ere ed ) =

:
Wao @ (@) gee
See
ees
ast:
i}
ite |
fe ; mei cone =
a H oe

: DAVID DAMSCHRODER

= HASRCHMUOBNEYRT IN

145 146 183 154 158 -16l ign

 One of Western music’s great harmonists,
Franz Schubert created a wondrous and treasured
body of music that has retained its fascination to
this day. His innovative harmonic practice has
been a topic oflively discussion among analysts
_for generations. Harmony in Schubert presents
_ a fresh approach, yielding insightful readings of
a large and varied range of excerpts, as well as
‘readings of fifteen complete movements spanning
Schubert’s chamber, choral, orchestral, piano,
and vocal output. Damschroder reformulates the
apparatus for Roman-numeral harmonic analysis,
_ integrating his own speculations with various
strands of historical analytical thought, including
Schenkerian principles-and historical Perspectives.
In addition, he juxtaposes his readings of complete
movements by Schubert with discussions of how
they haye been interpreted by other Schubertian
analysts. The book sets a new direction for the
future of music analysis, proposing innovative
improvements on existing methodologies.
Harmony in Schubert

One of Western music’s great harmonists, Franz Schubert created a

wondrous and treasured body of music that has retained its fascina-
tion to this day. His innovative harmonic practice has been a topic
of lively discussion among analysts for generations. Harmony in
Schubert presents a fresh approach, yielding insightful readings of a
large and varied range of excerpts, as well as readings of fifteen com-
plete movements spanning Schubert’s chamber, choral, orchestral,
piano, and vocal output. Damschroder reformulates the apparatus
for Roman-numeral harmonic analysis, integrating his own specula-
tions with various strands of historical analytical thought, including
Schenkerian principles and historical perspectives. In addition, he
juxtaposes his readings of complete movements by Schubert with
discussions of how they have been interpreted by other Schubertian
analysts. The book sets a new direction for the future of music analy-
sis, proposing innovative improvements on existing methodologies.

DAVID DAMSCHRODER is Professor of Music Theory at the University

of Minnesota School of Music, where he teaches a range of courses
devoted to the analysis of eighteenth- and nineteenth-century music,
the repertoire that also serves as the focus for his scholarly writings
and performance activities on fortepiano. His previous books include
Thinking About Harmony: Historical Perspectives on Analysis, Music
Theory from Zarlino to Schenker, Listen and Sing, and Foundations of
Music and Musicianship.
 { beim auaet eye 1) “fe
ddl : of tefl!) daw) ad
7 : | adliul ) gebanreey MO
iw ewig 6 Stay
ia i pies prides. Lawormngyes
=
LEE Tl & eqguilen, co lowe ery
Li ei ed _ ana aul y erated’ AR
ory «l eqiawedm CAniirasnatl Fil
jie ye oor @o! GMI ep Cony Cente, :
4 pty - plage Sage \exreutandIp, ole int io i
ut Jie 2 opprey errs:
Mo bawink @) cnmiereey Gelpipay &
dav ahant cai wi atone ate pea
ant Jalon I quan att eal Aiit
Tag (Haiecehooey
h Qialttinw0. derey
& . ad .

Acmeyteld
seitwipuredl seatitenant
mer Hy QO @ EnAan! et whe ill
i ett Himeewiin hie alieagh
wives (Ube 10 wat aM te
whwow
et auorpeey ol Caiholegat ya,
1) Ga) ah ewig \i )
Paialon Meare int aM
ou

yd nae
xe -
7 7
Harmony in Schubert

DAVID DAMSCHRODER
The University of Minnesota

Nyack College - Baliey Library

One South Blvd.
Nyack, NY 10960

] CAMBRIDGE
BY UNIVERSITY PRESS
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,
Sao Paulo, Delhi, Dubai, Tokyo, Mexico City

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521764636

© David Damschroder 2010

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.

First published 2010

Printed in the United Kingdom at the University Press, Cambridge

A catalogue record for this publication is available from the British Library

ISBN 978-0-521-76463-6 Hardback

Cambridge University Press has no responsibility for the persistence or

accuracy of URLs for external or third-party internet websites referred to in
this publication, and does not guarantee that any content on such websites is,
or will remain, accurate or appropriate.
Contents

Preface [page vii]

PART ONE METHODOLOGICAL ORIENTATION

Harmonic progression [3]

Tonicized keys and chromatic chords [3]
Context determines function [8]
I-II [15]
IV-V_ [21]
Peculiar juxtapositions [26]
Unfurling of $chords [31]
“Ungeduld” from Die schéne Miillerin (D. 795/7) [36]

Linear progression [41]

Parallel progressions [41]
The ascending 5-6 sequence [44]
Circular progressions [52]
Chromatic lower-third and upper-third chords [56]
Piano Sonata in D Major (D. 850), movement 3, Trio [60]

Common prolongations and successions [65]

Tonic prolongation [65]
Dominant prolongation [71]
I-V_ [75]
I-IV [80]
-V_ [85]
Variation on a Waltz by Diabelli (D.718) [89]

Chords on bIJ, on HI, and from the parallel key [92]

III [92]
Modal mixture [101]
The lowered supertonic chord [106]
“Die Einsiedelei” (D. 337) [113]
“Am Flusse” (D. 160) [116]
“Erster Verlust” (D. 226) [121]
Landler in G# Minor (D. 790/6) [126]
vl Contents

PART TWO MASTERPIECES

“Ganymed” (D. 544) [135]

with Lawrence Kramer and Suzannah Clark

Quintet in A Major (“Trout? D. 667), movement 1 [149]

with David Beach

Symphony in B Minor (“Unfinished,’ D. 759), movement 1 [159]

with Richard Taruskin

Piano Sonata in A Minor (D. 784), movement 2 [179]

with Robert S. Hatten

“Die junge Nonne” (D. 828) [191]

with David Kopp

10 Four Impromptus (D. 899) [201]

with Charles Fisk

11 “Auf dem Flusse” from Winterreise (D.911, No.7) [236]

with David Lewin

12 Piano Sonata in Bb Major (D. 960), movement 1 [245]

with Richard L. Cohn

Epilogue [264]

Notes and references [267]

List of references to music examples [306]
Select bibliography [307]
Index of Schubert’ works [316]
Index of names and concepts [318]
Preface

How we think about harmony, what we write or say to convey our thoughts,
is a fundamental component of musicianship. For centuries the elucidation
of harmonic theory has been a compelling field of inquiry and a vital aspect
of musical training. The conscientious and diligent striving of numerous
thinkers over many generations has produced a wide array of method-
ologies and insights. The topic remains engaging today, its importance
undiminished for those who take a keen interest in tonal music.
Harmony in Schubert is addressed to musicians who strive to attain a
deep understanding of how harmony works, to seekers who are eager to
explore sophisticated speculations that may shed new light on a treasured
body of music. The goal of this study is to foster more enlightened perform-
ance of and more astute listening to music, especially that of Schubert and
his contemporaries.
A music analyst coordinates myriad individual judgments to form
a coherent conception. Much depends upon the ability to hear beyond
surface differences to the essence of things. Developing analytical insight
is a cumulative process, with each new analysis drawing upon the perspec-
tive attained through all previous endeavors. A challenging passage may
eventually yield its secrets when one comes to realize that it extends a more
straightforward formulation whose workings one understands. To maxi-
mize the benefits from comparing related music examples, this study draws
upon a finite body of repertoire: the music of Franz Schubert. By so limiting
our purview the hunches that are essential to sophisticated analysis can be
made with greater assurance, and one may hope that a gradually emerg-
ing perspective, evolving out of close study of hundreds of works from the
same pen, will offer insights that could not be achieved were one to explore
a more extensive and diverse corpus of music. My plan has been to forge
ahead in conceptualization about harmony through a study focused on
Schubert’s music (does a better candidate for such an undertaking exist?),
to share the results of my speculations with fellow seekers in this volume,
and then to extend the scope of what I develop here to a broader repertoire
~ both before and after Schubert - in future publications.
I expect that readers will have a strong interest in and significant prior
viii Preface

exposure to the main trends in tonal music analysis, including Schenkerian

analysis. Because my project seeks to extend the range of what music analy-
sis can accomplish, I am in constant dialogue with existing methodologies,
both historical and contemporary. Though I address many basic issues, I do
so from the perspective of re-orientation, rather than that of initial expo-
sure. For example, when I develop the notion that Ab-C-Eb-F# in the key
of C is an evolved supertonic chord, I will expect that readers already have
dealt with this pitch combination in some way and have noticed its similar-
ity in both construction and function to other chromatic chords, such as
A-C-Eb-F# and A-C-D-F#.
The two key ingredients in my synthesis are the analytical practices of
the eighteenth and nineteenth centuries, extensively documented in my
Thinking About Harmony: Historical Perspectives on Analysis (Cambridge
University Press, 2008; hereafter abbreviated as TAH), and the Schenkerian
perspective, especially that presented in Schenker’s Der freie Satz (Vienna:
Universal Edition, 1935; translation of Oswald Jonas’s 1956 edition by
Ernst Oster as Free Composition, New York: Longman, 1979; hereafter
abbreviated as FC). I draw upon many insights that were in circulation
around Schubert's time, though I inject levels of hierarchical thinking
that entered mainstream analytical practice only as a result of Schenker’s
writings a century later. Likewise I make extensive use of Schenkerian
principles, though my high regard for what some of Schenker’s predeces-
sors accomplished has led me to draw freely from their methods. Thus the
post-Schenkerian perspective that prevails in Harmony in Schubert derives
in part from pre-Schenkerian notions that I resuscitate (despite Schenker’s
rejection or indifference) because I find them indispensable, in part from
Schenker’s writings, and in part from my own speculations. In addition, I
critique an array of Schubert analyses by prominent modern thinkers. By
placing my analyses side by side with theirs, I hope that readers will come to
understand what issues are at stake, how our differing perspectives respond
to them, and why my reaction to the alternative methodologies that I
explore is sometimes unenthusiastic.
Harmony in Schubert is presented in two parts. Readers will want to
work carefully through the first part (Methodological Orientation) before
proceeding to the more challenging second part (Masterpieces). I have
attempted in the early chapters to provide a clear and detailed accounting of
my procedures, terminology, and symbols, which differ from conventional
practice in numerous ways. Once introduced, I assume that readers will
be able to correlate those generally straightforward instances with more
complex formulations in the works examined later.
 Preface

The Methodological Orientation intermixes explorations of numerous

individual topics with analyses of short movements by Schubert. These
materials are arranged in an order that offers both a logical development
of the analytical framework and a gradual increase in complexity, so that
readers may develop or extend their capacities expediently. I rely upon TAH
for grounding in pre-Schenkerian procedures, using the endnotes to alert
readers to enrichment available from that source. (Because it offers numer-
ous English translations and other aids to understanding, these authors
generally will be accessed via TAH rather than directly from their treatises.
TAH supplies full bibliographical data on the authors mentioned.) In the
same manner, I cite FC to clarify concepts derived from Schenker. In almost
all cases, musical scores or their reductions are provided where warranted.
The Masterpieces section (so named in homage to Schenker’s Das
Meisterwerk in der Musik) offers detailed analyses of a range of move-
ments by Schubert. To follow the discussion fully, readers will need to
obtain scores for these works, as well as the publications that contain the
alternative analyses (by authors such as Richard Cohn, David Lewin, and
Richard Taruskin) that I critique. These items should be easily accessible in
North American and British research libraries (a promise that necessitates
restricting our purview to works in English, though my own study has
extended beyond that boundary), and some are available on the Internet.
Though the Methodological Orientation section should be read first and
in the order presented, the chapters of the Masterpieces section, which is
arranged chronologically, may be read selectively or out of order. Some
readers may prefer to begin with the relatively straightforward chapters 8,
9, and 11, then continue with the more challenging chapters 6, 7, and 10
before concluding with chapters 5 and 12.
The juxtaposed analyses presented in the Masterpieces section will
allow readers to assess various contemporary methodologies and to chart
a course for their future analytical efforts. Though presented here in a
high-level formulation, Harmony in Schubert calls into question many
time-honored conventions of lower-level analytical pedagogy. It is a mani-
festo for a top-to-bottom transformation in the way musicians think about
harmony.
Success has been an onerous burden to harmonic analysis. On one hand,
the fact that a series of courses on harmony has become a mainstay of
academic training in music (at least in North America and Great Britain)
means that the curriculum often must accommodate students with weak
backgrounds and limited capacities. Methodologies requiring complex or
subtle thought too often are eschewed, with the unfortunate consequence
Preface

that many fine musicians are never exposed to harmonic analysis at its
best. On the other hand, because other branches of theoretical study such
as species counterpoint and figured-bass realization are often neglected,
analytical procedures are more dependent on harmonic explanations for a
range of musical phenomena than they ought to be. In my view harmonic
theory as understood by most musicians is not a suitable starting point
for advanced work. Consequently I cleared the slate when I began writing
Harmony in Schubert, readmitting concepts and procedures only upon
careful consideration of various options. Among the most notable casual-
ties in this process have been mainstays of elementary pedagogy such as
applied dominant chords (e.g., V/V), the use of capital and lower-case
Roman numerals to designate chordal quality (e.g., C Minor: i ii°° V’ i),
figured-bass notation to the right of a Roman numeral to indicate chordal
inversion (e.g., I°), chord nicknames (e.g., “French” “German,” “Ttalian,’
“Neapolitan”), and Roman-numeral labels for the internal chords of circu-
lar, parallel, and sequential progressions (e.g., 1 1V VII HI VIII V I). Why
must these traditional symbols and procedures be sacrificed? Because they
rest on faulty notions of what harmony is, how harmony absorbs chromati-
cism, how voice leading interacts with harmony, and how a harmony may
evolve as it is prolonged. Though I appreciate that my proposed reforms
will widen the gap between musical speculation and elementary pedagogy
as currently practiced, certainly one should not hold back advances in
the field to accommodate its initiates. Instead I call upon music theory’s
pedagogical branch to develop the resources to align itself better with the
speculative branch.

A number of students and colleagues (plus two anonymous reviewers)

have kindly encouraged and assisted me as I pursued this project. Two
former students offered extensive and indispensable aid as the book took
shape: Christopher Brody’s astute critiques of most of my draft chapters
have prompted many improvements in both content and tone; and Peter
Smucker’s setting of the music examples saved me from worry, transform-
ing my longhand jottings accurately and attractively. A sincere “Thank You”
to all!

Conventions regarding note relations, chords, keys, and Roman

numerals

Pitch simultaneities (such as C-E-G) are indicated using hyphens (-), while
pitch successions (such as C-E-G) are indicated using dashes (—). Direction
 Preface Xi

may be indicated in melodic succession: ascending as C<E<G, descending

as G>E>C. A black arrow may be used to indicate a descending-fifth rela-
tionship that is or emulates a V'”)-I succession, whereas an arrow in outline
may be used to indicate a succession from a chord of the augmented sixth
type: for example, C>F-D—G—>C; C-Ab-DGC.
Keys and chords are distinguished as follows: C Major (with a capital M)
is the key of C Major; C major (with a small m) is a C major chord.
Unless another analyst’s methodology is being discussed, Roman numer-
als are presented in capital letters regardless of a chord’s quality, though
modified by one or more accidentals if the chord is altered. Thus C Major:
I-II-V-I and not I-ii-V-I; and C Minor: I-II-V#-I# (closing on a major
tonic), not i-ii°-V-I. An accidental to the left of the numeral corresponds
to the chord’s root, to the right corresponds to its third. The modification of
the chordal fifth, seventh, or ninth requires the inclusion of the correspond-
ing Arabic numeral, as in C Minor: I]}*. The bullet symbol (+) indicates an
absent root. (For example, B-D-F in C Major will be analyzed as V7.)
Likewise a progression of chordal roots generally is presented in capital
letters (C-D-G-C), though on occasions when quality is a factor in the
discussion, a capital letter may refer to major quality, a small letter to minor
quality, and a small letter followed by a degree circle (°) to diminished
quality: for example, C-a~F—d-b°-G-e-C.
A bracket is used to connect the analytical notation for two musical
events that normally would follow one another but which in the context
under discussion occur at the same moment: for example, C Ft BIE when
an F#-A#-C# chord sounds with, rather than before, root B in a descending
circle of fifths.
Parentheses around a pitch in an analytical example indicate that it is not
actually present in the score, though it is understood. Parentheses around
analytical notation may refer to the expansion of a deeper-level harmony
(for example, when I is expanded by I-IV-V-I) or to the harmonic asser-
tion of a voice-leading phenomenon (for example, when the 6 phase of a
I>, as in C-E-G to C-E-A, asserts the harmonic role of VI). Open paren-
theses designate a voice-leading transition between two harmonies. For
example, I ( ) IV indicates that the chords between I and IV (perhaps a
circular, parallel, or sequential progression) do not themselves participate
in the harmonic progression, but instead serve to connect the harmonic I
and IV.
Schubert’s chordal spellings occasionally do not coincide with the struc-
turally appropriate spellings (for example, the substitution of easier-to-read
F#-A-C# for cumbersome G}-Bbb-Db). I generally will use the structurally
Xii Preface

appropriate spellings in my examples and commentaries, often placing

Schubert’s spellings within square brackets to assist readers in finding the
pitches in question within the score.
Because diverse musical contexts are analyzed using graphs, it is difficult
to pin down precise guidelines for how their notation should be crafted
and read. Many styles of “Schenkerian” notation have appeared since the
publication of FC (which itself does not present a single normative style). I
regard the creation of a reductive graph as an art, endeavoring to use nota-
tion that is as clear and informative as possible. In general, open noteheads
in my graphs represent deeper structural or harmonic events than filled-in
noteheads, while notes at the endpoints of beams or slurs are deeper than
internal notes. Notes connected to a beam by a stem are more integral to
the structure than those that are not. Especially in the early chapters I offer
abundant written commentary, which will give readers the opportunity
to develop facility in interpreting my graphic notation. Occasional anno-
tations using abbreviations indicate functions of individual pitches, as
follows:

CP chromatic passing note

IN incomplete neighboring note
N neighboring note
P passing note
W wobble
ant. anticipation
prg. progression
susp. suspension

Of course, the graphs will often incorporate Roman-numeral harmonic

analyses, and in this regard I often depart from Schenker’s practice. Because
it is innovative, Idocument my Roman-numeral usage very carefully as the
chapters unfold.
PART ONE

Methodological Orientation
 i [a
—— :

Gap
ais
Geis
-
2 ee a 7 aa Sit re, ie

foidetuadt bai

- een

= G7 ie

> om: ana

- ) oe aaa
~— a = lang sae -

a : -

- male emit eee ‘= L

Bt

aoe 2 a
. eae
- eae =

'S eae
@ agin
(>) ewan Gage 1S
a % >) Cuan, | tecke orf hae ——
SUSHI ANBAES10: ja,Labi Lai en Tey face bal var

: - : a 7 oe

:
- =

Hi

——
1 Harmonic progression

Tonicized keys and chromatic chords

Tonality’s scope encompasses both local and broad relationships. The

pitches in a phrase from the third movement of Schubert’s early Piano
Sonata in A Minor [1.1] form a local community. Their interactions
establish D Major as a tonal center - as a tonicized key within the broader
A Minor/Major tonality. The phrase traverses the tonal space from tonic
D to dominant A. Though the movement is in the key of A, during these
measures A is a subservient pitch. To persist in charting these chords’ inter-
actions within the movement’s tonic key would be counterproductive. For
example, the E minor chord of measure 62 functions as a supertonic in D
Major, not as a dominant in A Minor. A careful analysis will assess both
the inner workings of the D Major region and how D Major fits within the
broader tonal scheme. (An exploration of the latter topic at present would
take us beyond our current agenda.)
The phrase employs all seven of D Major's diatonic pitch classes plus
three of its five chromatic pitch classes: Ch, D#, and G#. Since every pitch
class is diatonic in fourteen of the twenty-four keys,' some analysts seek out
contexts in which chromatic chords may be interpreted diatonically.* For
example, Schubert’s chromatic G# (measures 63 and 64) might be inter-
preted diatonically as 7 in the key of A Major, a notion embedded within
the label “V/V” (the dominant of the dominant). I eschew that practice,
instead deploying symbols that account for local transformative processes
within the governing key. Here Schubert’s bass G (measures 61 and 62), a
diatonic member of D Major’s supertonic chord, is chromatically inflected
to G# in pursuit of the dominant root. The harmonic activity involves the
connection of just two roots: E to A. To analyze the passage as

m. 61-62 63-64 65-66

D Major: II VEN TON Te

gives the misleading impression that three harmonic events occur and that
the tonal center shifts from D to A and then back to D. I propose instead
Harmony in Schubert

pure
1.1 Piano Sonata in A Minor (D. 537), mvmt. 3, mm. 59-66.

that a transformation occurs within the supertonic as the succession to V

draws near:
m. 61-62 63-64 65-66

6 67 4
yy 8)

D Major: I [aes
7——
aE
6-5
ees
The alterations to the supertonic chord result in dominant emulation.
Though the supertonic’s transformed state mimics the construction of a V’,
the chord remains a supertonic.* An arrow may be used to acknowledge a
fifth-relationship in which dominant emulation is operative (e.g., IV or
EA).
The following guidelines should prove useful in interpreting analytical
symbols such as those employed above:

(1) A Roman numeral corresponds to the scale degree of a chord’s root

within the prevailing key, indicated at the left edge of the row of
numerals. (This analytical practice, extensively documented in TAH,
began in the eighteenth century and was developed especially in
German music pedagogy during the nineteenth century. It is often
referred to by the German term Stufentheorie, or scale-step theory.)
Capital Roman numerals are employed exclusively. Only chords with a
harmonic function are so labeled. (Thus I proceeds directly to V: there
is no intervening I in measure 65.)
(2) If the third above a root is modified, the corresponding accidental is
placed to the right of the Roman numeral. In the II chord above, the
third above root E is transformed from diatonic Gk to chromatic Gf.
The notation “4—#” acknowledges this shift. Diatonic and chromatic
relate to the stated key —- here D Major - not to the composer’s key
signature, which will seldom shift during a tonicization.
Modifications to other chord members above the root are displayed
via accidentals to the right of Arabic numerals corresponding to those
chord members, counting upwards from the root, regardless of the
 Harmonic progression

1 2 3 4 5 6 7 8 9 10 11

:
eS SS 9
;
Os
ae 7
—
9k
DUN Gorell we Ha TT Aa Ty Tp 113)
Oe nmTr wsab Ni

1.2 The II family of chords in D Major.

inversion in which the chord appears. An absent root is indicated by a

bullet symbol («).4
(4) For the II chord of measures 61 through 64, the symbol “8——7” indi-
cates the stepwise descent to the chordal seventh. The prolongation of
a harmony may involve several stages of transformation. Typically a
chord becomes more dissonant and more chromatic over time.> If a
seventh is present in a chord from the outset, the number 7 generally
will be placed to the right of its Roman numeral.
(5) A chord’s positioning - root position, first inversion, etc. - is indi-
cated using conventional figured-bass numbers placed above the
Roman numeral (just below the bass pitches if a score or reduction is
provided).° Often such numbers will be omitted even when a chord
is inverted if that information is not germane to the topic under
discussion.’

The most common variants of II in D Major are displayed in 1.2.

Alterations include omission of the root, raising of the third, lowering of the
fifth, and the addition of a seventh or ninth. In measures 62 and 63 of 1.1,
Schubert proceeds from Chord 1 to Chord 4 (both inverted).
Whereas other analytical systems offer a hodgepodge of incommensura-
ble symbols (ii, V/V, vii°’/V, Ger*®, etc.) that mask the functional commo-
nality among similar chords, here the label II is shared by all chords built
on the same root, with adjustable components to the right of the numeral
noting the chord’s specific constitution. This composite symbol will always
be formulated in terms of the chord in root position, even when the root is
absent. In general, a chordal variant will be referred to not via a name such
as “Chord 8” or “French augmented sixth chord,’ but through the unique
combination of symbols that appears to the right of the Roman numeral.*
A chief asset of this notation is that chordal evolution from one variant
to another (a pervasive feature of Schubert's music) can be precisely and
efficiently charted, maintaining a single parent-chord designation (here
II) to the left of the symbols that indicate the shifting chordal constitution
(here are
Harmony in Schubert

The chromatic chord in measure 61 has a complex derivation. Its potency

results in part from the clash of two distinct entities. Bass G announces the
arrival of the supertonic (in first inversion), secured only in the following
measure.” None of the other pitches of measure 61 assist G in its mission.
Instead D#-F#-A-C§ extends the opening tonic. A 5-6 shift (for example,
D-F#-A to D-F¥-B) is a common means of facilitating the connection of
two chords whose roots form an ascending second.'° The procedure fosters
a gradual changeover of pitch content and avoids the potential hazard of
parallel fifths:

A - B - B
Fe - FR - G
Dea Das Ee
D Major: I e iat

In this context the D-F#-B chord (either in : position or unfurled into

> position) may be referred to as tonic’s 6 phase, rather than as VI, to
emphasize its voice-leading genesis and the hierarchical relationship with
its parent 5-phase chord. However, in a process that I call assertion, the
6-phase chord may come to life, undergoing transformations like those
we encountered in the II chord on its way to V. In the context under
discussion, the pitch B is transformed from a voice-leading anticipation
into a chordal root. The resulting entity is tonic’s lower-third chord. That
status is acknowledged by a Roman-numeral label within parentheses, as
follows:

Pony = akon a eal

DIM Ey a Ss eee ie
aa la)
Since the BE succession (leading to II) parallels the EA succession
explored above, the B chord may be fashioned in an array of variants
equivalent to those shown for the II chord in 1.2. In measures 59 through
61, Schubert skips over B-D-F#, the consonant, diatonic 6 phase of I>,
utilizing only its dissonant, chromatic variant D#-FH-A-Cy (a transposi-
tion of 1.2, Chord 7),' positioning it not before the arrival of II, as would
be normative, but in collision with it.'2 A bracket is deployed above to
indicate that two separate syntactic events are merged into the same
moment in time. The bullet symbol indicates that root B is absent. This
is an important point: the 6 phase’s eponymous 6 is not always present in
a chord’ evolved state. Usually when Schubert employs a ninth, the root
will be absent.
 Harmonic progression

m. 59 60 61 62 63-64 65-66

Di Major f= 11;

1.3 Analysis of Piano Sonata in A Minor (D. 537), mvmt. 3, mm, 59-66.

In an attempt to create a diatonic context for as much of the chromatic

content as possible, many analysts would employ Roman-numeral labels
conceived not only in relation to D, but also in relation to E and to A,
as in
m. 59-60 61 62 63-64 65-66

D Major: I Vike7 oil VWiN- V

My analysis instead embraces chromaticism within the governing key as
one of the principal devices of chordal evolution. The entire phrase is in
D Major, even though some of its chords contain pitches that are not dia-
tonic within the key. Though the VI and II chords are transformed so as to
emulate dominants, only one chord within the phrase is a dominant. In my
view analysts who deploy an “applied” V (or vii°) label for tonic’s asserted 6
phase and for the major supertonic confuse chordal function and chordal
quality. The term “dominant emulation” and the arrow symbol (—) that I
advocate allow analysts to acknowledge certain types of chordal evolution
without losing sight of a chord’s non-dominant function within its broader
context.
Schubert’s voice leading is a bit quirky. As the phrase begins, a descend-
ing melodic third emanating from the initial F# is followed by a reinstate-
ment of F# during measure 61 [1.3]. F#’s successor E (measure 62) seems to
be the starting point for a similar prolongational strategy: D (measure 64)
is a descending passing note that, prodded by the diminished fifth formed
with G#, we expect will lead to C#. Yet C# does not occur in this register. E
is reinstated (measure 66), but the E>D>... line simply peters out. (Observe
that C# does occur in a lower register in measure 66.) Later, in a varied
repetition of the phrase transposed to E Major, the descent succeeds con-
spicuously in reaching the leading tone, D# (measure 135). The high B of
measure 63, which punctures the broader F#>E (3>2) melodic descent dis-
played in 1.3, may be understood as an inner voice temporarily hoisted to
the top of the texture. That melodic contour has a strong unifying impact,
Harmony in Schubert

Sw Vel
&
“i

ae == SSS e= 4 He

1.4 Comparison of Piano Sonata in A Minor (D. 537), mvmt. 3, mm. 5-8 and 63-66.

for it mimics thematic material from early in the movement [1.4] and goes
on to become a pervasive feature of measures 71 through 110."
Finally, we should acknowledge that in pursuing a harmonic analysis we
sort out pitches that perform a local embellishing role, giving most subse-
quent attention to the remaining structural pitches. Though E in measure
59 and C# in measures 60 and 64 would be missed if omitted, their role
is not structural. Likewise, D and F# in measure 65 are embellishments.
Though tonic chords do occasionally appear in second inversion, the chord
of measure 65 does not function as a tonic. Instead it consists of the domi-
nant root (A, in two registers), a suspension (D), and a neighboring note
(F#, in two registers). Depending on the level of detail warranted in a given
analysis, passages such as measures 65 and 66 may be labeled as V$->, or
simply as V. Numbers to the right of a Roman numeral always refer to inter-
vals above the root. To acknowledge the second inversion of a dominant
chord, one would instead place the figures above the numeral.

Context determines function

“Hagars Klage” is one of Schubert’s earliest surviving compositions. The

confidence and creativity displayed therein signal that its composer was
no ordinary teenager. A keyboard interlude [1.5] reconfirms a cadence
that has just occurred. C Major's § third is traversed at two levels: as a
space-opening ascent to 3 (C<D<E) within the initial tonic expanse, and
as a third-progression (E>D>C) that shapes the entire excerpt. The latter
is marked in both 1.5 and 1.6. After 3 and after 2 Schubert extends the
melody upwards, bringing pitches that reside in the interior of the basic
structure to the top of the texture: Gt above 3 in measure 249; A and C
above 3’s implied neighbor F in measure 250; G above 2 in measure 251.
In a major key, the I<V span typically incorporates either II or IV. The
passage from the Piano Sonata in A Minor [1.1] follows the path I<II<V,
 Harmonic progression

247 3 5 1

S
7
S
Ja
= 3

—F
SaaS
oe 433
oe
z ; mi is
-
C Major: I. ih ee, I

1.5 “Hagars Klage” (D. 5), mm. 247-252.

meee e249 250 251 252

“A A Nn
3
|
IN Z 1

S\N BSA
i

<

1.6 Analysis of “Hagars Klage” (D. 5), mm. 249-252.

while the passage from “Hagars Klage” proceeds as I<IV<vV. In both cases a
5-6 shift facilitates the internal ascending-second root succession:

D Major: | I°® U V
C Major: I Vest V

In “Hagars Klage” IV’s 6 phase (F-A-D) is represented by the potent

chromatic variant F#-A-C-E} (measure 250). This chord’s relationship to
IV is identical to that of D#-F#-A-C in measure 61 of the sonata to the
preceding I.
The I chord in a I-IV succession is susceptible to dominant emulation
similar to that which often enlivens II in a II-V succession. In the Piano
Sonata in A Minor [1.1], II (E-G-B) is transformed through a raised third
(G#) and an added minor seventh (D). For the I chord in “Hagars Klage,’ a
similar state of emulation would result from the addition of B}. That does
not occur, however. Schubert chooses a different means of propelling the
10 Harmony in Schubert

Model 1 Model 2

=
CP
—————————

ae

f
C Major: I> : IV

1.7 Models for the connection of Iand IV.

root succession of an ascending fourth: he raises the chord’s fifth, thereby

emulating an augmented dominant.'* Here soprano G#, emanating from
a G lower in the texture, passes chromatically to IV’s A [1.7, Model 1].
The construction becomes more complex as I’s diatonic seventh and ninth
are engaged, passing downwards to IV’s third and fifth, respectively [1.7,
Model 2]. As often happens when the ninth sounds, Schubert omits the
chordal root [1.5, second half of measure 249]. E behaves as a tonic chord’s
major third ought to behave: it ascends by half step to the root of IV. There
is a danger of losing track of the foundational C<F ascending-fourth suc-
cession. Many analysts would label this chord as V;/VI, an interpretation
that unjustifiably makes the IV chord that follows seem out of place, a
“deceptive” resolution of its predecessor. Depending upon its context, an
E-G#-B-D chord might in fact resolve to an A chord, in which case E func-
tions as its root. If it instead resolves to an F chord, then it may be appropri-
ate to interpret E as a chordal third.'* Here E-G#-B-D functions as tonic’s
chromaticized upper-third chord,'® an entity that may precede, follow, or
represent a harmony rooted a third lower. Whether or not a pitch such
as this E asserts itself as an independent structural root is an important
analytical question.'” Too often the casual application of a Roman-numeral
label (such as Vi/VI here) claims such an assertion when the composer
may have had no such intention. As mentioned earlier, chordal evolution
generally works in the direction of increased chromaticism and increased
dissonance. Particularly as the ninth is added, the chord may well become
overburdened. One should not regard the root as sacrosanct: it is in fact the
pitch Schubert is most likely to jettison in such situations.
Third-related chords both below and above a principal chord occur fre-
quently in Schubert's music.'* Thus two prominent generative principles
have been introduced at the outset of this study: lower-third chords come
about through a 5-6 shift; upper-third chords come about through the
 Harmonic progression

Model 1 Model 2

6e
oo
() gioed —
|
——| ONS
oo ae tog HO
=|
|———

CMajor:I V2 JI C Major:1 Va—: I

Model 3

<=
2
==
es a
)
C Major: I Poa
7
ed

1.8 I-V-I incorporating variants of the dominant’s upper-third chord.

addition of the chordal seventh and omission of the root. Third-related

chords can maintain an alliance even if chromaticism is concurrently
introduced. Whereas the affinity between G-B-D-F and B-D-F in C Major,
for example, was acknowledged in the eighteenth century,” a similar rela-
tionship between D-F#-A and lower-third chord B-D#-F# (represented
by Dz-F2-A-C in 1.1) in D Major and between C-E-G and upper-third
chord E-Gé-B (represented by E-G2-B-D in 1.5) in C Major likewise can be
proposed despite the chromatic fluctuations.”
Schubert often chromaticizes the dominant’s upper-third chord when
composing in a major key. The instability of the diatonic triad on 7 (e.g.,
B-D-F in C Major) likely induced his frequent adoption of chromatic vari-
ants. In 1.8 the diatonic version (Model 1) and two prominent alternatives
are displayed.*’ These models propose that what I call V-space begins with
V’s upper-third chord. Though Model 1 is serviceable, its emotional impact
is limited. In contrast, Model 2 is bright, with its juxtaposition of two major
triads a half step apart, unavailable in any diatonic context. Both Dé and
Fé play a connective role. The label V/III would be misleading: why should
downward-tending Dé be christened as an upward-tending leading tone?
In contrast, Model 3 is subdued. The flat realm is called into play, and the
B>—Bs juxtaposition comes across as a release from a somber fate.”
Models 2 and 3 reveal one worrisome potentiality in the deployment
of the dominant’s chromatic upper-third chords following tonic: parallel
fifths. Schubert clearly was aware of the danger. In “Schwertlied” [1.9], he
traverses the space between C and B via a deformed chunk of the circle of
descending fifths (C-F2—4B), thereby preventing untoward voice leading.
(The F# and B chords collide in measure 14. Compare with the collision of
Harmony in Schubert

11

2S as = E =
hab' mei - ne Fred-de dran.
das freut damSchwer-te — sehr."

== mee Hur -_ rah!

——— Hur -_ rah!
Hur -_ rah!
Hur -_ rah! Hur -_ rah! Hur -_ rah!

4
te SS# + é ay
6% 7]
SS
ir if fe
—
kz
d a a a ASS = eo.
2 zy 2 2 4 = se7 4
Fe B G (Cc

1.9 “Schwertlied” (D. 170), mm. 10/11-17.

B and E chords in 1.1, measure 61.) In “Auf dem Strom” [1.10], the E>D#
juxtaposition is softened through the delay of D#5* by means of a §, yet
even this conventional dominant indicator is not sufficient to offset the
downward continuation. D#-Fx-A# functions as an upper-third chord of
dominant B, and not as the dominant of a G# chord that does not material-
ize. During the first movement of the Piano Sonata in C Minor [1.11], Eb
and Db chords occur in direct succession. Though the }?fifth of measure 46
connects at a fundamental level to the }?fifth of measure 48, the intervening
melody fills in the descending third B}>Ab>G. Consequently Bb and Ab are
not directly juxtaposed.
Though an upper-third chord may be prolonged or even tonicized, its
duration in these three examples is brief. It functions in each case as a
preliminary phase of the impending dominant, and thus as the onset of
V-space. One or two of the dominant’s pitches are in place from the outset,
while the remaining pitch(es) - corresponding to the filled-in noteheads
of 1.8, Models 2 and 3 - come to order once the root arrives and asserts its
 Harmonic progression 13

x
|
|
' | 4
< YS =
++
| 1
e
||
Het

ai

——
oe

———

—.
4 P
=i

PP
=
ee

t
&

Z¢4

tas g — S— — SS I ee

|= ER TEE Ee #\a°* a Soe

a
P PP
=a"
ess - a =
ro e
oa == -

1.10 “Auf dem Strom” (D. 943), mm. 6-11.

44 . a - +
—_. Ra

t < = —— ws = [aa ae a
4 +
=e =o a —— e os pF:
© v wetter
Rnceeeeoie | a = Ses ix Pp
| ae ee
es EEE pa Re», le Pe a
Bare oe —— a | =

1.11 Piano Sonata in C Minor (D. 958), mvmt. 1, mm. 44-53.

14 Harmony in Schubert

authority. Generally I will place the Roman-numeral label V at the outset

of the V-space (as in 1.8), followed by symbols that chart the chordal evo-
lution resulting in full attainment of the dominant.” If such a chord does
not proceed eventually to the dominant, then likely it is not the dominant’
upper-third chord.” For example, it might instead participate in a circle of
descending fifths (C-F-Bb-Eb-. . .) or serve as V announcing a tonicized
mediant region (Bb—E} or BE).
Of course several viable options exist as the successor of almost any
chord. The extent to which the paths not taken are explored and remarked
upon varies among analysts. For example, the D}-F-A}b chord of 1.11, meas-
ures 48 through 50, does serve as dominant Bb’s upper-third chord, but it
could have served as a dominant announcing a tonicized mediant region
(Gb Major) and in fact does behave in that manner in measure 71 (there con-
necting Eb Minor and G} Major). Because chords may play multiple roles
within musical syntax, potential interpretations that the composer might
not have intended may emerge in the minds of listeners. Analysts must
learn to cope with this dilemma. I recommend a practice in which interpre-
tations are confirmed through consideration of the broader context, taking
into account pitches that sound after the chord in question.”° Other analysts
may take a chronological approach, processing only the data received up
to the moment of a chord’s sounding in coming to terms with the situation
and thus contending with a wider array of unrealized hypotheses, since less
data from the composition is used in determining how a chord functions.”®
In most cases I find that methodology to be unproductive, particularly after
repeated hearings of a work. For comparison, consider an example of com-
munication in English. One might develop a specific notion regarding what
a speaker is conveying upon hearing the following syllables:

“T attended a mass...”

that is either confirmed or contradicted by the continuation. The sentence

might in fact be

“I attended a Mass celebrated by Pope Benedict XVI during his visit to

New York in 2008”;
or instead

“I attended a massive demonstration protesting the Roman Catholic

Church’s prohibition of birth control.”

Clearly the speaker intends one or the other of these interpretations of the
syllable “mass” — not both. That meaning is discernible only by taking into
 Harmonic progression 15

account content that follows the syllable. Anyone who initially misinter-
prets that syllable would immediately relinquish the faulty hypothesis as
nonsense upon comprehending the broader context, and not dwell on it.
Some readers may argue that our mechanisms for the processing of English
and of music are sufficiently distinct as to disqualify this comparison.”’ (In
that the succession from one chord to the next is generally much slower
than that of words, there is more time for auditors to develop alternative
hypotheses before the context clarifies the structure. Composers might
take advantage of that situation as a compositional device.) Yet in order to
proceed in the analytical enterprise one must take a stand on the issue. I
advocate that one not dwell on potential chordal interpretations that are
incompatible with what follows, making accommodation for special cases
in which an event is hinted at in an early statement but blossoms only later
(such as D}’s resolution to Gb in the Piano Sonata in C Minor, mentioned
above).°

I-II

I<II

Some musicians regard II as no more than an offshoot of IV.’? Sometimes

that viewpoint may prove instructive, as in the asserted 6 phase of IV°°. In
other contexts, however, II is derived directly from root 3. Internal to EW,
IV invites a specific set of responses: a composer might modify I to produce
a more focused pull towards the subdominant (for example, I*-”’), and a
5-6 shift will often assist in navigating the IV<V span. The same responses
— differently positioned — often occur when II comes between I and V: a
5-6 shift will often assist in navigating I<II, and II may be transformed to
pull more strongly towards V (for example, II!"{). I<II and I<IV are equally
viable, yet different, approaches to the same goal, V.
From a major tonic, a 5-6 shift is an efficient means of directing the
progression towards the supertonic. Model 1 in 1.12 shows its simplest
instantiation. The chromatic variant of Model 2 is a common dominant-
emulating alternative. In a minor-key context, 5-6 leads more effectively to
LII (Model 3)? unless drastic chromatic modifications are made (Model 4).
In a passage from the first Landler in D Major [1.13] from his set of
Zwolf Landler, Schubert fills in the tonic 5-6 chromatically (A<A#<B),
with an A#-C#-E-G chord (measure 10) coming between tonic’s 5 and 6
phases. The supertonic, decked out with major third and minor seventh,
follows directly thereafter, leading to V at the cadence. Just twenty Deutsch
16 Harmony in Schubert

Model 1 Model2

6 Se =a
mee
C Major: I° II
ried
C Major: Iij—+# Il
(= VI)

Model 3 Model 4

qe
b; ‘ae
a a ¢ 5h
: 2 ; eng
C Minor: P—° bil C.Minor: 1k——} IP
(- BVI?)
1.12 Models for I°-°-II.

numbers later, in the last movement of his String Quartet in D Minor,

Schubert gives this progression a more audacious realization [1.14]. The
passages relate as follows:

Landler Quartet
m. 9 m. 726 Tonic D Major is stated. In the Quartet,
D Major displaces the movement's D
Minor key for the concluding Prestissimo
section, extending the cadential tonic
(with Picardy third) into a region of
forty-eight measures.
m. 10 ms 727. Behaving as a leading tone and forming
dissonant intervals with E and G, the
Landler’s A¥ raises the expectation that a
B minor chord will follow. The Quartet’s
starker C major chord results from voice
leading in contrary motion connecting
tonic’s 5 and 6 phases.
m. 11 m. 728 In the Landler, tonic’s 6-phase chord
is diatonic: D-F#-B. It is preceded by
its evolved upper-fifth chord. In the
Quartet, chordal evolution is at play:
B-D#-F# more dynamically prepares II
(via dominant emulation), asserting itself
within a three-measure expansion of
VI#, culminating in the further evolved
 Harmonic progression 17

9 ae aa Fame —

3 — Ce eee
; Ze 2 ;

lone = =r z= see
=_ ars
D Major: [> ' ue i ==-ve d
ai)

1.13 Landler in D Major from Zwélf Landler (D. 790/1), mm. 9-16.

726
‘ :
gg |ora 6 SEs 6
ss6ser ee.
ee
rrr sntSt eee
ea ass es RE ae acer
Ve eR Be Ee ale Le
a ! !
= He es ys = =

= SSS ee
9

D Major: 1° )-6 we v8 —
10 9: °
8 if
(= VI3 )

1.14 String Quartet No. 14 in D Minor (D. 810), mvmt. 4, mm. 726-733.

state, D#-F#-A-Cy. A connective chord

(B’s upper-fifth chord) comes between
these two statements of the evolved B
chord.*!
m2 m. 731 Whereas the II chord is converted to
major and supplied with a seventh
(dominant emulation) in the Landler, in
the Quartet it is further transformed into a
diminished seventh chord (E-G#-B-D into
G#-B-D-Fs). The downward momentum
of the bass persists up to this point.
mm. 13-16 mm. 732-733 After the relentless bass descent in the
Quartet, the relative stability of the
dominant finally takes hold. The Landler
18 Harmony in Schubert

17 54 one) 1
|
=I ca =e:
—o
=e
oe = |
Nim-mer trag'ich lan - ger die - ser Lei - den Last.
Off -ne mir den Him - mel, mil - der, git’ - ger Gott!

=o NF SSS a
ercresc. SS
: are
¥ =
Bar
e x iS Z==S tyIar = D5 aS oSG =e: = =

2 s Il
fs) VeeCex
Sa
B Minor: ———
(¢—$——
;
(= #v1?

1.15 “Der Leidende’” (D. 432a), mm. 17-21.

achieves the dominant less flamboyantly,

via the fifth E>A in the bass. In both the
Landler and the Quartet, the dominant’s
seventh emerges.

The minor-key trajectory for 5-6 displayed in 1.12, Model 4, serves as

the foundation for a passage from “Der Leidende” [1.15]. As was also the
case in the Landler and the Quartet, the 5 and 6 phases of I°-* do not reside
side by side. Since the 6 (G#) occurs in the bass, the melodic filling-in of
third B>G# is an appealing option. This motion in turn invites a parallel
descent a tenth higher (thus >2°%@4). The intervening chord” softens the
harshness of three simultaneous chromatic shifts (as occurs in the model
of 1.12), as does the common tone F¥. Because the chromaticized 6 phase
of the I°-° emulates a dominant (G#—), II must be consonant (with fifth
G# in place of diatonic Gk).°° Yet a G4 inflection does occur fleetingly
in measure 19 before the cadential V#—I. The vocal melody contains
some structural holes. From a voice-leading perspective, the vocal C# in
measure 18 should pass down to B# (as does the keyboard’s C#), yet F#
sounds instead. This F# is, however, an essential pitch: though it arrives
later than usual (during the 6 rather than the 5 phase of I5-°), it is the

structurally aberrant as well: C# is missing in measure 20, in both the vocal

and the keyboard lines. In an idealized voice leading, thirds would proceed
downwards to the cadence:
 Harmonic progression 19

i mit Kraft Ee

SSS
men. Gross ist Je -
es =!

ho - vah,der Herr, gross ist Je-

== ee
vah,der Herr.

1.16 “Die Allmacht” (D. 852), mm. 70-83.

Behe se D are(Chir iB
Perk. o(B)r Ad B.

Whereas Schubert unfolds the D<F# and E>C# thirds in measures 17

through 19, the following two thirds are not so meticulously presented:
neither B nor C# (shown within parentheses above) occurs in the melody.

I>II

Root-position I and II generally will relate via the bass ascent of a second,
not the descent of a seventh. Thus the closing measures of “Die Allmacht”
[1.16] are exceptional. The eponymous All-Pervading Power is not con-
fined to small or diatonic spaces. The bass fills a mighty range, while
chromatic pitches permeate the texture. Schubert even extends from C
major through G) major, tonic’s antipode,™ calling to mind another text:
20 Harmony in Schubert

m 70 75 76 Tdi 78 VD

= Se
(Q) 2 2 a
Z

i‘ i‘

MW i‘
y
i) Q

5
C Major: loa,

1.17 Analysis of “Die Allmacht” (D. 852), mm. 70-79.

“I am Alpha and Omega, the beginning and the end, the first and the
last?
The Gb chord is marked fortissimo, and it supports the final word of
the text (“Herr? which will be supported by a tonic C major harmony when
repeated five measures later). Yet that moment is neither the end of the
crescendo, nor of the chromatic ascent in the keyboard melody, nor of
the stepwise descent in the bass. Those markers coordinate with the chord
succession from A major (inverted and with seventh) to D minor in the
following two measures. Model 2 of 1.12 displays that very succession; 1.17
places these chords within a foundational harmonic structure.
The first soprano pitch in 1.17 is G. In Schubert’s composition G
arrives at a later moment, in the keyboard melody of measure 75. The
; third, a component both of the opening tonic and of the Aj, chord of
measure 75, is unfolded over time. (Compare with F#’s belated arrival
in “Der Leidende” [1.15].) Model 1 of 1.18 shows how the melodic
ascent E<G, achieving 5 in the soprano, coordinates with the chromatic
inflection of C to C#. Model 2 displays an expanded context for tonic’s
6 phase, with the bass leading downwards, rather than upwards, towards
goal D. In Model 3, the A chords of measures 72 and 75 are connected
via passing notes.*°
Schubert's composition deviates from this normative construction by
one half step in one chord: the model’s E in measure 72 is displaced by an
anticipation of measure 73’s chromatic E#. Through this rambunctious
deformation, tonic’s 6 phase is not stabilized at the outset of its prolonga-
tion. In fact, Schubert creates some uncertainty regarding exactly which
trajectory he is pursuing en route to V: is it 1° I, as proposed in 1.17,
or is it I” IV, as a reading of A-C-E# in measure 72 as A-C-F (Schubert's
spelling) would suggest? Only the broader context of the entire phrase con-
firms that II serves as the principal connector between I and V. Schubert’s
spellings of convenience in measures 72 through 74 make the passage seem
 Harmonic progression

Model 1 Model 2 Model 3

meee 5p 876 m. 70 Tz 75 76 m. 70 15> IB EN sy US
—_ ——— yo ——— ces

6 =|
aN

GMa
oe e¢

=
=
oS) FP

“i
eos

pe
wr SS
oS oe

Il
ee

Cas
ee
5
id => yo -
a- =

II
(= Vii)7 &—7
(= VIF—4) (= VIE- == 4)
1.18 Analysis of “Die Allmacht” (D. 852), mm. 70-76.

Massig
3

jee Eo “@ 3
SS SS
é £6 || 4S a i
e > ro / oe

G Major: : Leg IT vo} I

(=IV° : VI)
(Hy)

1.19 “An die Nachtigall” (D. 497), mm. 1-8.

more eccentric than it in fact is. The proposal in 1.18 is that there is no
diminished third in the bass of measures 74 and 75, nor is there a Dbh-C#
enharmonic shift. Apparently a reluctance to notate measure 72 as A-C-E#
led to the spellings using flats in the following measures, thereby requiring
a shift back to the sharp side in measure 75.

