bassoon, the oboe’s pedagogical associate, who refuses
to support its partner. As a result, we could imagine a version of example 9.4
transposed down a fifth. In this version, the crossed-out and negated third
statement is a D-form of the motive. Accordingly, the oboe’s error duplicates
the flute’s error in the exposition. Not only does it repeat in measure 217 a
transposed form of the initial error in measure 64, but in doing so, it specifi-
cally negates a D-form of the motive, the form it wished the flute had played in
the first place.
The oboe’s mistake also emerges from a very different dramatic motiva-
tion than the one committed by the flute in measure 64. In a sense, we might
imagine that the oboe has been distracted by playing (as it should) an F-form
of the motive in measures 207 and 215. Obviously, the oboe has on its mind
something like, “This is where the F-form belongs!” or “This is the form of
the motive that caused all that trouble earlier!” In the exposition, the flute’s
error seems to have emerged from its individuality, either refusing to construct
something conventional or not really knowing how. The oboe’s error here is
more serious: it is its own insistence on conventional logic (in this particular
context, the convention of transposing the exposition’s second theme in the
recapitulation) that brings about the very mistake committed earlier by the
flute (suitably transposed, of course).
Example 9.7 allows us to witness the oboe’s final humiliation in Satz 4. After
the low strings provide their synopsis of the logic brought about by the oboe’s
mistake, it is the oboe itself, on its own, that re-creates its error by playing the
B♭-form of the motive, suitably transformed from a descent through a fifth to a
descent through a fourth. (The reappearance of the low strings in this role has
the effect of drawing them into the dramatis personae, in the part of bailiffs
reading out the charges.) In measure 235, the oboe attempts to provide its own
correction. But it cannot. It continues to play f2, the final pitch of its descent,

Example 9.7. Second theme, recapitulation, mm. 230–38
rather than g2, the pitch it ought to have played in the first place. Perhaps it
cannot sort out what is right and what is wrong in its music at measure 217. Or
perhaps it has become disabled as it meditates upon f2, the pitch that initiated
the entire affair. Restoration is left to the bassoon, the oboe’s former teaching
assistant, and the clarinet, the two instruments that withheld music in exam-
ples 9.5 and 9.6.28
There are two questions in particular that interest us here: first, the flute’s
motivation for playing the F-form in the exposition; and second, the appearance
in the second theme’s recapitulation of the clarinet, a character not obviously
present in the exposition. Going back to example 9.3, we recall that in mea-
sure 62 an A-minor triad, functioning as the subdominant parallel of G major,
replaces the dominant of G major from measure 54, given earlier in example
9.2. We discussed how this alone produced a very different harmonic logic in
Satz 2 than in Satz 1. And we observed how the harmonic logic of Satz 2 in the
exposition also becomes the harmonic logic of Satz 1 in the recapitulation. But
the revised progression in example 9.3 also has a more local meaning in the First

Symphony: the symphony’s well-known slow introduction begins with three pairs
of chords. The final two pairs are just the progression we see in example 9.3.
And in the introduction, the final G-major triad is immediately supplemented
with an F♮, transforming the chord into the dominant of C major. So when the
flute plays the F♮-form in measure 64, it does so because the harmonic logic
in example 9.3 simulates the opening of the symphony (although in the intro-
duction the progression functions as D–Tp–D/D–D rather than as T–Sp–D–T
in ex. 9.3). The introduction begins by overreaching on the subdominant side
of C (C7–F), avoiding the tonic C by replacing it with its Parallelklang (again, A
minor), and then overreaching on the dominant side (D7–G). The dominant
orientation characterizes the harmonic work of Satz 1 in the exposition; the sub-
dominant orientation characterizes the harmonic work of Satz 2 in the exposi-
tion (and of both Sätze in the recapitulation), thanks in particular to the flute’s
error. Referring again to the dramatic content of Satz 2, we now understand that
the flute has been set up to make its mistake in measure 64.
So who is responsible for replacing D major with A minor in measure 62?
Not just one instrument, surely. Looking back to example 9.1, however, we
see in measure 62 that indeed it is one instrument alone that brings about
the change in harmony: the first clarinet, who accompanies the humiliated
flute in example 9.5 (the recapitulation of Satz 1), then abandons it in mea-
sure 208.
❧ ❧ ❧
We might ask how our reading reflects on the formal issues raised earlier in
connection with the analyses by Caplin and by Hepokoski and Darcy. On the
one hand, our reading seems to correspond better (but not conclusively so)
with Hepokoski and Darcy’s view of the second theme as four Sätze extend-
ing from measures 53–88 than it does with Caplin’s view of a group of three
subordinate themes extending from measures 53–100; yet we might also argue
that our reading is helped a great deal by Caplin’s understanding of the third
Satz in Hepokoski and Darcy’s as an expansion of the cadential progression
of the consequent (Satz 2) rather than as a stable, relatively independent for-
mal unit on par with Sätze 1 and 2. On the other hand, our reading, which is
figural, narrative, or allegorical in orientation and character, simply does not
engage the same frame of reference as the earlier analyses, which are both for-
mal (or to be more exact, form-functional or form-typological) in orientation.
In this sense, we have isolated only a part of the musical structure at hand, a
play within a play, nested within a larger formal dynamic, to which it can only
contribute but which it cannot itself determine. Nevertheless, we might draw
from our reading a particular methodological conclusion about formal analy-
sis, namely, that certain passages or elements may be fruitfully understood in

the first instance as (very loosely speaking) hermeneutically important rather

than purely along formal lines. In this regard, analyses of form in the tradi-
tional sense, regardless of the way form is understood, are disadvantaged by
their quite reasonable concern to define the nature of convention, so that they
become less open to idiosyncratic or personal elements. This is not to say that
the conventional-personal dynamic need be absent entirely from traditional
formal analysis. Caplin’s analysis of the music at hand, to give one example,
forms a component in his study of changes in Beethoven’s symphonic shapes.29
In this case, the evaluation of the personal, however, takes as its object lifelong
development rather than the individual work.
Notes
1. Donald Francis Tovey, Essays in Musical Analysis: Symphonies and Other Orchestral
Works (New York: Oxford University Press, 1989), 38.
2. Hector Berlioz, A Critical Study of Beethoven’s Nine Symphonies, trans. Edwin Evans
(Urbana and Chicago: University of Illinois Press, 2000), 29.
3. Ibid., 29–30.
4. Hermann Kretschmar, Führer durch den Konzertsaal, 5th ed. (Leipzig: Breitkopf und
Härtel, 1919), 1:192 (my translation).
5. Carl Schachter, “Beethoven’s First and Mozart’s Last: Echoes of K. 551 in the
First Movement of Opus 21,” in Mozart Studies, ed. Cliff Eisen (New York: Oxford
University Press, 1991), 227–51; Elaine Sisman, “‘The Spirit of Mozart from Haydn’s
Hands’: Beethoven’s Musical Inheritance,” in The Cambridge Companion to Beethoven,
ed. Glenn Stanley (Cambridge: Cambridge University Press, 2005), 45–63; and
Daniel Heartz, Mozart, Haydn and Early Beethoven: 1781–1802 (New York: Norton,
2009), 786.
6. James Hepokoski and Warren Darcy, Elements of Sonata Theory: Norms, Types, and
2006), 125–29.
7. Erwin Ratz, Einführung in die musikalische Formenlehre: Über Formprinzipien in den
Inventionen und Fugen J. S. Bachs und ihre Bedeutung für die Kompositionstechnik
Beethovens (Vienna: Österreichischer Bundesverlag für Unterricht, Wissenschaft
und Kunst, 1951), 23.
8. Ibid., 22–24.
9. William E. Caplin, “Structural Expansion in Beethoven’s Symphonic Forms,” in
Beethoven’s Compositional Process, ed. William Kinderman (Lincoln and London:
University of Nebraska Press, 1991), 27. Later in the essay, Caplin amplifies some-
what on the source of the additive technique, characterizing it as “a compositional
procedure employed by Haydn and even more frequently by Mozart” (ibid., 32).
10. Hepokoski and Darcy, Elements of Sonata Theory, 125.
11. William E. Caplin, “What Are Formal Functions?” in Caplin, James Hepokoski, and
James Webster, Musical Form, Forms & Formenlehre: Three Methodological Reflections, ed.
Pieter Bergé (Leuven: Leuven University Press, 2009), 31.
12. The performance, part of Mengelberg’s last prewar cycle of Beethoven sympho-
nies, is available on a number of labels: Philips LP 6767 003, Philips LP 416 200-2,

Andromeda CD 5040, Opus KURA 2015, Grammafono 2000 78 032/36, and Pearl
0074.
13. Richard Taruskin, Text and Act: Essays on Music and Performance (New York: Oxford
University Press, 1995), 220–21.
14. It is worth pointing out that Mengelberg is not the only conductor to treat the
passage along these lines. Furtwängler also inserts a ritardando at this point in his
recorded performances of the First Symphony, and Pfitzner, in a recording released
in 1928, maintains the tempo into the cadence at measure 77, but begins Satz 4 with
a much slower pace and then, like Mengelberg, accelerates into the cadence of Satz
4 in measure 88. Pfitzner’s recording (with the Berlin Philharmonic) is available on
Naxos 8.110927; Furtwängler’s recording is also published by Naxos (9.80014).
15. Taruskin speculates that Mengelberg’s tempo choices here are motivated by “‘the
equation of crescendo with acceleration and diminuendo with ritard,’” which “is
often taken as paradigmatic for ‘Romantic’ (or in Virgil Thomson’s well-known
opinion, for ‘European’) interpretation generally.” Taruskin, Text and Act, 221.
Taruskin is citing Will Crutchfield, “Brahms by Those Who Knew Him,” Opus 2, no.
5 (1986): 12–21.
16. In this way, Mengelberg’s performance helps us to appreciate Caplin’s characteriza-
tion of the passage as “medial,” “so the theme seems to start, in some sense, already
in its middle” (“What Are Formal Functions?” 31); although perhaps not in the
purely functional sense that Caplin has in mind.
17. Kretschmar, Führer, 192.
18. Sisman, “The Spirit of Music from Haydn’s Hands,” 55.
19. Berlioz, A Critical Study, 30.
20. Hugo Riemann, Catechism of Orchestration: Introduction to Instrumentation, trans.
anon. (London: Augener, 1910), 33, 36; originally published as Katechismus der
Musikinstrumente (Instrumentationslehre) (Leipzig: Max Hesses Verlag, 1897).
21. Ibid., 9. See also Jeffrey De Thorne, “Klangvertretung: On Riemann’s Aesthetic
Theory of Orchestration” (paper delivered at the Annual Meeting of the Society
for Music Theory, Nashville, TN, November 7, 2008). Subjectivation is a central
notion in Riemann’s aesthetics. In Catechism of Musical Aesthetics he writes:
When Fr. von Hausegger says that music is first and above all expres-
sion, and when Arthur Seidl lays stress on “entering, with one’s feel-
ing, into the forms of movement of music,” that, certainly, is something
akin to the Subjectivation of music on which I insist. Still I believe
that by distinguishing the elementary, which arises simply from the
impulse to impart oneself (Hausegger’s “Music as Expression”), from
the formal, which belongs to the impulse to play (which according
to Hanslick is everything), and from the characteristic, which in itself
is alien to music and is drawn into its sphere only by the impulse
to imitate (music not as the expression of the subject but as expres-
sion of an imagined object)—I have set up something that is worth
developing. For the listener there result principally two altogether
different ways of perceiving music: in one, music is felt as the mani-
festation of one’s own will (complete subjectivation); while in the
other it is, partly at least, objectivated by the imagination. But the
more, in absolute music, the formal preponderates over the elemen-
tary, that is to say, the less music is felt, and the more it is made—the
more imperfectly shall we subjectivate it, the more it will remain out-

side of us; on the other hand, imitative music, in spite of scene and
program,—if only it does not proceed too restlessly, but rather leaves
time for the expression of sentiment of the represented beings to
develop (that is to assume form)—can affect us so sympathetically
that, for moments at least, we can completely subjectivate it and
identify ourselves with the object represented.
Catechism of Musical Aesthetics, trans. H. Bewerunge, 2nd ed. (London: Augener,
1895), iv; originally published as Die Elemente der musikalischen Aesthetik (Berlin: W.
Spemann, 1900).
22. Many other commentators point out that the motive that opens Satz 1 recalls the
opening of the first theme itself, the element that binds the slow introduction to
the Allegro, which is true but not very interesting.
23. In A Classic Turn of Phrase: Music and the Psychology of Convention (Philadelphia:
University of Pennsylvania Press, 1988), Robert Gjerdingen tracks the appearance
of instances of the changing-note schema from the mid-eighteenth to the late nine-
teenth century and finds among other things that its frequency peaks in the 1770s.
Its appearance here, then, suggests that the second theme is a nostalgic element in
the symphony. That would be true even if the nostalgia were in service of an hom-
age to the School of Haydn.
24. Caplin, “Structural Expansion,” 35.
25. The cognitive assertion is open to demonstration. Find an audience of at least one
person. Think of key and play a subdominant triad and dominant triad (in any
order): the audience will be able to sing the relevant tonic. The exercise may be
repeated successfully in any key regardless of the relation of one key to the next.
On the other hand, triads alleging to serve as I and V in one key may as easily serve
as IV and I in another key. So that if one plays, say, a C triad followed by an F triad,
an audience may develop the impression of I–IV as easily as V–I.
26. We can understand the transformation of the motive from a descent through a
fourth rather than a fifth both as an omission (of the motive’s final pitch, B♭ in
this case) and as a reversal (of the first and last pitches of the C-form, the form that
extends from C to F).
27. Quoted in Carl Dahlhaus, Ludwig van Beethoven: Approaches to His Music, trans. Mary
Whittall (Oxford: Clarendon Press, 1991), 81. This is certainly true in the first
movement of Beethoven’s op. 2, no. 1 (where in the exposition, chords in the first
theme articulate every beat but the first of each measure, whereas in the recapitula-
tion they articulate the first beat of the measure very strongly) and the first move-
ment of his Fifth Symphony, where the presentation cycle of the second theme
articulates three statements of the basic idea in the exposition and four statements
in the recapitulation.
28. We might be able to insert more content into the drama we have described in the
second theme by investigating the associative dynamics of the instruments involved.
Riemann, in his Catechism of Orchestration, tells us that woodwinds are particularly
rich in such associations, pointing out that the flute is the shadowy and disembod-
ied (33). The oboe, he says, is gendered, and gendered in two ways: in its feminine
mode, it represents sexual naiveté; in its masculine mode, it relates to the topic of
the Pastoral, evoking the shepherd’s shawm.
29. See Caplin, “Structural Expansion,” 27.

Chapter Ten
Laborious Homecomings
The “Ongoing Reprise”
from Clementi to Brahms
Giorgio Sanguinetti
The idea of “homecoming” is a traditional metaphor for the recapitulation

section in sonata form, one that is deeply rooted in the society and culture
of late eighteenth- and early nineteenth- century Europe. Both Nicholas
Marston and Janet Schmalfeldt associate the idea with the political and
cultural climate of post-1815 Vienna, when the oppression of Metternich’s
restoration made many citizens think that “home” was the only place safe
from spies and police.1 Chadwick Jenkins goes back to Homer as a possible
source for the homecoming metaphor. Homer’s poetry was rediscovered in
European culture during the eighteenth century, and in particular influ-
enced the culture of Britain and Germany.2
While the title of the second of Homer’s epics, the Odyssey, has become
a synonym for a long and eventful journey, it is sometimes overlooked that
Ulysses’s return to Ithaca is anything but triumphal. He returns by night, while
still asleep, delivered by the Phaeacians to a hidden shore. He is disguised,
dressed as a beggar, and only his decrepit hound Argus is able to recognize
him without the help of a sign. His true identity is disclosed only gradually:
first to his son Telemachus, then to Eumaeus the swineherd, to Eurycleia the
nurse (who discovers his scar), and finally—but only after he has slaughtered
the Suitors—to his wife, Penelope. So, one might say that the exact moment
of Ulysses’s return is when he, asleep and unconscious, touches the shore of
Ithaca, but this is hardly a moment to be underscored with clashing cymbals.
His homecoming is an ongoing process that develops over time and that, as
Jenkins put it, is “necessarily anticlimactic.”3
There is another aspect that distinguishes Ulysses’s return from the reestab-
lishment of a pristine condition. Not only is home different after the journey

318 giorgio sanguinetti
(there were no Suitors before Ulysses’s departure), but also the hero himself
has changed. Dante powerfully captures these changes in his “sequel” to the
Odyssey in Canto 26 of the Inferno:4
Quando mi diparti’ da Circe, che sottrasse

me più d’un anno là presso a Gaeta,
prima che sì Enëa la nomasse,
né dolcezza di figlio, né la pieta

del vecchio padre, né ’l debito amore
lo qual dovea Penelopè far lieta,
vincer potero dentro a me l’ardore

ch’i’ ebbi a divenir del mondo esperto
e de li vizi umani e del valore;
ma misi me per l’alto mare aperto

sol con un legno e con quella compagna
picciola da la qual non fui diserto.
[When I sailed away from Circe, who’d beguiled me

to stay more than a year there, near Gaeta—
before Aeneas gave that place a name—
neither my fondness for my son nor pity

for my old father nor the love I owed
Penelope, which would have gladdened her,
was able to defeat in me the longing

I had to gain experience of the world
and of the vices and the worth of men.
Therefore, I set out on the open sea

with but one ship and that small company
of those who never had deserted me.]
According to Dante, on this second journey, made after his return from
Troy, Ulysses’s ship was wrecked and he drowned with all his companions.
The fact that he felt compelled to leave Ithaca once more shows how different
Ulysses is now from the man who once set out for Troy: his desire for knowl-
edge cannot be soothed, and he finds himself incapable of enjoying the quiet
life that he had so long dreamed of.
Homecoming also may have its harbingers. During the journey the arrival
is awaited, imagined, and anticipated. And the closer one comes to home,
the more frequent and the more recognizable the signals of approach are.

laborious homecomings 319
Gradual, ongoing changes in climate, vegetation, landscape, and architecture

are the heralds of arrival. Indeed, the approach can be so gradual that the
moment of arrival is overlooked, and the awareness that one is home may arise
only in retrospect. In his book La tregua, Primo Levi wonderfully describes the
slow epiphany of home toward the end of his laborious return from Auschwitz.5
❧ ❧ ❧
In this chapter I try to identify a musical counterpart to Ulysses’s return to

Ithaca, one that I will call the “ongoing reprise.”6 By this term, I indicate the
onset of a recapitulation in a sonata-form movement that gives the impression
of a gradual, processual homecoming—of something that develops over time
but does not occur at any moment.7 In so doing, I aim to illustrate the idea of
the ongoing reprise and its concomitant phenomena—phenomena that are
similar to those I have just described in Ulysses’s own homecoming: namely,
the musical equivalents to the anticlimactic nature of the return, the impos-
sibility of a return to the pristine condition, and the “harbingers.”
In order to better define the ongoing reprise, we may want to consider its
opposite, the reprise with “double return”—that is, the simultaneous return
of the main theme and the home key.8 In traditional accounts of sonata
form, the double return (hereafter abbreviated DR) is the most remarkable
event in the recapitulation.9 It symbolizes the reestablishment of order, the
long-awaited moment when, after the destabilization in the second part of
the exposition and the meanderings of the development, the first theme tri-
umphantly returns in all its splendor. According to Peter Hoyt, the DR has
the same function as synecdoche in rhetoric: it stands for, and represents, the
totality of the recapitulation.10
If the DR is the norm, it is a norm that is often flouted. The first theme
sometimes reappears in a key other than the tonic, or it does come back in
the home key, but too early and in a context that obviously is still part of the
development. Occasionally it does not appear at all, and the home key returns
in conjunction with later material, such as the transition or the subordinate
theme. In this latter case, the first theme may never be heard again, or it may
return after the subordinate theme. In some cases, the return of the home key
occurs so late that it even bypasses the subordinate theme, and is secured only
by the authentic perfect cadence that brings the sonata proper to a conclu-
sion. Several definitions and labels have been coined to bring some order in
this chaotic universe: a tentative list of recapitulatory teratology might include
“false,” “premature,” “failed,” disjunct,” “mirror,” “reversed,” and “truncated”
recapitulations, to mention only the most common types. James Hepokoski and
Warren Darcy’s Elements of Sonata Theory has attempted to wipe out the “old”
categories of nonnormative recapitulations using the new category of “Type 2”
sonata.11 However, the ongoing reprise differs from a false recapitulation, or

from any other kind of non-DR recapitulation. Indeed, an ongoing recapitula-

tion may have a DR: but when it arrives, it passes unnoticed (as his arrival did
for the sleeping Ulysses upon his return to Ithaca).
Ongoing reprises also have some connection with retransitions, since both
signal the arrival “home.” In William E. Caplin’s terminology, retransitions are
“those passages that modulate back to the home key in preparation for the
return of some previous opening material. . . . Most typically, a retransition is a
complete phrase, or even a full theme-like unit, that follows the cadential artic-
ulation of a development key.”12 And, shortly after: “A retransition sometimes
starts with reference to the opening material from the main theme, usually in
the development key just confirmed by a prior half cadence. This effect can
be characterized as a false recapitulation. Eventually the music returns to the
home key for the true recapitulation.”13 If we compare the retransition with
the ongoing reprise, we might say that the latter resembles a retransition that
includes references to the main theme in an unstable tonal context and is not
followed by a true recapitulation.
Another common aspect of the ongoing reprise is that developmental
“rhetoric” persists after the onset of the main theme: the home key is then
stabilized only later, with the transition or (at the latest) with the arrival of
the subordinate theme. From a narrative point of view, this musical aspect
mirrors the hero’s inability to settle back in at the return from his journey, as
narrated by Dante.14
To summarize: (1) ongoing reprises cannot be said to occur “at” a given
moment: they occur progressively; (2) the exact moment cannot be deter-
mined, not even afterward (a posteriori); (3) “ongoing” recapitulations are
preceded by “harbingers” (or signals); and (4) they are often anticlimactic.
In this chapter, I will discuss at some length three examples, spanning almost
a century, of ongoing recapitulations from Muzio Clementi to Franz Schubert
and Johannes Brahms, and I will mention some other cases more fleetingly.
Although many of my examples are already well known for their unusual reca-
pitulation and have been discussed in the literature, I hope that my perspective
may shed new light on them.
Muzio Clementi: Piano Sonata in

G Major, op. 40, no. 1, mvt. 1
During his long career Clementi employed an impressive array of strategies for
his recapitulations, including the Type 2 sonata, which was a standard format
during his early years as a composer.15 A deviation from the Type 2 occurs in
the first movement of Clementi’s Piano Sonata in E-flat Major, op. 23, no. 3
(1790). At the beginning of the development, the main theme appears on a
diminished six-three harmony (vii°6/vi in C minor), but only the two-measure

basic idea is repeated; the third measure already brings in material from the
transition, while the harmony lands on a ten-measure standing on the domi-
nant in C minor (vi of the home key). After a fermata the main theme appears
again, this time on a stable harmony and in C major (VI♯ of the home key).
In this key, the main theme matches the exposition exactly for ten measures
(eight-measure theme and the beginning of the transition). The transition
is then expanded (twenty-four measures against fifteen in the exposition) to
effect the modulation back to E-flat major; at this point, the subordinate theme
appears in the home key, and the recapitulation is definitively “on track.”
So, even if we have accepted the appearance of the basic idea after the
repeat sign as beginning a developmental rotation, when the same theme
appears off tonic, verbatim, and in its entirety, and especially when it is fol-
lowed by the transition, we can no longer regard this movement as a Type 2
sonata: the second rotation is not continued so as to include the return of the
secondary theme but instead gives way to a third rotation (beginning with the
C-major statement).16 But this is not a standard, “textbook” sonata either (the
kind labeled by Hepokoski and Darcy “Type 3”), because there is no DR. What
I hear is a series of attempts at recapitulation. At first, the main theme is stated
off tonic, on an unstable harmony, and in a much-compressed version; then
the same theme is stated again, and in its entirety, over a stable harmony, but
again off tonic. Only when the subordinate theme appears in the tonic do we
have the feeling that “home” is attained.
Another intriguing case of premature, off-tonic recapitulation occurs in the
first movement of Clementi’s Piano Sonata in D Major, op. 16 (“La Chasse”;
1787). After the repeat sign, a motive loosely derived from the codetta (clos-
ing section) initiates a chain of dominant-seventh chords (V of VI♯–V of ii);
and there follows a passage ending on the dominant of ♭VII (C major) whose
rhythm and design derive from the subordinate theme. The main theme
returns in this most unusual key (the subtonic), and from that point on, all the
elements reappear in the same order as in the exposition. The main theme,
unchanged, remains stably in the subtonic, but the transition is expanded and
attains the home key’s dominant only after twenty-five measures (against four-
teen in the exposition); as before, the recapitulation is firmly on track only
with the beginning of the subordinate theme.17
In this case, as in the previous one, some “ongoing” aspects are present in
the recapitulation: the lack of DR, multiple returns of the main theme, and the
persistence of developmental activity after the last thematic return. However,
those cases lack the narrative and emotional implications of the “real” ongo-
ing reprises. As my first example of these, I will consider the first movement of
Clementi’s Piano Sonata in G Major, op. 40, no. 1 (1802).
The main theme (MT), shown in example 10.1, is a regular eight-measure
period, whose antecedent is composed of a strong, self-confident basic idea
and a gentle contrasting idea ending with an imperfect authentic cadence

Example 10.1. Clementi, Piano Sonata in G Major, op. 40, no. 1, mvt. 1, mm. 1–8
(IAC) on 3^. The G-major triad is stressed emphatically on the downbeat of the
first measure, but in the right hand only, without the bass—a circumstance that
will have interesting consequences in the recapitulation. After the opening
period, the movement rushes on into a vigorous transition ending on a half
cadence in the home key (not shown in ex. 10.1). But, instead of a subordinate
theme, the main theme returns, now in a hesitant manner: after its antecedent
phrase, the transition begins again, and this time the dominant of V is reached,
followed by an expanded standing on the dominant.18 Only at measure 37
does the subordinate theme finally enter.
The development, a moto perpetuo in triplets that is based on material from
the preceding closing section, appears to end in measure 110, on the domi-
nant of F-sharp minor. However, four measures before the fermata, another
dominant, V7 of G (minor) is reached by means of an ascending sequence;
more surprisingly still, this chord reveals itself as an augmented-sixth chord
moving to the dominant of F-sharp minor.
This “wrong turn” is a signal that the music has gotten lost (ex. 10.2 is a
voice-leading graph showing a summary of the development and the reprise).
In the key of F-sharp minor (♯vii of the home key), the main theme makes a
first, timid appearance (m. 111; the ongoing reprise is shown in ex. 10.3). The
theme lasts only four beats and is followed by a rest of the same length. The
second attempt (mm. 113–17) turns out to be more successful: now transposed
up a fourth, the theme holds out for the entire length of the antecedent (four

Example 10.2. Piano Sonata, op. 40, no. 1, mvt. 1: Voice-leading graph of the
development and “ongoing reprise”
measures, ending with a half cadence), but it soon swerves into a contempla-
tive mood and is significantly altered with respect to its original form. The third
attempt (mm. 118–25) begins in the key of B minor but is derailed midway and
lands instead on D major. As if realizing its mistake, the theme attempts to
recover the minor mode (fourth attempt, mm. 125–28), but finds itself again
in the wrong key, E minor: the fourth attempt thus results in a transposition
of the second attempt, but it is still far from a credible antecedent. The fifth
attempt starts in E minor, but in the second measure, the dominant of the
home key is precariously attained (m. 131). Over this dominant, the main
theme makes its hurried entrance, but in its haste, it mistakes the correct tim-
ing, so that the entrance is not as effective as it should have been. Entering as it
does, in the middle of the cadential process, its impact as a DR reprise is greatly
weakened, though not exactly anticlimactic. While in the exposition the right-
hand chords allowed us to clearly perceive a stable G major on the downbeat
of the first measure despite the absence of the left hand, here the bass retains
a D below the G-major chord, and as a result we hear a G-major 46 prolong-
ing the dominant chord that precedes and follows it. In this way, the effec-
tive tonal resolution is postponed until measure 135. This reprise, the fortissimo

Example 10.3. Piano Sonata, op. 40, no. 1, mvt. 1, mm. 110–35
notwithstanding (the sudden outburst sounds more like a mark of uncertainty

than a resolution), is unconvincing, and in fact, the developmental quality of
what has come before returns in the transition. As example 10.2 shows, the
recapitulatory process begins in the extremely remote key of F-sharp minor.
From this point on, successive attempts gradually approach the home-key
dominant through a descending-fifth progression (C-sharp–F-sharp–B–E) that
ends on ii of G major (from which the cadential process begins). Whatever the

composer’s intention may have been, this meandering, erratic return is a strik-
ing musical counterpart to Ulysses’s journey.
Schubert: Sonata in C Major, D. 840 (“Reliquie”)

In her book In the Process of Becoming, Schmalfeldt devotes a great deal of atten-
tion to recapitulations. Her reading of the first movement of Schubert’s Piano
Sonata in A Minor, op. 42 (D. 845) relates closely to my idea of the ongoing
reprise, though it differs in certain details. In particular, Schmalfeldt reads the
F-sharp-minor appearance of the main theme (m. 145) in the development as
the beginning of a “core” (the third of its kind in this sonata).19 The core is
the central event in many classical developments, and is also central in Caplin’s
description of the development.20
However, I do not really hear the rhetorical features of a core in this pas-
sage. As Caplin writes, “the core . . . projects an emotional quality of instability,
restlessness, and dramatic conflict. The dynamic level is usually forte, and the
general character is often one of Sturm und Drang.”21 In Schubert’s sonata the
dynamic level of the “core” is piano, and the passage conveys a feeling of uncer-
tain wandering rather than one of increased dramatic tension. Obviously, there
can always be nonnormative cores, but I propose another reading of the pas-
sage instead: namely, as the first harbinger of an ongoing reprise. Schmalfeldt
also reads the second occurrence of the main theme in A minor (m. 151) as
the beginning of a sequence, and as a “true” recapitulation that retrospectively
reveals itself as “false.”22 To my mind, this is the second harbinger; but even
accepting this as a “real” recapitulation (because it is a DR) would only make
the ongoing reprise that much more fascinating. In fact, since the main theme
is in the middle of a sequence, and in addition is incomplete, the point of
reprise is crossed unconsciously, exactly like Odysseus’s return to Ithaca. What
begins in measure 186 is not the recapitulation, but the moment we come to
realize that the threshold of the reprise has already been crossed; but again,
even in retrospect we cannot tell where (or when) that happened.
I now turn to another Schubert sonata movement, one very similar to opus
42 in its general idea, and no less extreme in its recapitulatory strategy: the first
movement of the unfinished Sonata in C Major, D. 840 (“Reliquie”). Nothing
in this movement seems to aim in the right direction; everything appears to
lose its way in a blind, erratic meandering—as in a pessimistic rendition of the
Odyssey. The main theme itself (mm. 1–27, shown in ex. 10. 4) is a complex one:
a compound period, with a hybrid 3 (compound basic idea + continuation,
mm. 1–8) as its large antecedent, and an enormously expanded consequent
(mm. 9–27).23 The lowering of 6 ^
to ♭6^ in measure 13 diverts the consequent’s
continuation to ♭VI (or more precisely, the dominant of the Neapolitan),
where it stays with growing anxiety for nine measures, until a last-moment

Example 10.4. Schubert, Sonata D. 840 (“Reliquie”), mvt. 1: main theme
enharmonic transformation into an augmented-sixth chord brings the con-

sequent to its proper cadential 46 for an emphatic conclusion by means of an
elided perfect authentic cadence (PAC) in the tonic.24
The transition begins with a forte restatement of the main theme in the
home key (recalling a symphonic gesture typical of Haydn) and spends a
great amount of energy going nowhere. In fact, the transition comes back to
♭VI (this time heard as a tonic), which is expanded again for nine measures

through an octave ascent by minor thirds (A♭–B♮–D–F–A♭, mm. 35–39).

But this enormous effort has no consequence: ♭VI plunges dramatically into
the dominant of the home key once again, and then an emphatic caesura-fill
following the I:HC medial caesura increases the tension of this dominant by
bringing in the seventh in measure 46 and the minor ninth a measure later.
Then, in measure 51, something stunning happens: the dominant ninth trans-
forms itself into the cadential 46 of B minor (vii♯5), the key in which the subor-
dinate theme appears.25
The new theme seems less than completely comfortable in B minor. The
theme, in which the upper voice circles around F♯, consists in an (apparently)
endless succession of antecedents unable to find a consequent. The first ante-
cedent clumsily modulates to the relative major (the dominant of III is miss-
ing, and replaced by an unlikely III36). The next attempt to find a consequent
ends with an HC, as does the third; the fourth attempt eventually concludes
with a PAC in measure 71, but on a different tonic: G major, the “proper” tonic
of a subordinate theme, but not the one in which the theme had begun.
Although the top voice now clings to G, it seems unable to accept that
note as the tonic of the subordinate key, and tries to find a way to move down.
Apparently unable to set itself free from F♯, however, it frantically repeats the
G–F♯ step (almost a motive by now) in connection to G–A and G–A♭ (the two
neighboring motions in the main theme) as an inner voice (mm. 75–80).26
Finally, F♮ having been found in measure 80, a new descent can take place:
and, in fact, a cadential progression begins from 3^ in the bass—but again in
the wrong key!27 A corrective is introduced in the form of a deceptive cadence
(mm. 83–84): the correct key is now established, and this amazing fifth pro-
gression—in which 5^–4^–3^ are in F major, and 3^–2^–1^ in G major—finally lands in
the proper key of G major with a second PAC.
If we would accept this as the true conclusion of the exposition, what follows
should be a closing section: but now the subordinate theme appears again in
its proper key and form—a true period, with antecedent and consequent, and
performing a perfectly recognizable fifth descent (5 ^ ^
in m. 88, 4 in m. 93, ♮3
^
,
^ ^
2 in m. 96, and 1 in m. 97). Feeling at home, the theme dances into a frenzy
of excitement, as in measures 89–90 and 96–101. Only an unexpected ♭3^ in
the consequent (mm. 94–95) casts a shadow on this joy, to be dissipated by
the diatonic correction (accompanied by another surge of dancing triplets)
two measures later. The conclusion of this late appearance of the subordinate
theme leads to a third, and this time completely satisfying, PAC that is followed
by a short codetta.
I have described the exposition of the “Reliquie” sonata in such detail not
only because it is an extreme case of formal and tonal wandering but also
because it creates the expectation of a comparably extreme recapitulation.
Shortly before the repeat sign, the bass alone, having reached G after the last
of several reiterated PACs, moves down a third to E. This pitch connects back

to the repetition of the main theme (first-time repeat) and forward as V of the
statement of the main theme in the key of VI♯ (A major).
The main theme obsessively permeates the entire development section,
which opens in A major with that theme sounded above a pulsing pedal point
of triplets (an extension into the development of the “dancing” triplet rhythm
that ends the exposition). The character of the main theme is now “more
relaxed yet also somewhat hesitant and anticipatory.”28 These are the words
Caplin uses to characterize the “precore” (or the introduction to the core),
and it is a description that is entirely appropriate for this transformation of the
main theme. Part of the new character is no doubt due to the unexpected shift
to the third-related key of A major, but equally significant is the G♮ in measure
116, which produces a minor seventh above the tonic triad. Laura Krämer has
referred to this as a “farewell seventh” (Abschiedsseptime), a figure associated
with the plagal cadence and with the galant schema that Robert Gjerdingen
has called the “Quiescenza”: both figures are subdominant-oriented and are
typical features of what Caplin calls “postcadential function.”29
The core is based on the progressive liquidation and reduction of an initial
six-measure model (mm. 115–20), which is restated once, then reduced to four
measures (mm. 127–30), restated again, and then further reduced to one mea-
sure in one of the most impressive dissonant sequences in classical music: five
dominant-seventh chords a minor third apart are lined up in ascending direc-
tion, producing an astonishing prolongation of the dominant of B minor, with
an octatonic upper line as a by-product.
The sequence leads to a standing on the dominant of B minor (vii♯5) con-
sisting of a liquidation of thematic material and a countergeneric recessive
dynamic, from fortissimo (m. 138) to pianissimo (mm. 146ff.; see ex. 10.5).30
After a cadence in B major, the main theme returns in this key, in the “fare-
well” version that opened the development. At first, its reappearance could
be taken for an extreme case of off-tonic recapitulation (on VII♯♯35), but the
theme dissolves over a chromatic ascending bass that leads a third upward, to
D major; and the perception of a reprise vanishes as well. Given its position at
the (presumed) end of the development, the theme acquires the meaning of
a foreshadowing, a harbinger of the expected recapitulation; but at the same
time, it is also a recollection, pacified and almost motionless, of the extreme
tumult of the main-theme-based development.31
In measure 162, in a precariously attained D major (there is no cadence to
secure it), a second element of the main theme—the continuation—appears
in an abridged version (corresponding to a cut connecting mm. 7 and 20).
The addition of a minor seventh in measure 166 (already hinted at in m. 160)
seems to aim for a cadence to G, but the ensuing cadence veers instead to a
startling F major (IV), where the continuation vanishes and in so doing reveals
its true nature as a harbinger, the second so far. The main theme starts over
again in F, this time as a six-measure antecedent followed by a compressed

Example 10.5. Sonata D. 840 (“Reliquie”), mvt. 1: ongoing reprise, mm. 146–83
(continued)

version of the consequent’s continuation (compare mm. 174–82 with mm.

14–27). This is the longest restatement of the main theme so far, and the key in
which it appears (IV of the home key) could well lead one to take this passage
as the onset of a subdominant recapitulation—not an uncommon procedure
for Schubert. I do not rule out this possibility; in fact, it would sit well with
the interpretation of the section as an “ongoing reprise” that I am proposing.
Coming as it does after two harbingers, and with the “main theme” thus dis-
torted, the (supposed) reprise is almost beyond recognition: the threshold of
the recapitulation has been crossed without our realizing it. Still, my preferred
interpretation of this thematic statement is as third harbinger.