IV-V
IV<V with 5-6 shift

When I-IV inaugurates a harmonic progression, the extension of IV via

a 5-6 shift often serves to direct the progression towards V. (Otherwise a
circular progression of descending perfect fifths might ensue: e.g., C-F-
Bb-. . .) If the 6 phase asserts itself, a second fifth-relationship is established
(e.g, C-F/D-G in C Major). Dominant emulation may enhance one or
both of these relationships (C-~F/D—G).
The chords that begin “An die Nachtigall” [1.19] form an auxiliary pro-
gression — that is, a progression that starts away from tonic.*° Only gradually
Harmony in Schubert

12 3 gO
| mo 2 rx ia ial o . ri Ns

oe ——— + 7 + as ee
zelt ver - birg in dei - ner grii - mnenHil - le die
ab; doch leicht er-wor - ben, aus dem Schoo - sse der

pi Sa
ea
a Pe

ial
=

es
aa crese.
fad ein: eee e: ga
oh |
Ab Major: I° us a

15 Beutls Be
(ae ==
Lie - ben-dendem Aug' der Welt!
Got - ter fallt das Gliick her- ab.

— } eS : =e oye zt :
Gite 909 = : fee ha fen
~ Coa ga
: R 2 Pare aN - al ;
pes eff : be — —
Ban me Y
( Sevier. I
1.20 “Das Geheimniss” (D. 250), mm. 12-17.

do we comprehend that the work is in G Major, and not C Major. The A

in measure 3, as C’s sixth (thus IV°~°), is a conventional move within a
chord that is here unconventionally positioned at the starting point of a
progression. Though Schubert frames this introduction to his G Major
composition with a descending G-to-G octave scale (the circled notes), the
harmonization of which was of special fascination to musicians for centu-
ries, the chords he employs and the presence of Fb in place of F# in measure
2 offer a welcome contrast to the ho-hum models of harmony manuals.
Whereas the interlude from “Hagars Klage” [1.5] ascends a third to 3 and
then descends 3>2>1, “An die Nachtigall” descends a sixth to 3 and then
descends 3>2>1.
In a passage from “Das Geheimniss” [1.20], the 8-75 motion that
expands tonic engages several other chords (measures 12 through 14).
The stepwise-ascending soprano fills in the diminished fifth ©. The bass
 Harmonic progression es:

Model 1 Model 2 Model 3

p p
cerery
a= (Sas
ry aa e id g 2 78
: ‘ N
Ab Major: Iv>~° Ab Major: IV? ° A> Major: [V2-( =e

1.21 Analysis of “Das Geheimniss” (D. 250), end of m. 14 through beat 2 of m. 15.

supports most of those notes a tenth below. Schubert responds to this C<G}
diminished fifth by melodically unfolding its major-third resolution F>D}
in measures 14 and 15. (See the circled and boxed notes in the score.) The
primary strand, which begins as C<Db, proceeds downwards by step to Ab
for the cadence.’
In 1.21 three models for the F>D> connection between melodic strands
in “Das Geheimniss” are displayed. Model 1 shows IV>* embellished only
by passing note E>. In Model 2, dissonant Eb is made locally consonant
through the coordinated deployment of neighboring note C in the bass.
In the context of the phrase, this C-E>-A> chord does not convey a tonic
function. Its subsidiary status does not prevent Schubert from adding yet
another hierarchical layer, coming between IV and the C-E}>-A> chord
(Model 3). The G of this inserted chord forms a dissonant minor seventh
with soprano F, a destabilization that propels the melodic descent from
F through E> to D>. Model 3 also incorporates a chromatic inflection: F
mutates to chromatic F> en route to the upcoming E>. (Compare with the
G%>Gz inflection in 1.15, measure 19.)
Consider again the connection between the two chords of measure 250
in “Hagars Klage” [1.5]. The shift from F to F# is an ascending melodic
gesture that targets the upcoming dominant root G. Schubert deploys these
chords (transposed up a step) in the last of his Zwanzig Walzer for Piano
[1.22, measures 9 and 12]. The presentation of IV in first inversion opens
up some space, so that instead of a direct G<G# bass motion, a chromatic
line descends from B to Gé. As in “Das Geheimniss,’ the two phases of IV>~*
consequently do not occur in adjacent positions, but instead are connected
via a passing chord (measure 11). The downward momentum proves to be
unstoppable. Though bass G# normally would ascend to dominant root A,
after the chromatic descent B>B»>A>G# Schubert continues with Gs and
then F#, representing — in unexpected inversions — the V and I that we
expect after IV°~®. Because V’s minor ninth is added and its root is omitted,
two diminished seventh chords appear in succession.
24 Harmony in Schubert

(- 17)
9%

1.22 Walz in D Major from Zwanzig Walzer for Piano (D. 146/20), mm. 0/1-16.

IV’s upper-fifth chord

Though the variants of IV°~° offer a rich creative resource for the con-
nection of IV and V, other options are available. For example, IV*” is a
straightforward way to harmonize a melody that descends by step from
the subdominant’s 4 to the dominant’s 2. A passage from Schubert's late
Piano Sonata in C Minor [1.23] develops this strategy. The melodic descent
from incomplete neighbor Ab (= 4) to F in measures 42 and 43 coordinates
with the succession from IV to V.** In 1.24 three stages in its evolution are
shown. In Model 1, a straightforward I-IV-V progression harmonizes Bam
with an intervening incomplete neighbor. Model 2 displays the passing
note corresponding to IV’s seventh. Taking the harmonic support into
account, one discerns that the second G is not a restoration of the initial
3 but instead a note generated by melodic forces. (Thus a slur connects Ab
and F, not G and G.) Finally, in Model 3 the subdominant’s root drops out as
 Harmonic progression US:

2
SSS -s tan a5
pp a ee

Beg teen Eas ot emeeeee’

=

.
o-0 ©

7—
;

E> Major: I

1.23 Piano Sonata in C Minor (D. 958), mvmt. 1, mm. 39/40-43.

Model 1

3. IN 2
Ge
e
Z 2
a
ZZ
OD
{

Eb Major: I IV Vv

Model 2

sy
Bhasin iP 5/2
A = S ZZ
e = OT

E> Major: I Vee en V.

Model 3

3 IN D

Eb Major:I1 IV ( ) V

1.24 Analysis of Piano Sonata in C Minor (D. 958), mvmt. 1, mm. 40-43.
26 Harmony in Schubert

the seventh arrives, with the chordal fifth and an anticipation of the domi-
nant’s root as support. Though an upper-third chord (as in 1.5, measure
249) may convincingly extend the impact of its parent chord, generally an
upper-fifth chord offers a contrast.*? It will either lead back to the preceding
chord or serve as a link to the next chord of the harmonic progression. In
1.19, IV’s upper-fifth chord G-B-D-F (measure 2) leads back to IV, and thus
the analysis displays a prolongation of IV.“° In 1.23, IV’s upper-fifth chord
E}-G-B} links IV and V. The open parentheses in the Roman-numeral row
indicate the no-longer-IV, not-yet-V state of affairs. In neither case does
IV’s upper-fifth chord function as tonic. It instead offers consonant support
for passing note G.

Peculiar juxtapositions

Schubert's music exudes vibrancy and creativity that are wondrous to

behold, even if those laudable qualities sometimes result in perplexing
musical structures. An imperfect fit between analytical tools and the music
under scrutiny accounts for why some of Schubert's procedures have
remained elusive. Yet sometimes his writing seems to invite interpretive
decisions that ultimately prove to be untenable. On occasion the surface
design may appear to be in conflict with the underlying structure. One type
of conflict will be explored here: the juxtaposition of two musical events
that one might have good reason to regard as equivalent, but which in the
broader context of the work do not function as such.
Schubert opens the Scherzo from his late Piano Sonata in A Major with
an enigmatic presentation of the tonic key [1.25]. Two seemingly equiva-
lent events are juxtaposed: in measures 5 through 8, an A major chord
proceeds up a step to a B minor chord; then, in measures 9 through 12, a G
major chord proceeds up a step to an A major chord [1.26, Hypothesis 1a].
The A and G chords that begin these four-measure units are in § position
and thus are well suited to serve as the starting points for brief descending
parallel progression whose perimeter chords convey a sense of chromati-
cized 5-6 expansion.*' In such constructions the 5 phase generally is hierar-
chically deeper than the 6 phase. (See Hypothesis 1b.) The metrical context
reinforces this interpretation: measures 5 and 9 are hypermetric downbeats.
Yet there is something very disturbing about this seemingly straightforward
analysis, for it represents a topsy-turvy world in which A Major's leading
tone, G#, is subordinate to chromatic Gk. Despite the tidy parallelism in the
construction, I suggest that an alternative guiding principle prevails. The
 Harmonic progression

Allegro vivace
'

1.25 Piano Sonata in A Major (D. 959), mvmt. 3, mm. 1-16.

Hypothesis la
m. 5 6 7 8 9 10 1] 12

ph we
J)
ry ASSN
ey WR, we wh it
Hypothesis 1b

& 6 B G 6 A

Hypothesis 2
aati
i.
Z ne bg zs,
a Z
Z
$2
Z
3 g
oot
3S g%
{
7
NMajornl hen Va! I
(= VIF)

1.26 Analyses of Piano Sonata in A Major (D. 959), mvmt. 3, mm. 5-12.
28 Harmony in Schubert

(SSS ee
stiss,. dass es dir vél-lig glich. Und vor ihr lag ein Jing - ling auf den
letz - te, schwache Kraft ent- wich. Da kam ein Jiin - ling hiilf-reichihr ge-

= 2 ———
SOS <See SS? Nig ——
f
oH
a z = =

patualaas
ee: 4

' :
: 8
an (
G Major:V é :
Vv’ VI:
Jey

er
i

— ae ——s
$

—= + oe 7 y
Knie - en, er schien sie sanft an sei-ne Brust zu zie - hen,
flo - gen, er sprang ihr nach und trug sie aus den Wo - gen,

Ss rrrrrresy Wbiitites ss ain

ot 3 eee 7s
ae fang aa 2
Is, i v’

1.27. “Das war ich” (D. 174), mm. 8-13.

first four-measure unit (measures 5 through 8) leads conventionally from I

to II. In that context a 5-6 expansion of tonic is a commonplace. But from
II the logical continuation - what listeners will expect - is the dominant,
not the subtonic. Thus G major (measure 9) should come across as an
intermediary event, and not as a chord on the same hierarchical plane as I
and II. In my view it initiates the progression’s V-space. (Compare with 1.8,
Model 3, and 1.11.) Once V is fully in place in measure 11, the progression’s
focus is clarified (Hypothesis 2). The surface parallelism of Schubert’s
four-measure units thus is in conflict with the foundational I-II-V-I
harmonic progression that shapes the passage.
A passage from “Das war ich” [1.27] is similarly perplexing. On the
surface measures 12 and 13 appear as a transposition down a step of meas-
ures 10 and 11, giving rise to a potential interpretation of the passage as

m. 10 ill 12 13

A Minor TV Vit ule «GMajoriel Vpaaiy ol

 Harmonic progression 29

The fortepiano markings in measures 10 and 12 enhance the parallelism.

Yet the foundational harmonic progression proceeds along an altogether
different path. Observe that the keyboard melody in measures 9 through
12 is shaped as a descending chromatic line — F#>Fk>E>Eb>D — below
which the span A>G>F¥# and above which (in the vocal melody) the span
D>C occur. These spans all reinforce a prolonged V*~’, which overrides the
surface parallelism. II comes between the perimeter V chords, allowing V’s
seventh C to be introduced in a consonant context (as a tenth) in measure
11 before its dissonant phase begins when root D returns.‘ The deeper
elements of the progression do not correlate with the two-measure pat-
terns: the first V chord occurs even before the first statement of the pattern
begins, while the second (V’) is interior to the pattern’s transposed repeti-
tion (as is the V in 1.25, measure 11). The fortepiano markings accentuate
moments of chromatic passing motion in the keyboard melody: F#>F4>E
and E>E}>D. The latter is supported by II’s upper-third chord.
The Rondo in B Minor for Violin and Piano contains a powerful juxta-
position of two potential dominant chords: F#-D-C (an inverted V’ of G)
in measure 380, and F#-A#-E (V’ of B) in measure 382 [1.28]. Both chords
are presented sforzando at the peak of a crescendo coordinated with lines
moving in contrary motion (diatonic in measures 379-380, chromatic in
measures 381-382), and both are prolonged during a decrescendo to piano.
The passage contains several expansions of individual chords. Model 1 of
1.29 employs slurs, open and closed noteheads (indicating structural and
connective pitches), and crossed diagonal lines (indicating voice exchanges)
to display how these chords are expanded. The initiating G-D and closing
D-G phases of the progression behave in a conventional manner (isolated
in Model 2). The content of measures 381 through 385, which amounts to
a parenthetical passage, dramatically extends and potentially deviates from
that secure tonal path, hinting at resolution to B Major. Yet, as Model 3
shows, the F#” chord may be interpreted as an evolved D” chord, with absent
root (that is, as one of the D dominant’s chromatic upper-third chords).
The A# and C# function as wobbly notes. A wobbly note temporarily takes
on a chromatic inflection that is later revoked.*° The § chord prolonged in
measures 383 through 385 functions as a conventional passing chord whose
continuation is elided (as displayed in Model 4).
Though the Rondo’s surface juxtaposition of what seem to be conflict-
ing dominant seventh chords suggests potential continuation either in G
Major or in B Major, in the broader perspective the D chord controls the
F# chord — or, more strongly worded, the F# chord is the D chord, in an
evolved state refashioned through chromatic adjustments, the addition
30 Harmony in Schubert

decresc.

,2 te # i: Ss a
: decresc.
—- =

as cbs peer 5

papa eae aE
te Teilioe 23% eetore'

1.28 Rondo in B Minor for Violin and Piano (D. 895), mm. 377-389.

of the chordal ninth, and the omission of the root.” We encountered a

similar phenomenon in “Hagars Klage” [1.5], where tonic, in its domi-
nant emulation phase, evolves from C-E-G to E-G#-B-D as it approaches
F-A-C. The middle chord is identical in structure to the Rondo chord of
measure 382. In “Hagars Klage” it leads directly to the F chord, whereas
in the Rondo it reverts to a D’ before succession to the goal G chord.
Schubert may occasionally take this notion one step further by employing
the modified ninth chord only - that is, without ever sounding the root.
The Rondo’s F#-A#-C#-E chord also appears in measure 12 of “Auflosung,”
where its potential to resolve to a B chord likewise is held in abeyance
[1.30]. Its placement between two tonic G chords invites an interpretation
like that in the Rondo, as a dominant ninth chord with absent root and
 Harmonic progression il

382 383 384 385 386 387 388 389

ce oe i

eee
D G

Model2
ee a

S 7) 7 “4

G Major: I We I

Model 3

¥ |
=H
28 g
G Major: vi i :
oy -)

Model 4

a zuSee —<—_
SeeSe
|
7 ————o
5¢ 4 34

1.29 Analysis of Rondo in B Minor for Violin and Piano (D. 895), mm. 379-389.

wobbly fifth and seventh. In this case, the A and C that serve as founda-
tion for the A# and C# wobbles are elided.**

Unfurling of§chords

If a chord appears in second inversion, the pitch a fourth above the bass
serves as its root, the generator of the bass and of the sixth above the bass.
It is relatively uncommon for a harmonic function to be represented by a
chord in second inversion. Most § chords instead fall within the domain
of embellishment: passing, neighboring, anticipating, or suspending notes
displace or connect harmonic pitches. In such contexts the fourth above the
bass is not a root. In our discussion of 1.1, the label V$-> was proposed for
measures 65 and 66. This label indicates that the V chord’s fifth and third
are embellished — not that the chord is inverted.
Sy) Harmony in Schubert

—_— . oe oe

og
goss 5 = 2 !
denn die_ Glu- then der Won - ne

pees gece = “26 ee,

ite ee f we

; —
Ea 3 : E 4
A a a
G Major: I

q2 re

ee ee ver-
- gen mein Ge - bein;

a 9—2 c = d steve =

= = © to? nreazieta oye é 4

PP
= _—— ——

a8 : eee aon
Gg

3
7) a FG o =.

“i I

1.30 “Auflosung” (D. 807), mm. 9-14.

Embellishing § chords may themselves be extended through inversion,

thus appearing in 3or $position. The figured-bass symbol }and the notion
of root position are not equivalent. The former is a neutral description
stating that above a chord’s lowest pitch a third and a fifth sound; while
the latter implies, in addition, that the third and fifth are derived from -
dependent upon - the bass/root.
Two fourths occur above bass A during the first measure of “Der
Wanderer” [1.31]. Soprano D on beat 3 is a rearticulated suspension,
extending the supertonic’s seventh into the domain of the dominant. After
beat 4, D is one of two concurrent neighboring notes: F# embellishing E, and
D embellishing C# (positioned an octave lower in measure 2). The resulting
A-F#-A-D is not a tonic chord in second inversion. The dominant function
persists, even if only its root is sounding at this moment. E and C¢ return at
the downbeat of measure 2, followed again by embellishing neighbors. Since
the persistence of bass A may by now begin to cause tedium, Schubert here
 Harmonic progression

Langsam

D Major: I it haae a

1.31 “Der Wanderer” (D. 649), mm. 1-2.

Ruhig, zart
= = ee aN
Sa Der A-bendschlei-ert Flur und Hain in trau-lich hol - de Damm- rung ein;
Die Wo-gen- fluth hallt Schlum- mer- klang, die Bau-me lis-peln A - bend - sang;

as
4 ————————— | | —_— N

5
——
— 4
——
° P F
—
id it
i i oa3
sempre pp SS
2 @ 2 Se
-s

SS Se ee
- rd s 7 nd o—#

——
, ==]
G Major: I V35 yay
Q= 7)

1.32 “Geist der Liebe” (D. 414), mm. 0/1-4.

unfurls the §chord, so that A and the neighboring notes F# and D appear first
in §and then in 3position. These chords do not represent tonic harmony any
more than does the §chord of measure 1. Just as the 5-6 shift may generate a
lower-third chord, a >-° shift may generate a lower-fifth chord.
The tonic is prolonged during the opening two measures of “Geist der
Liebe” [1.32]. The thirds G<B and B<D are presented unembellished
(measure 1) and then with internal passing notes (measure 2). The domi-
nant chord that follows accommodates neighboring note G (embellishing
F#) and passing note B (filling in the dominant seventh’s A<C third) in
the upper lines of measure 4. Instead of presenting D-G-B in § position,
however, Schubert doubles G in the bass. The resulting chord is not a
tonic. Though a leading tone generally resolves to the tonic pitch, the G
of measure 4 is not that resolution. The dominant harmony persists until
measure 5.*”
 34 Harmony in Schubert

Allegro moderato

Violino I Glee Sh Ze i= = ce f Fo
e
PP ="
4 t
ViolinoII] oy@—> = foto His = 2—a =
PP
= = = = === : eo =o
Viola [ete
PP

Soprano ets —- eee == 2 a : EE = = = |

PE
Cre - doin u - num De - um, Pa - tremom- ni - po - ten - tem,

aito |eB =P SS SSS

PE
fz = f= E= S e - = (Z cox =
Tenore ioere [ | = :
PP ;
Cre - doin u - num De - um, Pa - tremom- ni - po - ten - tem,

ow SSS
|

sempre staccato ; ; ;
a a SSS] So = —= = Sr = ==
Silane [eases serene ai a ea
LHL
G Major: I ey

1.33 Mass No. 2 in G Major (D. 167), Credo, mm. 1-8.

The notion of the dominant’ third and fifth gradually falling into place
by descending step over a prolonged root (V$-3, as in 1.1), for which the
terms cadential ¢ and accented $ are often employed, prevails even if such
a ° chord is unfurled into $ or } position, as in the Credo from Schubert's
Mass No. 2 in G Major [1.33]. In both the upper string and vocal parts, G
and B enter as accented passing notes that persist for a full measure above
bass D (measure 7). In the continuo line the §chord is unfurled: D>B>G<B.
For a moment G-B-D, which in another context (such as the Credo’ first
measure) could function as tonic in root position, sounds. In this context
that chord is an unfurled cadential §.*°
It may even happen that a cadential §chord never appears as a §. In “Am
Feierabend” from Die schéne Miillerin [1.34], the bass pursues a direct
stepwise path towards E, A Majors dominant pitch: Gf<A<B<C<D<
. . (These notes are circled in the example.) Instead of continuing to E,
however, the bass descends by step from D to C for an unfurled cadential °
chord. This modification likely results from the fact that the vocal melody is
itself ascending from D to E, which is unconventional for this point in the
large-scale structure. The melody emanating from E>D in measure 71 more
 Harmonic progression BD

69

— - 2 Ny)
—— =e =
rs
S —— 7 a 7 ze |
dass die sché - ne Miil le - rin merk - te
2 er tle pet oy Ne PES « eft fe
2

—= ie *

A Minor: V]7———

mei - nen, mei - _ nen treu - - en Sinn,

a
Js = gg 3 Se ase Se aS SriCaer me ==
———| ——_
pe =ee
—_—_—_*. yt

G 81
ml Ht Ve 3 I
7 8 7

Model 1 Model 2
m. 69 71 2 B m. 69 71 2 B

a ee
5 nw A A A A nw nN A A

as Sap oN 54 (3) 2) 4

<r gs Taam S ONE Sere ee S

6 6 6
5 5 Sh

AMinor:I I UI I A Minor: I Il UO < ih

BD —

1.35 Analysis of “Am Feierabend” from Die schéne Miillerin (D. 795/5), mm. 69-73.

typically would descend via C and B in measure 72 to A in measure 73 for

a straightforward close [1.35, Model 1].°' The vocal E in measure 72 [1.35,
Model 2] highlights the potent word treuen (faithful, devoted). Though C
and B are absent from the vocal melody during these cadential measures
and during the repeat of the phrase, which follows immediately thereafter,
C>B>A occurs several times — in both the vocal and piano lines - in the
concluding measures of the lied.
36 Harmony in Schubert

“Ungeduld” from Die schéne Miillerin (D. 795/7)

Die schone Miillerin probes the realm of feelings relating to love: ulti-
mately unrequited and tragic, though in “Ungeduld” (Impatience, the
seventh of twenty poems that Schubert set for the cycle) still ebullient
and hopeful [1.36]. The piano introduction abounds with perky disso-
nances that crop up unexpectedly and quickly dissipate, enhancing the
momentum developed by the persistent triplets. The harmonic progres-
sion likewise supports the forward flow, with its numerous descending-
fifth motions:
m. 1 2 3 5 6 Wi 8

[NQO—As

F# BOE A— D° 8
B> E

The tonic-expanding progression of measures 1 through 5 presses towards

IV. The local dominant prolonged in measures 3 through 5 sacrifices
its leading tone in anticipation of the following tonic’s minor seventh:
G#>Gh.°? IV undergoes two evolutionary transformations: first a 5-6
expansion™! (D-F#-A to D-F#-B in measures 6 and 7); then, just before
the dominant arrives (measure 8), dominant emulation (D#-F#-A-B). The
introduction thus projects the following harmonic structure:
m. 1 2 5 4 5 6 7 8

A Major: I es Lye aap NE

e ie ae ee (= 11)
In four of the introduction’s eight measures the keyboard melody
presents a downbeat descending second, with dissonance resolving to
consonance: F#>E, E>D, C#>B, and A>G# (covered by F#). The vocal
melody of measures 9 through 18 is even more fully saturated with such
embellishments, on various beats. The A-D*°-E progression of meas-
ures 1 through 8 is reprised in measures 9 through 12, while the local
A°*°-B-E-A of measures 1 through 5 appears in measures 14 through 18.
(A#-C#-E-Gb in measure 15 is a chromaticized 6 phase of A°~°.)
The C# chord of measure 12 is curious, likely causing attentive listeners to
wonder if Schubert is traversing a segment of the ascending 5-6 sequence
[1.37].°° Given Schubert’s continuation in measure 13, however, we under-
stand in retrospect that measure 12 does not replicate the construction
of measure 10 up a step, despite the similarities. Instead C#-E#-(G#) is
 Harmonic progression 37

a passing chord within the realm of dominant E, which is prolonged by

means of several concurrent linear connections among the dominant’s
intervals: B<D, G#>E, and E>B [1.38].°°
Beginning at measure 19 the contents sharply contrast the preceding
material. The descending seconds cease, as does the swift pace of chordal
change. A single B chord is prolonged (in two phases: with seventh and
raised third and then with, in addition, absent root and minor ninth) in
measures 19 through 22, which include several unprecedented half and
dotted half notes in the vocal melody. In all four stanzas this passage sup-
ports the exultant text, “Dein ist mein Herz und soll es ewig bleiben” (My
heart is yours and ever shall be). In that parts of the text are repeated,
this one line occupies nearly as much musical time as all five preceding
lines of each stanza, and thus its conviction is powerfully emphasized.
Melodically lively, several chords are arpeggiated in full, and a high A is
twice attained. Yet underlying this vigorous activity is a basic progression
that moves diligently towards closure [1.39]. The analytical model tidies
up the flamboyant melody, placing the high As in the normative register
as connectors between B and G# (first as the supertonic’s seventh, then as
a suspension resolving to the dominant’s third), and likewise curtailing
the venturesome bass of measure 23, restoring dominant root E for the
conventional cadential § that underlies Schubert's bold unfurling. This
is not to suggest that Schubert overstepped propriety in his writing; but
instead to assert that his creative endeavors develop out of an exemplary
basic structure.
Measure 25, not shown in the model, contains a parenthetical passage.
The path towards closure upon which Schubert was already embarked is
insufficient to contain the protagonist’s joy and commitment, and so we
hear an emphatic reinforcement, with another approach to the dominant
root in the bass (C#<D<E) and another approach to 2 in the soprano
(E>D>C#>B). The redundancy in both words and musical structure exudes
conviction. The parenthetical addition echoes beyond the singer's final
utterance after the last stanza.
Harmony in Schubert

dé i Etwas geschwind

Sno

eg
p>
Dans’ re
?
A Ft B E

_—
:

iSesel== aa
e YS

BERS Sescepa
ot stet a =

A D

a4
= eens See Ss
schnitt’ esgern in al - le Rin - den ein, ich grtib' es gern in je - den
mécht' mir zie - henei - nen jun - gen Staar, bis dass er sprach' die Wor - te

ae =e
pts
oe Ss
SS
== = t = ‘t =re

f é
—E————— jae gill

A D B

12 ‘

(oa
4 : :

Kie - sel- stein, ich mécht' es sa'n auf je - des fri - sche Beet, mit
rein und klar, bis er sie sprach' mit mei - nes Mun - des Klang, mit

(se ©ee Pare tere ee ee ce ee ae

gine ee === ===
E A

1.36 “Ungeduld” from Die schéne Miillerin (D. 795/7).

 Harmonic progression 3)

SSS SS ee

Kres - sen- sa - men,der es schnell ver - rath,

s

auf
\\——e:

je - den wei - ssen Zet - tel

aS ——o

mei - nes Her - zens vol - lem, hei - ssem a

ge
“a sang’ er hell durchih -wes
gt
= aESdES==
rut
———— ¥ = ==
Fe E

18

tits =o SoH
Saas aSy a ; 5
e
mécht' ich's schrei - ben: Dein ist mein Herz, dein ist mein
fs‘ia - ster schei - ben: Dein ist mein Herz, dein ist mein

ts ss= eerie
== t B
, =]
; =e2SSSe5
ceee
6 $323. dorrrrre
=———
3

pags —— pe te 2 Be ae: Z z
. 2 Sete =

A B

tte =
ee
7 Zz z aoa
:
a
Ne
~~
Herz und soll es e Witgae =e - wig__
Herz und soll es e Wig, ue - wig__
_—— —— = = Se SSet ee
a:
a
oan — z
$5555 E eee nent zt

4 . 3 3

ST cea
5 o- o—-—* ‘ ¥ S eH
= SSS ts

o =

te sagas = aa
a
raya
p7
aac

1.36 (continued)
40 Harmony in Schubert

m. 9 (1G) AP (13)

ay Z ss Z 2
23 Ee {|
6 5 6 5—6 5
A Major: I ( ) VI

1.37 Analysis of “Ungeduld” from Die schéne Miillerin (D. 795/7), mm. 9-12, with
hypothetical continuation.

rat, 8) 10 1] 12 13

S (ee ae
eee —— ie
A Major: I [Veen CV
(= i)
1.38 Analysis of “Ungeduld” from Die schéne Miillerin (D. 795/7), mm. 9-13.

m 180 19) 21 23 24 26
i ome 1
—
‘ ot He Se 2
; 4h
aye be oe (I =
o

ae 7a a
AgMajor: te hr V4 3 I

1.39 Analysis of “Ungeduld” from Die schéne Miillerin (D. 795/7), mm. 18-26.
2 Linear progression

Parallel progressions

The passage from “Nachthymne’ presented in 2.1 occurs during the traver-
sal of the tonal space between the D Major tonic and its dominant. B major,
a chromaticized 6 phase of tonic (= VI#), arrives in measure 36. At first it
is tonicized, but soon (measure 39) added dissonance propels the chord
towards II# (E major). That resolution occurs and is reconfirmed (through
repetition of the B’—E chord succession and of the corresponding text)
at the end of the excerpt. The initial confirmation of B major and the con-
cluding B’—E succession employ no especially noteworthy techniques. In
contrast, the conversion of the B major chord into a dissonant, E-seeking
entity is a magical moment, one that Schubert invites us to savor through a
slowing of the tempo (Langsamer werdend). Dominant emulation, already
initiated through the B chord’s major quality, is heightened in measure
39, with D#-F#-A-Cs. Instead of proceeding to an immediate resolution,
however, Schubert allows the lines to float gently downwards, in a parallel
progression (or chordal glide). This is an astonishing effect, for it seems as
if the resolutional tendency of one of music’s most potent chords is being
disregarded. Bass D# “should” resolve upwards to E, but instead Ds, then
C#, and then C follow. The trick is given away in 2.2: because a diminished
seventh chord contains nothing but minor thirds (or their enharmonic
equivalents), if all voices slither downwards three half steps the original
pitch classes are reencountered. (Some suspensions occur in one of the
inner strands.) Thus in retrospect Schubert’s device turns out to be a pro-
longation of the initial diminished seventh chord, not its repudiation.’ The
upward-tending bass D# reappears as E} in measure 42 and, after a restora-
tion of the D# spelling, resolves in the middle of measure 44. Neither of the
internal diminished seventh chords asserts itself as a harmonic entity.
In a passage from the “Wanderer” Fantasie in C Major for Piano [2.3],
eleven diminished seventh chords appear in direct ascending succession
(measures 101 through 104), with in this case no suspensions to offset
the uniformity of the sweep. Instead we encounter raw power and numer-
ous sforzandi. A diminished seventh chord spelled C#-E-G-B} emerges
42 Harmony in Schubert

34
te

ter nach dir.

SS SS FSS —=ES FS == SS SS
oe Severe Sececde| feccdsdssts a

ThA
NT}

Langsamer werdend.
2

An je-nem

tte

det
ie)

i 8

in &a!

=I
al
P
—
5
N|

||
oe

mo}
Hl}
og
‘Oo

D
is)

me
Uh)
2
eia}

|
N

oO

N
oS
-

g
Mg
‘

ry
¢

3
nywoos
K3a

z
ein Schat
- ten brin - get den kith len-den Kranz, brin-get den kiih len-den Kranz.

Agi © Nachthymne” (D. 687 )s mm. 34 —45.

94
4
=
Cc
D Major: VI}
<

)
2.2 Analysis of «“ “Nachthymne” (D. 687), mm. 38-43,
 Linear progression 43

Pemconee teGhe 18tt Gait hisses tS As a a eee ee

ft, ic eS ie. fs 2 ig

= posse Pases== ESS See, = f Snes

P —_
si 22 a cresc. £ s ms i
> = = — ai =e 2 ~4 Z <

101 We b
i Pg Hd
;
——— a =
esi _——| £ e if a te E
pee areas : eEes=—= aoe reewaee

= =
pale t oa

——— ee
“qian ‘ehcp
eS Se ae

C Major: inadi TL,

2.4 Analysis of Fantasie in C Major (“Wanderer,” D. 760), mvmt. 4, mm. 98-106.

in measure 101 after a three-measure expansion of the C Major tonic.

Whereas the spelling E-G-B>-D}> would suggest a direct extension of the
C tonic (absent root, lowered seventh and ninth), Schubert's choice corre-
sponds instead to root A. The succession from a C Major tonic chord to its
6 phase, A-C-E (placed in parentheses because its consonant, diatonic state
is elided), is displayed in 2.4. The initial A chord appears only in its evolved
state C#-E-G-Bb (measure 101). An ascending parallel progression serves
as connection to G#-B-D-F (measure 105), which embellishes the A chord.
44 Harmony in Schubert

Though a descent of two half steps or an ascent of one half step would be
sufficient to lead to the appropriate pitch classes, Schubert here ascends ten
half steps.
Analysts who attempt to make harmonic sense of each chord within a
glide inevitably will be frustrated, because a parallel progression is a linear,
not a harmonic, operation. Only its endpoints participate in the broader
progression.
Schubert’s glides make use of other chord types as well, especially $
chords. In measures 22 and 23 of “Die Erwartung,’ the harmonic succession
from I to first-inversion IV is followed by a descending diatonic parallel
progression that spans a seventh, landing on first-inversion V [2.5]. One
of the glide’s internal chords is incomplete (Bb-D-Bb, in place of Bb-D-G),
offsetting potential tedium. Later the glide’s first three chords are reprised,
this time at a more luxuriant pace and emphasized through the repetition
of the word zarte (tender, delicate). The delicate foot mentioned in the text
is that of the beloved one, whose steps Schubert conveys through stepwise-
descending chordal glides. In measures 24 through 26 the sense of IV>°
(an untapped potentiality during the first two beats of measure 23) emerges
strongly. Because the 5-phase IV chord appears in first inversion, the glide
of descending § chords effortlessly leads to IV’s 6 phase, which, asserting
itself as II, unfurls into 3position before the progression continues with V
and I.

The ascending 5-6 sequence

Some chords are bearers of a harmonic function; others are not. A central
point in our discussion of “Nachthymne’” [2.1] was that its four consecutive
diminished seventh chords do not all play a harmonic role: the enharmoni-
cally equivalent first and fourth chords participate in the passage’s defining
harmonic progression, while the two internal chords play a connective
role. A similar diversity of roles is inherent in the context of a sequential
progression. The chords that parade past are exactly the same entities that
Schubert might engage at other times in forming harmonic progressions.
In the context of a sequence, however, their harmonic potentialities are not
asserted.
In Schubert's Symphony No. 3 in D Major, the passage that concludes the
minuet sections of the minuet and trio form emphasizes the movement's
D Major tonic in measures 41-43, 48, and 54-56 [2.6]. The addition of
Ck to the tonic chord in measure 49 and the G major chord that follows
 Linear progression 45

22 5 4

as Es

gen
y =a
SSS
wenn sei - ne sché-ne
Biir- de, leicht be - wegt,
Le
der zar
:

- te

Gry fot Foal eae ree a “ee Sores

f= ean

ee om a ee ae
Cf (ZS 2

ae ae
oft
vg y ee
SiS eae
SS BS a = —
See Se
=
; 6 6 6 6
Bb Major: I ee Veen
en IV ( ys aL )

33
6? 2222
te; der zar - te Fuss
eo St e
zum _ Sitz_der Lie -
oi
be
= es a

ot eee ype
ee eee °
ms on == ——=S}| as

» ef be bebe
aa
Be
= of —— =
eS eS
Se
Be— a : zi :
ee
eh tos 8 7—
) Me ere
; (11)
Bs
28
: (Gaay

tragt.

Sa z =
6 ce Saari ae =
(OX

eae g eo
a o

2.5 “Die Erwartung” (D. 159), mm. 22/23-29.

encourage a harmonic interpretation, I*”*—IV.’ Yet as the progression

continues we come to understand that, instead, a sequence has been initi-
ated. Thus what we may have thought was certain becomes uncertain, and
we suspend judgment about how these chords relate until we know where
the progression is leading. The sequential ascent ultimately reaches B minor
(measure 52). That arrival point is highly conventional: it is the unfurled 6
phase of I~.
 2.6 Symphony No. 3 in D Major (D. 200), mvmt. 3, mm. 40/41-56.
th

Pa es as sme . alt
(Cit ( BS
i) Ap
4 rN
it
i I
| het ota
a sits iis dallas Mlle : <i a
CH oll elles
qa] Goll GW Tit CNH “All =A ci is Q
oH we UUaE Ny re Ty Sa
“ we fee Mbades Hile. ||] ot i & Hille
fla, |f Hill» |f f Malls |! ' hy Te ST] SH
alll
;
Cui Cray Th
UID 1 +t CH fe rT
tH ™ il finn & [Mlle
inhi a ttt LI ahd a
“Aa Ht at, MUL fillesHlth I): HS Tae
NY nt} tl ow ly Ae City GE | Hi ( » Uh ¢
ane Med
hi
We
i ih
We NY
| Mh alll, ae it it .) nt i te il
ee
i tt I Hf
oe aah
ie } } Ped 4 alll 1 } Hs [MS
sll geal] shrms Fr cs
Tbe,gta, HATS GHALSS sii &
AM Int ant tk AY NY
tnt NY
BELG TE
» eu ty ttt oH He Te
| | | all Ie NY
Harmony in Schubert

IL!
ay ( |) ( Db [| ) tt tip
HS Un Sips es aS “UT ae a - TH 1] 1] 1 HT
we nen.
Np all, She She Bis
aie)
Hee
L
Hu a Me oll a 1 hl a i] 1 Hts, 1 J a |e & ng
A a. dl. ge
46
 Linear progression 47

Model 1 Model 2 Model 3

Gis — : 2
2] ==
= $ 8 3

D Major: °° D Major: P—° D Major

1°:°
Model 4
— 4
=
& Grass e el.=
eles ¥
DimlEr SG: Ae eB
Model 5
ma. P4S, 49.650 9. 51 52 53 54
ae Br ly a 2

Ge
ee
ie 3 ; tg
ee =
eeS 24
ee= Zl

D Major:
. —
I ( je
—6
av
1 (O==5)
et

2.7 Analysis of Symphony No. 3 in D Major (D. 200), mvmt. 3, mm. 48-54.

A derivation for the sequential passage is proposed in 2.7. Given that

II#-V—I concludes the phrase, the opening I is ripe for expansion via a
5-6 shift (Model 1), a common antecedent of II. As with any chord con-
taining a sixth and a third above its bass, the 6 phase may be unfurled into
> position (Model 2). Though the bass succession from D to B normally
would transpire as D>B (the shortest route), a space-opening D<B is also
feasible (Model 3). The ascending 5-6 sequence is an attractive means of
traversing that space, though its full course would be quite extended and
risk tedium:

D°*-6 F>-6 |sive G>6 A>?6 B

Schubert here abbreviates this trajectory: since the sequence’s first (D°) and
sixth (F#°) chords contain the same pitch classes, one may meld those two
chords and omit the sequential cycles that normally would come between
them (Model 4).* Schubert fortifies this progression with added chromati-
cism, dissonance, and the unfurling of the § chords, imbuing the passage
with grace and a gentle power (Model 5).
Some readers might be inclined to analyze this sequential progression
in the same manner as a harmonic progression, perhaps even invoking
a number of subordinate keys (G Major, A Major, B Minor) before even
the phrase’s supertonic chord arrives in measure 52. The perspective
that I advocate focuses instead on how chords work in the service of the
48 Harmony in Schubert

mal 2 3) a4: Sie aG 117 118 119 120

ee ee
aa 2 e ‘os fa

—z—s ote ee Z |

SSS
:
P = © See
a
ee Z

AG(Co). Dp eb AE
Ab Major: I°~( eine Ne

2.8 Analysis of Symphony No. 4 in C Minor (“Tragic,” D. 417), mvmt. 1, mm.

112-120.

broader tonal plan. The observation that a pitch behaves in the manner
of a leading tone does not necessitate that its chord be interpreted as V
in some key. Instead, any chord in a descending-fifth relationship with its
successor may be molded through chromaticism and added dissonance so
as to emulate the dominant’s character. In a sequential progression, which
may contain numerous descending fifths (or ascending fourths), a recur-
ring voice-leading pattern — not harmonic thinking — guides the chordal
trajectory.
Schubert reprises this sequential progression in the first movement of his
Symphony No. 4 in C Minor [2.8]. Here even richer chromaticism than that
from the Third Symphony is employed, and yet the analysis remains firmly
grounded in the key of Ab Major.
Though an ascending 5-6 sequence could be pursued for an entire
octave (or more), Schubert's artistic sensibilities generally steered him
towards use of the technique in more moderate doses. In the first move-
ment of his Piano Sonata in Bb Major, he traverses the span I<IV, adding
some novel twists along the way [2.9]. The passage quoted begins with
a straightforward I-II-V’-I harmonic progression. The I<II connection
incorporates a 5-6 expansion of I. Its 6 phase is unfurled (bass D), made
dissonant (added C), and chromaticized (F# in place of F) in the service
of dominant emulation.° As the passage continues, the initial 5-6 shift
(connecting chords on F and G) is replicated in an ascending trajectory:
after G comes Ab, A, and finally Bb. (After the initial F-G whole step,
the sequence proceeds by half steps to its goal. I classify such a sequence
as idiosyncratic, in contrast to a diatonic sequence, which proceeds by
diatonic degrees, or an obstinate sequence, which proceeds entirely in
half steps or entirely in whole steps.) Consequently the latter G chord
(measure 103) does not function as II, but instead is an internal compo-
nent within a sequential ascent. Observe that the procedure is simplified
in the latter cycles: the pattern’s descending third is filled in at first, in
 Linear progression 49

e THe §apps
=e es
Sae
& £2oo vi
So aaa

ora
= = 5—tea==s = —SS
= Se
F Major: I° 6 v! I =
Na!
103 ‘ 2 es Sas

Baers
eS \ igs og ite 2 #
: =e = 4 yi = —=— = =
LS —— crese. It mf == ———

De S~ a iz: i S # —
eee: oe
os = é fe % : © 7 =
= <5
F> 6 Ge 6 Ab> 6 A? 6 Bb

aaa | ease bY
108
esses ge
o =F qe

P ee —
b = ai -g ie
o re ze + 2 rane 2

Meee il

2.9 Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 99-110.

m. 103 104 105 106

$15 —42 B
ol
bg be be > tele

a ae Ge iNieee —— Bb

F Major: I ( adel:

2.10 Analysis of Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 103-106.

measures 103 (F>Eb>D) and 103-104 (G>F>Eb), but the next two thirds
(Ab>Fb[E] and Ay>F) are presented in direct succession. Observe also
that a higher register is gradually attained. One of the bass's “descending
thirds” (G>E}b) sounds as an ascending thirteenth, while another (A>F)
sounds as an ascending twentieth. The essential elements of the sequence
are presented in a normalized register in 2.10.
Analysts who take a harmonic approach to sequential progressions will
be concerned about the implications of the enharmonic shift that must be
50 Harmony in Schubert

From a major triad, diatonic From a minor triad, diatonic

6 aes _A o=aea
5——— 6 5 5——— 6 5)

From a major triad, whole step From a minor triad, half step
with dominant emulation with dominant emulation

ab DZ
6 ie 6 | re
5—— 6 5 5 —— 6 5)

From a major triad, half step From a minor triad, whole step
with dominant emulation with dominant emulation

6a 6 Sas
5 6 5) 5 6 5)

From a major triad, stretched

ae
with dominant emulation

Sao 5

2.11 Models for an ascending 5-6.

invoked here to land on Bb rather than Cb. I propose treating sequential

progressions in the same manner as parallel progressions, which (as we
have seen) may ascend or descend by half steps. Though the notation is
cumbersome, there is a free linear flow between meaningful points within
the broader harmonic trajectory (here I and IV). Ascending in minor
seconds from G in our notational system proceeds as G<Ab<Bbb<Cbb
(a problem that must be fixed). If our notation were instead a neutral
modulo 12 system (with C = 0 through B = 11), that ascent might appear
as 7<8<9<10 (not a problem). Our notational system, which contains seven
basic elements — noteheads, letter names — per octave, was developed with
diatonic relationships in mind. The notation that Schubert was compelled
to use for 2.9 is inelegant: though it succeeds in informing the performer
which keys to push when, he was required to engage in corrective measures
to correlate the local sequential and framing harmonic initiatives.®
The first movement of Schubert's Piano Sonata in D Major offers an
astonishing example of sequential deployment over a long span. Consider
first the models of ascending 5-6 cycles shown in 2.11, which displays
both diatonic and chromatically modified progressions. Usually dominant
emulation in the 6 phase requires a chromatic adjustment - either in that
 Linear progression Sl

m. sy Sey IHG Swi 117 118 124 125) 132 133 140 141

m3 M2 M2
D Major:
[°=

142 143 144 148 150 Si 52

fa Hale saa baltaligile)= 'g 4 2 NS a

|e 2 F#

—7h ah eae

(= 5s)
instead of .
m. 1 141 142 143 144
4
=i
Gt 3 ge
elision

D Major:I

2.12 Analysis of Piano Sonata in D Major (D. 850), mvmt. 1, mm. 1-152.

6 phase, or in the chord that follows if it is to emulate the tonic that the
preceding chord targets, or in both.’ The models in the second and third
rows of the example confirm that both half- and whole-step motions with
dominant emulation (C<D}> and C<D) are feasible starting from both
major and minor triads. The final major-triad model, a “stretched” version
of the procedure, traverses a minor third (C<E}). Like the final minor-triad
model, its first two chords share no common tones.
The intention of 2.12 is not to be a rigorous reductive analysis of
Schubert’s score. Instead it demonstrates how an extended sequential
expansion of tonic permeates the exposition and development sections.®
The sequential progression is emphatically presented in the example, with
a normalized soprano ascent and abundant § and § chords of dominant-
emulating character. The core idea of the tonic expansion is the notion
that an ascending 5-6 sequence’ first and sixth chords share the same
pitch classes: D5-(° E>-° F#°)*. (I referred to this relationship earlier, with
regard to 2.7, Model 4.) Schubert expands this D<F# tonic third into a
tenth — measure 1 to measure 144 - and utilizes cycle spans like those
presented in 2.11: three minor thirds, then two major seconds, then three
52 Harmony in Schubert

minor seconds.’ The half-step cycles not only ascend more gradually, but
also proceed more quickly, at the rate of one cycle per measure (measures
141 through 144). The more frenetic pace even induces the elision of F#°
at the downbeat of measure 144, an event that pushes F#-A-D (the reitera-
tion of tonic) into the limelight. The long-prolonged tonic then takes on
dominant-emulating characteristics (measure 148), leading the progression
onwards to IV and then V.

Circular progressions

Occasionally all seven of a key’s diatonic chords will sound one after the
other. We encountered that event in 2.5, where a descending parallel pro-
gression of chords in §position transpires within a single measure, connect-
ing IV and V. Likewise the familiar circle offifths progressions — ascending
or descending diatonically - will touch upon these same chords, though
in different orderings. There is a hint in 2.5 at yet another strategy for
traversal: the glide’s first, third, fifth, and seventh chords are metrically and
durationally emphasized, outlining a segment of the descending circle of
thirds progression: E>>C>A>E. The full circle is rarely encountered, though
Schubert employs one in his Characteristic March No. 2 in C Major [2.13].
In this diatonic rendering each chord shares two common tones with those
that surround it. A passing note fills in each bass melodic third. The bass
confirms tonic by extending from C back to C. Each of the upper strands is
a filling-in of one of the tonic chord’s intervals: C<E, G<C, and E<G.
Any circular progression balances two contrasting operating princi-
ples: the maintenance of diatonic pitch content, and the maintenance
of a pattern of chord qualities. For example, a descending circle of fifths
in C Major begins CE. At that point the composer confronts a fork in
the road. A Bb major chord logically belongs next, with FB} maintain-
ing both the interval between adjacent roots and the innate dominant
emulation that has already occurred in CF. In contrast, the diatonic
B chord is deficient, attained via a dissonant melodic interval (F>B or
its inversion) and containing a dissonant simultaneity (j)- Yethitticwa
member of the diatonic club, and accepting it despite its faults is often
preferable to the obstinate alternative, a string of increasingly chromatic
chords (C-+>F—>Bb—E}—Ab—Dh-. . .) that simply delays the correc-
tive (for example, D>~-G->C may occur in this context) or instead seem-
ingly leads ever further from tonic. Dbb, the progression’s thirteenth
chord, is often hailed in harmony textbooks as tonic through enharmonic
 Linear progression Do

ks ~|

ezaedaete: cae aS ee Se
p_ $s t# $5 £6 a a et
Jp Chesca

NS vu eT fF Mt
YW
han Bhd
= ee f° i @

a eee ———
= =
P crese.. — ¥ |

aay SS ea :
C Major: I

2.13 Characteristic March No. 2 in C Major for Four-Hand Piano (D. 968b [886]),
mm. 287-290.

reinterpretation, as I think composers who pursued this construction likely

intended.'° Under that premise the notation, rather than the conception,
is the complicating factor. The tonal system, with its inherent bias in favor
of the seven diatonic pitch classes, is severely challenged by such circular
progressions, such that an enharmonic corrective becomes mandatory. If
the diatonic bias is eliminated through use of the modulo 12 system (intro-
duced above), then the circle of fifths could be traversed without inter-
vention as 05-1038 16114-9270. Alternatively,
some listeners may interpret the progression as a spiral that spins further
and further from tonic. From that perspective a composition may begin in
C Major but end in Dbb Major. I suggest that when a progression of twelve,
six, four, or three equivalent motions returns to the enharmonic equivalent
of the starting chord, the composer is relinquishing the diatonic seven-note
system and entering a chromatic twelve-note system, which conventional
music notation inevitably will accommodate only with difficulty.'! As is
also the case with parallel and sequential progressions, I do not regard the
interior regions of circular progressions to be harmonically motivated, and
thus do not employ Roman numerals in their analysis.’
In a circle of thirds there is an uneasy interface between the diatonic pro-
gression and the obstinate maintenance of the initiating chordal relation-
ships (an alternation of major and minor triads), which would inevitably
traverse some chromatic terrain and induce some curious modal shifts
(descending C-a~F-d-Bb-g-Eb-—c), and perhaps spiral further and further
from tonic (ascending C-e-G-b-D-f#-A-c#-E-g#-B-. . .).
Harmony in Schubert

=
ye os
nh {xi

lass die Lei . den - schaf - ten

W

SE
Teme =
: 2 322 e Sree
< ~e

C Major: I (

49 S °

ee sau. - sen im me -tal - le-nen Ak-kord,

=!<SSeka ; : 7 |

DE a = = : rs

Oo *s OF F3 @*" =e
ae eae ae
2.14 “Aus Heliopolis 1” (D. 754), mm. 47-51.

The descending circle of thirds progression that Schubert deploys in “Aus

Heliopolis 11” [2.14] differs from the Characteristic March's progression in
several respects.'* Observe that the circle is not complete, but instead pro-
ceeds only so far as G (measure 50), at which point the harmonic succession
from V to I concludes the phrase. This alteration offsets one of the principal
weaknesses of the circle of thirds: because each bass melodic interval is a
third, cadential effects are limited.'* Also note that the B chord is modi-
fied. Instead of its normal diminished quality, here (measure 49) it shines
in major splendor: its third and fifth are both raised. The F# proceeds to the
following chord’s G, invigorating the keyboard melody’s E<G linear ascent
over the course of the progression. The chromatic inflection of D to D# is
less purposeful. Since Ds is restored immediately thereafter, D# functions
as a wobbly note, adding some color without compromising the B chord’s
role within the circular progression (rather than announcing the impend-
ing arrival of an E chord, as some analysts might suggest).!° The A chord’s
C# (measure 48) is also a wobbly note. In all, five of the six chords that
constitute the circular progression connecting I and V are of major quality.
 Linear progression 55

id ere 29 3] #3. 185 37 39

g * 43
a S|
as

Cae!
$ = : — m2)
2) 3 t
P : Sse 7-) = :
G Minor: V3z( alan;

2.15 Analysis of Divertissement a la Hongroise for Four-Hand Piano (D. 818), mvmt. 3,
mm. 27-39.

A mirror image of the “Aus Heliopolis II” structure appears in the

Divertissement a la Hongroise [2.15]. Whereas the lied descends five
thirds from the tonic chord, the Divertissement ascends five thirds from
the dominant chord, reaching an unfurling of the dominant’s embellish-
ing § in measure 37. The routing of the ascending circle deviates from a
diatonic path: instead of G Minor’s d~F—a°-c-E}~g or tonicized D Minor’s
d-F-a—C-e°-g, Schubert begins with two consecutive minor chords,
thereby reorienting the entire ascent, which proceeds as d-f—Ab—c-E}-g.
As a consequence, no diminished triads are encountered.
Without going into detail at present, mention should be made of two
circle-of-thirds variants that were of special fascination to Schubert. The
normative pattern for the progression is an alternation of major and minor
chords, always maintaining two common tones, as in

C-E-G
E-G-B
G-B-D “etc:

Proceeding in this manner one must eventually either break the pattern to
stay within the diatonic confines of the key (e.g., major G-B-D to dimin-
ished B-D-F) or extend beyond the diatonic pitch collection to maintain
the pattern (e.g., major G-B-D to minor B-D-F#). In contrast, Schubert's
ascending circle in the Divertissement [2.15] contains a chromatic succes-
sion at the outset: from D-F-A to F-A}-C. A continuation of that pattern
results in a progression of minor chords: d-f-ab/g#-b-d (or, in modulo
12, 2-5-8-11-2). (The location of the enharmonic shift in the context of
modulo 7 notation is of no consequence. It is arbitrary: f or ef, ab or gt, ch
or b, ebb or d.) The other novel variant is the segmentation of the octave
into three major thirds: C-E-G#/A}-C (or 0-4-8-0).'° These progressions
may be pursued in either the ascending or descending direction. In my
view they function in the manner of a parenthetical passage: the endpoints
56 Harmony in Schubert

participate in the broader harmonic initiative, while the internal compo-

nents are linearly conceived within twelve-note chromatic space (rather
than seven-note diatonic space). Passing through an “enharmonic seam’ is
merely a notational matter, the result of using an ill-suited set of symbols
developed to accommodate seven-note diatonic composition. Analysts
who interpret circular progressions in the same manner as harmonic pro-
gressions face difficulties that disappear if one accepts that a twelve-note
conceptual space is being forced into a seven-note notational space.

Chromatic lower-third and upper-third chords

The words “mit unwiderstehlicher Sehnsucht” (with overpowering longing)

are the climactic moment in Ladislaus Pyrker’s poem, “Das Heimweh”
(Homesickness). A mountain lad has been transplanted to a town in the
plain. He seeks refuge in the hills beyond its walls, where he contemplates
his native mountain peaks, visible in the distance. Schubert sets these words
dramatically, juxtaposing A major and F minor chords at the apex of a
crescendo extending beyond fortissimo [2.16, measures 188 through 191].
The 5-6 shift is the generator of lower-third chords. The 6 phase may
either set a progression in motion (as in °° I) or serve as a local embel-
lishment (as in I>-*°), In 2.17 the possibilities are shown for consonant 5-6
realization of the tonic chord in C Major: diatonic C-E-A and dominant-
emulating C#-E-A, plus two more potent versions whose alliance with the
parallel minor key (also shown in the example) is more pronounced than
that with the initiating tonic. In many cases a lower-third chord has no
independent harmonic meaning, since the pitch corresponding to 6 may
function as an anticipation or as a neighbor without asserting itself as a
root. This chapter segment’s examples display such chords as the conse-
quence of melodic embellishment, in which neighboring motion (a dia-
tonic or chromaticized 6) and up to two wobbles (chromatic inflections of
the root, the third, or both) occur concurrently.
“Das Heimweh” deploys the A major chord’s third - and most extreme
- chromatic variant.'” All three pitches shift by one half step: one as a
neighbor, two as wobbles. Yet the lied is in the key of A Minor, as is evident
both at the beginning and at the end of 2.16, and thus the A major chord
of measure 188 is itself chromatic. Instead of presenting the direct connec-
tion of A minor and F minor chords, in which only one chromatic pitch (a
wobble from A to Ab) is engaged, Schubert instead swerves first to A major
(chromatic in the sharp direction), whose relation to F minor dynamically
 Linear progression

s =
z + Bu 9— 5 —
- if 7 t t t
— ee ~_

dopeteded

DP ud = t aa 4% = epee

a
[se oe7Soe v=
zieht ihn da - hin m

2.16 “Das Heimweh” (D. 851), mm. 182-193.

gy Harmony in Schubert

Diatonic Chromatic Chromatic Chromatic

Variant #1 Variant #2 Variant #3

coirGg 386 AES ae

Diatonic Chromatic Chromatic GEhromatic

Variant #1 Variant #2 Variant #3

2.17 Diatonic 5-6 and three chromatic variants.

90 2\ -
a P

Se ist das Herz, se-ligdoch hal -tenwillich den Schmerz,

ob auch zer - spal-ten mir

|Fei == re

esa =
Ab Major:

2.18 “Fille der Liebe” (D. 854), mm. 89/90-93.

conveys the text’s notion of “overpowering longing.” The F minor lower-

third chord is itself extended in measure 189 by neighboring notes applied
to its Ab and C, embellishment that recurs in the context of the tonic chord
in measures 191 through 193.
In a similar passage from “Fille der Liebe” [2.18], a major tonic prevails
throughout. An F> [E] minor lower-third chord brusquely contrasts the
Ab major tonic chord.'® As in “Das Heimweh,’ the dynamic range peaks
beyond fortissimo and then quickly decreases to piano, the neighboring
chord is itself embellished by neighboring notes, and the text is emotionally
intense: “Though my heart is splitting |I will regard the pain as a blessing”
Their enharmonic respellings mask the roles of these pitches: Fb [E] is a
chromatic neighboring note, while C} [Bk] and Ab} [G] are wobbly notes.
In “Lied der Mignon” a chromatic lower-third chord embellishes the
dominant, rather than the tonic [2.19]. Dominant F# major and lower-third
D minor (measure 38) have no common tones, and thus their relationship
corresponds to the third chromatic variant displayed in 2.17. As has been
 Linear progression 59

36

° . iS SSSr ty +5 — fya ae-

oe re
io 9 an
‘3 os
aS aie —+ > ee ar oe
Zwar lebt' ich oh - ne Sorg und Mii - he, doch fihlt' ich tie-fen Schmerz ge- nung, vor Kum-mer
— — “= —
es . e ; a> +s £ e +ce ty i. i5O—s 4S ¢ 4
a ee = =
Oe ee a os z at + ~e st ae 2g = =
PP fo a —— Zs Wi mie BO es eos

sy = N é i aA ar ree
es
= =—2=: 7
oe a
See= ot ==:
es ee
=
— —
eee
= Ss aase 5
4
= qe 3

1)79: 8
Vi=,Z 6
——<=

Se — ————————— : = :
= meee aes = StF
7 74 4 ==
zu - frii - he; macht mich auf e - wig, auf e¢ - wig wie-der jung!

ee eee eee eee

$ 4b. td ig g 2+
{+s = —s = se a o
1 2 =

a e@ e ° a=: =e: + Z :
¢ 4 — o er soa et sf
=: 2=

5 —6: 5 ae 7
2 4 2 I I: V4 3 I

2.19 “Lied der Mignon” (D. 877/3), mm. 35/36-44.

the case in our other examples, a crescendo leads into the 6-phase chord,
which here arrives at the peak fortissimo moment. Also typical are the
words that this musical gesture supports: “yet I felt intense grief enough”
In that the somber sentiment continues in the poetic text beyond that point
(“from sorrow I aged too soon’), it is fitting that the chromatic Ds persists
during the embellishment of the dominant in the following measures. Only
in the middle of measure 41, in support of the prayerful “make me eternally
young!” is B Major fully restored.
The degree of assertiveness displayed by the lower-third chords presented
in these examples is minimal. Though each chord is itself embellished briefly
(in the passages corresponding to the open parentheses in the analyses
below the scores) and thus represents something more than a mere 5<6>5
extension of the preceding tonic or dominant chord, they do not take on a
life of their own (as harmonic entities) to lead the progression in a new direc-
tion. Since these 6-phase chords revert to the preceding 5-phase chords,
separate Roman numerals have not been formulated for the analyses.
Upper-third chords likewise may appear in an array of chromatic
60 Harmony in Schubert

Diatonic Chromatic Chromatic Chromatic

Variant#1 Variant #2 Variant #3

C Minor oN g Zt Bats Be ios el

Diatonic Chromatic Chromatic Chromatic
Variant #1 Variant #2 Variant #3

2.20 Diatonic upper-third relationship and three chromatic variants.

variants equivalent to those for lower-third chords. The chart of upper-

third chords for C-E-G and C-E}-G is presented in 2.20.’ (Compare with
2.17.) Chromatic upper-third chords are featured in the Trio from the third
movement of Schubert’s Piano Sonata in D Major, to which we turn our
attention next.

Piano Sonata in D Major (D. 850), movement 3, Trio

The Trio from the third movement of Schubert’s Piano Sonata in D Major
offers a diverse landscape of musical events [2.21]. Whereas the opening
tonic expansion (measures 118 through the downbeat of measure 126)
occurs on a rather flat plateau, the musical traveler soon encounters more
jagged terrain (in coordination with a crescendo that reaches fortissimo),
peaking at measure 151. The drop-off thereafter is sudden and stark. A
smooth melodic surface pervades the following measures; only unisons
and seconds occur in the melody of measures 154 through the downbeat
of measure 165. Though the line eventually soars once again, the second
peak (measure 177) is at a pianissimo dynamic level — more ethereal than
dramatic — leading to a cadence of sweet calmness.
Though Schubert had control over many of the parameters that give
the Trio its special character and impact, its basic shape is generic [2.22].
The melodic path from opening tonic to closing tonic generally will span
a descending interval from tonic’s third or fifth to its root (here By
2 will be supported by V. Between I and V in a major key, one generally
encounters II or IV. A special case of this structure, which Schenker calls
interruption, is displayed in 2.22. Instead of a direct connection between
cand |, Schubert reprises the opening material during which a secondary
descent to | occurs. The two vertical lines after 2 above the staff indicate
 Linear progression 61

{Me ie
|
The,
oF, ye: |

ec |An: ees
[eal
o c= o aE

vest Sere cites

oe
[—o—a@ _— e
wY
P legato <<
Ss
—— YY a8
ete ————~
ES h.
see
Th
="
Pages
333
=e
=eee 3oce
a = ae
JP
3 ¢ B-
ome

| ey vee
Fe
Y
i( *B)|

é if
2 it

tt
Gi &

=
Cait4 TO

yc dt
12 9

Baty
ae ee ——— ea
(4

tT
A

“UL
s(AnD

It
TP

ona
Nl
8

Ja
f
tt th)
tt hy

oe es
rae>

ba 2 be @ @ )be}

ee
ra
z
+: === BE $ 5s_s—s
cTresc.
hod Ht

| emul
a fai nT=i
wlll

hint

Hg
tt.
Lat

In in
ie

ny
LI

Athy

e-
e-
3|
=
r)

ett:
Mr
aie!

eee!

yJe
ll

rey

jANI)
rt

Wi,
=

ths
te:

ine
te-

“.

.
ny

:e 159

e444 Se
Pn
ee oe ere ee 8|| ees

5S!
4

-
tae
ee ete eee FG

Camre (RIA
aeins (as wiltCHar
(

: ae
“.
170 a titahd)
+h

Gt (Sai=
GH
& SK 3

mihih (saa ene Sale (-«

‘els wll? 2 wk
it ry vntor) qe

i)
C
ae
&« @ieo)

N q
= =| < a
Bhdhiu
|in =a)
4 1
SQaa) Ss =Ss
62 Harmony in Schubert

m. 119 153 159 163 179 180

Sa Pale 1G spy a
4
FJ
\
aE ee

SS SS ee
reoH Le —.
es a
G Major: I IV. V I

2.22 Analysis of Piano Sonata in D Major (D. 850), mvmt. 3, Trio.

Model 1 Model2

LON IAS SiS SS ss lisi7. m. 119 149 SiS: 157,

Mee

3 IN 3 IN
2 e
wy AN | a 2 3

=H [De
—oe ;
=f

G Major: I IV GMajor l= SV

2.23 Analysis of Piano Sonata in D Major (D. 850), mvmt. 3, Trio, mm. 119-157.

the interruption. Note that 2 and the delayed 1 are nevertheless connected
by the beam.”° The composition’s surface departs from this foundational
model in various ways, as will be explained below. Indeed, the soprano
of measure 153 is not C, the bass of measure 159 is not D, and the bass of
measure 163 is not G. In addition, registers have been normalized: the high
pitches of the final chords have been placed down an octave.
Model 1 of 2.23 shows the rising and falling melodic lines that arpeggiate
the pitches of I and IV. This unfolding of chords leads to the non-alignment
of IV’s root C and melody pitch C (which appear together in 2.22). In fact,
by the time the melody has reached C, another developmental process is
already under way: the 5-6 expansion of IV. (See 2.23, Model 2.) Thus the
chord displayed for measure 153 in 2.22 never actually occurs as presented.
It would have been fastidious for Schubert to await melody pitch C’s arrival
before allowing inner-voice A to displace G, as indicated parenthetically in
Model 2. And yet one may claim that, at a deep level, the IV chord of 2.22
serves as the foundation for the surface progression.
Model 2 of 2.23 displays another device that readers may well have
 Linear progression 63

Model 1 Model2
OR IS6. 37 AAS TAS A
m. 119 NG YF/ 145 149 ~
4 4 B (aa (fe gS
f g “= (> Z - ig
= o 7 (7% A oe
e ts 4

7)
=) # Z 2 a or a
o- O 2
G Cc FY B G G F¥ B

Model3

m. 119 PATS | Viteye 141 145 148 149 151

G Major: I8—( et 75

2.24 Analysis of Piano Sonata in D Major (D. 850), mvmt. 3, Trio, mm. 119-151.

expected: in a I-IV succession, I often takes on a dominant-emulating

aspect, here achieved via Fs. In this case the G>Ft line is filled in chromati-
cally, as G>F#>Fs. Tonic’s upper-third chord supports F#.?! Note something
rather disconcerting: thirty measures separate the tonic and its upper-third
chord. Though the tonic is prolonged for about a fourth of that time, various
other chords that are not acknowledged in this model constitute most of the
region. Model 1 in 2.24 explores the connection between the G tonic and its
upper-third chord. That connection takes the form of a descending circle
of fifths. Though we christen this pattern as “descending fifths,” in fact such
progressions generally alternate between descending fifths and ascend-
ing fourths so as not to extend beyond a conventional range. In this case
Schubert is vitally interested in extending that range in the upward direction
(as was described above), so he takes the melodic descending fifth E>A#
(measures 137-145) and rotates it into an ascending fourth, as shown in
Model 2. This model also displays how an arpeggiation from chordal third
to fifth extends each of the first three chords. Finally, Model 3 reveals how
Schubert takes advantage of an opportunity for a vivid expansion. Why har-
monize both B and D witha G chord, or both E and G with a C chord? In the
first case, Schubert deploys G’s upper-fifth chord,”* while in the second case
one of C’s chromatic upper-third chords takes a turn. (A complete listing of
C’s consonant upper-third chords appears in 2.20.)
64 Harmony in Schubert

=
4 ye ey 4 ei
# fe —4 i

N N
G Major. G Major: VY—————_

2.25 Embellishment of tonic and dominant chords in G Major.

Only this broad perspective allows us to make sense of Schubert's proce-

dure here. One might be struck by the juxtaposition of three major chords
— C, Eb, and Gb - in measures 137 through 148. Each is presented boidly,
and the span is an antipodal diminished fifth. The discussion above clarifies
that (1) G} is actually F#; (2) the antipodal C-F# here functions in one of its
few normative diatonic contexts, within a circle of fifths; (3) the Eb chord is
not on the same hierarchical level as the C and F# [Gb] chords;”° and (4) the
C and F# [G}] chords both function as interior elements within a broader
circular initiative: G-C-F#-B.
The transition back to the tonic following the background dominant of
measure 159 utilizes a device from earlier in the Trio that has not yet been
mentioned. Schubert engages in some wondrous local embellishments of
individual chords. The first instance occurs in measures 130 through 133,
where dominant D major and its upper-third chord F# major are juxtaposed.
(This relationship corresponds to Chromatic Variant #1 in 2.20.) Later, in
measures 174 through 177, tonic G Major and upper-third chord B major
are similarly juxtaposed.” Their linear derivation is displayed in 2.25: the
chords are built using one common tone, one diatonic half step (a neighbor-
ing note), and one chromatic half step (a wobbly note). Schubert calls upon
such a chord in the delicate transition between the dominant (measures 159
through 162) and the return of the opening melody. Whereas the soprano
melody starting at measure 163 corresponds to that starting at measure 119,
Schubert’s harmonic support only gradually reaches tonic. Dominant D
major (in 5position) and tonic G Major are connected by an intermediary B
major chord, one of tonic’s upper-third chords. (We might say that I-space
begins in measure 163.) A smooth melodic surface was mentioned earlier as
a characteristic of this region. The soprano’s melodic seconds (E>D>C>B>A
in measures 153 through 162) are matched by those in the bass (C>B>A>G
in measures 153 through 165). Though the first inversion of the tonic would
be a more common initiation of I-space (we teach our students that V in .
position resolves to I in §position), here the B major chord not only offers
a novel alternative for the context of a gradual bass descent but also relates
motivically to other third-relations earlier in the Trio.
3 Common prolongations and successions

Tonic prolongation
Tonic’s lower-third and upper-third chords
The second movement from Schubert’s Piano Sonata in A Major opens
with a fifteen-measure evocation of D Major [3.1]. The harmonic content
is exceedingly plain. Tonic’s prolongation in measures 1 through 5 and
8 through 13 (like the dominant’s prolongation in measures 6 and 7,
which will be addressed later in this chapter) engages embellishing
pitches. Though many analysts use Roman numerals - sometimes very
awkwardly - for neighboring motions, that practice will be eschewed
here, unless the neighboring chord asserts itself beyond its embellishing
role. For example, compare measures 1 through 5 of the sonata move-
ment with the first six measures of the String Quintet in C Major [3.2].
Both passages involve a major tonic chord embellished by three neigh-
boring or wobbly notes, with one pitch held as a common tone (root C
in the quintet, fifth A — locally embellished by B - in the sonata). A label
such as 21V” for the quintet’s second chord may describe pitch content,
but it leaves one clueless regarding function, and in fact would make
Schubert's resolution to I seem eccentric (antipodal Ff-C). Though the
label V’ for the chord of the sonata’s third and fourth measures is less
objectionable, the same procedure ~ that of multiple neighbors — is at
work. So, just as the sonata movements first chord is not VI but instead
an embellished I,' the chord of its third and fourth measures might be
interpreted as a multiply embellished I.
Whereas tonic’s root and third are embellished in the Andante’s first
phrase, the second phrase focuses on its fifth. B served as a local embel-
lishment of A during the first phrase. That event is promoted from the sub-
chordal to the chordal level in measures 9 through 12, where B joins tonic’s
root D and third F%. Even here, B is destined to return to A. B-D-F# is
tonic’s diatonic lower-third chord, the internal component of a P~°° expan-
sion. (See 2.17.) The 6 phase is unfurled into 3position, and tonic’s return is
presented in §rather than ?position (measure 13).
Tonic’s upper-third chord likewise may assist during a prolongation.
66 Harmony in Schubert

ae
Andante

or [3 4 3
1
1 72
—_
ON, Say hil ae:
8
4 =< arin Sere. =: ; $3 a = Lae:
el
be hdd : cio
pee eae p
aug gs Lz x: «
PE Poe
FS

pares iitie ti =o
ee See a ESETSF at HeesSS 4 Ey) See,

Dp 6 5 Vo

3.1 Piano Sonata in A Major (D. 664), mvmt. 2, mm. 1-15.

Allegro ma non troppo

To: oes ay ' = s

3 g
ety
Violino I Ge cer > =
Pp f
e
2
= =
cs
Violino II
6 Yu; +} 8——— 8
ifs
P

ai =
S bS E= 5
Viola Hee
ip at ip
Pp +
eS ena
oO aes ioe ?
Violoncello I ee
P
12 if pest e
5 = =
Violoncello II Bae =

C Major: I
5
Pa

I >4 4
6
4¢
or

3.2 String Quintet in C Major (D. 956), mvmt. 1, mm. 1-6.

In “An Emma’ a major tonic chord’s succession to IV is propelled by the

addition of a minor seventh (Eb) [3.3, Model 1]. Model 2, derived from
Schubert's score, fills in the melodic F>Eb major second as F>E>Eb. E is
supported by tonic’s third (A) and wobbly fifth (C#); root F is suppressed.?
 Common prolongations and successions

Model 1 Model2

ig 3) B2 33 34

o- a5 a |

Saad ae 6 —
a Q g
a= = 3 a =
= a= a
F Major: I Ls Iehi, F Major: I >i IV
e

3.3. Analysis of “An Emma’ (third version, D. 113), mm. 31-34.

Trio 78 Wee

rb ] —— J i: = = = T

>) 2 ae —— = o = ed

a : : ue t 4 =i r=

; : SS es,
Dy 5 A +

6
D> Major: I [von ey

(= II!)
3.4 Moment Musical in Ab Major (D. 780/6), mm. 77/78-81.

In this case the chromaticized chordal fifth appears in the bass. Other
instances of inversion will be addressed below.

Inversional diversity

When a bass line ascends from tonic’s root to its third, inversional diversity
within the tonic prolongation is not the only consequence. In addition, an
upward trajectory is established, one that often pushes onwards to 4 and
then 5. Sometimes the 1<3 connection will be complemented by 3>1 in
another voice, forming a voice exchange, as in the Trio of Schubert's Moment
Musical in Ab Major [3.4]. The pitches in measure 79 correspond to those
in measures 3 and 4 of 3.1. In both works 4 and 7 function as neighboring
notes, leading back to tonic’s third and root, respectively. Yet whereas 3 also
functions as a neighbor in the sonata, in the Moment Musical it functions
as a passing note in both outer voices. Neither 3.1 nor 3.4 displays a har-
monic label for the chord that comes between the two tonic chords, since
it does not assert itself beyond its connective role. I-space is extended until
the arrival of IV.
68 Harmony in Schubert

a aa
ye ee res wearers | ; E 2% #
2 st
@twegi@o |). 2 = = = - ie Se

he ne eas | |
gies: ee ees s e =

3.5 Menuet in C Major (D. 995/1), mm. 0/1-8.

Schubert's Menuet in C Major merges notions of inversion, voice exchange,

and transfer to a higher register [3.5]. The basic gist of the tonic prolonga-
tion is that two successive applications of voice exchange will leave one
exactly where one started [3.6, Model 1]. One may even expand bass C<E>C
into a full-octave arpeggiation (C>G>E>C), as shown in Model 2.* This idea
is merged with yet another agenda: that of moving gradually higher. That
initiative begins with soprano G on beat 2 of measure 3. The unfolding of the
tonic’s three pitches in the melody of measures 0 through 2 ascends: G<C<E.
A conventional response would be to descend from measure 3’s D: D>B>G.
Schubert instead proceeds upwards: D<G<B. Consequently inner-voice B
is hoisted up an octave, as shown in Model 3 (measure 6). The remainder of
the upper structure transpires in that register as well. The bass responds by
transforming the E-C link of measure 7, represented by a descending third
in Model 2, into an ascending thirteenth (Model 3).
A more extended voice-leading connection unfolds in the opening
measures of “Todtengraber-Weise” [3.7]. The structure begins as a
straightforward ascending 5-6 sequence, with the common chromatic
adjustment G4 applied in measures 3 and 4 to avoid the diminished
G#-B-D chord that otherwise would result.> Left unattended, that progres-
sion would proceed to IV or beyond [3.8]. Schubert makes the slightest
of adjustments, yet one that has a major impact upon the course of the
ascent. After G#-B-E, already a 6 phase, he raises the sixth E another half
step to E# [3.9]. Whereas G#-B-E (to which Schubert adds D) emulates the
dominant of A, G-B-E# (to which Schubert adds C#) emulates the domi-
nant of F#, to which it resolves. Three lines moving in tandem from one
 Common prolongations and successions 69

Model 1 Model 2

W> N>

ee )
pee XX = = Z
ia]

|
6 V
I ——————-
C Major:

Model 3
=? pels
ae
A
2 oe ees
&
Zz
2
A
poe

=a
ry

== a

“
I ini 6
C Major:

3.6 Analysis of Menuet in C Major (D. 995/1), mm. 0/1-8.

Ziemlich langsam

seas
haig CLESC. |) —<——

e¢ lecets

F# Minor: I

3.7 “Todtengraber-Weise” (D. 869), mm. 0/1-6.

3.8 Model for an ascending 5-6 sequence in F# Minor.

70 Harmony in Schubert

oL iRae is
eo! Teh UY
ke

F¢ Minor: I

3.9 Analysis of “Todtengraber-Weise” (D. 869), mm. 1-6.

to the next nodal point of the tonic chord are shown in 3.9: FR<A, A<C#,
and C#<F¥. Since the upper line must cover more ground than the lower
ones, it becomes aggressive in measure 5, so that all three lines arrive at
their destinations concurrently.

The circle of thirds

Two topics of chapter 2 - circular progressions and chromatic lower-third

and upper-third chords — merge in an innovative tonic prolongation in the
Sanctus from Schubert’s Mass No. 6 in Eb Major. Consider again the third
chromatic variant of major °° from 2.17: one half-step neighbor and two
wobbly notes induce a shift from a C major chord to an Ab minor chord.
That succession is transposed into E} Major in 3.10, Model 1.° (A resolu-
tion chord is supplied as well.) In Model 2 the resolution is divided into two
phases: first G and B} return, with retained Ebb; then Eb returns. Hearing
this example, one can sense how the prolonged Ebb behaves as a suspen-
sion, with upward resolution to E>. This progression retains the essence of
the tripartite °° neighbor embellishment. Yet the notes within brackets
reveal that the moment of partial resolution between tonics chromatic
lower-third chord and tonic is itself the enharmonic equivalent of tonic’s
diatonic upper-third chord. In this way a pathway is forged between two
distinct tonal initiatives. The separate prolongations

Tonic —- Lower Third- Tonic and Tonic - Upper Third - Tonic

are joined in the progression

Tonic — Lower Third - Upper Third - Tonic.

This juxtaposition may offer some insight into how a composer such as
Schubert made the transition into the realm of modulo 12 composition.
What began conceptually as a chordal embellishment involving a neighbor
and two wobbly notes (Model 1), with no enharmonic respelling required,
at first breaks apart into a two-part resolution (Model 2), in which the
 Common prolongations and successions Bil

Model 1 Model 2

by = a .2
ae ee (ae Bae Pere
; See ieee iS) 5 b
E> Major: 1 aa Eb Major: I. : ; ;

Model3
m. 1 3 5 7

= Bo g
3—
22s 2)

6b 5)
Eb Major: 1}, >
bb
5
b

3.10 Analysis of Mass No. 6 in Eb Major (D. 950), Sanctus, mm. 1-7.

coincidental identity in sound between a partially resolved lower-third

chord and the upper-third chord ultimately leads to an equal subdivision
of the octave (Model 3, in the contour of the Sanctus), made emphatic
through the insertion of passing notes that fill in a whole-tone scale.’ The
inelegance of Schubert’s Ch-to-By enharmonic conversion would not occur
in numerical modulo 12 notation: 3>1>11>9>7>5>3. Because the internal
chords do not assert themselves harmonically, Roman-numeral labels
would be pointless and misleading. The entire progression projects a single
harmony — namely, tonic. At the conclusion of the prolongation, tonic’s
third is lowered in anticipation of the Cb major chord that will follow.

Dominant prolongation
Local passing and neighboring motions

The tonic chord generally is built using three pitch classes. The domi-
nant often expands to four: root, third, fifth, and seventh. (In the context
of dominant emulation, a tonic chord often takes on a minor seventh as
well.) With such abundant pathways for melodic connection, the dominant
chord offers a wide range of mix-and-match options. For example, a domi-
nant prolongation in “Das Heimweh” [2.16] includes a B<D connection
between fifth and seventh, which coordinates with the G#<B (third<fifth)
ascent in measures 185 through 187. In this case the upward momentum
72 Harmony in Schubert

6 SS
Dein

(eae
ae aa=
Me ai el es Reel

(eo
=f — ee
meer 2
G Major: V :

3.11 “Liebesrauch” (D. 179), m. 10.

pushes even beyond these nodal points, extending to tonic’s fifth and
third. In “Liebesrauch” [3.11] three concurrent motions occur: ascending
root<third (D<F#) and third<fifth (F#<A) against the chromatic descend-
ing octave>seventh (D>C). The concurrent stepwise ascent of three thirds
(root<third, third<fifth, and fifth<seventh) is more difficult to achieve due
to the threat of parallel fifths. Yet Schubert manages this in his Piano Sonata
in A Major [3.1, measures 6-7] by separating the E and F# of the E<F#<G
soprano line. F# arrives only at the end of measure 7, beat 1, as a thirty-
second note. These measures offer a vivid example of what Schenker calls
Ubergreifen (reaching-over). The basic idea is that E descends to D (across
the measures 6-7 bar line), then F# “reaches over” it. This procedure is then
repeated a step higher: F# descends to E, and G “reaches over.” In this latter
instance the descending second and the emergence of the upper pitch occur
concurrently.®
In one distinctive instance of dominant prolongation, called an omnibus
for obscure reasons,’ the spans of root<third and third>root are concur-
rently and chromatically traversed, against the fifth and seventh held in
other voices. In the first movement of his Piano Sonata in A Minor (D.
845), Schubert connects dominant G’s root and third in this manner
during the build-up for the arrival of the tonicized mediant key, C Major:
G<G#<A<Bb<B in the soprano, B>Bb>A>Ab}>G in the bass, with prolonged
D and F in between [3.12, measures 36 and 37]. This excerpt also contains
neighboring motions embellishing individual chord members. Neighbors
A} and A¢ [Bb] are emphasized in measures 34 through 36.!° These notes
are of purely melodic import, thrice embellishing dominant root G and
third B before taking on duties within the omnibus (where Ab, now ascend-
ing, is respelled as G#, and where Bb indeed functions as a B} — not as
an A#). The forward thrust developed by the omnibus leads beyond the
boundaries of the G dominant chord, extending to ry Whereas in the earlier
 Common prolongations and successions 73

| —

sé
tedee ote
zs
C Major:V’
l

omnibus

3.12 Piano Sonata in A Minor (D. 845), mvmt. 1, mm. 34-39.

measures the neighboring motions descended from the dominant’s third

and ascended from its root, now the opposite relationships prevail, and the
interior pitches are affected as well.

V—X—V

A chord that comes between two dominant chords may assert itself force-
fully enough to warrant harmonic interpretation. Given its context, such a
chord generally will be one that leads smoothly back to V. In “Das war ich”
[1.27] the most basic event of the dominant prolongation in measures 8
through 12 is the emergence of a seventh: V*’” [3.13, Model 1]. C, which is
dissonant against root D, is not dissonant in the context of the dominant’s
upper-fifth chord, A-C-E."’ A chordal seventh (such as the C above D here)
may be introduced in a consonant context during such an upper-fifth chord
(thus ©), followed by the reemergence of the dominant root and concurrent
activation of the dissonance (Model 2). Whether A-C-E in Model 2 is better
interpreted as a II harmony or instead as the dominant’ fifth and seventh
sounding with neighboring embellishment of its root and third depends
upon how a composer projects the event. In “Das war ich” the supertonic,
heralded by a dominant-emulating VIj, clearly asserts itself [3.13], and so
II is appropriate as an analytical symbol.
In the Piano Sonata in A Major (D. 664), an E major dominant chord
concludes the first movement’s exposition. This E chord extends through
the development as well [3.14]. Schubert utilizes a segment of the descend-
ing circle of fifths to move from dominant E to subdominant D (measure
61). 3.15 displays how the first descending fifth is traversed via two thirds:
E>C#>A. The second fifth occurs in direct motion, as A’—D. In an expan-
sion of IV>°, retained fifth A (displayed as an open notehead) eventually
joins company with D#, Fs, and C4.'* Schubert positions these notes in a
74 Harmony in Schubert

Model 1 Model 2

8 11 13 m. 8 ite 13
m.

a od <#
i)

G Major: Wear I G Major: Vet eee iT

Ev mv)
3.13 Analysis of “Das war ich” (D. 174), mm. 8-13.

m. 50 51 56 57 59 61 63 64 65
Qa Te art age aoe ean a |

oe
S
ie = 2 g
———— =

my aS ta
2
oD
a abo ae he
ee
os
bes

3.14 Synopsis of Piano Sonata in A Major (D. 664), mvmt. 1, mm. 50-65.

m. 50 57 59 61 64 65

2 eS Se
ae =

CP

A Major: V_ ( hy ee
94
(114, )
:
3.15 Analysis of Piano Sonata in A Major (D. 664), mvmt. 1, mm. 50-65.

conventional augmented sixth configuration, with Fy in the bass [3.14,

measure 64}.
Because V-I is the most basic of all harmonic progressions, any suc-
cessor of V that does not evoke tonic!? represents a backwards step,
generally requiring that the dominant be restored. This backwards-and-
forwards motion can be the defining initiative for an extended passage of a
composition, particularly within a development section.
 Common prolongations and successions VE

I-V

I<V via the ascending 5-6 sequence

The tonic and dominant roots, when presented as an ascending perfect

fifth, often are connected via an intermediary harmonic scale degree that
segments that tonal space into two distinct phases (for example, I<II<V or
I<IV<V). During a sequential connection, in contrast, the path is smooth.
The young Schubert allows an ascending 5-6 sequence free reign in the first
movement of his Symphony No. 1 in D Major (D. 82), connecting I and V
in the tonicized dominant key, A Major:
m. 97 101 105 109 113

A>-6 B-$ Cr D>-6 E>

A Major: I( ee
Such cautious, essentially diatonic writing would be atypical of Schubert in
later works. His creative impulse resisted standing to the side as a passage of
his composition emerged via sequential autopilot. He likes to steer the ship
himself, venturing into novel tonal terrains.
“Suleika I” exemplifies such a creative venture [3.16]. The sequential
passage transpires in the tonicized key of F# Major. A model for the dia-
tonic ascending 5-6 sequence spanning I<V appears in open noteheads in
3.17 (upper staff). Even when the 6 phase of each sequential cycle is chro-
maticized in accord with dominant emulation (indicated by the model's
filled-in noteheads), the progression is neither harmonic within F# Major
(I-II-III-IV-V) nor a series of modulations (F# Major: I; G? Minor: Vx-I;
A# Minor: Vx-I; etc.), but instead a linear connection between the I and
V harmonies.
Anticipating Schubert's penchant for modal shifts, consider next a
model for the same sequence in the context of F# Minor in 3.17 (lower
staff). Because the diatonic chord on G# is of diminished quality, a
rerouting via G4 is proposed. The example shows the ascent both without
and with dominant-emulating inflections (the latter indicated by the
model’s filled-in noteheads). Comparing these normative, “autopilot” pro-
gressions with the extraordinary sequential progression of “Suleika I” in
3.17 (middle staff), we discover that Schubert freely utilizes segments of
both the major- and minor-mode models, while in the middle region he
defies them both: whereas F# Minor offers an A major triad and F# Major
offers an AZ minor triad, Schubert wondrously traverses both A minor and
A major! (To do so in a sequential context requires some enharmonic
reinterpretation in the seven-note notation system: Bb = Af, etc.) The
76 Harmony in Schubert

84 x
a re = \— 2 — z H z

Bes Sp ——— = =
wei-ter zie - hen! Die - ne Freund - den und Be - trib - ten.
‘ Und sokannstdu
——— == ———— 3 7 See
Ge o oe aaa? eo oe ¢ ¢ =e: eo e e 4|
< SS Een SS oe SS . —— eel

yt re ts oe" e« |
=e

38 :

pete
4 ©I to oo A —
=, aa
f z cs
ee — Seat e us = ey =
; Und sokannstdu wei-ter zie - hen! Die - ne Freun - denund Be - trib - ten.

——— o_o _ SS SS SS SS
ches- ws ees faite ete eee oraeees : sEEzeroe!
oe e
Q ———__ Cres; ——— ———— —___

+ a aS PS |
Bae o — e ——_ o — = a ——

92 -

ar =F — Pee.
I$ int He
———
Dort, dort, wo ho he Mauern glii hen

Ze eiesese: ee ee ee ee
i © 0° -@ © —¢ = oe 6 oe aa=s # E
© >=——.__crese es
CPS i —— ee. SSS Pa _|
—=
a ial T
Be te a: | —— be

96

4 $e
ua 2 2 2 te ro te ce z rs

e omen ait
dort, find’ ich bald den Viel-ge - lieb - - = ten,
4 4 eo a ef e ° e °

6 Pefter et 2S SS fas stain,’3!

5 ————— = == = — =
cre - scen - do - Wa ER Jp
LL @ te te 2 a te oe
pete P= AE —o. 9 —__o—e

a 3 == = ———-U —

3.16 “Suleika I” (D. 720), mm. 84-99.

“Vielgeliebten” (much-beloved one) of the poem is found not by travers-

ing life's conventional thoroughfares, but instead through a singular and
breathtaking path.

I>V via the descending circle of fifths

The alternative sixth and seventh scale degrees in minor keys are enshrined
in a distinctive version of the minor scale. Raised 6 and raised 7 come
between 5 and 8 in the “ascending melodic minor” scale, giving these mod-
ified pitches an aura of legitimacy. The reverse may occur in major keys.
Because the leading tone points upwards, a viable means of connecting
8>5
 Common prolongations and successions 77

Progression in F# Major (open noteheads = diatonic)

Se
Fe Major: Fs ot a VY ow Q AN

Progression corresponding to "Suleika I," mm. 84-98 (chordal sevenths omitted)

m. 84 85 88 92 93 94 9 ‘i 97-98
: | 2: be 2 g

Gi
e
= z oe eat fe He i
# G a Bb B Ce
; (A?)
Fz Major: I ( yeas

pe
Progression in Fz Minor (open noteheads = diatonic; Gs substitution for G2)

Z — ;
NSS SS oe Od
a
{tess

Fz Minor: f2 > b ce

3.17 Sequential progressions traversing I<V.

in major is via lowered 7 and lowered 6 - components of what we might

christen as the “descending melodic major” scale.!°
In its diatonic state the descending circle of fifths is problematic in a
major key. In C Major, C>F is followed by the B diminished triad. This defi-
cient element inconveniently takes its turn before the sequence’s forward
momentum has fully developed. Freed of diatonic constraints, F, a major
triad, would press onwards to Bs), not to B. So why not do so, postponing the
necessary corrective melodic diminished fifth until later? Such thinking may
have influenced the young Schubert when he composed “Erinnerungen”
[3.18]. From the tonic B) major chord, whose dominant-emulating capac-
ity is activated through the addition of Ab, the progression leads first to an
E> major chord. From E> Schubert leads not to A diminished, but instead
to Ab major. On the way to Ab he transforms E>-G-B> into G-B)-D)-F)
(its F} spelled as Ey because it proceeds not directly to goal pitch E> but to
E}’s upper neighbor F at the downbeat of measure 25).'° Once Ab major is
secured (with G> added to enhance dominant emulation), the progression
continues onwards to D} and then G}. Thus, though the key is Bb Major, the
tonal trajectory of Bb “Descending Melodic Major” is traversed, approach-
ing dominant root F not via A and G, but instead via Ab and Gp:

Bh Eb Ab Db Gb...
Harmony in Schubert

Sa b> Ya i2 way peo—t 5

ae a = = =
== =:

zii-cken: Wie schén,wie schén,wie —sch6n! =

ys riefstdu voll_Ent

: ive e mae: eo é i: 2
<a

&e —
aS ane y; eal
SS eee

ot :
SATE og Pet
= : _——> be:

Bb Eb Ab Db
Bb Major:I (

3.18 “Erinnerungen” (D. 98), mm. 23-29.

Even in this trajectory one diminished fifth is required to prevent an

extended circle (. ..-Gb-—C)b—Fb-Bbb-. . .) and to secure the goal dominant F.
Schubert makes this adjustment at }6-2 (Gb-C). The C chord is here repre-
sented by Ek-Gb-Bb-D}). (Compare with 1.2, Chord 10.)
The circle in “Erinnerungen” is propelled by dominant emulation. In
each of its first four links (B},—Eb—Ab—D}—G}), a diminished fifth or
augmented fourth pushes towards resolution in the circle’s next chord.
(These resolutions are indicated by arrows in 3.18.) Gb and C form the
weak link, not suitable for dominant emulation. The concluding CF link
is distinctive, offering an augmented sixth playing its conventional role as
herald of the dominant. The concluding chords of a circular progression
of fifths often begin to take on a harmonic character. Despite the persist-
ent dominant emulation in the early and middle regions of the sequence,
none of those chords actually is a dominant, because none of the chords to
which they resolve substantiates itself as a tonic. In contrast the C chord of
measure 28, though still internal to the circular progression, does behave in
accordance with a normative harmonic function, which may be indicated
within parentheses below the principal analysis.
In the first movement of the String Quintet in C Major [3.19], from the
opposite end of his composing career, Schubert deploys the propulsive force
of the augmented sixth repeatedly in a circular progression much like that
in “Erinnerungen.’ Often at the outset of a circle of fifths the tonic will be
prolonged via a local progression to its upper-fifth chord and back: that is,
instead of beginning abruptly, as C+F—Bb—. . ., the progression warms up
with a tonic expansion, as C-G->C-5F-4Bb—. . .!” In the quintet Schubert
takes that strategy one step further, developing that notion into a tonic-
prolonging harmonic progression I-II#-V-I at the outset.!8 A concurrent
 Common prolongations and successions 79

PESTS 1G25 Sf be 2g
oe cae ra - =
Gh e ‘ ae =

ie |= oll tae — P i ti |
g 3 = i be. a
Gt S22 o- #8=

Pah eee gy = ia Jaeee

154 f t 2 = z = = z3
Pp = ul Se Pp Kt

ee 4 z t= = i = = = = es

oe = =
flalee a.doe ee = =
elo ay aoe aa obo
i Sf pres P cres¢c-—= fo

C Major: I
Ob 8

ad i V3
421
Ag . be
6 : : r . = 2
be Sz Wi ke fi Fe é

et as a
=. = =e
45 = be = a 2 z cS ? zs oe = a —

=
=
2 S a A.=
SaaS crese

PS ae ne = Nps = A
———————— crese.

F Bb
(
I)

ie be ry de be be aa ee Siecl yo
=e ee = sce v
A cresc. It a
fi Fi
C2 a Pa SS r=
S a e Re =e ? oo oie =
Laecresc. Sard Hi a es ie
2 | arty, 7 il Sz ‘e

{oer err ceerae uae ae aS ft

a
Kc
Tip Bie
cresc

Bat ei = = ——— ae
are ROP ie = i ee ‘fz

a = ie pee a=
Eb Ab D G at:
tres
Ob

- I )

3.19 String Quintet in C Major (D. 956), mvmt. 1, mm. 414-428.

80 Harmony in Schubert

initiative has a pronounced impact, however. V-space is expanded via

one of the dominant’s chromatic upper-third chords, Bb-D)b-F (measures
420-421), and when G is restored (measure 422) Schubert does not quite
shake off the chromaticism: Bb (a wobbly note) yields to By and F serves as
the chordal seventh, but D} remains, not yielding to the dominants diatonic
Ds. That deformity is replicated in every second chord of the quintet’s circle
so that, in contrast to the “Erinnerungen” model’s emulation of dominant
sevenths, the augmented sixth resolution is emulated: C(-D—~G™)C-
Fe>Bb-Eb>Ab-D&G. The additional chromaticism imbues every other
chord with a sense of arrival on a dominant. (Even the C chord of measure
422 may seem like a dominant.) But neither C, nor Bb, nor A} substantiates
itself as a dominant. Only the concluding G chord does, and thus only the D
chord of measures 425 and 426 (emphasized both durationally and dynami-
cally) may justifiably be christened as II, as displayed within parentheses in
319:

I-IV

I>IV

A major tonic chord balances two contrasting potentialities: as goal of

various harmonic motions (particularly as resolution of chords in which
the key’s leading tone is operative) it embodies stability, the cessation of
motion. Yet if the composition is to proceed forward in a new tonal trajec-
tory, then tonic is also the foundation from which the initiating sally will be
launched. In this latter capacity it is advantageous that I is built in a way that
emulates a dominant. (In a minor key, tonic’s third may be raised to attain
that resemblance.) That makes IV a straightforward target. Dominant emu-
lation is intensified if a minor seventh joins the ranks. Whereas C-E-G in
C Major may serve as a cadential resting point, C-E-G-B} forcefully surges
towards IV.'?
A gentle and moving phrase in A} Major from the second movement of
Schubert's Piano Sonata in C Minor [3.20] pursues a ce trajectory [3.21,
Model 1]. Upper neighbor D}, a detour in the melodic descent from 3 (C),
is supported by IV (Model 2).’° Three of the tonic chord’s intervals are
traversed during the I*” prolongation in measure 11: C>Ab, Eb<Gb, and
Ab>E> (Model 3), leading inexorably to IV. Schubert's crescendo marking
coincides with these traversals and supports my interpretation of the
F-Ab-C chord, which in measure 3 served as tonic’s 6 phase, as a passing
chord between the foundational 8 and evolved 7} states of the Ab tonic.2!
 Common prolongations and successions 81

9 aes

eesagenia agp peo

f AF

i Se ee
pe
e
ts
Pt
C

3.20 Piano Sonata in C Minor (D. 958), mvmt. 2, mm. 9-14.

Model 1 Model 2
m. 9-11 13 14 m. 9-11 12 13 14
nw A Nn A A N
o 2 1 3 IN 2 1

Oa N Bot ohh OQ)

AN
|
Ow N
~

> er &

Ab Major: 1°‘ DES

3.21 Analysis of Piano Sonata in C Minor (D. 958), mvmt. 2, mm. 9-14.

Alternatively the descending perfect fifth of I>IV may be divided into two
thirds. As in any circle of thirds (compare with 2.14), the chord on 6 may
contain a wobbly note. In “Die Liebe hat gelogen” [3.22] an A major chord
occurs directly after the C tonic chord. Whereas such a succession would
often lead onward to II (as a chromaticized I°-°), here the chromaticism
seems coloristic and idiosyncratic, since Schubert immediately restores C4
in the IV chord. There is a temporary breakdown in the cooperative spirit
among chords: normally during a C>A>F descent the pitch A would forsake
its own partials (A A E A Cf...) and conform to C Major's diatonic pitch
collection. Yet here A uncompromisingly retains C#. Thus major chords on
C, A, and F are juxtaposed. The latter two are emphasized dynamically.
82 Harmony in Schubert

ee Seer ——
be - tro-gen,ach, be-tro-gen hat al - les
S
mich
ic um- her! 1

os ffxJSP
Eger i =

C Major: I ( )

3.22 “Die Liebe hat gelogen” (D. 751), mm. 4/5-7.

Moderato

rs ro rt
v SSs8

F Major: Le

3.23 “Schlaflied” (D. 527), mm. 1-4.