As in the exposition, the consequent leads to an emphatic dominant sev-

enth, transforms it into an augmented-sixth chord and resolves it eventually
to a cadential 46 (in m. 179). But this time the 46 is on A minor, and the fol-
lowing dominant seventh is that of C major: the transition follows, unchanged
for twelve measures. The cadence’s swerve from A minor to C major is almost
unbearably crude: it is as though a panicked Schubert, having lost any hope
of finding “home,” forced the tonal path to recover, desperately, at the last
moment, the DR.32 Example 10.6 provides a synopsis of the bass line of the
development and the beginning of the recapitulation, including the thematic
entries. Is there any single measure that one might label as the onset of the
recapitulation? In my opinion, the answer is “no.” As Richard Kramer wrote, “if
one should quibble—as many do—about the precise moment at which ‘reca-
pitulation’ sets in, that would be to quibble about the wrong issue.”33
Obviously, there is no DR in this sonata. But even if we separate out its two
components—thematic and tonal return—their independent identification is
no less problematic. There is no thematic return, because the theme has been
Example 10.6. Sonata D. 840 (“Reliquie”): bass-line synthesis of the development

and beginning of the recapitulation, with thematic entries

present in the development all along, often in a tense, dramatic, and even tor-
tured way (as in mm. 127–38). The dreamlike appearance of the Quiescenza
version of the theme in B major hardly qualifies as the “real” recapitulation
for at least three reasons: (1) it is in the key of the leading tone; (2) only the
first measure is similar to the real theme; and (3) what follows makes it clear
that we are still in the development.34 Similar arguments may be invoked
against the other “forebodings” that begin in measures 162, 169, and 172.
In measure 182, by contrast, the main theme does in fact return; but this
passage is a verbatim replica of the “Haydnesque” transition, identical to the
one in the exposition for no less than twelve measures. Since the onset of
the transition coincides with the tonal return, we might decide that the reca-
pitulation begins here: but then, we would ignore the fact that we have been
hearing rhetorically plausible appearances of the main theme for thirty-two
measures already. Alternatively, we might decide that the “on-track” transi-
tion makes us retrospectively hear one of the several statements of the main
theme as the real reprise: but which one? Thus, there is no way we can say:
“this is where the recapitulation begins.”
Brahms, Sonata for Piano and

Violin in G Major, op. 78, mvt. 1
Peter H. Smith epitomizes one of Brahms’s most important recapitulatory strat-
egies as the tendency of the development to overlap with the beginning of the
recapitulation, in such a way that the distinction between them is blurred.35
Smith classifies Brahms’s recapitulations according to the extent and nature
of these overlaps. He places them along a continuum from the more conven-
tional, DR recapitulations (on the left side) to extreme cases of fusion with the
development (on the right side).36 The first movement of the Violin Sonata in
G Major, op. 78 (completed in June 1879), which would seem to have a reason-
ably clear reprise, is not very far from the left side of Smith’s continuum; how-
ever, I will argue that its reprise is more apparent than real.37
The essential prerequisite for any attempt to identify a recapitulation is to
find where the main theme begins, and this is the first difficulty with this sonata.
In fact, the theme has three possible beginnings. The first is the downbeat of
measure 1, with the first G-major chord of the piano (marked “x” in ex. 10.7).
The second is the fifth quarter note of the same measure, with the violin’s two
upbeats (marked “y”); the third is the arpeggiated figure on the downbeat of
measure 2, which sets the theme fully in motion (“z”).38 Each of the three pos-
sibilities has good points in its favor. The piano chord is clearly introductory,
but in the course of the sonata it takes on a significant thematic function. The
violin’s upbeats not only create an obvious connection with the sonata’s finale
but also are programmatically linked with two of Brahms’s songs, “Regenlied”

Example 10.7. Brahms, Sonata for Piano and Violin in G Major, op. 78, mvt. 1,
mm. 1–9
and “Nachklang” (op. 59, nos. 3 and 4).39 But the downbeat of measure 2 is
the moment when, as we shall see, the recapitulation comes to be “on track.”
The main theme proper is a ternary form—A (mm. 1–10), B (mm. 11–20),
and A‫( ׳‬from m. 20)—in which A‫ ׳‬merges with the transition (A‫–׳‬transi-
tion: mm. 20–29). A is a Brahmsian version of the classical hybrid 2 (ante-
cedent plus cadential) in which the antecedent ends with a plagal cadence.
It is linked to B by a two-measure extension that introduces a descending
scalar motive and that will have some interesting consequences.40 The con-
trasting middle section avoids the resolution of the dominant-seventh har-
mony reached at the end of A, and instead prolongs that harmony, the bass
moving chromatically to the (tonicized) third, then to the fifth of D major.
This bass motion is associated with a multilevel metrical conflict originating
from a clear caesura in both instruments in the middle of measure 12, which
destabilizes the meter; the consequences are only resolved completely at the
beginning of A‫׳‬.

There is, however, a feature of the main theme that, although not as imme-
diately evident as the motives, has a deep influence on the way the recapitu-
lation will be attained. The first four measures are based on what we might
call a “plagal” cadenza doppia: that is, a cadenza doppia where the structural bass
is a tonic rather than a dominant.41 The model is shown in example 10.8a,
whereas example 10.8b shows the same cadence expanded through a “cast-
ing out” of the root of the 46 chord and the addition of a 9–8 suspension in the
upper voice.42
The main stations of the sonata’s route are signaled by returns of the main
theme, which appears also in framing functions. The clearest tonally stable
return of the main theme—with the two instruments exchanging their roles—
occurs at the beginning of the development, where it creates an obvious
instance of developmental rotation. As always with ongoing reprises, the devel-
opment is inextricably merged with the recapitulation.
The ongoing reprise takes up twenty-three measures, from measure 134 to
measure 156 (see ex. 10.9). The standing on the dominant ends with a dra-
matic PAC on the minor tonic in measure 134, which prevents the return of
the major-mode main theme; instead, the piano’s left hand and the violin
engage in a canon on a lower neighboring-note motive. From the beginning of
the standing on the dominant in measure 127 to the Tempo I in measure 156,
the harmony dwells on the minor tonic, G minor; in fact, the tonal structure
of this passage is that of an enormously expanded cadenza doppia in G minor.
(The basic scheme is shown in ex. 10.10: on the second beat, the bass of V
is embellished by a neighboring note.)43 Between the cadenza doppia proper
and its final resolution in measure 156, the cadence is repeated twice (in mm.
134–41 and 142–55), both times omitting the initial V35 harmony (the repeti-
tions are shown in parentheses in the example). In both incomplete cadenze
doppie the two instruments engage in a canon on a lower neighboring-note
motive that derives from the conclusion of a chorale-like postcadential theme
(second half of m. 62), but which ultimately originates in the rhythmic impulse
of the subordinate theme’s beginning (“con anima”: m. 36, first half). Both
canons end with a clear allusion to the main theme, and in that sense function
as harbingers.
When the first canon breaks after four measures, the rhythm slows down,
leading to a musical void: a minor-mode harbinger of the main theme’s basic
idea materializes over a prolongation of the dominant harmony. The canon
then returns, with the parts exchanged, and so does the harbinger, for the sec-
ond time, after six measures. But when the basic idea is concluded, the har-
mony is no longer the dominant: a mysterious chord progression leads to a
profound anticlimax (m. 150), where the harbinger comes for the third time,
dolce, leading to a sequence above a descending circle of fifths in the bass.
Measure 150 is a moment of absolute stillness and suspension, almost evocative
of Ulysses’s slumber during his own homecoming. From this point through to

Example 10.8. “Plagal” cadenza doppia and its expansion
Example 10.9. Sonata for Piano and Violin, op. 78, mvt. 1: ongoing reprise,
mm. 133–58
(continued)

Example 10.10. Sonata for Piano and Violin, op. 78, mvt. 1: basic scheme of the
cadenza doppia, mm. 128–56

the Tempo I (m. 156), the music slowly feels the attraction of the G major’s
tonal gravity through the circle of fifths (therefore the cryptic indication poco a
poco means, in my opinion, an accelerando, not a ritardando!).
Example 10.11 shows my interpretation of measures 128–56 as an expanded
cadenza doppia (compare ex. 10.10). The presence of a modified cadenza doppia
(with 36 on B♭) at the end of the development (m. 133) is perhaps more sig-
nificant than the motivic or thematic signals scattered through the no-man’s-
land between the end of the development and the moment when we feel that
the recapitulation is safely on the track. The “plagal” cadenza doppia provides
the chordal frame for the main theme; by basing the motivic signals of the
main theme’s reprise on it, Brahms creates a deeper kind of recapitulatory har-
binger. That Brahms’s strategy was deliberate is made evident by a small but
significant detail: part “e” of the development ends in measures 125–26 with
an explicit cadenza doppia, thus suggesting, through a sophisticated usage of
linkage technique, the key to understanding one of the most striking aspects
of this reprise.44
If one were to single out an “official” moment for the recapitulation, the
obvious choice would be the Tempo I at measure 156.45 Apart from the desta-
bilizing seventh added to the tonic harmony in that place, this is quite patently
the tonal return: but what about a return of the theme? As I said earlier, there
are three “beginnings” of the main theme: the solo piano chords (“x”), the
Example 10.11. Sonata for Piano and Violin, op. 78, mvt. 1: voice-leading graph of
mm. 128–56

violin’s dotted upbeats (“y”), and the arpeggiato figure (“z”). The Tempo
I recapitulation begins with the arpeggiated “z” motive, ruling out entirely
the other two motives. But these motives do appear earlier. The solo piano
chords set the scene for the first two harbingers (mm. 140 and 148); and the
dotted upbeats, besides opening the violin’s three harbingers, palpitate won-
derfully in the piano’s left hand during the descending-fifth sequence (mm.
151–52). When we arrive at the Tempo I, we do not mark the beginning of
the recapitulation there; rather, we feel the awareness of already being in the
midst of an extended, ongoing recapitulatory process. Quite paradoxically,
in this sonata it might be possible to single out the moment of the reprise. It
is the least likely candidate: the moment of absolute stillness at measure 150.
But the identification is possible only if we recognize both its anticlimactic
and utterly countergeneric character and its connection to the literary model
of the Homeric poem.
❧ ❧ ❧
Perhaps the few examples I have discussed in this chapter do not represent a
large category of recapitulations: a piece of music such as the “Reliquie” sonata
is so extreme that one can hardly find any counterpart to it. However, I think
that the idea of ongoing reprises may help us to better understand the com-
plexity of the recapitulation process. At least, it can relieve us from the moral
imperative of finding at all costs a specific moment for raising the “reprise”
flag: something that does not occur in the literary examples of journeys and in
their precursor, the Odyssey.
Notes
1. Nicholas Marston, “Schubert’s Homecomings,” Journal of the Royal Musical
Association 125 (2000): 248–70; Janet Schmalfeldt, In the Process of Becoming: Analytic
and Philosophical Perspectives on Form in Early Nineteenth-Century Music (New York:
Oxford University Press, 2011), 227–57.
2. Pim den Boer, “Homer in Modern Europe,” European Review 15 (2007): 171–85.
3. Chadwick Jenkins, “Recapitulation as Process: The Augmented-Second Tetrachord
in the First Movement of Haydn’s Op. 33, no. 5,” Studia Musicologica 51 (2010):
347–67. According to Jenkins, Ulysses’s “true” homecoming coincides with “his res-
toration to his proper position as the recognized patriarch of Ithaca” (349), and
Homer devotes to it only half of book 23; therefore, it is “necessarily anticlimactic
and yet . . . no less necessary” (ibid.). I agree with Jenkins that if one had to single
out a moment of “true” homecoming, this cannot be Ulysses’s delivery, asleep, on
the shores of Ithaca. But it is also a fact that if we consider coming home as a pro-
cess, it takes much more time to develop; precisely, from book 13 to 23, more than
half of the entire poem.

4. “Inferno 26: The Quest” (trans. Allen Mandelbaum), in Digital Dante, ed.
Teodolinda Barolini, http://digitaldante.columbia.edu/dante/divine-comedy/
inferno/inferno-26/ (accessed April 23, 2015).
5. Primo Levi, La tregua (Turin: Einaudi 1963). Translated into English by Stuart
Woolf as The Reawakening (La tregua): A Liberated Prisoner’s Long March Home through
East Europe (Boston: Little, Brown, 1965).
6. I distinguish between “recapitulation” (as a formal unit) and “reprise” (the moment
when the recapitulation begins).
7. The processual aspects of recapitulation have been explored by Schmalfeldt. In particu-
lar, her discussions of Clementi’s F-minor Sonata and the first movements of Schubert’s
A-minor Piano Sonata, op. 42, come to conclusions that are comparable to my idea of
ongoing reprise. See Schmalfeldt, In the Process of Becoming, 73–80, 126–29.
8. As far as I am aware, this term was first introduced by James Webster, who defined
it as “the simultaneous return to the opening theme and to the tonic which con-
stitutes and defines the beginning of the recapitulation in sonata form.” See James
Webster, “Binary Variants of Sonata Forms in Early Haydn Instrumental Music,” in
Joseph Haydn: Bericht über den internationalen Joseph Haydn Kongress (Wien, Hofburg,
5–12 September 1982), ed. Eva Badura-Skoda (Munich: Henle, 1986), 127.
9. See James Webster, “Sonata Form,” in The New Grove Dictionary of Music and
Musicians, 2nd ed., ed. Stanley Sadie (London: Macmillan, 2001), 23:688.
10. Peter A. Hoyt, “The ‘False Recapitulation’ and the Conventions of Sonata Form”
(PhD diss., University of Pennsylvania,1999), 16.
11. “Type 2” is one of Hepokoski and Darcy’s five sonata types; the standard, DR sonata
is Type 3. See James Hepokoski and Warren Darcy, Elements of Sonata Theory: Norms,
Types, and Deformations in the Late-Eighteenth-Century Sonata (New York: Oxford
12. William E. Caplin, Classical Form: A Theory of Formal Functions for the Instrumental Music
of Haydn, Mozart, and Beethoven (New York: Oxford University Press, 1998), 157.
13. Ibid., 159.
14. One important caveat is to avoid confusion between ongoing reprises and Type
2 sonatas in which the second “rotation” encompasses the development and the
recapitulation. In Type 2 sonatas, the tonal return arrives, as a rule, together with
the subordinate theme, whereas in the ongoing reprise it tends to come much
earlier—within the transition, or even during the main theme itself. In addition,
during an ongoing reprise the flow of the development rotation is interrupted by
signals (“harbingers”) that introduce elements of the main theme well after their
appearance at the beginning of the development.
15. Clementi’s first piano sonata, WO 14 in G Major (written in 1768 at age sixteen,
when still in Rome) is a textbook Type 2 sonata.
16. A developmental rotation is a cyclic repetition of the same material already pre-
sented in the exposition, and in the same order, encompassing the development
and the recapitulation. It is an essential feature of the Type 2 sonata. See Hepokoski
17. “Premature” recapitulations of this kind are discussed in L. Poundie Burstein,
“True of False? Re-Assessing the Voice-Leading Role of Haydn’s So-Called ‘False
Recapitulations,’” Journal of Schenkerian Studies 5, no. 1 (2011): 1–36. However,
Burstein’s essay deals with the strongest kind of “premature” recapitulations: those
with a DR.

18. In Hepokoski and Darcy’s terms, this would be called a “trimodular block”
(TMB)—that is, a formal type resulting from a double medial caesura occurring
before the essential expositional closure (EEC). A TMB consists of three elements:
the theme after the first caesura (TM1), the passage leading to the second caesura
and that caesura itself (TM2), and the following new theme (TM3). See Hepokoski
19. Schmalfeldt, In the Process of Becoming, 127.
20. Caplin draws the idea of “core” from Erwin Ratz’s concept of Kern der Durchführung
as discussed in the latter’s Einführung in die musikalische Formenlehre, 3rd rev. ed.
(Vienna: Universal, 1973).
22. Schmalfeldt, In the Process of Becoming, 127.
23. On compound periods, see Caplin, Classical Form, 65–69.
24. The enharmonic transformation is never made explicit by Schubert, but it is retro-
spectively inferred by the listener (or performer) on the arrival on V46. However, the
dominant seventh of D-flat is so obsessively repeated that an experienced listener
might suspect it as early as in measure 22, and perhaps even earlier.
25. This is an instance of modulating caesura-fill. In their book Hepokoski and
Darcy mention two instances of this rare kind, both by Schubert: the first move-
ments of the Symphony no. 8 in B Minor (“Unfinished”), D. 759, and of the
Piano Sonata in C Major, D. 279. Hepokoski and Darcy, Elements of Sonata Theory,
29 and 29n8.
26. This cadence, although formally a PAC, is not satisfying because it lacks a complete
^
descending fifth progression in the theme (3 is missing between m. 69 and m. 70).
27. A cadential progression is “a progression that confirms a tonality by bringing its
fundamental harmonic functions” (Caplin, Classical Form, 253). A strong cadential
progression is that beginning with the tonic in first inversion and ascending step-
^ ^ ^ ^
wise to the dominant before reaching the tonic (3–4–5–1): “combined with a pre-
dominant built over the fourth scale degree in the bass, the cadential I6 initiates a
powerful ascending melodic motion toward the fifth scale degree, which supports
the root-position dominant, the linchpin of the cadential progression.” William E.
Caplin, “The Classical Cadence: Conception and Misconceptions,” Journal of the
American Musicological Society 57 (2004): 71.
29. Laura Krämer, “Die ‘Abschiedsseptime’ und ihre Transformation bei Schubert und
Brahms,” Musik & Ästhetik 56 (2010): 60–71. Robert Gjerdingen, Music in the Galant
Style (New York: Oxford University Press), 2007, 181–95. Caplin, Classical Form, 16.
30. The “recessive” standing on the dominant is described by Caplin in Classical Form,
261n22. The concept of “recessive dynamic” is drawn from Wallace Berry, Structural
Functions in Music (New York: Dover, 1987), 7.
31. “Rather, one wants to admire the apparent contradiction of a return to C major
approached through the back door and the affirmation of B major, with its lumi-
nous D♯ sounding the Romantic sixth in ecstatic response to twenty-four big
measures on a dominant F♯. The grand gestures of resolution, removed to this
exotic key, are tinged in the sublime.” Richard Kramer, Unfinished Music (New York:
Oxford University Press, 2009), 353.
32. Only in the coda will Schubert be able to retrieve the regular harmonic progression
ii–V7–I for this passage.
33. Kramer, Unfinished Music, 353.

34. “The opening theme coalesces into something visionary, a coherent phrase (fusing
what, at the opening of the movement, seem four discrete pieces of a puzzle not yet
in composite, four phrases in search of a source) now enhanced with a new inner
voice that makes audible the silent ‘innere Stimme’ that Schubert’s poets are for-
ever invoking.” Kramer, Unfinished Music, 353.
35. “One of the chief means by which Brahms sustained the esthetic legitimacy of reca-
pitulatory conventions was through a consistent blurring of the articulation at the
beginning of the reprise. In place of the instantaneous crossover from the develop-
ment to the recapitulation characteristic of late eighteenth-century sonata forms,
Brahms favors an extended and formally ambiguous overlap that continues the
motivic and harmonic-contrapuntal process of the retransition, while simultane-
ously introducing elements of a gradually emerging, large-scale restatement.” Peter
H. Smith, “Formal Ambiguity and Large-Scale Tonal Structure in Brahms’s Sonata-
Form Recapitulations” (PhD diss., Yale University, 1992), 1–2.
36. Brahms developed his recapitulatory techniques over many years. Smith describes
this process in the two versions of the Piano Trio in B Major, op. 8, in “Formal
Ambiguity,” 109–20.
37. “Though there is no real formal ambiguity regarding the actual point of recapitu-
lation in either op. 26 or op. 78, the latter occupies a position somewhat further
along the continuum of recapitulatory possibilities due to the suppression of the
tonic scale step at bar 156” (ibid., 104).
38. For Smith, “the theme of op. 78 floats in only at the end of the first bar, diminish-
ing the assertiveness of the initial tonic” (ibid., 105).
39. See Dillon Parmer, “Brahms, Songs Quotations, and Secret Programs,” 19th-Century
Music, 19 (1995): 161–90.
41. The cadenza doppia (or “double cadence”) is a cadence that uses four metrical units
(beats) on scale degree V. In the simplest kind the four beats bear the following
chords: 35, 46, 45, 35, but there is also a variant with the (prepared) seventh on the first
beat. Double cadences are susceptible to several diminutions, as described (for
instance) by Francesco Gasparini in his treatise L’armonico pratico al cimbalo (Venice,
1708). For a detailed account of double cadences see my The Art of Partimento:
History, Theory and Practice (New York: Oxford University Press, 2012); and Johannes
Menke, “Die Familie der cadenza doppia,” Zeitschrift der Gesellschaft für Musiktheorie 8
(2011): 389–405.
42. “Casting out the root” is the translation suggested by William Rothstein
of Schenker’s term “Auswerfen des Grundtones”; see William Rothstein,
“Transformations of Cadential Formulae in the Music of Corelli,” in Essays from the
Third International Schenker Symposium, ed. Allen Cadwallader (Hildesheim: Olms,
2006), 245–78. In particular, Rothstein’s example 5 (p. 255) is a cadenza doppia
with the root “cast out.” Cadenze doppie with the root “cast out” are also shown in
Gasparini, L’armonico pratico, 48. Note, however, that Gasparini’s name for cadenze
doppie is cadenze maggiori.
43. Neighboring notes on the second beats of double cadences are shown in Gasparini,
L’armonico pratico, 48.
44. “Linkage technique” is a Schenkerian concept introduced by Oswald Jonas: it
“means that a new phrase takes as its initial idea the end of the immediately pre-
ceding one and then continues independently, either in the same formal unit . . .
or to initiate a new section.” Oswald Jonas, Introduction to the Theory of Heinrich

Schenker: The Nature of the Musical Work of Art, trans. and ed. John Rothgeb (New
York: Longman 1982), 7–8.
45. “In the G-major Violin Sonata, op. 78, a V–I resolution at the structural downbeat
(mm. 133–34) initiates the final section of the development, which prolongs the
minor tonic on the foreground and contains subsequent internal resolutions to
that chord at points of formal articulation (see mm. 140, 141–42, and 148). When
the recapitulation does enter at m. 156, Brahms adds a minor seventh to the
G-major triad, which thereby functions as an applied dominant to IV rather than
as a return to the tonic Stufe. In opus 78, the rhetorical power of the reprise is
diminished by both the harmonic structure of the retransition and the absence of
the structural tonic at the entrance of the large-scale thematic restatement.” Smith,
“Formal Ambiguity,” 103.

Part Six
Schoenberg and Beyond

Chapter Eleven
Dominant Tunnels,
Form, and Program in
Schoenberg’s Verklärte Nacht
Julie Pedneault-Deslauriers
“Zwei Menschen”: Schoenberg and Dehmel

The premiere of Schoenberg’s sextet Verklärte Nacht on March 18, 1902, elic-
ited mixed reviews. Most critics were impressed by the work’s inventiveness
and richness of instrumental colors and sound, and they hailed the young
Schoenberg as a composer to be reckoned with. Others, however, chastised
the sextet on account of its post-Wagnerian harmonic language—rife with
jarring chromaticism, dissonant counterpoint, and relentless sequences—
which they perceived as contrived, harsh, or simply ugly. Several writers also
took issue with the work’s programmatic dimension. In addition to a degree
of general resistance to the idea that chamber music, the last standing bas-
tion of “absolute music,” might yield to the modern taste for programs,
some doubted whether the intricacies of Richard Dehmel’s “Verklärte
Nacht” could lend themselves to musical treatment, while others objected
to what they considered its dubious morality. In the poem, a woman con-
fesses to her beloved that before she met him, she had given herself to a
man whom she did not love because she longed for fulfilment and mother-
hood. She now carries this man’s child. Magnanimously, her companion
vows that their mutual love will transfigure the child and make it his own:
“you will bear it for me, from me.” In his review for the Neue Freie Presse,
Richard Heuberger doused Schoenberg’s transfigurative flame with icy cyn-
icism, deriding the sextet’s “overly-tolerant two-father system for compos-
ing and hearing” and thus heaping contempt both on the program and the
Excerpts from Arnold Schoenberg’s Verklärte Nacht, op. 4 © 1899 by Dreililien-Verlag

Richard Birnbach, Munich, are reproduced by permission of Dreililien-Verlag Richard
Birnbach and Belmont Music Publishers.

346 julie pedneault-deslauriers
work’s admixture of the programmatic and the absolute: “Is this to be told
in music without words? Schoenberg credits composition with such definite
expressive ability!”1
Heuberger’s doubts were compounded by the fact that the public did not
even receive a copy of the poem, contrary to both custom and Schoenberg’s
wishes. Indeed, lest Dehmel’s “Verklärte Nacht” prejudice the audience against
the musical work, the Arbeiter Zeitung critic Josef Scheu suggested, tongue in
cheek, that the concert organizers had “shied away from desecrating the con-
cert program with this chapter taken from the Gospel of free love, and only the
critics, who evidently cannot become more corrupt than they already are, were
given copies of the sinful poem.”2 The controversial lines read:
Two people walk through the bare, cold woods;

The moon runs along, they gaze at it.
The moon runs over tall oaks,
No little cloud dulls the heavenly light,
Into which the black points reach.
A woman’s voice speaks:
I bear a child, and not by you,

I walk in sin alongside you.
I have gone seriously astray.
I believed no longer in good fortune,
Yet still had a great longing
For a full life, for a mother’s happiness
And duty—then I became reckless;
Horror-stricken, I let my sex be embraced
By a stranger
And even blessed myself for it.
Now life has taken its revenge:
Now I have met you, oh, you.
She walks with clumsy gait.

She gazes upwards, the moon runs along;
Her somber glance drinks in the light.
A man’s voice speaks:
The child that you conceived,

Let it be no burden to your soul;
Oh, look, how clear the universe glitters!
There is a radiance about everything;
You drift along with me on a cold sea,
Yet a special warmth glimmers
From you in me, from me in you.
It will transfigure the strange child,
You will bear it for me, from me;

dominant tunnels, form, and program 347
You have brought the radiance into me,

You have made me a child myself.
He holds her around her strong hips.

Their breaths mingle in the air.
Two people walk through the high, clear night.3
“Verklärte Nacht” appeared in Dehmel’s collection Weib und Welt, published

in 1896.4 Though Dehmel is a relatively unknown figure today, one whose work
is often dismissed as unrefined and uninspired,5 the notes that Schoenberg
wrote to accompany the 1950 Columbia recording of his opus 4 underscore
that such was not always the case. “At the end of the nineteenth century,” the
composer reminisced, “the foremost representatives of the ‘Zeitgeist’ in poetry
were Detlev von Liliencron, Hugo von Hofmannsthal, and Richard Dehmel.”6
The poet was a member of the intellectual coterie that gave birth to the art
and literature magazine Pan; Strauss, Zemlinsky, Webern, Reger, and others set
his poetry. Weib und Welt ensured Dehmel’s position at the vanguard of mod-
ern lyric verse and earned him a success mixed with (and perhaps fueled by)
scandal: its free amalgam of religiosity and eroticism resulted in a trial for blas-
phemy and immorality. Indeed, “Verklärte Nacht” manifests important facets
of Dehmel’s poetics, most importantly a weltanschauung rooted in the sanctifi-
cation of a defiantly expressed sexuality.7
Since the premiere of Schoenberg’s sextet, no consensus about the work’s
form has emerged. A number of scholars have addressed this issue by mapping
the layout of the poem onto that of the sextet. A decade after Webern called
the work a “free fantasy,” Egon Wellesz divided the work into five parts that cor-
respond to the poem’s five sections: “The structure of Verklärte Nacht, in accor-
dance with the poem, is made up of five sections, in which the first, third, and
fifth are of more epic nature and so portray the deep feelings of the people
wandering about in the cold moonlit night. The second contains the passion-
ate plaint of the woman, the fourth the sustained answer of the man, which
shows much depth and warmth of understanding.”8 Carl Dahlhaus and Walter
Bailey also propose a five-part, rondo-like form. Richard Swift and Michael
Cherlin hear a pair of consecutive sonata forms, whereas Wilhelm Pfannkuch
detects an overarching sonata design with an interpolated adagio movement.
Walter Frisch, on the other hand, has raised objections to such Formenlehre
approaches and concentrated instead on dense motivic transformations and
large-scale harmonic strategies.9
My own contribution to the issue of form and its relationship to the pro-
gram focuses on phrase structure and on strategies of cadential articulation,
deflection, and especially expansion, thereby drawing on Formenlehre con-
cepts developed out of Schoenberg’s own theoretical writings. Rather than
revisiting the top-down approach to form that characterizes most existing

studies of the sextet, I privilege close analysis of selected interthematic events,

which I interpret as signposts within the sextet’s large-scale form, and which
I connect to Schoenberg’s comments on the thematic, motivic, and pro-
grammatic layout of opus 4. More precisely, I borrow an important concept
from William E. Caplin’s theory of formal functions, namely the expanded
cadential progression, and I extend its application to a work that combines
an advanced harmonic language with familiar strategies in formal design at
the phrase level. The core of this analysis thus revolves around what perhaps
constitutes the work’s most characteristic harmonic progression: an elabo-
rate cadential formula that contains an ambiguous ninth chord in fourth
inversion. This harmonic-formal unit makes several increasingly varied reap-
pearances throughout Verklärte Nacht, and its transformations both articulate
the transfigurative narrative of the work and play a role in the sextet’s large-
scale layout. Before delving into these analytical considerations, however,
and in order to shed light on the inflammatory social issues that Schoenberg
confronted musically when he selected Dehmel’s poem for his sextet’s silent
libretto, I will begin by situating the work’s program against contemporane-
ous discourses concerning the prevailing patriarchal model of the family
and fatherhood, and suggest that it challenges and offers an alternative to
this model. Then, addressing aspects of both local concluding functions and
large-scale structure, I will offer a hermeneutic reading of Schoenberg’s inge-
nious tactics for cadential expansion and of their consequences for the sex-
tet’s overall formal layout.
Dehmel’s “Verklärte Nacht” and the

Patriarchal Family at the Turn of the Century
In his 1950 program note, Schoenberg offered a provocative interpretation
of Dehmel’s poem. In that text, which associates precise passages from the
sextet with sections of the poem, the composer modifies Dehmel’s story:
the woman, Schoenberg writes, “had married a man whom she did not love”;
“unhappy and lonely in this marriage [m. 50],” she nevertheless “forced
herself to remain faithful [mm. 75–76], and finally . . . she is now with child
from a man she does not love [mm. 105–15]”; she accuses herself of a “great
sin [mm. 138–39].”10 Not only does Schoenberg recast Dehmel’s narrative in
the context of an extramarital affair (in the poem the woman is apparently
unmarried), but he also neatly reverses the traditional poles of matrimonial
virtue and sin. In Schoenberg’s scenario, the woman’s moral offense lies not
in her conjugal infidelity but rather in her having submitted without love
to the prevailing social codes that prescribe marriage as the most desirable
path for women; this scenario also seems to advocate illegitimate parenthood
and the rejection of matrimony. Of course, it is entirely possible, as Cherlin

suggests, that Schoenberg simply misremembered the nuances of Dehmel’s

poem.11 Equally conceivable, however, is that Schoenberg based his interpre-
tation on a knowledge of Dehmel’s personal circumstances: the poet and his
lover Ida Auerbach were themselves in the midst of what became a highly
public affair, one that saw Dehmel assume a paternal role in relation to
Auerbach’s young son.12
I cite Schoenberg’s programmatic commentary here because it touches
on several sensitive questions about what constituted a moral sexual union at
the time the work was composed (and even at the time Schoenberg wrote his
program note), namely, the potentially oppressive dimensions of marriage,
the nature of paternity, and what factors determined the legitimacy of chil-
dren. To begin with, Schoenberg’s program subverts one fundamental tenet of
European social order: the primacy of the patriarchal family, wherein paternity
was defined by biological kinship between a father and his child born within
matrimony. In Verklärte Nacht, legitimate fatherhood ensues neither from mar-
riage nor from ties of blood, but rather from the love that unites man, woman,
and child. To give an idea of the kinds of anxieties that surrounded the ques-
tion of biological lineage in the late nineteenth century, we may briefly turn to
a work that predates Dehmel’s poem by a decade, August Strindberg’s play The
Father (1887). Here, the Captain is driven to madness by doubts about whether
he is in fact his child’s biological father. Obsessed by the suspicion that “the
mother is closer to the child, as it has been discovered that no one can tell
for a certainty who the father of a child is,” the Captain agonizes: “Didn’t you
ever feel ridiculous as a father? I know of nothing so ludicrous as . . . to hear
a father talk about his children. ‘My wife’s children,’ he ought to say. Did you
ever feel how false your position was?”13 Indeed, heredity means so much for
the Captain that he declares: “if the child is not mine I have no control over
her and don’t want to have any.”14
The Father and Verklärte Nacht were by no means exceptional in thematiz-
ing the patriarchal family; indeed, it would be no overstatement to say that
Schoenberg’s program struck a raw social nerve. By the time Schoenberg
wrote his sextet, the history and nature of the patriarchal family had become
a topic of debate and critical study among anthropologists, archaeologists,
sociologists, and feminists, with little consensus as to whether it represented a
summit of social evolution—as anthropologists argued in the wake of Johann
Bachofen’s Das Mutterrecht (1861)—or a degrading contract motivated by mer-
cantile preoccupations—as Friedrich Engels maintained.15 In The Origin of the
Family, Private Property and the State (1884), Engels lumped together the rise of
monogamy, patriarchy, and capitalism, while asserting that the monogamous
family “is based on the supremacy of the man, the express purpose being
to produce children of undisputed paternity; such paternity is demanded
because these children are later to come into their father’s property as his
natural heirs.”16 Engels (like August Bebel) emphasized the deplorable

consequences of this for women, who remained in a state of social, economic,

and sexual subordination.17 Society nevertheless considered marriage the best
course of action for a woman, as it warded off the stigma of spinsterhood and
its attendant poverty. The situation that Schoenberg describes in his program
note reflects these attitudes: marriage was at best a stopgap for the unnamed
woman, through which she tried to escape a life devoid of meaning (the poem
tells us that she “had a great longing for a full life”). Proponents of a differ-
ent type of conjugal relationships included the Swedish feminist Ellen Key
(1849–1926) and Auguste Fickert (1855–1910), the latter a founding member
of the Allgemeiner Österreichischer Frauenverein. These activists insisted that only
love between two partners, and not marital status, could make a sexual union
moral, and they believed that children born out of wedlock ought not to face
social opprobrium.18
In Schoenberg’s program, the man’s response to the woman’s confession
implies a critique of the conventional matrimonial model and also articu-
lates a new perspective on what modern fatherhood might entail. This fact
in itself is significant, for as Rosa Mayreder (1858–1938), the prominent
turn-of-the century feminist writer, observed, the nature of paternity received
relatively little attention in contemporary discourses: whereas nobody con-
tested the importance of motherhood in a woman’s life, “the correlate in
the male psyche, the feelings and images that connect fathers with their chil-
dren, is, in contrast, kept far in the background.”19 Mayreder deplored the
socially prevalent view of fatherhood, namely, its construction solely in terms
of social status and grounding in relationships of ownership and authority.
For Mayreder, fathers should reconnect with a primordial but now lost pater-
nal instinct that, adapted to modern times, could supplant the traditional
equation of fatherhood with power relationships, thereby eliminating the
need for oppressive patriarchal institutions and transforming the dynamics
of the couple. Mayreder’s vision of fatherhood thus voids the obsessive need
to ensure the legitimacy of one’s children and emphasizes the affective bond
between man and woman, and man and child.
The idea of fatherhood put forth both in Dehmel’s “Verklärte Nacht”
and in Schoenberg’s interpretation of it links up in important respects with
Mayreder’s views. It too stems from a sexual morality that is independent of
the politics of marriage and renounces traditional notions of ownership
and authority. It would be naive, of course, not to acknowledge that the text
hardly escapes the most conventionally Christian trope of the sinful (and
subordinate) woman being forgiven by the godly (and superordinate) man.
Nevertheless, the poem’s valorization of sincere love between the sexes as the
proper basis for bringing forth children offers a progressive alternative to the
traditional idea of the paterfamilias possessing wife and children, and enables
the transfiguration of the familial relationships that initially hold between the
characters of the sextet.