I<IV

The first two measures of “Schlaflied” establish F Major as the tonic, with
the filling-in of an * third in the melody [3.23]. In the following measure E}
ignites the surge towards IV. Once IV is achieved, the phrase continues to V
and then I in a conventional manner.”
The leading tone sometimes descends directly to tonic’ minor seventh
(8-7-}7) instead of returning to the tonic pitch (as in 8-7-8).25 Schubert’s
resolution of leading tone E to the tonic pitch F before the addition of E>
in “Schlaflied” [3.23, measures 2 and 3] is downright fastidious compared
with the convenient shortcut A>G#>G_ that he pursues in measures 1
through 5 of “Ungeduld” from Die schéne Miillerin [1.36], in A Major. The
leading tone’s conventional resolution is elided, and the thrust towards I'V
is enhanced.
Schubert occasionally harmonizes the 8>7>)7 motion in a more novel
fashion, employing tonic’s upper-third chord as support for 7. Ina passage
 Common prolongations and successions 83

a =
oe
ah FD Cem eS
wy, Fe
;

———— = iH ==: +
5
ce
a
ee oe +s 4e .
dis
SS = —"
——- s = = 1 a = SS
z ra + T Tt Zz ZF 2, ER oe

1= ae

J. ——— So SS =
SCsae ee ae|a peace eae deel ¢

eS
a
SSrs ee ee =
Phi ee

4 7 6—5
Z 1V¥—— Vi4e3" ft

3.24 String Quintet in C Major (D. 956), mvmt. 1, mm. 137-142.

in G Major from the first movement of the String Quintet in C Major

[3.24], the descending melody G>F#>Fs is so harmonized. Inner-voice D
wobbles to Dé during the upper-third chord, but diatonic Dz obediently
reappears before the succession to IV. This upper-third chord may instead
proceed directly to IV, as we saw in “Hagars Klage” [1.5].
One time-honored means of traversing I<IV is via an ascending 5-6
sequence [3.25], which often is modified via dominant emulation so that
each 6-phase chord resolves dynamically into the next 5-phase chord
[3.26]. On one hand, each diatonic triad on the path between I and IV
is traversed (B><C<D<E>); on the other hand, a special characteristic of
this sequence type (as is displayed in 2.7, Model 4) is that its sixth chord
restores the pitch content of its first chord. With the addition of A», this
span projects the harmonic intent I* ”, making the arrival of IV inevitable.
In “Der Sanger, Schubert subtly adjusts this pattern so that A>, tonic By's
minor seventh, arrives at the earliest possible moment [3.27], rather than
downward from As (as in 3.26). Even more striking is Schubert's strategy in
“Aus Heliopolis II” [3.28], where the sequence is calibrated to ascend obsti-
nately, one half step per cycle (C**® D5”* D**.. .). Proceeding in this way
for the full span of I<IV could cause tedium, and so once some momentum
has been developed and the gravitational field of tonic escaped, the 5 phases
are elided and the sequence morphs into a glide: ® chords on D, Dé, and E
are juxtaposed, leading dynamically to IV.
In an excerpt from Schubert's String Quartet No. 14 in D Minor [3.29],
tonic D Minor is prolonged initially via a clash of four neighboring notes
forming a diminished seventh chord (C#-E-G-B>) against a tonic pedal.
84 Harmony in Schubert

=
Bb Major: I ( SN

3.25 Ascending 5-6 sequence.

= eee
D Eb
Bb C
Bb Major: 1° ( DP aly:
Vs.
Bb Major: 1° Pine BY

3.26 Chromaticized ascending 5-6 sequence.

ni, (30 SMisT! ) 1624 Misamis “ss * FFI6

=‘ a ee ee$9
pee

Bb Major: ls iy el,

3.27 Analysis of “Der Sanger” (D. 149), mm. 130-136.

mh DM MO Bl BS Bie 28 41

a ‘
rh —gael
| b
Peet —
C Minor: i° a IV

3.28 Analysis of “Aus Heliopolis IP (D. 754), mm. 26-41.

The second time this neighboring chord resolves, the tonic takes on a
surprising new aspect: instead of staid D-F-A, as before, a fortissimo D-F#-
A-C-E} variant sounds (measure 176). Though one might expect that this
intensely dominant-emulating chord will lead directly to G, an inverted
E} chord sounds instead in measure 178. The upward path continues with
such apparent oddities until reaching the major subdominant in measure
181. Perhaps the most useful conceptual approach to the passage is that
shown in 3.30, Model 1: a straightforward § glide in chromatic ascent.
Some of the chords in fact appear in § position in the score. Others are
unfurled, with added dissonance (Model 2). Their ninths may be under-
stood as anticipations of the § chords that follow, while their sevenths are
 Common prolongations and successions

pe
BRP
ta
ole= te
Swe
EEZESTdhe
| ;
:
oS —S= = = ee at p af :
: Rec |
i a = mo

Pp St ‘

a ae Ss
= Z _ :

If
oS f — T Se —s —— — sf te |fe ie

pp® weve see ee v a Vues sees : : oo A

D Sf
D Minor: I (
178 = = - AGS, Ss
rae é So xa =e eae le AN ET ete- ott =
s > } = ae = T St ui vm 7 =

ee
see = Eee o | ots, 2
eve eee
A -~ he Se ie he he: ie: ie -
B t beet rs ete 2 i ize et ad S #
— amd

= (be
ae be van
f A re iG(ie fhe
aoe elhe OG
& ia fits ==
eS i — oH == Z : E * pre

Eb E EF Fe
Vier Vi
9
© 1!)

3.29 String Quartet No. 14 in D Minor (D. 810), mvymt. 1, mm. 173-182.

passing notes (Model 3). These highly charged chords sound in place of,
rather than after, the $ chords that they represent. The ascending trajec-
tory concludes with the major subdominant chord, at which point a new
initiative — a variant of the 5-6 shift (as shown in 3.29) — serves to connect
IV_ and V#.

II-V

When the roots of II and V are arranged as an ascending fourth, the inter-
vening fourth scale degree, the II chord’s third, often plays a prominent
role in the bass. II may be presented in first inversion or represented by its
upper-third chord, and the fourth scale degree may be raised chromati-
cally so that II emulates a dominant. Both 4 and #4 in the tonicized key of
D Major serve as bass in the third movement of Schubert’s early Piano
Sonata in A Minor [1.1]. The third and fifth of II engage in a voice exchange
86 Harmony in Schubert

Model 1

me 173 176 178 180 181

Ge
SI —= 3 ——— a= ——|
Model2
a a
ay : Spe “ = fy z— |
$
D Minor: I ( ) Iv4

Model3

elided chord anticipation of next 6

eh ce

hos ae
e
passing note
unfurling of :

3.30 Analysis of String Quartet No. 14 in D Minor (D. 810) , mvmt. 1, mm. 173-181.

ee
See
3.31 Piano Sonata in A Minor (D. 845), mvmt. 1, mm. 40-50.

(between bass and upper inner voice) in measure 71 of “Am Feierabend”

from Die schéne Miillerin [1.34].
When the roots of II and V are arranged as a descending fifth, a new pos-
sibility for expansion emerges: the fifth may be traversed using a segment of
the descending circle of thirds progression. Because the diatonic chord on
 Common prolongations and successions

Model 1 Model 2
m 0 48 49 50 m 0 48 49
=
——— =
es s!
Vou boa

ho — ——

ce
|
|
NI |
% | |
4
itt
tt}——
t a

CMajo:I Vv’ VI 2 vy C Major:I1 WV’ VI

Model 3
m 40 43 44 47 48
2 Ff DE ca 7
b = = ———— =

Du, a =—-— j
= Pac oS >

C Major: °° Ot a) Vv’
( VIZ)
3.32 Analysis of Piano Sonata in A Minor (D. 845), mvmt. 1, mm. 40-50.

Tina major key is of diminished quality (for example, supertonic D minor

through B diminished to dominant G major in the context of C Major), the
circle of thirds often traverses an alternative route (e.g., By-D-F in place of
B-D-F). Bo gives this chord a distinctive character within C Major without
asserting a harmonic implication beyond the scope of the prevailing
circular connection.
In Schubert's Piano Sonata in A Minor (D. 845), the opening movement's
exposition contains a tonicization of the mediant key, C Major, beginning
at measure 40 [3.31]. The basic plan for the phrase of measures 40 through
50 is displayed in 3.32, Model 1, wherein the melodic span from 5 to 2 in
C Major is traversed, supported by a harmonic progression that extends
from I to V.% In Model 2, the initial motion from ?to 4: is realized with
inner-voice B placed above outer-voice F, a topsy-turvy situation that rights
itself again in measure 49. Model 3, which shows in detail how Schubert
traverses the span from I to V’, incorporates not only passing note A filling
in the G<B third in the melody, but also the harmonic support of II. (A
5-6 expansion of I facilitates the arrival of I.) The melodic whole step A<B
(supported by II>V) is chromatically filled in as A<B><B, in coordination
with descending thirds in the bass.
In the succession from II to V, II's third, fifth, and seventh occasionally
sound without its root. In the second movement of the Piano Sonata in D
88 Harmony in Schubert

Con moto
legato E == 7 PSSs 7

é Tn 2 “ig 88 “8 a |

ane a

BLnizezite SSse=e
3.33 Piano Sonata in D Major (D. 850), mvmt. 2, mm. 0/1-4.

a tempo i ae eee. a te fe ha <=)

toaaf tater sreei atieefet Pe pci-e 25

e Ee = =
a eo het Se

—
:
yor
ait
gifs = a ae
A Major: I li—( yy I

3.34 Piano Sonata in D Major (D. 850), mvmt. 2, mm. 84/85-88.

Major, the D-F#-A chord at the latter half of measure 2 [3.33] should be
interpreted as II’ in A Major (that is, as the supertonic’s upper-third chord),
not as IV. Root B sounds prominently on beat 2, as resolution of both C#
(soprano) and of A# (inner voice).*° Then a voice exchange between the
chordal fifth and seventh occurs, during which root B is suppressed. As a
result, seventh A migrates downwards an octave. Its resolution pitch, G#, is
placed even lower, in the bass at the beginning of measure 3.
In 3.31, Schubert avoided the diminished chord on the leading tone
during the circle of thirds connecting II and V. In the context of a chordal
glide, however, that chord (in $ position) may assume its rightful position
as one of three chords that come between IJ and V. In another passage from
the second movement of his Piano Sonata in D Major [3.34], Schubert con-
nects I] in A Major (represented by D-F#-B) and V (represented by G#-B-E)
on the lower staff of measures 86 and 87 by means of gliding § chords. The
motivating force for the progression is linear, not harmonic. The passage
does more than just this, however, for in addition to the descending motion
in three voices, a contrasting ascending motion transpires in the soprano
(B<C#<D<D#<E).”’ For the most part, the soprano doubles one or another
of the left-hand pitches. The provocative A-C#-F#-D# chord at the end of
measure 86 results from the concurrent pursuit of multiple linear initiatives.
The well-established motion from II to V, first presented in measures 2 and 3
[3.33], shines through even when a dissonant, chromatic element is added.28
 Common prolongations and successions 89

3.35 Variation on a Waltz by Diabelli (D. 718).

Variation on a Waltz by Diabelli (D. 718)

Unlike Beethoven, Schubert complied with Diabelli’s instruction to create

but one variation for the waltz he was circulating among notable com-
posers. Schubert’s decision to cast his contribution [3.35] in the parallel
minor key (C Minor) results in a tonal plan somewhat at variance with the
model's. Though melodic arpeggiations, chords rearticulated on each beat
Harmony in Schubert

16 17-20 21-30 3] 32
Til sl eels

(a. 229) 1
rN Nn AN

3 > ll

6 = ) $ d el
—— ee
i eS
ez ———

C Major: ae otal: v° iY Sah

(<u) os Ve Vl)

3.36 Analysis of Diabelli’s Waltz in C Major.

over several measures, sequential or circular progressions with internal

two-chord repetitions, and the cadential formula confirm an alliance with
Diabelli’s waltz, the prominence of tonic’s lower- and upper-third chords in
Schubert’s variation reminds us how compositional choices in the minor
mode often differ from those in the major.
The model displayed in 3.36 corresponds to Diabelli’s Waltz in C Major.
The structure is conventional,” though some surface features (not displayed
in the graph) add special interest: the initial I-° (C major to A minor) is
navigated via a segment of the chromaticized ascending 5-6 sequence (C*/
E°—>F5-F#°—G°-G#°— A°), with each 6 phase modified in accordance
with dominant emulation so that, for example, E° appears as Bb-C-E(-G); a
similar sequential motion connects I and V within the tonic prolongation
of measures 24 through 29 (C°/E°-F°-F#°G°); and the closing descent
(measure 29 onwards) occurs in a higher register.
The model for Schubert's variation [3.37] is startlingly different from
3.36 in one respect: the DG root succession that closes the first half in the
C Major version is moved up a half step to E,—A} in Schubert's C Minor
version. Thus one of the three principal chords in my analysis of Diabelli’s
conception (those with outer voices attached to the principal beams in
the graph) does not occur at that critical juncture: 3.37 shows that 3 (the
Kopfton) remains unperturbed throughout this region. Eventually the tonic
returns (as in the C Major version), and so the submediant ultimately serves
as tonic’ lower-third chord, in a broad neighboring context (I>*). There
is something beguiling about measures 17 through 20 of Schubert’s version,
for the Ab chord takes on dominant-emulating characteristics even more
 Common prolongations and successions

me ese 12 S14 15 16 17-20 21-24 27-28 29 30 31 By)

A
rN Aw
3 2 1

C Minor: I° S oe ies 1 evan eal

(= Ab Major: ve ‘ I )

3.37 Analysis of Schubert's Variation on a Waltz by Diabelli (D. 718).

pronounced than those that Diabelli employs with the bona fide dominant
G in the corresponding passage of his version. Yet Ab— does not proceed
to Db. Tonic returns instead.*° As was also the case with the return in
measure 8,°' the C minor tonic chord is displaced by a dominant-emulating
variant (here C-Es-G-B}-D)) that propels a circular progression to tonic’s
upper-third chord, which arrives in measure 27.°* Consequently dominant-
emulating chords on A} and C are juxtaposed. (Because 3.37 displays a
relatively deep level of structure, foreground Es is not shown at measures
8-9 or 21-24.)
One of the thornier questions when graphing an interruption structure
concerns the positioning of the background dominant and 2.3 Two promi-
nent dominant chords occur in 3.36: that of measures 16 through 20, and
that of measure 31. The model presents the first instance as a background
event, a conception in agreement with a number of Schenker’s graphs.**
In Schubert’s variation, the employment of Ab rather than G as root in
measures 16 through 20 removes the interruption. His deployment of
tonic’s lower-third chord results in a more extended prolongation of the
initial tonic and of 3.3° Measure 31 contains the only dominant and the
only 2 available for duty at the background level. Consequently Diabelli’s
waltz and Schubert’s variation are at considerable variance in their deep
structures.
- Chords on }II, on III, and from the parallel key

Ill

I-III via the circle of fifths

In the scheme of seven diatonic chords configured as a circle of fifths, the

tonic and the mediant are at some distance: three descending fifths, or four
ascending fifths. The descending route is particularly appealing in a minor
key, since all the fifths between adjacent roots are of perfect quality, whereas
the same trajectory in a major key would encounter a melodic diminished
fifth (between 4 and 7) and a diminished triad on 7.1 The opposite situa-
tion prevails in the ascending circle: a progression of ascending fifths from
a major tonic leads effectively to the mediant, whereas from a minor tonic
a diminished fifth separates 3 and 6, and the chord on 2 is of diminished
quality.
Because of its size, any root succession of a fifth invites intervening
chords, such as those connecting roots C and F in “Der Jager” from Die
schone Miillerin [4.1]. A similar structure guides the opening measures
of “Von Ida” [4.2], a work of an entirely different character. After two
measures of delicate alternation between tonic chords in 3 and 3 position,
a descending glide emanates from the §. Whereas “Der Jager” employs
four ° chords in the initial fifth-descent, with an underlying rhythm that
evokes the rugged hunter,” in “Von Ida” Schubert engages all five diatonic $
chords that fall within the F>B} span in an even and peaceful flow, evoking
the dawning of a beautiful day. This idyll would be rudely shattered if,
after such a lovely and relaxed F>Bb fifth, the remaining links of the circle
(Bb-Eb-A}) were accomplished by the downbeat of measure 4, interpreting
Bb-G-F at the end of measure 3 as representative of E} (with absent root
and added ninth) and the following C-Ab-Eb as representative of Ab} (in first
inversion). Instead, the proportions of this passage match those of “Der
Jager”: the Bb chord is prolonged for two beats (G is a passing note), while
in the «5 embellishment of the Eb chord, the ;is unfurled as a §,leading to an
inversion of the Eb’ chord. These two phrases, which exude such contrast-
ing characters, are identical in their foundational structures.
 Chords on bII, on III, and from the parallel key

oS
a
5 s 6 SS@ SS
oe eo e oi
o|*7 #
7 J eo t 2 ee
ra o—e rd

Was sucht denn der Ja- geram Miihl - bach hier? bleib’, tro ~ tzi- ger J&é- ger, in dei-nem Re - vier
Doch bes - ser, dublei-bestim Wal-de da- zu, und lie- ssest die Mith-len und Mil-ler_ in Ruh’;

Sle ee
= Sas SSS aS
ie le ¥ aes =~? Z—= eee = f
2 sled —
E ( ) F—— Bb— Eb
(eG Gaus)
C Minor: I ( ) ai

4.1 “Der Jager” from Die schéne Miillerin (D. 795/14), mm. 4/5-8.

Klagend
3

=7 ee
7
Der Mor-gen bliiht: der O- sten gliiht; es a-chelt aus dem diin - nen Flor
Auf wel-cher Flur, auf wes-sen Spur, so fen von I - den wallst__ du _ itzt,
| FE TE

o i (+ eo _* —ae o —=< i & #

sempre pp haa oe

Deeg 3 t
= Gog 7 Gas

Bae Vea ee eee i

6 SSeS Ss ee

FE ( ) Bb Eb Ab
GF Eb DC Bb)
F Minor: I ( )

4.2 “Von Ida” (D. 228), mm. 0/1-4.

Within the realm of tonal travels the fork in the road at the subdomi-
nant is particularly intriguing and consequential. To be sure, the direct
succession IV<V (often with IV expanded via a 5-6 shift) is expedient and
frequently encountered. Yet especially in a minor key, continuation via
descending fifths until III is attained is an appealing alternative. (Thus there
is a question of whether the subdominant functions within a harmonic pro-
gression — as IV - or instead within a circular progression.) In his Moment
Musical in C# Minor, Schubert eventually attains the mediant, though his
first thrust in that direction is diverted back to tonic at the last moment.
The opening eight-measure phrase establishes C# Minor as the tonic.
In the next phrase [4.3], tonic C#-E-G# leads gradually to C#-E-Gh-A#
94 Harmony in Schubert

pad
tia : = fe
SSf = oo
ae
ct p

=
igty—it — 2S pate a: sae ee
==5 ——— rr a 4
e rz e
# ° == = 2 _@ @ t=
eo He | 2 —

B G# Ct Fe B E

4.3 Moment Musical in C# Minor (D. 780/4), mm. 9-17.

Model 1 Model2
m. Om Ie r(13) CLA 815 m. o> 10s Tee wae
4
o — ase + fue ee a ee

on eS Ee =5
[Oe eo ia)

Cl= ee Ge Ce Cl a Gree

Model3
m. O12 8 14 15

tate
! 5

Ci HS BruGinG
ah ae 9
C# Minor: I°~2 1V4:—, Vel
¢ — a

4.4 Analysis of Moment Musical in C# Minor (D. 780/4), mm. 9-15.

(measure 13). Although a diminished seventh chord can resolve in many

ways, its spelling and context here suggest that the chord stems from root
F#. (Compare Models 1 and 2 of 4.4.) In the context of a potential arrival
of the mediant key, E Major, F£ would be the second of four chords in a
circular progression of fifths (C#-F#-B-E). Further mutation ensues, as
 Chords on SII, on III, and from the parallel key S)5,

C Major: I (

4.5 Analysis of Symphony No. 9 (7) in C Major (“The Great” D. 944), mvmt. 3, mm.
229-235.

bass Ck pushes more urgently towards B, which arrives in measure 14 [4.4,

Model 3].
Yet the E Major mediant key is not achieved quite yet. Consider again the

of
fork in the road at F#. Two conventional routes are:

Gt C#

B E

Schubert moves from C# and F# to B, emphasizing B’s dominant-emulating

character through its arrival via augmented-sixth resolution, and then
veers back to G# and C#. In retrospect we interpret the B chord as the
onset of V-space (as displayed in Model 3). The C#-F#-B-E progression is
carried out successfully on a second try, in the final measures of the phrase
(measures 15 through 17).
In the ascending direction the circle of fifths outlines a clear and uncom-
plicated path between a major tonic and the mediant chord or region.
Schubert often enlivens that routine trajectory with chromaticism, going so
far as to bleach out the distinctions among chord qualities during a passage
from the third movement of his “Great” Symphony in C Major [4.5]. Every
chord is of major quality, with descending chromatic passing notes offer-
ing a belated minor flavor to the G, D, and A chords. The mediant arrives
in measure 233, dividing the I<V tonal space into two thirds. Its G# is a
wobble, continuing the major-chord saturation that has persisted through-
out the circular progression. In this context it does not signal an impending
succession to the submediant.
The label III has historically served as the harmonic analysis of chords
performing a wide range of functions. I propose that this chord sometimes
functions within I-space, as the tonic’s upper-third chord [1.5, 2.23 (Model
2), 3.3 (Model 2), 3.24]; sometimes within V-space, as the dominant's
lower-third chord [2.19]; and sometimes as a true mediant - a chord that
96 Harmony in Schubert

(ae :
ote See Ne
be = i

CMinor: I P! Va’ Il---

(= EbMajor: Vv’ I)

4.6 Model for a I-VII-III progression in C Minor.

asserts itself as an individual entity on the path between the tonic and the
dominant [4.5].

I-III via III's upper-fifth chord

In the Moment Musical in C# Minor [4.3], Schubert's juxtaposition of B major

and G# major chords in measure 14 is a particularly poignant moment. The
choice between two paths is being negotiated not by the vying E and C? goals
themselves, but by chords capable of propelling the progression to those goals.
The transformative power vested in that descending third can be reversed:
the ascending third from the dominant is an efficient way to steer a progres-
sion towards the mediant. In the context of a major tonic, the diminished
chord on the leading tone is ineffectual for such a strategy. Instead, one of
two chromatic modifications may be employed: either the subtonic will
target the lowered mediant (e.g. B)Eb in C Major) or the major chord on
the leading tone will target the diatonic mediant (B—E in C Major). If the
goal mediant is tonicized, then the subtonic or altered leading-tone chord
might be interpreted as V in that key.°
The model of 4.6 displays the smooth voice leading that often prevails in
a progression employing the subtonic. In the second chord the leading tone
may retain an upward tendency, as tonic C’s lower neighbor. By proceeding
downward an additional half step to subtonic Bb (with Bg thus serving as a
passing note), the mediant, which may be tonicized, is targeted. Employing
the subtonic in the context of C Major rather than C Minor, only one pitch
of this model would change: Es in place of Eb in the first chord. The goal
would remain E} major, the lowered mediant.
Returning again to the Moment Musical in C# Minor, Schubert employs
the parallel major key, spelled as Db Major, for the B section of the move-
ments A,-B-A, ternary form. (This B section is itself organized as an |:a,
:|: b a, :| ternary - a.k.a. “rounded binary” - form, the a, section of which
is presented in 4.7.) Whereas chords rooted on B (upper fifth of E) and
G# (upper fifth of C%) were juxtaposed in measure 14, enharmonically
 Chords on $II, on III, and from the parallel key

Car
oa
eet
s See
er
pr esteearierr ee=Ser=:
mip S$ 61S 6 tle ¢ gis ¢ gle ¢ ¢ le ¢ 6)
SS
D> Major:I

ee e
eS fmt ae
a I ae
air
—
*t
7s
a
#8 ;

at oe PP
pZ t a S e
2 #2 lo oe

ore eae”
ei rae sar
( ) pv” bi
(= D> Minor: Vi’ IIL)

4.7 Moment Musical in C# Minor (D. 780/4), mm. 62-73.

equivalent Ab and C) chords — in that order — are juxtaposed in measures

70 through 73. These chords are transposed representatives of the second
and third chords of 4.6. The progression continues to Fb (bIII*), which
concludes the a, section. This turn of events gives a minor-key hew to the
entire B section, first hinted at by the Fb neighboring notes of measures
70 through 72. One might assume that Schubert chose Db Major for the
middle section to provide a contrast to the C# Minor of the outer sections.
Yet F) Major is none other than E Major, the minor-key mediant likewise
attained in measure 17 of the A, section [4.3]. The return of Db Major late
in the B section (measure 93) is a surprise, because the pitch Fb has been so
persistent since measure 70.
Schubert employs a tonic-mediant connection in “Die Berge” [4.8] that,
in its broad parameters, follows a normative path:
m. 25 34 38 39

G D Fe b
|
I-V in G Major

| |
V in G Major to V# in B Minor

V#-I in B Minor
98 Harmony in Schubert

a = eS See wie
{—b

e-wig
Bald musser stau-nend mer- ken,

ee
Eueeee Se | — Ls

a aoe aa ig! abel

paar_——_—— 3
aS eee
modal shift

29

6G = = : set
pe —e : sao e aac
as
dann strebt in si - chern

al

P
——|

Eb
Re a Cc
5-6!
(= A)
34
$ E
Se Ss === = :

Gatcughieaies bbe Ne
Wer-ken seingan - zes Thun, ver-biin - det, vom Grun - de nie zu

“t—e ——— a tet 2 — —— ———

are cresc. i
—= er | — —=
= Z ae
: 30 ad ie be ote == ae
— ne
ate Fe Fz

D (dominant of G?) E (dominant of A?)

38

Gn = = =e =: o = —s =e = =

ayea
wan - ken, und baut wie Fel - sen den Bau der Ge- dan- ze

rete —— Se
4

oer Pe be iz, ; Zz Be fsPA |Z4

pte
SSS SSSet es
= SS
F¢ (dominant of B!) B ene

4.8 “Die Berge” (D. 634), mm. 24-42,

 Chords on bII, on III, and from the parallel key 2s)

Yet the first two successions are accomplished in innovative ways. The
former begins with a modal shift from G Major to G Minor, whose sub-
mediant Eb major (the 6 phase of I>~6) arrives soon thereafter. This event
is duly celebrated by a cadential progression (measures 30 through 32).
Schubert employs § chords reminiscent of those in measure 7 of “Der
Jager” [4.1] to lead downwards a third towards IV} in measures 32 and 33.
In place of subdominant Eb-G-C we find a mutant Eb-G-A-C¢ chord in the
second half of measure 33. (Pitch A is the 6 of IV*-°, which sounds along
with the 5, while C# is a conventional dominant-leaning chromatic inflec-
tion.) This progression presses very insistently towards dominant D major.
Yet Schubert does not rest once this I>V path has been traversed. D’s
arrival is not strongly marked; it is not even a point of punctuation in the
poetic text. Indeed, an ascending initiative is imminent. In some extraor-
dinary writing, amidst a crescendo ultimately to fortissimo, Schubert
pursues a strategy of hoisting the material targeting dominant D up a step,
first to goal E, and then up another step to goal F# (measures 34 through
38). Because the initial path towards D so vividly proclaims an immi-
nent dominant arrival, D, E, and F# in turn each seems to be the rightful
dominant — at least until the next claimant arrives. By embracing F# and
bestowing upon it the full trappings of dominant legitimacy (seventh E and
even a fleeting tinge of ninth G in measure 38), Schubert precipitates the
succession to B Minor.’

I<III via sequence or parallel progression

As we observed in “Die Berge” [4.8], Schubert may navigate the ascending

third between I and III by traversing the ascending third between their
respective dominants. More directly, the tonic and mediant chords may
themselves be so connected. A diatonic ascending 5-6 sequence starting
on tonic reaches the mediant at its fifth chord (e.g., C>® D®® E?), though
E> may seem an awkward stopping point, since leading-tone B (its upward
tendency enhanced by the ? interval in D°) does not ascend to C, continu-
ing the 5-6 motion. Added chromaticism in the service of dominant emu-
lation may direct the progression decisively towards III, as in Schubert's
“Gesang der Geister iiber den Wassern’ [4.9].
Schubert glides from I to III in a passage from the first movement of his
String Quartet in D Minor [4.10]. The mechanics of the voice leading are
shown in 4.11. This I<III connection sets the stage for his I<IV glide later in
the movement, which we encountered in 3.29.
100 Harmony in Schubert

m. 134 135 136

2 - 4
> bf | Ka i
5— 6 5 —=16 )

C Major: I ( )I

714), mm. 134-136.

4.9 Analysis of “Gesang der Geister iiber den Wassern” (D.

6 2 = ™ a ze
p ff |e fe |e — fe |ferescfe |
5 + —}

o |5 to ‘J Seri a ao e
6 p F ta cresc.

SS Vn 3 3
SSS 3 3
a
4
2 p i ee = £ 2 z JH
p : cresc.

D Minor: |———————- ( pL amen Via

4.10 String Quartet No. 14 in D Minor (D. 810), mvmt. 1, mm. 15-19.

m, 15 17

aus at at

Di 6 SEP ee
38
D Minor: I ( )

4.11 Analysis of String Quartet No. 14 in D Minor (D. 810), mvmt. 1, mm. 15-18.

III<V in direct succession or via the fourth scale degree

As its name suggests, the mediant chord often serves to divide the I<V
span into two smaller segments. V may follow directly after III, as in the
excerpt from Schubert's C Major Symphony displayed in 4.5; or, the III<V
span may be traversed more gradually. An ascending 5-6 sequence offers a
simple means of connection. In “Tranenregen” from Die schéne Miillerin,
a descending circle of fifths serves to connect the I-III span (rendered as a
descending sixth from A to Cz in the bass), while a chromaticized variant
 Chords on bII, on III, and from the parallel key 101

25
Gis s =
yy 7 gc “NS
et Z peas “2 ia
Da gin-gendie Au-genmir ii - ber, dawardesim Spie-gelso kraus, sie
———

eee
mats==
A

A Minor: I

29
- 4 \ _
!
| \
s
=, dt Se SS y2
e y = 7
sprach: es kommt ein Re - gen, a = de; ich geh' nach Haus.

4.12 “Tranenregen” from Die schéne Miillerin (D. 795/10), mm. 24/25-32.

of the ascending 5-6 sequence connects the III<V span. (See 4.12.) In other
cases the fourth scale degree may play a more potent role in the bass, to the
extent that its chord (generally IV or an inverted II) is structurally deeper
than III. In the first case, the mediant serves as a divider of the I<V span
(thus I<III<V), whereas in the second it serves as tonic’s upper-third chord,
extending I-space en route to a II or IV chord.

Modal mixture

A key is comprised of a tonic pitch and a specific set of auxiliary diatonic

pitches. To refer to “the key of C Major” is to organize the twelve pitch
classes within three categories: tonic (C), the remaining diatonic members
102 Harmony in Schubert

B Major. 2 ——— V I:
B Minor: I V? I

4.13 Analysis of “Die bése Farbe” from Die schone Miillerin (D. 795/17), Introduction
(mm. 1-4).

(D, E, EG, A, B), and the chromatic non-members (C#/Db, D#/Eb, F#/Gb,
G#/A}b, A#/B}). The non-members may appear as “guests” within a key,
fulfilling duties similar to those of their diatonic counterparts, though gen-
erally more emphatically or with heightened sensitivity. Three of the chro-
matic pitch classes have a special status in C Major due to their membership
in the parallel minor key, C Minor: Eb, Ab, and Bb. Though Db is not
diatonic in C Minor, its frequent occurrence there gives it a minor-mode
flavor within C Major as well. Raised pitches in a major key come about for
a variety of reasons, most often through dominant emulation or toniciza-
tion. Of course, in a minor key the raising of the third, sixth, and seventh
scale degrees is reciprocal to their lowering in the context of a major key.
The intermixing of pitches from parallel keys is often referred to as modal
mixture.
Towards the end of Schubert's twenty-song cycle Die schéne Miillerin, the
miller lad finds it ever harder to endure his situation, for the lovely miller
maid, to whom his attentions are directed, is clearly enamored of the hunter.
The poet Wilhelm Miiller represents the hunter through his characteristic
color, green. In “Die bose Farbe” (‘The Vile Color), the miller lad’s attempts to
sustain a positive outlook are repeatedly thwarted, represented by Schubert
through phrases that, beginning in B Major, are gradually infiltrated by guest
pitches from B Minor. Even in the keyboard introduction [4.13], tonic’s dia-
tonic third is displaced by its lowered counterpart (D# to Dk).
B Major is restored at the onset of the vocal line, and an entire phrase
(progressing from I to V in measures 5 through 8) transpires without
parallel-minor inflection. Yet the second vocal phrase veers decisively in
that direction, as shown in 4.14. Diatonic I>~* leads into II# (measure 11).
This II# is embellished by a $chord that in its initial phase Schubert unfurls
into $ position, at which point A4 inaugurates the modal shift. The domi-
nant that arrives at the phrase ending — also the stanza ending - is that of B
Minor. (The dominant’s major third is restored later in measure 12, during
the segue into measure 13’s B Major tonic.)
 Chords on sll, on ILI, and from the parallel key 103

m. 5 8 9 10 1] 12

3
yg *
—
RN
|g
a NINE =a as

B Major: I eg ook ie |e je Ve
4.14 Analysis of “Die bése Farbe” from Die schéne Miillerin (D. 795/17), stanza 1
(mm. 5-12).

m 13 16 17 18 19 20 21 22
7, Ge
g t9 —-_ L

Dfey? a Fs —¥— 5-49 +f

Sit =e OD O Sey) a
ow a vig <a =

(5)—6: ee!
eee ee VO. IV |Ve
8—7 56
=I V i) (= HI°’)

4.15 Analysis of “Die bése Farbe” from Die schéne Miillerin (D. 795/17), stanza 2
(mm. 13-22).

The opening of the second stanza (measures 13 through 16) matches that
of the first (measures 5 through 8), with motion from I to V, though now the
V is prolonged for an additional measure, with resolution to a minor tonic
in measure 18 [4.15]. If the dimensions of the first stanza and the conven-
tions of antecedent/consequent construction are to be maintained, measure
20 should be the stanza-ending measure. Schubert prevents closure there by
employing a first-inversion minor tonic chord and by covering the melodic
close B (= 1) in the keyboard line with a vocal F$. These evasions motivate
a two-measure phrase extension, during which a continuing expression
of the parallel minor key emanates from the lowered supertonic chord
(E-G+-C:), here interpreted as a chromaticized 6 phase of IV.” The stanza
ends with a bizarre surprise: after four measures during which pitches of B
Minor clearly supplant those of B Major, the concluding tonic in measure
22 is B-D2-F2!
Dz is ofcentral importance at the outset of the next stanza. Along with
As, it imbues tonic with a dominant-emulating character, initiating a
descending circle of fifths that pursues the minor-mode course of B~E-Az—
Ds (in lieu of the awkward major-mode B-E-Az-D8), reaching the lowered
mediant at measure 26 [4.16]. A 5-6 motion on Dz propels the progression
onwards to the dominant, attained in the stanza’s sixth measure and pro-
longed for two additional measures via coordinated neighboring motions
below prolonged C2. Given the pervasive minor-mode infiltration within
104 Harmony in Schubert

25 26 pai 28 e
m. DS 24

Do Bao bg EE aS
a
Barer ‘<7 — 6 ee
L 2 eo + @
t= :
B E Ad Dh Rega
B Major: I ( Pil 2 ey
B Minor: Ft ( \. UII 2 oata

4,16 Analysis of “Die bése Farbe” from Die schone Miillerin (D. 795/17), stanza 3
(mm. 23-30).

this stanza, one may question whether B Major should be retained as the
key for the analysis. An alternative B Minor version of the symbols is dis-
played in 4.16.° In fact, since the remainder of the song corresponds mainly
to the structures displayed in 4.13 through 4.16, and even ends with the
major-to-minor transition of 4.13, one might well question whether “Die
bése Farbe” is actually in B Major after all.” That conundrum corresponds
to the consternation in the miller lad’s mind at this point in his story: can he
win over the miller maid, or will his life become pointless through failure to
do so? Three songs later his death (a suicide by drowning) is conveyed via a
gentle lament in the key of E Major.
Schubert’s Moment Musical in Ab Major is rich in chromatically modi-
fied chords. At the outset the tonic chord is embellished by three neighbor-
ing notes, including an upper neighbor to its third in the melody: C<D}>C
[4.17]. When the dominant chord’s fifth is similarly embellished in meas-
ures 4 through 6, Schubert employs B)<C)>B}, maintaining the half-step
relationship that had prevailed in the tonic passage, instead of diatonic
Bb<Ck>Bb. A full-fledged Eb minor chord occurs during measures 6 and 7.
The broader tonal plan likewise betrays a borrowing from the parallel
minor, though a contorted one. As the form’s A, section concludes, the
Ab tonic chord adopts chromatic Gb as its seventh (an instance of domi-
nant emulation), leading to a cadence on the subdominant in measure 17.
The role of root Db within the tonal scheme remains uncertain, though
the soprano Db} clearly derives from the persistent C<D}>C neighboring
motive of the preceding measures [4.18]. At first root Dh seems to be a
link within a descending circle of fifths: A>>Db—>Gb. G} minor (spelled
as F# minor in the score) arrives at measure 18 and is tonicized through
measure 31, thus encompassing the entire B, section. On the one hand,
the Gb root is justified by the progression of descending fifths, though the
chord’s quality is minor. (More on that presently.) On the other hand, once
dominant Eb arrives in measure 34 and decisively leads back to Ab Major,
 Chords on bII, on II, and from the parallel key 105

Andantino.

Ab Major: I

ates
5

=a , Es Pare
Bae
rz ie: -
oe s 53 ——s!
a ee ed
Se = =p

oe =e a 2 z
pee ere. rae
a a rr i a oo
, a Vo=3
1V32 he)
(= Eb Major: I:

(GC 14.)

4.17 Moment Musical in Ab Major (D. 780/2), mm. 0/1-8.

oe 16, Sie MS: 33 34” 36

‘gl; 4
6 Poe Ghgy, BS
a g= Cer:Lie de
Ab Major: [8—” re I

Ay B, Transition A,

4.18 Synopsis of Moment Musical in Ab Major (D. 780/2), mm. 0/1-36.

the aberrational G} minor chord may be understood as the third chromatic

variant of Eb major’s upper-third chord, and thus as the onset of V-space.*
(The relationship between Eb-G-Bb and its chromatic upper-third chord,
G}-Bbb-D}, is camouflaged by the enharmonic spelling of convenience,
F#-A-C#.)
By any standard the keys of Ab Major and Gb Minor are distantly related.”
Their key signatures differ by five flats (four versus nine). Clearly Schubert
intends for the tonal region of the B, section, delicately stated at a pianissimo
dynamic level, to serve as a potent contrast to the Ab Major of the A, and A,
sections that surround it. Yet something even more daring is in store later in
106 Harmony in Schubert

the work. The A, section, which begins at the upbeat to measure 36, possesses
a concluding section not found in A,, starting after the perfect authentic
cadence in measure 47. Consequently first-time listeners likely will surmise
that the piece ends at measure 55, as the supposed concluding section comes
to a close. To their surprise, without preparation like that of measure 17 and
at a forte dynamic level, Gb [F#] Minor comes crashing in again. An entire B,
section is thereby launched. Schubert compensates for that outburst by devis-
ing a more gradual return to Ab. Whereas the initial redirection towards Ab
Major (measures 31 through 35) involved two concurrent chromatic shifts:

Geb Dbe ito

Le Ges
in this latter case Gb Minor first is transformed into Gb Major (the diatonic
subtonic in Ab Major’s parallel minor key, measure 69), after which the E}
dominant arrives in measure 72:

Gb Bbb Db to
Gb Bb Db to
Eb Gk Bb Db
(See 4.19.) Only during this second traversal of the B material is Bbb’s role
as a wobbly note conventionally resolved. That wobble permeated the entire
first statement of the B material. By eventually embracing the normative
subtonic during the second B section, Schubert honors the traditional
interface between parallel keys. (Only a chromatic key can play the role
that Gb Minor and G} Major play in this work, because Ab Major’s diatonic
G-Bb-D}, a dissonant triad, cannot be tonicized.) The juxtaposition of Ab
Major and Gb Minor (measures 55 and 56) suggests an extreme emotional
torment. As listeners we internalize that pain, and thus we experience relief
when Gb Major emerges. In fact, a rare sweetness results from the progres-
sion of major triads: Gb major (measure 69), Eb major with minor seventh
(measures 72 and 73), and finally the Ab major tonic (measures 73 and
beyond). This progression is unavailable in a purely diatonic context. It aptly
demonstrates the widened resources — and extended emotional range — that
become available through the mixing of major- and minor-mode pitches.

The lowered supertonic chord

Most harmonic progressions may occur in either a major- or a minor-mode

context. Such progressions sometimes undergo chromatic adjustment
 Chords on SII, on III, and from the parallel key 107

SSD
on ESerphet TE Serepsat

anlar
etett tertenteriet
= eS Se Ere
Cae= ef ial 4
: PP.

SS = jaa

4.19 Moment Musical in Ab Major (D. 780/2), mm. 65/66-77.

Ze Pp (lo — 83—— 7 —
by
OO] ANS
Y/N
ae oh 7 |
Lni ian —
I C Minor: [°° (DIT Vs I

4.20 Models for the P-°-II-V’-I progression.

when they migrate from one mode to the other, as in the cadence from
dominant to tonic in a minor key, where the leading tone substitutes for the
subtonic pitch. Progressions involving the supertonic - such as ?-°-II-V’-I
in major [4.20, Model 1] - likewise are often modified when deployed in a
minor key [4.20, Model 2]. The 6 phase of I may or may not contain a chro-
matic C# in the major-key version; but when it does, dominant emulation
becomes a dynamic factor. The C#-E-A chord makes the following D chord
seem inevitable. In a minor key the diatonic pitch content of [>-§
P~° is naturally
imbued with dominant emulation in its 6 phase (C-E}-A}), but the chord
108 Harmony in Schubert

GE iby 3) AX. 44 45 46 47 48 49) 53) 57

ant, 25) A RO) Sa sigh

a i= : be gz ga
_——— Se
He
a:
oe ba

6 5 6 5 6 :
3) (5)

C Minor: I° ( )®-6 bite V5

4.21 Synopsis of Piano Sonata in C Minor (D. 958), mvmt. 4, mm. 29=97.

to which it would resolve is not diatonic. Instead it targets a chromatic goal,

the major chord on the lowered second scale degree (Db-F-Ab). That chord
often substitutes for the diatonic II, which is of diminished quality. Though
LII has been interpreted in a variety of ways,'° the progression of 4.20,
Model 2, reveals that it is inherent in the arrangement of pitches within a
minor key to lead logically and effortlessly to bII. Once ensconced in minor,
transfer to the context of a major key likewise becomes feasible.
When bII and V are juxtaposed, as in 4.20, Model 2, one recognizes that
the lowered second scale degree is a wobble - a displacement of the dia-
tonic 2, which is a member of the hierarchically deeper dominant harmony.
Rarely will 42 and 2 be juxtaposed in the soprano melody, however.
Composers generally proceed from the lowered 2 via downward motion
to the leading tone, perhaps with an intervening passing note, such as
Db>C>Bk in C Minor. Yet even in that context a structural 2 is understood
to be operative during V3."!
An impressive extension of P° bII V/ occurs in the fourth movement
of Schubert’s Piano Sonata in C Minor [4.21]. At first it may seem as if the
D> major chord of measure 29 functions harmonically as }II. Yet in the
broader context that eventually takes shape, we realize that a linear initia-
tive extends for over an octave. This initiative begins as an ascending 5-6
sequence but eventually mutates into a glide of §chords (some of which are
extended via their upper-fifth chords, as indicated by the filled-in noteheads
in the example).'* The correspondences with 4.20, Model 2, are indicated by
asterisks.
The bII sonority comes about in two principal ways: either through the
wobble of a minor-key II’s root down a half step, as in the C Minor Sonata
excerpt discussed above, or through the raising of its IV’s fifth by a half step,
as in “Der Miller und der Bach” from Die schéne Miillerin [4.22].!3 The
excerpt'’s three motions to V# are progressively more ambitious. (See 4.23.)
In the first, the tonic root persists as a pedal during the dominant chord,
which barely asserts itself.!4 In the second, the dominant attains some inde-
pendence from tonic (measure 9), functioning within a progression guided
by a G>D<G bass arpeggiation. The C that precedes this D root would most
 Chords on $II, on III, and from the parallel key 109

Massig (Der Miiller) N 4 4

== eee
SS Wo eintreu-es Her-ze in
ee
Lie - be ver- geht,
ee,
eS da_ wel - ken die
— t t y t ——
;

s a 3 +3 2 = = : =)
L | Ss :

8
Z a. | o—| = 7 =e
Sale Ae = Se ree
Li-lien auf je - dem Beet; da muss indie Wol-ken der Woll- mond

sa Fa — Se = :
is ae 2 ine i r aa
SSS SSS SS SSS
5 = { : a T |
: z @ : :

14
=e = Shes ra
comes et peter Sr eee ——
geh'n, da-mit sei-ne Thra-nen die Men- schen nicht seh'n; da

Be = SSS
Se | iss
¥ :
i—t } | |

— te =a v-

4.22 “Der Miller und der Bach” from Die schéne Miillerin (D. 795/19), mm. 1-19.

normatively support a IV chord, as shown in 4.23. The migration of IV’s

fifth G up a half step to Ab produces the characteristic bII sonority, yet in
this context the A} functions essentially as a neighboring embellishment
preceding that strand’s descent to the leading tone, F#.'° This dependence
upon the subdominant is expressed in analytical notation by the fact that
IV appears as the principal harmony, with the bII variant displayed (grudg-
ingly) within parentheses. The third motion to the dominant, the first in
which V# serves as a phrase-ending goal, employs IV (5II) in a similar way,
while what precedes and follows that chord is more fully developed. III
(measure 14) is a robust presence, attained via a descending circle of fifths.
The connection between III and IV is achieved via a 5-6 shift whose 6 phase
is unfurled into ? position in the score. The IV (with fifth elided) likewise
enjoys a 5-6 expansion. Its 6 phase, C-Eb-A, undergoes transformation
110 Harmony in Schubert

Model 1
m. ] 4 6

Le
Py qf

a oD a
G Minor: I :
(= Vi)

Model2
m 1 8 9 10

2 eS
eae p bw

ae
ae i
GMinor™! Ivo” Vit
(= bI)

Model3
m. 10 14 15 16 ig? ie

3 ee =e Ss i
6 = 3 = —z o —— fe ag

G Minor: I( aT? eae V8 OREN

(= bi)
(= M55)

4.23 Analysis of “Der Miller und der Bach” from Die schone Miillerin (D. 795/19),
mm. 1-19.

resulting in C#-E4-G-Bb. Before that 6 phase, Ab embellishes the IV chord’s

fifth, as it did in measure 8.
Without attempting a full accounting of the myriad strategies Schubert
deployed to attain the bII sonority, an excerpt in A Major from his Piano
Sonata in D Major [4.24] illustrates his extraordinary creativity in that
regard. Here the diminished triad A#-C#-E is a critical linchpin. That sonority
first occurs in measure 15. Model 1 of 4.25 shows how the chord represents
a chromatically modified, dominant-emulating 6 phase of I. As expected, it
leads to II. V follows thereafter (measure 18) in conventional fashion. At first
it appears that this V will be prolonged via the concurrent linear filling-in of
two of its thirds: G#<B and B<D (realized in 4.25, Model 2). Yet observe that
measure 20 is identical in sound to measure 15. In Schubert's revised spell-
ing, the chord contains both an augmented second and an augmented fourth.
These two dissonant intervals resolve by ascending half step in measure 21,
 Chords on $II, on III, and from the parallel key 111

atempo___13
a

4.24 Piano Sonata in D Major (D. 850), mvmt. 2, mm. 12/13-29.

over stable bass Bb. (See 4.25, Model 3.) The sonority that results is bII°',
which is boldly tonicized through measure 27, after which the dominant
returns, leading ultimately to tonic. The dominant return is itself quite inter-
esting. Three phases of harmonic chordal content occur during measure 27:
112 Harmony in Schubert

Model 1

git
16

ss
es

A Major: Ee

Model 2: plus hypothetical continuation

m. 18 19 20 te

[i
# (a)

: = -7
A Major: vou

Model3
m. 18 19 20 P| 27

A Major: von yet

ev ( ) bie Vo
4.25 Analysis of Piano Sonata in D Major (D. 850), mvmt. 2, mm. 14-27.

Bb-D-Fb, bII># in root position.

B)-D-F4-G#, achord that merges the leading tone and the pitches
of bII*. The dominant’s third, lowered fifth, seventh,
and minor ninth are present, tending towards a tonic
resolution. This formulation, particularly with Bb
positioned in the bass, is a tonic-resolving version of a
familiar “augmented sixth” chord.
G#-B-D-E, with B} replaced by diatonic Bk and with ninth F4
lowered to root E, much of the tension of the preceding
chord is released. The chord is the first inversion of V’.

The duration of the bII* prolongation in 4.24 is striking. The second

model of 4.25 proposes a dominant that persists for four measures. The
nearly seven-measure expansion of }II°* (that is, about half of the total
content between the tonics of measures 15 and 28) within the broader
dominant prolongation gives durational priority to a chord that is struc-
turally not even a part of the passage’s basic A~-B-E-A trajectory (marked
 Chords on bII, on III, and from the parallel key 13

as I-II-V’-I in 4.24). Apparently Schubert was entranced by the E-B)-E

antipodal relationship he had created (a contortion of the more conven-
tional diatonic upper-fifth prolongation of the dominant, as in 3.13) and
endeavored to increase the tension already inherent in that juxtaposi-
tion by holding out on Bb for as long as possible before returning to E in
measure 27.

“Die Einsiedelei” (D. 337)

“Die Einsiedelei” (The Hermitage) [4.26] describes a retreat deep in an

oak forest. Though a performance of the chorus requires at least four male
voices, the poem’ protagonist is the singular ich (I), not the plural wir (we).
The transition from G Minor to G Major near the end relates not so much
to the Gebiisch (thicket) of stanza 1 or the Seufzer (sigh) of stanza 2, but
to the exhilarating sense of freedom that oneness with nature induces, as
described at the end of stanza 3.
Though the pitch class represented by the raised fourth or lowered fifth
scale degree occurs frequently in music (for example, as a member of II#), a
chord rooted there is a rarity. It is tonic’s antipode. Beethoven touches upon
it in the Funeral March of his Piano Sonata in Ab Major, op. 26, as does
Liszt in his funereal “Il penseroso” - The Pensive One - from the Italian
volume of his Années de pélerinage.'® In both cases it is achieved through
an upward trajectory of two minor thirds from tonic. Schubert likewise
grazes Db in the context of G Minor in “Die Einsiedelei.” All three of these
works are tonally closed in a conventional sense despite their bold initial
trajectories.'’ Since Beethoven was a model for both Schubert and Liszt,
it is perhaps not coincidental that all three compositions work their way
upwards to the antipode in the same manner: first a conventional progres-
sion from a minor tonic to its diatonic upper-third chord, and then the
same strategy applied to this chord (converted to minor), leading upwards
another minor third. In “Die Einsiedelei” the opening measures traverse
the path from g to Bb. That content, transposed, then connects bb and Db.
A proposal for how these chords function in their context is offered in
4.27. Tonic G Minor is prolonged initially via its upper-third chord, Bb
major. Though Schubert devotes two measures to establishing the Bb chord,
its role within the structure is as yet uncertain: will tonic G return, confirm-
ing an embellishing role for the B> chord; or will B} assert itself as III? The
answer comes in measure 5, where Bb (now with minor hue) is projected by
the same structure that was employed to project tonic in the first phrase.'*
114 Harmony in Schubert

Ziemlich geschwind 5
aa mf = \, > po
mp, Ss 1)
So = 2-e a yee i
Tenoree I Ge g e a} a, Sie
}
2 one
t _ 4 —
i 7
t t
——a— 7
7

1. Es rie-selt klar.und we-hend ein_Quellim Ei-chen wald, da_wiahlich ein sam ge-hend mir_

m, Pp i! a Ni P
: Tenore II Gs =
= SSS
z ze
=i ‘
Nhe s — = 5 aac ae =e oep
2.Zwar dii-ster_ ist_und trii-berdie na-he Wii-ste- nei, al - leinnur de - sto lie -berder

mf ~ FF ste #P, _mf, TIPE

z = Sie = Pe
Basso :
SS' I S,Z
ane ;ees
2 eer
Sue, =o- —7os f 2 5 a:= Pe oe

3. Wie. sichdas Herzer-we-tertim en-gendich-ten Wald, den 6-den Triib-sinn hei -tertder
mf = Pp mh
Na Sie
Sew AN
Basso II eg 4 ae — te ie Fuses : ——

; be iP ave cresc. LES REN =

= Z SS == z f—
—
@ =a aa == a
mei-nen Auf - ent alt. Mir die-net zur_ Ka - pel - le ein Grétt-chenduf - tig frisch__ — zu_
Pp ee cresc. “ P eX mf.
ft ae oe. — oo
= == me Sees : == S|
ra 5

stil-len Fan - ta - sei._—_— Da ruh' ich oft im dich - ten, be - bliim-ten Hei - de- kraut;__ hoch
l p crese.| —<—~ fe lan Lip
eae ~
es 2]732S = f— SEs
se ———-—-— ae ae ee
= i 7 —— =
trau-te Schat- ten bald.__——- Kein ii - ber-leg - ner Spa - her er-forschthier mei-ne Spur; __ hier
yc : = = ; —— ! crese. Pa mf
g : be z t i —— z = z= —-s
p—ha —a —# 2 7 +26 ire F a } oe

13 ‘
ch —— : o EP 2 oe e eo PPI
6 Sa a SF
mei - ner Klaus - ner - zel - le ver - schlun - ge - nes Ge - _ biisch, Zu
as 5 haa —— ——— ide PPP

ee 5 = wehn die schwan - ken Fich

————
~- ten
== und___ stdh_ - nen Seuf - zer
——— laut, hoch
s

———— : — =e
ES aoa PP ~~ Ppp

p ———— _— é i :
bin ich frei und nd, (= her der____ Ein - falt und Na - tur, hier
s a PP PPP.
ere = = ———
> oe 2 ss
SSS
2 —

TG <
ae e = x — =S—
Z Wy —
6 § 7 4 : +3S 2 eo = ae nt
——" a—- :
|

mei - ner Klaus - ner - zel - e ver- schlun - ge - nes Ge - biisch.

——_ fz mf
ee
Se =a ———— = ss !
= z= a
| ef
wehn dieschwan-ken Fich - ten und stéh - nen Seuf - zer — laut.
——— Ce mf

= bin ich frei und na -

=oe her = — der
=
Ein - falt und Na - tur.

==. S ————— ————-= zi

a £ fe =
7s

4.26 “Die Einsiedelei” (D. 337).

 Chords on bII, on III, and from the parallel key 115

m. 1 Den 5 6 TEs y 10 tik 1)

5 5 5
a ae

| == ee
= Pe
e = 1%
z
2. oe
——
eee
fer

g e o—te 2
ar ee eee i SS |
P 8 7b
G Minor: I~ ‘ Il; ea lve Vi

4.27 Analysis of “Die Einsiedelei” (D. 337), mm. 1-12.

Soon thereafter Bb’s upper-third chord, Db major, emerges; and again we

are uncertain whether its role will be embellishing or harmonic — or cir-
cular, given that a pattern ascending in minor thirds has emerged. In this
case the upper-third chord dissipates: Bb returns (its quality now major, its
diatonic state in G Minor) at the upbeat to measure 9, and from that point
the progression presses forward inexorably towards the dominant.!®
The opening musical content is reprised beginning at the upbeat to
measure 13. Unlike the earlier passage, tonic — eventually inflected to G
Major — remains the focus for the remainder of the stanza. Though the
second tenor line descends by step from 5 to | in measures 15 and 16, the
first tenor remains on 5 at that cadence, thereby preventing full closure.
Schubert proceeds with another version of the final phrase, one that is sig-
nificantly more cadential in its effect, employing the conventional formula
m. 20 21

Sear he oa
GMaors 1 i “Vie 1

Measures 17 through 19 offer a fresh reworking of a notion that has been

invoked several times already. In measures 2 and 14, C# intrudes upon a
tonic arpeggiation (G<B}<D). Though bold and memorable, this C# is but
a melodic embellishment, an upward-resolving appoggiatura. Its presence
hints at the impending antipodal D} chord without stating it. In Schubert's
later reworking, the direction of the initiating motion is reversed: G<Bb<C#
is replaced by G>Eb>C#. Note that G, Eb, and C# are three members of a
familiar chord - one that Schubert has employed as recently as measure 11.
In both cases the altered II (presented with an augmented sixth simultane-
ity in measure 11 and with a melodic diminished third in measures 17 and
116 Harmony in Schubert

18) heralds the dominant (measures 12 and 19). This dramatic moment
is emphasized both dynamically and through metrical expansion. The
C#<D second, to which Schubert devotes two beats in measures 2 and 14,
becomes a two-measure event in measures 18 and 19.

“Am Flusse” (D. 160)

The two keyboard measures at the outset of Schubert's first of two settings
of Goethe’s “Am Flusse” introduce not only the lied’s key and mood, but also
some of its principal structural devices [4.28]. The tonic chord’s three pitch
classes are embedded within the descending melody (F>D>A), establish-
ing a distinctive profile that the vocal line immediately begins to traverse
in reverse [4.29]. Given the prominence of soprano F in measure 1, the
soprano A<D fourth in measures 2 and 3 and the reiteration of that D in
measures 5 and 8 come across as interior structural elements, an interpre-
tation that is confirmed when F is attained in measure 10, concluding the
opening tonic prolongation.
The B}>A half step of measures 1 and 2 likewise returns in the vocal
melody. The complete neighbor figure A<Bb>A, expanded into a double
neighbor through the incorporation of G# before the return of A, transpires
within the tonic prolongation of measures 3 through 5, while A<B>>A
(middle of measure 5, end of measure 6, beginning of measure 7) appears
in the context of I-II-V#, closely related to the I-IV-V# of the introductory
measures.
Another element of the introduction is reversed during the first stanza:
tonics } fifth is expressed as D<E<F<G<A in the bass of measures 1 and 2,
then as A>G>F>E>D in the keyboard melody of measures 3 through 8. The
descending version is displayed in 4.30. (The open treble-clef noteheads
denote the keyboard pitches that are reinforced by the vocal line.) Whether
in the context of I-IV-V# or I-II-V$-I, the stepwise filling-in of 4 is a
powerful expression of the D Minor tonality.
The first stanza’s invocation of Knabe (boy) and Madchen (girl) recalls
a lost past, when the content of the protagonist’s now-discarded lieder
coincided with the enchanting and flower-filled world around him. At first
Schubert allows that memory to sway his song’s mood: the image of the
Knabe singing with enchantment coincides with a shift from the minor
tonic to its upper-third chord, of major quality (measure 10). Yet that
moment is fleeting. The image of the Madchen amidst blossoming flowers
coincides with a restored tonic moving decisively to its minor dominant
 Chords on II, on III, and from the parallel key 17

Wehmiithig 3
= ——e es = == i
Ge it = 5 5 eas p veers sh
EN Ver-flie - sset, viel - ge-lieb-te — Lie-der,zum Mee - re der_Ver-ges - sen -

Gee hs SP $33
e999 “Sea tas
P=]
|a |p

S55 Sa
=— Ss ———————————— NES Se Sa \- —— : =

heit! kein Kna - be sing'ent-ziickteuch wie - der, kein Mid - chen in der Blt - then -

wa
Ne —s
S42 |
ele ile}

+n ooh rye+ Wi bE2g r baarze ae

haar we
SS fF + ren
| ve
Ex
= i t —
7 - z + ~, La 7 - rs 2 a rs

SS = Mi f = te a Ny
Za (a a ea
SSS
(er n

x a BAe a
|98 tnur von mei-ner
wt] Lie - ben; nun spricht sie mei - ner Treu-e Hohn. Ihr
~
= == | eso Sr
cy “et ST ae a aes a fare is f

i Wd

f
3 é3

SSS
| SSS7 = x
7 Sas t
- oa a
—
7 =7
wart ins Was. - serein-ge- schrei - ben: so auch mit ihm
fliesst__denn da - von; ihr

SSeS
a
na
ee
cs
ae y a
w=
{ o r

=SH
Se ao
a = =
eo
: == =

wart ins Was - ser ein - ge - schrie-ben, so fliesst denn auch— mit
5 z a

oe cy ieSerif iae morSet Jkfer

él @ oie elie

cull

4.28 “Am Flusse” (D. 160).

[4.31]. The contrast between these images and the course of the music
conveys the desolation that has replaced the robust life-spirit of which the
lieder are a token, necessitating their destruction. The text explicitly negates
these fond images: kein Knabe (no boy), kein Madchen (no girl).
The first stanzas eleven measures conform to a traditional antecedent
118 Harmony in Schubert

m. 1 2 7 3 10

SSS
eto Ts 7 : - *

—SS=
D Minor: J-— IV V# I

4.29 Analysis of “Am Flusse” (D. 160), mm. 1-10.

m 3 6 i 8

Oh to
= iW
ae. 2 = =

DMinor: [| [I V4-3

oe
s—7
D Minor: I TGs eV:

4.31 Analysis of “Am Flusse” (D. 160), mm. 10-13.

construction: an engaging prolongation of the tonic (measures 3 through

10) followed by a strong motion to the dominant, here via a dominant-
emulating supertonic (measures 11 through 13). Its consequent counterpart
is also eleven measures in length, but occurs only after an expansion and
transformation of the dominant. Thus the potential binary construction
m. 3 13 18 27 28

A1 A2
 Chords on $II, on II, and from the parallel key 119

anchors the ternary construction

m 3 13 17 18 PLY 28

I Vi , I Vi |
A, B A,
The first of the poem’s two stanzas coincides with the A, section, while the B
and A, sections each are assigned half of the second stanza, with a repetition
of the closing lines creating the textual bulk that allows the A, section to match
the A, section in length. This division of the poetic content ideally suits the
form, for only the first half of the second stanza refers directly to the source of
woe, the mocking sweetheart. The outer A sections focus on the consequence
of that mocking: the destruction of vestiges from that happier period.
The minor dominant of measure 13 is especially susceptible to two types
of transformation: the conversion to major quality, and the addition of a
dissonant minor seventh. These transformations both occur during the B
section. Schubert's attainment of these goals is stated in a most pronounced
manner at the fateful word Hohn (mockery, disdain) in measure 17, under-
scored through four musical devices: the vocal G is the high point of the
song's melody, an incidental dissonance (Bb descending to A) hovers in the
bass, the moment concludes several beats of marked dynamic emphasis,
and a fermata sets the moment off from the normal metrical flow. How
these measures might have evolved in Schubert's imagination is suggested
in 4.32. Model 1 shows the essence. In Model 2 the dominant is tonicized:
whereas D-A-D guides the tonal structure of the work as a whole, A-E-A
controls that of the B section. The chordal content is further extended in
Model 3, where F serves as an upper neighbor to pi supported by IV in A
Minor. Finally, in Model 4 the I>IV motion is segmented into two descend-
ing thirds in the bass (A>F>D), while soprano F is extended through Vi,
whose root E is suppressed as the cascade of descending thirds continues
in the bass. F’s maintenance during the E chord allows for a further upward
thrust to G at the return of the A chord.”
Thus far Schubert’s large-scale structure is conventional: 3 descends to
2 in coordination with the progression from I to V [4.31]; then the fourth
scale degree (the dominant’s seventh) rises above 2, preparing the return of
3 as its upper neighbor (4.32, Model 4].”! Yet the remainder of the song does
not fully realize the structural plan that it appears Schubert has been pursu-
ing, for 3 never sounds again in the upper vocal register. In the A, section
F (3) emerged in the vocal line only after eight measures of tonic prolonga-
tion [4.29, measures 2/3 through 10]. In the A, section Schubert turns his
back on this convention. Thus in addition to the means already mentioned
120 Harmony in Schubert

Model 2 Model 3
Model 1

py pS a =
sce a ge. keel
i
DsMinora aD Minor:Vi—( J=i D Minor: V:— ( =

© AMinor:1 V{] I?) € AMinor:I IV Vi I)

Model4

2 N
ae aa

es EE
-e

e ee ee ee

ge —— S|
er
D Minor: V }— ( rag
3 9
(AMinor:
1° ° IV Vi By)

4.32 Analysis of “Am Flusse” (D. 160), mm. 13-17.

by which the Vj chord of measure 17 is emphasized, another potent one

must be added: its dissonant seventh is never resolved in the stated register.
Though the water engulfs the discarded lieder, memories of the ill-fated
love will never totally subside. Structurally, the interrupted 2 finds closure
in the A, section's D (= is without the conventional reiteration of 3-2-1.
The potent and much-emphasized dissonant G of measure 17 resolves only
in a lower register, on the D tonic’s internal F in measure 18.
The concluding A, section begins, as did A,, with a tonic-prolonging
harmonic progression coordinated with a descending melodic progression
(A to D) in the keyboard melody. The model of 4.30 still holds, so long
as we incorporate one crucial alteration: the final chord’s bass is B}, not
D (measure 23). In an important sense the A, section is a synthesis of aN
and B, for now in the tonic key Schubert pursues the harmonic progres-
sion of the tonicized dominant key. Model 4 of 4.32 shows the bass motion
A-F-D-(E)-A. Now in measures 18 through 28 Schubert restates that
progression in the tonic key: D-Bb-G-A-D [4.33]. Lacking the B section’s
 Chords on bII, on III, and from the parallel key 121

m 18 23 26 27 28

qj
igie
; Se ie
it
2 a o 4] cA) 2

Diino eve Vee eT

4.33 Analysis of “Am Flusse” (D. 160), mm. 18-28.

conversion from minor to major and from consonance to dissonance, the

A, section wholeheartedly confirms the minor tonic - and thus also the
persona’ unhappy fate. The vocal melody again traverses - again imper-
fectly, though now differently — the descending melodic span from A to D,
while the keyboard melody starting in measure 24 is restricted to the tonic
pitch and its upper and lower half steps (D<E}>D>C#<D, reminiscent of
A<Bb>A>G#<A in the vocal line of measures 3 through 5 and taken up in
the postlude that follows).
Chromaticism enhances the lied’s impact. In measures 6 and 21, a chro-
matic passing motion connects E and D (the latter being itself a passing
note within an E>C# third). Though for a moment the pitches of bII sound
together, that label would not be appropriate in this context. In measure
26 a chromatic neighboring motion embellishes the G chord’s fifth, D. The
pitches of bII emerge again, now as the 6 phase of [V>-. Does 2 assert itself
as a root? Or is the harmonic connection better represented as IV-V# (thus
matching the IV-V# succession in A Minor during measure 16)? Perhaps
these questions have no definitive answer. Yet their very asking loosens the
hold of traditional — literalist - analytical practice, with its inherent ten-
dency to impose the label bII indiscriminately. By toppling that tradition,
alternative readings that may penetrate more deeply into the essence of
harmonic practice might result.

“Erster Verlust” (D. 226)

Three chords play precisely etched roles within Schubert's “ Erster Verlust,’
the setting of a poem by Goethe [4.34]. The title announces a “first loss,”
which has followed upon the “first love” mentioned in the poem's second
line. I propose that the state of Loss corresponds to tonic F minor, while the
radiant mediant, Ab major, represents the sweet time of Love. These chords
interrelate during measures 1 through 9 and 17 through 21, in which the
122 Harmony in Schubert

Sehr langsam, wehmiithig

o a 5 aa
ee eee = ——
Pe a

“, = ; iaF i sché - inn

men. =
Ta-ge, = je-ne r =
eee me = be,
Li ach,
+ < bringt nur
Ach, wer bringt die
—— I> | —_

SS ees
oz +
e | — —=
PP ip ea 2 ee ile he

prgr= aes ae
e (Os On| bce AL = =

g - oo s=n
ae a 253 es ——~ — a a z z Si5 5 o

i ei - ne Stun-de je - ner hol ~ den Zeit zu- riick! Ein = sam nahr' - ich mei - ne
eS Fae
= ——— e o—-# Bs
(gatas 5 Bao2 see eee ee ee Sees za Ss
e Copp __¥
VeSC. aS ee

BE
: aoe
a =
—— t
=e
t
—— 65 @ z
“352 ,
=.tie===

Fete er Sooo eae ee

12 .
in ~

: Wun - de, und mit stets he neUm ter Kla - ge traur' ich um's ver lor’ - ne Gliick.

2s —s = ——T =
mf FI Fa
p oa ==see a ttt $ = -
e a Ee Se et

17 me
> <—> Fae —— = =
> r= r a rd r

ces ia = == = = = s t = a 2
Ach, wer bringt die sch6 - nen. Ta - ge, wer je-ne hol-de Zeit. zu- riick!

' sates ast > A z sy — =

e Es . =e Sars S

po peeb iz . o Lox He ‘ = He & ==s z

[SZ

4.34 “Erster Verlust” (D. 226).

persona dreams of a resumption of Love. Thus I<III corresponds to Hope.”

Between those sections the dominant, C Major, which directs the progres-
sion back to F Minor, imposes a return to the harsh reality of the situation.
The persona accepts and laments lost happiness. Thus V4 corresponds to
Negation. After a second expression of Hope in measures 17 through 21,
the final measure, for keyboard alone, imposes a second Negation, more
objective because external to the persona’s own consciousness, and contin-
uing directly to a close on tonic, a third and irrevocable statement of Loss.
The song thus has four phases, arranged within a ternary construction:
 Chords on SII, on III, and from the parallel key 123

IN
am BERiy
oe —— 4
Se bt SS == |

F Minor: P ° ;
Ab Major: IV" ° v3=3

4.35 Analysis of “Erster Verlust” (D. 226), mm. 1-5.

Persona: Hope - Negation, Hope

Keyboard: Hope - Negation, Hope - Negation
I<Ill - Vi; Tobe «Vas
aS B AS

A: From [ to III

In a minor key the subtonic chord often heralds the mediant. It might even
be interpreted as V in a tonicized mediant key. On the other hand its pitches
may serve as the onset of V-space. (Compare with 1.8, Model 3.) Measures
2 and 3 of “Erster Verlust” reveal how closely allied the subtonic and
dominant chords can be: E}-G-Bb-D) loses its potential tendency towards
mediant Ab with a single half-step shift, to Es-G-Bb-Db, the diminished
seventh variant of V/.*° When a chord on 47 occurs, listeners do not know
whether a continuation to V —- Negation, confirming Loss - or to III - Love
~ will be pursued. Thus Schubert's dramatic (because unexpected so early in
the lied) and dynamically emphasized Eb’ chord in measure 2 is intriguing.
For now Hope is held in check: the subtonic dissolves into the dominant,
resulting in a three-measure expansion of the opening tonic (Loss).
A more assertive thrust towards the mediant follows immediately. The
D}<Eb bass motion of measures 4 and 5, with the Db chord extended via a
5-6 shift and the E} chord embellished by a $-3, distinctively suggests a IV<V
harmonic succession, seemingly initiating a tonicization of the mediant. (See
4.35.) Once again an attempt is made to resist that Hope, for measure 6 begins
with the very Ey-G-B}-D} chord that thwarted the Ab arrival in measure 3.
This time that strategy is ineffectual, for a contrasting continuation tran-
spires: against common tone D} in the melody, the remaining pitches of the
diminished seventh chord resolve into the third and fifth of a D} chord,”*
124 Harmony in Schubert

a= ———
pb: o——ee
46
a5
Ab Major: IV” ‘ Vi-3

IN 4 3

_——
a

See |

[Dee —— ——
V
Ab Major: IV>~ ‘ leet Vv! I

4.36 Analysis of “Erster Verlust” (D. 226), mm. 4-5 and 6-9.

triggering a reiteration (in expanded form) of the Db<Eb succession first

presented in measures 4 and 5.”° (See 4.36.) This time goal Ab is achieved, in
measure 9. Note in 4.36 the contrasting chordal support for passing note E)
in measures 4 and 8. In the earlier instance, E) and A concurrently perform
passing roles — one diatonic, one chromatic. The more elaborate formulation
of measures 6 through 9 provides consonant support for passing Eb and the
passing C below it, via IV’s upper-fifth chord (Ab). (Compare with the similar
upper-fifth chords explored on p. 25.) In fact, the [V°~° expansion here is very
similar to a passage from “Das Geheimniss” presented in 1.21 (measures 14
and 15). (“Erster Verlust” and “Das Geheimniss” were written at about the
same time, and both passages are in Ab Major.) The first half of measure 8 in
“Erster Verlust” and measure 15 in “Das Geheimniss” are virtually identical.
In “Erster Verlust,’ the passing chord is prolonged for two additional beats
before proceeding to the 6 phase of IV.

B: The prolongation of V4
The poetic text of measures 10 through 16 starkly contrasts the more
positive sentiments of the preceding measures. An honest acknowledg-
ment of Loss and of the consequent emotional wound conveys the actual
state of affairs, with Negation (C Major) replacing Hope (the motion to Ab
 Chords on 41, on IIL, and from the parallel key 125

m 10 11

eee aeee
FS,

Evie) V¥3—_———

4.37 Analysis of “Erster Verlust” (D. 226), mm. 10-11.

Model 1

ies
st Sa ee eee
Bua (Ol

F Minor: Vi d ° Z

Model 2

2;me
iG= be 2 Z be

m 10 12 13 14 15 16

Dy Il
be (2)
as Fe: 7 eT = =
Fes
- @ Hg ato Z Z 5?
| 5] 5]
, he ae: a g =
é ZZ
Ebb Se ZZ

4.38 Analysis of “Erster Verlust” (D. 226), mm. 10-16.

Major). The C Major dominant in measures 10 and 11 is heavily laden with

“sighing” descending seconds [4.37].°° The progression that connects the
C chords of measures 1] and 16 is derived from a conventional dominant-
prolonging construct [4.38, Model 1]. As the : fourth is traversed in the
upper voice, a neighboring motion arises in the inner voice. Though an a
diminished fifth often resolves to an 4? minor third, in this context both F
and Ab help extend the C major dominant chord. Schubert confirms the
non-tonic nature of that moment by activating the pitch C, with motion
126 Harmony in Schubert

phe == =
6 = Ls 7 EL Z

[Dee fe 2 ; &
; en
y Le
a)
(=|F Bb _Eb_ Ab)
7
F Minor: I ( ) vi I

4.39 Analysis of “Erster Verlust” (D. 226), mm. 17-22.

both upwards and downwards by a half step (Model 2).”” The resulting Bk-
D}-F-Ab is an evolved G chord. Thus C-G-C prolongs V+ in these meas-
ures in much the same way that F-C-F prolongs I in measures 1 through
3. Model 3 shows how the dissonant soprano pitch B} is presented first as a
consonance (measure 13) before the C chord is restored (measure 14). To
attain the Bb chord, Schubert deploys a short segment of the descending
circle of fifths (C->F—>Bb). Perhaps the threat of parallel fifths prevented
him from sounding the model's G for the concluding vocal pitch (measure
16), though that pitch is presented by the keyboard.** The C that sounds
instead is by far the lowest vocal pitch in the composition. It supports the
potent word Gliick (happiness), a happiness that is verlor’ne (lost) and thus
set low and isolated, concluding the B section.

A,: Condensed trajectory and closure

Despite the expansive Negation, Hope returns with the reprise of the initial
tonic material, where the content of measures 1 through 9 is condensed
into five measures. (See 4.39, measures 17 through 21.) The persona seems
to want to let things rest there, leaving the window of Hope open. The key-
board steps in, however, asserting a Negation that is gentle, concise, and
definitive (measure 22). The last chord, a mere whisper, brings the matter -
and the lied - to a close.

Landler in G# Minor (D. 790/6)

Among Schubert's Zwélf Landler (D. 790), only three open with a root-
position tonic chord. The set’s sixth Landler, in G# Minor, instead offers a
 Chords on $II, on III, and from the parallel key 127

d ivi se = A
(aire eare s beat ‘ = ee ee
ry) P al ——

tly |4. Sie Dae. ‘LA

DER ee = : : i

“pa — ¢ dei. Mi,

: = ES if =
(pS
SSS Se
isuaeiaen kd Pana Pil daa
4 ee (aay te ee ab Nos at ae

fF Fe Ola P ;
Bee : Ss . % a — I
meres A = le i

!
aes ier ig S aie

e = o—_-—_ eo id

Fable aa a
7a i

Deets t 4 oe ike ia . z oe e = -
oa = a = . 5

4.40 Landler in G# Minor (D. 790/6).

most curious concoction of pitches in its opening measure: dissonant yet

not of dominant flavor (several of the other Landler open with a domi-
nant or dominant-emulating chord), and with some uncertainty regarding
exactly which pitches are chord members [4.40]. (In particular, is D$ a
suspension resolving to C#, or is C# a passing note between B and the D¢#
in measure 2?) In that the preceding Landler ends on tonic B Major, the
upbeat D#s may seem to imply a continuation of B-D#-F#; thus measure
1 could be interpreted as B Major's II or IV chord. Or, given that the per-
former is instructed to play the opening eight measures of the Landler
twice, during the repeat the D#s might seem to represent the D#-Fx-A# of
measure 8, restored in measure 2 after a neighboring chord in measure 1.
Or, given that the Landler turns out to be in G# Minor, D# might represent
the fifth of tonic G#-B-D#. What a conundrum!
128 Harmony in Schubert

Nineteenth-century musicians were attuned to the notion of multiple

meaning (Mehrdeutigkeit), acknowledging that a pitch (such as D#) may
function as a chordal root (of D#-Fx-A#), third (of B-D#-F#), or fifth (of
Gt-B-D#).2° One might have good cause to interpret the Landler’s opening
upbeat in each of these ways. Yet a performer might instead assess the early
measures using data from later measures and play the work with a particu-
lar conviction in mind. For example, consider measure 2’s Fx. If the domi-
nant of measure 8 is to be prolonged through the repeat of measure 2, then
measure 1’s G4 would need to be projected as an upper neighbor. (Likewise
for E and B.) On the other hand, with reference to the key of G# Minor
that indisputably emerges as the Landler’s tonal center, the melodic G#>D#
fourth (eleventh) of measures 1 through 8 is a defining interval, suggesting
that Fx should be interpreted as subordinate to G#. A performer can deploy
dynamics, rubato, and articulation in ways that would promote one or the
other of these conceptions. But can one have it both ways at the same time?
Despite inklings of other (mutually contradictory) chordal alliances, I
regard the tonal function of measure 1 as the traditional one: tonic in the
key of the work. In measure 3, where a dominant-emulating I chord (G#-B#-
D#-F#-Ak) leads to the following measure’s IV chord, D#>B¢ is an arpeggia-
tion from the chordal fifth to the third. The clarity of that measure helps
one to interpret measure 1: D# and B serve as chord members, while C# is a
passing note. Schubert's daring opening depends in part on an expectation
that listeners will recognize the strategy in which tonic gradually assumes
dominant-emulating characteristics en route to IV [4.41, Model 1]. In this
case the 8-7 motion is filled in chromatically: GE>F# becomes G#>Fx>F#,
a context in which the hierarchy is clear and fixed (Model 2). In a potent
deformation of this construction the tonic root is withheld from the bass
until F# arrives, while concurrently E, the upper neighbor of tonic’s fifth
D#, sounds with the opening chord (Model 3). The progression is mystify-
ing because it begins after its normative starting point. Instead of a staid
stating of tonic followed by embellishment, a neighboring note has impetu-
ously imposed itself upon tonic even at the outset. The gist of the musical
structure can be discerned only in retrospect.*° During the listening process
various hypotheses regarding the function of measure 1 likely will have
emerged, with one of them ultimately fitting much better into the broader
scheme than the others.
The content of the opening eight measures - already heard twice as a
result of the repeat sign — returns in measures 9 through 16 in an enhanced
and modified presentation. Note that now the G#>Fx>F# melodic motion
sounds in tandem with B>A#>A4 a third higher (measures 9 through 11).*!
 Chords on $II, on III, and from the parallel key 129

Model 1 Model2
m 1 3 4 m. 1 2 3 4
J CP
a
a
| 4 Pie
Gite 2 zo? [aaa a8 ey
= oo = N =
oO (O) O

abe eee : ie 7
G# Minor: I;—: IV G# Minor: I: mI

Model3
m 1 2 3 4
ss
tty a a
3 ss a
21) aaa, = =
(3 NJ ow

Se
Sevier, |= >=—,

4.41 Analysis of Landler in G# Minor (D. 790/6), mm. 1-4.

Two features of measure 9 confirm my interpretation of measure 1: first, the

fuller chord, with the third ?* held in the right hand for the entire measure,
confirms beyond doubt that C# is a passing note; and second, the bass more
straightforwardly expresses tonic (in first inversion), now without neighbor
E
The two phrases — measures 1-8 and 9-16 — begin and end similarly, yet
their internal content is not the same. Both involve a descending soprano
line (a descending fourth — expanded into an eleventh - from | in the first
phrase; a descending sixth from 3 in the second). Since the latter melodic
descent has more terrain to cover, the second phrase’s path to IV is more
extended. The structural models of 4.42 show how in the first phrase IV
arrives at measure 4 and, supporting soprano E (which is transferred down
an octave), continues through measure 7; whereas in the second phrase the
parallel spot (measure 12) is treated not as IV but instead as an internal
component within a chromaticized descending circle of fifths that leads
to III (an extension of I-space), at which point the soprano descent has
reached F#. IV again supports E, from which a local E>D#>C# descent
emanates before the broader line’s culminating D# arrives. (The cadential
melodic E>D# in both phrases reprises the bass E>D# of measures 1 and
2. It actually appears in the bass in the latter phrase.) Though one might be
tempted to hear a return of tonic at measure 6, that G# chord occurs in the
midst of a process involving an octave connection between two Es, and thus
130 Harmony in Schubert

Model 1: mm. 1-8

m. 1 3 tl 5 6 7 8

a ; a
Q —— Gel ae
cet SEs S

ee pega ee CP
53 es Vi
G# Minor: I; —+ IV—(

Model 2: mm. 9-16

m. 9 ll ye ale} 14 15 16

Gt Minor: I ( \1 Vs —— Vi

4.42 Analysis of Landler in G# Minor (D. 790/6), mm. 1-8 and 9-16.

Model 1 Model 2
m. 4 a 8 m. 4 6 a 8

ok hek |

beg bile = aA

Gt Minor: ve V* G# Minor: TV8-6 )77) v«

4.43 Analysis of Landler in G# Minor (D. 790/6), mm. 4-8.

it is hardly suitable for such an interpretation. As 4.43 shows, the G# chord

serves as the subdominant’s upper-fifth chord, a means of introducing the
subdominant’s seventh B in a consonance context, after which the return of
bass C# activates B’s dissonant effect. (Compare with 1.24.)
After sixteen eventful measures — two-thirds of the entire Landler — the
opening tonic prolongation is still in progress. The soprano G# and B
initiation points of descending motions (measures 1 and 9) belong to a
deeper initiative of ascending arpeggiation that continues to the high D#
of measure 21 [4.44]. As often happens in such situations, the bass motion
towards the dominant root begins before the soprano tonic arpeggiation
 Chords on bII, on III, and from the parallel key 131

(= I—V«, I— Ve)
G# Minor: _[_——————- ( ie 0 ve I

4.44 Analysis of Landler in G# Minor (D. 790/6).

has concluded. Thus D# (= 5) sounds not with root Gf, its generator,
but with mediant B, which here serves to extend I-space. The transition to
the mediant in measures 17 through 20 refocuses the local dominant in a
strategy similar to that shown in 4.8. Here G#’s dominant D# (leading tone
Fx in measure 17) glides up a third to B’s dominant F# (leading tone A# in
measure 19).
With the soprano arpeggiation’s goal D# finally achieved in measure 21
and the bass embarked upon its upward course to the dominant root, the
composition continues without a hitch. The mediant is extended by a 5-6
motion before the bass proceeds to C# (harmonized as II in first inver-
sion, supporting 4). The following Vx and I achieve harmonic closure in
conjunction with the soprano descent to it
|
;

== j “Testht

x (ied ional

t) j Sbulinti
oa
inte
re A tte auive

BOTT fil innileanr :

le doth a walinals VES

bile (9 Lease nis

ih coumyead ata
4 rete: sad air
‘ a iis Netiagge
on bi uid -yyeited rest
let. sagt] RD BO Fis
rUnwreritte? snl ttae

}j ,
- ra Ay

G anhe
:
= } ' = «<a
i
& — + ot (is

gil ie pet eryt

Mi
1 aale sappy
PART TWO

Masterpieces
 OWT THAD
~

steyistesM

= 2)

S
i
= \ Pe
=>

.'
—
Lite

- =

i ' f 7

y/ i]
~
ied

< be
> 7
=a ~
= —
— >> =
AC eee x. = .
Re ge — )
- —
my - 2 a
— - -

- * ae
a
——
5 “Ganymed” (D. 544)
with Lawrence Kramer and Suzannah Clark

Occasionally Schubert will begin and end a lied in different keys.

“Ganymed” is one of his most celebrated works of this type. Such struc-
tures of course pose unique analytical challenges. Fortunately in this case
the poem that Schubert sets, Goethe's “Ganymed,” offers a compelling
rationale for the shift, in that Ganymede is lifted out of his earthly exist-
ence and carried off to Olympus - like Kansas to the Land of Oz, with no
return. The subject is distinctive also in that it was a rare context in which
the subject of homosexuality could be broached openly in Schubert’s time.
Neither Goethe nor Schubert flinches from depicting the ecstatic bond
between Zeus and Ganymede. A careful assessment of the lied has attained
new urgency in recent decades, as musicians and the musical public
grapple with the suggestion that Schubert’s own inclinations may have
been towards men. Given that the weakened health of his final years was
a direct outcome of his sexual exploits and likely influenced his choice of
projects and the urgency with which he completed them, Schubert's private
life has become a matter of vital interest. Alas, there is very little that one
can pin down, and the clues that have emerged have been interpreted in
contrasting ways.' Might we tease out some insights on this matter from
his music?? “Ganymed” - a work with sexual overtones, written when the
composer was still in robust health (1817) - might be an especially good
place to look.
Lawrence Kramer has twice addressed issues in “Ganymed,’ first in
a general assessment of Schubert's lied composition, and later (after
Schubert’s sexuality became a topic of mainstream interest) in a study
devoted to several works by Schubert with homosexual associations.’ In
keeping with the topic of my book, I will focus especially on Kramer's
detailed harmonic analysis in my comments below. I am not, however, the
first person to respond to Kramer. In an extended review article, Suzannah
Clark not only evaluates Kramer’s perspective on “Ganymed,’ but also
offers an extended analysis of her own.* I will proceed by first presenting
my reading of the entire lied, without reference to these other perspectives.
Then I will examine each section of the lied again, commenting first on
Kramer’s and then on Clark's reading.
Wo Harmony in Schubert

My reading of “Ganymed”

In “Ganymed” Schubert erects an extraordinary compositional structure to

serve as musical framework for Goethe's extraordinary poem. The Trojan
youth Ganymede led a charmed earthly life. When we first encounter him
in the lied he is a shepherd, unaware of the fate that soon will befall him — to
be snatched up by a smitten Zeus (in the guise of an eagle) and transported
to Olympus, where he will serve as royal cupbearer and attain immortality.
Goethe presents Ganymede in the throes of enthusiasm for the springtime
splendor that surrounds him, a landscape notably unpopulated by other
humans. The grass, flowers, breeze, and nightingale are Ganymede'’s loving
companions.
The lied’s broad tonal trajectory before the fateful swoop (with liftoff
at the forte “ach!” in measure 71) is an obstinate circular progression of
descending perfect fifths, shown in 5.1. (Further details will be provided
below.) Three major chords are offered as potential tonics: Ab major
(measure 1), G> major (measure 46, followed by a keyboard interlude),
and F> major (notated as E major in measure 56, also followed by a key-
board interlude). Once F> arrives, it becomes clear that Gb is internal to
the descending sweep, rather than the goal of a harmonic progression. The
relationship between A> and F> is never adequately clarified, because the
tonal structure in which these chords participate is, like Ganymede’s earthly
life, cut short.° Schubert might have intended this downward trajectory as
a musical evocation of the poem's literally “down to earth” imagery: “Ah, on
your breast | I lie, pining, And your blossoms, your grass |Press upon my
heart.” (“Ach, an deinem Busen | Lieg ich, schmachte, |Und deine Blumen,
dein Gras |Drangen sich an mein Herz.’)
The half-step motion from C> to D5} [B to C] in measures 70 and 71
signals that the prevailing tonal structure (involving Ab and F) [E]) is being
abandoned, as another (involving a new and unrelated tonic, G}} [F]) is
embraced. This tonal shift corresponds to Ganymede’s transference from
the earthly to the heavenly realm. Schubert’s enharmonic spellings of con-
venience mask the true state of affairs: the half-step motion in fact connects
F> Major (signature of eight flats) and Gbb Major (signature of thirteen flats)
via their respective dominants.’ The latter key, though hopelessly unwieldy
in its proper spelling, is ideally suited to convey a domain far from any
earthly habitat. Schubert's strategy for its attainment is straightforward yet
audacious. During the initial A> region, dominant Eb is prolonged in meas-
ures 20 through 27. A chord plays a key role in this prolongation. (See 5.2,
Model 1.) The prolongation of F5’s dominant in measures 68 through 70
 “Ganymed” 137

m 1-5 44 45 46 a Be

Ee ey ee eT
S 8 5) 8 5

Bae Z : PE =
oS am oO

(Ns es Gh Cs

5.1 Analysis of “Ganymed” (D. 544), mm. 1-56.

Model 1 Model2
mae 20-24: 250 27 m. 68 7 71 74

ee
SSS Pap ay
Saas
, Saama! : d fl
Ab Major: V2 soe 3 Fb Major: V3 = 6
Set)
Gb> Major: V

5.2 Analysis of “Ganymed” (D. 544), mm. 20-27 and 68-74.

(Model 2) similarly calls upon a °” chord. In this case, however, something

extraordinary occurs: whereas we assume that Fb and Abb [E and G] in
measure 70 are unstable pitches that will resolve downward, as did A} and
C} in Model 1, they instead persist while the root ascends a half step to Db}
[C]. In this way two unrelated keys — Fh Major and Gbb Major - are boldly
juxtaposed.®
Now let’s consider the Ab-to-G} segment of 5.1 in greater detail. Model
1 of 5.3 shows how a chromatic variant of Ab’s upper-third chord (which
includes a wobble from C to C>) comes between the consonant Ab-C-E}
chord and its potent dominant-emulating variant, C-Eb-G}-Bb}. Schubert
spells Bb) as Ay in measure 44 because this downward-tending pitch
ascends to Bb (a member of an embellishing ¢ chord) in measure 45. The
model displays the more basic downward resolution. Model 2 shows
additional chordal activity that supports a descending stepwise line in
the soprano. Note in particular that both the initial Ab chord and the C)
upper-third chord are extended via their upper-fifth chords, and that the
path from A} to C} is a circular progression (Ab-Db-G}-C)). If the Gb in
measure 46 were the lied’s tonic, the broad trajectory of this model could
be interpreted as a harmonic progression: I+V—l. The perspective of 5.1
reveals that the chordal progression instead is a circular construction. Thus
138 Harmony in Schubert

44 45 46

Model2
m. 1-18 20-27 28-29 30 31-40 43 44 45 46

5.3 Analysis of “Ganymed” (D. 544), mm. 1-46.

harmonic labels are not employed, in keeping with the analytical prac-
tice introduced in chapter 2. (Under normal circumstances the initial Ab
chord might be labeled as I. Yet, as indicated above, the lied’s tonal focus is
sufficiently elusive as to make such a decisive label inappropriate.)
Note the similar structures during the keyboard introduction and the
initial vocal lines: a descending fourth from 5 coordinating with har-
monic activity that leads to the dominant - a local dominant that works
to strengthen the preceding tonic. (See 5.4. Though I am hesitant to label
Ab as the lied’s tonic, clearly the opening region tonicizes A} - even if it
should later turn out to be interpreted as F)’s upper-third chord.) The local
harmonic progressions are contrasting: in the introduction I*” leads to
IV, whereas in the first stanza I°-° leads to II (over dominant pedal).? My
reading consistently demotes the local upper-fifth chords (in measures 2, 4,
9, 11, 13, and 15) to embellishing status, not as participants in the harmonic
progression. Even so, there is some interaction between their functioning
and the prevailing harmonic progression: Dy against Ab and C in measure
4 blossoms into a II;chord in measure 19.
The most tentative moment in this conception is the V chord of measure
7, whose pitches appear within parentheses. Schubert has truncated what
we might assume will be an eight-measure phrase into seven measures.
 “Ganymed” ISS

8-10
m. hes eet} 6 W 12-14 16 17 18 19 20-27

Ab Major:ie
1°” meee ve et = ey Le ee
(= VIZ)

5.4 Analysis of “Ganymed” (D. 544), mm. 1-27.

Beginning at measure 5, several straightforward unfoldings occur in the

melody: Ab<C<E} | Db>B) | C>Ab. We may expect that either G<Bb or
Bb>G (or perhaps Bb>Eb), supported by V, will occur during an eighth
measure. Instead a rather nondescript inverted V is squeezed into the
final beat of measure 7. My model [5.4] projects the norm that Schubert
has evaded.'° The expansive, tonicized V of measures 20 through 27
compensated for this neglect.
The Gb [F#] chord in measure 46 comes across as Ab’s equal, and
potentially its successor: it is likewise of major quality, with fifth residing
in the soprano [5.1], restored after the descent to inner-voice Gb [5.3,
Model 2]. Schubert celebrates its attainment and depicts the contentment
in Ganymede’s heart (“Herz”) during the keyboard interlude of measures
46 through 49, in which a I'0-?-IV-V?-I progression transpires over a
G} pedal. The Ch-E}-G}-Bbb [B#-D#-F#-A] chord of measure 46, which
embellishes the Gb chord (as a “common-tone diminished seventh”
chord), is transformed during what initially seems to be a repeat of this
phrase into a chord that functions as ch?” [B”], thereby directing the
deep circular progression along on its route towards Fb [E], as shown in
the last two chords of 5.1. Fb likewise is of major quality, with fifth resid-
ing in the soprano, and thus also may seem like the arrival of the lied’s
tonic. The Fb chord’s C)>Fb [B>E] fifth is traversed (its pitches presented
sometimes by the keyboard, sometimes by the singer, and sometimes
by both) during a prolongation in measure 50 through 56 [5.5]. A key-
board interlude, depicting an outdoor landscape with morning breeze
(“Morgenwind”) and nightingale call (“Ruft drein die Nachtigall”),
follows the cadence on Fb. This Edenic locale is the setting for the central
140 Harmony in Schubert

ae

9 sae

Fb Major: 1° ‘ Veo ap Th gee I

5.5 Analysis of “Ganymed” (D. 544), mm. 50-56.

m 56 63 «64 68 70 71 3 75

Fb 5 Gbb: 5
N
peasy a 2
6 ee a ss trae ary
ais

bb D5 t H

oe ug 8—— 7b — 16%
F> Major: I IV Vas

Gbb Major: aie t I

5.6 Analysis of “Ganymed” (D. 544), mm. 56-75.

event of the lied: Zeus’s snatching up of the boy. In 5.6 the shift displayed
in 5.2, Model 2, is placed within its broader tonal context. Note that once
again the goal chord in measure 75 is of major quality, with fifth residing
in the soprano.
Though measures 75 through 116 constitute over one-third of the
lied, just two structural ideas, which appear in alternation, provide their
content.'! The poem oscillates between references to Ganymede’s ascent
(“Hinauf strebts,” “In eurem Schofe | Aufwarts!”) and to downward
motions of the heavens embracing Ganymede (“Es schweben die Wolken
|Abwarts, die Wolken |Neigen sich”). During the former, Schubert com-
poses passages that lead upwards sequentially, eventually attaining IV
(employing, either rigorously or loosely, an ascending 5-6 sequence, with
chromaticism and unfurlings of the 6-phase chords), while during the latter
and during the salutation “Allliebender Vater!” the lines descend, supported
 “Ganymed” 141

by a progression from IV°-* to VS (and ultimately, in measures 110 and

116, to I). These passages are distributed as follows:

Descents: mm. 79-84 mm. 92-99 mm. 106-116

Ascents: mm. 75-78 mm. 85-91 mm. 100-105

This construction suggests mutual excitement concerning the event, as

Ganymede embraces and is embraced by Zeus (“Umfangend umfangen!”).
Since the final vocal line descends, Schubert manages to close his struc-
ture with a conventional motion to 1. The 5 of measure 75 is prolonged
through various subsidiary descending fourths and thirds until 4 in
measure 113, 3 in measure 114, 2 in measure 115, and finally 1 in measure
116 bring the structure to a close. Yet the alternation principle persists
even beyond this cadence, as an ascending two-octave span transpires
during the brief coda.

Kramer’s and Clark’s readings of “Ganymed”

Measures 1-27

Kramer employs reductive strategies to develop a perspective on how

measures 1 through 27 are structured. The numerous chords of the lied’s
opening region boil down to just three chords in graph A of his example 4.5
(1998, p. 123). (A natural was inadvertently omitted beside the D notehead
in the second chord.) These chords likewise appear prominently in my 5.4,
which offers greater detail in both the harmonic and melodic domains.
Kramer’s perfunctory soprano line apparently is intended only as an indica-
tion of pitch content, with no further Schenkerian implication. One should
not equate Kramer’s methodology here to rigorous Schenkerian analysis,
despite the similarity in appearance. Only in measures 71 through 116 is
such rigor in evidence.
The E} chord in Kramer's graph (measure 20) seems to me to be visually
overemphasized. (Compare Kramer's reading with my 5.3, Model 2.) As
tonic’s upper-fifth chord it is a local event, despite its brief tonicization. Just
as measure 2 confirms measure 1 in a local fifth-relationship, measures 20
through 27 confirm measures | through 18.
Finally, the graph’s Ch notehead (corresponding to measures 25 and 26) is
curious. I am not sure why C} appears but not the inner-voice Ab that works
in tandem with it. (Compare with my 5.2, Model 1.) Though the graph
lacks a slur here, one would assume that Kramer regards C)} as an ascending
passing note connecting Bb and Db. The score, like my 5.2, confirms that
142 Harmony in Schubert

Ch instead descends to Bb. There is no Bb for C) to descend to in Kramer's

graph, suggesting a haphazard reductive process — a problem that will recur
in later passages as well.
Responding to Kramer's analysis, Clark proposes an alternative reading
of “Ganymed” in her example 4 (pp. 234-235). She presents the motion
from tonic to dominant (measures | through 20) even more sparsely than
does Kramer, omitting the IIk chord of measure 19. Her soprano is an
improvement in that it begins on Eb - in fact, the Ey>-Db—C)—B> line of my
5.2, Model 1, is fully accounted for in her graph. However, her accounting
seems to me flawed. I cannot conceive of measures 24 through 27 other
than as a prolongation of an E} chord, with internal * embellishment.’
Clark instead regards that % chord as a modally modified restoration of
the opening Ab chord, describing it as “an Ab minor, second-inversion
harmony” (p. 232) and binding both its bass Eb and soprano C) to pitches
of the opening tonic chord in her graph, thereby asserting that it is
hierarchically deeper than the intervening Eb’ chord.