A Dominant Tunnel
While the mention of Verklärung usually elicits descriptions of Schoenberg’s
glittering moonlight music, I will focus instead on how a striking cadential
progression, which holds significant formal, harmonic, and programmatic
implications, brings about important dramatic turning points in opus 4. This
distinctive progression sounds for the first time when, after an introductory sehr
langsam, an agitated viola melody launches the section that Schoenberg relates
to the woman’s avowal of guilt (ex. 11.1). Writing about Schoenberg’s next
instrumental work, Pelleas und Melisande, Steven Vande Moortele has observed
that it offers “superb examples of how well Formenlehre categories developed out
of Schoenberg’s theoretical work after his death apply to Schoenberg’s own
music.”20 The same can be said of Verklärte Nacht. Measures 29–45 comprise two
sections (mm. 29–33 and 34–45) that both exhibit the melodic characteristics
of a sentence: a presentation consisting of a basic idea and its repetition, and
a continuation featuring an intensification of surface activity as well as frag-
mentation. I will call these units Sentences 1 and 2.21 Together, they express
a complete harmonic progression with an initiating tonic and a lengthy pre-
dominant prolongation, followed by an attempt at an authentic cadence when
the dominant enters at measure 41. The theme can therefore be said to articu-
late an expanded cadential progression (ECP). As defined by Caplin, an ECP
occurs when any or all of the constituents within a progression leading to a
cadence are inflated through prolongation, metrical expansion, or decelera-
tion of harmonic rhythm, such that they support one or more full phrases.22
Schoenberg’s particular use of this device, of course, reflects the fact that he
was composing in 1899 rather than in 1799. When an ECP occurs in a classical
sentence, it typically underlies the complete continuation phrase (whereas a
normative cadential idea occupies a continuation’s last two measures). Here,
however, Schoenberg employs the ECP to give a broader harmonic sweep to
the whole theme and build tremendous momentum toward the emphatic
cadential dominant of measure 41. (For this reason, a melodic scale-degree
3^ in the bass, which usually marks the onset of the drive to the cadence, is
omitted.) Following Schoenberg, who identifies Sentence 2 as “the woman[’s]
confess[ing] a tragedy to the man in a dramatic outburst,” I will refer to the
entire passage spanning measures 29 to 45 as the “confession theme.”23
The ECP, however, takes an unexpected turn at measures 41–45, a strik-
ing musical moment marked by abrupt rhythmic, textural, and motivic liq-
uidation. Neither of the cadential dominants at measures 41 and 45 resolves
to the tonic; rather, dissonant chords redirect the harmony into ambiguous
territory. I shall call this passage a “dominant tunnel” on account of its dis-
orienting, labyrinthine quality. Immediately following the first dominant, the
tunnel’s most distinctive harmony sounds at measure 42: an A-flat ninth chord
in fourth inversion, marked by an asterisk in example 11.1 (and to which a

Example 11.1. Schoenberg, “confession theme” with dominant tunnel 1.
Schoenberg’s program note: “The woman confesses a tragedy to the man in a
dramatic outburst.”
(continued)


disgruntled Schoenberg attributed the rejection of his sextet by the Wiener

Tonkünstlerverein.)24 A rhythmic reduction is given in example 11.2. For this
enigmatic chord, David Lewin has suggested two harmonic progenitors (to
use a metaphor in keeping with the program of the work).25 Examples 11.3a
and 11.4a summarize the ninth chord’s affinities with a German-sixth chord
ornamented by chromatic passing tones, on the one hand, and with a simi-
larly decorated deceptive submediant, on the other. In Schoenberg’s score, of
course, the bass B♭ moves up to B♮ before either a German sixth or a subme-
diant materializes (exx. 11.3b and 11.4b). In Lewin’s view, the ninth chord’s
functional ambivalence mirrors the woman’s emotional conflict: as an aug-
mented sixth, it pulls back to the dominant at measure 45, thus expressing “the
[woman’s] urge to avoid revealing her secret,” while as a submediant, it “yearns
away from V,” reflecting her sense that “she must force the matter to a head.”26
If expanded dominants abound in normative ECPs, they do not, as a
rule, disrupt an underlying sense of dominant harmony. By contrast, once
Schoenberg departs from the dominant at measure 41, we veer onto an unpre-
dictable harmonic path that by no means promises to return to V. Nevertheless,
measures 41–45 certainly evince the “sense of heightened drama inherent in
the gesture [of expanding V]”27 that Caplin detects in several ECPs, and retro-
spectively at least, the creeping voice leading as well as the textural, registral,
and textural uniformity provide powerful elements of continuity that connect
the two V chords within the same cadential trajectory. And since the twofold
Example 11.2. Motion to and from the ninth chord, rhythmic reduction,
mm. 41–45
Example 11.3. (a) Hypothetical derivation for the ninth chord: the German sixth
(after Lewin); (b) Schoenberg’s score (reduction)

Example 11.4. (a) Hypothetical derivation for the ninth chord: the submediant
(after Lewin); (b) Schoenberg’s score (reduction)
iteration of the cadential dominant arouses strong expectations of closure,

the ultimate absence of a cadence is doubly conspicuous. Indeed, the sec-
ond dominant also fails to resolve to a tonic; rather, it leads deceptively to a
diminished-seventh chord (F♯–A–C–E♭ in m. 46). The F♯ and the A suggest
that the chord might be heard as a chromatic tonic substitute, but F♯ is soon
reinterpreted as the sixth degree of B-flat minor, the tonality of the next theme
(which Schoenberg associates with the woman’s unhappy marriage).28 The
painstakingly dug harmonic tunnel of measures 41–45, in sum, bores a chasm
of psychological, tonal, and formal irresolution.
Transfigured Cadential Progressions

Through the remainder of the piece, Schoenberg subjects the dominant
tunnel to a series of reconfigurations, each reinterpreting its harmonic and
programmatic ambiguities. The tunnel next sounds at measure 181 within a
condensed version of the confession theme, positioned at the end of the sec-
tion that Schoenberg, Wellesz, Cherlin, and Swift relate to the portion of the
program in which the woman speaks (ex. 11.5). Here Sentence 1 is missing
while Sentence 2 emerges out of a highly unstable harmonic context. At mea-
sure 179, an emphatic viio43/V directs the harmony back to the dominant of D
minor: might this mark the onset of an ECP that could successfully bring about
cadential closure? At first, measures 181 and following appear to replicate
measures 41–45: the same ninth chord (voiced in a higher register), the same
creeping chromatic lines, and the same cryptic harmonic progression engen-
der the same sense of limbo. But Schoenberg offers a couple of twists. First, in
an instance of thematic telescoping, he integrates the main melodic idea of
Sentence 1 into the dominant tunnel of Sentence 2. Second, the outcome of
this tunnel differs dramatically from the progression’s initial occurrence. After
the cadential 46 of measure 181, we expect the dominant to sound again at mea-
sure 185, replicating the return to V of measure 45. But if (following Lewin)

Example 11.5. Return of the confession theme, dominant tunnel 2
(continued)

we interpreted the tortuous return to V at measure 45 as the expression of the

woman’s reluctance to confess, a second dominant here would seem unneces-
sary since the woman has already divulged her secret. And indeed Schoenberg
dispenses with the chord altogether, replacing it with a beat of dead silence in
measure 188. Nor will the following measures supply the long-awaited cadence.
Rather, Schoenberg lunges directly into a bridge to the sextet’s second half, a
section in B-flat minor where implacable, hammering chords (mm. 202–11)
represent the woman walking “in desperation . . . beside the man with whom
she has fallen in love, fearing his verdict will destroy her.”29 The first half of the
sextet thus ends without cadential or tonal resolution; the fizzling of the ECP
into anguished muteness eloquently signals the end of the woman’s confession
and generates suspense: how will the man answer?
If the tunnel has characterized the woman up to this point, in the remain-
der of the sextet it integrates materials related to the program’s two other char-
acters: the man and the child. Before turning to the third tunnel, let us address
these materials. In many respects, the radiant D-major theme at measure 229
(ex. 11.6) marks a turning point in the work. It is the beginning of Swift’s and
Cherlin’s second sonata form and of Pfannkuch’s interpolated adagio move-
ment; even Frisch, generally reluctant to propose Formenlehre divisions, parti-
tions the work into two sections at this juncture. This is also a pivotal moment
from the point of view of dramatic evolution. According to Schoenberg, the
cello theme represents the lover’s charitable answer to the woman: “The voice
of a man speaks, a man whose generosity is as sublime as his love.”30 This

Example 11.6. Chorale theme. Schoenberg’s program note: “The voice of a man
speaks, a man whose generosity is as sublime as his love.”
warm, declamatory theme resembles a chorale with its homophonic texture,

conspicuously diatonic idiom, short phrases demarcated by cadences similar
to fermatas, and its plagal cadence in measures 232–33—all with the result that
the man’s response takes on a quasi-divine glow. (I will subsequently refer to
this theme as the “chorale theme.”) The man’s answer dramatically reorients
the direction of the work’s narrative, and future occurrences of the dominant
tunnel will bear the marks of this reorientation.
Several pages of scintillating moonlight music in F-sharp major follow
the chorale theme. Then, toward the end of a long section that expresses
“‘the warmth that flows from one of us into the other,’ the warmth of love”
(Schoenberg is quoting Dehmel), a new theme sounds in D-flat major, a theme
that the composer identifies with the child’s Verklärung (mm. 320–22, ex. 11.7a).
In Schoenberg’s words, it “corresponds to the man’s dignified resolution: this
warmth ‘will transfigure your child,’ so as to become ‘my own.’”31 This passage,
transfiguration motive 1, soon spawns a closely related idea, transfiguration
motive 2 (ex. 11.7b, mm. 332–36). Both motives involve the juxtaposition of
the dominants of D flat and D, a pairing that subtly relates to the dominant tun-
nel by picking up an important motive of the confession theme: the irregular

Example 11.7a. Transfiguration motive 1, mm. 320–22, 323–24. Schoenberg’s

program note: “another new theme, which corresponds to the man’s dignified
resolution: this warmth ‘will transfigure your child,’ so as to become ‘my own.’”
resolution of V over melodic degree ♭6^ in the bass. Moreover, the successive
dominants in both motives recall a harmonic scenario proposed by Lewin for
the initial tunnel, that is, the progression from V to a German sixth. Indeed, as
Frisch has remarked, the pitches of D major’s dominant seventh enharmonically
spell the German sixth of D-flat major (A–C♯–E–G in D major becomes B♭♭–
D♭–F♭–G).32 Schoenberg builds up tremendous momentum through repetition
in measures 320–24 and 332–36 as the augmented-sixth chord keeps the music
circling back to V7 in D-flat major in unsuccessful attempts to arrive at a cadence;
just as in the original tunnel, the cadential potential of the dominant remains
frustrated. All of this, of course, takes place in the “wrong” key: it is a cadence in
D major, not D-flat major, that is needed to effect definitive tonal and dramatic
closure. Accordingly, at measure 336, the German sixth of D-flat serves as a pivot
back into D major, and transfiguration motive 2 appears transposed up a semi-
tone and so suggests that cadential resolution might finally be close at hand.
The return of the tonic major coincides both with a contrapuntal combina-
tion of tremendous programmatic significance and with the next appearance
of the tunnel. At measures 338–44, Schoenberg conflates materials associated
with all three of the poem’s characters: the woman, the child she carries, and
the man. Example 11.8 shows how transfiguration motive 1 (which, as we

Example 11.7b. Transfiguration motive 2, mm. 332–33, 334–35, 335–36
recall, Schoenberg linked to the “strange child”) elides with the chorale theme
in measure 340 as the cello reiterates the man’s declaration of paternity. This
elision is supported by an ECP whose dominant prolongation strongly evokes
the sextet’s first dominant tunnel (and its associations with the woman).
Although significantly abridged and rendered diatonic, the initial tunnel is
recognizable through its emphasis on melodic degrees 3^ and 2 ^
in the first vio-
lin, its Spartan rhythm and texture, and its chromatically rising bass from
scale degree 5 ^
through ♭6^ to ♮6^. At this point, the tensions that have accumu-
lated over the preceding measures finally diffuse, as Schoenberg replaces the
ninth chord by a much less dissonant deceptive verticality over the bass B♭.
This overdetermined moment not only turns the child of a sinful union (to
paraphrase early critics) into the child of a loving one (and thus, by Dehmel’s
standards, a moral one) but also turns a lover into a father and deepens and
consecrates the relationship between the man and the woman.
Yet for all its significance as a moment of dramatic denouement, the con-
junction of themes in measures 338–44 does not reach a strong, authentic
cadence in D major: the concluding tonic of the chorale theme (m. 343) is
brought about by IV, not by a dominant. This absence of authentic cadential
closure motivates yet another reworking of the tunnel in measures 363–69. As
example 11.9 shows, the ascending chromatic bass remains, but Schoenberg

Example 11.8. Dominant tunnel 3, transfiguration motive 1, chorale theme

Example 11.9. Dominant tunnel 4 and cadential resolution

now dissipates much of the tunnel’s harmonic ambiguity and thus further solid-
ifies the music’s hold on D major. This time, the bass’s lowered sixth degree,
spelled as A♯ rather than B♭ (m. 363), sustains a secondary dominant of vi.
Driving home the cadential dominant, Schoenberg gives a threefold state-
ment of the cadential 3^–2^ in the first violin (mm. 365, 366, 368–69), with 2 ^
sup-
33
ported by root-position dominant-seventh chords. As a finishing touch, the
composer decorates the root of V with a reminder of the persistent 5^–♭6^ bass
motive: the neighboring figure A–B♭–A in measure 369 transforms what was
harmonic angst into a delicate ornament, following which Schoenberg finally
allows the dominant to proceed to the tonic in measure 370.
Bruno Walter, in preparing to conduct a 1943 performance of the arrange-
ment for string orchestra, detected a certain redundancy in the sextet’s sec-
ond half and suggested a cut of some fifty measures to rebalance the work.
Schoenberg flatly refused, motivated perhaps, as Frisch suggests, by a desire
to maintain broader tonal processes.34 But programmatic considerations may
have weighed just as heavily. For one, Walter’s cut would have eliminated our
example 11.8, where musical materials related to all three characters sound
together in counterpoint—a climactic moment in the piece’s transfiguration
narrative. Moreover, the dominant tunnel’s role in another large-scale formal
process might also help explain Schoenberg’s reluctance to abridge his sextet.
Brian Alegant and Don McLean have coined the term “closing parallelism”
to describe the common nineteenth-century technique of concluding a given
formal section with material that has served to conclude one or more ear-
lier units. Closing parallelisms typically involve “passages [that] are normally
cadential or post-cadential in formal-harmonic function.” They are “heard nei-
ther as mere recurrence nor as recapitulation, but rather as a weighted ending
again—a kind of rhyming strategy.”35 Although their morphologically flexible
harmony repeatedly deflects closure, Schoenberg’s dominant tunnels never-
theless create such rhyming effects throughout the sextet. They can therefore
help us address thorny questions about the sextet’s large-scale form. Analysts
often describe opus 4 as a laboratory for a technique Schoenberg would refine
in later instrumental works: the integration of multiple movements within a
single, continuous form. To address compositions that express an overarching
sonata form and a multimovement cycle at the same hierarchical level, Vande
Moortele has developed a theory of “two-dimensional sonata form”; he gives
the first Chamber Symphony, Op. 9, and Pelleas und Melisande as examples in
Schoenberg’s oeuvre. Although Verklärte Nacht hardly resembles a sonata form
and does not express a complete four-movement sonata cycle, it nevertheless
projects elements of these two dimensions, and the recurring dominant tun-
nels mediate between them.
Table 11.1 takes a bird’s-eye view of the sextet and combines aspects of
Cherlin’s, Pfannkuch’s, Frisch’s, and Swift’s analyses. Its first part functions as a
sonata form—one that, however, lacks a proper recapitulation and a home-key

perfect authentic cadence, and therefore remains open-ended. A slow move-

ment ensues in the second part (along with Frisch, I find the two-sonata-form
reading too restrictive for the music’s thematic and tonal fluidity), which
exhibits a pronounced degree of formal independence, a wealth of new
themes, and a closed tonal scheme in D major. In Vande Moortele’s termi-
nology, its onset marks a point of “dimensional disconnection”; that is, this
fresh formal unit seems relatively autonomous and bracketed off from what
came before, prompting the listener to perceive it as belonging to a separate
formal dimension, namely, that of the multimovement design. Its appear-
ance also encourages us to hear the incomplete sonata form that preceded it
both as a first movement and as a unit that initiates an overarching form, the
unfinished business of which will be addressed once the embedded move-
ment has concluded.
Vande Moortele notes that composers typically integrate these embedded
movements into the overarching form, thereby tightening the cyclic effect
of the whole. In Verklärte Nacht, the dominant tunnel’s rhyming effect serves
this purpose. These insistent “endings again” (to recall Alegant and McLean)
are not only pregnant with programmatic meaning; they also sound at junc-
tures that delineate the two dimensions of large-scale formal organization that
operate in the work. As table 11.1 shows, the tunnel concludes the overarch-
ing sonata form’s main theme, which Schoenberg’s program relates to the
woman’s sense of guilt (ex. 11.1 above). It then rounds off the whole section
expressing the woman’s confession at the end of the sextet’s first half (ex.
11.5), but what began as a recapitulatory gesture morphs into the transition
to the slow movement. Together with the chorale theme and transfiguration
motive 1, a modified tunnel sounds at the apex of a lengthy climactic buildup
(ex. 11.8) that ushers in that slow movement’s epilogue. Schoenberg then
evokes the dominant tunnel to bring the slow movement to a close with a PAC
in D major (ex. 11.9). By sounding (or suggesting itself) at key endpoints in
the syntactically incomplete sonata form and the slow movement, the tunnel
creates a large-scale formal rhyme across music belonging to the two distinct
formal dimensions and unites them thereby.
One final tunnel remains, which effects an apotheosis of the processes we
have thus far addressed: measures 390–401 bring tonal and formal closure to
the overarching form and integrate material from the interpolated slow move-
ment into that overarching dimension. Here (ex. 11.10), the tunnel appears
just as it was in the sextet’s first half, intact with its converging lines and ninth
chord. It thus links back to the unfulfilled recapitulation of the initial sonata
form. But whereas there it failed to achieve closure, here it consummates the
function of recapitulation by arriving at a PAC. Moreover, Schoenberg grafts
the tunnel onto transfiguration motive 1 and other materials from the slow move-
ment, thus extending its recapitulatory effect to material from the embedded
movement. In sum, it makes explicit the “rhyme” that clinches the effect of

Table 11.1. Dominant tunnels and large-scale form
Part I Part II

Sonata form Interpolated slow mvt. (Pfannkuch)
(Swift, Intro. Exposition Develop- Intro.= A B C Epilogue Coda

Cherlin, (1−28) (29−131) Trans. Subordi- ment Chorale Th.
Pfannkuch) Main theme (50−104) nate th. (132−228)
(29−49) (105−31)
Dominant DT 1 DT 2 DT 3 (340−42) DT 4 DT 5
tunnels (41−45) (181–87) Supports (363−69) (cf. DT 1)
(DTs) Implies transfigura- Concludes (393−400)
return to tion motive 1 the slow PAC; tonal,
D minor, and chorale mvt. with a formal, and
but further th.; leads to PAC program-
develop- concluding matic
ment leads section of slow closure of
the music mvt. the work.
toward
the slow
movement
Over- Open-ended sonata form Recapitula-

arching tion
form
Multi- Slow First movement Second movement No final

movement intro. movement,
cycle but final
“rhyme”
with DT 5
9/30/2015 7:54:08 PM
Example 11.10. Dominant tunnel 5 (cf. dominant tunnel 1) and PAC in D major

its closing parallelism and definitively voids the need for another tunnel to
appear: for tonal, formal, and programmatic reasons, there is no doubt that
when the cadential dominant enters at measure 393, it will finally lead to a
tonic, and a major one at that—literally, the light at the end of the tunnel.
This last harmonic-formal unit represents the telos of the piece on a number
of different levels: the dominant tunnel finally leads to authentic cadential artic-
ulation; chromaticism resolves into diatonicism; and the original minor-mode
dominant prolongation returns in D major. Schoenberg’s denouement resonates
with two of Western culture’s archetypal instances of transfiguration—one bibli-
cal, the other operatic—and I will arrive at my own final cadence by exploring
these parallels. Three of the four Gospels tell of how Jesus was transfigured in
front of a group of apostles: “His face shone like the sun, and his clothes became
as white as the light.”36 The biblical notion of transfiguration is relevant because
it embodies a moment where God claims Jesus for his son: “Behold, a bright
cloud overshadowed them: and behold a voice out of the cloud, which said, This
is my beloved Son, in whom I am well pleased” (Matthew 17:5). This claiming
and revealing of spiritual filiation runs close to the kind of paternity that Verklärte
Nacht upholds, and we might read Schoenberg’s conclusion as the gesture that
finalizes the process of transfiguration. The sextet draws on a venerable sym-
phonic pedigree that includes Beethoven’s Fifth and Ninth Symphonies, as well
as Schumann’s Fourth, Brahms’s First, and Bruckner’s Third and Eighth, all of
which trace a minor-to-major, darkness-to-light trajectory. (Beethoven’s Ninth,
Schumann’s Fourth, and Bruckner’s Third also share the key of D with Verklärte
Nacht.) In Schoenberg’s composition, as we have noted, highly modified forms
of the dominant tunnel, originally in D minor, sound in D major throughout the
second part of the work. At measure 391, however, the dominant prolongation
returns in its original guise, but with one important change: its mode. Just as the
voice of God claims Jesus as his son, D major—the key of “The Father” in the sex-
tet—here claims the original form of this progression. And Schoenberg lingers
insistently on this modal transfiguration, drawing out over four measures (mm.
397–400) the only element in the tunnel that unequivocally conveys mode: the
^ ^ ^
repeated melodic degree 3 (as part of the 3–2 over the cadential 46–35).
The sextet’s narrative also makes points of contact with the quintessential
operatic transfiguration, that of Isolde.37 A systematic examination of these
parallels would exceed the scope of this study, but we can briefly note how, in
both works, transfiguration involves an intimate communion with a sensuous,
all-enveloping environment: the “blissful fragrances” and resounding “aerial
waves” that Isolde perceives recall the shimmering glow that bathes Dehmel’s
couple when the man marvels: “Oh, look, how clear the universe glitters! There
is a radiance about everything.”38 In Verklärte Nacht, the protagonists unite as “a
special warmth glimmers / From you in me, from me in you,” while in Tristan,
a “wondrous, glorious tune . . . from him flowing, through me pouring” dis-
solves the lovers’ individual consciousnesses.

Night imagery, of course, is paramount in both works, where night repre-

sents a realm of truth and revelation that contrasts with the festering deception
of daylight. Yet there are also differences. For the romantics, night was a source
of human consciousness, or, as Dieter Borchmeyer writes, “life is the flashing
foam on the wave of night, created by it and returning to it.”39 Tristan as a
whole runs a teleological course toward night as death and as consummation
of desire. But only as transfigured, disembodied spirits will the couple be eter-
nally united. By contrast, the transfiguration in Verklärte Nacht anchors the lov-
ers firmly to a terrestrial, carnal life (and love), in the form of the child to be
born and the lovers’ relationship to it. We might say that Tristan and Verklärte
Nacht ride “life’s flashing foam” on opposite segments of the wave: Verklärte
Nacht on the crest and Tristan on the trough.
Finally, both works draw upon a similar harmonic strategy to consummate
their respective transfigurations: the resolution of highly ambiguous chromat-
icism into diatonicism. As Alegant and McLean explain, in Tristan “the fun-
damental pre-dominant versus dominant ambiguity of the Tristan chord . . .
returns in the final bars of the Liebestod, where it is inflected in an unambigu-
ously transfigurative plagalism that ‘resolves’ the principal idea of the drama
and transcendentally frames the work.”40 Unlike this concluding Tristan chord,
the aberrant ninth chord of Schoenberg’s sextet is not ultimately inflected or
absorbed into a different progression. Nevertheless, contrary to the first part
of the sextet, it ultimately figures in an ECP that confirms rather than thwarts
cadential expectations. This long-delayed closure puts in relief the creative har-
monic and formal treatment that the dominant tunnel receives over the course
of the work. Schoenberg underscores the dramatic turning points of Dehmel’s
transgressive poem with a progression that itself defies a host of conventions
associated with cadential function. The tunnel’s chromatic bass comes to par-
ticipate in a variety of original cadential formulae that dodge closure and pro-
pel the sextet ahead. Among the numerous aspects of Verklärte Nacht that mark
the sextet as a watershed in Schoenberg’s tonal output, the daring harmonic
and formal ambiguities of the dominant tunnel certainly deserve emphasis.
The tunnel’s very last appearance, after which only a tonic pedal remains, is
laden with form-functional, tonal, and programmatic weight, which only rings
in its proper fullness when we allow the meanings accumulated through the
process of transfiguration to be heard resounding within it.
Notes
1. Richard Heuberger, review of Verklärte Nacht, Neue Freie Presse, March 24, 1902; trans.
in Walter B. Bailey, Programmatic Elements in the Works of Arnold Schoenberg (Ann
Arbor, MI: UMI Research Press, 1984), 10–12.
2. Josef Scheu, Arbeiter Zeitung, March 27, 1902; quoted by Dorothee Schubel in sec-
tion 6, series B, vol. 22 of Arnold Schönberg: Sämtliche Werke (Mainz: Schott; Vienna:

Universal, 2000), 89 (my translation). For a thorough overview of Verklärte Nacht’s

early critical reception, see Esteban Buch, Le Cas Schoenberg: Naissance de l’avant-
garde musicale (Paris: Gallimard, 2006), 66–75.
3. Translation slightly modified from Walter Frisch, The Early Works of Arnold Schoenberg,
1893–1908 (Berkeley: University of California Press, 1993), 111.
4. Richard Dehmel, Weib und Welt (Berlin: Schuster und Loeffler, 1896). A number of
subsequent editions do not feature “Verklärte Nacht,” probably because Dehmel
included it in his Zwei Menschen: Roman in Romanzen of 1903.
5. Robert Vilain writes that “it was and is by no means universally acknowledged that
Dehmel’s effusions represent poetry of a particularly high quality.” See Vilain,
“Schoenberg and German Poetry,” in Schoenberg and Words: The Modernist Years, ed.
Charlotte M. Cross and Russell A. Berman (New York: Garland, 2000), 6.
6. Arnold Schoenberg, program note for Verklärte Nacht, The Music of Arnold Schoenberg,
vol. 2, CBC Symphony, Robert Craft, Columbia Records M2S 694, 1950. The pro-
gram note also appears in Joseph Auner, A Schoenberg Reader: Documents of A Life
(New Haven, CT: Yale University Press, 2003), 38–40.
7. On Dehmel’s life and works, see Horst Fritz, Literarischer Jugendstil und
Expressionismus: Zur Kunsttheorie, Dichtung und Wirkung Richard Dehmels
(Stuttgart: J. B. Metzler, 1969); and Sabine Henning, ed., WRWlt—o Urakkord:
Die Welten des Richard Dehmel (Herzberg: Bautz, 1995). Walter Frisch has under-
scored Dehmel’s crucial influence on the development of Schoenberg’s musi-
cal language by attributing the stylistic and technical advances of 1897–99 to
his discovery of Weib und Welt, which the composer found riveting. See Frisch,
The Early Works of Schoenberg, as well as Frisch, “Schoenberg and the Poetry of
Richard Dehmel,” Journal of the Arnold Schoenberg Institute 9 (1986): 137–79. See
also Vilain, “Schoenberg and German Poetry,” 1–30; and Martina Sichardt, “Zur
Bedeutung der Dichtung Richard Dehmels für die Liedkomposition um 1900,”
in Neue Musik und Tradition: Festschrift Rudolf Stephan zum 65. Geburtstag, ed. Josef
Kuckertz et al. (Laaber: Laaber, 1990), 365–88. Julius Bab offers the perspec-
tive of a near-contemporary in Richard Dehmel, die Geschichte eines Lebens-Werkes
(Leipzig: Hässel, 1926).
8. Anton Webern, “Schönbergs Musik,” in Arnold Schönberg (Munich: Piper, 1912), 23;
Egon Wellesz, Arnold Schönberg, trans. W. H. Kerridge (1921; repr. New York: Da
Capo Press, 1969), 67.
9. Carl Dahlhaus, “Schoenberg and Programme Music,” in Schoenberg and the New
Music: Essays, trans. Derrick Puffett and Alfred Clayton (Cambridge: Cambridge
University Press, 1987), 96–97; Bailey, Programmatic Elements in the Works of Arnold
Schoenberg, 35; Richard Swift, “I/XII/99: Tonal Relations in Schoenberg’s Verklärte
Nacht,” 19th-Century Music 1 (1977): 3–14; Michael Cherlin, “Schoenberg and the
Tradition of Chamber Music for Strings,” in The Cambridge Companion to Schoenberg,
ed. Jennifer Shaw and Joseph Auner (Cambridge: Cambridge University Press,
2010), 30–52; Wilhelm Pfannkuch, “Zu Thematik und Form in Schönbergs
Streichsextett,” in Festschrift Friedrich Blume zum 70. Geburtstag, ed. Anna Amalie
Abert and Wilhelm Pfannkuch (Kassel: Bärenreiter, 1963), 258–71; Frisch, The
Early Works of Schoenberg, 109–39. Catherine Dale surveys several formal analy-
ses in Schoenberg’s Chamber Symphonies: The Crystallization and Rediscovery of a Style
(Aldershot: Ashgate, 2000), 57–58.
10. Schoenberg, program note for Verklärte Nacht (emphasis added).
11. Cherlin, “Schoenberg and Chamber Music,” 42.

12. “Verklärte Nacht” was inspired by the poet’s relationship with Auerbach (née
Coblenz), whom he met in 1895. Earlier that year, Ida had complied with her
father’s wish that she marry Leopold Auerbach, and soon thereafter she became
pregnant. The marriage was unhappy, and she began an affair with Dehmel (him-
self at the time married to Paula Oppenheimer). Ida and Dehmel eventually mar-
ried in 1901. The attitude of the man in the poem toward the unborn child was
in all likelihood modeled on Dehmel’s own. In a letter to Ida dating from 1898,
Dehmel asked that her son Heinz-Lux call him “Uncle” instead of “Herr Richard,”
and adds that hopefully, the child will soon call him something else: “Später . . .
wird er schon anders sagen. Nit, Mutter Isi?” After Dehmel and Ida had mar-
ried, Heinz-Lux took the name of Auerbach-Dehmel. Dehmel and Ida became
something of an iconic couple; Rilke’s Stundenbuch (1905) would bear the dedi-
cation “An Zwei Menschen,” after Dehmel’s eponymous work. See Dehmel to Ida
Auerbach, September 19, 1898, in Richard Dehmel, Ausgewählte Briefe aus den Jahren
1883 bis 1902 (Berlin: Fischer, 1923), 284.
13. August Strindberg, The Father, in Plays: The Father, Countess Julie, The Outlaw, The
Stronger, trans. Edith and Warner Oland (Boston: J. W. Luce, 1912), 27, 39–40.
14. Ibid., 46.
15. See Ann Taylor Allen, “Patriarchy and Its Discontents: The Debate on the Origins
of the Family in the German-Speaking World, 1860–1930,” in Germany at the Fin
de Siècle: Culture, Politics, and Ideas, ed. Suzanne Marchand and David Lindenfeld
(Baton Rouge: Louisiana State University Press, 2004), 81–101.
16. Friedrich Engels, The Origin of the Family, Private Property and the State, trans. Alick
West (New York: International, 1942 [1884]), 55.
17. In his Woman under Socialism (1879), Bebel opined that “monogamous marriage,
which flows from the bourgeois system of production and prosperity, is one of the
most important cornerstones of bourgeois or capitalist society; whether, however,
such marriage is in accord with natural wants and with a healthy development of
human society, is another question.” August Bebel, Woman under Socialism, trans.
Daniel De Leon (New York: New York Labor News Press, 1904), 85.
18. Harriet Anderson, Utopian Feminism: Women’s Movements in Fin-de-Siècle Vienna (New
Haven, CT: Yale University Press, 1992), 69.
19. Rosa Mayreder, “The Crisis of Fatherhood,” in Gender and Culture, trans. Pamela
S. Saur (Riverside, CA: Ariadne Press, 2009), 29. See also “Das Problem der
Väterlichkeit,” in Krise der Väterlichkeit, ed. Käthe Braun-Prager (Graz: Stiasny
Verlag, 1963), 93–101. For a critical discussion of Mayreder’s views on fatherhood,
see Anderson, Utopian Feminism, 173–75.
Movement Instrumental Works by Liszt, Strauss, Schoenberg, and Zemlinsky (Leuven:
Leuven University Press, 2009), 110.
21. At a higher functional level, Sentences 1 and 2 together may be understood to
form a hybrid theme, with Sentence 1 functioning as a compound basic idea (an
antecedent without a cadence) and Sentence 2 as a continuation.
22. See William E. Caplin, Classical Form: A Theory of Formal Functions for the Instrumental
254; Caplin, “The ‘Expanded Cadential Progression’: A Category for the Analysis
of Classical Form,” Journal of Musicological Research 7 (1987): 215–57; and Caplin,
“Harmonic Variants of the Expanded Cadential Progression,” in A Composition as
Problem II, ed. Mart Humal (Tallinn: Eesti Muusikaakadeemia, 1999), 49–71.

23. Schoenberg, program note for Verklärte Nacht (text accompanying his example 4).
24. See Schoenberg, Theory of Harmony, trans. Roy E. Carter (Berkeley: University of
California Press, 1978; repr. 2010), 346. This uncommon sonority has elicited a
great deal of analytical attention. Peter Schubert, for example, has demonstrated
how it arises from the convergence of five chromatic lines in contrary motion,
and Ethan Haimo has observed that at the time Schoenberg wrote his sextet, he
appeared to have been especially interested in exploring inverted-ninth sonori-
ties. Haimo points to other inverted-ninth chords in the work (mm. 91, 104, 110,
and 341). None of these chords, however, shares the motivic saliency of the ninth
chord of measure 41. See Peter Schubert, “‘A New Epoch of Polyphonic Style’:
Schoenberg on Chords and Lines,” Music Analysis 12 (1993): 289–319; and Ethan
Haimo, Schoenberg’s Transformation of Musical Language (Cambridge: Cambridge
25. David Lewin, “On the ‘Ninth-Chord in Fourth Inversion from Verklärte Nacht,”
Journal of the Arnold Schoenberg Institute 10 (1987): 45–64.
26. Ibid., 57.
27. Caplin, “The ‘Expanded Cadential Progression,’” 227.
28. Schoenberg, program note for Verklärte Nacht (see his example 5).
29. See example 9 in Schoenberg’s program note for Verklärte Nacht. Haimo calls this
passage the “walking theme” in Schoenberg’s Transformation of Musical Language, 39.
^ ^
Note that the 6–5 motion between the pitches F♯/G♭–F that effected the modula-
tion to vi back in measures 46–50 (ex. 11.1) is now compressed into a neighboring
figure that decorates the recitative-like melody of measures 188 and following. The
^ ^ ^ ^
bass of these measures, in turn, articulates 5–♯5/♭6–♮6 motions in the tonicized
areas of E-flat (mm. 189–90) and D-flat (mm. 191–92, not shown in the example).
^ ^
Starting with the viola’s entry in measure 2, the melodic degrees 5 and 6 play a cru-
cial role throughout the work.
30. Schoenberg, program note for Verklärte Nacht (see his example 10).
31. Ibid., examples 15 and 16.
32. Frisch, The Early Works of Schoenberg, 135.
^
33. I understand the support to scale degree 3 in measures 365, 366, and 368 (as well
^
as in mm. 399–400 in ex. 11.10 below) as a V6 chord in which 3 is borrowed from a
6
cadential 4 structure. This chord, however, could also be understood as a cadential
6
4 in F-sharp, the key to which the original chorale theme cadences. The half-step
descent from D to C♯ involved in the D-major/F-sharp-minor ambiguity, moreover,
reproduces the D/D-flat conflict that is fundamental to large parts of the sextet.
34. Frisch, The Early Works of Schoenberg, 129–39.
35. Brian Alegant and Don McLean, “On the Nature of Structural Framing,” Nineteenth-
Century Music Review 4 (2007): 6. The authors draw an analogy between closing
parallelism and the rhetorical figure of epistrophe, quoting from T. S. Eliot’s Waste
Land: “Do / You know nothing? Do you see nothing? Do you remember / Nothing?”
36. The German Verklärung conveys much more fully than the English “transfiguration”
connotations of light, brightness, and revelation. Verklären is the word used in Luther’s
Bible of 1545: “Und er ward verklärt vor ihnen, und sein Angesicht leuchtete wie die
Sonne, und seine Kleider wurden weiß wie ein Licht” (Matthaeus 17:2).
37. Camilla Bork has briefly touched on this matter in “‘Tod und Verklärung’: Isoldes
Liebestod als Modell künstlerischer Schlussgestaltung,” in Zukunftsbilder: Richard
Wagners Revolution und ihre Folgen in Kunst und Politik, ed. Hermann Danuser and
Herfried Münkler (Schliengen: Argus, 2002), 161–78, esp. 171–74.

38. Translation of “Isolde’s Transfiguration” by Andrew Porter in Robert Bailey, ed.,

Prelude and Transfiguration from Tristan and Isolde (New York: Norton, 1985), 96–97.
39. Dieter Borchmeyer, “The World in a Dying Light: Tristan und Isolde and the Myth
of Night,” in Richard Wagner: Theory and Theater, trans. Stewart Spencer (Oxford:
Clarendon Press, 1991), 338–39.
40. Alegant and McLean, “Structural Framing,” 26.

Chapter Twelve
Form and Serial

Function in Leibowitz’s
Trois poèmes de Pierre Reverdy
Christoph Neidhöfer and Peter Schubert
Trois poèmes de Pierre Reverdy, one of the last works by René Leibowitz (1913–72),
is a set of three short songs for four voices and piano.1 Composed in 1971, a
year before the composer’s death, the work raises interesting questions about
form and style. We will first consider a few excerpts to acquaint the reader
with the unique and unmistakably French flavor of Leibowitz’s musical lan-
guage. We will then discuss Leibowitz the theorist and proceed to analyze his
Trois poèmes, calling upon compositional and analytical concepts discussed in
Leibowitz’s unpublished treatise on form in serial music, “Traité de la composi-
tion avec douze sons” (ca. 1950).
Example 12.1 shows the beginning of the first movement. The opening
piano phrase in measure 1 sounds disjunct and Webernian. It serves to intro-
duce the choral phrase that follows in measures 2–3, which is in contrast very
smooth, with much oblique and stepwise motion. This pairing of a piano solo
measure with a continuing choral phrase is replicated in the following mea-
sures (4–6), suggesting a simple principle of form-building and raising ques-
tions about what will follow.
Example 12.2 shows the beginning of the second movement. Its texture
(double canon) in a simple declamatory rhythm recalls examples from the
choral music of Mendelssohn, Brahms, Schoenberg, and Webern.2 In contrast
to the opening of the first movement, whose phrases are clearly demarcated,
the phrases here almost all overlap.
Excerpts from René Leibowitz’s Trois poèmes de Pierre Reverdy © 1976 by Mobart Music,
renewed, are reproduced by permission of European American Music Distributors
Company, sole US and Canadian agent for Mobart Music Publishers. Excerpts from manu-
script materials held at the Paul Sacher Foundation, Basel, are reproduced by permission.

Example 12.1. Leibowitz, Trois poèmes de Pierre Reverdy, mvt. 1, “Son de cloche,” mm.
1–8, with serial analysis shown below the score (brackets group pitch classes that enter
simultaneously); theme, mm. 1–6, and cadential/intermediary structure, mm. 7–8

form and serial function 375
Example 12.2. Trois poèmes de Pierre Reverdy, mvt. 2, “Air,” mm. 1–5
The texture of the third movement is completely different. The piano open-
ing shown in example 12.3a makes a clear reference to the first Gymnopédie of
Erik Satie. Finally, the end of the third movement, shown in example 12.3b,
includes a succession of major and minor triads, concluding the work on an
extended tonal harmony with a distinctly “French” sound (a major triad in
second inversion with added augmented fourth in the last measure). These
excerpts cause us to wonder: how can the relatively conservative use of a single
twelve-tone row be made to assume such different musical characters; how do
the final chords at the end of the third movement come about; and how does
Leibowitz articulate form in such different environments?