Measures 28-31

Kramer employs slurs skillfully to show that a chord progression from

Db through G} to Cb occurs during measures 28 through 31. The Roman
numeral I suggests that Cb is tonicized. One might reasonably wonder,
however, where Db comes from, since there is no attempt to correlate this
group of chords with what precedes it. I would attribute the problem to
Kramer's emphasis on the dominant of measure 20. I instead interpret
the Ab-Eb trajectory of the opening twenty-seven measures as a means of
prolonging the initial Ab. It is from that Ab that D> emerges, initiating a
circle-of-fifths progression leading to Cb in measure 31 (as shown in my
5.3, Model 2).
Kramer more successfully displays this broader perspective in his graph B
(also in his example 4.5), where a beam connects Ab and C} and the Roman
numerals I and bIII appear. (Note also that the intervening bass Eb now is
presented as a stemless, filled-in notehead without a Roman numeral, a
modification that I applaud.) Kramer is making an important and in my
view correct judgment by forging an intimate bond between the opening
Ab tonic and the C) chord of measure 31, despite the chromatic shift. (My
2.20 offers a framework for this relationship.) Though I am perplexed by
the strange interface between upper- and lower-staff pitches in his example
B (for example, are we to read the bIII chord’s soprano pitch as Ab?), I do
not doubt but that one could create a more cogent graph that retains the
 “Ganymed” 143

perspective displayed in the current graph’s first four bass pitches, the first
three of which I regard as correctly analyzed. The Gb of measure 46 will be
assessed below.
Clark's interpretation of this passage does not succeed as Kramer’s does.
In his graph A, measures 28 and 29 are bounded by slurs that reveal a voice
exchange between the outer voices (??xf), allowing the reader to com-
prehend the broader Db-G}-C} root progression. Clark’s graph instead
contains a dubious dotted slur connecting the bass Gbs in measures 29 and
30. (Her figure below bass G} at measure 30 should read }7, not 7.) Unlike
Clark, Kramer understands that the first bass Gb and concurrent soprano
C (measure 29, beat 1) function as neighboring notes: the compound
augmented fourth resolves into the ?? sixth on the next beat. In addition,
whereas the soprano line in Kramer’s graph is coherent and useful, that in
Clark's graph is not. She displays no new content between Cy at the begin-
ning of measure 29 and C} in measure 31. Though one normally would read
such notation as a two-measure prolongation of Cy leading to Cb, certainly
she does not suggest that Cy persists during the Gb” chord. Clark seems
concerned only with charting the course of a Ck/C) wobble in the soprano,
at the expense of intelligible interface with the bass.

Measures 31-46

With the arrival of Gb in measure 46, Kramer’s and my views begin to

diverge sharply. In both his graphs A and B, Kramer asserts that the
principal expansion of the initial A} tonic is the motion to its chromatic
upper-third chord.'’ I instead regard the principal expansion as motion
to Ab’s chromatic lower-third chord, Fb [E] in measures 50 through 60.
(My 2.17 offers a framework for this relationship.) The upper-third Cb
is an episode within a prolongation of Ab that extends from the opening
of the lied through measure 44 [5.3]. Whereas Kramer's graphs correlate
G} with the Eb of measure 20 (both as dominants of preceding chords), I
instead correlate Gb with Ab, within a descending trajectory that will con-
tinue another step to F} [E], as displayed in 5.1. I regard the keyboard epi-
sodes after the Gb and F) arrivals as a strong confirmation that Schubert's
basic strategy is a descent, in stages, from Ab to Fb. The C-E}-Gb-Bbb[A5]
chord of measure 44 is a restoration of tonic Ab, now in a dominant-
emulating variant.
Again in this region, Clark’s reading is not an improvement upon
Kramer's. My 5.3, Model 2, proposes that the basic progression during the
first forty-six measures should be read as follows:
144 Harmony in Schubert

Ab (Cb) Ab Db Gb

The circular nature of this progression is apparent. In Kramer's reading,

though the return of Ab after Ch is omitted, both D} and G} are acknowledged.
Clark, on the other hand, jumps directly from C) to Gb: no content is dis-
played between measures 31 and 46. At what structural level is Clark intend-
ing to represent the lied? On one hand, foreground details are displayed for
measure 29. On the other hand, in the same graph measure 45’s root Db - one
of only five roots displayed in my 5.1 as the basic structure for fifty-six meas-
ures of music - is absent. Though her graph lacks Roman numerals, Clark's
commentary corrects Kramer’s reading of measure 46 as a dominant. She
asserts instead that it functions as a tonic (p. 232). Whereas I mention above
that Ab-D}-G} might be interpreted as II+V—I in Gb Major, in the broader
perspective G} is a step along the linear path between A} and F) [5.1].

Measures 46-56

New lines of the poetic text commence at the upbeat to measure 50. An
inverted E major chord sounds on the first downbeat, and the phrase
cadences on an E major chord in measure 56, after which a keyboard inter-
lude prolongs that chord for several measures. In my view, the only viable
interpretation of this passage is as a tonicization of E Major [5.5], which I
display as Fb Major for reasons explained above. The initial E chord oddly
has been omitted from Kramer’s graph. He instead jumps to the B’ chord of
measure 52, which he labels as I in the key of Cb [B] Major, apparently pro-
longed since measure 31. (If B is I, why is it preceded by A in measure 51,
rather than by A#?) Kramer further proposes that the goal E chord serves
not as I in E, but instead as V in an upcoming tonicization of the key of A,
a notion I will consider below. Why did Schubert convert to an E Major key
signature if he were not intent upon tonicizing E? Kramer’s and my analyses
have by now diverged irreconcilably.
Nevertheless, we both employ the notion of circular progression by
descending fifths. Mine extends from measure | through measure 56:

Ape = "Db = GhFRY P= "Ch [BR] Ev ED,

as displayed in 5.1. His extends from measure 46 through measure 64:

Fe sed 2b 2 an
(as described in 1998, p. 122, and displayed in his example 4.5 in 1998,
pxl2zs):
 “Ganymed” 145

Though Clark’s assertion of a brief B Major tonicization during the F#

Major prolongation of measures 46 through 49 seems to me excessive,“ her
reading of measures 50 through 56 as a tonicization of Fb [E] Major concurs
with my interpretation of the passage. Her graph (example 4) shows exactly
the same bass pitches as my 5.5, which more fully interprets the harmonic
progression and hierarchical relationships. As in previous passages, Clark’s
graph does not fully engage the soprano. My version shows how the upper
line strongly projects the tonicization through the descending traversal of
the key’s Ch>F) [B>E] fifth.

Measures 56-69

Two flaws mar Kramer's graph A in the region from measure 56 through
measure 69. First, he forgets that Schubert has written the passage in a
four-sharp, rather than a three-sharp, signature: the chord of measure 63
in Kramer’s graph lacks a natural beside the D notehead. I assert that the
lied in fact has been in Fb [E] Major since measure 50 and will remain
in that key through measure 69. Second, he acknowledges - through a B
notehead in the bass and the measure number “68” - a chord that he elects
not to analyze. That B’ chord is, of course, the dominant of E. The E-A-B
root progression of measures 56 through 69 corresponds to I**” IV V7 in
the key of E Major. (See 5.6, which is notated in F} Major.) As often occurs
in Schubert's music, I takes on dominant-emulating characteristics as it
approaches IV. (Thus I-IV becomes I-IV - not VI.) Kramer's neglect of
measures 68 and 69, like the omission of the E chord in measure 50 and of
the At in measure 51, gives the impression that he is discounting composi-
tional events that happen not to conform with the conception he is advocat-
ing. If, as I propose, the essential harmonic activity of measures 56 through
69 is the traversal of the bass fifth from E to B, then root A (represented by
inverted chords in measures 64 through 68) is hierarchically dependent
upon those two perimeter chords. Yet Kramer's graph B displays that A as
one of the four most significant roots in the entire lied.
Kramer describes the lied’s harmonic progression as “little more than
nonsensical from the Classical standpoint evoked by the opening measures”
and refers to a “melting away of tonal coherence” (1986, p. 224).'° Though
the lied is admittedly unusual, in my view its novelty has little to do with
what Kramer shows in his two graphs. I regard the passage from measure
1 through measure 70 to be the start of a structure that I trust would be
absolutely coherent if allowed to conclude.'® Schubert audaciously aban-
dons this structure. Beginning at measure 71 a second structure, itself fully
146 Harmony in Schubert

coherent, emerges. Kramer’s accusations of nonsense and incoherence stem

from a failure to comprehend just how extraordinary the event of measures
70-71 is. Consequently he does not interpret the lied as the juxtaposition
of two independent structures corresponding to the earthly and heavenly
phases of Ganymede’s story. He explicitly rejects the notion of disjunction:
“The transfer of desire from a female to a male love object, from nature to
divinity, suggests a poem of radical disjunction . . . But a look at the poem
reveals that this discontinuity is actually the articulation of an underlying
continuity” (1986, pp. 224-225). Even if Kramer is capable of interpret-
ing the poem as a weaving together of “nature and the divine into a larger
unity” (1986, p. 225), my analysis of the lied suggests that Schubert instead
was intent upon emphasizing the discontinuity.
Clark likewise misconstrues the hierarchical relationship among the
roots E, A, and B in measures 56 through 69. (Compare this portion of her
graph with the I-IV-V progression that opens my 5.6.) Though she does
not describe root A as a tonic,'” as does Kramer, she uses an open notehead
— one of only six in her entire graph - to display its third, bass Ck (measure
66), and interprets the following B as a neighbor between this C and that in
measure 71.'8

Measures 68-75

As I have proposed in other contexts as well, Schubert is inclined to

perform the mechanics of a switch from one tonic to another via their
respective dominant chords: a tonic proceeds to its dominant, this domi-
nant segues into a different dominant, that new dominant resolves to a
new tonic. This procedure is operative in measures 68 through 75 as a B’
dominant, following an E tonic, segues into a C’ dominant, preceding an
F tonic. (See my 5.6, presented, without resort to Schubert’s enharmonic
respellings, as Fh Major to Gbb Major.) Indeed the juxtaposition of B and
C dominants is highly unusual. Yet this is the very moment of transition in
the poetic text from Ganymede’ earthly to heavenly life. The story Kramer
tells is much different. Whereas I place the shift at measure 71, the discon-
tinuity in Kramer's graph B occurs much earlier: between measure 46 (bass
Gb) and measure 56 (bass E). Whereas I correlate F} [E] Major and Gb} [F]
Major, the open-notehead chords of my 5.6, Kramer instead makes E Majer
dependent upon A Minor (his iii). As I have stated above, this seems to me
a reversal of the intended hierarchy: the A chord functions as subdominant
within a tonic-to-dominant trajectory in E Major.
Though Clark here makes no reference to the equivalent $ chord in
 “Ganymed” 147

measures 25 and 26 (discussed above), she suggests that the ° chord in

measure 70 “resolves (though not satisfactorily)” the B’ chord of the
preceding measures (p. 233). Thus she is consistent in her views, but in
my opinion consistently mistaken. Instead, 5.2 proposes a discontinuity
between measures 24-27 and measures 68-74. The former is a normative
context for a $ chord, whereas the latter is a highly idiosyncratic structure
whose startling Cb<Dbb [B<C] bass rise is a vivid musical depiction of the
principal event of the lied, the moment when Ganymede is plucked up by
Zeus. Clark's reading offers no correlation between the poetic and pitch
dramas: her open-notehead C arrives during the preceding keyboard inter-
lude (measure 66), at a spot characterized by continuity with the preceding
material (descending stepwise fifths in every measure from 60 through 67)
and in the midst of a diminuendo.

Measures 75-121

Kramer describes the onset of tonal stability at measure 75 as follows: “the

arrival at F major, then, splits the song harmonically into a wandering half
and a centered one, and so articulates the gap between mundane and divine
rapture that Ganymed quickly, inexplicably, orgasmically crosses” (1986, p.
228). Orgasmically? It is unclear whether Kramer is asserting an explicitly
sexual reading to the passage, or whether that word instead is being used
in the more general sense of unrestrained excitement. In his book, which
followed in the wake of the huge brouhaha within the musicological com-
munity concerning the possibility that Schubert was homosexual, Kramer
advises that “it has become important to ask . .. how the music may or may
not project sexual perspectives or sexual subject positions, regardless of
whether its expressive content is explicitly sexual” (1998, p. 94). My brief
response to these issues is that the poem's imagery of an eagle hoisting a
youth to the heavens is in fact an embrace, but no more. (The individual
voices of Schubert’s sequential passages during the F Major region of the
lied work in close alliance. A Zeus/Ganymede unity, not two separate
entities, glides upwards through the sky.) I perceive nothing in Schubert's
writing within “Ganymed, other than his willingness to set a text that cel-
ebrates a male-to-male bond, that might reflect his own sexual proclivities.
This lied ends an augmented second [minor third] from where it begins;
Die schéne Miillerin, which is clearly about heterosexual passion, ends an
augmented fourth from where it begins. Schubert uses extraordinarily crea-
tive means to convey the wide panorama of situations that the poetry he
sets evokes. “Ganymed” is no exception.
148 Harmony in Schubert

Though his graph A displays measures 75 through the end in somewhat

unorthodox notation (in particular, the lack of connection among some
of the bass noteheads),!? Kramer displays a convincing close for the lied.
Though the route of the background descent that I prefer is somewhat dif-
ferent (as stated above, I hear background 4 in measure 1 13), this is an issue
about which analysts might reasonably disagree.
Clark’s graph displays the F Major region in greater detail than it does
the preceding regions.” It is helpfully arranged on three systems so that
the three cadential points (measures 94, 110, and 116) appear in vertical
alignment. (Why, though, is the soprano at measure 94 labeled 6-5? This is
instead a conventional half-cadence on eSantecedent of the descent through
2 to 1 at measures 115-116.) Some of the graph’s details seem peculiar to
me. The boxed Ck-C# connections (measures 85-87 and 100-102) strike
me as misleading: if at one point the bass passes from C to D via C# and at
a later point C occurs in the soprano, the C# and the latter C are not related;
and the C from which soprano C# emerges in measure 102 is not present
in the graph. As has been the case often in earlier passages of her analysis,
the soprano in this region is spottily represented. Why are the C#<D and
E<F half steps in measures 102 and 103 represented, but the preceding
A<B} and Bs<C half steps in the same metrical position in measures 100
and 101 not? The graph works hard against promoting the interpretation
of the passage that I think needs to be made: as a prolongation of tonic F
Major (measures 100 through 103), which takes on dominant-emulating
characteristics leading to IV in measure 104. Its essence derives from the
ascending 5-6 sequence, as follows:
m. 100 101 102 103
A Dp Dio Cant? Dae aE
F Major: |

The progression starts off running, with dissonant Eb present at the outset.
Unfortunately a printer’s error further muddles the passage: the stemmed
bass note in measure 103 of Clark’s graph should be F, not G. Her sugges-
tion that the recurrent motions to the subdominant “frustrate the sense of
V-I rather than articulating it, resulting in “a vaguely troubled conclusion”
(p. 233), is countered by my argument that upward and downward motions
(referring explicitly to events described in the poem) occur in alternation.
Indeed, the upward lines are highly charged. Yet the descending IV-V(-I)
phases are nurturing, benevolent. I do not share Clark’s concern that Bb
might emerge as the final tonic of the lied. In their broader context the
passages from F to Bb clearly project I>I1V, not VoL
6 Quintet in A Major (“Trout,” D. 667),
movement 1
with David Beach

The structural dynamic of an exposition and of a recapitulation are not the

same. Whereas both generally will begin with a tonic prolongation, the
exposition is an open-ended structure (in a major key, usually pursuing
the tonal route of I followed by II#-V), whereas the recapitulation will
both begin and end on the tonic (I followed by VI). An alternative to this
scenario is encountered occasionally: the recapitulation may be presented
as IV-V—I, a transposition of the exposition’s I-II#-V structure. This
so-called “subdominant recapitulation” places the primary theme in an
awkward tonal context. Does the subdominant root serve as a lower neigh-
bor within a prolonged dominant that extends well into the recapitulation?
Or is the recapitulation’s basic structure IV-I? Either way, we enter a topsy-
turvy world in which the recapitulation’s conventional tonic-embracing
I-V—JI structure is absent. Whether one regards this as a good idea or not,
enough movements of this type (including some by Schubert) have come
down to us that our close attention is warranted.'
In one of several articles devoted to Schubert's music, David Beach con-
siders some of these movements.” He proposes that “it is . . . necessary to
[explore] the relationship between what is often referred to as design — the
broad category which includes not only surface compositional elements
like motive and rhythm, but also features of large-scale organization such
as form and key succession - and underlying structure” (p. 1). I will both
assess his reading of the first movement from Schubert's “Trout” Quintet
(D. 667) and offer an alternative explanation for how the subdominant
that opens the recapitulation functions within the broader structure of the
movement as a whole.

Exposition

What ought we to make of the fact that the expositions introductory

region (measures | through 26) ends exactly as it begins — that the tonic
A Major arpeggiation of measures 1 and 2 recurs in measures 25 and
26? I propose that Schubert imbues the opening of the quintet with a
sense of cyclic repetition.’ We first hear the concluding two measures
150 Harmony in Schubert

of a primordial cycle. Measures 3 through 26 form the next cycle, and

measures 27 through 39 yet another. We expect that the cycle beginning
in measure 40 will conclude with a perfect authentic cadence in meas-
ures 60 and 61, following the dominant of measure 59. In that case the
cadence is withheld. Observe how the cadential tonics become gradu-
ally more vibrant: measure 2 is a mere tripled A, while measure 26 is a
pulsating A-C#-E chord. Measure 39 adds a dash of B>A bravura. From
this perspective the first chord in Beach's graph (his example 7) is per-
forming a double duty, representing both the end of one cycle and the
beginning of the next. I suggest that instead a distinction should be made
between soprano A (resolution) in measures 1 and 2 and C¢ (initiation)
in measures 6 and 10, and that what Beach calls “Th[eme] 1” begins not
in measure 25 but in measure 27.*
The chromatic pitch C4 occurs prominently in measures 11 through 22.
Though he does not coin a term corresponding to my “wobble,” Beach's
description is congruent with that notion: “The primary feature of this idea
... is the modal change C#-C4, followed by a brief recapturing of the C#
before the descent to local closure” (p. 10). In this case, however, I do not
perceive a wobble. The Cy instead is a chromatic passing note connecting
C# and B. The C# of measure 23 is a local embellishment of the following
B, not a restoration of the C# that preceded C4. In fact, Ck is prolonged
through two phases. First it is supported by tonic’s lower-third chord (in
its second chromatic variant). At this point the prospect of a C# restoration
- and thus the interpretation of Ck as a wobble — is still alive. Yet when
D# sounds in measure 20 the root shifts to B. (Compare with Chord 10 in
1.2.) In that context Cy is a chordal ninth, a downward-tending pitch. The
following C# (measure 23) is an accented neighbor, a temporary diversion
from the inevitable C4>B resolution.°
Beach proposes that measures 40 through 58 present a varied repetition
of this pattern, which he displays in letter notation as C# (Ck B A). (The
parentheses denote a subsidiary, C#-prolonging motion to inner voice A.)
I perceive no connection between C# and Ck. Beach himself draws a slur
connecting D and B in his soprano. Ck, which resides within that third, is
a product of dominant emulation: D*-*4G}. It does not emanate from Cd.
Though this D chord (prolonged from measure 48 through measure 54) is
quite interesting, Beach does not devote much attention to the local har-
monic progression of this region, given his broader agenda. In fact, he com-
presses two separate statements of the theme into a single representation in
his graph, choosing to display the second, more complex version. A closer
look now will reveal the magic of Schubert’s harmonic writing and generate
 Quintet in A Major 15]

a proposal for amending Beach's graph. (In preparation for this discussion,
a review of my commentary to 4.4 may prove useful.)
Particularly when its quality is major, a chord built on a key’s fourth
scale degree possesses multiple potential successors. Of course, V is a
common choice (thus A~-D-E when A is tonic). Yet an alternative circular
continuation (for example, A-D-G+-Cz) is also viable. To negate the latter
potentiality, IV often undergoes alteration via a 5-6 shift (thus IV>*-V).
Schubert pursues this course during the first presentation of Theme 1:
IV arrives in measure 35 and is reiterated in measure 36, yet before the
latter measure is over, IV’s 6 phase sounds, with B-D-F# represented by
dominant-emulating D2-F2-A-Cz. (Compare with 1.22.)° Model 1 of 6.1
shows the essence of this construction. In the second presentation of Theme
1, the arrival on IV is even more emphatic, with statements on three suc-
cessive downbeats (measures 48 through 50). Here also what at first may
seem to be IV’s 6 phase emerges: D2-F2-A-Cs sounds in measure 52 and
again at the end of measure 54. (This is the point where a change in Beach's
graph is warranted: bass FZ (not Fz) in measure 54 should be the endpoint
of the slur from D and of the diagonal-line connection with the soprano
F2 of measures 48-50.) Schubert here exploits a most intriguing feature
of the tonal system. As his earlier statement of the theme demonstrates,
D-F2-A to D2-F2-A-Cz (measure 36) is a powerful means of directing the
IV towards V. However, in the second statement he takes advantage of the
extraordinary coincidence that this version of IV’s 6 phase (D#-F2-A-Cx) is
the enharmonic equivalent of one of IV’s dominant-emulating variants (F2-
A-Cz-E5[D2}).7 A G chord (with °embellishment) in fact emerges in meas-
ures 55 and 56, and since it is also dominant-emulating, one may surmise
that an A~>D-—G=-—. . . circular progression will lead onward to Cz. At
the last possible moment Schubert pulls out of that trajectory, with the Gs
chord serving as a chromatic variant of dominant E’s upper-third chord.
(Compare with 1.11.) Model 2 of 6.1 shows the essence of this construc-
tion. Perhaps in acknowledgment of the rocky road that led to the A tonic
of measure 58, Schubert inaugurates another brief traversal of I-IV-V-...
(measures 58-59), thereby enhancing the sense of closure,’ but in this case
the cadence is pre-empted. (See 6.1, Model 3, the only instance among
the three graphs in which diatonic C is employed during the connection
between D and B in the soprano.) Finally a restored A (measure 62) serves
as a foundation for a contrasting trajectory: I-II2—V leads into the exposi-
tion’s dominant region, as graphed by Beach.’
What Beach labels as “Th[eme] 2” transpires between measures 64
and 84. An amplification of Beach’s reading is given in 6.2. Note first
152 Harmony in Schubert

Model 1 Model 2
m. AG EASES 4 Dao TESS
m. 3%) S35) XO, ol 38

a 7
A Major: I IV°-° VI A.Major wil ie |Vaaty da\jocan!
94
€ 7)

Model 3
m. 58 59 x

poise weit

=a 7
A Major: eam
I Vie

6.1 Analysis of Quintet in A Major (“Trout,’ D. 667), mvmt. 1, mm. 33-38, 46-58, and
58-60.

that the Bt of measure 68 behaves in the same manner as E# in measures

34-35 and 47-49. The situation is somewhat more complex here, yet in
a way that corresponds very closely to Schubert’s early “Hagars Klage.”
(Compare measures 64 through 70 in 6.2 with measures 247 through 251
in 1.5.) The 6 phase of IV in measure 69 correlates with that in measure
36. (Compare with 6.1, Model 1.) My graph contrasts Beach’s in one
important regard: I hear two motions to an inner voice emanating from
D (B). First a third is traversed in measures 70 through 75; then a fifth in
measures 75 through 84. Consequently I again question the motive that
Beach represents using letter names between the staves of his graph. This
time, I do acknowledge a wobble (G#>G4<G#). G4 fosters the motion
from E to upper-third Gk (measure 80), which asserts itself and calls to
mind the near-attainment of A’s upper-third Ck in measures 46 through
57. The Gk<G# bass motion at the end of that passage corresponds to the
D<D# bass motion in measures 81 and 82. At various points B (2) is reit-
erated. This prolonged B inaugurates the keyboard interlude that begins
in measure 84.
 Quintet in A Major 153

m. 64 68 69 70 73 74 75 80 81 82 83 84

gauss
Ce
ae -prg. 5-prg.
a at ° +The «> | eta a ae

a —— ee
pt Se == be e

(= G Major:I IV V -)
10 —9 :
E Major: 12:2; IV?
3 =
Vearecs I W()v3 TsVavieel
9b

(=ITj,)

6.2 Analysis of Quintet in A Major (“Trout;’ D. 667), mvmt. 1, mm. 64-84.

Beach's graph of measures 84 through 93 (his “Th[eme] 3”) corresponds

to FC, fig. 109b, which he acknowledges. Though Beach suggests that the
third theme “is derived from the second” (p. 11, n. 13), curiously only his
reading of the third theme contains a descending fifth-progression from 2,
a motion that I regard as the melodic foundation for both themes, though
with contrasting harmonic support. (Note especially how G# coordinates
with local tonic root E in measure 82 but appears in the midst of a IV°°
motion in measure 91.) The theme is repeated, in an abbreviated form, by
the full ensemble in measures 93 through 100.
Beach does not address the remainder of the exposition. (It is acknow-
ledged by the notation “etc.” in his graph.) Given the general sentiment
that Schubert is sometimes long-winded, one may well wonder what the
additional content might offer to attentive listeners beyond yet another
powerful cadence at measure 143. From the perspective of harmonic
practice, these measures meaningfully extend Schubert's exploration of
third-relations and adumbrate the chordal trajectory that will be pursued
during the development section. Chords that reside outside a key’s dia-
tonic collection expand the emotional range of a work. In this movement
Schubert calls upon an extended range at the very outset, through the A>F4
juxtaposition in measures | through 18. (The structurally deeper chord is
underlined.) Later, in measures 55 through 57, the dominant is projected
through a similar, though opposite, third-relationship: G4>E. These juxta-
positions attract attention because they are unconventional occurrences
~ at least until composers such as Rossini and Schubert began using them
extensively. The bass succession from G to G# (measures 56-57, repeated
in measures 60-61) comes across as an abrupt change of trajectory, in this
154 Harmony in Schubert

case a vigorous return to the tonic key, curtailing a more venturesome tonal
exploration.
Once the E Major dominant key is established, third-relations extend into
that context as well. Expanding upon the G4>E upper-third relationship
presented in the tonic region, the E>B[D#*] tonic-to-dominant trajectory of
measures 79 through 82 is complemented by a parallel relationship between
Gs and Dé a third higher, thus invoking both E<Gk and D§>B. (6.2 displays
how the motion from E to Gk is negotiated through the prior wobble of the
E chord’s third from G# to Gk.) During the latter portion of the expositions
dominant region this agenda is potently expanded. First the A>Fs of the
movement's opening measures is reiterated in the context of the dominant
key, as E>Ch in measures 100-109. (In a truncated transposition of the earlier
passage, F#-B-E completes that progression.) Placing this third-relationship
within a dominant context is vitally important to Schubert’s agenda for the
development section, which likewise begins with an E>C4 succession.!? Then
in measures 129 through 140, E Major's dominant B is expanded as B<D4>B.
The D chord is itself tonicized in measures 136 through 140.

Development and recapitulation

A recurrent notion from the exposition inaugurates the development

section: a descending major third from the dominant root (E>C). Beach
offers no advice concerning how one should interpret the Ck chord or any
other event before the development’s concluding A’ (measures 203-209). I
propose an alternative perspective, one in which the development’ interior
organization plays a critical role in how one interprets the D Major region
that begins the recapitulation (measures 210ff.). As a third hypothesis
regarding how the movement is structured, 6.3 is intended to stand beside
Models a and b in Beach's example 8. (More detailed readings of passages
from this model are presented in 6.4 and 6.5.)
Beachs Model a (example 8) displays an important slur connecting the
bass E noteheads that precede and follow the development. In that context
his D chord (labeled as IV in a separate row above the more basic V*7
analysis) offers a consonant context for the introduction of upper-voice D,
which will serve as the dominant’s seventh when E is restored.!! See 6.3,
which retains that slur, but proposes a different set of interior chords as
facilitators of the dominant prolongation. This reading is more radical than
either of Beach’s in that the D Major start of the recapitulation (measure
210), a “deep” moment in terms of the movement’s design, is interpreted
as internal to a linear initiative that connects Ab and E (to be explored in
 Quintet in A Major 155

m. 6 63 «64 150/158 189 248 249

W> N> 1

ese tara a
pe: de : ae F: 2p,

912 ==—
A Major: I al

6.3 Analysis of Quintet in A Major (“Trout,’ D. 667), mvmt. 1.

m. LSOR MGS 70173) 179) 18012 188) 189

Ch i > Bb saa Eb > Ab

6.4 Analysis of Quintet in A Major (“Trout,’ D. 667), mvmt. 1, mm. 150-189.

m. MID" NO ae I Ie 196 197 199 200 203 210 248

oy
$ bree ZZ go eg

Oe
Peotiibe! Mca weo ne esHee ee ne
bowtie aR e_ te ‘Ss g
Ab ( GE

6.5 Analysis of Quintet in A Major (“Trout,’ D. 667), mvmt. 1, mm. 189-248.

detail below). The E>Cy third-relationship that opens the development is

followed by a C>A} third-relationship. The completion of an octave con-
nection via circular progression would result from one additional descend-
ing third: Ab>FDb/E. Schubert arrives at E not by descending this third, but
instead by ascending an augmented fifth in a mix of whole and half steps
(recouping the descent of the two major thirds), within which an extended
prolongation of D occurs, coinciding with the onset of the recapitulation.
An arrow in 6.3 acknowledges this ascending trajectory.
156 Harmony in Schubert

Within the large-scale circular progression of descending major thirds, a

circular progression of descending perfect fifths connects the third-related
Cand Ab chords [6.4].!2 The descending melodic line from Cy to Ab extends
upper-voice Ck (a neighbor to the dominants 3, B), while the C4>Ab bass,
coming after dominant root E, strongly suggests that Ab will lead onwards
to a restoration of F}/E. (Though the juxtaposition of two major thirds is
not a common occurrence in music of this era, when it does occur a third
major third is a conventional continuation. An enharmonic adjustment
acknowledges a tacit acceptance of an augmented seventh - E>C>A}b>Eb -
as the equivalent of an octave.)!> Schubert’s upward course from Ab to E is
an atypical variant of this notion, yet only that course passes through the
D Major that constitutes the subdominant initiation of the recapitulation.
A model of this idiosyncratic ascending trajectory is offered in 6.5. Most of
the 5-phase chords are separated by a $,§,or §chord. Some of these 6-phase
chords are dominant emulating, while others are not.
The juxtaposition of Beach’s and my hypotheses" highlights the extraor-
dinary subtlety that the analysis of Schubert’s music often requires. The
crux of the matter is the structural depth of the A’ chord in measures
203-209. In 6.5 it is displayed as a 6-phase chord, preparing the next
5-phase chord within an extended ascending trajectory (thus akin to the
chords of measures 190, 192, 196, and 199, despite its longer duration). If
this traversal of Ab<E in fact represents the concluding major third of an
E>Cs>A}>F)/E circular progression, then the A’ chord and the following
extended D region are structurally dependent upon that broader initia-
tive. Though I find the extraordinary divergence between the conventional
sonata design and this circle-of-thirds middleground structure straddling
the divide between the development and recapitulation to be disconcerting,
I offer this perspective as a provocative alternative to Beach’s two conserva-
tive hypotheses. In the abstract Beach’s proposals for the basic structure
are viable as well, though by neglecting to explore their interface with
the specific content of this particular development’s middleground chord
progression he does not offer readers a sufficient analytical framework for
understanding. In particular, how are the development’s circular opening
(E>C}>A}) and its A<D close to be integrated? Granted, the ascending 5-6
sequence displayed in 6.5 is unusual both in its construction and its place-
ment. Yet the sweep from Ab to D proceeds without a hitch. (The D<E con-
tinuation contrasts the measures leading up to D: the intervening 6 phase
has been elided. Consequently the hypothesis of an Ab<E span might seem
eccentric to some readers.) I have difficulty pulling the A chord (measures
203-209) out of its sequential context and assigning it the role that Beach
 Quintet in A Major 157

does. In 6.5 the D Major prolongation that begins in measure 210 is interior
to a broader linear initiative. Without the design necessity of marking the
onset of the recapitulation, the structure that precedes that moment could
continue unimpeded to the attainment of goal E. In fact the C<C# bass
in measures 245 and 246 recalls, in a truncated form, the portion of the
development's sequence corresponding to measures 197 through 209 (now
with 5-phase chords separated by a whole step rather than by half steps). In
this latter context we hear C<C#<D<E in direct succession — one measure
per bass note. Could this be atonement for the extraordinarily stretched
D Major prolongation that accommodated the statement of the primary
theme? Could the earlier motion to D Major be regarded as an eventu-
ally rejected “Try 1,” and the latter motion through D to E as an ultimately
triumphant “Try 2”?!
Such a structure alters the shape of a sonata-form movement. With the
delay in the tonic’s return, a greater sense of symmetry emerges: the first
part of the exposition correlates with the second part of the recapitulation.
This tonal relationship prevails even though a thematic correlation still
binds the first part of the exposition and the first part of the recapitulation.
In this movement the connective role of D Major within the middleground
Ab<E chord progression thwarts the tonal unity that generally reinforces
the thematic unity. The very A-B—E passage that steers the exposition
away from tonic in measures 62 through 64 reemerges in measures 247
through 249, transposed as D-E—A, to bring about the restoration of the
tonic. (This correlation may explain why D is the only 5-phase chord in 6.5
that is not extended via a 6 phase.) The conventional synchronization of
tonal and thematic markers imbues the starting point of most sonata recap-
itulations with an overwhelming sense of resolve. In this movement the
non-coincidence of these two markers results in a more diffuse boundary.
The region between measure 210 and measure 249 concurrently completes
some of the business of the development and initiates some of the business
of the recapitulation, just as a suspension allows the content of one chord to
linger into the domain of its successor.
Though I have quibbled over some details of Beach’s reading, my central
concern with his presentation is that he has left his readers in the lurch
regarding the development section, whose structure we may reasonably
expect will be a significant factor in making sense of the recapitulation’s
unconventional D Major starting point (which is, after all, what motivated
Beach to analyze this particular movement).'® If the ascending bass in 6.5
were to terminate on D rather than E, how exactly would antipodal Ab
(measure 189) interact with goal D? Beach does not address this issue.
158 Harmony in Schubert

Instead he places an inverted A major chord within parentheses and labels

it “Dev[elopment]” (his example 8, Model a). Clearly this is inadequate.
Though conceivably this is a reasonable boiling down of the development's
content to its essence, it is imperative that we be able to retrace the process
by which such a crucial decision has been made through study of more
detailed analytical graphs. As things stand, I am unable to reconcile my
perceptions with the conception that Beach proposes.
7 Symphony in B Minor (“Unfinished,” D. 759),
movement 1
with Richard Taruskin

My reading of the “Unfinished”

Richard Taruskin’s perspective on Schubert may reach more musicians

than that of any other contemporary author, given its prominent venue.’
Laudably his history of Western music embraces a vigorous analytical
agenda, and Schubert's music is awarded the prominent position that it
deserves. His discussion of the first movement from Schubert's Symphony
in B Minor (“Unfinished”) is extensive and sophisticated.” Yet my analysis
of that movement is at variance with Taruskin’s in many respects. I will first
present my reading in detail, without reference to Taruskin’s, and then will
test various assertions that he makes against that analytical framework.

Introduction (measures 1-8)

Ten portentous pitches introduce one of the most celebrated and beloved
of all symphonic first movements. Ascending and descending thirds -
outlining B minor (B<D) and G major (B>G) chords — delineate the 5 and 6
phases of I?~* before continuation to the dominant [7.1, Model 2]. Schubert's
melody is a modest expansion of an interrupted third-progression, which
accounts for six of its ten pitches (compare Models 1 and 2). Two of the
others (FZ and G) correspond to the 5 and 6 of I-~*, while the remaining two
(CZ and A) are passing notes. Model 2 in fact previews the structure of the
entire sonata-form exposition: tonic is projected during measures 9 through
38, the submediant takes its turn in measures 42 through 104, and the domi-
nant is achieved in measures 107 (¢ embellishment, unfurled in measure
108) through 109 (inverted ,).* Though tonic’s minor quality and the lines
gradual descent evoke a somber mood,’ the final pitch, F#, offers an unex-
pected change of direction and spirit, a welcome spurt of vitalizing energy.

The exposition’s tonic region (measures 9-38)

Features from the introduction are woven into the musical fabric as the
exposition gets under way in measure 9. The B<C2<D third of measures
160 Harmony in Schubert

Model 1 Model 2

B Minor: I V# B Minor: I° : Vi

7.1 Analysis of Symphony No. 8 in B Minor (“Unfinished,’ D. 759), mvmt. 1, mm. 1-8.

m 9 10 ll

pense eine
irBo
N

a —
G = | =
eZ Re* f=
ee ee Se eee

pay r ?
o
Ey =

8 7h 8
B Minor: Ie 4 2

7.2 Analysis of Symphony No. 8 in B Minor (“Unfinished,” D. 759), mvmt. 1, mm.

9-11.

1 and 2 recurs in the second half of measure 9, along with D<E<F#. F#

proceeds to G in measure 10, followed by the return of F# in measure 11,
echoing the introduction’s F#<G>F# motion. The B>G contour of meas-
ures 3 and 4 appears in the bass of measure 10 and will extend to D (as in
measure 5) in measure 17.
A condensed view of measures 9 through 11 is provided in 7.2. (Schubert
inverts the model's lower set of thirds in measure 10 into sixths, thereby
placing A#<B<C¢ in the prominent position that it will assume in the oboe
and clarinet melody. Compare with measure 14.) Four neighboring notes
embellish the tonic’s pitches. Against pedal B these neighboring pitches are
bound to return to tonic pitches. But when bass B is displaced by G at the
end of measure 10, other potential continuations become feasible, due to
the enharmonic versatility of the diminished seventh chord. Though the
tonic returns in measures 11, 13, and 15, in measure 17 it is displaced by
its upper-third chord, D major, when upward-tending A# is transformed
into downward-tending B} in measure 16. This D chord asserts itself as
 Symphony in B Minor (“Unfinished”) 161

m 6 17 20 II
DD) 26 29 ||
31 34 38
ras

3}
€3 2 1)
auc Seo

rae
3.
a ot
:
:2
< be
ie
a =

B Minor: [I Il Vi I

7.3. Analysis of Symphony No. 8 in B Minor (“Unfinished? D. 759), mvmt. 1, mm.

Seis).

=
III through a tonicization in measures 17 through 20 (I-II Apo
i); /

though here the major mode is less radiant than it often is due to the chro-
matically deflated neighboring notes Fy (measure 18) and Bb (measure 19).
The basic structure of measures 13 through 38, displayed in 7.3, is rich
in associations with the introduction. Here the ascending soprano C# and
descending bass G (corresponding to the introduction’s second and seventh
pitches) are juxtaposed, producing the strong thrust towards D major
mentioned above.° (The mediant here and the introduction’s submediant,
which will resurface later in the exposition, both support 3 and both lead
onwards to V#.) There is some hesitancy in the melody: B<C# in measure
14 at first fails to rise to D, which arrives only on the second try (measure
17). D major eventually leads onwards to the dominant, F#, which supports
2. The progression is interrupted at this point (measures 20 and 21), with
the poignant melodic flutter between G and F# in the horn part recalling
the 5-6 shift from the introduction and recapturing the melody’s F# from
measures 13 and 15. The consequent phrase that begins in measure 22 at
first likewise progresses no further than the dominant (measures 29 and
30), despite the more dynamic sequential ascent that transpires in place
of the antecedent phrase’s mediant tonicization [7.4]. Schubert picks up
the thread again with III in measure 31, this time allowing his dominant
to extend from measure 34 through measure 37 before the long-awaited
cadence on tonic in measure 38.’
The sequence in measures 26 through 29 reveals Schubert's characteristic
creativity. Whereas more typically the ascent would be via 5-6 cycles, here $
chords occur during the 6 phases. Schubert daringly underscores the inter-
nal ascending fourths (e.g., D-G in measures 26 and 27) by incorporating
162 Harmony in Schubert

5 6 5— 6 5
3—4 34— 4 34
B Minor: II ) Vi I

7.4 Analysis of Symphony No. 8 in B Minor (“Unfinished,” D. 759), mvmt. 1, mm.

26-38.

a dominant-emulating C4 within the D chord, thereby generating some

uncertainty as to whether the D or the G chord is in fact the 5-phase chord.
From a broader perspective, however, it is clear that the sequence ascends
from D to F# (mediant to dominant, reiterating the direct D-F# succession
of the preceding phrase). Yet something is afoot here, for DG, passed
through within the sequence, is reprised in measures 38 through 42, where
it accomplishes the crucial tonal shift from the initial tonic region into the
exposition’s extended submediant region, G Major, a tonicization of tonic’s
asserted 6 phase. Perhaps in order to minimize the variance from the intro-
duction’s contour, the transition is breathtakingly brief. The D-F#-A-C
chord of measure 41 substantiates the A>F# melodic third of measure 4.

The exposition’s submediant region (measures 42-104)

The simple yet perfect melody that opens the G Major region is timeless,
seemingly stemming from a distant past and extending through eternity.
Schubert enhances the sense of familiarity by integrating some of its fea-
tures with those of the introduction. The cellos and basses descend G>D in
measures 4 and 5. That interval, again performed by the cellos, inaugurates
the G Major region. In fact, the melody incorporates a retrograde state-
ment of the introduction’s fifth through eighth pitches [7.5]. The founda-
tional structure displayed in 7.6 reveals a further relationship: the pitches
B>A>F#<G from its upper line correspond to the fourth through seventh
pitches of the introduction.
Goal pitch G, which overlaps the second entry of the theme at the down-
beat of measure 53, is withheld at the end of the restatement: in place of
its expected cadential arrival in measure 62 Schubert imposes a measure
 Symphony in B Minor (“Unfinished”) 163

m. 4

rz; =e Fi
oD
| J
SSS See ae

m 44
Yt

7.5 Comparison of Symphony No. 8 in B Minor (“Unfinished? D. 759), mvmt. 1, mm.

4—5 and 44-45.

m 44 49 51 Sy 53
53 58 60 61

3
3 2 1)

f = t SS

caeie
Cy) : te
eae | a

zs —— 2 ZZ
pie zi 2 =i
| Y
Guar Ty em |
¢ VIZ)
7.6 Analysis of Symphony No. 8 in B Minor (“Unfinished, D. 759), mvymt. 1, mm.
44-61.

of silence. Though the previous descent to G (measure 53) follows the

blueprint of the introduction, recall that in measures 4 through 6 the down-
ward motion continues to D. Must our pleasant tune also descend into the
depths? The suppression of tonic G at measure 62 allows the progression
to continue to an imperfect authentic cadence in measure 80, coordinat-
ing with a continuing downward motion of the melody to D. (Schubert
here places that line in an upper register, at times covered by other chord
members.) In 7.7 its content is divided into five phases, separated by bar
lines. The melodic sixth B>D, which corresponds to the span from the
introduction’ fourth through eight pitches, unifies the progression from
measure 53 through measure 80. In the first phase, the progression restates
measures 44 through 52. In phase two, a new harmonic initiative crashes
in upon the hesitant-to-close preceding phase, extending the melodic line
below the G tonic pitch, in the process contorting it in accordance with the
164 Harmony in Schubert

78 79 80
m 3 58 60 «61 63 71-73 74 75 76 Fil

4
aa aoe Ot 5 eo
| 7a Z 2 oS: eo
2 2 | iz $2 ——2 Re A) ke
elision

G Major: I IV} '

6
vy7 I
8)

7.7 Analysis of Symphony No. 8 in B Minor (“Unfinished,” D. 759), mvt. 1, mm. 53-80.

7.8 Complete omnibus expanding C” and analysis of Symphony No. 8 in B Minor

(“Unfinished, D. 759), mvmt. 1, mm. 63-71.

subdominant key (with F4 rather than G Major’s diatonic F#). The third
phase stalls melodically while accomplishing a conventional 5-6 expansion
of IV. The fourth phase brings in the dominant, with the melodic descent’s
final pitch D sounding, but in a dissonant context. (I designate the domi-
nant as a phase in its own right to emphasize its importance within the pro-
gression despite the fact that, unlike the preceding I, IV, and II chords, its
leading tone plays no role in its presentation.)® The fifth phase resolves that
dissonance, cadencing on the G tonic.
In 7.8 it is proposed that the astonishing C minor chord of measures 63
and 64 stems from the omnibus voice-leading technique (introduced in
chapter 3). Schubert's C-E}-G melds components from the C” omnibus’s
first two chords. His tendency towards deformation persists at the close,
with the refusal of C# to descend to Cy. Consequently a shift from C root to
A root over the course of measures 63 through 71 is perceived (confirmed
by the A chord of measure 73), echoing the G>E third of measures 48
and 49 (or 57 and 58). In fact, the theme’s harmonization (G>E<A>D<G
in 7.6) becomes the principal harmonic initiative for the subdominant
 Symphony in B Minor (“Unfinished”) 165

prolongation in the second segment of 7.7 (Cy>A<D>G<Cz with the A and

G chords inverted results in the bass progression Ch<C#<D>B<Cz).
The grandly slow progression of measures 53 through 80 is followed by
two confirmative phrases for full orchestra that more quickly retrace much
of the same harmonic terrain. Whereas the initial progression proceeds (in
accordance with 7.7) as

G E> a DO) iceeA Se ede Gee CO Fe an D34G

that of measures 80 through 89 proceeds as

G eS MA Nee G*> C D’> G

and that of measures 89 through 93 proceeds even more synoptically as

GF (6 E’> a D’> G
In the first of the latter progressions the C minor chord is replaced by C
major (which 7.8 suggests should be regarded as the pre-deformation
norm), while in the second the initial G chord exhibits dominant-emulating
tendencies that lead smoothly and directly into the C chord, eliminating the
aberrational D>C root succession of the preceding versions resulting from
the elision of the initial G tonicization’s conclusion. Perhaps Schubert is
suggesting that if one can endure the vicissitudes of life, eventually things
will work out. The ° fourth, emphasized in the middle region of the intro-
duction [7.1, Model 2] and traversed during the G Major theme, is arpeggi-
ated in the melody between the cadential points of these latter progressions:
D (measure 89) to G (measure 93). The latter pitch of course is a required
element of a perfect authentic cadence. Measure 93 is the termination point
of this thematic region.
The opening idea of the G Major theme returns in measures 93
through 104, now outfitted with extended dominant pedals and with
potent chordal variants: instead of E-G#-B-D during tonic’s 6 phase (as
in measure 49), Schubert now employs G#-B-D-F4 (measure 96), and
the diatonic supertonic A-C-E (measure 51) is intensified as C#-E-G-B
(end of measure 97). There is little time to savor the cadence at measure
104, since a resolute B, which sounds throughout the orchestra on beat
2, serves as a potent reminder that G, though tonicized, is the 6 of I°°.
The exposition thus far has proceeded so far as the introduction’s seventh
pitch. (Compare with 7.1, Model 2, end of measure 4. In both cases the
chord B-D-G is to be understood. The G of measure 104, beat 1, is pro-
longed through to its restatement in measure 106 and then descends to
the dominant root, F#.)
166 Harmony in Schubert

m. CT) 6 m. 106 107 108 109

2: Zi—— i =: 2
+
= |

A Aw A A

= SS
3 2 3 2

2. OD $ $ t¢
—— a = a

B Minor: VI V3 B Minor: VI V3 3

7.9 Comparison of Symphony No. 8 in B Minor (“Unfinished,” D. 759), mvmt. 1, mm.

4-6 and 106-109.

The exposition’s dominant region (measures 104-109)

Certainly we might expect that an exposition that thus far has pursued the
contour of the introduction’s melody may likewise continue to the intro-
duction’s concluding dominant chord. In 7.9 the introduction’s seventh
through tenth pitches and the bass melody of measures 106 through 109
are compared. Observe that the latter is a permutation of the former. In
the introduction Schubert juxtaposes descending and ascending fourths
(corresponding to the fifths of VI and V#, shown in a model below the
melody pitches in 7.9). At the end of the exposition, his unfolding of
the structure proceeds by first presenting one strand (G>F#) and then the
other (D>C#).!° The adjacent F# and D offer a context for an unfurled |
embellishment of the dominant (likewise shown in a model below those
pitches in the example). The passage is mesmerizing in part because, as in
the introduction, the structural soprano (622) occurs below the dominant’s
root (F#).!!

The development (measures 110-217)

The dissonant seventh in the exposition’s dominant goal chord propels the
return to the tonic — either the B-D-F# of the introductory theme, inaugu-
rating the repeat presentation of the exposition, or the dominant-emulating
B-D#-F#-A of measure 110b, inaugurating the development. The latter
chord focuses the tonal thrust of the entire exposition in the direction of
IV, which duly arrives in measure 114 and asserts its significance through
a presentation of the introduction theme (modified) under its auspices. In
the broad scheme of the development this IV holds sway until the domi-
nant begins to emerge in measure 202. The prolongational strategies that
Schubert deploys result in a fascinating whirl of activity within the domain
 Symphony in B Minor (“Unfinished”) 167

m. 13-40 110b 114-192 202 205

3 2 Il
IN

——"
Fel rane
te

i
B Minor: 15,—; IV ==

7.10 Analysis of Symphony No. 8 in B Minor (“Unfinished, D. 759), mvmt. 1, mm.

13-205.

Model 1 Model 2
92 199 202 205

“fe
ooh) Cn NO z | Tr # 4)

= Za g Zz H ware Z. az e

B Minor: IV ( oe B Minor: IV ( ea

7.11 Analysis of Symphony No. 8 in B Minor (“Unfinished, D. 759), mvmt. 1, mm.

192-205.

of IV, with articulation points at measures 170 and 192. The broad struc-
ture, displayed in 7.10, is straightforward, its most novel element being the
substitution of a $embellishment for the more common § at the dominant’s
arrival. Observe Schubert's subtle respelling of measure 199’s diminished
seventh chord in measure 201 [7.11]. The “cadential $ chord” that follows
(Model 2) is unfurled into } position in measure 204.'* Even though a {
chord does not perform its conventional role before the dominant }, it
emerges afterwards, through a series of Vj-{"? neighboring motions in
measures 210 through 212, followed by the motivic G-F# (9-8) in the flute
line, derived from the introduction’s melody and echoing measures 20 and
One
The prolongation of IV (E Minor) within the development is divided
into two phases, bordered by the articulations of that chord in measures
114, 170, and 192. The first phase evolves out of the E Minor variant of the
introduction’s theme. As we have seen, the basic harmonic contour of the
168 Harmony in Schubert

E Minor: 13 1175 4 $ ae
$

7.12 Analysis of Symphony No. 8 in B Minor (“Unfinished,” D. 759), mvmt. 1, mm.

114-170.

exposition (first expressed in the introduction’s theme) is b-G-—F#. That

would translate into e-C-B in the key of E Minor. Though measures 114
through 122 traverse the e>C span, Schubert barely allows the C chord
to stabilize. AZ soon emerges above C-E-G, giving the chord a supertonic
focus. One very consequential shift occurs during the theme's presentation:
the E>C motion is expanded from a third (as in B>G in measures | through
4) into a tenth (measures 117-122). That action motivates a compensat-
ing reaction: before E Minor’s supertonic chord proceeds to the dominant
(a long-delayed goal, like the submediant’s continuation to the dominant
during the exposition), it is hoisted up an octave, in coordination with
the chromatic inflection of its fifth from C4 to C#. How that initiative (to
be explored in detail below) fits within the broader harmonic progression
is displayed in 7.12. Here the deep structural melody’s incomplete neigh-
boring note E, prominently displayed in 7.10, is the goal of a subsidiary
third-descent G>F#>E that transpires during the connection of the E minor
chords of measures 114 and 170.
The extraordinary supertonic prolongation of measures 125 through 162
is grounded upon a straightforward basic structure: an ascending parallel
progression. Model 1 of 7.13 displays its simplest manifestation. Observe
that this progression of ascending § chords is obstinate: a major second
separates the corresponding pitches of all adjacent chords, and each chord
is of the same quality: major. Consequently the progression has one chord
fewer than would a diatonic glide (which would mix whole- and half-step
ascents), and an enharmonic respelling is required at some point in order
 Symphony in B Minor (“Unfinished”) 169

Models 1 and 2

eo a
+

SS= Ske
eeSS Ss
eee
Model3
m. 125 129 130 131 132 133 134 145 146 153 154 158 161 162 166 169
at pr ieeant =

ICP REECE cP
Aap eee9 ade
E Minor: II2, pines
F

Model4
=
Ge = — 9 =
526 HG 5
3—4 3 ——4 3

7.13 Analysis of Symphony No. 8 in B Minor (“Unfinished, D. 759), mvmt. 1, mm.

125-169.

not to end up on G###-Bx-Ex instead of A#-C#-F#.'° Schubert ups the ante,

however, for instead of employing A#-C#-F# in measure 125 (already a
dominant-emulating substitute for E Minor’s diatonic A-C4-F#), he deploys
A#-C4-E-G, which also emphatically seeks resolution on dominant B major
(F#5>B rather than F#—B).
Whereas gliding §chords are relatively common, Schubert’ glide (based
on 7.13, Model 2) is remarkable. Observe in Model 2 that one of the chords
(with bass A) has shifted up a half step. That comes about through a poten-
tial curtailing of the glide. Take a close look at this model's first and fourth
chords: all their corresponding pitches are separated by an augmented
fourth, and yet they potentially represent the same harmonic function.
With a bit of enharmonic reinterpretation (Dx = E, Fx = G), these chords
are nearly identical:

A# Ch E G_ versus
Ee tcAt ie Cherk to which Schubert adds the ninth G, thus
Eteeed Ciel C
170 Harmony in Schubert

The upper © third of this latter chord mimics the rhythmic shape of the
opening theme, against which it sounds:

j FH G (mm. 114-115, also 134-135 and beyond)

G Fe E (mm. 134-135 and beyond)

Will this magnificent chord, with such potent supertonic potentiality

(within tonicized E Minor) stored within, break out of the glide initiative,
assert itself harmonically, and proceed directly to dominant B? Or will
all that striving fail, and the upward glide continue? The ascending linear
motion eventually wins out. However, from this point onward its trajectory
is transformed. First, consider the extra note that was added to the chord of
measure 134 (spelled as G above to explore a potential resolution that does
not materialize, but as Fx and labeled as an anticipation in 7.13, Model 3,
reflecting its role within the ascending trajectory). Since this Fx defies the
ascent by whole steps that has prevailed thus far (A#<B#<Cx<Dx<. . .),
the glide adjusts to accommodate it, moving upwards three, rather than
two, half steps in measure 145 (bass F#<A). This extra-large step requires
an extra-small step in compensation (at measure 154) because the goal is
twelve, not thirteen, half steps from the starting point. Second, note that
the glide has given way to a sequential progression.'* Model 4 of 7.13 shows
the foundation upon which the additional content is based. Compare that
model with the nearly identical sequential initiative during the exposition
[7.4, measures 26 through 29], which ascends to F#, dominant in B Minor,
rather than to F#, dominant-emulating supertonic in the tonicized key of E
Minor. Third, the chords in measures 154 and 162 do not maintain exactly
the same characteristics as those proposed in Model 2. Though the goal
chord of the octave ascent (measure 162) remains a supertonic, its evolved
state is slightly altered from the supertonic that began the progression in
measure 125. The progression continues as F#—>B, not as FRB.
The return of E Minor at measure 170 is marked by a restatement of
the introduction’s theme, followed by its fragmentation within sequential
progressions. Beginning in measure 176 Schubert works in uniform four-
measure units. From E Minor, the opening motto of the theme (adjusted
to E<F#<G>F#) leads upwards a fifth to B. The process is replicated on B,
leading upwards another fifth to F#. This ascending passage thus works as
follows:

m. 176 178 180 182 184

9 FY B Ci’ FF
 Symphony in B Minor (“Unfinished”) al

The transposed middle section of the theme (FE>E>C#<D>A>G#, meas-

ures 184-187) now leads downward by fifths, from F# to B, and then to E,
as follows:
m. 184-186 187 188-190 191 192

ae <—C#’ B>-6-5 FH’ E

Observe that the two sequential passages are in a retrograde relationship:

what goes up comes down. The juxtapositions between measures 187 and
188 and between measures 191 and 192 are of special interest. For example,
F minor is followed by its dominant C¥’, and then B minor sounds. That is
a close approximation of the unusual progression encountered in measures
53 through 63 of the exposition: G major followed by its dominant D’, and
then (after the one-measure silence) C minor.
The E-G-B chord of measure 192 is essentially the same E-G-B chord
that was introduced in measure 114, despite the extraordinary adventures
that have transpired in the meantime. As with the F# and B chords of the
preceding sequence, it likewise is extended via a 6 phase before root Ff,
dominant in the B Minor key of the movement, arrives in measure 202,
followed three measures later by the remaining notes of the dominant triad
(as displayed in 7.10) and in measure 213 by its seventh. The agenda of the
development has been fulfilled. The recapitulation is on the horizon.

The recapitulation and coda (measures 218-368)

In many sonata-form movements the recapitulation offers few analytical

challenges. The restructuring of the exposition’s tonal plan is often predicta-
ble, and large chunks of exposition material will be restated exactly as before
or in a routine transposition. The first movement of Schubert's Symphony
in B Minor is exceptional in this regard. As we have seen, the exposition
consists of three distinct regions (b-G-F#). Transforming b-G into b-B
would be awkward both because the exposition provides little transitional
material to separate two broad expanses centered on tonic B and because
the continuation down a half step (as in G-F#) would need to be suppressed.
Restructuring the recapitulation as b-Ck-B, which would deploy the half-
step descent to a closing tonic, would be equally problematic, due to the
chromatic relationship of Cy to what surrounds it. Besides, a transposition
of the exposition’s F# region, so brief and so clearly of dominant character,
would be ineffectual as a culminating tonic. Rejecting such dubious alterna-
tives, Schubert instead reprises the broad tonal shape of the exposition (b-
F#), reserving the final projection of tonic for the coda. Thus this movement
172 Harmony in Schubert

Model 1 Model 2
Mm 222-226) 252) 258 298 324 325-327
m 13-17 44. 107

W> N>

5 5 5
B Minor: I> ° Vi B Minor: I ( hivl Vi
(= I)
(=)V)
7.14 Analysis of Symphony No. 8 in B Minor (“Unfinished,” D. 759), mvmt. 1, mm.
13-106 and 222-327.

takes on an extraordinary cyclic character: the introduction, exposition,

development, and recapitulation each proceeds from b (or B) to F#. Between
those pillars his writing is tonally audacious and motivically spectacular.
The basic structure of the exposition is summarized in 7.14, Model 1.
(Parallel fifths - @ to S — are averted through the repositioning and unfold-
ing of those pitches during measures 106 through 109.) In the recapitula-
tion a significant shift, which initially may seem puzzling, occurs in the
passage leading up to the cadence in measure 252. During the exposition
that cadence reinforces tonic B (measure 38). In the recapitulation the
passage instead cadences on F#. How Schubert achieves this shift will be
explored below. Model 2 of 7.14 reveals its consequences for the tonal
plan. Whereas one descending-third motion transpires during the expo-
sition (I°°, reiterating the descending-third motive of the introduction’s
cello melody), during the recapitulation that device becomes the guiding
principle for a broader descending trajectory. Not just one, but three suc-
cessive descending-third motions transpire - F# to D, D to B, and B to
G - before the dominant arrives. Successive roots arpeggiate the tonic
triad: Ff>D>B. The first third, F#>D, corresponds to the B>G motion in
measures 38 through 42. Consequently the theme that was first presented
in the submediant key, G Major, now resides in the mediant key, D Major,
creating an interesting and uncommon symmetry. The second third, D>B,
transpires during the reprise of the structure presented in 7.7. Whereas in
the exposition that progression both begins and ends in G Major, in the
recapitulation (measures 267 through 298) it undergoes some slippage in
measures 275 through 279, so that the chord before the full-measure rest
has extended down a third from what the exact transposition of the exposi-
tion's model would have dictated. Thus that structure begins in D Major but
 Symphony in B Minor (“Unfinished”) 3

Model1

m. yn wh 19 20 I
22 26 26 29 II
31 34 38
a b
be — G * S 2 a tH
I J Ht J
c d
ime ik —. srs

Model2
m. 222 226 228 230 231 235 237 239 240 243
245 248 252

SS> — : a
a b ale ob
° = &
a Oo:

a eee | Ss |
c c d

Cz Minor: III! No.

E Minor: I aut
Fe Minor: I Vv!

7.15 Analysis of Symphony No. 8 in B Minor (“Unfinished, D. 759), mvmt. 1, mm.

13-38 and 222-252.

ends in B Major, the latter incorporating the wobbly pitch D#. Schubert in
fact has managed to transform the exposition’s b-G into b-B (see the dotted
slur connecting two bass Bs in 7.14, Model 2), but in a way that provides
abundant tonal contrast. The third descending third, B>G, occurs in meas-
ures 322 through 324. This phase of the recapitulation incorporates a swift
traversal of the exposition’s (and introduction’s) I°-V. (Compare 7.14,
Model 1 and the final chords of Model 2.) The continuation from B is not to
A%, as adherence to the exposition’s G-F# trajectory might have achieved,
but instead to F#, thereby maintaining the tonal integrity of the recapitula-
tion, though exceptionally leading to V2 at its endpoint.
How did Schubert manage to cadence on F# rather than on B in measure
252? He deploys what I call seismic composition — that is, a progression
that, despite unexpected jolts that repeatedly move it off track, nevertheless
perseveres in completing its trajectory on that new track.'° The exposition
model for this progression (introduced in 7.3) is displayed in 7.15, Model
1. Several salient features are marked by letters. The general contour of the
root motion is upwards from tonic to dominant via the mediant, followed
by an undivided motion back down. ‘The letter c corresponds to the span
of I<III. Recall from 7.3 that a C#-E-G-B) chord comes between those
174 Harmony in Schubert

more elemental chords. The III<V# span is labeled d. Recall from 7.4 that a
sequential progression is employed to connect those more elemental chords.
The motion from the mediant to its upper fifth is labeled a, while the return
to the mediant is labeled b. Model 2 displays how these features fare in a
seismic context. The first three pitches follow the contour of Model 1, with
c followed by a. Then the first tremor occurs: the pitch A is replaced by B
(boxed), and the “return to the mediant” designated by b now leads to E, not
D. Though Schubert could have proceeded from E major through G# major
to C# minor, concluding the progression up just one step, he instead treats
the E chord (of minor quality) as a new starting point, continuing with c
rather than d (thus corresponding to I<II] rather than I1<V#). That turn of
events shifts the tonal center upwards another two steps. (Thus a vertically
aligned I and III are displayed within a box in the example.) The G mediant
is prolonged via another ascending fifth (a), at which point another seismic
shift occurs: D is replaced by E. When A arrives (via b) in measure 240, it is
allowed to function as a mediant (unlike the E chord of measure 231), and
dominant C# is achieved via d. The model concludes, as does Model 1, with
a descending fifth (C#>F#). The extra height - a full perfect fifth, added in
three increments — gives Schubert the space to descend three thirds, rather
than just one, during measures 252 through 324 [7.14, Model 2].
Because the introduction, the exposition, the development, and the
recapitulation all have ended on the dominant, the coda, which begins in
measure 328, is charged with the role of securing the final tonic. That is
accomplished through cadences in measure 336, 352, and 366. A modest
counter to the cadential initiative is instilled by the violin line of measures
338 through 341, which brings back a potent chromatic element from the
development (measures 122 through 125), though here the E# appears
in a less complex chord, a straightforward II# that is eventually followed
by V# (measures 348 through 351, with the cadential ; initially unfurled
into $ position). The melody ultimately attains the simple form of a B<D
third during I, followed by an F#<A# third during V#, and finally B for the
cadential I.'° The B resolves all the residual tension caused by the dangling
melodic C#s (= 2) at the close of the earlier sections of the movement.

Taruskin’s reading of the “Unfinished”

Though Taruskin does not attempt as thorough an analytical presentation

as mine above, one should assume that his analytical observations would fit
within a more detailed formulation even where none is explicitly presented.
 Symphony in B Minor (“Unfinished”) 175

Introduction (p. 109)

Taruskin correctly understands the melody’s basic contour as represent-

ing “an embellished plunge from tonic to dominant.” Yet his premise
that Schubert has composed a “mysteriously ill-defined beginning” based
on “the descending-tetrachord motif” (B>A>G>F#) misses two crucial
aspects of what I suggest Schubert wanted to impart through these ten
pitches. First, the B>F# tetrachord is not a smooth stepwise descent, but
instead is divided into a third plus a second, with a perceptible emphasis
on G major, which is fully arpeggiated: D>B>G>D [7.1 and 7.9]. At this
early point G major is no more than the 6 phase of tonic: we do not know
whether it will assert itself beyond its voice-leading origins. Though such
an assertion is not confirmed until measure 42, Schubert sets up a context
for its realization as early as measure 4, and listeners should be cognizant
of this potential plot twist from that point onward. Second, the mysterious
character of the theme derives in large part from the positioning of the
quintessential B25 “soprano” line (which coordinates with Taruskin’s I-V#
harmonic plunge) at the low edge of the melody’s contour. A topsy-turvy —
and thus mysterious — world emerges. Though the parentheses around the
C# notehead in 7.1, Model 1, lack the punch of Taruskin’s vivid vocabulary,
they get to the core of why the melody has the effect that it does. Model 2,
with upward- and downward-pointing stems in a rare convergence, offers
specific data on what has transpired, with the descending octave transfer
of D (and continuation to C# in that lower register) and soulful F#<G>F#
neighboring motion predominating.

The exposition’s tonic region (pp. 109-110)

Taruskin suggests that the texture of the tonic opening is “reminiscent -

perhaps deliberately reminiscent” of passages by Mozart and Rossini. The
full cadence on tonic in measure 38 is singled out as a novelty in symphonic
music up to this time. Yet how measures 9 through 12 relate so richly both
to the introductory theme and to the following B Minor theme is left unex-
amined. (For example, the F#<G>F# neighboring motion is articulated over
successive downbeats.) Taruskin’s commentary makes it seem that Schubert
simply started churning out tonal energy in the manner of Mozart and
Rossini, rather than thoughtfully developing content organically connected
to its context. The theme itself is barely addressed. Schubert's structure
is not a conventional antecedent/consequent pair. The two halves differ
substantially in their internal content. The failure to close around measure
176 Harmony in Schubert

30 and the resulting repetition of some elements of the structure make the
cadence at measure 38 all the more satisfying. Though Taruskin indicates
that the cadence occurs “demonstratively;’ he does not explore how hard-
won that victory was.
Taruskin describes the transition as “an obviously ‘patched-in’ four-bar
linkup” and wonders if Schubert might suffer from a “lack of interest in
transitions.” Though he explores some relationships with the introduc-
tory theme, he misses the most important point: the theme's melodic
A>F#<G>D (measures 4 and 5) is transformed into harmonic D-F#-A-C
toG-B-D (measures 41 and 42). This is no “patch-in” at all, but instead a
conscientious filling-out of the plan articulated at the outset!

The exposition’s submediant region (pp. 110-112)

Taruskin proposes a popular song (his example 34[2]-16B) as a “possible

source” for the theme (a good point), but he makes no mention of the
compelling case of derivation from the introduction’s theme [7.5], nor
does he correlate the theme’s harmonization (analyzed in 7.6) with that
of the C Minor/Major region that occurs later (measures 63 through 76,
displayed as the second “measure” of 7.7), preferring instead to interpret
the latter as falling within “a series of attempted, then frustrated, circles
of fifths.” Of particular concern is his assertion of an “implied cycle of
major thirds” comprising B Minor (the tonic region of the movement),
G Major (the region that begins in measure 42), and Eb Major (the chord
of measure 67), claiming this to be a Schubert “specialty. (He is referring
to progressions such as that displayed in 3.10.) In 7.8 we view another
Schubert specialty: the omnibus, which he often deforms, as here. Bass
Eb cannot be both a passing note within a C<E4 expanse (measures 63
through 71) and the goal of a cycle of descending major thirds (B>G>E}).
In addition, Taruskin claims that the diminished seventh chord of
measure 71 “resolves” two measures later. In my view, C#-E-G-Bb to
A-C#-E-G is no resolution, but merely a shift within the prolongation
of a single harmony: from one to another variant of VI# (in the key of
C). (His example 34[2]-9B, to which he directs his readers at this point,
shows a resolution in which the dominant’s seventh in E} Major leads not
downwards to the tonic chord’s third, a universal prescription of voice
leading during Schubert's lifetime, but instead upwards, with diminished
to perfect fifth: A to Fe) The resolution occurs not two, but three measures
later, on the D minor chord in measure 74 (the root succession being
A-D). Taruskin seems to have no patience for the careful analysis of
 Symphony in B Minor (“Unfinished”) ee,

such large numbers of chords as occur in this region. (Instead he informs

readers that “a little development section” is in progress.) He makes no
mention of the articulation points on G Major at measures 53, 80, 89,
and 93, nor of the similarities among the progressions that occur between
those points (as I do on p. 165, above).

The exposition’s dominant region (p. 110)

The passage that I interpret as the harmonic goal of the exposition — the
arrival on dominant F# major — Taruskin brands as a “quickie transi-
tion.” I propose that the seventh through tenth pitches of the introduc-
tion's melody are the basis for measures 106 through 109 [7.9]. Though
Taruskin describes the introduction as an “embellished plunge from tonic
to dominant,’ he does not correlate that trajectory with the structure that
unfolds over the course of the entire exposition. Though he acknowledges a
“long-held unison” on B, he does not probe the context in which that pitch
functions: the simultaneity Ft in measure 107 is a dissonant fourth, which
resolves into the dominant’s consonant a third two measures later. As he
notes, the G submediant region is curiously mammoth compared with the
opening B tonic region, and likewise, I would add, compared with the con-
cluding F# dominant region. The exposition nevertheless pursues a trajec-
tory from tonic to dominant, as did the introduction’s theme. All ten of the
introduction’s pitches are prominently articulated within the exposition,
with an intriguing permutation at the end:

m. 14 16 17 38 41 4] 42 108 109 107

BonCiel
Dab. Ade riiGroDaben@to Ure

The development (pp. 112-113)

The perspective on glides, sequences, and circular progressions that I advo-

cate (presented in chapter 2, above) offers a freedom that many analysts are
not prepared to accept: chords that in some contexts evoke specific har-
monic functions may in other contexts keep those potentialities unrealized.
My notion of assertion acknowledges that those capacities may be either
active or in remission. When in remission, a chord’s role may be starkly
different from what transpires when its harmonic tendency is asserted. In
my practice, assigning a Roman numeral to a chord signifies a presumption
that its harmonic tendency is being asserted. If a chord ultimately behaves
in a different way (for example, within a glide or sequential progression),
178 Harmony in Schubert

the deployment of such a label does a disservice, for it makes the chordal
flow seem aberrational.
Taruskin’s commentary on the development is loaded with words that
evoke aberration: “unconventionally handled, “deceptive,” “series of feints,”
“unconventional placement and conduct, “one of the most violent deceptive
cadences Schubert ever attempted,” “thwarted, “preempted,” “diversionary
ploy? “unwanted cadence,” “frustrated, “forestalled,’ “unconventional,”
“redirects” Granted, Schubert tantalizes listeners with some potentiali-
ties that do not materialize. What Taruskin misses is that other initiatives,
which place many of these allegedly aberrational incidents within a rational
framework, are operative. Certainly Taruskin would be horrified by my
models in 7.13, for it would seem to him that they neutralize the wild whirl
that Schubert hurls upon us. I counter that only because such a coherent
structure guides the progression does Schubert have the freedom to taunt
us at the surface level.

The recapitulation and coda (p. 113)

Taruskin correctly notes that the first theme deviates from its exposition
model, with a cadence on the dominant rather than on the tonic. One
sentence accounts for the remainder of the recapitulation: “But however
unconventional this [the first theme's deviation] may appear, it is done for
the sake of conformity to convention, for it enables the recapitulation of the
second theme in the customary mediant key (D major).’ This statement is
misleading in several ways. First, it is highly unconventional for the second
theme to be presented in the mediant key during a recapitulation. Second,
the tonal purpose of a mediant region within a minor-key exposition is
generally to fill in an ascending sweep from the tonic to the dominant. Here
the mediant is part of a descending tonal trajectory. (See 7.14, Model 2.)
Concerning the coda, Taruskin’s main interest is to correlate the “nostalgic
nods at the subdominant” with the emphasis on E Minor during the develop-
ment. I suggest instead that the E and G in measures 358-359 and 362-363
are neighboring embellishments of the tonic third and fifth, and that they
derive not from the development, but instead from the very outset of the
exposition. (Compare with 7.2, measure 10. In the coda Schubert suppresses
the invigorating ae third that coordinates with° in the earlier passage.)
The gulf between Taruskin’s and my conceptions of this movement could
hardly be wider. In principle I am grateful that his project, which he calls a
“history” of Western music, is so deeply steeped in analytical inquiries. Yet
in this case J cannot endorse the outcome.
8 Piano Sonata in A Minor (D. 784), movement 2
with Robert S. Hatten

Robert S. Hatten has long been devoted to exploring strategies for teasing
out layers of musical meaning that are often neglected in traditional ana-
lytical approaches.’ His Interpreting Musical Gestures, Topics, and Tropes:
Mozart, Beethoven, Schubert (2004) offers a wide range of strategies
and insights, including some that are presented in the context of music
by Schubert. I wholeheartedly endorse an intention of his book that he
expresses in the following words:

Furthermore, although musical gestures are often made distinctive through spe-
cific articulations, dynamics, and pacing or timing — and given unique shape by
the systematic potential of rhythm and meter, texture, and timbre - they cannot be
fully described without reference to the more “syntactic” levels of music structure
and process. These syntactic levels are shaped by the overlapping disciplines of
counterpoint, harmony and voice leading, phrase structure and form, and, in many
cases, motivic developing variation. (p. 2)

Given the focus of my book, I am especially interested in exploring how

Hatten fulfills his intention to integrate his insights on meaning with the
syntactic parameters that he lists. The second movement of Schubert's
Sonata in A Minor (D. 784) will serve as the venue for this assessment.
Hatten’s analysis of the movement is brief - fewer than 1,000 words, plus
a few markings in a score excerpt. My analysis (including the endnotes)
dwarfs his in bulk — nearly 5,000 words, plus an array of music examples.
Consequently one may wonder just how detailed an analysis this move-
ment warrants. Are there insights Hatten has omitted that one ought to
know in order to interpret this movement successfully? Or am I wasting
paper, ink, and my readers’ time?

The opening tonic theme (measures 1-18)

The D of measure 1 is an extraordinary pitch. It is an embellishment

that falls within a straightforward tonic arpeggiation: C<F<A<C.* The
peak C is thereby diverted to the downbeat of measure 2. There is some-
thing luxuriant about this melody. It conveys the sense that we are in no
180 Harmony in Schubert

hurry. We have the time to enjoy a deep upward stretch as we greet a

new dawn. D participates in several broader relationships as well. Note
that the chordal support is removed during the second half of measure
1. Hatten consequently regards the melody as “vulnerably exposed,’
recalling a “tragic gesture” from the first movement (p. 194). I impute a
more specific structural implication to this absence of support: A<D is
independent of any explicit reference to the tonic triad. (In particular,
D does not clash with an inner-voice C.) In that light the melody of
measure 1 presents a challenge: two ascending fourths (C<F and A<D),
representing the tonic chord (F-A-C) and its 6 phase (D-F-A), are juxta-
posed. At this point the latter chord is held in check. The story changes
in measures 5 and 6, where chordal support reinforces the 6 phase and
where the unfolded A<D fourth is followed by D<F<A, thereby com-
pleting a full-octave arpeggiation. Though it is not made manifest until
measure 6, the relationship between tonic F major and its 6 phase, D
minor, mirrors the broad relationship between the sonatas A Minor
home key and its 6 phase, F Major, during the first two movements.
Finally, note that the D>C neighboring motion of measures 1 and 2 is
contorted into C<D}>C during measure 4. This gesture will recur regu-
larly.* The life-embracing opening phrase must coexist with such less
salutary undercurrents. Is it Satan, as Hatten suggests?? Perhaps. Or, it
could be an acknowledgment by the composer that even when basking
in the radiance of good feelings, the prospect of recurrent ill health
and early death lurks just beneath the surface.° The whispered sordini
gesture may represent the beckoning hand of Death, not only cutting
short the cadential tonic chord, but also threatening to bring an early
end to the composer’s life.’
The internal organization of the opening phrase is elusive. Though
there are only five chords, three hypotheses concerning their hierarchical
relationship should be considered:

Bass: | ee itl 1) Ral tell

Elypothesis: (>) Dy) F
Hypothesis 2: F E F
Hypothesis 3: F° vy Ey AE

Hypothesis 1 emphasizes the symmetrical down-and-up motion of the

bass, which might be interpreted as a projection of [5-65 (with 6-phase
chord unfurled). In my view the second E is undervalued in this reading.
As the theme continues the corresponding cadential progressions are
 Piano Sonata in A Minor 181

7 8 9 12 1314 15 16 17 18

SS —SSS=
Sot ae
F Major: I V1

8.1 Analysis of Piano Sonata in A Minor (D. 784), mvmt. 2, mm. 1-18.

more fully developed, with a root-position dominant chord occurring

in measures 7, 14, and 17. Demotion of the first phrase’s latter E to a
mere passing phenomenon would create a curious discontinuity among
the phrases. Hypothesis 2 embraces bass E in its initial statement, with
D serving as a lower neighbor. That reading contrasts Hypothesis 3, in
which the first E instead passes to D. My preference for the latter reading
is based in part on Schubert's use of rests in measures 1 and 2. These
silences in the chordal accompaniment focus attention on the melodic
pitches A and G, which sound in unexpected purity. Those pitches lead
by step to F at the downbeat of measure 3. Thus parallel tenths prevail
in the outer voices over three measures: #2 2 [). In this interpretation,
tonic’s A-to-F melodic third is shadowed below, which in turn calls to
mind another passage by Schubert, from the Piano Sonata in D Major
(D. 850), movement 3 [2.21, measures 126 and 127]. The clarity of the
motion in that context helps one come to terms with the opening meas-
ures of the F Major Andante.
Pondering such subtle hints may help one in developing a coherent
interpretation of a passage. Without this careful consideration, a performer
might simply play the pitches in the metrical positions indicated and hope
for the best. With it, a sense of conviction emerges: measure 2 is part of a
downward trajectory that is transcended during the second half of measure
3. That conviction informs my graph of the theme [8.1]. Note how often the
A>G>F third is traversed before the cadence in measure 18.°
The Db neighbor in measure 4 resides within a more fundamental
C>B4<C interjection that embellishes 5 in a manner corresponding to the
E>D#<E embellishment of A Minor’s 5 during the opening measures of
the first movement. This motion is further fleshed out when the opening
182 Harmony in Schubert

m. (7) (8)

hice

ow ==2 E
Waal

wae =|
F Major: ees ili V

8.2 Alternative content for Piano Sonata in A Minor (D. 784), mvmt. 2, mm. 7-8.

theme recurs in the dominant key starting in measure 31. A more robust
interjection derived from that of measure 4 occurs in measure 34, where
not only #4 (in C Major), but also #3 and 6 embellish members of the C
chord, forming a so-called “common-tone diminished seventh” chord. This
fuller rendition is deployed again once tonic F is restored, in measures 45
and 48. The fast-note upper neighboring motion (A<B}b>A) is diatonic. A
local C<D}>C (as in measure 4) would have been odd in the vicinity of the
broader D neighbor in the tenor register. Schubert's concurrent Bb upper
neighbor of A and Bs lower neighbor of C is one of those magical moments
that ought not to work but somehow does.
The openings of the first two phrases differ in that D sounds briefly as a
pitch in measure 1, whereas a full-fledged D chord emerges during measures
5 and 6. The implications of this chord will be worked out as the phrase
continues. Two potential trajectories come to mind: that D is a temporary
displacement of F's fifth, C; or, that D will propel the progression further
from tonic F, leading perhaps to II or to IV, both of which incorporate the
D. Schubert's creativity in exploring this situation draws upon the fact that
F and D form a minor third, and thus this context is ideally suited to draw
upon the diminished seventh chord’s mehrdeutig potential. Spelled as C#-E-
G-Bb, that chord fosters a prolongation of tonic’s 6-phase D chord. Spelled
as E-G-Bb-Db, it fosters a restoration of the tonic F chord. The powerful
A-C#-E-G-B} chord in measures 6 and 7 thus contains a dual potential.
It might be interpreted as C#-E-G-B} above root A, a chord that points
towards D. (How that potential might be realized is explored in 8.2, with
a continuation to IIk and V.) Or it might be interpreted as E-G-Bb-Db[C#]
pointing towards F over pedal A, the F tonic chord’s third. (Compare with
1.1, measures 61 and 62. This quandary will resurface in the Andante’s
coda, where Schubert extends it further. Readers might wish to view 8.10,
Model 1, at this point.) Because both F and D chords have preceded this
 Piano Sonata in A Minor 183

m 9 10 i] 12

\ be
—————=————
s Ss Ze a, i ~— a (oe)

é Sa PSS
eeaell ican |
ed
= ———+*
ee rv ——
S 4] = zt D 7
ees y,

y) eee
ee ee
: 6 z
F Major: V I, 3

8.3 Analysis of Piano Sonata in A Minor (D. 784), mvmt. 2, mm. 9-12.

dissonant chord, and because a D chord may either embellish an F chord or

lead away from it, listeners ought to keep both continuation routes in mind.
(Therefore conventional analyses such as V? in D Minor and V?%/vi in F
Major, which adhere to Schubert's chordal spelling using C# rather than Db,
seem to me too prescriptive in this context.)? His resolution in measure 7
is in fact to an inverted F chord, preceding a perfect authentic cadence in F
Major. (See 8.1.) Hatten does not address this chord in its tonic context, but
does comment on it during this passage’s restatement in the dominant key,
where it is presented as E-B-D-F-A> in measures 36 and 37.° He observ-
antly comments that “a dissonant B diminished-seventh chord over E pulls
us back to the realm of C major just in time for the cadence” (p. 198).
The theme's longest uninterrupted segment occurs during measures
9 through 15. It consists of two distinct phases. The first is an expanded
restatement of the second chord of measure 3 and the first chord of measure
4. Those chords reside at the left and right edges of 8.3. In between is a §
embellishment of the tonic 3,which is a prominent feature of the first move-
ment (starting in measures 9 and 10).’’ During a parenthetical passage
these pitches unfold upwards, as shown in 8.3, which expands upon the
broader view of 8.1. The further layer of neighboring embellishment that
prolongs both of the parenthetical chords is not displayed in 8.3: accented
neighboring chords resembling the one that occurred in measure 7 occupy
the downbeats of both measure 11 and measure 12. Once the root-position
tonic chord arrives, the passage’s second phase begins, closely following
the model of measures 1 through 4. The close on soprano 3 in measure 15
motivates another attempt at full closure, which Schubert undertakes, an
octave higher, as the theme's final segment. A perfect authentic cadence is
achieved in measure 18, as shown in 8.1.
184 Harmony in Schubert

m. 19 20 21 23

hal ( |

adie —D>b —Bb —>Gb

F? ( ps a6

8.4 Analysis of Piano Sonata in A Minor (D. 784), mvmt. 2, mm. 19-23.

m. 21 22 23 24 25

a
a ee a ae
ass

cama a
ipo—t + iS ——

ar EO! gaelGy

8.5 Analysis of Piano Sonata in A Minor (D. 784), mvmt. 2, mm. 21-25.

Transition to the dominant (measures 19-30)

Though this movement is not in a traditional sonata form with exposition,

development, and recapitulation, the transition section that ensues during
measures 19 through 30 performs the same role as that which often occurs
during the exposition of a major-key sonata: it leads from the tonic (F
Major) to a medial caesura on IIs (the G major chord of measures 29 and
30), announcing the impending arrival of a tonicized dominant region (C
Major, measures 31ff.). Such transitions often are the venue for extraor-
dinary compositional creativity, and here Schubert does not disappoint.
Hatten describes the mood as “grandiose,” “heroic,” “powerful,” “tragic” (pp.
197-198).
The transfer of the Death motive from embellishment of the tonic chord’s
fifth to its root!? in measures 18 and 19 facilitates its truncated continu-
ation as embellishment of a series of chordal roots descending in thirds:
F>D}>Bb>Gb in measures 19 through 23. This contour is displayed, with
considerable imaginative filling-in of the internal chords, in 8.4.'° A hairpin
symbol indicates that other content comes between the chords of measures
21 and 23. This content, a sequential progression that resides within the
broader progression of measures 19 through 23, is shown in 8.5. Once the Gb
chord arrives, it is confirmed via a succinct restatement of the D)<Gb succes-
sion during measure 25 (also displayed in 8.5). Hatten’s and my readings are
 Piano Sonata in A Minor 185

m. 19 21 23 26 x

6X areas
ag Ba

8.6 Hypothetical structural model for Piano Sonata in A Minor (D. 784), mvmt. 2,
mm. 19-26.

Model 1 Model 2

= ee,
|
ey
Model 3 Model 4

m. LOD 246 28 29

$2275, tapaemicn
2

Sa

F Major: I ( )1

8.7 Analysis of Piano Sonata in A Minor (D. 784), mvmt. 2, mm. 19-29.

in conflict with regard to the relationship among the chords of this region.
His analysis of the descending sequence is “I-V-vi-iii-IV-I” (p. 197), which
suggests that he would want to extend the bass slur connecting D} and G} in
8.5 to the F of measure 25, with G} arriving later in measure 25. I propose
instead that G} arrives in measure 23, with a prolongation into measure 25.
Given the transition section’s broad F-to-G trajectory, the Gb arrival in
measure 23 is a suitable interior moment, dividing the whole-step path
into two half steps. However, the fact that G4 is the imminent goal is not
yet apparent. Thus Schubert was able to project two distinct potential tra-
jectories concurrently. One of these paths, an idiosyncratic sequence with
5-phase chords on F, Gb, and Ab (leading quite far afield from F Major), is
displayed in 8.6. An alternative path is displayed in 8.7, Model 1: an obsti-
nate sequence ascending in half steps (F<G}<G) and incorporating the
enharmonic reinterpretation that is inevitable in such contexts.'* Schubert's
creative energies are focused on its fourth chord. Three additional models
document my reading of his strategy. In Model 2, a common heightening
of intensity occurs: D-F#-A is transformed into F#-A-C-Eb. In Model 3,
a neighboring diminished seventh chord precedes the F#-A-C-Eb chord.
At this point in the score, Schubert seems to be pondering which route to
take. He wavers between F} and Eb in the soprano of measure 26. If Eb is
accepted as the successor of F), then the chord of measure 26 corresponds
186 Harmony in Schubert

to the fourth chord of 8.6, and the trajectory of 8.7 would need to be aban-
doned. However, Fb [Es] rises again at the downbeat of measure 27. Yet
even now the half-step descent into F#-A-C-E}, proposed in Model 3, does
not occur, That resolution is achieved by instead ascending two half steps,
as shown in Model 4. (Since the diminished seventh chord is made up of
minor thirds, descending one half step and ascending two half steps lead to
the same chord.) In the score this ascending trajectory is pursued against
a pedal G. Consequently the sequence’s last two chords collide. (Compare
with the chordal collisions in 1.1, measure 61, and 1.9, measure 14.) The
upward course not only intensifies the resolution (as do the forzando mark-
ings and the crescendo), but also leads to the goal soprano pitch G, succes-
sor of the first theme’s A (which is displayed in 8.1 as the Kopfton 3).

Thematic statement beginning in the dominant key (measures

31-48)

Though there are exceptions, an opening theme in tonic F Major and a

transition ending on IIs generally will be followed by a new theme in the
dominant key, C Major. Despite its more ornate accompaniment, the theme
starting in measure 31 is clearly the F Major theme again, transposed into
C Major. Yet Schubert is not done surprising us: this theme migrates back
into F Major even as it is being presented. Observe in 8.1 that the theme is
built with some redundancy - a potent A~G-F descent occurs not only at
its end, but also at its middle, in measures 6 through 8. This earlier A-G-F
corresponds to the E-D-C displayed in 8.8, measures 31 through 38, com-
pleting a descending fifth-progression from the G of measure 29. Rather
than reiterating the descent to C during the theme's remaining measures,
Schubert instead juxtaposes this C goal chord and the C’ chord of measure
9, which in measure 39 induces the move back into F Major. The remainder
of the theme closely follows the precedent shown in 8.1, though now the
concluding A-G-F third (measures 42 through 48) hooks up with the back-
ground structure at its F rather than its A, due to the prior prolongation of
background G, where the descent was interrupted.!°
Consequently the movement's form is highly unusual. It seems at first
that a four-part form is being pursued:!°

RO ea res
ONE
(cScon nn rencP ell OM ain
Yet what transpires once the B, section is under way amounts to a merger of
the B, and B, sections (the first half of B, followed by the second half of iBeyls
 Piano Sonata in A Minor 187

m 6 17 18 29 fh. = Sil Sime Sa eA AS

3 5| 2544

ar te = ==

Ss
= 2 fe {

pt
rec Lela
F Major; 9. ——— qi —/7 y8——7_ 7
8.8 Analysis of Piano Sonata in A Minor (D. 784), mvmt. 2.

with all the formal components in between eliminated. (Schubert might

have called upon the A theme to serve also as the B theme so that measures
31 through 48 fulfill some of the typical characteristics of the B,, A,, and B,
sections!) Though 8.8 appears very conventional as a structure, the interface
between the thematic statements and this structure is exceptional, in that
one statement of the theme (measures 31 through 48) encompasses both
the interrupted V and the restored I.
Hatten’s main interest within this region is twofold: (1) the further
development of the “serpent” motive (which I have addressed as part of my
commentary on the tonic region, above); and (2) the E-B-D-F-A} chord of
measures 36 and 37 (which might have been addressed during his commen-
tary on the tonic region, since this chord first appears in measures 6 and 7).
He is curiously silent about the extraordinary fact that a theme that we first
heard in the context of a single key (tonic) now is heard bridging two keys
(dominant and tonic). His commentary downplays the event: “modulatory
sequencing... returns us to F major” (p. 198). Linstead perceive the theme
as consisting of two chunks, both of which prolong tonic during the first
statement. During the second statement the first chunk appears transposed
into the dominant key. Because the second chunk begins with a V’ chord
(measures 9 and 39) there is no need for an elaborate retransition into the
tonic key. Measure 9/39 may follow either a tonic chord (measure 8) or a
dominant chord (measure 38).
This breathtaking deformation of the form invites some speculation
concerning what meaning Schubert might have intended to convey. During
measures 31 through 38 listeners well versed in the conventions of musical
form would have good reason to assume that the movement is less than
halfway finished. Yet at measure 48 the structure is complete, with only a
coda to follow. Before Schubert contracted syphilis he likely assumed that
188 Harmony in Schubert

—D —>F —>Ab
8.9 Three enharmonically equivalent diminished seventh chords.

his life was less than halfway finished. (Beethoven was over twice his age,
for example.) Could the unconventional curtailment of this movement
represent the outcome he feared for his own life, once that fateful diagnosis
was hurled upon him? The “tragic” mood that Hatten notes emanates not
from words, but from combinations and successions of pitches. Particularly
when pitch relations deviate from their conventional formulations, one
may find structural reasons that confirm the intuitive sense one has about
what the music is evoking.

Coda (measures 49-66)

Concerning the coda, Hatten suggests that “there are further expressive
harmonic digressions” and even proposes that “the movement has appar-
ently lost its way” (p. 198).'7 I seek greater precision in my explanations of
harmonic practice. From my perspective Schubert is masterfully and con-
currently engaging with two of the nineteenth century’s predominant com-
positional themes: enharmonically equivalent diminished seventh chords,
and lower- and upper-third chords.
Three different spellings for the same set of four pitch classes are shown
in 8.9. Generally a diminished seventh chord consists of the third, fifth,
seventh, and ninth of a chord whose root is absent. (Compare with 1.2,
Chord 7.) If one accepts the given spellings, these three chords would be
understood as derived from roots A, C, and E}, respectively. Yet in their
compositional usage, a mehrdeutig potential is tapped when such a chord
functions in a way that contradicts its spelling. (Of course, listeners are
not cognizant of the composer's spelling. Yet assumptions often can be
made based upon how one approaches such a chord — assumptions that
are upset when the chord resolves as if spelled another way.) Adding a fifth
pitch (such as A below C#-E-G-B}) generally will impose a tonal focus that
negates the mehrdeutig potential. Schubert has discovered an ingenious
loophole that defies that assumption. In 8.10, Model 1, we see how placing
an A below C#-E-G-B} may confirm its tendency to resolve to D. However, a
second potential resolution is displayed as well, one in which the resolution
proceeds instead to F. This is the ruse that Schubert employed in measures
 Piano Sonata in A Minor

Model 1 Model 2

=+2) eS alike.
ee
e s Fae
=e rae Sh —» Ab

8.10 Mehrdeutigkeit between pairs of enharmonically equivalent chords.

6 and 7 of the Andante, discussed above. Because there it was preceded by

a D minor chord, the spelling A-~C#-E-G-Bp suggests a tonicizing progres-
sion in D Minor (as I-V?- . . ., which we might expect will lead back to I).
Yet Schubert treats the chord as if it were spelled as A-E-G-B)-D}, as in the
second half of Model 1. This relationship is projected during the opening
theme, and later is repeated in the context of the dominant (measures 36
and 37), confirming that even a five-note chord can possess mehrdeutig
potential.
In the coda Schubert goes further. Model 2 of 8.10 proposes another
mehrdeutig relationship between five-note chords, constructed by trans-
posing Model 1 up a minor third. Now tonic F and one of its chromatic
upper-third chords are potential resolutions of the same five-note chord.
The interactions between tonic and its lower- and upper-third chords, as
expressed in these two models, are at the heart of Schubert’s extraordinary
writing during the coda. The chord of measure 52, like that of measures 6
and 7, proposes to lead into the domain of the submediant (according to
the first alternative displayed in Model 1). It turns out that F wins out after
all. Then the passage from measures 56 through 58 proposes to lead into
the domain of the lowered mediant. (In this case the chord of measure 57
corresponds to the second alternative displayed in Model 2.) Just before
that resolution, however, the progression comes to a standstill (beat 1 of
measure 59). A mehrdeutig diminished seventh sonority — shared by chords
that resolve to D, to E, and to A> — sheepishly follows. Then the content of
measures 52 and 53 is repeated, leading ultimately (after the fermata separ-
ating measures 61 and 62) to tonic FE. The final five measures present tonic
without further impediments, recalling the opening theme and extending
the final tonic chord by an extra measure.
If the hand of Death indeed hovers over the movement and if the
abnormally early formal closure in measure 48 corresponds to Schubert's
premonition of an early demise, then the coda may represent his attempts
at a reprieve. Is there a way to reinvigorate the tonal motion? Can the
inevitable tonic closure — inevitable demise — be held off for a while? Tonic’s
lower- and upper-third chords are among Schubert's principal means of
190 Harmony in Schubert

launching a progression away from tonic. Countering Hatten’s suggestion

that “the movement has apparently lost its way,” I propose that Schubert is
pursuing, in alternation, two routes by which he might — through enhar-
monic trickery - dupe Death, who beckons urgently in measures 49-50
and 54-55. Despite Schubert's persistence, the attempts are not successful.
Tonic is confirmed, and the movement closes with a death knell.'®
9 “Die junge Nonne” (D. 828)
with David Kopp

David Kopp has developed new strategies for music analysis that focus
especially on relationships between adjacent chords. His reductions of
a compositions chordal progressions, shown in music notation or with
alphabet letters indicating chordal roots, are annotated using letters such
as D (for a dominant relationship of adescending perfect fifth) and M (for
a mediant relationship of a descending major third) between adjacent (or
sometimes non-adjacent) roots. (The symbol ~! reverses the direction: D"!
indicates an ascending perfect fifth, thus the subdominant relationship.)
Though his music examples lack Roman numerals, his written commentar-
ies incorporate scale-step thinking, including tonicization.
Kopp cites over a dozen compositions by Schubert in his book, which
concludes with an analysis of Schubert’s “Die junge Nonne,’ a work that he
chose in part to demonstrate how he handles third relations.’ In my com-
mentary below I will compare the insights that he and I achieve through
our contrasting perspectives, but I will not attempt a thorough critique of
his analytical methodology as a whole. (I admit from the outset, however,
that I find the absence of Roman numerals in his music examples an unnec-
essary hindrance to comprehension. I often inserted them as I read his
discussion of a passage, without harm to his perspective.)