Example 12.3a. Trois poèmes de Pierre Reverdy, mvt. 3, “Soleil,” mm. 1–8, with
serial analysis
Example 12.3b. Trois poèmes de Pierre Reverdy, mvt. 3, “Soleil,” mm. 22−28 (end of
movement)

Leibowitz as Theorist
Although René Leibowitz was a prolific composer, with ninety-two numbered
works including five full-length operas, he is known today primarily through
his writings on the music of the Second Viennese School. We will use his
groundbreaking theoretical work on twelve-tone technique and form as a start-
ing point for the analysis of his own music, drawing on his “Traité de la compo-
sition avec douze sons,” which continues the analytical work from his previous
books on the music of the Second Viennese School.3 Leibowitz’s view of form,
like that of William E. Caplin, stems from Schoenberg.4
Exactly when and how Leibowitz encountered the form-functional termi-
nology of the Schoenberg school is difficult to reconstruct. At the time of
the publication of Leibowitz’s first book, Schoenberg et son école (1947, preface
dated May 1946), Leibowitz was familiar with Schoenberg’s Models for Beginners
in Composition (1942), which introduces many of the basic form-functional
concepts in the context of tonal music. Writings mentioned by Leibowitz
in the “Traité” that address some of these concepts in the context of twelve-
tone music include the article “On the Spontaneity of Schoenberg’s Music”
by Heinrich Jalowetz (1944) and Schoenberg’s article “Composition with
Twelve Tones,” which was first published in 1949 in Leibowitz’s translation.5
(Schoenberg’s article identifies the antecedent–consequent phrase in the first
theme from the first movement of his Wind Quintet, op. 26.) Leibowitz started
corresponding with Schoenberg after the war and visited him in Los Angeles
in 1947–48 and 1950. In Los Angeles Leibowitz assisted Schoenberg, writing
out the full score of A Survivor from Warsaw.6 Claims that Leibowitz had studied
with Webern between 1930 and 1933 and that he had met Schoenberg in 1931
could so far not be confirmed.7
Leibowitz was also in contact with Webern’s student Leopold Spinner
(1906–80) and Schoenberg’s student and former assistant Josef Rufer (1893–
1985). He conducted the 1949 premiere of Spinner’s Piano Concerto (version
for chamber orchestra, 1948), and in a letter of March 12, 1949, Spinner pro-
vided Leibowitz with an outline of the form of his work, using Schoenbergian/
Webernian form-functional terminology.8 Rufer and Leibowitz began to cor-
respond in 1947. Their correspondence hit a sour tone in 1950 as the two
argued over two points concerning Schoenberg’s music. Rufer disagreed with
Leibowitz’s characterization of Schoenberg’s String Trio, op. 45 as tending
toward “total athematicism” and also rejected Leibowitz’s idea that Schoenberg
must have used more than one twelve-tone series in Moses und Aron.9 Rufer
wrote to Schoenberg for clarification and related Schoenberg’s answer to
Leibowitz. Schoenberg stated (as quoted by Rufer): “Leibowitz’s athematic
music: this goes back forty years when, for a short period of time, I claimed
as much. But I retracted this claim soon thereafter because coherence in
music can rely on nothing other than motives, their transformations, and

378 christoph neidhöfer and peter schubert
developments. . . . Not everything is not gold that does not glitter, and some-
thing can be thematic while not looking like it by far.”10 Concerning the use
of more than one series, Schoenberg apparently replied (again as reported in
Rufer’s letter): “It does not seem right to me to use more than one row, but the
main thing is whether the music is good.”11
Schoenberg’s theory, as developed by Caplin, can classify any segment of a
classical tonal piece, and often allows us, merely by looking at a few measures,
to identify which section of a piece a given segment must come from. Leibowitz
applies the same form-functional categories in his writings and, we argue, in
his music. However, his claim that row formations can take on functions com-
parable to those projected by triadic harmonic progressions is a risky business,
and not everything he observed in this respect is clear and precise. Leibowitz
acknowledges as much by emphasizing that the comparisons to tonal music are
analogical and metaphorical.12
Applying form-functional concepts and categories to this music has great
analytical appeal for two reasons: first, in a context where traditional harmony
is absent, they allow us to understand form as an integral whole comprising not
only pitch relations but also many other dimensions of a piece. Second, they
are consonant with the heritage of classical form in the twelve-tone music of
Schoenberg, Berg, and Webern. We will attempt to use the analogies Leibowitz
draws between formal functions in tonal music and those in serial music to see
whether, and to what extent, his own twelve-tone music—which he modeled
closely on the music of Schoenberg and Webern in particular—reflected these
functional differences.
Leibowitz on Structures Closes

Leibowitz borrows the Schoenbergian terms “sentence” (phrase) and “period”
(période) as well as concepts associated with them. The “Traité de la compo-
sition avec douze sons” contains mostly analyses of twelve-tone pieces by
Schoenberg and Webern, and also examines a few pieces by other composers
(Erich Itor Kahn, Leopold Spinner, and Leibowitz himself). In the first few
analyses, excerpts from Schoenberg’s and Webern’s music are aligned with pas-
sages by Haydn, Mozart, and Beethoven in order to show the phrase-structural
parallels between the twelve-tone and tonal pieces. In his examples the modèle
(in the sentence) and the proposition (in the period) both correspond to the
“basic idea” in Caplin’s terminology, and usually consist of a statement of the
row or a subset of it.
In Leibowitz’s theory, “serial functions” are those manipulations that are
applied to the series to differentiate the segments of a piece and smaller for-
mal units. For example, in his analysis of the opening theme from Webern’s
Concerto, op. 24, Leibowitz shows how each formal component of the

sentential theme (modèle, répétition du modèle, modèle de la réduction, variation de la

réduction, cadence) is characterized by a full statement of a series, and how the
choice of series (especially as concerns their particular transpositions) over-
all metaphorically mirrors a tonal progression.13 Leibowitz defines serial func-
tions according to: (1) which parts of the series are used when and where; (2)
how rigorously series are used; (3) the transpositional relationships between
series; and (4) the partitioning (tronçonnement) of the series.14 Concerning the
latter, of particular interest to Leibowitz are symmetrical versus asymmetrical
partitionings of the series, and the way in which each partition determines a
formal component of a structure.15 Thus for Leibowitz, the partitioning of the
series is a key element in defining the theme, and a departure from a charac-
teristic partitioning signals a move to another structural area.16
One thing that makes identifying formal functions in serial music so chal-
lenging is the immense variability of segment length. For instance, the caden-
tial segment, most often included within the first eight measures in a tonal
period, can be a massive extension in twelve-tone music, as Leibowitz shows
in his analysis of the beginning of Webern’s opus 29.17 This variation in unit
length makes it difficult to formulate strict criteria for the identification of
phrase structures, and often makes it hard to apply Leibowitz’s analytical strat-
egy to other pieces, including his own, as we cannot necessarily rely on sym-
metrical phrase lengths as basic formal units.
Leibowitz begins his “Traité” by defining “closed structures” in twelve-tone
music. They “are characterized by complete unfolding of one or several forms
of the row, as well as by generally simple partitioning of the series whose model
is consistently maintained.”18 He concludes:
(1) All these structures, to whichever type they belong, and whatever their
degree of simplicity or complexity, are characterized dodecaphonically by a
simple and rigorous manipulation of the row. (2) This manipulation implies:
(a) a specific partitioning (generally symmetrical) of the row forms, a par-
titioning that is consistently maintained throughout the structure; (b) the
determination of various segments or sections [of a movement] by complete
partitions or complete unfolding of the row forms. (3) If several forms of
the series are used, their transpositional relationship remains constant. Thus
Schoenberg generally uses two forms (prime and inversion), realizing a fifth-
relation. . . . The rigorous adherence to a transpositional relationship estab-
lishes a specific and characteristic harmony which permits repetitions and reprises
(the equivalent of a tonal region), while abandoning such a relationship
(which as we will see appears in other sections of the work), creates a different
harmony, thanks to which it is possible to characterize new structures (trans-
positions, contrasting sections, secondary themes, codas, etc. . . .), structures
that make possible the overall articulation of the musical work.19
Leibowitz’s definition of closed structures bears a strong resemblance to

Caplin’s Schoenbergian definition of “tight-knit” form: “A formal organization

characterized by the use of conventional theme-types, harmonic–tonal stabil-

ity, a symmetrical grouping structure, form-functional efficiency, and a unity of
melodic–motivic material.”20 The consistent partitionings that Leibowitz insists
on recall Caplin’s “symmetrical grouping structure.” Leibowitz finds anteced-
ent–consequent relations in serial music supported by the use of related row
forms (e.g., prime followed by inversion), in what resembles, in twelve-tone
terms, a Caplinian “conventional theme type”; and his reference to a charac-
teristic harmony recalls Caplin’s “harmonic–tonal stability.”
The Theme of the First Movement

The opening of “Son de cloche” (ex. 12.1) clearly exemplifies Leibowitz’s defi-
nition of a closed structure. The first measure contains a complete statement of
the row, and therefore corresponds to a significant segment. We label this row
RI-2, in accordance with Leibowitz’s row table for the work, shown in figure
12.1. The next two measures also contain a complete row form, P-1, sung by
the voices with some piano doublings. (It will become clear later why Leibowitz
considered this to be the prime form of the series.) The instrumentation and
lack of overlap between phrases articulate the segments clearly. We propose
that the piano has the basic idea (m. 1), the voices the contrasting idea (mm.
2–3). These first three measures thus form the antecedent of a period, and the
next three measures the consequent. That the consequent should consist of
similar gestures as the antecedent while inverting the row forms is also a fea-
ture of the consequent phrase in the opening of Webern’s opus 27, third move-
ment, as Leibowitz points out in his own analysis of that movement.21
The consequent is slightly different from the antecedent in the way the
piano row links up with the vocal row. The vocal statement (mm. 5–6), how-
ever, is partitioned exactly as it was in the antecedent (mm. 2–3), albeit using
the inversion of the series (I-4 instead of P-1), bestowing closed thematic status
on the first six measures. The partitioning of the first vocal statement is of great
importance, as it is repeated exactly not only here in the consequent but three
more times later on (we call it a “ritornello”), and it is one of only two ways of
partitioning twelve notes into four segments each containing a unique number
of elements. Here each voice part has either 1, 2, 4, or 5 notes (the other such
partitioning is 1, 2, 3, and 6 notes). Leibowitz realizes this partitioning in two
different ways, analyzed in example 12.4. The melodies are mostly consistent:
the melodic dyad always forms a semitone and the tetrachord is always a chro-
matically ascending or descending segment (all of these are realized as literal
semitone progressions); however, the melodic pentachords represent two dif-
ferent set types, that is, one type of prime form [01236] in examples 4a–c and
another of prime form [01367] in examples 4d–e. This is a subtle difference,
especially in light of the high saturation overall in semitone motion, but it

Example 12.4a. Distributions of the series in the ritornelli of the first movement,
P-1, mm. 2–3
Example 12.4b. Distributions of the series in the ritornelli of the first movement,
I-4, mm. 5–6
Example 12.4c. Distributions of the series in the ritornelli of the first movement,
I-7, mm. 27−28
Example 12.4d. Distributions of the series in the ritornelli of the first movement,
R-3, mm. 10−11
Example 12.4e. Distributions of the series in the ritornelli of the first movement,
RI-2, mm. 12−14

represents a variation in serial function nevertheless, one that results from two
different distributions of segments of one, two, four, and five elements. We shall
call the distribution in examples 4a–c distribution 1 and that in examples 4d–e
distribution 2. Distribution describes the assignment of the members of the
series to the different voices.22
Distribution allows the composer to derive a new line not composed of adja-
cent members of the row or to obtain verticalities containing intervals not adja-
cent in the row. An example of a derived melody occurs in the excerpt shown in
example 12.5 from the first movement of Schoenberg’s Fourth String Quartet,
op. 37, which Leibowitz analyzes in the “Traité.” Here Schoenberg partitions
the series in the second theme to derive a chromatically descending tetrachord
in the cello. In his discussion of this theme, Leibowitz points out that the viola
plays only a fragment of the series, that is, the pitch classes of order positions 3,
4, 5, 6, 7, 9, 10, and 11.23 Although he does not point out specifically that the
cello plays a descending chromatic tetrachord formed by the remaining pitch
classes from the series (order positions 1, 2, 8, and 12), Leibowitz must have
been aware of it as a guiding principle behind Schoenberg’s distribution of the
series between the cello and viola parts.24
New chords result from the polyphonization of the row into different voices,
allowing for simultaneities between potentially any members of the row.25 When
two notes are sounded together, their order constraints are temporarily hidden,
and a special, local type of partitioning occurs briefly. In the first ritornello from
opus 92 (mm. 2–3, ex. 12.1), Leibowitz sounds the C♯ (order number 1) against
all of the remaining eleven notes of the series. The last chord in the aggregate
consists of notes of order numbers 1, 10, 11, and 12, a minor chord with an
added second. The other four ritornelli likewise end with notes of the same
order numbers, creating different chord types depending on whether the row
has been inverted or retrograded or both; that is, a major chord in first inver-
sion with an added fourth in measure 6 (ex. 12.1) and measure 28 (ex. 12.8),
and whole-tone sonorities in measures 11 and 14 (prime form [0248], formed
by the last pitch classes of each of the four segments in exx. 12.4d and 12.4e). In
sum, distribution 1 occurs in the opening antecedent-consequent phrases (mm.
2–3 and 5–6) and does not reappear until the very end of the movement (mm.
27–28), providing closure. Distribution 2 is used in the varied repetition of the
theme in measures 9–14. The variation here thus entails, among other things,
a subtle change in the reading of the series; as we will see, this repetition is also
varied by other means.
Leibowitz on Cadence and on Structures Intermédiaires

Leibowitz does not dedicate a separate discussion to cadential function in the
“Traité,” but the topic comes up in several of his analyses. Key concepts in this

Example 12.5. Schoenberg, Fourth String Quartet, op. 37, mvt. 1, mm. 66–68, with
serial analysis
context are “liquidation” and “neutralization”; they are hard to distinguish in

Leibowitz’s writing because these terms are often used with reference to the
same sections and the same attributes. For Leibowitz, two features in particu-
lar contribute to cadential function: (1) immediate repetition of a few notes,
which has the effect of slowing down the relentless presentation of aggre-
gates;26 and (2) chordal presentation where this has not been a feature of pre-
vious closed formal sections.27
Liquidation is a feature not only of cadences but also of intermediary (i.e.,
transitional) structures, which arise from “(1) the need for the composer to get
rid of his obligations regarding the closed structure he has just presented; (2)
the need to open the musical discourse in order to ‘get on with other things’;
(3) the introduction of a new idea (of a transitory nature) permitting the intro-
duction of a contrasting structure pure and simple (the central section of the
movement, the second theme, etc.).”28 These structures, he explains, occur
in three stages. The first is the repetition of the theme (which, “it goes with-
out saying, must be varied so as not to once more close the structure”).29 In
the second stage, either at the end of the repetition of the theme or immedi-
ately thereafter, “one immediately undertakes the neutralization of the musical
discourse which often ends up presenting only one aspect (neutral) of equal
[rhythmic] values.”30 Finally, the third stage “takes as its point of departure
the equal values presented in the neutralization, values that often allow the
deduction of a modèle de transition [transitional basic idea]. This new idea

(which should no longer have much in common with the theme) is most often
repeated many times in succession until complete saturation [is achieved]. It
[the new idea] leads—sometimes almost imperceptibly, sometimes suddenly—
to the contrasting structure.”31
It may seem alarming to find cadences sharing essential features with transi-
tions. However, once we consider that liquidation in transitions in tonal music
is most often achieved through the reduction to “generic” material (scales,
arpeggios, repeated notes), and that cadences are also generic, conventional-
ized formulas, the parallel may seem less surprising. In both cases, the motives
characteristic to a given section of a piece are eliminated.
Leibowitz explains intermediary structures as follows:
The unique characteristics of various intermediary structures (liquidation,

neutralization, dissolution, and transition) have their unique characteristics
at the level of serial functions. Here the most salient points are: (1) the use
of new serial forms and transpositional relationships (just as in tonal music
new regions are used); (2) new partitionings (which correspond to harmonic
changes in tonal music); (3) a freer and above all a less “complete” treat-
ment of serial forms in order that structures stay open (this corresponds to
the absence of symmetry in the tonal unfolding and to the absence of overly
decisive cadences in the corresponding sections of tonal music); (4) the use
(generally) of a larger number of serial forms (which corresponds to a path
through a larger number of tonal regions). In sum, these functions turn out
to be less stable and less symmetrical than those of closed structures.32
Many of the terms in this paragraph recall Caplin’s definition of the

Schoenbergian term “loose”: “A formal organization characterized by the use
of non-conventional thematic structures, harmonic–tonal instability (modu-
lation, chromaticism), an asymmetrical grouping structure, phrase-structural
extension and expansion, form-functional redundancy, and a diversity of
melodic–motivic material.”33
The first section of the first movement of Trois poèmes (mm. 1–15) presents
intermediary and cadential structures as Leibowitz describes them in the “Traité.”
The opening antecedent–consequent phrase (mm. 1–6) is followed by a caden-
tial segment in measures 7–8 (ex. 12.1) whose vocal/piano phrase sounds at first
like another vocal “ritornello” phrase, but it is built differently, partitioned as 3,
4, 3, 2 among the vocal parts. This repartitioning defines the segment as some-
thing new, something looser. It also contains a certain amount of liquidation, the
two-note repetition in the bass part. This may seem like a trivial detail, but the
new partitioning and the motivic repetition taken together differentiate these
two measures from the first six. Thus this is a cadential/transitional phrase pre-
paring the repetition of the theme that starts in measure 9. Instead of being
included in the consequent as in tonal music, the cadential segment follows the
antecedent–consequent six-measure phrase.34

Leibowitz often insists that repetition be varied. In “Son de cloche” the rep-
etition of the theme (mm. 9–14) is varied by a different introductory piano
phrase, different row forms, a different distribution that produces the changed
pentachord melody we have noted earlier (exx. 12.4d–e), and a liquidation in
the form of a repeated two-note figure in the alto in measure 11 (not shown).35
The large opening A section of the movement concludes with a new tran-
sitional/cadential phrase in measure 15 with pickup (see ex. 12.6): liquida-
tion results from the gesture of two four-note chords in the piano across the
measure, the first occurrence of such attacks in the movement. This has the
effect of “neutralizing” rhythmic activity. This liquidation generates the modèle
de transition, which is the following two chords in measure 15 between piano
and voices. This third stage leads to the middle section of the movement in a
new tempo (poco più scorrevole, m. 16), a secondary structure in this small ter-
nary (ABA‫ )׳‬form.
Leibowitz on Structures Secondaires

Secondary structures are contrasting sections that are “less symmetrical and
less simple” than closed structures but more closed than intermediary struc-
tures.36 Serial loosening plays a central role in this process, with the introduc-
tion of new row forms, new transpositions and new partitioning, liquidation,
and neutralization. Aside from the new tempo, the B section of “Son de cloche”
(mm. 16–26) is set off by the longest piano solo so far (ex. 12.6). This solo fits
Leibowitz’s description of how a transitional structure leads into a secondary
structure, with the “new idea . . . repeated several times in succession . . . [lead-
ing] often almost imperceptibly . . . into the contrasting structure.”37 In this
case, the new idea is the chordal presentation in measure 15 of three four-note
chords that leads into the secondary structure that begins in measure 16 with
the four three-note chords. This piano texture differs considerably from the
earlier piano gestures: four simple triads, alternating major and minor.
What serial functions can account for these four triads? They can be said
to result from two row forms interleaved as shown in example 12.7. The first
hexachord of RI-9 is followed by a clear statement of R-0 in its entirety. Then
the rest of RI-9 follows, with the discrete trichords reversed. In addition, in
a further loosening of serial procedure, the second trichord is reordered.38
One of the hexachords of the row does contain two discrete simple triads,
but they have never been allowed to sound in isolation until now. Further,
the interruption by R-0 makes it possible to follow the two simple triads with
two more. The fragmentation that results from this interpolation is clearly a
new procedure, one that prevents the row from being heard in a complete
form (a requirement of closed forms). We will return to this passage with
another serial reading later on.

Example 12.6. Trois poèmes de Pierre Reverdy, mvt. 1, “Son de cloche,” mm. 14−20. Three
wrong pitches in the published score have been corrected here following Leibowitz’s
manuscript fair copy. Collection René Leibowitz, Paul Sacher Foundation.

Another loosening strategy of this passage consists in the reshuffling of the

internal ordering of the series in measures 17–18. One of the features that
makes this phrase especially “messy” (from a serial point of view) is the rep-
etition of pairs of notes. As we have seen, such repetition is a feature of liq-
uidation. The complexity and looseness of the B section continues with new
partitions of the series and voicings, leading to the long climactic soprano solo
in measures 24–26 (not shown here).
The reprise A‫ ׳‬in measures 27–29 (ex. 12.8) is made backward, the vocal
quartet preceding the piano solo phrase. As mentioned in connection with
example 12.4, voices are partitioned again into 1, 2, 4, and 5 pitches as they
Example 12.7. Interleaving of two series in mm. 16−18 illustrated
Example 12.8. Trois poèmes de Pierre Reverdy, mvt. 1, “Son de cloche,” mm. 27−29
(end of movement)

were in the four statements in measures 1–6 and 9–14, making the reprise obvi-
ous. Finally, the piano also repeats its own initial partitioning (cf. ex. 12.1).
Features of the Overall Form of the First Movement

While some of the elements we have discussed so far play out at the local
level, others have large-scale formal implications. We will now have a closer
look at the latter. Let us begin with properties of the series that structure the
large-scale form. Example 12.9 lists the series that open the first and middle
section of the movement, RI-2 and RI-9 respectively. The latter is a transpo-
sition at pitch-class interval 7 of the former. In other words, the middle sec-
tion starts out in what Leibowitz would call a different region, analogous to
the dominant in tonal structures. A comparison of the two regions, as illus-
trated to the right in the example, reveals the following property: RI-2 com-
bines the two (complementary) hexatonic hexachords that we have labeled
1A and 1B. This hexachord type (the “Ode-to-Napoleon” hexachord, also
used in works such as Webern’s opus 24) with the prime form [014589]
can be realized at only four distinct transpositional levels. They include,
in addition to 1A and 1B, the two complementary hexachords 2A and 2B
that constitute series RI-9. In other words, there are only two distinct ways
of pairing complementary hexatonic hexachords. We will call the combina-
tion of 1A and 1B region 1 and the combination of 2A and 2B region 2.
Each of the forty-eight forms of our series will thus either represent region
1 or region 2.39
Table 12.1 summarizes the formal architecture of the first movement, indi-
cating the performance forces, the forms of the series used, their regions,
and partitions of the series of particular interest. Given that Leibowitz
assumes an analogy between transpositions of keys in tonal music and trans-
positions of series in twelve-tone music, two more features of formal signifi-
cance stand out in the first movement as a whole. The regions that we have
defined classify into two broad categories all transpositions and their retro-
grades, inversions, and retrograde-inversions of the series of opus 92, based
on the (unordered) content of the hexachords. From this perspective we
can recognize another aspect of the form of the first movement. As table
12.1 shows, the A section uses almost exclusively region 1 whereas the B sec-
tion uses mostly region 2. Region 2 occurs in A only once, in the first half
of the last phrase (m. 12), followed by region 1 in the second half of that
phrase (mm. 13–14). This splitting of one phrase into two different regions
then recurs in the final two phrases of the B section (mm. 22–23 and 24–26,
respectively) and again at the very end of the movement, in the reprise A‫׳‬
(mm. 27–29). In other words, the ends of sections A and B, and all of section
A‫ ׳‬are characterized by oscillations between regions.

As already noted, the beginning of the middle section B (ex. 12.6, mm.
16–17) presents a succession of major and minor triads, and later of augmented
triads, that cannot be obtained straightforwardly from the ordered twelve-tone
series. (Any form of the series always contains two augmented triads in one
ordered hexachord and one major and one minor triad in the other.) This
particular succession of triads can be derived via an interleaving of two row
forms, as we have shown in example 12.7. However, in an alternate reading,
the triads can also be obtained via an internal reordering of the hexachords
of the series. This reading is illustrated in example 12.10. Example 12.10a
shows the two hexachords of region 2 (2A and 2B) in normal order, now each
partitioned into a major and a minor triad. These are the triads with which
Leibowitz opens the middle section, in the order 1–4 as numbered (compare
with ex. 12.6, m. 16). This in turn is followed by a different partitioning of
the same hexachords into augmented triads exclusively, as shown in example
12.10b (compare with ex. 12.6, m. 17 with pickup). In short, Leibowitz intro-
duces new serial functions at a salient moment in the form of the movement,
to mark the beginning of the contrasting middle section, and he will use the
same strategy to mark the end of the last movement to be discussed later.
Example 12.9. RI-2, RI-9, and their discrete hexachords
Example 12.10a. Repartitioning of the hexachords into all major and minor triads,
mvt. 1, m. 16
Example 12.10b. Repartitioning of the hexachords into all augmented triads,

mvt. 1, mm. 16−17

Table 12.1. Overview of the form of the first movement
A
Measures 1 2–3 4 5–6 7–8 9 10–11 12 13 (w/ 15

pickup)–14
Series RI-2 P-1 R-7 I-4 R-7 I-0 R-3 R-2 RI-2 RI-6
Region 1 1 1 1 1 1 1 2 1 1
Voice-leading 4512 2154 5241 1425

partitions that (inversion (inversion
occur more than of that in of that in
once (# of pcs per mm. 2–3) mm. 10–11)
part, SATB)
Orchestration pno voices + pno voices + voices + pno pno voices + pno voices + pno voices + pno
pno pno pno
form functional prop. contraste prop. contraste prop. contraste prop. contraste
elements
antecedent consequent antecedent consequent
cadential/ cadential/
theme intermediary repeat of theme intermediary
structure structure
9/30/2015 7:54:15 PM
B A′
Measures 16 17 (w/ 19 (w/ 21 (w/ 22–23 24 (w/ 27–28 29
pickup)–18 pickup)–20 pickup) pickup)–26
Series Trichordal Trichordal R-2 I-1 R-8, P-3 P-7, I-3, R-6 I-7 R-e
partition of partition (into
aggregate (into 4 augmented
2 major and 2 triads),
minor triads), replaces series
replaces series
Region 2 2 2 2 2, 1 1, 2, 2 2 1
Partitions 2154 (same as in
(occurring mm. 5–6)
more than
once)
Orchestration pno voices + pno voices + voices + voices + voices + pno voices + pno pno
pno pno pno
9/30/2015 7:54:15 PM
Leibowitz’s Sketch for the

Opening of the First Movement
Figure 12.1 reproduces Leibowitz’s sketch for the opening of the work. The
lower half of the sketch (shown on the second page) contains Leibowitz’s row
table.40 The upper half shows how he first drafted the first vocal ritornello
(mm. 2–3 of the final version) before working out the opening piano gesture.
In the first two measures of the sketch, Leibowitz partitioned P-1 into lines
of 4, 5, 1, and 2 pitch classes as in the final version, realized in 43. Following
the double bar, he reworked the same material in cut time, diversifying the
rhythmic alignment of the voices and assigning the text. To the right of the
following double bar, Leibowitz notated the series just used (P-1) and one of
its inversions that exchanges the two hexachords (I-2). Underneath this, from
a retrograde reading of this inversion he drafted a first version for the piano
opening (sounding a succession of 2, 1, 2, 4, 3 attacks). Below and further
to the left (next to the underlined title of the poem), Leibowitz placed this
material, in cut time, presumably to precede the vocal ritornello, and then,
further below, slightly changed the alignment of attacks as in the final ver-
sion (producing a succession of 2, 2, 1, 4, 3 attacks). To the right, following
the double bar and in preparation for the working out of the second vocal
ritornello, Leibowitz paired P-1 with I-6. The opening pitch classes of these
two series (C♯ and F♯, respectively) form ic5, a relationship often found in
pairs of series in Schoenberg’s twelve-tone music. Using I-6, Leibowitz then
drafted the second vocal ritornello (“Le vent . . .”). In the final version, this
pitch structure is then transposed down a whole tone (see mm. 5–6 in ex.
12.1), with changed rhythms.
What may have motivated this transposition? We suggest that looking at
the succession of hexachords in the larger antecedent–consequent phrase
may provide the answer. This phrase, measures 1–6 in the final version, con-
tinuously alternates hexachords 1A and 1B, characterizing region 1, as labeled
in example 12.1 (a property that incidentally holds for the entire A section,
only briefly interrupted in m. 12 where region 2 occurs). In Leibowitz’s first
draft using I-6 (fig. 12.1), on the other hand, the succession of hexachords at
“Le vent” is 1B followed by 1A, which would break the larger pattern of con-
tinuously alternating hexachords 1A and 1B. We conjecture that maintaining
a pattern of alternating hexachords throughout the entire A section imbues
Leibowitz’s thematic material with a special consistency.
Second Movement: Double Canon

The second movement is a double canon in inversion almost all the way
through. Leibowitz writes about canons in the “Traité”: “Canons and fugatos

Figure 12.1. Leibowitz’s sketch for “Son de cloche” including his row table. Collection René Leibowitz, Paul Sacher Foundation.
(continued)
9/30/2015 7:54:15 PM
Figure 12.1.—(concluded)

do not have—inside a musical composition—as specific locations as the other

structures we have studied up to now. They can be found everywhere, so to
speak, at any moment in a piece—in the capacity of introduction, exposition,
transition, development, reprise, or coda—they can even constitute entire move-
ments.”41 This characterization suggests that phrase-structural laws do not apply
here in the same way as in other types of structures. In keeping with the canonic
texture, phrase boundaries in this movement are less clear than in the first: most
of the time, a phrase ending in one pair of voices overlaps with the onset of a
new phrase in the other voice pair, as for instance in measure 3 (see ex. 12.2
above). Nevertheless, the movement has a clear overall shape because, as we will
see, Leibowitz uses serial functions to make an arch form that involves various
dimensions, including partitioning of the series, hexachordal loosening, and
choice of axis of inversion. In addition, he distinguishes the end by some subtle
and some not-so-subtle means that make a striking climax.
The movement consists of six phrases, each of which distinguishes itself with
respect to serial function. This is summarized in example 12.11, which shows
the partitioning of the pitch material of the entire movement. The example
does not preserve the vertical alignment of the pitches as in Leibowitz’s score,
but simply notates the pitches of each series in consecutive order, showing how
they are distributed between voices. Each series constitutes a phrase. Most of
the time (with the exception of m. 8) Leibowitz superimposes two phrases,
one in each voice pair, in canon in inversion. The boundaries of phrases are
marked by the double bars, the division of each series into its two hexachords
by the dotted bars.
Most series are distributed between two voice parts, and the first and last
double canon statements are distinguished by being further broken up (e.g.,
the line starting in the soprano is continued in the tenor; see the top staff in ex.
12.11), creating pillars of a large arch. The middle (m. 8) is unique in that it
distributes only one series among four voice parts. In addition, the movement
starts out and ends with an element of asymmetry: I-9 and P-4 in measures 1–3,
and R-8 in measures 3–6 are each partitioned into lines of 7 and 5 pitches,
respectively (these cardinalities are shown in the margins). The remaining
pairs of lines, with the exception of measure 8 (unique, as noted), on the other
hand, partition the series into 6+6 pitches, until the very end, measures 11–13,
where the partition of 7+5 returns. This reminds us of Leibowitz’s analysis of an
ABA‫ ׳‬theme type in Schoenberg, the theme from the Variations for Orchestra,
op. 31, where he observes that the B section partitions the series (symmetri-
cally) into 6+6 after it had been partitioned (asymmetrically) into 5+4+3 in
the A section.42 As Leibowitz observes elsewhere, while the middle section of a
“lied” ABA‫ ׳‬form is often more complex than the outer sections, the opposite
can occur too.43 From a partitional point of view, this is in fact the case for the
third and fifth phrases of the second movement of opus 92, which are charac-
terized by 6+6 partitions.

Example 12.11. Serial counterpoint of the six phrases in the second movement

In the third through fifth phrases the realization of the series is symmetri-
cal in the sense that within each phrase the series are distributed the same
way between lines. In measures 6–7, the voice leading in the upper pair is
exactly mirrored in that of the lower pair;44 in measure 8 (with pickup) the two
ordered hexachords are divided the same way between soprano/bass and alto/
tenor, respectively; and in measures 9–11 the voice leading is again exactly mir-
rored between the two canonic pairs (save for the final fifth in the alto to be
discussed shortly). In the outer first, second, and sixth phrases, on the other
hand, the voice leading differs in certain places between the canonic pairs.
These places are marked by the dotted boxes in example 12.11.45 Overall, this
kind of variation in the realm of serial function produces its own distinct nar-
rative in the course of the movement: perfect symmetry of partitioning (both
in terms of segmentation of order positions and the resulting succession of
interval classes in the corresponding voices) only arrives in the third and fifth
phrases. However, Leibowitz introduces yet another novelty in the latter, by
reading the second hexachord of each series backward, the only time this hap-
pens in this movement, providing its own serial formal marker just before the
start of the last phrase.
Leibowitz uses distributional techniques to give the melodies varying
shapes, especially as concerns their beginnings and ends. Example 12.11
illustrates how he achieves this. As marked by brackets, straight lines, and
slurs, the particular distribution of the series between voices serves to fore-
ground certain melodic intervals that are not available between successive
members of the series. For instance, the ascending major third A–C♯ at the
beginning of the soprano line is not available between the first two notes of
an inversion (that interval would be a descending major third). Similarly, the
descending perfect fourth E–B at the end of this first phrase (in the tenor)
cannot be obtained from the last two pitch classes of any form of the series.
(Series start and end with minor and major thirds.) As the intervals marked
at the beginnings and ends of all phrases (except for m. 8) show, Leibowitz’s
distribution of the series between voices is guided by his preference for cer-
tain intervals in these locations. These are the major third, perfect fourth,
and minor second, ascending or descending. In other words, we never hear
melodic major seconds, minor thirds (themselves available at the beginning
or end of the series), and tritones in these places (except for the minor third
in the soprano of the exceptional m. 8). This intervallic consistency, achieved
through the particular partitioning, provides melodic unity beyond the suc-
cession of consecutive intervals in the series.
The prevalence of these three intervals takes a remarkable turn at the end
of the movement where, as highlighted by the wiggly lines, the major third and
perfect fourth are melodically inverted into their complementary intervals of
minor sixth and perfect fifth. This enlargement, subtly foreshadowed by the
descending perfect fifth f1–B♭ that concludes the fifth phrase in the alto (m.

11), creates a melodic explosion that delivers a strong formal marker, com-
bined with the sudden expansion of the soprano voice into the highest regis-
ter. (The highest pitch of the movement thus far, f♯2, is now superseded by g2
and the final b2.)
Finally, the end of the movement (A‫ )׳‬is distinguished from the begin-
ning (A) by yet another subtle feature in the serial structure. In both sections
Leibowitz superimposes series such that (near-) simultaneously sounding hexa-
chords are complementary (see alignment of hexachords 2A and 2B in table
12.2, in the row marked “Series and order of hexachords”; they are comple-
mentary via index numbers 1 and 5, as shown in the row marked “Index # of
inversion”). However, although the second phrases of A and A‫ ׳‬use the same
transpositions of rows, those in A‫ ׳‬are sounded backward; and although the
first phrases of A and A‫ ׳‬use the same index number (1), they use different
transpositions of rows. Again, a particular moment of the form (the end of the
movement) is marked by a new feature in the serial narrative. The movement
closes with a climactic gesture (dramatic expansion of melodic intervals and
upper pitch range) within a serial structure that is a variant of that from the
opening of the movement.
A Sketch for the Second Movement

Figure 12.2 reproduces an excerpt from a sketch that shows Leibowitz’s inven-
tory of the series used in the second movement (save for the last two).46 One of
Leibowitz’s indications is particularly telling here, namely, the mention of an
“inversion for final cadence” (in English!). It is possible that Leibowitz meant
the inversion of the row directly above, I-3; an inversion at the preferred inter-
val of a fifth or fourth would yield P-8 or P-t. In fact, P-8 is the row heard in the
soprano and alto in the last phrase, which with I-9 provides the “final cadence”
of the movement.
The Third Movement: Unité motivique

We have seen the beginning and end of the third movement in examples
12.3a–b. The piano opening is remarkable for its strange partitioning (see the
series marked in ex. 12.3a) and the amount of repeated pitch material, out of
order (the chords in the piano right hand and the Bs in the alto line), both
of which signal loosening. We recognize these features as too “open” for the
beginning of a piece, and therefore we can conclude that (1) the beginning
of the third movement is to be understood as a secondary structure to the pre-
ceding movements, and (2) that there is likely to be further loosening in the
course of this movement. As we will see, the movement all but self-destructs!

Table 12.2. Overview of the form of the second movement
A B Aʹ
Measures 0−3 4 (w/ pickup)-6 6-7 8 (w/ pickup) 9−11 12 (w/ pickup)−13
Tempo Poco mosso rit. Tranquillo poco rit. Meno mosso Ancora più lento
(𝅘𝅥𝅮 = 112) (𝅘𝅥 = 𝅘𝅥) (𝅘𝅥𝅮 = 84), rit. (𝅘𝅥𝅮 = 72), rit.
Series and I-9 (2B/2A) R-8 (2B/2A) RI-9 I-3 (2A/2B) P-6 (2B/2A) P-8 (2A/2B)
order of P-4 (2A/2B) RI-9 (2A/2B) (2A/2B) I-7 (2A/2B) I-9 (2B/2A)
hexachords R-8 (2B/2A)
Partition of 7+5 7+5 6+6 2+2+4+4 6+6 5+7

each series 5+7 6+6 6+6 6+6 7+5
above
Index # of 1 5 5 N/A 1 5
inversion
Region 2 2 2 2 2 2
intermediary
structure
(followed
by caesura
between mm.
8 and 9)
9/30/2015 7:54:17 PM
Figure 12.2. Sketch for the second movement. Collection René Leibowitz, Paul Sacher Foundation.
9/30/2015 7:54:17 PM
The liquidation is augmented straightaway with repeated chords in the piano

and an oscillating semitone motion in the vocal bass part in measures 8–10, not
shown here. The repetitions become so thick that confident identification of
row forms becomes difficult. By loosening up the serial functions in this way,
Leibowitz is able to acquire what he calls unité motivique (one of three kinds of
unity that apply in small forms, discussed in chapter 11 of the “Traité”).
Leibowitz’s analysis of Webern’s opus 25, no. 1 (Wie bin ich froh), invokes
motivic unity as follows:
The forms of this complex are not signs of a specific partitioning called upon
to become operative in the course of the movement. Thus the voice part is
most free in that respect, and it turns out to be absolutely impossible to sub-
divide it into segments corresponding to any serial partitioning whatever. It is
pretty much the same for the accompaniment model: only the triplets almost
always correspond to certain specific partitions of the row. These are the first
and the last three-note partitions as well as those that constitute notes 5, 6,
and 7 of the rows. Note, however, that these partitions constitute variants of
a single motive and it is only for that reason that they have been respected.
Furthermore, neither the two-note motive nor the four-note chord managed
to be thematized as to their intervals, given that they always arise from differ-
ent fragments of the rows. Only their motivic significance counts.47
In the third movement of Leibowitz’s opus 92, in measure 14 (ex. 12.12),

we find an example of motivic repetition. The model begins with the salient
twelfth from soprano F to tenor B♭. This model is repeated in measures 18–19,
from the soprano F♯ to the bass B♮. The two patterns beginning with the inter-
val of a perfect twelfth are striking, as they have the same contour and similar
rhythm. The patterns comprise four notes each. Both patterns are even spelled
with the same solfège syllables, fa–si–mi–ré in French, or alphabet letters in
English: F–B–E–D, inflected as F–B♭–E–D♯ in measures 14–16 and F♯–B–
E♭–D in measures 18–19.
This concept of “motivic unity” permits varied repetition within the rigor of
the strict application of the twelve-tone system. Such unity results from similar
contour and rhythmic features, and need not be derived from corresponding
members of a series; that is, they are not the product of “serial forms.” The
combined melodies of soprano and bass arise from order numbers 1, 3, 9, and
12 the first time and 7, 8, 9, 10 the second (see order numbers indicated on
the score). Once again, derived material comes to the fore, as it did in the
chromatic melodic segments of the first movement ritornelli and the linear
thirds at the end of phrases in the second movement.
The conclusion of the third movement repartitions the series within
a larger liquidation process that brings the work as a whole to a close. The
repartitioning uses the same strategy we have seen in the middle section of
the first movement, where the “loosening” of the serial material marked the

formal function of a secondary structure (see the earlier exx. 12.6, 12.7, and
12.11). At the end of the third movement, this process of “loosening” now
serves a concluding function. The particular partitioning is illustrated in exam-
ple 12.13. Here Leibowitz moves from region 1 to region 2 and back again to
region 1. In the last phrase (mm. 25–28, see ex. 12.3b above) the loosening is
taken yet one step further: Leibowitz switches between the hexachords from
which the simple triads are drawn, thus finally liquidating the identity of the
hexachords themselves. The first chord (B major, m. 25, third quarter) comes
from hexachord 1B, the second (A minor, m. 26) from hexachord 1A, and the
third and fourth chords (G minor and D-flat major in mm. 27–28, respectively)
again from hexachords 1B and 1A in this order (compare with ex. 12.13c).
Gesturally, the repeating of notes in homophonic rhythm in three voices (ex.
12.3b, m. 26) and in the soprano (mm. 27–28) slows down and “neutralizes”
the melodic activity, adding further to a sense of closure. The soprano note g1
is held over into the last measure where it adds an augmented fourth to the
D-flat-major triad, creating a modal flavor. (This chord is one of the signature
harmonies in Messiaen’s second mode of limited transposition.)
Example 12.12. Trois poèmes de Pierre Reverdy, mvt. 3, “Soleil,” mm. 14−21, with order
numbers in series RI-3 and P-9 shown on the score
(continued)

Example 12.13a. Repartitioning of the hexachords into all augmented triads, mvt.
3, mm. 22–24 (region 1)
Example 12.13b. Repartitioning of the hexachords into all augmented triads, mvt.
3, mm. 24–25 (region 2)
Example 12.13c. Repartitioning of the hexachords into all major and minor triads,
mvt. 3, mm. 25–28 (region 1)

Conclusion
One of the most valuable lessons we learn from Leibowitz’s writings is that
we should attend to details that might seem trivial or capricious. The salient
long notes in the ritornelli of the first movement belong to a partitioning
that is thematic, and whose repetition is a means of formal organization; the
immediate repetitions of two-note fragments (e.g., in mm. 7–8 of the first
movement) are carefully used to signal the liquidation that prepares new
formal areas; and “motivic unity,” which depends on contour and rhythm
more than on exact intervals (mm. 14–16 and 18–19 of the third movement),
is a new and flexible structural principle altogether, one that creates—in
Schoenberg’s term—coherence.48 Features like these give richness to each
movement and to the set as a whole.
It is possible to consider the three movements as a single “whole” that
gradually develops from something very tightly constructed to something
much looser. The melodic semitone is one element we can track in support
of this observation. The semitone is prevalent in the first movement: two of
the lines in the ritornello segment consist entirely of semitonal motions (the
two-note segment and the tetrachordal melody, ex. 12.4). These do not arise
from adjacencies in the row, but result from the particular distributions. In
the ritornello, Leibowitz realizes melodically all three semitones contained in
the unordered second hexachord (in each of exx. 12.4a–e three voices end
with a semitone). Likewise, many of the motions from one row to the next
contain semitonal movements (see, e.g., the link between R-7 and I-4, E♭ to
E and B to C, in ex. 12.1, mm. 4–5). This is because between the two hexa-
chords of the same region are six possible semitone connections. The semi-
tone occurs less prominently in the second movement but still plays a role.
As we noted, every phrase either begins or ends with a semitone motion in at
least one voice. In the final movement the semitone continues to lose ground
as a motivic ingredient until the remarkable final chords, when whole-tone
connections rise to the surface.
Form-functional elements evolve over the course of the three movements.
The rigor and clarity of the opening of the first movement give way to a slight
loosening already in the repetition of the theme, and then to further row-
cracking in the B section. The reprise of the A section is appropriately varied,
but is too short to support a full-scale structure, and thus demands a continu-
ation. The second movement, by being canonic, presents a different kind of
structure altogether that introduces many new melodic ideas. The third move-
ment is governed by the more flexible principle of “motivic unity.” It takes the
reordered rows from the beginning of the B section in the first movement and
makes the four major and minor triads the ultimate goal of the developmental
process in the third movement.

The writings and compositions of René Leibowitz contribute a rich chapter

to the history of music in the middle of the past century. Taken together, they
represent a line of serial thought that continues the aesthetic of the Second
Viennese School with its strong links to the music of the past. We hope that our
analysis of Leibowitz’s last published work will awaken further interest in this
remarkable musician and theorist.
Notes
We would like to thank the students in a 2011 McGill graduate seminar on Webern,
as well as our colleague Jonathan Wild, for their comments on an earlier draft of
this chapter. We are also grateful to Federico Andreoni and Corey Stevens for their
help in acquiring materials. Our special thanks go to Heidy Zimmermann at the
Paul Sacher Foundation.
1. The work was published in 1976 (Hillsdale, NY: Boelke-Bomart), in the series
“20th-Century Choral Music Series” (general editor: Jacques-Louis Monod). It was
premiered by the New Calliope Singers, directed by Peter Schubert, in 1977 in
Carnegie Recital Hall.
2. Choral movements that contain double canons similar to this one include Webern’s
Entflieht auf leichten Kähnen, op. 2, Schoenberg’s Satire, op. 28, no. 3 (e.g., starting
at the pickup to m. 59 and the pickup to m. 108), Brahms’s “Beherzigung,” op. 93a,
no. 6 (mm. 16ff.), and Mendelssohn’s “Die Nachtigall,” op. 59, no. 4.
3. René Leibowitz, “Traité de la composition avec douze sons.” The introduction of
the typed manuscript is dated February 1950. We are grateful to Will Ogdon, who
sent us a copy of this text in 2002. For a description of the treatise, see Will Ogdon,
“Concerning an Unpublished Treatise of René Leibowitz,” Journal of the Arnold
Schoenberg Institute 2 (1977): 34–41.
Music of Haydn, Mozart, and Beethoven (New York: Oxford University Press, 1998);
Arnold Schoenberg, Fundamentals of Musical Composition, ed. Gerald Strang with
the collaboration of Leonard Stein (London: Faber and Faber, 1967); Schoenberg,
Models for Beginners in Composition, ed. Leonard Stein (Pacific Palisades, CA: Belmont
Music, 1972); and Schoenberg, Structural Functions of Harmony, ed. Leonard Stein
(New York: Norton, 1969).
5. René Leibowitz, Schoenberg et son école: l’étape contemporaine du langage musical (Paris:
Janin, 1947). Schoenberg, Models for Beginners in Composition. See Heinrich Jalowetz,
“On the Spontaneity of Schoenberg’s Music,” Musical Quarterly 30, no. 4 (1944):
385–408; and Arnold Schoenberg, “La composition à douze sons,” trans. René
Leibowitz, in Polyphonie 4 (1949): 7–31. See also Schoenberg, “Composition with
Twelve Tones,” in Style and Idea, ed. Leonard Stein, trans. Leo Black (Berkeley:
University of California Press, 1975), 214–45.
6. See Therese Muxeneder, “‘I saw it in my imagination.’ Zur Textwerdung von
Arnold Schönbergs A Survivor from Warsaw,” in Arnold Schönberg in seinen Schriften:
Verzeichnis-Fragen-Editorisches, ed. Hartmut Krones (Vienna: Böhlau, 2011), 255.
7. See Reinhard Kapp, “Die Schatten des Urbilds des Doubles. Vorsichtige
Annäherung an eine Figur: René Leibowitz,” Musiktheorie 2 (1987): 16–17; and

Sabine Meine, Ein Zwölftöner in Paris: Studien zu Biographie und Wirkung von René
Leibowitz (1913–1972) (Augsburg: Wißner, 2000), 41–42.
8. A facsimile of this manuscript is reproduced in Regina Busch, Leopold Spinner
(Bonn: Boosey and Hawkes, 1987), 86–89.
9. Leibowitz’s comment on the Trio appears in Introduction à la musique de douze sons
(Paris: L’Arche, 1949), 319.
10. “Leibowitz’ athematische Musik: das geht vierzig Jahre zurück, wo ich, für
kurze Zeit das behauptet habe. Ich habe es aber bald widerrufen, da ja der
Zusammenhang in der Musik auf nichts anderem beruhen kann, als auf Motiven,
deren Verwandlungen und Entwicklungen. . . . Es ist nicht alles kein Gold, was
nicht glänzt, und es kann etwas thematisch sein, was bei Weitem nicht so aussieht.”
Letter from Rufer to Leibowitz of April 25, 1950, in the Collection René Leibowitz,
Paul Sacher Foundation. In this chapter all translations are ours.
11. “Es kommt mir nicht richtig vor, mehr als eine Reihe zu verwenden, aber
die Hauptsache ist doch, ob die Musik gut ist.” For a study of the correspon-
dence between Schoenberg, Leibowitz, and Rufer see Gianmario Borio,
“Zwölftontechnik und Formenlehre: Die Abhandlungen von René Leibowitz
und Josef Rufer,” in Arnold Schönberg (1874–1951): Autorschaft als historische
Konstruktion: Vorgänger, Zeitgenossen, Nachfolger und Interpreten, ed. Andreas
Meyer and Ullrich Scheideler (Stuttgart: Metzler, 2001), 287–321. For a dis-
cussion of Leibowitz’s form-functional analyses of Schoenberg’s and Webern’s
music in the “Traité” and an application of Leibowitz’s analytical method to
the Minuet from Schoenberg’s Suite for Piano, op. 25, and Piano Piece, op.
33b, see John MacKay, “Series, Form and Function: Comments on the Analytical
Legacy of René Leibowitz and Aspects of Tonal Form in the Twelve-Tone Music
of Schoenberg and Webern,” ex tempore 8 (1996): 92–131. See also John MacKay,
“On Tonality and Tonal Form in the Serial Music of Arnold Schoenberg,”
Canadian University Music Review 8 (1987): 62–77.
12. Comparing the sentential openings of the first movement of Webern’s opus 24
and of Beethoven’s Piano Sonata in F Minor, op. 2, no. 1, Leibowitz lays out the
five row forms of the former and says they are “metaphorically comparable to
the tonal schema tonic–dominant–dominant–tonic–dominant” (comparable méta-
phoriquement au schéma tonal: Tonique–Dominante–Dominante–Tonique–Dominante).
“Traité,” 13. Later he says: “As always the student should look for comparisons
with tonal music and in that way check for the structural and functional analo-
gies.” (Comme toujours l’élève devra chercher des comparaisons avec la musique tonale et
vérifier ainsi les analogies structurelles et fonctionnelles.) Ibid., 57. In our transcriptions
of portions of the treatise, minor misspellings have been corrected and accents
missing in the typescript have been added.
13. Leibowitz’s terms roughly correspond to statement, response, successive states of
fragmentation, and cadence (ibid., 13).
14. Ibid., 12, 14, 18–20.
15. “From the point of view of serial functions, the principal characteristic is the parti-
tioning of the two forms into three groups of four notes, each partition being used
to determine a complete section of the structure, melodically and harmonically.”
(Du point de vue des fonctions sérielles la caractéristique principale est le tronçonnement en
trois groupes de quatre sons des deux formes, chaque tronçon servant à déterminer, mélodique-
ment et harmoniquement, une section complète de la structure.) Ibid., 20.

16. We use the term “partitioning” here—as Leibowitz does (tronçonnement)—in the
most general sense, that is, with respect to segmenting a twelve-tone row in some
way. We will return to questions of terminology below.
17. Leibowitz, “Traité,” 17–18.
18. “Les structures closes de la musique de douze sons sont caractérisées par des
déroulements complets d’un ou de plusieurs formes de la série, ainsi que par des
tronçonnements sériels généralement simples dont le modèle est toujours main-
tenu avec constance.” Ibid., 10.
19. “1o. Toutes ces structures, à quelque type qu’elles appartiennent, quelque soit leur
degré de simplicité ou de complexité, se caractérisent dodécaphoniquement par
un maniement simple et rigoureux de la série. 2o. Ce maniement implique: a) un
tronçonnement spécifique (généralement symétrique) des formes sérielles, tron-
çonnement qui se trouve maintenu avec constance tout au long de la structure;
b) la détermination des divers segments ou sections par des tronçons complets ou
des déroulements complets de formes sérielles. 3o. Si plusieurs formes de la série
se trouvent utilisées, leur rapport transpositionel reste constant. Ainsi Schoenberg
se sert généralement de deux formes (original et renversement) réalisant le rap-
port de quinte. . . . Cette dernière constatation est de la plus haute importance car
le maintient rigoureux d’un rapport transpositionel fixe établit une harmonie spéci-
fique et caractéristique qui permet les répétitions et les reprises (c’est l’équivalent d’une
région tonale), alors que l’abandon d’un tel rapport (qui se fait jour, comme nous
le verrons, dans d’autres sections de l’œuvre) crée une harmonie différente, grâce à
laquelle il est possible de caractériser des structures nouvelles (transpositions, sec-
tions contrastantes, thèmes secondaires, codas, etc. . . .), structures qui rendent
possible l’articulation générale de l’œuvre musicale.” Ibid., 25–26.
21. Leibowitz, “Traité,” 16.
22. For present purposes, we make the following terminological distinctions: “Partition”
or “partitioning” in the general sense refers to the segmentation of a twelve-tone row
into smaller (complementary) sets. The cardinalities of the beamed sets in examples
12.4a–e arise from the same partition of twelve into four parts (that partition is 5421). A
twelve-tone series can be divided into imbedded segments of 5, 4, 2, and 1 elements
respectively in a large number of ways. We shall call each of these ways a specific dis-
tribution. From the 83,160 possible ways of selecting four (complementary) segments
of 5, 4, 2, and 1 elements, respectively, from a twelve-tone row, Leibowitz has chosen
two, distribution 1 (exx. 12.4a–c) and distribution 2 (exx. 12.4d–e) for his realizations.
Applied to the specific series of opus 92, the two distributions preserve (under TnI)
three of the four subsets as noted. A distribution corresponds to what Daniel Starr
calls a partition of a series into imbedded segments. See Daniel Starr, “Derivation and
Polyphony,” Perspectives of New Music 23 (1984): 214. For a table of the 77 partitions
of twelve into twelve or fewer parts see John Riordan, An Introduction to Combinatorial
Analysis (New York: Wiley, 1958), cited in Andrew Mead, An Introduction to the Music of
Milton Babbitt (Princeton, NJ: Princeton University Press, 1994), 32.
23. Leibowitz, “Traité,” 37.
24. Gianmario Borio discusses this property in “Zwölftontechnik und Formenlehre,”
301.
25. We borrow the term “polyphonization” from Starr, “Derivation and Polyphony,”
213. For another discussion of chords freely obtained in a rigorously serial context,

see Peter Schubert, “A Multidetermined Moment in Milton Babbitt’s ‘Three

Cultivated Choruses,’” Theory and Practice 19 (1994): 57–82.
26. Leibowitz’s discussion of repeated notes includes the following comments. Of
Schoenberg’s Prelude, op. 44, he observes that the cadential segment “condenses
the harmony into a single chord made of all the notes of the second hexachord
inverted while the melody unfolds the six notes of the second hexachord of the
original (bassoon m. 11), the last four notes being repeated by the harp” (condense
l’harmonie en un seul accord fait de tous les sons du deuxième tronçon de la forme renver-
sée, alors que la mélodie déroule les six sons du deuxième tronçon de l’original (basson ms.
11), les quatre derniers sons se trouvant répétés par la harpe). “Traité,” 15. Of the open-
ing of Schoenberg’s Fantasy, op. 47, he notes “the very strong caesura (mm. 3–4)
obtained thanks to the repetition of the A♭ and G♭ in the melody and the triplet
figure in the accompaniment” (la très forte césure [ms. 3–4] obtenue grâce à la répétition
des sons la bémol et sol bémol dans la mélodie et de la figure en triolets de l’accompagnement).
Ibid., 18.
27. Leibowitz’s discussion of cadential function includes the following observation. In
Webern’s opus 24, mm. 9–10, the cadential segment consists of a complete row
form; what is cadential, he says, is that “Webern condenses his harmony by telescop-
ing the notes so as to make chords.” (Webern condense ici son harmonie en télescopant les
sons de manière à former des accords.) Ibid., 13. Chords have not been heard until that
moment.
28. “1o) la nécessité pour le compositeur de se débarrasser de ses obligations vis-à-vis
de la structure close qu’il vient d’exposer; 2o) la nécessité d’ouvrir le discours musi-
cal afin de pouvoir ‘passer à autre chose’; 3o) l’introduction d’une nouvelle idée
(de caractère transitoire) permettant l’introduction pure et simple d’une structure
contrastante (partie centrale du mouvement, deuxième thème, etc.).” Ibid., 27.
29. “Il va de soi que la répétition doit être variée afin de ne pas, encore une fois, fermer
la structure.” Ibid., 28.
30. “l’on entreprend immédiatement une neutralisation du discours musical qui finit
souvent par ne plus présenter qu’un aspect (neutre) de valeurs égales.” Ibid., 28.
31. “La troisième étape prend généralement son point de départ dans les valeurs égales
que propose la neutralisation, valeurs qui permettent souvent la déduction d’un
modèle de transition. Cette idée nouvelle (qui ne doit plus présenter grand’chose
[sic] de commun avec le thème) se trouve le plus souvent répétée plusieurs fois de
suite jusqu’à saturation complète. Elle mène—parfois presqu’imperceptiblement,
parfois de manière brusque—à la structure contrastante.” Ibid., 28.
32. “Les caractéristiques propres des diverses structures intermédiaires (de liquidation,
de neutralisation, de dissolution et de transition) ont leurs caractéristiques propres
sur le plan des fonctions sérielles. Ici les points les plus saillants sont: 1o Utilisation
de formes sérielles et de rapports transpositionels nouveaux (de même que dans
la musique tonale de nouvelles régions se trouvent utilisées; 2o tronçonnements
nouveaux (qui correspondent à des changements harmoniques dans la musique
tonale); 3o traitement plus libre et surtout moins ‘complet’ des formes sérielles
afin que les structures restent ouvertes (cela correspond à l’absence de symétrie
du parcours tonal et à l’absence des cadences trop affirmées dans les sections cor-
respondantes de la musique tonale); 4o utilisation (généralement) d’un plus grand
nombre de formes sérielles (ce qui correspond à des parcours à travers un plus
grand nombre de régions tonales). En résumé, ces fonctions s’avèrent moins sta-
bles et moins symétriques que ne sont celles des structures closes.” Ibid., 34.

33. Caplin, Classical Form, 255. Leibowitz uses the word délié (loose) several times (e.g.,
in connection with Lied form, where the A section is compact and the B section is
loose). “Traité,” 20.
34. In his analysis of Webern’s opus 27, mvt. 3, Leibowitz finds that the cadence
“reaches major proportions” (“Traité,” 16). The opening of this movement is
aligned with that of the first movement from Haydn’s Piano Sonata in E-flat Major,
Hob. XVI:49, to show that the cadence in Webern corresponds to an expanded
cadential progression, in Caplin’s terms.
35. Compare Leibowitz’s analysis of Webern’s opus 27, mvt. 3, where, in the conse-
quent, “la symétrie est garantie par le tronçonnement identique dans les deux cas,
alors que la variation résulte du renversement sériel et de la registration différente
dans les deux cas” (symmetry is provided by identical partitioning in both cases,
whereas variation arises from serial inversion and the different use of register in
both cases.” “Traité,” 16). See also p. 28 on the repetition of the theme in Webern
opus 24, as well as p. 53 where he says: “Les qualités essentielles des reprises doivent
être: cohérence et nouveauté” (the essential qualities of repetition must be coherence
and novelty—his emphasis).
36. Both intermediate and secondary structures are based on “fonctions sérielles moins
symétriques et moins simples que celles des structures closes” (less symmetrical and
less simple serial functions than those in closed structures). Leibowitz, “Traité,”
35).
37. Ibid., 28.
38. F precedes D♭ in the left hand of measure 17. See example 12.6.
39. A region, as defined for the specific context here, corresponds to the combina-
tion of two complementary “hexatonic systems” in the terminology of Richard
Cohn. Region 1 combines what he calls the Western/Eastern systems, region 2
the Northern/Southern systems. See Richard Cohn, “Maximally Smooth Cycles,
Hexatonic Systems, and the Analysis of Late-Romantic Triadic Progressions,” Music
Analysis 15 (1996): 17–18. David Lewin, in his analysis of “hexachord transposi-
tions” in Schoenberg’s Fantasy, op. 47, called a particular transposition of one of
the (unordered) hexachords from the series and its complement an “area.” See
David Lewin, “A Study of Hexachord Levels in Schoenberg’s Violin Fantasy,” in
Perspectives on Schoenberg and Stravinsky, ed. Benjamin Boretz and Edward T. Cone
(Princeton, NJ: Princeton University Press, 1968), 78–92. Schoenberg’s series is
built from complementary hexachords of type [023468], which can be transposed
to yield a total of twelve different areas. Leibowitz’s series is built from complemen-
tary hexachords of all-combinatorial type [014589], which can be transposed to
yield only two different areas, called here regions 1 and 2.
40. Leibowitz indicates transpositional levels with Roman numerals (for prime forms,
I = starting on C♯, II = starting on D, etc.) and Arabic numbers (for inversions, 1 =
starting on C♯, 2 = starting on D, etc.). The arrows above the staves mark most of
the rows he used in the first movement.
41. “Canons et fugatos n’ont pas—au sein d’une composition musicale—des ‘emplace-
ments’ aussi spécifiques que ceux des autres structures que nous avons étudiées
jusqu’à présent. Ils peuvent se rencontrer partout, pour ainsi dire, à n’importe quel
moment d’une œuvre—en tant qu’introduction, exposition, transition, développe-
ment, reprise ou coda—ils peuvent même constituer des mouvements entiers.”
Leibowitz, “Traité,” 82.
42. Ibid., 23.

43. Ibid., 21–22.

44. In terms of register, there is one deviation from an exact mirroring, in the d1 of the
soprano and E♭ of the bass.
45. In the sixth phrase the D♭ in the lowest line does not conform to the series. It
should be D♮.
46. The sketch begins with a draft for measure 11 of the first movement. To the right of
the first double bar, in the top two staves that are braced together, we find the pairs
of series used in measures 1–3. To the right of those are the pairs of series for mea-
sures 3–6 and 6–7. On the third staff, following the first double bar, is the row used
in measure 8 (with pickup). To the right of this, after the crossed-out fragment, we
find the pair of rows used in measures 9–11. The series written at the bottom of the
sketch are used in the third movement.
47. “Les formes de ce complexe ne témoignent pas d’un tronçonnement spécifique
appelé à se fonctionnaliser au cours du morceau. C’est ainsi que la partie de chant
est des plus libres à cet égard et il s’avère absolument impossible de la subdiviser
en des segments correspondants à un tronçonnement sériel quel qu’il soit. Il en
va à peu près de même pour le modèle d’accompagnement dont seuls les triolets
correspondent presque toujours à certains tronçons spécifiques de la série. Ce sont
les premiers et les derniers tronçons de trois sons ainsi que ceux qui constituent les
sons 5, 6, 7 de deux formes sérielles. Il est à noter, cependant, que ces tronçons con-
stituent des variantes d’un seul motif et ce n’est que pour cela qu’ils on été respectés.
Par ailleurs, ni le motif de deux sons, ni l’accord de quatre sons n’arrivent à se thé-
matiser du point de vue des intervalles, étant donné qu’ils sont issus de fragments
toujours différents des formes sérielles. Leur signification motivique est la seule qui
compte.” Leibowitz, “Traité,” 91.
48. Leibowitz mentions coherence in his brief summary of Schoenberg’s manuscript-
in-process, which he calls “Der musikalische Gedanke und seine Darstellung
(Lehre vom Zusammenhang der musikalischen Gedanken)” in the introduction
of “Traité” (4): “De même les structures diverses d’un morceau de musique doi-
vent être coordonées de telle sorte que l’ensemble du morceau constitue un tout
cohérent” (Likewise, the different structures of a piece of music have to be coor-
dinated such that the entire piece constitutes a coherent whole”—underline in the
original). It is unclear which of the several manuscripts of Schoenberg Leibowitz
was referring to. Schoenberg had been working on a number of manuscripts on
the musical idea. See Arnold Schoenberg, The Musical Idea and the Logic, Technique,
and Art of Its Presentation, ed., trans., and with a commentary by Patricia Carpenter
and Severine Neff (Bloomington: Indiana University Press, 2006), xix–xx.

Chapter Thirteen
The Philosopher as Theorist

Adorno’s materiale Formenlehre
Steven Vande Moortele
Practitioners of the theory of formal functions have never been shy to acknowl-
edge the origins of their enterprise in Schoenbergian Formenlehre of the mid-
twentieth century. Right at the beginning of the preface to Classical Form,
William E. Caplin emphasizes his debt to Schoenberg and Erwin Ratz—as well
as to Carl Dahlhaus—and reference to Ratz in particular is made in almost
every one of the book’s chapters.1 The influence of Schoenbergian thought
on form (which also includes writings by Josef Rufer, Erwin Stein, and Anton
Webern) manifests itself in many of the theory’s central concepts and termi-
nology, and even the very notion of “formal function” draws directly upon
Ratz’s funktionelle Formenlehre.2 Not coincidentally, Caplin’s earliest publication
on classical form appeared in German.3
Like Caplin, Janet Schmalfeldt—his principal interlocutor in the early
development of the theory of formal functions in the 1980s—repeatedly
invokes Schoenberg, Ratz, and Dahlhaus in her writings on musical form.4
One further author she mentions, but about whom Caplin remains silent,
is Theodor W. Adorno. Schmalfeldt first brings up Adorno in her 1995
article on Beethoven’s “Tempest” Sonata.5 There, he figures as part of the
philosophical strand of what she calls the “Beethoven-Hegelian” tradition—a
strand that is paralleled by a music-theoretical one leading from A. B. Marx
to Dahlhaus. Yet it is not difficult to see how Adorno, as both a philosopher
and a writer on music (not to mention a composition student of Alban
Berg in the 1920s), really stands at the crossroads between these two lines.
Schmalfeldt does not explore this matter in her 1995 article, but she does in
her later book, In the Process of Becoming. She writes:
Scholars of both Adorno and Dahlhaus have tended either to neglect or to

disparage the capacities of these two for genuine music-analytical insight—a

412 steven vande moortele
position to which I take strong exception; in particular, Adorno’s unfinished

Beethoven fragments reveal a philosopher-musician grappling over the course
of his maturity with highly sophisticated aspects of Beethoven’s music. But
[my] study employs analytic techniques developed beyond the lifetimes, and
maybe even the interests, of both Adorno and Dahlhaus; and so the presence
of both recedes as the book proceeds.6
While Adorno’s intellectual influence thus permeates Schmalfeldt’s book in

a broad sense, she may be underestimating the extent to which her own ana-
lytical work—and the theory of formal functions in general—resonates with
some of Adorno’s more technical writings on music. During the last decade of
his life, Adorno indeed repeatedly wrote about what he called “eine materiale
Formenlehre.” Although potentially his most important contribution to the
discipline of music theory, this “material theory of form” remains fragmentary
and was never developed into an analytical system. Its reception by the music-
theoretical community has been minimal; Schmalfeldt briefly mentions it, but
does not elaborate.7
My purpose in this chapter is to reconstruct and contextualize Adorno’s
theory of musical form, and in so doing to clarify some of the prehistory of
current-day Formenlehre, especially in its Caplinian and Schmalfeldtian guises.
In what follows I proceed in three steps. I begin by investigating some of the
reasons for the neglect of the theoretical side of Adorno’s thinking; this serves
simultaneously to refamiliarize readers with some of the basics about Adorno
and his work. Next I reconstruct the materiale Formenlehre from Adorno’s writ-
ings and situate it in the context of his philosophy and that of twentieth-century
Formenlehre. Finally, I illustrate Adorno’s theory by means of his fragmentary
analysis of the first movement of Beethoven’s Eroica. For most of the chapter, I
trust that correspondences between Adorno and modern form-functional the-
ory are obvious enough not to require additional emphasis. Only in the con-
cluding paragraphs do I explicitly draw parallels between Adorno’s and Caplin
and Schmalfeldt’s work on musical form.
Musicology’s Adorno, Music Theory’s Adorno

Several interrelated factors have impeded the reception of the theoretical side of
Adorno’s work in North America. At the core of these stands the general context
in which American musicology was first exposed to Adorno. When, in the mid-
to late 1970s and early 1980s, Rose Rosengard Subotnik “discovered” Adorno for
American musicology, she did so as part of a critique of the discipline’s empir-
icist bias. This critique contains the seeds of what would later become known
as the “new” (or “cultural”) musicology: “American musicology,” in Subotnik’s
diagnosis, “is excellent at pursuing those kinds of truth that lend themselves to
the scientific model of verification. . . . It is far less good at grappling with those

the philosopher as theorist 413
humanistic sorts of truth that . . . evoke large numbers of divergent interpreta-

tions, no one of which can consider itself privileged or complete.”8
In this context, Adorno is presented as an alternative. In contrast to the
positivist ideology espoused by (“old”) American musicology, Adorno’s writ-
ings on music postulate a direct link between music and society and thus prom-
ise access to music’s cultural meaning (Subotnik’s “humanistic sorts of truth”).
Highlighting the dialectical nature of the relationship between musical auton-
omy and cultural meaning, Subotnik summarizes the core of Adorno’s aesthet-
ics in three points: “Western art has tended toward increasing autonomy from
society; . . . the more autonomous the work of art is, the more deeply it embod-
ies the most profound social tendencies of its time; and . . . proper analysis can
decipher the social meaning of artistic structure so as to criticize art and soci-
ety simultaneously.”9
Readers may be surprised at the emphasis this summary puts on the “proper
analysis . . . of artistic structure.” For when cultural musicologists—Subotnik
herself as well as others in her wake—apply ideas inspired by Adorno to the
cultural interpretation of music, such analysis quickly recedes into the back-
ground. To be sure, the writings of many of these authors include analysis or
at least offer a detailed description of the music that is being discussed. Yet
the emphasis invariably rests on the mediation between music and society.10
American musicology in the 1980s drew on Adorno in order to open up new
avenues of scholarly inquiry into music, not to remind itself that close analysis
of musical structure remains a sine qua non for those who want to walk those
new paths with some authority.
In this way, new musicology has performed a remarkable swerve from the
attitude Adorno himself propagated. In the largely extemporized lecture
“Zum Problem der musikalischen Analyse,” given less than six months before
his death and published posthumously in 1982, Adorno insists that analysis
is an essential component of the cultural interpretation of music. “Works,”
he writes, “need analysis for their ‘truth content’ to be revealed,” and criti-
cism that is not founded in analysis “remains stuck with disconnected impres-
sions, and thus . . . deserves to be regarded with utmost suspicion.”11 It hardly
seems a stretch to claim that for Adorno, the dialectics of music’s autonomy
and its social meaning—the more autonomous the former, the more clearly
it expresses the latter—carries over into analysis. The more deeply and even
fanatically one immerses oneself into a work’s technical structure, the more
closely one approaches its truth content; the shortest way to music’s cultural
meaning is the one that seems to lead farthest from it. This, I believe, is how
one can understand enigmatic statements such as “analysis is more than merely
‘the facts,’ but it is so only and solely by virtue of going beyond the simple facts
by absorbing itself into them.”12
New musicology, in short, has one-sidedly adopted Adorno’s dialectical pair-
ing of autonomy and social meaning but largely ignored that of analysis and

social meaning. This is not a critique of new musicology, nor is it a defense

of Adorno against new musicology. As I will suggest near the end of this sec-
tion, the paradoxical claim that the shortest route to music’s meaning leads
through a single-minded focus on music’s internal structure may very well be
unfounded.13 But it should be clear that the new musicological project, in spite
of its appropriation of Adorno as one of its heroes, is more alien to Adorno’s
own concerns than has generally been acknowledged.
New musicology’s idiosyncratic take on Adorno is relevant to my concerns
because it has had a direct impact on music theory in two very different ways.
On the one hand, the association of Adorno with new musicology has likely
deterred more orthodox music theorists from engaging with Adorno’s work
altogether. On the other hand, and more importantly, theoretical work that
does refer to Adorno seems to have essentially copied the moves of new musi-
cology. In these writings, Adorno’s ideas are used as a way to transcend mere
analysis and go beyond the technical explanation of music by connecting it to
a cultural or philosophical interpretation. Typically, Adorno is brought into
play only after the analysis has taken place; the analytical methodology itself
remains largely unaffected by his work.14
An important contributing factor is the problematic position of analysis in
Adorno’s own writings. As Ludwig Holtmeier has observed, there is a “remark-
able disconnect between the status Adorno grants music analysis and the way
in which he himself relates to it in his own work.”15 This is not the place to
mount a full critique of Adorno’s analytical practice, especially since oth-
ers have done so already.16 The problems can be summarized in three main
points: (1) Adorno’s published writings offer surprisingly little in the way of
technical analysis: what analysis there is generally remains fragmentary, and
sustained analyses of entire pieces or portions thereof are exceedingly rare; (2)
the methodological basis of Adorno’s analyses is unsound; and (3) the results
of his analyses often appear unconvincing, biased, or downright faulty.
The mixed quality of Adorno’s published analyses poses a double-layered
problem. The immediate difficulty is obvious: there is no “Adornian” analyt-
ical technique, method, or procedure that one can copy. As Julian Johnson
has observed, “if anyone has been holding his or her breath . . . waiting for
someone to reveal that, somehow hidden in the Adorno oeuvre, there is a
blueprint for a new method of analysis . . . now is the time to breathe out and
accept the bad news: it isn’t there.”17 The underlying problem is less obvious,
but even more fundamental, and it becomes especially pressing if one consid-
ers Adorno’s cultural interpretations of music to be valid (valid, that is, even
though themselves culturally contingent). If the analysis is mediocre, but the
cultural conclusions are nonetheless sound, what does that say about the use-
fulness or even the necessity of analysis? Indeed, it may seem that Adorno’s
own practice, in which successful cultural interpretations of music are pos-
sible regardless of the accuracy of the analysis, can be used to disprove the

position he himself defends in theory: that the quality of cultural interpreta-

tion is dependent on the analysis that leads to it. I hasten to step back from
this disciplinary abyss, but there is a price to be paid. If, in the remainder of
this chapter, I stay far from any cultural analysis and treat Adorno’s materiale
Formenlehre as what it purports to be, namely, a theory of music, I present an
image of Adorno that is no less one-sided than the one that new musicology
has constructed.
Adorno the Philosopher

It would be an exaggeration to claim that Adorno’s theory of music works in
the same way as his philosophy. Nonetheless, some familiarity with the latter is
helpful to the understanding of the former, for Adorno the theorist operates
with the same dialectical mode of thought that is so characteristic of Adorno
the philosopher. In spite of the term’s centrality to Adorno’s oeuvre, it is
extremely difficult to provide a nontrivial definition of what, exactly, his ver-
sion of dialectics is. Adorno himself always remained reluctant to define the
concept. Even in the Einführung in die Dialektik—the posthumous transcriptions
of a course he taught at Frankfurt University in the summer term of 1958—
he approaches the term only circumferentially. A “dialectical operation,” he
explains in one of the first lectures, consists in “confronting the concept with
what it signifies until certain difficulties arise between that concept and what
it refers to, [difficulties] that make it necessary to modify the concept in one
way or another as the thinking moves forward.”18 What is crucial here is that
concepts are not static but are set in motion; it is no coincidence that Adorno
in this context cites Hegel’s famous dictum of the Bewegung der Begriffe (move-
ment of concepts).19 In addition, Adorno claims that dialectics is not merely a
“method of thinking,” but also an attempt “to do justice to an essential charac-
teristic of the object under consideration itself.”20 Dialectics is, in other words,
not only a way of conceptualizing reality but also a conceptual reflection of a
property of reality itself.
Both essential characteristics of dialectics—its dynamic nature and its
concreteness—inform Adorno’s materiale Formenlehre at a broad, general
level. More specifically, the theory’s firm embeddedness in his philosophy
(and especially in the philosophy of art laid out in the posthumous Ästhetische
Theorie) crystalizes in the use of three interrelated conceptual pairs: “form”
and “material,” “universal” and “particular,” and “whole” and “part.”21 These
concepts are as central to the materiale Formenlehre as they are to Adorno’s
aesthetics in general.
At the heart of the materiale Formenlehre stands a paradox that is expressed
by the term itself.22 Traditionally, and at least since Aristotle, “material” and
“form” have been conceived as opposites, much in the way that “form” and

“content” are. At first sight, Adorno adheres to this traditional view: material
is that which is “formed”; form is the “forming” of material.23 However, the
seemingly fixed categories of form and material start to blur as soon as one
considers them more closely. For one thing, material does not conceptually
precede form: it is “always already” (pre-)formed. Material, in other words,
should not be confused with raw material (Stoff); it is essentially cultural, not
natural. Second, the binary opposition material/form does not coincide with
the binary opposition form/content. The content of form is not the material
itself, but that which is happening to the material: the processes the material
undergoes. For Adorno, the content of form is ein Werden, a becoming. If, how-
ever, the content of form is a process (rather than the material that is being
subjected to that process), the very distinction between form and content
seems to evaporate; form and content (as the process of forming the material)
become indistinguishable: form is not only the result of that process; it is also
that process itself.
An additional complication—although one that is easier to disentangle—is
that form itself can become material. In order to understand this, we need to
bring into play a second conceptual pair: that of the “universal” and the “par-
ticular.” Form in art exists in two different ways. On the one hand, it exists as
universal (or abstract) form, similar to an “ideal type” in Max Weber’s sense.
This is the level of “textbook” forms—in music, “sonata form,” “the sentence,”
and so forth. On the other hand, form exists as particular form: the individual
form of a specific work (in music, say, the first movement of Beethoven’s Piano
Sonata in F Minor, op. 2, no. 1, or its main theme). These two kinds of form
are, of course, interdependent: forms as universals are schemata, and forms as
particulars are exemplars of those schemata. Moreover, general form has no
existence in reality outside of the particular forms. But at the same time—and
this is where a dialectical tension arises—no particular form entirely coincides
with its universal form; no work is ever one with its type. Rather, works use their
type as a background against which they assert their individuality. In this sense,
particular forms use general forms as their material.
Similar in structure to the dialectic of the general and the particular is that
of the “whole” and the “part.” In any engagement with reality, the whole is not
a given, but comes into being only through its parts; it is essentially the result
of a process. At the same time, the parts can be understood only in light of the
whole, without which they remain meaningless. For Adorno, the relationship
between part and whole virtually exemplifies dialectics:
The difficulty . . . is that although Whole and Part can only be understood
in relation to one another, the Whole is not given in any positive way when
you have a Part and the other way round, when you think the Whole, its Parts
are not given in any positive way. . . . In dialectics, the relationship between
Whole and Part is not one of mere subsumption, . . . but it is a dynamic

relationship: both moments produce each other reciprocally and are not
there . . . at the same time.24
This perspective is particularly relevant for musical form: since music

unfolds over time, it is impossible to stand back and see the whole: at any given
moment, that whole is only present in the form of its parts. At the same time,
the individual parts point beyond themselves toward the whole: in a musical
form, a given formal unit is what it is exactly because of the form (the whole)
that gradually emerges.
Adorno the Theorist

The materiale Formenlehre: First Principles
Adorno discusses his material theory of form in several writings and public
addresses on music from the 1960s. The term is first introduced in the chap-
ter “Charaktere” in Mahler: Eine musikalische Physiognomik (1960), which also
includes the most substantial discussion of the theory. The same ideas reappear
in varying constellations in Adorno’s speech for the centenary of Mahler’s birth
(the “Wiener Gedenkrede” of 1960), the Darmstadt lecture “Vers une musique
informelle” (1961), the Einleitung in die Musiksoziologie (1962), the chapter
“Analyse und Berg” in Berg: Der Meister des kleinsten Übergangs (1968), and the
1969 lecture “Zum Problem der musikalischen Analyse” (cited above).25 In
addition, several analytical notes that Adorno prepared in the 1950s for the
book on Beethoven that he never wrote are informed by the same ideas with-
out, however, using the phrase materiale Formenlehre explicitly.26 The connection
between the theory and its philosophical infrastructure, finally, is provided
mainly in the Ästhetische Theorie.27 In all of these discussions, one is struck by
the ease with which Adorno moves back and forth within a century and a half
of German music written between the French Revolution and World War II—
music, say, from Beethoven to Schoenberg. In my reconstruction of Adorno’s
theory, I will follow a similar path, starting with Mahler, then moving forward
to Schoenberg and Berg and back to Beethoven.
The material theory of form is inextricably linked to Adorno’s interpre-
tation of Mahler. In his Mahler monograph, it is introduced in the context
of the new analytical categories that he develops for that composer’s music:
Suspension, Durchbruch (breakthrough) and its opposite Weltlauf (course of the
world), and Erfüllung (fulfillment) along with its opposites Zusammenbruch (col-
lapse) or Katastrophe. Many of these categories, of course, have since become
common currency in Mahler scholarship. Adorno coins them to come to terms
with certain recurring “characters” (hence the chapter’s title) in Mahler’s
music that are both formal procedures and vehicles for extramusical meaning.

At the same time, these new analytical categories are what triggers the material
theory of form. It is worth citing Adorno at length:
Mahlerian categories like suspension or fulfillment suggest an idea that could

contribute, beyond the scope of his work, to endowing music with speech
through theory: the idea of a material theory of form, that is, the deduc-
tion of formal categories from their meaning. This is neglected by academic
Formenlehre, which operates with abstract classifications such as main theme,
transition, subordinate theme, and closing theme without understanding
these units in terms of their function. In Mahler the usual abstract formal
categories are overlaid with material ones; sometimes the former become
specifically the bearer of meaning; sometimes material formal principles are
constituted beside or below the abstract ones, which, while continuing to
provide the framework and to support the unity, no longer themselves supply
musical coherence.28
In interpreting this paragraph, it is important to understand the relation-

ship between new and traditional categories of form. Adorno begins by dis-
tinguishing between his “Mahlerian” formal categories, which are the primary
subject of his material theory of form, and traditional Formenlehre categories
such as main theme, transition, subordinate theme, and closing group. But he
immediately qualifies this distinction, explaining that in Mahler, both sets of
categories overlap. In spite of their novelty, therefore, Adorno’s categories are
not unrelated to more traditional ones. On the contrary: as he specifies a few
pages later, Mahler “drew his functional characters . . . from the stock of tradi-
tional music.”29 This connection between strategies that are typical of Mahler
and more general formal procedures is crucial, because it allows Adorno to
state that the relevance of his theory goes beyond Mahler’s oeuvre. The per-
spective that Adorno initially develops in relation to Mahler may, in other
words, be applied to the analysis of other—formally less extravagant—music.
The categories he mentions in relation to the material theory of form in “Zum
Problem der musikalischen Analyse,” for instance, are not the novel categories
from the Mahler volume, but general formal functions such as “statement,”
“continuation,” “contrast,” “dissolution,” “succession,” “elaboration,” “return,”
and “modified return.”30
Adorno illustrates the interconnectedness of old and new for several of
his Mahlerian formal categories. The clearest example is his discussion of
Zusammenbruch. Taking his cue from Ratz’s analysis of the end of the develop-
ment in the first movement of Mahler’s Ninth Symphony, Adorno describes this
category as the situation in which “the music collapses as in a kaleidoscope.”31
Its origins, he claims, lie in the “fields of disintegration” (Auflösungsfelder) that
sometimes occur in closing groups of classical sonata forms and in which “the
thematic contours dissolve in a play of tones, for example on the dominant,
which operates more or less like a formula.”32 In a similar manner, Adorno

establishes a connection between his category of Erfüllung—“the fulfillment

of a musical context by something fundamentally new”—and the “Abgesang of
bar form.”33 He provides two examples: the end of the exposition in the first
movement of Mahler’s Third Symphony and the end of the recapitulation of
the Finale of the Sixth.34 Even though for Adorno’s remaining three categories
(Suspension, Weltlauf, and Durchbruch) the connection to established formal cate-
gories is more tenuous—and in one case even nonexistent—the important point
is that Adorno never tires of emphasizing their ancestry in, and their fundamen-
tal compatibility with, the formal vocabulary of classical and romantic music.35
What Mahlerian and traditional categories of form have in common,
according to Adorno, is that both can be understood as functions, that is to
say, in terms of the work that individual formal units do in relation to the form
as a whole.36 The difference between both categories of form, then, is a differ-
ence in degree rather than in kind: Mahler’s music expresses its formal func-
tions more clearly than most other music does. Formal units in Mahler virtually
coincide with their formal functions: they are what they do.37 Adorno goes so
far as to accord passages in Mahler’s music a metaphorical voice that enables
them to say what they are or what they do: “Elaborations seem to say: this is an
elaboration”; and “The Marcato . . . announces: ‘Here I am.’” Mahler’s music,
Adorno writes, “calls its forms, as it were, by their names.”38
Adorno and Schoenbergian Formenlehre

Several aspects of Adorno’s discussion of his material theory of form help to
define its position in the broader landscape of twentieth-century Formenlehre
traditions. First, there is the emphasis on formal function and on form-
functional clarity. Significantly, Adorno’s main objection to the “academic”
Formenlehre of his day is not that it uses traditional terms such as “main theme,”
“transition,” “subordinate theme,” “closing group,” and so on—terms, after all,
that Adorno also uses himself—but that it employs these labels without grasp-
ing them as functions. Needless to say, this focus on formal function as an
analytical criterion is not new: Adorno inherited it directly from Schoenberg’s
variant of Formenlehre. Adorno was, of course, a lifelong critical supporter of
Schoenberg and his school. Although Schoenberg’s own main theoretical
text on musical form, the Fundamentals of Musical Composition, was published
at the very end of Adorno’s life, and although there is no evidence that Berg,
Adorno’s composition teacher, passed on any of Schoenberg’s theoretical ideas
to his student, it is beyond doubt that Adorno had a wealth of opportunities
to familiarize himself with Schoenberg’s ideas about musical form.39 He had
privileged access to a vivid oral tradition promulgated by different generations
and branches of the Schoenberg School. By the 1960s, moreover, at least one
Schoenbergian Formenlehre had appeared in print: Ratz’s Einführung in die musi-
kalische Formenlehre, of which Adorno owned a copy.40

The concept of formal function is, of course, front and center in Ratz’s the-
ory. In his preface, Ratz describes his theory as a funktionelle Formenlehre whose
aim is to describe “the means that allow the individual parts of a composition
to fulfill their function (e.g., transition, subordinate theme, development, etc.)
in the formal construction, similar to the different organs in a living organ-
ism.”41 To be sure, and as Hermann Danuser has pointed out, Adorno rejects
the organicist outlook that was a central aesthetic tenet for Ratz (as well as
for Schoenberg), but apart from that, his understanding of formal functions is
very close to Ratz’s.42
Second, Adorno’s emphasis on formal function operates in tandem
with a problematized understanding of the relationship between the whole
and its parts in musical form. He does not view analysis as a top-down pro-
cess that leads from the whole to the parts out of which that whole is com-
posed. Although he does not deem the form-functional requirements of the
whole irrelevant, Adorno proposes instead to deduce a unit’s function from
its internal organization, complementing the traditional top-down view with
a bottom-up approach to musical form; one might speak in this respect of a
unit’s “material” formal function. This perspective is never made explicit in the
Mahler book, but it becomes crystal clear in the “Wiener Gedenkrede.” There,
Adorno writes that “since the music has been composed from bottom to top, it
must be heard from bottom to top. One must abandon oneself to the flow of
the work, from one chapter to the next, as with a story when you do not know
how it is going to end.” The music’s logic, he continues, “follows from the orga-
nization and definition of the individual figures, rather than from an abstract,
preordained design.”43
This perspective differs fundamentally from the “academic” Formenlehren
that would have been available to Adorno at the time of his writing.44 The
distinction is less self-evident than it seems. Ironically, a bottom-up approach
is exactly what determines the structure of any German Formenlehre of the
early and mid-twentieth century: in their ordering of materials, all these
are modeled on Adolf Bernhard Marx’s prototypical exposé in the third
volume of Die Lehre von der musikalischen Komposition, which leads from the
smallest formal units to ever larger and more complex ones—from a single
motive to a full-movement (sonata) form.45 Later in the nineteenth cen-
tury, there was a tendency to understand all of these formal units “tectoni-
cally,” meaning that analytical decisions were informed more by the size
and number of formal units as articulated by cadences and—especially—
by thematic-motivic content than by the harmonic and phrase-structural
processes that take place within those units. At any level in the form, one
expects a given unit to be followed by a balancing unit at the same level,
as in the antecedent and consequent of a period. At the larger scale, form
is seen as a conceptually preexisting vessel that is filled with material but
remains essentially unaffected by it.

Third, and most radically, Adorno’s claim that Mahler’s music “calls
its forms by their names” amounts to saying that music can analyze itself.
Adorno merely hints at this possibility in his Mahler book, but he develops it
more fully in his monograph on Alban Berg. There, he writes explicitly that
Berg’s music “to a certain extent effects its own analysis. . . . In Berg’s mature
works ultimately every phrase or partial entity not only divulges with com-
plete clarity to cognitive understanding its formal function, but also makes
that formal function so emphatic a part of the directly perceived phenom-
enon that a concluding phrase declares: I am a concluding phrase; and a
continuation: I am a continuation.”46 The best illustration of Adorno’s point
might be that it is indeed much easier to explain his analytical categories
deictically by means of examples than conceptually by means of definitions;
the music in question (e.g., many of the passages from Mahler’s symphonies
mentioned above) “says” what it is in a much more direct and eloquent way
than analytical prose is able to do.
Nominalism and Tension

It is no coincidence that Adorno finds the highest degree of form-functional
clarity in the works of Mahler and Berg. In their music, he argues, “extreme
determinacy of the composition” has become a historical necessity. The form
has to be organized bottom-up (from within the different parts) because the
large-scale formal scheme—the whole—has lost the ability to provide coher-
ence. “The less the music is articulated by [the conventions of] musical lan-
guage,” writes Adorno, “the more must it ensure its own articulation.”47 In
his Ästhetische Theorie, Adorno refers to this phenomenon as the “advance of
nominalism” in art, which is both a necessity and an aesthetic ideal, albeit one
about which Adorno remains ambivalent.48 A “nominalistic artwork” is one he
defines as being “organized from below to above, not by having principles of
organization foisted on it.”49 In this context, the term “nominalism” can be
taken quite literally, in the sense that units “call themselves by their names.”
But at the same time, of course, Adorno uses the term in its traditional philo-
sophical sense: when every piece organizes itself bottom-up rather than being
organized top-down, there are no longer any universals, but only particulars,
so that each work of art becomes its own type.
The extreme instance of such nominalism in music is what Adorno refers
to as musique informelle. As is well known, he introduces this notion as a uto-
pian vision of what the new music of the 1960s should aspire to; in this sense,
the term has been adopted and popularized most notably by Gianmario
Borio.50 But Adorno also uses it in relation to some of Schoenberg’s atonal
works of 1909, especially the last of the Three Piano Pieces, op. 11; the last
of the Five Orchestral Pieces, op. 16; the monodrama Erwartung, op. 17; and
the Four Orchestral Songs, op. 22. Adorno defines musique informelle as “a

type of music that has discarded all forms that are external or abstract to it or
that confront it in an inflexible way. Entirely free of anything that has been
superimposed on it heteronomously and that is alien to it, [such a music] is
nonetheless objectively compelling in itself and does not constitute itself in
terms of external laws.”51 The result is a large-scale form that is unique for
every work and that is generated from within (or bottom-up) rather than
from outside (or top-down).
Such extreme nominalism, however, is highly exceptional—perhaps
even hypothetical. For Adorno, the completely nominalistic work of art
is a contradiction in terms: “No artwork left blindly to itself possesses the
power of organization that would set up binding boundaries for itself. . . .
By being something made, artworks acquire [an] element of organization,
of being something directed, in the dramaturgical sense.”52 Individual parts
always require the “intervention of the guiding hand” (der Eingriff der lenk-
enden Hand), and behind that hand there is a mind that keeps an eye on
the whole.53 Context, in other words, continues to play a role. No matter
how well-defined their material formal function is, units rarely appear in a
vacuum. “Usually,” writes Adorno, “they are also determined by their relation
to what precedes them.”54
Accordingly, it would be wrong to say that Adorno’s materiale Formenlehre sim-
ply substitutes a bottom-up view of musical form for a top-down one. As he
explains in “Zum Problem der musikalischen Analyse,” the two perspectives
are complementary.55 Moreover, the relative emancipation of the parts from
the form prescribed by the whole does not necessarily lead to novel large-scale
forms. Even when the whole depends on the parts as much as the other way
around, this fact does not automatically result in a conflict between the two.
In Mahler’s music, for instance, a Suspension may still function as an episode in
the development, an Auflösungsfeld may still function as a closing group, and
so forth. Perhaps Adorno’s clearest example of this situation is the first move-
ment of Schoenberg’s Wind Quintet, op. 26. Although Adorno mentions the
Quintet only in passing in Mahler, it is enlightening to read his much earlier
essay from 1928 on that work in light of his later writings. The 1928 discussion
of the Quintet’s form is cast mainly in terms of pitch versus thematic-motivic
organization. Many aspects of it nonetheless resonate well with the materiale
Formenlehre. In twelve-tone music, Adorno argues, sonata form as a predisposed
scheme is no longer available. The result is that in the Quintet, “the sonata
form follows solely from the thematic relations. . . . It has stopped functioning
as an isolated determining principle [that lies] on top of the individual musical
events; it has moved into them.”56
However, if units are capable of articulating their function from within,
this also means that they can express a function that contradicts their position
in the form as a whole. Provided that they exhibit sufficient form-functional
clarity in their internal organization, units retain the capacity to express their

formal function regardless of their position in the whole. For instance, an

Erfüllung will continue to express its function even when it does not appear as
the Abgesang of a Barform, and a Zusammenbruch can appear anywhere in a form,
even though it is a descendant of the Auflösungsfeld typical of a closing group.
For both situations, Adorno provides an example from the first movement of
Mahler’s Ninth Symphony. The “fortissimo repetition of the main theme” near
the beginning of the movement is an example of a dislocated Erfüllung, the
end of the development an example of a dislocated Zusammenbruch.57
The result is a tension between the whole and the form-functional plan it
implies on the one hand, and the parts and the formal functions they express,
on the other. For Adorno, this tension between the parts and the whole—in
effect, the impossibility of their reconciliation—is at the heart of nineteenth-
century music after Beethoven. This tension also necessitates what Adorno
calls “construction”: “the form of works that is no longer imposed on them
ready-made yet does not arise directly out of them either, but rather originates
in its reflection through subjective reason.”58
Adorno’s Eroica analysis

Given the paucity of sustained analysis in Adorno’s published writings, it
is surprising to see how much analysis may be found in the fragments on
Beethoven that Adorno compiled over a period of almost twenty years
between 1938 and 1956. To be sure, nothing in these fragments comes close
to being a systematic or complete analysis; at most, they are analytical notes
of a preparatory nature that might eventually have led to such an analysis.59
Nonetheless, the analytical information on a number of works is substantial
enough to allow us to conjecture what such an analysis might have looked
like, especially when read in tandem with some of the central ideas from the
materiale Formenlehre. One of these works is the first movement of the Eroica,
on which Adorno’s notes provide some five pages of analytical information,
mostly concerning the exposition.60
Table 13.1 provides a schematic overview of Adorno’s reading of the Eroica
exposition. On the surface, there is little that seems remarkable. Modern-day
analysts will likely prefer to hear the tutti restatement of the main theme at
measure 37 as the beginning of the transition, but Adorno’s decision to post-
pone it to measure 45 is in line with Ratz’s practice of distinguishing between
an angegangene Wiederholung (incipient repetition) of the main theme and the
actual transition.61 Also open to debate is Adorno’s decision to hear measure
57 as the beginning of the subordinate-theme group—the obvious alternative
being, of course, measure 83. Adorno is fully aware of this: as we will see, the
“doubt over the beginning of the second subject,” as he calls it, is one of the
central themes in his analysis.62

Table 13.1. Adorno’s analysis of the first-movement exposition of Beethoven’s

Eroica: schematic overview
12 336 37–44 45–56 57108 109–47 148–53
Intro- [Main [angegan- Transition Subordinate- Closing [Retransi-

duction theme] gene Wiederho- (p. 155) theme group group tion]
(p. 151) lung] (p. 152) (p. 155)
Note: The page numbers refer to Adorno’s analytical notes.
Presented in the format of table 13.1, Adorno’s interpretation does not

differ significantly from any other traditional Formenlehre account of the
same exposition: it is static and thus neglects the processual aspect of the
form. Even from the disorganized arrangement of Adorno’s preparatory
notes, however, it becomes clear that it is precisely this processual perspec-
tive that he tries to bring out in his analysis: the form comes into being
through the interplay between the makeup and content of the individual
formal units (the parts) and the form of the entire movement (the whole).
It clearly transpires from Adorno’s wording that he primarily understands
the relationship between parts and whole as emerging from the smallest
parts: the internal organization of formal units at the local level articulates
the form in its entirety. But at the same time, tension arises between the
large-scale form that some individual formal units project and the inter-
nal organization of other formal units that are at odds with the position
they occupy in the overarching form. It is this tension that accounts for the
form’s dynamic character.
Rearranging Adorno’s scattered analytical notes permits a hypothetical
reconstruction of his interpretation, which itself takes the form of a two-step
process. In a first move, Adorno observes that in certain cases the character-
istics of formal units correspond to the role those units play in the form as
a whole. What leads Adorno to call measure 45 the beginning of the transi-
tion, for instance, is that in his view, the four-measure unit beginning there
has all the characteristics of a “transition model”; literally giving voice to the
music, he writes: “I am a transition model.”63 And the passage from measure
109, for Adorno, unambiguously marks the beginning of the closing group,
again because of its intrinsic qualities (as opposed to, say, proportional or
cadential considerations, which would probably lead one to locate the closing
group later in the movement). Moments like this furnish points of orienta-
tion in an exposition whose proliferation of thematic shapes might otherwise
appear cluttered.64

In a second step, Adorno notes that the other units are at odds with this
larger scheme. This tension becomes a central idea in his discussion of the
subordinate-theme group. Adorno never questions that the subordinate-
theme group begins in measure 57—at least from the point of view of the
form as a whole. That measures 57–108 occupy the subordinate-theme posi-
tion is not only clear because of the exposition’s tonal plan—at measure 57,
“the tonic of [the dominant] B-flat major enters”—but is also dependent on
the fact that the previous unit is intrinsically a transition and the following
one intrinsically a closing group. The music sounding at measures 57–64,
however, fails to live up to the expectations raised by its position: it does not
constitute a “true” subordinate theme. Adorno lists three (interdependent)
reasons for this: (a) the theme is kept “somewhat noncommittal,”65 (b) last-
ing a mere eight measures, it is too short in comparison to the overall propor-
tions of the movement, and (c) it is interrupted by a much more memorable
figure at measure 65.66
The theme that appears at measure 83 is much more acceptable as a sub-
ordinate theme. However, because it is introduced so late in the subordinate-
theme group, Adorno argues, it does not sound like a beginning but like a
consequent, more specifically, like the “consequent phrase to a non-existent
antecedent phrase.”67 Even though Adorno notes a motivic connection to
measures 57–64, it seems fair to say that measures 83–108 mainly project their
function from within, that is, regardless of their surroundings: they are a con-
sequent, even though no antecedent precedes them.68 From the perspective
of this consequent, the theme at measure 57 is a placeholder that merely hints
at the actual subordinate theme that never really materializes; “the ‘theme,’”
Adorno writes, “is left out.”69 The music in measures 57–64 does not fulfill the
expectations imposed upon it by the form as a whole; the listener is deceived
and once again, Adorno gives voice to the music itself: “So you thought I was a
subordinate theme!”70 The characteristic closing group at measure 109, more-
over, makes it clear that there is no hope that an actual subordinate theme
will be introduced later in the exposition.71 Table 13.2 attempts to capture the
processual nature of Adorno’s analysis.
The vacuum created in the subordinate theme group, Adorno concludes,
is filled in by the new theme in the development: “The new theme is the song
theme that had been omitted.”72 The new theme is, in other words, a dislocated
subordinate theme that projects its function through its intrinsic characteristics
and in spite of its position in the overall form.73 Adorno sees his interpretation
confirmed when the new theme—as the “real” subordinate theme—receives
“its own recapitulation” in the movement’s coda (mm. 585ff).74
❧ ❧ ❧

Table 13.2. Relation between the whole and the parts in measures 45–155
of the first movement of Beethoven’s Eroica, according to Adorno’s
analytical notes
Transition → Subordinate-
“whole”: theme No subordinate- Closing group
position theme position
↑ ↕ ↕ ↑
“parts”: “Material” No “material” “Material” “Material”

transition subordinate subordinate closing
theme theme group
mm. 45–56 57–82 83–108 109–55
Notes: Upward arrows: the form of the whole is articulated by the material characteristics of
the parts; right arrows: one unit follows logically from the previous within the whole; double-
pointing arrows: there is a conflict between a unit’s contextual function (its position within
the whole) and its material function.
Readers familiar with the writings of Caplin and Schmalfeldt will have noticed
the many aspects of Adorno’s materiale Formenlehre that resonate with central
concerns of the modern theory of formal functions. In part, of course, this
is merely the consequence of both theories’ common origin in Schoenberg’s
thinking about musical form. Adorno, like Caplin and Schmalfeldt, not only
draws much of his terminology from Schoenberg and Ratz (with the excep-
tion, of course, of those terms coined specifically for the analysis of Mahler’s
music) but also takes the central notion of “formal function” from them.
In itself, this serves as a useful reminder not only that the notion of “formal
function,” although popularized in North America by Caplin, has a long his-
tory in European thinking about musical form, but also that Schoenbergian
Formenlehre was more diverse than the orthodox Schoenberg–Ratz tradition.
There are, however, more specific connections between Adorno and the
theory of formal functions, both in its Caplinian and Schmalfeldtian incarna-
tions. I want to single out two of these by way of conclusion. Schmalfeldt has
suggested that Adorno and Caplin share a concern for terminological specific-
ity. “Caplin’s sharply-defined form-functional categories,” she writes, “fulfill a
long-unmet need expressed by Adorno. . . . One cannot help imagining that
Adorno would have been interested in Caplin’s achievement.”75 This is more
than just a matter of terminology. One of the distinguishing characteristics
of Adorno’s materiale Formenlehre is the problematization of the relationship
between the part and the whole, in which a top-down view of musical form
is complemented by an equally important bottom-up perspective. In empha-
sizing the capacity of individual formal units to express their function—and,
more generally, in distinguishing between the function projected by those

units themselves and that imposed by the form as a whole—Adorno’s mate-

riale Formenlehre foreshadows the distinction that Caplin has formalized as that
between a unit’s “contextual” and “intrinsic” formal function. Whereas contex-
tual formal function depends on “a musical passage’s immediate surroundings
in order to determine its temporal location,” the identification of intrinsic for-
mal function “involves comparing the passage’s constituent elements with a set
of norms—implicitly acquired by mere exposure to music or explicitly learned
through formal musical training—that link certain musical characteristics to
specific formal functions.”76 To be sure, Caplin is less interested than Adorno
in thematizing the tension that may arise from the clash between intrinsic and
contextual formal functionality. Nonetheless, in Classical Form he briefly dis-
cusses the situation of “formal ‘dissonance,’” which emerges when “a given
function is actually placed differently from its expressed temporal position.”77
The connections between Adorno’s materiale Formenlehre and Schmalfeldt’s
work are even more obvious. To both, the notion that musical form is a process
in an emphatic sense is central. Like Schmalfeldt, Adorno the analyst has a
distinct preference for the term “becoming,” and more than one of the moves
he makes in his Eroica analysis amounts to what Schmalfeldt would call “ret-
rospective reinterpretation.” Indeed, Adorno’s analysis as a whole—recast, of
course, in an updated terminology—would not have seemed out of place in
Schmalfeldt’s In the Process of Becoming. As I have argued above, an emphasis on
the dynamic and the processual characterizes not only Adorno’s view of music
but also his thinking in general. It is from Adorno’s philosophy, and from
Hegelian and post-Hegelian philosophy in general, that Schmalfeldt takes
the cue for her analytical work. But it testifies to the validity of her approach
that an emphasis on the processual nature of musical form is not just a conse-
quence that can be drawn from Adorno’s philosophical stance, but is integral
to his own understanding of musical form as well.
Notes
Music of Haydn, Mozart, and Beethoven (New York: Oxford University Press, 1998), v,
3. Caplin also briefly addresses the Schoenbergian influence in his “methodologi-
cal reflection” in “What Are Formal Functions,” in Caplin, James Hepokoski, and
James Webster, Musical Form, Forms & Formenlehre: Three Methodological Reflections,
ed. Pieter Bergé (Leuven: Leuven University Press, 2009), 25. A more detailed
discussion was given in an unpublished paper “Schoenberg’s Theories of Form:
Innovations, Limitations, Amplifications,” read at the conference “Schoenberg’s
Legacy on Form,” McGill University, Montreal, May 17–18, 2008. The seminal texts
of Schoenbergian Formenlehre are Schoenberg’s Fundamentals of Musical Composition,
ed. Gerald Strang and Leonard Stein (London: Faber and Faber, 1967) and
Erwin Ratz, Einführung in die musikalische Formenlehre: Über Formprinzipien in den in
den Inventionen und Fugen J. S. Bachs und ihre Bedeutung für die Kompositionstechnik

Beethovens (Vienna: Österreichischer Bundesverlag, 1951; 3rd rev. ed.: Universal

Edition, 1973).
2. Josef Rufer, Composition with Twelve Tones, trans. Humphrey Searle (London: Barrie
and Rockliff, 1954); Erwin Stein, Form and Performance (London: Faber and Faber,
1962); Anton Webern, The Path to New Music, ed. Willi Reich (Bryn Mawr, PA:
Theodore Presser, 1963); and Über musikalische Formen, ed. Neil Boynton (Mainz:
Schott, 2002). For Ratz on funktionelle Formenlehre, see Einführung, 9 and 56, as well
as below.
3. William E. Caplin, “Funktionale Komponenten im achttaktigen Satz,” Musiktheorie 1
(1986): 239–60.
4. See, for example, Janet Schmalfeldt, “Towards a Reconciliation of Schenkerian
Concepts with Traditional and Recent Theories of Musical Form,” Music Analysis 10
(1991): 233–87.
5. Janet Schmalfeldt, “Form as the Process of Becoming: The Beethoven-Hegelian
Tradition and the Tempest Sonata,” Beethoven Forum 4 (1995): 37–71.
6. Janet Schmalfeldt, In the Process of Becoming: Analytic and Philosophical Perspectives on
Form in Early Nineteenth-Century Music (New York: Oxford University Press, 2011),
ix–x.
7. Ibid., 16. I am aware of only one substantial discussion of the materiale Formenlehre
in English: Max Paddison’s Adorno’s Aesthetics of Music (Cambridge: Cambridge
University Press, 1993), 174–82. Although the present chapter will inevitably
rehearse some of Paddison’s points, it is aimed more toward an elucidation of the
technical aspects of the materiale Formenlehre than toward its interpretation against
the background of Adorno’s philosophy as a whole. For a more recent assessment
of Adorno’s materiale Formenlehre in German, see Hermann Danuser, “‘Materiale
Formenlehre’: Ein Beitrag Theodor W. Adornos zur Theorie der Musik,” in
Musikalische Analyse und kritische Theorie: Zu Adornos Philosophie der Musik, ed. Adolf
Nowak and Markus Fahlbusch (Tutzing: Schneider, 2007), 19–49.
8. Rose Rosengard Subotnik, “The Role of Ideology in the Study of Western Music,”
Journal of Musicology 2 (1983): 9. See also her “Adorno and the New Musicology,”
in Adorno: A Critical Reader, ed. Nigel Gibson and Andrew Rubin (Malden, MA:
Blackwell, 2002), 234–54.
9. Rose Rosengard Subotnik, “Adorno’s Diagnosis of Beethoven’s Late Style: Early
Symptom of a Fatal Condition,” Journal of the American Musicological Society 29
(1976): 242.
10. One classic example is Susan McClary, “Narrative Agendas in ‘Absolute’ Music:
Identity and Difference in Brahms’s Third Symphony,” in Musicology and Difference:
Gender and Sexuality in Music Scholarship, ed. Ruth E. Solie (Berkeley: University of
California Press, 1995), 326–44.
11. Theodor W. Adorno, “Zum Problem der musikalischen Analyse,” Frankfurter
Adorno-Blätter 7 (2001): 78, 79. English translation (ET) by Max Paddison as “On
the Problem of Musical Analysis,” Music Analysis 1 (1982): 176. The translation was
published before the original.
12. Ibid., 80 (ET, 177). The potential implications of this statement are far-reaching.
À la limite, they might mean that truth content will always have to remain unspo-
ken or can perhaps be expressed only in analytical terms: as soon as one reverts
to plain, nontechnical language, and thus recedes from the immersion into the
depths of the work’s technical structure, one is already distancing oneself again
from the work’s truth content.

13. For an early diagnosis of the gap between technical and sociological interpretation
in Adorno’s analyses, see Carl Dahlhaus, “The Musical Work of Art as a Subject of
Sociology,” in Schoenberg and the New Music, trans. Derrick Puffet and Alfred Clayton
(Cambridge: Cambridge University Press, 1987), 234–47.
14. A recent example is Michael Spitzer, Music as Philosophy: Adorno and Beethoven’s Late
Style (Bloomington: Indiana University Press, 2006). Although Spitzer’s general
framework is Adornian, his actual analyses are not; in fact, in his introduction, he
explains that his method consists in “employ[ing] the facts of modern musicology
and analysis as an illumination of Adorno’s ideas” (5).
15. “[ein] merkwürdige[s] Mißverhältnis zwischen der Stellung, die Adorno der musi-
kalischen Analyse einräumt, und der Art und Weise, wie er selbst sich in seinem
Werke zu ihr verhält.” Ludwig Holtmeier, “Analyzing Adorno–Adorno Analyzing,”
in Adorno im Widerstreit: Zur Präsenz seines Denkens, ed. Wolfram Ette, Günter Figal,
Richard Klein, and Günter Peters (Freiburg: Alber, 2004), 184. On the ambiguous
position analysis holds in Adorno’s thinking, see also Max Paddison, “Immanent
Critique or Musical Stocktaking? Adorno and the Problem of Musical Analysis,” in
Gibson and Rubin, Adorno: A Critical Reader, 209–33.
16. See, for example, Diether de la Motte, “Adornos musikalische Analysen,” in Adorno
und die Musik, ed. Otto Kolleritsch (Vienna: Universal, 1979), 52–63; Holtmeier,
“Analyzing Adorno–Adorno Analyzing”; Julian Johnson, “Vers une analyse
informelle,” in Nowak and Fahlbusch, Musikalische Analyse und kritische Theorie,
102–13.
17. Julian Johnson, “Analysis in Adorno’s Aesthetics of Music,” Music Analysis 14
(1995): 300.
18. “den Begriff mit dem von ihm gemeinten so lange zu konfrontieren, bis sich zeigt,
daß sich zwischen einem solchen Begriff und der von ihm gemeinten Sache gewisse
Schwierigkeiten herstellen, die dann dazu nötigen, den Begriff mit dem Fortgang
des Denkens in einer gewissen Weise zu verändern.” Adorno, Nachgelassene Schriften,
ed. Christoph Ziermann, part 4, vol. 2, Einführung in die Dialektik (Frankfurt:
Suhrkamp, 2010), 18.
19. Georg Wilhelm Friedrich Hegel, Werke, ed. Eva Moldenhauer and Karl Markus
Michel, vol. 3, Phänomenologie des Geistes (Frankfurt-am-Main: Suhrkamp, 1986), 38.
20. “Methode des Denkens, . . . einer Wesenseigentümlichkeit der zu betrachtenden
Sache selbst gerecht zu werden.” Adorno, Einführung in die Dialektik, 13.
21. Adorno, Gesammelte Schriften, ed. Gretel Adorno and Rolf Tiedemann, vol. 7
(Frankfurt-am-Main: Suhrkamp, 1970); English trans. by Robert Hullot-Kentor
as Aesthetic Theory (Minneapolis: University of Minnesota Press, 1997). In what
follows, and unless otherwise indicated, all references to Adorno’s writings are
first to the German edition in the Gesammelte Schriften, ed. Rolf Tiedemann,
Gretel Adorno, Susan Buck-Morss, and Klaus Schulz (Frankfurt-am-Main:
Suhrkamp, 1970–1986) (GS), then to the English translation (ET).
22. This and the following two paragraphs draw heavily on Paddison, Adorno’s Aesthetics
of Music, 149–56.
23. Adorno, Ästhetische Theorie, 222 (ET, 147).
24. Die Schwierigkeit . . . ist die, da zwar Ganzes und Teil nur durch Beziehung aufein-
ander begriffen werden können, daß aber jeweils, wenn Sie den Teil haben, Ihnen
das Ganze keineswegs positiv gegeben ist und daß umgekehrt, wenn Sie das Ganze
denken, damit keineswegs seine Teile positiv gegeben worden sind . . . in die
Dialektik ist das Verhältnis von Ganzem und Teil kein bloßes der Subsumtion, . . .

sondern es ist ein dynamisches Verhältnis: Diese beiden Momente produzieren sich
wechselseitig gegenseitig und sind nicht gleichsam . . . gleichseitig miteinander
da.” Adorno, Einführung in die Dialektik, 129.
25. Adorno, Mahler: Eine musikalische Physiognomik (GS 13:149–319), 193–200 (English
trans. by Edmund Jephcott as Mahler: A Musical Physiognomy [Chicago: University
of Chicago Press, 1992]); “[Mahler:] Wiener Gedenkrede 1960,” in Quasi una fan-
tasia (GS 16:323–38), 329 (English trans. by Rodney Livingstone in Quasi una fan-
tasia: Essays on Modern Music [London: Verso, 1998], 81–97); “Vers une musique
informelle,” in Quasi una fantasia (GS 16:493–540), 503–5 (English trans. in Quasi
una fantasia, 269–322); Einleitung in die Musiksoziologie: Zwölf theoretische Vorlesungen
(GS 14:169–433), 243–45 (English trans. by E. B. Ashton as Introduction to the Sociology
of Music [New York: Seabury Press, 1976]); Berg: Der Meister des kleinsten Übergangs
(GS 13:321–494), 373–74 (English translation by Juliane Brand and Christopher
Hailey as Berg: Master of the Smallest Link [Cambridge: Cambridge University Press,
1991]); “Zum Problem der musikalischen Analyse” (see above note 11).
26. Adorno, Beethoven: Philosophie der Musik, ed. Rolf Tiedemann (Frankfurt: Suhrkamp,
1993); English trans. by Edmund Jephcott as Beethoven: The Philosophy of Music
(Stanford, CA: Stanford University Press, 1998). See especially fragments nos. 44
and 233.
27. Adorno, Ästhetische Theorie, esp. 211–26, 296–305, and 326–34 (ET, 140–51, 199–
205, 219–25).
28. Adorno, Mahler, 193–94 (ET, 44–45; translation modified).
29. Ibid., 197 (ET, 48).
30. Adorno, “Zum Problem der musikalischen Analyse,” 88 (ET, 185; translation
modified).
31. Adorno, Mahler, 194 (ET, 45; translation modified). The reference to Ratz is to
“Zum Formproblem bei Gustav Mahler: Eine Analyse des ersten Satzes der Neunten
Symphonie,” Die Musikforschung 8 (1955): 176.
32. Adorno, Mahler, 194 (ET, 45). By “on the dominant,” Adorno obviously means
“over a tonic pedal in the dominant key.” Adorno does not provide any classical
examples, but he likely has in mind closing groups such as that in the first move-
ment of Beethoven’s Piano Sonata in A Major, op. 2, no. 2, or even the first move-
ment of the Pathétique, op. 13.
33. Ibid., 190 (ET, 41–42). Under the influence of Alfred Lorenz’s writings, Adorno
surely considered Abgesang a “traditional” category of form. See, for example,
Lorenz, Das Geheimnis der Form bei Richard Wagner, 4 vols. (Berlin: Hesse, 1924; repr.,
Tutzing: Schneider, 1966).
34. Adorno does not give any further specifications as to which passages he means. For
the Third, he may have in mind measures 351–62; for the Sixth, he is likely think-
ing of measures 754–72.
35. The category of Suspension still has a traditional formal analogue in the episode
that interrupts the expected course of a form (Adorno describes these as “exter-
ritorial” in Mahler, 190; ET, 41). By contrast, the connection Adorno sees between
Weltlauf and perpetuum mobile (Mahler, 154–55; ET, 6–7) is, although obvious, not
directly a matter of musical form. Durchbruch is alone among Adorno’s categories in
having no traditional analogue at all—in this sense (because it describes something
for which there is no other term), it is probably not a coincidence that it is the
only one of Adorno’s terms to have found its way into current analytical vocabu-
lary beyond Mahler scholarship. The classic instance of an Adornian Durchbruch

remains, of course, the passage immediately after the development of the first
movement of Mahler’s First Symphony (Mahler, 152–54 and 161–62; ET, 4–6, 13).
36. “Functional characters [are] what each individual part contributes to the form.”
Ibid., 197 (ET, 48).
37. Compare ibid., 197 (ET, 47): “Mahler’s characters . . . coincide with their emphatic
formal function.”
38. Ibid., 196–97 (ET, 47) and 197 (ET, 48).
39. Reminiscing about his lessons with Berg in the 1920s in his monograph on that
composer, Adorno states that they did not include “the study of form” (Formenlehre)
(Berg, 364; ET, 32).
40. This copy is preserved in Adorno’s “Nachlassbibliothek” at the Adorno Archiv in
Frankfurt-am-Main. Adorno also owned a copy of Schoenberg’s earlier (and much
more rudimentary) Models for Beginners in Composition (New York: Schirmer, 1943).
41. “Die Beschreibung der Mittel, die bewirken, daß die einzelnen Teile einer
Komposition die ihnen zukommende Funktion (also z.B. der Überleitung, des
Seitensatzes, der Durchführung usw.) im formalen Aufbau zu erfüllen vermö-
gen, ähnlich wie die verschiedenen Organe im lebenden Organismus.” Ratz,
Einführung, 9.
42. Danuser, “Materiale Formenlehre,” 29. To express this essential distinction,
Danuser refers to Ratz’s Formenlehre as funktionel and to Adorno’s as funktional.
43. Adorno, “Wiener Gedenkrede,” 329 (ET, 87; translation modified).
44. Perhaps the best-known such “academic” Formenlehre is Hugo Leichtentritt’s
Musikalische Formenlehre, first published with Breitkopf und Härtel in Leipzig in
1911 and arguably the most widespread textbook on musical form in Germany
for over half a century. By the time of Adorno’s death in 1969, it had been
reprinted five times, and an English translation as Musical Form first appeared
in 1951 (Cambridge, MA: Harvard University Press). Although I have not been
able to confirm that Adorno actually owned this book (no copy is preserved in
his Nachlassbibliothek), he certainly knew it because he refers to it in one of his
Beethoven fragments (151). The only traditional Formenlehre Adorno did own at
the time of his death is Christian Lobe’s Katechismus der Kompositionslehre (Leipzig:
Weber, 1882). Readers will also be interested to learn that he owned copies of sev-
eral of Heinrich Schenker’s writings and, as his penciled-in annotations show, also
perused them.
45. Adolf Bernhard Marx, Die Lehre von der musikalischen Komposition, praktisch-theoretisch,
vol. 3 (Leipzig: Breitkopf und Härtel, 1845).
46. Adorno, Berg, 373 (ET, 39).
47. Adorno, Mahler, 196, 197 (ET, 47, 48; translation modified). Ibid., 197.
48. Adorno, Ästhetische Theorie, 296 (ET, 199; translation modified).
49. Ibid., 327 (ET, 220).
50. Gianmario Borio, Musikalische Avantgarde um 1960: Entwurf einer Theorie der informel-
len Musik (Laaber: Laaber, 1990).
51. Adorno, “Vers une musique informelle,” 496 (ET, 272; translation modified).
53. Ibid., 329 (ET, 221).
54. Adorno, Mahler, 198 (ET, 48–49).
55. Adorno, “Zum Problem der musikalischen Analyse,” 85–86 (ET, 182). Adorno cites
Ratz in this respect (see, e.g., Einführung, 8).

56. “Die Sonatenform folgt [lediglich] aus den thematischen Relationen. . . . Sie
hat aufgehört, als objektives Bestimmungsprinzip oberhalb der musikalischen
Einzelereignissen isoliert zu gelten; sie ist hineingezogen in jene.” Adorno,
“Schönbergs Bläserquintett [1928],” in Moments musicaux (GS 17), 142, 144.
57. See Adorno, Mahler, 191, 194 (ET, 42, 45); translation modified. On this aspect
of the materiale Formenlehre, see also Arno Forchert, “Zur Auflösung traditioneller
Formkategorien in der Musik um 1900: Probleme formaler Organisation bei
Mahler und Strauss,” Archiv für Musikwissenschaft 32 (1975): 85–98.
59. Although one can reasonably doubt that much of the analysis would have survived
into a Beethoven monograph had Adorno ever completed it. For an introduction
to Adorno’s Beethoven, see Stephen Hinton, “Adorno’s Unfinished Beethoven,”
Beethoven Forum 5 (1996): 139–54; and Spitzer, Music as Philosophy, 44–70.
60. The relevant fragments are nos. 230–33. As far as I have been able to establish,
Adorno’s Nachlass does not contain any further analytical material on the Eroica;
his copy of the score (Adorno Archiv, Frankfurt-am-Main, NB Adorno 3714), for
instance, does not contain any annotations at all.
61. See Ratz, Einführung, 37: “Die Exposition [der Sonatenform] besteht aus dem
Hauptthema, der angegangenen Wiederholung des Hauptthemas, der Überleitung,
dem Seitensatz und dem Schlußsatz.” See also Ratz’s analysis of the first movement
of the Appassionata in the same volume (155–59). It should be noted, though, that
neither passage was included in the 1951 edition of Ratz’s book.
62. Adorno, Beethoven, 152 (ET, 102).
63. Ibid., 155 (ET, 104).
64. “The exposition, immensely rich in figures and quite unschematic in its intent,
nevertheless emphasizes the schema by means of certain characters used as if for
orientation” (ibid., 155; ET, 104). Note, incidentally, Adorno’s use of the term
“characters” (Charaktere) here, the same word that he would later use in his discus-
sion of the materiale Formenlehre in Mahler.
65. Adorno, Beethoven, 153 (ET, 103). Adorno does not specify what he means by “some-
what noncommittal.” One way of understanding his comment is that the passage
eschews a strong initiating function; in fact, it sounds more like a postcadential
“after-the-end” (as is made explicit at the passage’s final return immediately before
the end of the movement in mm. 677–84).
66. From the perspective of modern Formenlehre, it has been argued that the first
subordinate theme is not over by measure 64, but that measures 65–82 func-
tion instead as an expanded continuation to the presentation in measures 57–64
(which itself consists of two “loops”); see William E. Caplin, “Structural Expansion
in Beethoven’s Symphonic Forms,” in Beethoven’s Compositional Process, ed. William
Kinderman (Lincoln: University of Nebraska Press, 1991), 41. Although this inter-
pretation certainly rings true, one should not be tempted to explain away the sense
of interruption that measures 65 and following convey; if this is a continuation,
it certainly is not the one that one would have expected on the basis of measures
57–64.
67. Adorno, Beethoven, 153 (ET 103); Adorno’s emphasis.
68. Context, however, does play some role in Adorno’s analysis of measures 83–108:
they are a consequent partly because “so much has happened” since the beginning
of the subordinate-theme group. Ibid., 153.
69. Ibid., 153 (translation modified).

70. Ibid., 155 (ET, 233).

71. It is tempting to paraphrase Dahlhaus’s memorable formulation about the first
movement of the Tempest Sonata: measures 57–64 are not yet a theme; mm. 83–108
are one no longer. Carl Dahlhaus, Ludwig van Beethoven und seine Zeit (Laaber:
Laaber Verlag, 1987), 211–12; English trans. by Mary Whittall as Ludwig van
Beethoven: Approaches to His Music (Oxford: Oxford University Press, 1991). On this
matter, see also Schmalfeldt, In the Process of Becoming; and Steven Vande Moortele,
“In Search of Romantic Form,” Music Analysis 32 (2013): 417–23.
72. Adorno, Beethoven, 154 (ET, 103); Adorno’s emphasis.
73. Adorno mentions a motivic connection between the new theme’s lower voice and
that of measures 57–64. It is not clear what he means; perhaps he is thinking of
the descending line (B♭–A–A♭–G–F–E♭–D in mm. 57–59, G–F♯–E–D♯–D–C♯–B
in mm. 289–91) they have in common. He does not mention the more obvious piz-
zicato figure in the double basses that connects the new theme in the development
to measures 83 and following.
74. Adorno, Beethoven, 154 (ET, 103).
75. Schmalfeldt, In the Process of Becoming, 16–17.
76. Michel Vallières, Daphne Tan, William E. Caplin, and Stephen McAdams,
“Perception of Intrinsic Formal Functionality: An Empirical Investigation of
Mozart’s Materials,” Journal of Interdisciplinary Music Studies 3 (2009): 18.
77. Caplin, Classical Form, 111. Recast as “form-functional dissonance,” the notion also
appears in Caplin, “Teaching Classical Form: Strict Categories vs. Flexible Analysis,”
Dutch Journal of Music Theory 18 (2013): 133.

Afterword
Janet Schmalfeldt
I am deeply honored to have been asked to contribute an afterword for this

impressive volume, but the nature of my role remains elusive to me. Coming
as it does after so many fine words on a remarkably diverse series of topics, the
afterword—an “after-the-end,” to borrow William Caplin’s expression—hardly
needs to summarize the contents of individual chapters, already so succinctly
accomplished in the editors’ introduction. Nor should it be my place to offer
a review or critique of the collection; this will fall to future readers. Just the
same, I cannot resist opening with admiration for the arduous work that our
three editors and their contributors have undertaken in the development of
this project, and for the outstanding outcome they have achieved.
I have been described by Steven Vande Moortele (“The Philosopher as
Theorist,” p. 411) as Caplin’s “principal interlocutor” in the early develop-
ment of his treatise on Classical form, and so it seems fitting that I should
elaborate on this generous compliment. We all have moments in our lives
that remain unforgettable, down to the last detail, and one of them for me
was the arrival of Bill Caplin as a new assistant professor of music theory in
the School of Music at McGill University in the fall of 1978. I was waiting for
him at the old reception desk in the Strathcona Music Building—excited to
make his acquaintance, and relieved that, now three years into my first aca-
demic appointment, I would no longer be the sole full-time music theorist
within the school. Though probably exhausted from his travels, Bill began
almost immediately to talk about Carl Dahlhaus, having come directly from
seminars with him in Berlin. Our conversations took off from there, and, in
effect, they have never ended.
I had been hired at McGill largely on the basis of my Schenkerian and set-
theory credentials as a graduate student at Yale (remember!—that was in 1975),
and I was still struggling to complete my dissertation, about Alban Berg’s opera
Wozzeck. Bill, for his part, was three years away from completing his University
of Chicago dissertation, on “Theories of Harmonic-Metric Relationships from
Rameau to Riemann”; and so he was as yet very much involved with historical

after word 435
studies on harmony and meter (his first article, “Der Akzent des Anfangs: Zur
Theorie des musikalischen Taktes,” appeared in 1978).1 Thus our first years
together as colleagues often involved simply encouraging one another about
our respective dissertation goals. As a diversion from our dissertation anxieties,
we Americans eventually managed to find Canadian friends who were occa-
sionally willing to play poker; Bill would fold for most of the evening, waiting
for the rest of us to squander our gains, and then walk away with our money.
Before Bill had left Berlin, Dahlhaus had inspired him to begin thinking
very seriously about aspects of Classical form, especially through the writings
of—not Arnold Schoenberg (!)—but Erwin Ratz. Around 1979, Bill translated
the first chapter of Ratz’s Einführung in die musikalische Formenhlehre, and we
both began to explore Ratz’s work in our teaching.2 Facing heavy teaching
loads, with lots of students, we each taught sections of what are now called
Tonal Theory and Analysis I and II at McGill; Bill reminds me that, before he
had arrived, I had put together an anthology of pieces for analysis, which we
now both used. The chosen textbook at that time became unsatisfactory for
both of us, particularly on matters of form; thus we found ourselves endlessly
discussing the anthology repertoire, bickering but often agreeing about formal
issues, and trying to develop new approaches. We were still young enough then
to sustain animated phone conversations at three in the morning—moments I
have not forgotten.
My favorite courses at McGill were the ones on Nineteenth-Century Analysis
and on Twentieth-Century Analysis—both, blessedly, full-year courses, and
ones that Bill also taught. My tendency for the nineteenth-century course was
sometimes to include, as analytic subjects, solo piano works that I would per-
form that year at McGill. When I have stumbled since then on my old cop-
ies of those scores—for example, Schubert’s Sonata in B-flat, D. 960, Chopin’s
Polonaise-Fantasy—my annotations have told me that, in those early years at
McGill, I was still employing disturbingly antiquated ideas about form. But a
review of my “grade books” from those days reveals that assignments on “sen-
tence and period” had begun to creep into all of my courses by 1979–80. Why?
Because by then Caplin had discovered Schoenberg, and because we were both
absolutely certain that Schoenberg’s Fundamentals of Musical Composition could
serve as the point of departure for a refreshingly new Formenlehre.3
Bill’s breakthrough article was his “The ‘Expanded Cadential Progression’:
A Category for the Analysis of Classical Form,” not published until 1987, but
read in Ottawa as early as 1982.4 It was during his first sabbatical, in 1983–
84, that he began to write the initial chapters of his Classical Form, with me
as his backup reader, cheering him on. An unpublished early version of his
first seven chapters, plus two volumes of examples, worked their way into our
courses, and these were the materials that I took with me when I left McGill for
a position at Yale in the fall of 1987. With the greatest conviction, I introduced
Caplinian ideas about Classical form and formal function to my Yale students;

436 after word
I did the same during my subsequent years as a professor at Tufts University, as

well as—more recently—during my ten-week stint in the role of visiting pro-
fessor at the University of Chicago in spring 2014; in Chicago it was a special
pleasure to bring Caplin’s ideas into dialogue with aspects of James Hepokoski
and Warren Darcy’s Elements of Sonata Theory.5
Bill’s and my respective speaking engagements and publication records
seem to have begun to take on definition in the early 1990s—especially
Bill’s, with astonishing profusion, culminating in the appearance of his bril-
liant Classical Form in 1998, and now his Analyzing Classical Form: An Approach
for the Classroom (2013).6 It is hard for me to imagine participating in confer-
ences without Bill’s presence there as well; for example, we each gave papers
at conferences in Bad Sulzberg (2005), Tallinn (2006), and Freiburg (2007),
and we led workshops at the Mannes Institute for Advanced Studies in Music
Theory (2004; 2011). Regular long-distance phone conversations keep us in
touch, and annual meetings of the Society for Music Theory give us chances
to catch up with each other’s projects. In fact, we had great fun butting heads
with one another about Beethoven’s “Tempest” Sonata at the special Montréal
2009 Society for Music Theory session concerning my work.7 Those who know
my publications over these many years, beginning with my very first article, on
relationships between performance and analysis (1985), are well aware of the
debt I owe in countless ways to Caplin’s work. In short, our meeting in the fall
of 1978 retrospectively became one of the most important events in my career,
and I am grateful that Bill has regarded the event as seminal for him, too. I
thank him for our enduring friendship and mutual support.
❧ ❧ ❧
William Caplin will without a doubt be exceedingly pleased about this Festschrift
in his honor. That former graduate and postdoctoral students, current col-
leagues at McGill, and distinguished scholars in our field have come together
to produce this collection attests to Bill’s abiding commitment to teaching, and
to the persuasiveness of his influence. What might be especially fascinating to
Bill is that, in their separate ways, some of these authors address topics and
adopt modes of writing about music that are not predominant in Bill’s own
work—ones that reach beyond the obvious concerns in this volume to extend
the concept of formal functions through applications to nineteenth-century
and later repertoires as well as new genres. For example, various authors pro-
vide rich historical and social backgrounds that step outside of “the music
itself” in order to contextualize it. Andrew Deruchie comes to Camille Saint-
Saëns’s defense as a progenitor of cyclic instrumental techniques, in face of
the neglect that the composer suffered within the Franckiste milieu of fin-de-
siècle France. François de Médicis offers a fine overview of recent analytic lit-
erature on Schubert’s “heavenly length.” In their chapter on René Leibowitz’s

after word 437
twelve-tone music and theory, Christoph Neidhöfer and Peter Schubert hypo-
thetically reconstruct Leibowitz’s associations with Schoenberg, his school, and
his work. Julie Pedneault-Deslauriers’s study of Schoenberg’s Verklärte Nacht
includes an account of its mixed early reception, a synopsis of the status and
preoccupations of poet Richard Dehmel, a comparative summary of earlier
views about the form of Schoenberg’s work, and a penetrating discussion of
late nineteenth-century debates over the hot social issues of love, sexuality, and
the patriarchal family. A full command of background literature relevant to
each author’s topic is a characteristic of all the essays in this volume.
Caplin would be the first to acknowledge that hermeneutic, narrative, meta-
phorical, and philosophical approaches to Classical instrumental music have
not played a central role in his writings. (In Classical Form, he dares to offer
one endnote about Peter Schickele’s—P. D. Q. Bach’s—satire on Beethoven’s
Fifth Symphony, and he cautiously clarifies the formal functions of musical
“before-the-beginnings” and “after-the-ends” by analogy to a running race; but
he admits to remaining “squeamish” about those efforts to this day). By con-
trast, L. Poundie Burstein’s “twisted formal functions” within the main theme
of Haydn’s Symphony no. 59 would seem to encourage a hermeneutic interpre-
tation: after the “dynamic thrust” of the presentation, the continuation “sug-
gests uncertainty and wavering,” only to be swept away by the unhesitating
character of the rest of the exposition. For Nathan John Martin, on Mozart’s
sonata-form arias, and Steven Huebner, on parlante textures in Verdi’s operas,
it is not enough simply to consider the mechanics of text-setting in these vocal
repertoires; both authors explore psychological dramas that motivate choices
of form and texture. Huebner explains how “Verdi follows the sense of the
text” in Nabucco’s Abigaille–Nabucco duet; Martin boldly correlates Pamina’s
despair, in her “Ach ich fühl’s” of Die Zauberflöte, with the outrageous misogyny
of Sarastro and his priests.
Giorgio Sanguinetti and Henry Klumpenhouwer go further in the direc-
tion of bringing narratives to the music they discuss—of endowing music with
the ability to express human behavior. For Sanguinetti, the “ongoing reprise”
in the first movement of Schubert’s unfinished “Reliquie” Piano Sonata has
been prepared by the “extreme case of formal and tonal wandering” within the
movement’s exposition: “Nothing in this movement seems to aim in the right
direction; everything appears to lose its way in a blind, erratic meandering—as
in a pessimistic rendition of [Homer’s] Odyssey”; the transition “spends a great
amount of energy going nowhere” (“Laborious Homecomings,” p. 326). (As
an example of numerous links among chapters in this collection, the transi-
tion that Sanguinetti discusses might well provide another demonstration of
what Brian Black identifies as Schubert’s “deflected cadence transitions” in
his chapter.) Klumpenhouwer detects a “narrative of denial” in both Caplin’s
and Hepokoski and Darcy’s respective formal interpretations of the subordi-
nate group within the first movement of Beethoven’s First Symphony; turning

438 after word
to orchestration, and to Hugo Riemann’s association of individual woodwind

instruments with distinct, even gendered characters, Klumpenhouwer portrays
the drama of “denial” through a narrative in which “humiliating errors” on
the part of the flute (in the exposition) and the oboe (in the recapitulation)
need to be “corrected” by their woodwind peers, as pedagogical “reeducators.”
Klumpenhouwer concludes with a polite but firm suggestion that hermeneutic
readings—“figural, narrative, or allegorical”—can be important in offsetting
the tendency of purely formal analyses to focus on the nature of convention,
rather than on the idiosyncratic and personal.
The chapter about Leibowitz by Neidhöfer and Schubert and the chapter on
Theodor W. Adorno by Vande Moortele each bring historical and philosophi-
cal depth to Caplin’s own accounts of the origins of his theory of formal func-
tions in Schoenbergian Formenlehre. Neither Leibowitz nor Adorno has played
a role in Caplin’s work to date; but we learn from Neidhöfer and Schubert,
for example, that Schoenberg, Webern, Leibowitz, and other early twelve-tone
composers strove to capture the tonal-harmonic functions and effects of “basic
ideas” (modèles), sentences (phrases), periods (périodes), and contrasting formal
sections by delineating these with full statements of some discrete form of the
row or a partitioned subset of it. Vande Moortele greatly expands upon my own
brief proposal (in 2011) that Adorno’s “materiale Formenlehre” anticipates
Caplin’s theory of formal functions, and he emphasizes that Adorno’s dialectic
philosophical ideas are inextricably embedded within his material theory of
form.8 Here again we have a quasi-narrative, anthropomorphic interpretation
of formal functions: giving passages from Mahler’s symphonies a “metaphori-
cal voice,” Adorno proclaims: “The Marcato . . . announces: ‘Here I am’ . . .
Mahler’s music ‘calls its forms, as it were, by their names’” (“The Philosopher
as Theorist,” p. 419).
For one last example of the creative application of metaphor and narra-
tive to music, let me mention Pedneault-Deslauriers’s “dominant tunnel” in
Schoenberg’s Verklärte Nacht. The ever-evolving appearances of this potentially
precadential, dominant-prolonging progression not only help to articulate the
form of Schoenberg’s work but also are “transgressive,” like Dehmel’s poem
itself. If readers do not at first grasp the metaphor of “dominant tunnel,” the
light will dawn when they reach Pedneault-Deslauriers’s conclusive statement
about it: “when the cadential dominant enters at measure 393 . . . it will finally
lead to a tonic, and a major one at that—literally, the light at the end of the
tunnel” (“Dominant Tunnels,” p. 367).
I close here with references to several strong recommendations from
the contributors about how Caplin’s theory of formal functions might be
expanded and newly adapted when the goal is to apply the theory to music of
the postclassical, Romantic, and later tonal repertoires. In his chapter about
“sentences” in Schumann’s lieder, Harald Krebs carefully modifies and delimits
the requirements for the Classical sentence, as needed for the qualification

after word 439
of Romantically inspired sentences in Schumann’s songs. Brian Black reveals

that modulating “deflected cadence” transitions in Schubert’s sonata forms end
with a perfect authentic cadence in the subordinate key. Such transitions reach
the new key at the last minute, and with the deliberate effect of surprise. Thus
they do not fit the category of Caplin’s “fusion of transition and subordinate
theme”; rather, they defy the Classical convention in which transitions end on
the new dominant, and so they call for an expansion of the concept of transi-
tion. Finally, in his chapter on formal types and functions in postclassical piano
concerti, Julian Horton challenges both the “ubiquity” of Caplin’s instrumen-
tal “Viennese syntax” and its historical as well as geographical reach. Horton’s
“theoretical preliminaries” include a reconceptualization of Caplin’s threefold
distinction between grouping, function, and type, and an argument for the “flex-
ible application” of terms relative to inter- and intrathematic groupings. As
Horton demonstrates, such modifications of Caplinian theory are demanded by
new syntactic habits, cadential deferrals, novel arrangements of functions, and
the “promiscuous” proliferation of these (e.g., pluralities of continuations) in
postclassical concerti.
Horton’s powerful, manifesto-like conclusion puts the matter of nine-
teenth-century forms squarely on the table: “few authors start from the prem-
ise that a model of syntax needs to be established before formal analysis
can proceed. In short, the project on which Caplin embarks for Viennese
Classicism currently has no nineteenth-century counterpart” (“Formal Type
and Formal Function in the Postclassical Piano Concerto,” p. 118). Will a
treatise on nineteenth-century form, or perhaps on “Romantic form,” that is
comparable to Caplin’s work emerge some day? And could a project of this
kind embrace the “parallel historical agenda”—“the circumstances in recep-
tion history that condition the development of a form in any given genre”—
which Horton so passionately advocates (p. 119)? These are questions that
both Steven Vande Moortele and I have informally raised elsewhere;9 judging
from the multiplicity of revelations about forms in diverse nineteenth-cen-
tury genres that Formal Functions in Perspective offers, the goal of a comprehen-
sive treatise seems daunting. But contributors to this volume have pointed
in promising new directions, and William Caplin’s own past and current
work on “Romantic cadences” will surely play a principal role. The future of
studies on nineteenth-century and later forms awaits those who, like these
authors, are willing to move beyond the notion of Classical formal concepts
as arbiters of the new.
Notes
1. William E. Caplin, “Der Akzent des Anfangs: Zur Theorie des musikalischen
Taktes,” Zeitschrift für Musiktheorie 9 (1978): 17–28.

440 after word
2. Erwin Ratz, Einführung in die musikalische Formenlehre (Vienna: Österreichischer

Bundesverlag, 1951; 3rd. rev. ed.: Universal Edition, 1973).
3. Arnold Schoenberg, Fundamentals of Musical Composition, ed. Gerald Strang and
Leonard Stein (London: Faber and Faber, 1967).
4. William E. Caplin, “The ‘Expanded Cadential Progression’: A Category for the
Analysis of Classical Form,” Journal of Musicological Research 7 (1987): 215–57.
5. James Hepokoski and Warren Darcy, Elements of Sonata Theory: Norms, Types, and
2006).
Music of Haydn, Mozart, and Beethoven (New York: Oxford University Press, 1998);
Caplin, Analyzing Classical Form: An Approach for the Classroom (New York: Oxford
University Press, 2013).
7. The proceedings of this session were published in Music Theory Online 16, no. 2
(2010).
8. See Janet Schmalfeldt, In the Process of Becoming: Analytic and Philosophical Perspectives
on Form in Early Nineteenth-Century Music (New York: Oxford University Press, 2011),
16–17.
9. See Schmalfeldt, In the Process of Becoming, and Steven Vande Moortele, “In Search
of Romantic Form,” Music Analysis 32 (2013): 404–31.

Contributors
Brian Black is an associate professor of musicology at the University of
Lethbridge in Alberta. He has published articles in Schubert durch die Brille,
Intersections, and Intégral and has given papers at the annual meetings of
the American Musicological Society, the Society for Music Theory, and the
Canadian University Music Society.
L. Poundie Burstein is a professor of music theory at Hunter College and

the Graduate Center of the City University of New York. He has also taught at
Queens College, Mannes College of Music, and Columbia University. He is cur-
rently president of the Society for Music Theory.
Andrew Deruchie is a lecturer at the University of Otago, New Zealand. He

holds a PhD in musicology from McGill University and is the author of The
French Symphony at the Fin de Siècle: Style, Culture, and the Symphonic Tradition
(2013).
Julian Horton is a professor of music and head of the Department of Music

at Durham University, and president of the Society for Music Analysis. He was
educated at Trinity College, Cambridge, and has taught at University College
Dublin and King’s College London. He is the author of Bruckner’s Symphonies
(2004) and editor of The Cambridge Companion to the Symphony (2013). In 2012
he was awarded the Westrup Prize for his article “John Field and the Alternative
History of Concerto First-Movement Form.”
Steven Huebner is a professor of music at McGill University. His work focuses

on French and Italian music in the period 1850–1950. Recent essays include:
“Tristan’s Traces,” “Ravel’s Politics,” and “Wagner, Dujardin, and the Origins
of Stream of Consciousness Writing.” He was coeditor of the Cambridge Opera
Journal from 2007 to 2014, and has taught at McGill University since 1985.
Henry Klumpenhouwer is a professor of music theory at the Eastman School

of Music. He was formerly the editor of Music Theory Spectrum.
Harald Krebs is distinguished professor at the University of Victoria, and

former president of the Society for Music Theory. He is the author of Fantasy
Pieces: Metrical Dissonance in the Music of Robert Schumann (1999, winner of the

442 contributors
2002 Wallace Berry Award from the Society for Music Theory) and Josephine
Lang: Her Life and Songs (2007, coauthored by Sharon Krebs), and editor of two
collections of Lang’s songs and, with Dániel Péter Biró, a volume on Bartók’s
string quartets (2014).
Nathan John Martin is assistant professor of music at the University of

Michigan. He coedits Music Theory & Analysis, and his article “Rameau’s
Changing Views on Supposition and Suspension” won the 2014 Outstanding
Publication Award from the Society for Music Theory.
François de Médicis is a professor at the Université de Montréal. He has

coedited Musique et modernité en France, 1900 à 1945 (2006), and has published
articles in Acta Musicologica, Music and Letters, Saggiatore musicale, STM-Online,
Canadian University Music Review, the Enciclopedia della musica (Einaudi), and
in collections of articles published by Symétrie, Vrin, Dohr, the Sorbonne,
l’Harmattan, and Presses de l’Université de Montréal.
Christoph Neidhöfer is an associate professor and chair of the Department

of Music Research at McGill University. He holds undergraduate and master’s
diplomas in music theory, composition, and piano performance from the
Musikhochschule in Basel and a PhD in music theory from Harvard University.
His research focuses on twentieth- and twenty-first-century theory and analysis,
sketch studies, aesthetics of serialism, and tonal counterpoint.
Julie Pedneault-Deslauriers is an assistant professor of music theory at the

University of Ottawa. Her research centers on the late tonal and atonal works
of the Second Viennese School composers and on theories of musical form.
Giorgio Sanguinetti teaches at the University of Rome—Tor Vergata. His

areas of research cover the history of Italian music theory, Schenkerian analy-
sis, form, and opera analysis. As a pianist he has performed solo and in cham-
ber groups. His book The Art of Partimento: History, Theory and Practice (2012)
received the 2013 Wallace Berry Award from the Society for Music Theory.
Janet Schmalfeldt has taught at McGill University and at Yale; she joined the
Music Department at Tufts University in 1995. She is the author of a book on
Alban Berg’s Wozzeck and has published widely on late eighteenth- and early
nineteenth-century music. Her book In the Process of Becoming: Analytic and
Philosophical Perspectives on Form in Early Nineteenth-Century Music (2011) received
a 2012 ASCAP–Deems Taylor Award and the 2012 Wallace Berry Award from
the Society for Music Theory. Her performances as pianist have included solo,
chamber, and concerto music.

contributors 443
Peter Schubert studied theory and conducting with Nadia Boulanger and
holds a PhD in musicology from Columbia University. Currently a professor
at McGill’s Schulich School of Music, he is the author of Modal Counterpoint,
Renaissance Style, and Baroque Counterpoint (with Christoph Neidhöfer). In
Montreal he founded VivaVoce, which has released four CDs of Renaissance,
Romantic, and modern music. He is regularly invited to give lectures and work-
shops on improvisation.
Steven Vande Moortele is an assistant professor of music theory at the

University of Toronto. He is the author of Two-Dimensional Sonata Form: Form
and Cycle in Single-Movement Instrumental Works by Liszt, Strauss, Schoenberg, and
Zemlinsky (2009) and coeditor of the journal Music Theory & Analysis.

Index
An italic t following a page reference indicates a table. An italic f indicates a figure.
accentuation patterns, 198, 209–10 Balthazar, Scott, 252

additive technique, 300, 314n9 Barry, Barbara R., 219n3
Adorno, Theodor W., 411–27; and Basevi, Abramo: contribution of, 252–
Berg, 411, 417, 419, 431n39; and 53; discussions of parlante in Studio
Caplin, 426–27; on Mahler, 417–19, sulle opere di Giuseppe Verdi, 254–57,
420–21, 438; materiale Formenlehre, 7, 259, 262, 265–67, 269, 272, 281–82
411–27, 438; and new musicology, basic idea (main theme), 2–3, 16, 83t;
412–15; as a philosopher, 415–17; compound basic idea, 95
and Ratz, 419–20, 431n42; and Bebel, August, 349, 370n17
Schoenbergian Formenlehre, 419–21, Beethoven, Ludwig van: cadenzas, 112;
426–27; as a theorist, 412–15, 417– concerto forms, 79, 112, 118–19;
23; writing on Beethoven, 412, 417, cyclicism, 157, 158; Schubert com-
421, 423–25, 432n60 pared to, 198, 200–204, 219n3,
—publications: “Analyse und Berg,” 220n11; and Viennese classicism, 1,
421; Ästhetische Theorie, 415, 417, 4, 85, 165; and Viennese syntax, 77
421; Einführung in die Dialektik, —compositions: Piano Concerto no.
415; Einleitung in die Musiksozi- 3, op. 37, 122n39; Piano Sonata
ologie, 417; Mahler: Eine musika- in A Major, op. 2, no. 2, 430n32;
lische Physiognomik, 417; “Vers Piano Sonata in C Major (“Wald-
une musique informelle,” 417; stein”), op. 53, 200–202, 204;
“Wiener Gedenkrede,” 417, 420; Piano Sonata in C Minor (“Pathé-
“Zum Problem der musikalischen tique”), op. 13, 430n32; Piano
Analyse,” 413, 417, 418, 422 Sonata in D Minor (“Tempest”),
“after-the-end” schema, 21, 22, 23, 26, op. 31, no. 2, 106, 411, 433n71,
432n65, 434 436; Piano Sonata in E-flat Major,
Alegant, Brian, 363, 364, 368, 371n35 op. 31, no. 3, 201, 209; Piano
athematic music, 377–78 Sonata in F Minor, op. 2, no.
Auerbach, Ida (née Coblenz), 349, 1, 201, 209, 220n11, 221n20,
370n12 316n27, 406n12, 416; Piano
Auerbach, Leopold, 370n12 Trio in E-flat Major, op. 1, no.
1, 34n26; Symphony no. 1, 6–7,
Bach, Carl Philipp Emanuel, 21 295–316, 436, 437–38; (Caplin
Bachofen, Johann: Das Mutterrecht, 349 analysis), 7, 300, 302–3, 304, 306,
Bailey, Walter, 347 308, 313–14, 437; (Hepokoski–
BaileyShea, Matthew, 225, 226, 227, Darcy analysis), 7, 299–304, 305,
237–38, 243, 251n26 308, 313, 437; (Mengelberg’s

446 index
Beethoven, Ludwig van—(cont’d) Caplin, William E.: career and con-

performance), 302–3, 308, 309, tributions, 1–5, 434–36; terms and
314n12, 315nn14–16; Symphony categories (See under specific entries);
no. 3 (Eroica), 7, 220n11, 300, publications: “Der Akzent des
412, 423–25, 427; Symphony no. Anfangs,” 325; Analyzing Classical
5, 316n27, 367; Symphony no. 9, Form, 436; Classical Form, 1–2, 4, 411,
300, 367 435, 436; “The ‘Expanded Caden-
Bellini, Vincenzo, 253; Norma, 255 tial Progression,’” 435; “Structural
Berg, Alban, 378, 411, 417, 419, 421, Expansion in Beethoven’s Sym-
431n39; Wozzeck, 434 phonic Forms,” 314n9; “Theories
Berlioz, Hector, 157, 158, 295, 304 of Harmonic-Metric Relationships
Berry, Wallace, 340n30 from Rameau to Riemann,” 434
bifocal close, 69n12 Carraud, Gaston, 158
Black, Brian, 6, 165–97, 199, 437, Carter, Tim, 69n8
439 Chambellan, Charles, 158
Boito, Arrigo, 283 changing-note schema, 305, 316n23
Bonds, Mark Evan, 77 Chausson, Ernest, 123, 156
Borchmeyer, Dieter, 368 Cherlin, Michael, 347, 348, 355, 357,
Borio, Gianmario, 421 363, 365t
Boyce, William, 77 Chopin, Frédéric, 79; Piano Concerto,
Brahms, Johannes, 7, 124, 199, 218 op. 11, 11, 95, 98–99, 111; Piano
—compositions: “Beherzigung,” op. Concerto, op. 21, 103, 105–7, 111;
93a, no. 6, 405n2; “Nachklang,” Polonaise-Fantasy, op. 61, 435
op. 59, no. 4, 332; Piano Trio in Chusid, Martin, 206
B Major, op. 8, 341n36; “Regen- Clark, Suzannah, 197n32
lied,” op. 59 no. 3, 332; Sym- Clausius, Katharina, 68n4
phony no. 1, 367; Violin Sonata Clementi, Muzio, 7, 77, 320–25; com-
in G Major, op. 78, 332–38, positions: Piano Sonata in D Major
341nn37–38, 342n45 (“La Chasse”), op. 16, 321; Piano
Brendel, Alfred, 218, 222n7 Sonata in E-flat Major, op. 23, no.
Bruckner, Anton, 124; Symphony no. 3, 3, 320–21; Piano Sonata in F Minor,
367; Symphony no. 8, 367 op. 13, no. 6, 339n7; Piano Sonata
Budden, Julian, 252, 290 in G Major, op. 40, no. 1, 321–25;
Burstein, L. Poundie, 5, 11–36, 437 Piano Sonata in G Major, WO 14,
339n15
cadence, cadential function: cadential closed structures (structures closes),
progressions, 340n27; content vs. 378–80
function of, 18–19; continuation— closing parallelism, 363, 367, 371n35
cadences, 12–14; as a grouping, 82, coherence, 404, 410n48
83t; Leibowitz’s concepts of, 382–85; Cohn, Richard, 196–97nn28, 30,
syntactic function and prolongation, 409n39
103–10. See also deflected-cadence collapse (Zusammenbruch), 417, 418–19,
transitions; expanded cadential 423
progression Cone, Edward T., 215, 218, 221n23,
cadenza doppia (double cadence), 334– 222n30
37, 341nn41–42 conflation (of form and cycle), 78, 80,
cadenzas, 111–12, 126, 159n6 85, 112, 115, 118

index 447
continuation, continuation function, Deruchie, Andrew, 6, 123–61, 436

2, 3; as a grouping, 82, 83t; markers development, 200–205
of, 3 developmental rotation, 334, 339n16
continuation⇒cadence, 12–14, 28 developmental sequences, 198, 199–209
core, 14–15, 200–206, 325, 340n20 Dies, Albert Christoph, 31, 36n37
Coren, Daniel, 221n15 D’Indy, Vincent, 123, 124, 156–58,
Cramer, Johann Baptist, 78, 79, 95; 160n21
Piano Concerto, op. 10, 99–100, distributions, 380
102–3; Piano Concerto, op. 26, 103, Döhring, Sieghart, 56
104; Piano Concerto, op. 38, 103, dominant tunnels, 7, 351–68, 438
104; Piano Concerto, op. 51, 103, Donizetti, Gaetano, 256; Lucrezia Borgia,
104 257–59, 272
cultural (new) musicology, 412–14 double cadence (cadenza doppia), 334–
cycle grouping, 81t, 82 37, 341nn41–42
cyclicism and cyclic form(s), 123–61; double canons, 373, 392, 395–98, 405n2
definitions of, 124–25; D’Indy’s double return, 319–20, 339n8
views on, 157–58; Franck’s asso- double-function forms. See two-dimen-
ciation with, 123, 124, 156–58; in sional forms
postclassical concertos, 112–18; and Durchbruch, 417, 419, 430–31n35
socio-political change, 156; sonata Dussek, Jan Ladislav, 77, 78, 79, 86,
form interleaved with, 125–26 95, 110, 111; compositions: Piano
Czerny, Carl, 119, 122n39, 206, 208–9, Concerto, op. 14, 5, 86–88, 99–100,
221n20 101; Piano Concerto, op. 29, 90,
92; Piano Concerto, op. 49, 90, 93,
Da Ponte, Lorenzo, 37, 69n7 94–95; Piano Concerto, op. 70, 90,
Dahlhaus, Carl: Caplin influenced by, 94–95
1, 4, 411, 434, 435; and critique Dvořák, Antonín, 124
of virtuosity, 78; on Schoenberg’s
Verklärte Nacht, 347; on Schubert’s Elgar, Edward, 124
sonata forms, 195n13, 196n19; on Eliot, T. S.: Wasteland, 371n35
Tempest Sonata, 433n71 Engels, Friedrich, 349
Dante Alighieri: Inferno, 318 Erfüllung (fulfillment), 417, 419, 423
Danuser, Hermann, 420, 431n42 exocyclic form, 118
Darcy, Warren. See Hepokoski, James, expanded cadential progression: Cap-
and Warren Darcy lin’s category of, 1, 7, 348, 351; role
Debussy, Claude, 124, 158; String Quar- in subordinate-theme group, 209;
tet, 158 used by Beethoven, 306; used by
deflected-cadence transitions, 165–97; Dussek, 101; used by Hummel, 89;
in Schubert’s early works, 177–84; used by Mozart, 38, 52, 60; used by
in Schubert’s later works, 184–93; Schoenberg, 348, 351, 355, 357, 360,
stylistic significance of, 172–76. See 368; in works by Haydn and Webern,
also cadence 409n34. See also cadence, cadential
deformation (concept), 32n2, 60, 118 function
Dehmel, Richard: and Ida Auerbach,
349, 370n12; as a poet, 347, 369n5, Fallon, Daniel, 145
437; “Verklärte Nacht,” 7, 345–50, Fauré, Gabriel, 124
358, 360, 367, 368, 438 Fickert, Auguste, 350

448 index
Field, John, 78, 79, 86; compositions: galant style, 218, 328
Nocturne no. 12, 112; Piano Con- Gasparini, Francesco: L’armonico pratico
certo no. 1, 86, 87; Piano Concerto al cimbalo, 341nn41–43
no. 3, 111; Piano Concerto no. 7, genre (grouping level), 81t, 82
111, 112, 122n34 Gjerdingen, Robert, 4, 21, 23, 305,
folk-song sentence structure, 227, 228, 316n23, 328
230, 250n16 Gossec, François-Joseph, 21
form: relation to material and content, Gossett, Philip, 1
415–16; and whole vs. part, 416–17 Griesinger, Georg, 36n37
formal dissonance, 12–21, 33n17; at groupings, grouping structures, 80–85;
movement beginning, 16–21; in and closed structures, 379–80; inter-
movement middle or end, 12–15 and intrathematic levels, 81, 83t,
formal function, form-functional the- 112; large-scale, 81t, 83t; levels and
ory: and Adorno’s materiale Formen- hierarchy, 81t, 82, 83t; symmetrical,
lehre, 426–27; beginning-middle-end 380; uniform grouping patterns,
paradigm, 80–81; definitions and 198, 209–10, 218, 380
terminology, 2–4, 5, 80, 439; exten- Gut, Serge, 124
sion of, 225; formal categories as,
419–21; fusion of, 172; generalized Haimo, Ethan, 371nn24, 29
vs. level-specific, 82–83; hierarchical Hart, Brian, 124
mobility, 82–83; inter- and intrath- Haydn, Joseph: and Beethoven, 295,
ematic functions, 81; locations and 299, 304; formal function in sym-
time spans, 11–12; orientation of, phonies, 5, 11–36; musical humour,
2–3, 313–14; perspectives on, 4–5; 14; narrative or programmatic pro-
relationship of grouping, func- gression, 26, 30–31; operas, 225; use
tion, and type, 2–3, 80, 83–84, 439; of additive technique, 314n9; use
for serial compositions, 377–80; of Quiescenza, 21, 22–23; and Vien-
sonata-form transitions in, 166–67; nese syntax or classicism, 31, 37, 77,
top-down vs. bottom-up approaches, 165
420–21; and whole/parts relation- —compositions: Armida, 71n35;
ship, 420, 424–25 Piano Sonata in E-flat Major,
Formenlehre (German/Schoenbergian Hob. XVI:49, 409n34; Piano
Formenlehre), 1–2, 4, 7, 77, 125, 377– Sonata in E-flat Major, Hob.
78, 411–33 XVI:52, 22, 34n26; String Quartet
form-functional conflicts, 28–30 in G Major, Hob. III:41 (op. 33,
four-bar phrase tyranny, 215 no. 5), 18; Symphony no. 7, 31;
Franck, César, 123, 124, 156–58; com- Symphony no. 8, 22; Symphony
positions: Piano Quintet in F Minor, no. 22, 31; Symphony no. 26 in D
156; Piano Trio in F-sharp Minor, Minor (“Lamentatione”), 13, 31;
124; Prélude, aria et final, 156; Prélude, Symphony no. 28, 31; Symphony
choral et fugue, 156; String Quartet no. 36, 22; Symphony no. 39, 22;
in D Major, 156–57, 158; Symphony Symphony no. 43, 22; Symphony
in D Minor, 156; Violin Sonata in A no. 58, 22; Symphony no. 59
Major, 156 (“Feuersymphonie”), 12–14, 22,
Frisch, Walter, 347, 357, 359, 363–64, 437; Symphony no. 61, 23, 24,
369n7 35n31; Symphony no. 62, 22;
Furtwängler, Wilhelm, 315n14 Symphony no. 64 (“Tempora

index 449
mutantur”), 18–21; Symphony Humperdinck, Engelbert: Hänsel und

no. 65, 16–18, 26, 28–31; Sym- Gretel, 250n16
phony no. 73, 22; Symphony no. Hunter, Mary, 56
80, 14–15; Symphony no. 81, hypermeasures, 215, 221n23
21–27, 27t, 35n33; Symphony
no. 90, 18–19; Symphony no. 94, initiating functions, 16, 166, 172, 193,
22; Symphony no. 97, 295, 304; 432n65
Symphony no. 98, 22; Symphony inter- and intrathematic functions, 81,
no. 101, 22 83t, 132
Heartz, Daniel, 33n9, 67, 295 inter- and intrathematic groupings, 81t
Hegel, G. W. F., and Hegelian philoso- intermediary structures, 382–85
phy, 415, 427 interthematic events, 348
Henselt, Adolf, 79 intrathematic proliferation, 80, 85,
Hepokoski, James, 33n10, 121n26, 284 112–14, 118
Hepokoski, James, and Warren Darcy intrinsic formal function, 2–3
(Elements of Sonata Theory): on
Beethoven’s Symphony no. 1, 7, Jackendoff, Ray, 209
299–304, 301t, 305, 308, 313, 437; Jalowetz, Heinrich, 377
on closing section location, 110, Jenkins, Chadwick, 317, 338n3
121n27; dialogue/debate with Cap- Johnson, Julian, 414
lin, 5, 33n10, 110, 436; on modulat- Jonas, Oswald, 341n44
ing caesura-fill, 340n25; on Mozart Jones, Timothy, 138
concerto-sonata syntheses, 79; on Jullien, Adolphe, 158
rotation, 114, 120n14; on transitions
in classical practice, 166, 172, 194n3; Kahn, Erich Itor, 378
on trimodular block, 340n18; Type Kalkbrenner, Friedrich, 78, 79
2 sonatas, 319, 339n11, 339nn14–16; Katastrophe (category), 417
Type 3 sonatas, 321, 339n11 Keiser, Reinhard, 218
Herz, Henri (Heinrich), 79 Kerman, Joseph, 253–54, 265, 278–79,
Heuberger, Richard, 345 282
hexatonic systems, 388–99, 392, 398, Key, Ellen, 350
402–3, 409n39; “Ode-to-Napoleon” Kimbell, David, 253, 278
hexachord, 388 Klumpenhouwer, Henry, 6–7, 295–316,
Hinrichsen, Hans Joachim, 194n1 437–38
Hofmannsthal, Hugo von, 347 Koch, Heinrich Christoph, 35n28,
Höll, Hartmut, 251n31 36n35
Holtmeier, Ludwig, 414 Krämer, Laura, 328
homecoming metaphor, 317–19, 338n3 Kramer, Richard, 331, 340n31, 341n34
Homer: Odyssey, 317–18, 325, 334, Krebs, Harald, 6, 225–51, 438
338n3 Kretschmar, Hermann, 295, 300, 303–4
Horton, Julian, 5–6, 77–122, 439
Hoyt, Peter, 35n30, 319 Lalo, Edouard, 124
Huebner, Steven, 6, 252–92, 437 Leibowitz, René, 4, 373–405, 436–37;
Hummel, Johann Nepomuk, 79, on coherence, 410n48; as a com-
121n21, 122n29; Piano Concerto, poser, 373, 375, 377; links with
op. 89, 90, 91; Piano Concerto, op. Caplin, 7, 379–80, 438; links with
113, 88–89 Schoenberg, 377, 392, 395, 437; and

450 index
Leibowitz, René—(cont’d) loose structures: Caplin’s concept

Rufer, 377–78; on serial functions, of, 226, 227, 384; and Lied form,
378–79; as a theorist, 377–78; and 409n33; and lyric prototype, 227;
Webern, 373, 377, 378 in serial works, 385–88, 398, 401–2;
—“Traité de la composition avec and subordinate-theme function, 3,
douze sons,” 7, 373, 377, 378 16, 85, 100
—Trois poèmes de Pierre Reverdy, op. lyric prototype, 254, 259, 262, 264, 277
92, 7, 373–76; first movement
(“Son de cloche”), 373, 374, 380– Macdonald, Claudia, 113
82, 384–89, 390–91t, 393–94f; Mahler, Gustav, 124, 417–19, 420–23,
second movement (“Air”), 373, 426, 430–31n35, 438; Symphony
375, 392, 395–98; third move- no. 1, 431n35; Symphony no. 3,
ment (“Soleil”), 375–76, 398, 419, 430n34; Symphony no. 6, 419,
401–3; large-scale (overall) form, 430n34; Symphony no. 9, 418
388–91, 399t Mak, Su Yin, 219n3
Leichtentritt, Hugo: Musikalische For- Marnold, Jean, 158
menlehre, 431n44 Marston, Nicholas, 317
Lekeu, Guillaume, 123, 156 Martin, Nathan John, 5, 37–73, 225,
Lerdahl, Fred, 209 226, 238, 437
Levi, Primo: La tregua, 319 Marx, Adolf Bernhard, 119, 122n39,
Lewin, David, 354–55, 359, 409n39 206, 208–9, 411, 420
Liliencron, Detlev von, 347 material: Adorno’s concept of, 415–16
Lindeman, Stephan, 79 materiale Formenlehre (material theory of
linkage function, 83t form), 7, 411–33, 438
linkage technique, 341n44 Mattheson, Johann, 218
liquidation, thematic liquidation: in Mayreder, Rosa, 350
Chopin’s music, 107; in Hummel’s McClelland, Ryan, 219n3
music, 88–89; in Schubert’s music, McGill University: School of Music,
328; in serial music, 383–85, 387, 434–36
401–2 McLean, Don, 363, 364, 368, 371n35
Liszt, Franz: career, 78, 80; conflation Médicis, François de, 6, 198–222, 436
of form and cycle, 80, 85, 112, 115, Mendelssohn, Felix, 125, 157, 159n6,
125–26; and cyclicism, 125–26, 217, 222n27; compositions: “Die
144–45, 157, 158; and Saint-Saëns, Nachtigall,” op. 59, no. 4., 405;
125–26, 144–45; two-dimensional Piano Concerto no. 1 in G Minor,
forms, 78, 85, 112, 125, 134 op. 25, 79, 107–10; Piano Concerto
—compositions: Die Ideale, 125; no. 2 in D Minor, op. 40, 79; Violin
Piano Concerto no. 1 in E-flat Concerto in E Minor, op. 64, 159n6
Major, 78; Piano Concerto no. Mengelberg, Willem, 302–3, 308, 309,
2 in A Major, 78, 112, 115–18; 314n12, 315nn14–16
Piano Sonata in B Minor, 112, Messiaen, Olivier, 402
115, 125; Les Préludes, 125; Tasso, Metternich, Klemens von, 317
125 Moscheles, Ignaz, 78, 112, 125; Piano
Litolff, Henry Charles, 79 Concerto, op. 58, 95, 96–97, 112
Lobe, Christian: Katechismus der Kompo- motive, motivo: as formally generative,
sitionslehre, 431n44 120n13; motivic groupings, 81t;
Locke, Ralph, 155 motivic unity, 398, 401–3, 404; in

index 451
parlante texture, 256, 265; in serial 366–68; in Saint-Saëns’s cyclic works

compositions, 401 (per aspera ad astra), 145, 155
Mozart, Wolfgang Amadeus: Beethoven Neidhöfer, Christoph, 7, 373–410, 437,
influenced by, 209, 299, 300, 304; 438
and Czerny, 221n20; and Hummel, neutralization, 383–84, 385, 386,
121n21; operas, 225; piano concer- 408nn30–32
tos, 79–80, 111–12, 122n38; and Neuwirth, Markus, 35–36nn32–34
Schoenberg’s musical prose, 218; new (cultural) musicology, 412–14
Schubert compared to, 210, 215–16; nocturnes, 111, 112–13, 115
sonata-form opera arias, 5, 31–73, nominalism, 421–23
437; use of additive technique,
314n9; and Viennese syntax, 77 Odyssey (Homer), 7, 317–18, 325, 338,
—instrumental works: Piano Con- 437
certo in D Minor, K. 466, 121n21; Ogdon, Will, 405n3
Piano Sonata in C Major, K. 309, ongoing reprise, 317–38; in Brahms’s
82, 84; Piano Sonata in D Major, works, 332–38; in Clementi’s works,
K. 284, 216; Piano Sonata in F 321–25; definitions of, 319–20; vs.
Major, K. 332, 34n26, 220n11; double return (DR), 319; vs. false
Sonata in D Major for Piano or nonnormative recapitulation,
Four Hands, K. 123a/381, 209, 319–20; and retransition, 320; in
221n20; String Quartet in D Schubert’s works, 325–32; vs. Type 2
Minor, K. 421, 166–67, 168–69, sonata, 339n14
172; Symphony no. 40 in G operas: French operas, 295, 304;
Minor, K. 550, 33n17; Symphony Mozart’s sonata-form arias, 31–73;
no. 41 in C Major (“Jupiter”), K. parlante textures in, 252–91; sen-
551, 33n18, 295, 300, 304 tence types in, 225, 227; solita forma
—operas: La clemenza di Tito, 37, (usual form) for ensemble pieces,
45t, 46, 60; Così fan tutte, 38, 252; systematic analytical methods
46, 55–61, 69n7; Don Giovanni, for, 252. See also specific composers
69nn7, 12, 70n22, 73n42, 253; Die Oppenheimer, Paula, 370n12
Entführung aus dem Serail, 37, 38,
42, 43t, 46–55, 61, 67; Idomeneo, Paddison, Max, 428n7
37, 38–43, 42, 44t, 46, 67, 68; Le parallel forms, 114
nozze di Figaro, 37, 69n7, 69n11, Paris Commune, 156
70n22, 253; Die Zauberflöte, 5, 37, Parker, Roger, 290
38, 45, 46t, 60–67, 68, 69n12, 218, parlante, 252–91; definitions of, 252–54;
437 kinds of, 254–55, 256–57; longer
musical prose, 218 spans of, 282–83; and lyric proto-
musique informelle, 421–22 type, 254, 259; in Mozart’s operas,
253; orchestration for, 253, 255, 256,
narrative progression or trajectory: 264–65; phrases in, 265, 266–67;
formal functions combined with, solo voice vs. dialogue, 262; textures
313–14; in Haydn’s symphonies, of, 255–56; in Verdi’s early and mid-
26, 30–31; homecoming metaphor, dle operas, 265–77; in Verdi’s late
317–19, 320, 338n3, 437; narrative operas, 277–91
of denial, 302, 437–38; narrative parlante armonico, 256–57, 259, 262, 264,
of transfiguration, 348, 358–63, 272, 277, 283, 289

452 index
parlante di ripieno, 255 Ratz, Erwin: and Adorno, 419–20, 423,

parlante melodico, 256–57, 259, 269, 272, 431n42; and Caplin, 201, 340n20,
277 411, 435; Entwicklung (continuation
parlante misto, 256 module), 299; funktionelle Formen-
partition, partitioning, 379–80, 407nn6, lehre, 411, 420, 431n42; Kern der
22; repartitioning, 384 Durchführung, 340n20; on Mahler’s
Pedneault-Deslauriers, Julie, 7, 345–72, Symphony no. 9, 418; and Schoen-
437, 438 bergian Formenlehre, 4, 77, 201, 411,
periods, 82, 84, 201, 299–313, 378 423
perpetuum mobile, 430n35 Ravel, Maurice, 125, 158
Pfannkuch, Wilhelm, 347, 357, 363, 365t recessive dynamic, 49, 328, 340n30
Pfitzner, Hans, 315n14 recitative, recitativo, 253, 256, 259
phrase structures: balanced phrases, recitativo obbligato, 256, 259, 265–66
265, 266–67; lyric prototype, 254; recitativo secco, 256
in serial music, 394–97, 395; uni- Reger, Max, 347
form phraseology, 216–18, 222n30, Reicha, Anton, 309
331n26 re-presentation, 82, 83t
piano concertos, 85–118; analytical retransition, 83t, 86t, 219n8, 320
approaches to, 118–19; cadenzas Richards, Mark, 249n13
and virtuoso displays, 79, 111–12; Riemann, Hugo, 120n15, 197n28, 304–
cantabile idiom and nocturne style, 5, 308, 315n21, 316n28, 438
90, 111, 112–13; first-movement Ries, Ferdinand: Piano Concerto, op.
form, 78–79, 85; Mozartean forms, 132, 79
79–80; nineteenth-century perfor- Rilke, Rainer Maria: Stundenbuch,
mance repertoire, 79, 122n38; and 370n12
piano design innovations, 110–11; Rings, Steven, 56, 71n30
postclassical repertoire, 77–122; two- Roccatagliati, Alessandro, 252
dimensional form in, 85; virtuoso Rodgers, Stephen, 225, 226, 227, 233–
characteristics, 78. See also specific 34, 237, 238, 243
composers Roesner, Linda Correll, 115
political movements and events. See rondò (two-tempo rondò), 42, 45, 46t,
socio-political issues 69n16
polyphonization, 382, 407n25 Rönnau, Klaus, 194n1
postcadential function, 328 Ropartz, Guy, 123, 156
Powers, Harold, 252, 253, 265, 269, 272, Rosen, Charles: on Mozart’s operas,
277 37; on transitions, 173, 192, 197n29;
precore, 200, 328 on uniform phraseology, 215,
presentation, 2–3, 12–13, 16, 82, 83t 221n26
proliferation (intrathematic prolifera- Rosengard Subotnik, Rose, 412–13
tion), 80, 85, 112–14, 118 Rossini, Gioachino, 252, 253; Il barbiere
Proust, Marcel: À la recherche du temps di Siviglia, 255, 262; Semiramide, 255
perdu, 138 rotations, rotational form, 35n29, 114,
Puccini, Giacomo, 252, 253, 254, 278 120n14, 321, 339n14; developmental
Puri, Michael, 124, 125, 160n13 rotations, 334, 339n16
Rothstein, William, 110, 215, 249n14,
Quiescenza, 21–26, 34nn22–26, 328, 341n42
332 Rufer, Josef, 377–78, 406n10, 411

index 453
Saint-Saëns, Camille: abandonment of —compositions: Chamber Sym-

cyclic forms, 155–58; cyclic works, phony, op. 9, 363; Erwartung, op.
6, 123–61, 436; eclecticism, 124; 17, 421; Fantasy for Violin and
and Franck, 156; and Liszt, 125–26, Piano, op. 47, 408n26, 409n39;
144–45; parody of cyclicism, 156–57; Five Orchestral Pieces, op. 16,
views on Wagner, 156 421; Four Orchestral Songs, op.
—compositions: Cello Concerto no. 22, 421; Moses und Aron, 377;
1, 124, 126, 128–37, 130–31t, 157; Ode to Napoleon Buonaparte, op.
Cello Concerto no. 2, 157; Piano 41, 488; Pelleas und Melisande,
Concerto no. 2, 124; Piano Con- op. 5, 351, 363; Prelude, op.
certo no. 3, 124; Piano Concerto 44, 408n26; Satire, op. 28, no.
no. 4, 124, 137t, 138, 142–45, 143t, 3, 405n2; String Quartet, op.
149–50, 155; Piano Quintet in A 37, no. 4, 382, 385; String Trio,
Minor, 123; Samson et Dalila, 125; op. 45, 377; Suite for Piano, op.
String Quartet no. 1 in E Minor, 25, 406n11; A Survivor from War-
156–57, 160n21; Symphony no. 2, saw, op. 46, 377; Three Piano
123; Symphony no. 3 (“Organ”), Pieces, op. 11, 421; Variations for
124, 137t, 144–55; Violin Con- Orchestra, op. 31, 395; Verklärte
certo no. 1, 123, 124, 126–29, Nacht, op. 4, 7, 345–68, 437, 438;
127t, 131, 134; Violin Concerto Dehmel’s poem, 345–50; domi-
no. 2, 123; Violin Sonata no. 1, nant tunnels, 350–55; large-scale
124, 137–42, 137t, 139t, 145, 149 form, 347, 363–67, 365t; ninth
Salzer, Felix, 195n13 chord, 348, 351–55, 360, 364,
Samson, Jim, 78 368, 371n24; reviews and criti-
Sanguinetti, Giorgio, 7, 317–42, 437 cism, 345–48; transfiguration nar-
Satie, Erik: Gymnopédie no. 1, 375 rative, 348, 358–63, 366–68; Wind
Schachter, Carl, 295 Quintet, op. 26, 377, 422; Zwei
Schenker, Heinrich, and Schenker- Klavierstücke, op. 33b, 406n11
ian concepts, 3, 120n15, 341n42, —theoretical writings, 77, 201, 350,
341n44, 431n44 435, 438; “Composition with Twelve
Scheu, Josef, 346 Tones,” 377; Fundamentals of Musical
Schickele, Peter, 437 Composition, 419; Models for Beginners
Schiller, Friedrich, 219n3 in Composition, 377, 431n40
Schmalfeldt, Janet, 434–40; on Adorno, Schubert, Franz, 6; Beethoven compared
411–12, 426–27; and Caplin, 5, to, 198, 219n3; Brahms influenced
434–36; and development of form- by, 199; and cyclicism, 125; deflected-
functional theory, 5; on homecom- cadence transitions, 6, 165–97, 199,
ing metaphor, 317; on “one more 437, 439; early sonata forms, 177–84,
time” technique, 58, 195n9; on reca- 195–96n17; “heavenly length,” 198,
pitulations, 325, 339n7; theory of 216–18, 436; Mein Traum, 219n2;
functional transformation, 95 Mozart compared to, 210, 215–16;
Schoenberg, Arnold: cyclicism, 125; ongoing reprises, 7, 325–32; phrase-
form-functional concepts, 377–80; ology, 209–16; sketching methods,
and Leibowitz, 377–78, 392, 395, 222n32; Spohr compared to, 174–75;
437; on motives, 77, 120n13; on surprise modulations, 167–72, 194n8;
musical prose and expression, 217– symphonic sonata forms, 6; thematic
18; and Rufer, 377–78 sequences, 216–18

454 index
Schubert, Franz—(cont’d) Schubert, Peter, 7, 371n24, 373–410,

—compositions: Impromptu in C 437, 438
Minor, op. 90, no. 1, 186; Piano Schumann, Robert: and cyclicism,
Quintet in A Major, D. 667 (“Die 125, 157; and Field, 122n34; on
Forelle/The Trout”), 170–72, Schubert’s works, 198, 217–18; sen-
173, 205; Piano Sonata in A tence structures in vocal music, 6,
Major, D. 664, 184–87; Piano 225–41, 438–39; uniform phraseol-
Sonata in A Major, D. 959, 205, ogy, 221n26
217; Piano Sonata in A Minor, —instrumental works: Phantasie
D. 845 (op. 42), 339n7; Piano (1841), 113; Piano Concerto,
Sonata in B-flat Major, D. 960, op. 54, 5, 79, 112, 113–15; Piano
190–93, 195n11, 217, 435; Sonata no. 1 in F-sharp Minor,
Piano Sonata in C Major, D. op. 11, 221n26; Symphony no. 1
279, 340n25; Piano Sonata in (“Spring”), op. 38, 198
C Major (“Reliquie”), D. 840, —lieder: “An meinem Herzen,” op.
325–32, 437; Piano Sonata in C 42, no. 7 [Ex. 7.5], 230, 231, 234,
Minor, D. 958, 217, 218; Piano 237; “Auf einer Burg,” op. 39, no.
Sonata in D Major, D. 850, 215; 7, 230, 234; “Des Knaben Berg-
Piano Trio in B-flat Major, D. lied,” op. 79, no. 8, 230, 250n22;
898, 194n8, 205; Piano Trio in “Er ist’s,” op. 79, no. 23 [Ex. 7.6],
E-flat Major, D. 929, 186–88, 231, 232; “Erstes Grün,” op. 35,
190, 195n11; Die schöne Müllerin, no. 4 [Ex. 7.10], 236, 237; “Früh-
225, 226, 233–34, 238, 250n23; lings Ankunft,” op. 79, no. 19
String Quartet in A Minor [Ex. 7.7], 231, 232, 250n22; “Ich
(“Rosamunde”), D. 804, 205; wandelte unter den Bäumen,” op.
String Quartet in B-flat Major, 24, no, 3 [Ex. 7.12], 243–45; “Im
D. 112, 177–80; String Quar- wunderschönen Monat Mai,” op.
tet in C Major, D. 32, 195n17; 48, no. 1, 238; “In der Fremde,”
String Quartet in C Major, D. op. 39, no. 1 [Ex. 7.3], 228–29,
46, 195n17; String Quartet in 235; “Lied Lynceus des Türmers,”
D Major, D. 94, 195n17; String op. 79, no 27, 230, 250n22; “Nun
Quartet in D Minor (“Der Tod hast du mir den ersten Schmerz
und das Mädchen”), D. 810, 205; getan,” op. 42, no. 8, 235; “Räth-
String Quartet in G Major, D. 887, sel,” op. 25, no. 16 [Ex. 7.9],
194n8, 205; String Quartet in G 236, 237; “Seit ich ihn gesehen,”
Minor, D. 173, 180–84; String op. 42, no. 1 [Ex. 7.2], 227–28,
Quartet in G Minor/B-flat Major, 234; “Stille Liebe,” op. 35, no. 8
D. 18, 195n17; String Quintet in C [Ex. 7.13], 245–48; ; “Die wan-
Major, D. 956, 205; Symphony in delnde Glocke,” op. 79, no. 17,
B Minor (“Unfinished”), D. 759, 230, 250n22; “Wanderlied,” op.
195n15, 340n25; Symphony in C 35, no. 3 [Ex. 7.4], 229–30, 234,
Major (“The Great”), D. 944, 236; “Warte, warte, wilder Schiffs-
198, 205–9, 217; Symphony in mann,” op. 24, no. 6 [Ex. 7.11],
C Minor (“The Tragic”), D. 238–42; “Widmung,” op. 25, no.
417, 194n7, 199–205, 209–15; Vio- 1 [Ex. 7.8], 231–33, 234, 235–36,
lin Sonatina in G Minor, D. 408, 250n20; “Zigeunerliedchen,” op.
169–70, 173–74, 174t, 196n22 79, no 7, 230, 250n22

index 455
Second Viennese School, 7, 377 Steibelt, Daniel, 78, 79

secondary structures (structures secon- Stein, Erwin, 411
daires), 385–88 Sterndale Bennett, William, 78, 79;
Senici, Emanuele, 291 Piano Concerto no. 3, 95
sentence themes or structures: dra- Strauss, Johann, 218
matic expression, 237–38; in folk Strauss, Richard, 347; Symphonia domes-
songs, 227, 228, 230, 250n16; manic tica, 158
sentence, 238; mini-sentence, Stravinsky, Igor: Firebird, 138
251n29; monofold sentence, Strindberg, August: The Father, 349
249n13; normative vs. nonnorma- structures closes, 378–80
tive, 243; quatrains, 249n13; in structures intermédiaires, 382–85
Schumann’s lieder, 225, 227–33; in structures secondaires, 385–88
serial music, 378–79 subjectification, 315n21
serial (twelve-tone) technique, 373– subordinate-theme functions, 3, 194n2,
405; closed structures, 378–80; dis- 424–25; transition/subordinate theme
tributions, 382, 397; form-functional fusion, 166–67, 172–73, 194nn2–3
categories, 377–80; heritage of classi- Suspension, 417, 418, 419, 422, 430n35
cal form in, 378; Leibowitz’s writing Swift, Richard, 347, 355, 357, 363, 365t
on, 377–78; liquidation and neutral- symmetrical grouping structure, 380
ization, 383–85; polyphonization in, symphonic concerto, 79
382; transpositions, 388 syntax, 77–78, 80
Servières, Georges, 160n21
Sisman, Elaine, 32–33n9, 295, 304 Taruskin, Richard, 302, 315n15
Smith, Peter H., 332, 341nn35–38, 342n45 Taylor, Benedict, 123, 125
socio-political issues: Franco-Prussian Tchaikovsky, Pyotr Ilyich, 124, 156n9;
war, 156; French Revolution, 78, Violin Concerto, 159n6
417; revolutionary movements and Telemann, Georg Philipp, 218
state oppression, 78–79, 317; women thematic integration, 78
in patriarchal societies, 348–50, thematic liquidation. See liquidation
370nn12, 17 thematic logic, 78, 95
sonata form: alternate framework, 25t; thematic sequences, 198, 201–5, 216
concerto-sonata syntheses, 79; func- thematic syntax, 85–103
tions and grouping levels in, 83t; Thomson, Virgil, 315n15
major-key framework, 25t; in piano- top-down vs. bottom-up approaches,
concerto first movements, 79–80; 420–21, 422, 426
postclassical, 118–19; second thematic total form, 81t, 82, 83t
and closing sections, 299–300; sonata– Tovey, Donald Francis, 217, 218, 295
ternary, 60, 71n35; two-dimensional transitions: Caplin’s definition of,
forms, 125–26, 127t; unitary, 79, 113; 166–67, 195n12; in classical sonata
use of sequences in, 198 form, 166–67; deflected-cadence
sonata-ritornello form, 79 strategy, 165–97; and liquidation,
Spinner, Leopold, 377, 378; Piano Con- 383–84; modulations in, 166–72;
certo, 377 structures intermédiaires, 382–85; vs.
Spitzer, Michael, 429n14 subordinate-theme function, 166–
Spohr, Ludwig: String Quartet in G 67; transition/subordinate theme
Minor, op. 4, no. 2, 174–76 fusion, 166–67, 172–73, 194n2. See
Starr, Daniel, 407nn22, 25 also deflected-cadence transitions

456 index
trimodular block (TMB), 340n18 versi lirici, 159, 292n15

twelve-tone music. See serial (twelve- Viennese syntax, 77–80
tone) technique Vierhebigkeit, 215, 309
two-dimensional (double-function) Viertaktigkeit, 215
forms, 85, 112–18, 125–37, 127t, Vieuxtemps, Henri, 125
144, 363–64 Vilain, Robert, 369n5
Type 2 sonatas, 69n14, 319, 320–21, virtuosity, 78, 79
339nn14–16
Type 3 sonatas, 79, 321 Wagner, Richard: attitudes to, 156, 256;
“Quadratur der Tonsatzes,” 221n23;
Ulysses. See under Homer sentence structures of, 225, 226, 238,
uniform phraseology, 216–18, 222n30, 251n26; Siegfried, 251n26; Tristan und
331n26 Isolde, 367–68
unitary sonata form, 79, 113 Walter, Bruno, 363
unité motivique (motivic unity), 398, Wanhal, Johann Baptist, 21
401–3, 404 Weber, Carl Maria von, 79
universal vs. particular, 415, 416, 421 Weber, Max, 416
Webern, Anton: form-functional con-
Vallas, Léon, 157 cepts, 377–80, 411, 438; and Leibow-
Vande Moortele, Steven: on Adorno’s itz, 373, 377, 378; on Verklärte Nacht,
materiale Formenlehre, 7, 411–33, 438; 347
on exocyclic development, 118; and —compositions: Cantata no. 1,
Schmalfeldt, 434, 439; on Schoen- op. 29, 379; Concerto for Nine
berg’s Formenlehre, 351, 363; on two- Instruments, op. 24, 378, 406n12,
dimensional forms, 85, 112, 125, 408n27; Entflieht auf leichten Käh-
126, 144, 363–64 nen, op. 2, 405n2; Variations for
Verdi, Giuseppe: early and middle Piano, op. 27, 380, 409nn34–35;
works, 265–77; late works, 277–91 “Wie bin ich froh!” op. 25, no.
—operas: Aida, 283, 289, 290; Alzira, 1, 401
272; Un ballo in maschera, 255, Webster, James: on double return,
272–77, 278; Falstaff, 269, 277, 339n8; on Mozart’s opera arias, 37,
278, 283–91; I Lombardi alla prima 56, 68, 88; on Schubert’s sonata
crociata, 254, 255, 257, 264; Mac- forms, 180, 196n19, 199
beth, 268; Nabucco, 267, 283; (Abi- Wellesz, Egon, 347, 355
gaille–Nabucco duet), 183, 256, Weltlauf (course of the world), 417, 419,
257, 259–65, 272, 290, 437; Otello, 430n35
269, 277, 289, 290; act 3 trio, 278, Wheeldon, Marianne, 124, 158
283; (drinking scene and brawl whole vs. parts, 415, 420, 422–23,
[act 1]), 253, 278–82, 289, 290; 424–25
Rigoletto, 255, 268, 283, 290; Simon Wild, Jonathan, 405
Boccanegra: (Amelia–Simon duet), Will, Richard, 26n29
283, 289; (Fiesco–Simone duet), Winter, Robert, 69n12
269–72; La traviata, 253, 254, 255, Witkowski, Georges-Martin, 157
265–68, 278, 283; Il trovatore, 282; Wollenberg, Susan, 180, 194n1, 219n3
Les vêpres siciliennes, 255
Verklärung, 358, 371n36 Zusammenbruch, 417, 418–19, 423

5.8125 × 9.25 SPINE: 2 FLAPS: 0
in Perspective
Formal Functions
A mong the more striking developments in contemporary North American
music theory is the renewed importance of musical form (Formenlehre).
Formal Functions in Perspective presents thirteen studies that engage with musi-
cal form in a variety of ways. The essays, written by established and emerging
scholars from the United States, the United Kingdom, Canada, and the
European continent, run the chronological gamut from Haydn and Clementi
to Leibowitz and Adorno; they discuss Lieder, arias, and choral music as well as
symphonies, concerti, and chamber works; they treat Haydn’s humor and Saint-
Saëns’s politics, while discussions of particular pieces range from Mozart’s arias
to Schoenberg’s Verklärte Nacht. Running through the essays and connecting
them thematically is the central notion of formal function.
“Formal Functions in Perspective is a vital contribution to the body of scholarship
on musical form—and, in particular, to the recent work incorporating formal
function theory. ”
Formal Functions
—edward jurkowski , University of Lethbridge, Alberta
Contributors Brian Black, L. Poundie Burstein, Andrew Deruchie, Julian Horton,
Steven Huebner, Henry Klumpenhouwer, Harald Krebs, Nathan John Martin,
nathan j oh n marti n
juli e pedn eault-deslauri ers , an d
edited by steven van de moortele ,
François de Médicis, Christoph Neidhöfer, Julie Pedneault-Deslauriers, Giorgio
Sanguinetti, Janet Schmalfeldt, Peter Schubert, Steven Vande Moortele
steven vande moortele is assistant professor of music theory at the
in Perspective
University of Toronto. julie pedneault-deslauriers is assistant
professor of music at the University of Ottawa. nathan john martin is
assistant professor of music at the University of Michigan.
Cover photo: srapulsar38/123RF.com
Cover design by Ann Weinstock
668 Mt. Hope Avenue, Rochester, NY 14620-2731, USA

P.O. Box 9, Woodbridge, Suffolk IP12 3DF, UK
edited by steven van de moortele ,
www.urpress.com juli e pedn eault-des laur i ers , an d
nathan joh n marti n

Formal Functions in Perspective: Essays On Musical Form From Haydn To Adorno

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

Formal Functions in Perspective: Essays On Musical Form From Haydn To Adorno

Uploaded by

Copyright:

Available Formats

Formal Functions

Vande Moortele.indd i 9/30/2015 7:50:27 PM

Additional Titles of Interest

Analyzing Atonal Music: Pitch-Class Set Theory and Its Contexts

Analyzing Wagner’s Operas: Alfred Lorenz and German Nationalist Ideology

Aspects of Unity in J. S. Bach’s Partitas and Suites: An Analytical Study

Bach to Brahms: Essays on Musical Design and Structure

A Dance of Polar Opposites:

The French Symphony at the Fin de Siècle:

Music Theory in Concept and Practice

Reviving Haydn: New Appreciations in the Twentieth Century

A Theory of Music Analysis: On Segmentation and Associative Organization

A complete list of titles in the Eastman Studies in Music series

Vande Moortele.indd ii 9/30/2015 7:50:47 PM

Essays on Musical Form

Edited by Steven Vande Moortele,

Vande Moortele.indd iii 9/30/2015 7:50:47 PM

All rights reserved. Except as permitted under current legislation, no

First published 2015

University of Rochester Press

Library of Congress Cataloging-in-Publication Data

Formal functions in perspective : essays on musical form from Haydn to

This publication is printed on acid-free paper.

Vande Moortele.indd iv 9/30/2015 7:50:47 PM

Part One: Theoretical Studies in Haydn and Mozart

2 Mozart’s Sonata-Form Arias 37

Part Two: Nineteenth-Century Taxonomies

4 Saint-Saëns’s Cyclic Forms 123

Part Three: Schubert

6 “Heavenly Length” in Schubert’s Instrumental Music 198

Vande Moortele.indd v 9/30/2015 7:50:47 PM

Part Four: Text, Texture, and Form

8 Parlante Talk: Texture and Formal Function in the

Part Five: Analysis and Hermeneutics

10 Laborious Homecomings: The “Ongoing Reprise” from

Part Six: Schoenberg and Beyond

12 Form and Serial Function in Leibowitz’s

13 The Philosopher as Theorist: Adorno’s materiale Formenlehre 411

Notes on Contributors 441

Vande Moortele.indd vi 9/30/2015 7:50:47 PM

Toronto; Ottawa; New Haven

Vande Moortele.indd vii 9/30/2015 7:50:47 PM

Few writers have contributed as much to the revival of Formenlehre in current

Vande Moortele.indd 1 9/30/2015 7:50:47 PM

Vande Moortele.indd 2 9/30/2015 7:50:47 PM

in contrast, depends on markers internal to the unit in question: “This is

Vande Moortele.indd 3 9/30/2015 7:50:47 PM

comprising interrelated schemata (in something like Robert Gjerdingen’s

Vande Moortele.indd 4 9/30/2015 7:50:47 PM

Vande Moortele.indd 5 9/30/2015 7:50:47 PM

of Mozart’s piano concerti and highlights the implications of romantic con-

Vande Moortele.indd 6 9/30/2015 7:50:48 PM

the commission of an error and its subsequent correction in the woodwind

Vande Moortele.indd 7 9/30/2015 7:50:48 PM

Vande Moortele.indd 8 9/30/2015 7:50:48 PM

Vande Moortele.indd 9 9/30/2015 7:50:48 PM

Vande Moortele.indd 11 9/30/2015 7:50:48 PM

undermined by others—for instance, when a sense of beginning that is firmly

Deviating Middles and Ends

Vande Moortele.indd 12 9/30/2015 7:50:48 PM

Example 1.1. Haydn, Symphony no. 59 in A Major (“Feuersymphony,” 1768),

has described at length, the continuation⇒cadence of a typical sentence

Vande Moortele.indd 13 9/30/2015 7:50:48 PM