Stanzas 1 and 2

The first stanza of J. N. Craigher de Jachelutta’s “Die junge Nonne” evokes

a powerful and fearsome storm. Schubert employs a range of potent chro-
matic and dissonant chords in setting it. The less intense yet still menacing
introduction conveys to the listener that a storm is brewing and establishes
some foundational features of the upcoming harmonic progression. It is
shaped as an antecedent/consequent construction:
m. 1 3 5 6 7 8

F Minor: I Vz Pe ant ee Yee

The diatonic supertonic seventh chord, distinctive in its half-diminished

quality, supports the one melodic shift (to D>) that enlivens an otherwise
OD Harmony in Schubert

19-21 22-23 24-25 26 Di 28

m. 8-10 11-14 15-18

Eee ae ata ° 29. z 2 2

9 62-5
F Minor: I ( ) ee ee

9.1 Analysis of “Die junge Nonne” (D. 828), mm. 8-28.

unwavering soprano C, which in the context of this lied clearly represents

the tolling of a church bell. Variants of the supertonic (supporting both
Ds and its diatonic restoration to Db) are a conspicuous feature of the first
stanza’s progression (measures 22-25), as is the unfurling of the follow-
ing dominant’s ° embellishment (bass A} in measures 26, reprising that of
measure 7).
The first stanza’s progression develops slowly. Initially each two-measure
vocal statement accomplishes a minute chordal shift, coinciding with
textual moments of peak descriptive intensity - heulende Sturm (howling
storm), zittert das Haus (the house trembles), leuchtet der Blitz (the light-
ning flashes). Each shift is confirmed by the keyboard in a two-measure
response. Cumulatively these shifts establish a 5-6 sequence ascending
in half steps [9.1]. The obstinate character of the sequence makes attain-
ment of the diatonic II as the third 5-phase chord impossible, since its D}
has been passed through already. However, the variant of II that arrives in
measure 22 is more flamboyant than expected: shunning G-Bb-Ds (shown
in parentheses), Schubert employs B4-D4-F-Ab (with a suspension delay-
ing the F and with Ab spelled as G#). Its role as an evolved II is not clear
upon its initial sounding. In fact, it may seem to tonicize the F# chord of
measures 15 through 18 (as its evolved dominant, E#-G#-Bk-D). The low-
ering of all four pitches by a half step for the chord of measures 24 and 25
adds further consternation. Could this be the beginning of a descending
glide of diminished seventh chords, akin to 2.2? This time, no. By the end of
measure 25, where three of the voices ascend back up a half step, the chord
focuses more securely as an entity that we expect will resolve to dominant
C, which duly arrives in measure 26 and leads to an imperfect authentic
cadence at the stanza’s close. (See 9.1.)
 “Die junge Nonne” 193

Though Kopp’ interpretation of this passage is closely attuned to the

text and to the evocative character of the chords and their relationships,
as a harmonic analysis it does not cohere. Of all the chords that transpire
between the tonic of measures 8 through 10 and the dominant of meas-
ures 26 and 27, the most significant to the harmonic progression is that
supported by bass D) at the end of measure 25. (Many analysts call it the
“German augmented sixth chord,’ and under any label it is one of music’s
most potent heralds of the dominant.) There is no mention of this chord in
Kopps commentary. Though he carefully navigates the chord progression
of measures 8 through 21 (which contains two instances of third-relations,
his central concern), what happens next is left in question. He relates that
the chord of measures 19 through 21 resolves “irregularly” (p. 255), and he
acknowledges “a progression of two diminished-seventh chords” without
stating that three of the four pitches that descend bounce back immediately
(p. 256). A coherence is asserted in 9.1, achieved through the expansion
of the introduction’s I-II’-V4-I. Though his analysis does not deploy the
conventional harmonic labels, which would display the chords of measures
7, 22, and the last beat of 25 in woefully disparate manners (as ii’, vii°’/V,
and Ger*®), Kopp has formulated no alternative means of relating similar-
sounding and same-functioning chords. In contrast, my perspective fosters
an awareness of the various evolved states of a single harmonic function.
The family of II chords in F Minor includes G-Bb-Db-F (measure 7), Bk-D4-
F-Ab (measure 22), and Bk-D}-F-Ab (measure 25). In “Die junge Nonne,’
these chords all support scale degree 6, either diatonic Db or chromaticized
Ds.
The first five chords in 9.1 (through the parenthetical G minor chord)
constitute a conventional variant of the diatonic ascending 5-6 sequence.
I characterize it as obstinate because it proceeds in uniform half steps.’
(At this point in the poem, the weather's behavior is obstinate as well.)
Schubert’s rendition here is breathtaking because it proceeds at an exceed-
ingly slow pace. (The pace will quicken in the second stanza.) The 5- and
6-phase chords nevertheless emerge one after the other consistently until
the II chord arrives, at which point the sequential initiative concludes.
Kopp assumes that a scale-step analyst will apply Roman numerals during
the interior of such a progression. In his demonstration of how such an
analysis might appear, he includes a pivot-chord modulation (VI in F
Minor = V in F# Minor) for a chord that I regard as dominant-emulating
but not as a dominant. I argue instead that harmonic and sequential pro-
gressions have distinct characters and, in particular, different motivations
for succession from one chord to the next, and that the use of Roman
194 Harmony in Schubert

numerals is apt only in a harmonic progression. Kopp’s harmonic analysis

of the sequence leads him to comment on the “strong harmonic disjunc-
tion” (p. 254) between the “distantly related keys” F Minor and F# Minor
(p. 255). Acknowledging that Schubert may take us through some uncom-
mon chordal territory during a sequence, I instead place a harmonic focus
only on the trajectory’s endpoints. In these measures the F% minor chord is
passed through without being asserted as a harmony, and certainly not as a
key. The more pressing analytical concern is what to make of the chord on
which the sequence terminates (measure 22). A supertonic function is only
gradually confirmed. (Kopp offers no interpretation of that chord.) Finally,
due to the inherent bias of music’s notational system in favor of a key’s
seven diatonic pitch classes, enharmonic respellings are inevitable during
a 5-6 sequence that ascends in half steps. It makes no difference whether
what comes between F and G is spelled as F# or G}. (In fact, without enhar-
monic respelling, it would proceed as F*® Gb*° Abb.) Thus Kopp’ point
that the progression from “flat” F minor/D} major to “sharp” Ff minor/D
major “reflects” the text seems to me tenuous, despite his note 11 (p. 256).
I suggest that this sequence is conceived and executed as a means of con-
necting F minor and G minor chords (both on the “flat” side), developing
the introduction’s I-II succession (measures 5 and 6). Only at measure 22,
where a diminished seventh chord supports the word finster (dark), is that
initiatives premise violated. Though eventually we come to understand that
this chord is an evolved II, at the moment of its sounding we do not know
what to make of it.
The poet compares the storm imagery and the young nun’ life before her
Savior had instilled peace. Thus in her recounting of that earlier life during
the second stanza Schubert redeploys the first stanza’s harmonic progres-
sion, though now without the keyboard interludes during the sequen-
tial ascent. Only the stanza’s first line (measures 31 through 35), which
announces the shift from observation of the external world to a relating of
the nun’s former emotional state, departs from this structure. Those meas-
ures are closely allied with the introduction:

+ the tonic prolongation in measures 30 through 33 extends that of

measures 1, 2, and 5 (with a temporary F major inflection, § neigh-
boring embellishment, and the octave arpeggiation in the bass now
descending)
¢ the diatonic II’ of measure 34 (now inverted) is like that of measure 6
¢ the dominant of measure 34 (now with minor seventh) is like that of
measure 7
 “Die junge Nonne” 195

F Major: I

9.2 Analysis of “Die junge Nonne” (D. 828), mm. 52-61.

Stanza 3

A pronounced shift occurs in the text starting with the third stanza.
Whereas the first two stanzas are filled with foreboding (both the outdoor
environment is and the nun’s soul was in torment), a sense of rapture is
conveyed beginning in the third. Schubert supports that shift through a
conversion to the parallel major key, F Major, and through greater emphasis
on the singer’s heavenward upper register. (Except for measure 32, during a
passage that anticipates the shift to F Major, the melody has been bounded
by 6 up to this point. 7 and 8 occur frequently in the remainder of the work,
most notably in the cadential measures 60-61 and 82-83.)
Whereas the supertonic was the principal harmonic connection between
tonic and dominant in the earlier stanzas, [V assumes that role in the third
stanza [9.2]. As is often the case, dominant emulation assists during the
I-IV succession, and a 5-6 shift eases the path between IV and V. Vocal
ascents to 8 emanate from both this model’s 5 and 3.
Model 1 of 9.3 shows how a 5<8 connection coordinates with the initial
tonic’s prolongation. The arrival of goal F in the upper register coincides
with the addition of minor seventh E), triggering the succession to IV.
Despite the radiance of this high F and its prolongation into the IV-
space, the initial C (5) nevertheless connects with incomplete neighbor
D in measure 59. Model 2 provides a more detailed view of the ascending
motion from C to F, along with its chordal support. Though the local V of
measures 53 through 55 could have proceeded directly to I, Schubert calls
upon tonic’s upper-third chord to initiate I-space. This passage is virtu-
ally identical to the V-I succession (measures 159-165) in the Trio from
the Piano Sonata in D Major - just twenty-two Deutsch numbers later in
196 Harmony in Schubert

Model 1 Model 2
m 52 58) Sy Ss) 56 58
m.. “52 58 59

5 IN 5
aa ee
na? Tdi ae:
A 7 - y: 3 F es a

Oo oy * te
e Ser SIO

LS
4 DE= &
Se
12k Z o

[ee, f eee [
| | eee

; 7b
F Major: i. F Major: jig:
8 is We) =,
. (=I V bie :

9.3. Analysis of “Die junge Nonne” (D. 828), mm. 52-59.

Schubert’s oeuvre - that we explored in chapter 2 [2.21]. In both cases a

wobble occurs during the tonic’s upper-third chord.
A second ascent to 8 transpires during the cadential measures, 60 and 61,
coordinating with the words [der] ewigen Liebe getraut (wedded to eternal
love). Here the keyboard part dutifully completes the structural descent
(3-2-1) while the singer soars to 8. An astonishing aspect of this passage
is how the bass of measures 61 and 62 echoes this vocal ascent. When we
first heard a bass line of similar construction (measures 1 and 2), it evoked
menace and foreboding. Now, through God’s grace (according to the text’s
theological underpinnings), desolation has been transformed into bliss.°
How economically Schubert conveys this central conception of the poem!
Kopps analysis of this passage contains some admirable features. I
applaud his creation of broader hierarchical motions out of local chordal
connections. For example, though on one level he displays the harmonic
progression from root C in measure 53 to root F in measure 58 as D"!," D,
m, M, it is good to see the merger of those individual motions noted, via
a single D, beside the arrow above the bass staff in his fig. 9.12. I likewise
agree that the$ chords in measures 54 and 55 are rooted on C, not bass
G. Yet he fails to take the next step in hierarchical thinking: relating the
F chords of measures 52 and 58, so that the latter appears not merely as a
local applied dominant of the upcoming IV, but instead as the culmination
of a seven-measure expansion of tonic (I*-”), There is a perceptible sym-
metry here, with bass F<A<C opening up the space to the local dominant,
and bass C>A>F reverting to the tonic.‘ One might even suggest that
 “Die junge Nonne” 197

ascending passing note B} in the melody of measure 52 is acknowledged by

descending passing note B} in the bass at the end of measure 55, forming
a retrograde of the notion first sounded in the bass of measure | (there in
F Minor). That may help explain why Schubert chose to descend via thirds
rather than the more conventional fifth.
Kopp repeatedly refers to the chord on A (measures 56 and 57) as a “dom-
inant ninth” (pp. 208, 209, 257). I prefer to separate the notion of function
from that of chord quality, and thus coined the term “dominant emulation”
to describe a chord that is built like a dominant and possesses the tendency
to resolve by descending fifth, yet at a location within the key other than
V-I. Here I doubt even that intention on Schubert’s part. (It seems that
Kopp may agree, because though he calls the chord a “dominant,” he adds
that “it points to itself rather than to any tonic” (p. 258).) I likewise reject
its interpretation as a ninth chord. The vocal Bb is not a chord member,
but instead an embellishment, reiterating the Bb>A resolution of the bass
leading into measure 56. C# occurs in the context of a C<C#>C wobble,
an event that is exceedingly common during Schubert's descending-thirds
motions. Since C# does not behave as a leading tone, I refer to the chord in
a neutral manner, as one of tonic’s chromatic “upper-third” chords.
Kopp treats the cadential § chord of measure 60 (C-F-A) as rooted on
F, and thus his connection from subdominant to the following dominant
appears as D~? (two ascending perfect fifths: Bb<F<C) in his fig. 8.19.° In my
view, he misses a brief but critical maneuver that Schubert undertakes to
improve the voice leading between IV and V: a 5-6 shift. Kopp can propose
a root succession from Bb to F only by denying the importance of G, the last
pitch in the keyboard part of measure 59. The filling-in of the IV chord'’s
lower third (D>C>B)} in the vocal line and at the bottom of the treble-clef
keyboard part) during beats 3 and 4 of measure 59 coordinates with the
concurrent line Bb>A>G. I wholeheartedly accept this G as a displacement
of the IV chord’s fifth, E° Such a shift is a commonplace in Schubert’s music.
(Compare with 3.23, which also leads into IV via I*’”.) In my view C is the
only viable root for all of measure 60.

Stanza 4

Schubert incorporates a number of chords and progressions from earlier

passages of the lied in the final stanza. It is important to ponder what such
repetitions might imply, if anything, and to evaluate how these passages
function in the new context. The progression beginning at the upbeat to
measure 62 (especially the F-Ab-D} chord) is reminiscent of that which
198 Harmony in Schubert

began at the upbeat to measure 10; the chord of measures 69 and 71 reprises
that which appeared on beat 4 of measure 25; measures 73 and 74 are
derived from measures 31 and 32; and measures 80 through 83 close the
stanza in much the same way that measures 58 through 61 closed the third
stanza.
While still earthbound, the nun continues to encounter (though she is
not swayed by) the torments described in the initial stanzas. Thus in this
final stanza Schubert is justified in injecting some minor-mode character-
istics from the first half of the lied. The ascending sequence of measures
61 through 65, though not obstinate like that of measures 8 through 22
(compare the first two models in 9.4), passes through similar terrain. This
time it traverses a minor third, not a major second. This FA} span is fol-
lowed by a decisive Ab’—>D) resolution to the minor mode's submediant,
a chord that is extended via two rounds of minor-hued $~> embellishment
(measures 67 and 68, where D}-Gb-Bbb is enharmonically respelled and
unfurled into $ position). Of course, Dp-F-Ab needs only a By to become
the modified II chord that played such a crucial role in the first two stanzas
(measures 25 and 46). That is exactly the course that Schubert takes in
measure 69, and thus unexpectedly, via a different tonal routing, this
torment-signifying chord has reemerged. Yet the nun’s faith is unwaver-
ing; earthly tribulations do not deflect her heavenward gaze. This highly
dissonant chord miraculously dissipates: that is, instead of exerting its
usual control over the voice leading (as in measures 25 through 27, shown
in 9.1), a direct path to the radiant major tonic is opened up (measures
69 and 70, repeated in measures 71 and 72).’ How fittingly Schubert, as
Lord of chordal progressions, has responded to the plea, “liberate her [the
bride's] soul from earthly confinement”! Her prayer has been answered;
her tribulations are over. This marvel is celebrated through major-hued
‘3 embellishment in measures 73 and 74 (reminiscent of measures 31
and 32), leading directly into a sequential build-up to IV (the third model
of 9.4). (Note that this sequence is unusually sprightly once it gets going:
two chords are elided. Compare with the “stretched” cycle displayed at the
bottom of 2.11.) The remainder of this stanza resembles the third stanza,
including the singer’s heavenward closing ascent to 8, supporting the words
“to eternal heights.” The vocal melody of the final “Alleluja” is notable for its
clear C-D-C neighboring motion (a repudiation of the ominous D} neigh-
bor of the opening sections, beginning with measure 6) and for its refusal
to descend to the earthbound tonic.®
In the first stanza, just two chords sound in measures 11 through 18:
F-Ab-Db followed by F4-A-Cé. In my reading they are internal components
 “Die junge Nonne” 199

Model 1 Model 2
m. 8 iil 15 19 22 m. 61 63 oF yant. 65

a o $e
a b by am

z ]

= 6) I———= 6 5 ——5} 5S——= 6 5

Model 3
eh 15 78 Bison 80

rau ———_ bbe i

ee ne —
. moe

=] 6 5) 6 I=——H 5

9.4 Three sequential progressions from “Die junge Nonne” (D. 828).

of a sequential ascent, as displayed in 9.1 (and 9.4, Model 1). Kopp inter-
prets them as V-i in F# Minor. These two chords recur in the fourth
stanza: F-A>-D) sounds in measure 63, again in measures 65 through 68
(in 3 position, sometimes enharmonically respelled), and with the notable
addition of By yet again in measures 69 and 71. An inverted Gb-B}b-D}
[F#-A-C#] sounds in measures 67 and 68 without constituting a resolution
(D)—>G}[F#]). Kopp views this passage as an unsuccessful bid at the ton-
icization of F# Minor, trumped by an abrupt entry of the “disjunct” tonic, F
Major, in measure 70 (pp. 258-260). My contrasting reading of this passage
results from a more variegated interpretation of how the F-Ab-D} chord
functions in its several statements throughout this stanza. In measure 63 it
is again an internal component of a sequential ascent, as displayed in 9.4,
Model 2. The first stanza’s F# minor continuation is replaced by a Gb major
chord in measure 64, another internal element of the sequential ascent.’
Though his textual commentary and his Figure 9.12 do not seem to agree,
Kopp states that Schubert “lingers [on the D) chord of measure 63] to
establish D} major in its own right, by way of a chromatic bass progression
culminating in a strong cadence at m. 65” (p. 258). Thus our perspectives
for this passage clash:
m. 61 63 65

Kopp: F Db
(=Db Abo Db)
Damschroder: F( ) Abo Db

We agree that the pitches Gb and Bbb [F# and A] in measures 67 and 68
do not induce a succession away from the prolonged Db chord. (I regard
200 Harmony in Schubert

the event as derived from measures 31 and 32, with the unfurled bass
reminiscent of that in measure 26. Kopp calls it a “plagal tonicization of
C#/D}” (p. 258).) Then in measure 69 (not 70, as he states) By joins Db, F
and Ab. Kopp regards the chord that results to be “a Db dominant-seventh,”
not mentioning the enharmonic respelling required for this interpretation
(p. 258). Granted, Schubert does not always spell chords in the manner
our analyses suggest he should. Yet here the chord — which employs Bs,
not Ch — is a recurrence of a chord that has occurred earlier in the lied, in
measures 25 and 46. (As mentioned above, that chord is not addressed in
Kopp’s analysis of the first stanza.) In those earlier instances the resolution
leaves no doubt but that Schubert really meant By. Thus Kopp’s statement,
“the dominant seventh has clearly pointed toward F# minor as its tonic”
(p. 260) seems to me insufficiently tentative. (In his n. 15 on p. 260, Kopp
refers readers to another chapter of his book where he does assess the
chord’s potential interpretation as a “German augmented sixth” chord (p.
200).) Based on remembrance of the earlier passages, I suggest that atten-
tive listeners will expect yet another resolution to V in F Major. From the
stanza-opening F major to D} major is a large-scale I>-° motion (employing
a chromatic variant of the 6-phase chord). The latter chord’s dominant-
emulating potential might be asserted to direct the progression towards Gb
[F#]. On the other hand, it might (and in fact does) follow the conventional
harmonic course for tonic’s 6 phase by leading to II. One potent manifesta-
tion of II is Bk-Db-F-Ab.!°
Finally, what is one to make of the fact that two conspicuous chromatic
chords — A major in measures 56 and 57 and Db major in measures 65
(not 63) through 68 — are symmetrically equidistant from tonic F? Kopp
is intrigued by how Db “counterbalances” A, a flat “harmonic extreme”
after a sharp one (p. 258). At both points the text unites the bridegroom
(‘Brautigam’) and bride (“Braut’). I would give less weight to such a sym-
metrical relationship, because both of these chords are internal to broader
harmonic initiatives: V to I (expanded via arpeggiation as C-A-B), and I
to altered II (expanded via a 5-6 shift as F~-Db-G). They do not register as
“extreme” to me because V-I and I-II are such normative successions. The
symmetry is highlighted only if one takes the chords out of their contexts.
An insight I would offer in its place is that all the chords in the vicinity
of the words Brdutigam and Braut are of major quality (some with minor
seventh as well). The wobble to C# in measure 56 prevents a minor-chord
backdrop for the holy bridal couple, and the modal mixture (Ab—>Db in
measure 65) allows minor-key chords of major quality to occur in place of
their minor major-key equivalents.
10 | Four Impromptus (D. 899)
with Charles Fisk

Charles Fisk’s Returning Cycles: Contexts for the Interpretation of Schubert's

Impromptus and Last Sonatas (2001) offers a sumptuous banquet of com-
mentary on some of Schubert's most celebrated keyboard works. Fisk
ponders Schubert’s emotional state during his final years, finding intriguing
links between his texted and untexted compositions. Two central insights
guide his presentation: a pervasive “aura of Winterreise” in the late key-
board works, and a “cyclic return of specific melodic and harmonic material
... from one movement to another” (p. 1).
In that my agenda is more general than Fisk’s, I will focus on his overall
approach to analyzing Schubert’s music, without directly assessing his
speculations concerning the relationship between the keyboard works
and Winterreise. A sizeable body of music is needed to properly assess his
speculations concerning “returning cycles.” For this purpose I have chosen
the first set of Impromptus, D. 899, which Fisk analyzes in detail (though in
some cases not comprehensively, when certain passages do not contribute
to his argument).’ My analysis and my assessment of Fisk’s analysis are pre-
sented in alternation, section by section. Quotations from Fisk’s book and
my responses to those quotations are set off from the flow of my analysis
via shading.

Impromptu in C Minor

A bold dominant G opens the Impromptu in C Minor. It is as if we are

entering the world of the work at the concluding moment of a primordial
C-G structural cycle, whose traversal precedes the onset of the impromptu.
A new cycle begins promptly, with a solo melodic line. This region is con-
structed as a somber sixteen-measure theme (stated pianissimo and closing
in a perfect authentic cadence) followed by a variation, which builds even-
tually to fortissimo.? The non-tonic opening of a theme (such as Bk<D in
measures 1 and 2) is an occasional occurrence in Schubert's music, some-
times with less tonal focus than here. (See 1.19 and 4.40.) This impromptu's
theme is notable for its economical employment of musical resources.
202 Harmony in Schubert

e $d]
o>
aie a>
o||N)

t
N> —>
w

*X>“hh
: c(h
C Minor: I ( \m vi———I

10.1 Analysis of Impromptu in C Minor (D. 899/1), mm. 1/2-17.

Four unfolded thirds (Bp Ae , and %) account for most of the melodic
content. Fathoming how these thirds interrelate is complicated by the lack
of chordal support in some passages. Nevertheless the basic structure is, in
my view, conventional: the coordination of a 3-3-1 melodic descent and
the harmonic traversal of I-V4-I. (See 10.1, which omits the interruptions
that occur at two levels: phrase 2 completes what is begun during phrase 1,
and phrase 4 completes what is begun during phrase 3.) The initial I is pro-
longed via the presentation of & (marked by a bracket in Model 1) melodic-
ally as E>>D>C (Model 2), followed by an upper-third extension of bass C
to Eb, mimicked in tenths by the melody. This Eb supports tonic’s upper-
third chord, which asserts itself as III. The juxtaposition of minor tonic and
major mediant chords in measures 3/7 and 11/15 (here emphasized by the
similar approaches: GC and Bb—-E}) is a hallmark of composition in a
minor key. In this instance the mediant’s presence is brief. The bass contin-
ues its upward course to G for the cadential Vk-I in C Minor. An inverted
supertonic chord (not shown in 10.1) serves as intermediary between the
tonic/mediant expansion of I-space and the dominant. (This supertonic
is displayed in the more detailed rendering of the musical surface, corre-
sponding to the variation, in 10.2, Model 1. There a 5-6 shift in measure 27
extends the mediant.)
 Four Impromptus 203

Model 1
m. 23-25: 26 PAR 2 eS

mez

Gvimor. 1) () Tr? II NAD Ik

Model2
m. 42-47 48 50 51 52-53 54 55

3
(=[5 4 3 2 )

(ae
oy et ralid He iS H
iis

ee: ve 5

Boao NV
AP Minor fh Cina te! yy Py ay

10.2 Analysis of Impromptu in C Minor (D. 899/1), mm. 23-33 and 41/42-55.

“Not yet having played out its introductory role, this G [in measure 1]
now insists retroactively on that role by becoming the harmonic back-
ground, in place of the tonic, for virtually the entire theme” (p. 126).
Though Fisk and I probably have contrasting notions of what the
word “background” conveys in music analysis, my 10.1 offers a precise
and compelling display of how tonic C, and not dominant G, is the
theme’s controlling harmonic force. As I will show in the ensuing discus-
sion, a broad motion from C to G is a “returning cycle” throughout this
impromptu. Measure 1 corresponds to measures 87 and 160 - the end-
points of cycles. Tonic C in the opening theme grounds the first full cycle
that we hear. C must be established during the theme (and reinforced
in the variation) because the traversal of the path towards G begins
immediately after the cadence in measure 33. The sforzando A} chord of
measure 34 is a direct challenge to C’s hegemony.
204 Harmony in Schubert

The interpretation of Eb in measure 3 warrants close attention. Does the

repeated D of measure 2 lead into this E}, or is Eb an upper neighbor con-
necting the Ds in the preceding and following sub-phrases? As shown in 10.1
I favor the first reading: E} as Kopfton 3 is an indispensable component of the
conventional structure that I think Schubert is erecting here. Consequently
his wayward progression at the outset of the variation (measures 18 and 19) is
astonishing. Having followed a normative procedure in the theme, the varia-
tion opens with a daring variant. The G chords (measures 2 and 4) that in the
original theme performed contrasting roles are now united within a single
prolongation of the initiating dominant, expanded via the major supertonic
(represented by D-F#-A-C and by F#-A-C-E} in measures 18 and 19).° The
interruption structure’s redundancy facilitates this transgression: the 3 of
measure 23 saves the day by functioning in the conventional manner.’
The transition from C Minor to the submediant key, Ab Major, is accom-
plished in measures 34 through 41. The sforzando chords that open this
passage juxtapose the tonics of these keys and emphasize their shared
third. (Schubert provides a vivid demonstration of the theoretical notion
of Mehrdeutigkeit - multiple meaning - in this region: the * third appears
in the context of the tonic (measure 3), of the dominant’s } neighboring
embellishment (measure 4), of the supertonic (measure 19), and now of the
submediant (measures 33-34, 37-38, and more definitively 41-42).)° At
first the invitation to shift to Ab is declined, and tonic C is reconfirmed by a
cadence at measure 37. Then, at a piano dynamic level, the offer is repeated.
This time, even softer, Ab is embraced. The melodic Eb>Ab fifth (incorpo-
rating upper neighbor F) in measures 37 through 41 establishes the general
contour for the upcoming A} Major theme.

“At the end of the fourth period [measure 33] .. . a codetta initiates an
imposing ritual of C-minor closure. The cadential phrase has begun
again, as if in an echo, when a sudden inflection introducing Db, the
Neapolitan (m. 40), opens the opportunity for a deflection to Ab major.
The closing of the first A section is thus never fully completed, but is
instead elided with the beginning of a contrasting B section” (p. 126).
Fisk and I have differing notions regarding this regions topography.
He places measures 34-41 within the opening tonic expanse, whereas I
place them between the tonic and submediant expanses. Thus the swerve
to Ab is for him a transgression of the “ritual of closure” whereas for me
it is a mission accomplished, since I regard this passage as a transition,
not as a codetta. Fisk does not comment on the ingenious juxtaposition
that Schubert sets up in the transition’s Opening measures: clearly the
 Four Impromptus 205

A> major and C minor chords are closely related, but is Ab the lower-
third chord of C, or is C the upper-third chord of A}? That is the riddle
Schubert poses to initiate the move away from tonic. Though at first
the status quo prevails, ultimately Ab gains ascendancy. In fact, the
A><C<D><E><Ad progression in measures 38 through 41 (with C rep-
resented by its first inversion) is a harbinger of the Ab<C)<D}°<Eb<A}
progression that takes shape during the A} region (as we shall see in
10.2, Model 2). Though the D} chord of measure 40 has a “Neapolitan”
potentiality (which Schubert will tap adeptly in measures 114 and 118
to avert a repeat of the transition to A5), in the context of measures 38
through 41 it functions as IV in Ab Major.

The symbol I expresses an alliance between two distinct yet closely

related chords. The presentation of tonic’s 6-phase chord often is accom-
plished swiftly. Yet sometimes, as in the Ab tonicization here, a substantial,
thematicized prolongation occurs. Having already highlighted the shared
pitch content between C-E>-G and A>-C-E> (measures 33-35), Schubert
sacrifices a part of that bond — namely, their shared Cy - during much of
the A> region to establish a different alliance. Two structural models are
juxtaposed in 10.2. The first offers a detailed view of the chord progression
in the variation of the C Minor theme, once tonic belatedly arrives. Note
especially the bass contour and the chordal choices. The second model
shows how Schubert prolongs Ab Major. Through modal mixture the bass
contour and chordal choices are brought into conformity with those we
encountered in the variation.° Though it may seem peculiar that a Cb major
chord occurs in a work in the key of C Minor, it is only in that state that this
mediant (of A>) mimics the E> mediant of the tonic region.

“To be sure, the music digresses through a kind of echo into a contrast-
ing episode in A> major . . . This music evokes a dream, or perhaps only
the memory of a time when dreaming was still possible, in relation to
the opening’s bleak reality” (pp. 25-27). “Reverie, whether as reminis-
cence or fantasy, can intrude upon other kinds of awareness” (p. 126).
“Moreover, the harmony is no longer held to a single tonal center, but
instead moves as if in free exploration from Ab major, through modally
mixed harmonies to C) major, and back” (pp. 126-127).
Fisk’s assertion of a “free exploration” warrants the confirmation of
a detailed analysis, even if those details are not published. Details that
contradict Fisk’s assertion are provided in 10.2: these models reveal
206 Harmony in Schubert

that, in addition to the melodic and rhythmic features retained from the
opening tonic region (which Fisk acknowledges on pages 126-127), the
C and Ab regions share a deep structural bond. Likewise, Fisk’s descrip-
tion, “from Ab major... to Ch major, and back,’ seems to me an inaccur-
ate summary of the progression, since Ch is internal to an ascending
bass contour from A} to Eb (10.2, Model 2).’ (I acknowledge that some
readers may not be convinced by my reading of measure 52's Ab-C)-E}
as an unfurled 6-phase of III rather than as a return of I. Yet compare that
passage with bass E)>-C-F in measure 27 versus E>—F in measure 31.)
I welcome Fisk’s characterization of the passage as “reverie.” The story
that I imagine for this impromptu involves a valiant soldier returning
from war, suggested by the martial character of the opening theme.
The melody’s limited range and somber mood hint that the soldier is
seriously wounded. Yet in the Ab region he can expansively recall his
moments of glory. Eventually he becomes delirious (conveyed by the
odd succession of phrases in the final martial section) and expires
(measure 193). The coda mourns his heroic death.®

Though closure is achieved at measure 55, Schubert offers a concluding

phrase that reiterates, in slightly varied form, the structure of measures
42 through 46. Whereas the structural melody, displayed in a normalized
register in 10.2, Model 2, gradually migrates to a higher register (measures
51 through 53), this closing phrase confirms that shift by traversing the
entire Eb>Ab fifth in that upper register.
The Ab major prolongation is repeated beginning at the upbeat to
measure 61, with the melody repositioned at the bottom of the texture in
the first two phrases. (Again an element of theme and variation is at play.)
Now the concluding phrase is replaced by a more substantial passage (from
the upbeat to measure 75 through the middle of measure 82) that restores
the lower soprano register, though with some upward flourishes in meas-
ures 79 and 80. Here Fisk’s term “ritual of closure” might appropriately be
applied. Note one intriguing structural feature: in measures 75-76 and
79- au two adjacent§chords — both highly embellished- are juxtaposed (F
to e» in a construction similar to that displayed in 4.37). This juxtaposition
will recur, in the context of C Minor, during the return of the transition’s
“riddle” moment (3 to cin measures 112-113).
Despite its duration and its thematic potency, the Ab region is not the
endpoint of the broad structural initiative in which it resides. The remain-
der of that trajectory is accomplished with no frills, but instead with an
 Four Impromptus 207

insurmountable force that permeates the musical texture from top to

bottom. Tonic’s 6-phase chord (often an asserted or tonicized submediant)
characteristically proceeds to II. In minor keys a simple yet effective realiza-
tion of that succession occurs when #4 is added to the submediant’s pitches,
as in Ab-C-E} to Ab-C-E}-F# in C Minor (the latter chord being an evolved
D chord with absent root). That succession is the foundation for measures
83 through 86 of the impromptu, which culminates on an F#-C-E} chord,
built from the supertonic’s raised third, seventh, and ninth.’ Two layers
of connection occur: (Ab)-C-Eb-F (unfurled so that F resides in the bass)
comes between Ab-C-E) and (Ab)-C-Eb-F#;!° and passing G connects bass
pitches Ab and FE. As expected, the evolved D chord proceeds to dominant
G major (measure 87). The soul has been snatched out of its reverie, and the
C-G cycle that began with the C Minor theme concludes.

“In the return from A) major to C minor, the G renews its force through
agitated repetition” (p. 27). “Whereas [the] A [section]’s deflected cadence
has led to reverie, [the] B [section]’s full cadence now occasions something
like catastrophe . . . The entire five-measure phrase [measures 83 through
87] proceeds in closely conjunct parallel octaves, expanding upon and con-
firming the power of the impromptu’s opening G, which reinforces itself at
the cadence of the phrase with agitated, ‘Erlkonig’-like repetition” (p. 127).
Though Fisk and I are more or less in agreement regarding this
passage, our hierarchical organizations are subtly different. Whereas
Fisk reads the large-scale tonal plan as C-Ab-C (with G facilitating the
return of C), my reading is C-G, C (with Ab and D chords internal to the
progression that connects C and G)."’ The impromptu engages returning
cycles of this tonic-to-dominant trajectory:
m. 1 3 87 89 124/160 162 184 192

1 fet) Gy C10 G © fo G, Cthrough Gto C

Any mediant or submediant tonicization, exuding major-key splendor
in the context of a minor tonic, should be regarded as a temporary
reprieve: a return to the tonic or continuation to its dominant will almost
always occur. Though it may seem a “catastrophe” that this inevitable
course again prevails, only a deeper-level minor-to-major shift can sub-
stantively temper the emotive effect of the minor key, as happens later
(cadences in measures 168, 192, 197, and 201). The “Erlk6nig” associa-
tion is apposite, though in that context G serves as the tonic, whereas
here it is the dominant’s root. That work ends with the death of a young
male protagonist, which I suggest the impromptu also evokes.
208 Harmony in Schubert

Though the tonic’s return at measure 89 is a very significant structural

event, Schubert does not fully reiterate the C Minor theme. In its pared-
down version the interruption repetitions are eliminated, as is the motion
to tonic’s upper-third chord. Only the head motive and the G-C cadences
remain intact. The new variation that follows, beginning at the upbeat to
measure 96, does reprise the full structure, with several new twists in the
local harmonic coloring along the way. (Note, for example, that I-II from
measure 7 becomes 1-11. in measure 101.)

“Beneath the hammering Gs now comes a shadow of the opening theme

(... mm. 88-95) .. . The descending melodic sixth recalls, in grim
parody, the similar melodic motion of the B theme and the B codetta,
as well as the Eb-major moments of the A theme. This new left-hand
melody, already foreshadowing the actual return of the A theme, thus
also echoes the B themes as a musical emblem of the inescapable persist-
ence of specific memories” (pp. 127-129).
A full statement of the theme plus a new variation would require
thirty-two measures. Schubert resisted such an expansive reprise and
thus abbreviated the thematic statement. Fisk postpones the “actual
return of the theme” until the variation (measure 96). Whereas I pair
measure 2 with measure 88 and measure 17 with measure 96, Fisk
pairs measure 2 with measure 96, thereby distorting the cyclical alli-
ance. In addition, he makes bold and in my view untenable connections
among a range of descending sixths, proposing that E)>G in measures
89-90 corresponds to G>Bk in measures 11-12 and 15-16, to F>Ab in
measures 43-46, and to F>A} in measures 75-78. F in measures 43 and
75 functions as a neighbor to Eb, and thus E}>Ab, and not F>Ab, is the
meaningful interval.

Measure 112 is a surprise. Whereas in measure 34 A} Major was offered

as a successor to tonic C Minor, now a severe and discordant concoction
that cannot be tonicized mocks any hope of such a welcome reprieve. (As
mentioned above, the F-Ab-D (with pedal C) to Eb-G-C succession in
measures 112 and 113 — repeated with a flowing bass (representing Ab>F to
Eb>C) in measures 116 and 117 - is modeled on the succession of measures
75 and 76, transposed and modally adjusted.) Even the Db-F-A} chord of
measures 114 and 118, which in measure 40 helped confirm the shift to A}
Major, is here ineffectual. Consequently the transition ends where it began:
in C Minor. In that the initial transition was crafted specifically to realize
 Four Impromptus 209

a C-to-Ab shift, Schubert neutralizes it before proceeding, in a second

transitory passage beginning in measure 119, to one of C Minor’s more
somber allies, dominant G Minor. The Ab Major theme takes on a different
character in its new context, starting in measure 125. Yet recall that minor-
key elements infiltrated the Ab Major presentation as well. The especially
ebullient concluding phrase (corresponding to measures 56 through 60) is
omitted, but otherwise the region is closely allied to its predecessor, though
now already on G rather than in the tonal space between C and G. At some
point the G chord’s third, B}, must be raised, since By is needed to conclude
the large-scale C-G cycle on G major. (Compare with measure 87.) That
conversion would be very abrupt at measure 160, and so the deed is done
in measure 152. So even though G Minor displaces Ab Major, a bit of Ab’s
warmth is transferred to G during a reprise of the theme that began at the
upbeat to measure 75.

“Not only does G persist as a pedal through much of the second A

section; it also then overtakes the music of the B section, which returns
in the minor key of the opening G (... mm. 125ff.), the escape once
promised now denied” (p. 27). “The G, from the outset an overdeter-
mined dominant, now reveals itself as a kind of absolute, an inescap-
able presence with the power not merely to obstruct any escape from
C minor but to ensnare material from the major within the minor's
purview” (p. 130).
In that I regard the earlier B section as progressing from C to G via
A} and D, a paring-down of the structure would sacrifice the internal
chords/regions while retaining goal G. The “reverie” takes on a more
ominous character in its new tonal context, perhaps recalling the fateful
battle during which my story’s soldier is wounded. If such a reading
is credible (and certainly some readers will not find it so), then the
offbeat chords starting in measure 139 might represent enemy fire, with
our wounded protagonist collapsing in measure 148 and losing con-
sciousness. The sounds of battle no longer penetrate; a sweet G Major
eventually fills his frenzied mind.

The return of the C Minor theme, beginning at the upbeat to measure

161, is again truncated. In addition, a shift to C Major occurs during the
final cadence (measures 167-168). The variation that follows contrasts the
preceding variations in that its succession of phrases does not conform to
the structural model of the original theme (analyzed in 10.1), and in that
210 Harmony in Schubert

the mode wavers from major to minor and back to major again. A synopsis
of its various phrases follows:

measures 169-172 A normative first phrase, in C Major.

measures 173-176 An intensified rendition of the second phrase, in C
Major.
measures 177-180 Though the third phrase is expected, an aberrant
statement of the first phrase is presented in C
Minor. The harmonization is like that of measures
18-21.
measures 181-184 A normative third phrase, in C Minor. The
harmonization is like that of measures 26-29.
measures 185-193 An extensively modified fourth phrase, in C
Major. Tonic’s upper-third chord is absent.
The chord progression is an elaboration of
15-6_TI#-V-I.

“Again and again in the final A section the major mode will resurface,
but it will never break altogether free of the minor” (p. 130).
My story juxtaposes the pain and torment of serious injury (C
Minor) and the release through death (C Major). This final variation
conveys those delirious moments of struggle between life and death.
Death is the victor (measures 192-193). The soldier does break free of
his torment.
“The downward rush of [the last phrase’s] first two measures, from G
to A in the bass, and the V of ii harmony that the A supports, have no
precedent: (p. 132).
The downward rush (measures 185-186) is a somewhat acceler-
ated reprise of the descent in measures 100-101, restated with slight
variation in measures 173-174. Though “V of ii” is unprecedented,
the A-D succession (A-C#-E-G in measures 186-187 to F#-A-C-E} in
measure 190) varies the similar formulation in measure 101 (A-C-E-G
to Ab-C-D-F#).

The cycle is finally broken at the coda. The inaugural B&<D third in
measures 193-194 is preceded not by a bold dominant G, but instead by a
decisive tonic C (contrasting measures 1-2, 87-88, and 160-161). Despite
some lingering tinges of C Minor, the C Major tonic tolls six times in all
during the coda.
 Four Impromptus 211

“Even here, at the end of the impromptu, the major . . . emerges these
last times from the minor, like the image of a freedom or fulfillment
never achieved, or of a love that can still be born only of pain” (p. 133).
“The final turn to the major . . . is indecisive, still in the minor’s shadow;
the modal ambiguity of this ending suggests the impossibility of any full
tonal resolution for this music in its opening key” (p. 273).
In my story the E}-to-Ek shifts in these final measures no longer
signify a life-and-death struggle, but instead succinctly recall the strug-
gle that has just been lost. May this gallant soldier rest in peace.

Impromptu in E}

The second impromptu opens with a vivacity that starkly contrasts the
funereal close of the first.’* The Eb-G-Bb chord, which played a supporting
role in the context of C Minor and A} Major during the first impromptu,
now takes a star turn. Indefatigable triplet subdivisions invigorate the
melody, which in fact is so energized that it eventually soars above its
initially stated register. The contrasting ways in which three full cycles
(separated by bar lines) project the beamed G>F>E} structural descent
are shown in 10.3. The variances during measures 15 and 16 constitute a
mid-cycle upward shift that leaves the listener in doubt as to which register
will prevail. (Compare with the upward shift in measures 51 through 53 of
the first impromptu, discussed above.) The third cycle persists in the upper
register until the cadence, at which point the lower register is restored.
(In the first impromptu, measures 56 through 60 instead confirm the upper
register.)

“What is exceptional about [the opening phrase] is its immediate rep-

etition, not once, but twice. The second repetition suggests various
metaphorical equivalences: a blithe innocence, a somewhat maniacal
exuberance, or the turning of a kind of musical gear that cannot become
fully engaged” (p. 122).
Fisk does not mention the upward registral shift that occurs during
the second cycle. Will Schubert embrace the upper register, or will the
original register be restored? That question motivates the additional
repetition. Whereas Fisk’s characterization might well apply if Schubert
were merely repeating the same music over and over, that is not the
212 Harmony in Schubert

78 OM IS 14 SG I PAN 222
SG

:
m.

hi
E>Major: I I V’I
a)
je es
c
le Vaal

10.3 Analysis of Impromptu in Eb (D. 899/2), mm. 1-24.

m 25/32 33 34 35 36 38 «#440 «42 «43

N
b3

a ales
:

Eb Minor: I 0

10.4 Analysis of Impromptu in Eb (D. 899/2), mm. 25-43.

case here. The melody responds flexibly to unexpected circumstances

and eventually descends from the exciting yet ultimately unsustainable
upper register in favor of the more down-to-earth alternative. Not blithe, -
but aware; not maniacal, but responsible (at least for now).!°

The shift to Eb Minor at measure 25 signals the onset of a contrasting

b section. (The impromptu’s broad A, section, which transpires during
measures | through 82, divides into a ternary a,—b-a, form.) This shift is
more overt and extended than the modal mixture that permeates the first
impromptu’ B sections (as shown in 10.2, Model 2). Yet it is not in any way
problematic: a shift to the parallel key is one of various means by which a
composer may demarcate regions within a mevement.
Given the recent flirtations with the piano’s upper register, it is significant
that the b section is registrally subdued throughout: the root Eb in measure
25 is an octave lower than that of measure 1, and the octave connections
that shape the melodic lines - that is, Gb<G} in measures 25 through 43
 Four Impromptus V3}

m. 1-24 25-35 36-43 44-47 48-5]

Eb Major:I° Reh CRD Wan:

© EMino:I WM WV.
ay b

10.5 Analysis of Impromptu in E} (D. 899/2), mm. 1-51.

(connected by a dotted slur in 10.4), followed by A}<Ab a half step higher

in measures 44 through 51 - conform to the register of measures 1 through
8. The basic structure of the entire b section is displayed in 10.5.
Note especially one ofthe inner strands displayed in 10.5: Eb>Db>C)>Bb.
The C) serves as an extended displacement of the dominant’s Bb (as 9b-8).
Whereas Db} and C) function within a linear context here, Cb and Ds will
play disruptive roles in the melody when things begin to run amuck in the
a, section. (See the sforzando pitches in measures 72-73 and 75-76.)

“The turn to Gb major balances the turn of the preceding phrase from E} :
major to E> minor. This subsequent move to Gb major may initially seem
to be only an ingratiating exploratory digression, but it is also potentially
a dramatic response: an attempt to escape from the minor, to avert the
threat posed by E} minor to the impromptu’s opening Eb major” (pp.
46-47).
Fisk’s and my views here coincide. Tonics upper-third chord is
asserted as a tonicized key, while the bass E}<G) begins the march to
the dominant root B>. Though there is some temporary relief from the
tonic’s minor quality, the persistent Cb that embellishes the dominant
rekindles it, and only the E}) Major return at measure 52 removes the
“threat” (or so we think).

The presentation of the opening theme during the a, section, which begins
in measure 52, is truncated, as was the martial theme of the first impromptu
214 Harmony in Schubert

ho. Ee:
2 fe 4S) = ho VS

4,-———#—*
6 3h ae

2-2 —6—te ZZ 2%
pans SS =
E> Major: 1’
EI ( yee ole

10.6 Analysis of Impromptu in E} (D. 899/2), mm. 60-71.

during its recurrences. An upward shift is apparent already in the first cycle
(which corresponds to the opening statement’s second cycle). The situation
becomes critical in the next cycle, for at the very moment when the melody’s
initial G>G octave concludes (measure 64), Db unexpectedly arrives in the
bass, setting into motion a circular progression of descending fifths similar
to that which occurred at the head of the b section. Yet whereas that earlier
progression adheres (except for B}) to the diatonic chord qualities for E}
Minor (eb-ab—Db’ ~Gb—>C}’-f°-Bb’— eb), that which commences during
the a, section is more dynamic - dangerously so — due to several dominant-
emulating adjustments (E”—A}-D’>G-—>C’->f). This whirl of chordal
activity hurls the melody ever higher, without restraint, until it reaches the
highest key on the Schubert-era fortepiano in measure 69. The open note-
heads in 10.6 display the circular progression connecting | and II, followed
by V’-I, while the filled-in noteheads above show the actual location of the
soprano pitch for each chord. The spectacular ascent soon loses its wings:
a swift and agitated downward plunge, extending over four octaves in just
eight measures, crashes on a low Eb at the downbeat of measure 77. The
measures that followed the earlier circular progression pursued a rational
ascending melodic course: Gh<Cb>B)<D<E} (measures 32-35, shown in
10.4). Now some of these pitch classes form a disjunct downward trajectory,
as sforzando Gb>C)>D (measures 71-73, then again in measures 74-76).
During this dramatic descent the tonic shifts to minor quality.

“It is extravagant to entertain the notion that this might be the nightly
dance of the Erlking’s daughters — but might it, at least, be an alluring
dance emanating from some other pleasure garden?” (pp. 135-136).
Any speculation regarding extramusical associations is risky, and Fisk
prudently advises his readers about what he is doing. Yet one may wonder
 Four Impromptus

how well his story corresponds to what Schubert actually wrote. How
can the terrifying, minor-hued descent that ends the a, section have
anything to do with “pleasure”? My story is much different: Icarus,
whose overconfidence upon learning how to use the wings that his father
Daedalus had made for him led him to fly too close to the sun, plunges
downwards into the sea as the wax that binds the wings melts. How
could such a woeful turn of events not elicit a modal shift from major
to minor?
“In the Eb-Major Impromptu, the subversive pitches - Gb and Ch -
first occur not as momentary chromatic deviations but as components
of an Eb-minor passage to which they diatonically belong... They...
eventually precipitate a devastating climax” (p. 136).
Even without recourse to an external story, one might discern that the
principal event of the a, section is the registral collapse, not the modal
shift. The catalyst for the “devastating climax” is the rushing downward
motion from an extreme high point; the minor-inflected G> and C) from
measure 71 onwards are a response to that event already in progress.

Two analyses are juxtaposed in 10.7. They display the structural essence
of the broad A and B sections of the first and second impromptus. (The
register of some of the pitch content has been adjusted.) The similarities
between these models far outweigh the differences. The major-to-minor
shift near the end of the second impromptus A, section puts that structure
on an even footing with that of the first impromptu. Both bass lines descend
two thirds. Ebb, which first appears in measure 83 of the second impromptu,
is a wobble that is prolonged for nearly eighty measures. Note that Eb even-
tually is restored, and recall that even in the first impromptu the equivalent
region (Ab Major) was permeated by modal mixture from the parallel
minor key. (The prominent Cp) in 10.2, Model 2, corresponds to the second
impromptu’s Eb).)

Fisk does not comment on this evidence supporting his thesis that the
four impromptus are closely interrelated.

There is a close alliance between the bona fide Eb Minor region of the A,
section’s b subsection and the unexpected Eb Minor close of its a, subsec-
tion. Models 1 and 2 of 10.8 juxtapose melodic content from these pas-
sages. Something extraordinary happens next, during the approach to the
216 Harmony in Schubert

impromptu’s B section: the nadir Gb of measure 77 soon rises up an octave,

from which a melodic line ascends in arpeggiation (Model 3). Whereas
measures 33 and 34 offered C)>Bb, now the relationship is reversed to
Bb<Ch. (This is the very motion that constitutes the conventional 5-6 shift
from Eb Minor to Ch Major.) Were Schubert to follow the model exactly,
he would ascend further to Eb, perhaps preceded by Ds (as in Model 1).
The wobble to Eb} (D#’s enharmonic equivalent) in measure 85 replaces
this Eb. Consequently the last two pitches of Model 1 are both accounted
for in the Ebb of Model 3: what might be (in a different harmonization) the
lower neighbor of the latter pitch turns out to be that latter pitch’s chromatic
mutation.!° The stark Ch>D diminished seventh in measures 72-73 and
75-76 (Model 2) is transformed into the C)<Ebb minor third of measures
83-85. (Schubert’s notation of the entire B section in enharmonic B Minor
is a hindrance to visual comprehension, though of course it in no way
modifies the aural relationship.)
As we continue into the B section, I will begin employing Schubert's
B Minor spellings (representing C) Minor). First, a general orientation is
in order. Observe that in 10.7 the chord of measure 83 is displayed with
soprano Gb [F#]. That chord’s C} [B] and Ebb [D] are shown in 10.8, Model
3. The ascending trajectory continues upwards to G} [F#] in measure 87.'7
Though this F# is introduced during an F# chord, it is understood to persist
(though no longer sounding) when the B tonic is restored in the following
measure. F# is the initiation point for a local descent through E in measure
89 to D in measure 90. Thus the Gb<Gb trajectory of measures 77-81 con-
tinues upwards another octave in measures 81-87, thereby reinstating the
register of measure 74 — and of measure 1! Once that is accomplished, the B
section pursues cycles of descending F#>F# (representing Gb>G}b) octaves
that expand upon the cyclic G>G octaves from the A, section [10.3]: first
in measures 87-95 (upper F#) through 102-110 (lower F#), then in meas-
ures 115-118 (upper F#) through 124 (lower F#, followed by an additional
two-octave descent, thus reminiscent of the downward cascade of measures
71-77), and finally a motion from F¥ to B (measures 131-142 through
146-158). These descents are displayed prominently in 10.9.
Some additional comments concerning this region follow:

(1) The opening expansion of tonic B Minor incorporates an embellish-

ing chord in measure 84 that correlates with the chord in measure 112
of the first impromptu (there embellishing a first-inversion C Minor
tonic). This chord is unfurled in measure 89 of the second impromptu.
The chord of measures 103 and 107 embellishes dominant F#-A-C# in
 Four Impromptus PAU

Impromptu in C Minor
m. 3 41 86 87

E e
onee —
Di
i

m. 1 71 83 163 165

es sl — pg ue
Fee
Serre
a i?

Eb Major: I” ( ENE
E> Minor: ==" jie
(= VI»)
10.7 Analysis of Impromptu in C Minor (D. 899/1), mm. 3-87, and Impromptu in E}
(D. 899/2), mm. 1-165.

much the same way, though with a heightened concentration of dis-

sonance. That latter chord recurs in the context of tonic B Minor in
measures 147 and 151.
(2) 5-6 shifts facilitate various stepwise-ascending root motions: from
B to C# in measures 96-98 and 118-120, from F# to G# in measures
121-122 (where the 5 phase ofthe F# chord is elided), and from E to F#
in measures 140-142.
(3) Twice a broad descending octave proceeds as F#>D>C#>A>F# [10.9,
measures 87-124]. Compare the role of D (6 in the context of the goal
F# minor chord) here with that of Cb (6 in the context of Eb Minor)
in 10.8, Models 1 and 2. Schubert reiterates this octave connection
218 Harmony in Schubert

a
Eb Minor

Model 2

E> Minor
, ~s
bg
Model3
m. 81 82 83 85

(ps
ll

Cb Minor

10.8 Analysis of Impromptu in Eb (D. 899/2), mm. 32-85.

m. 81 83 87 90 95 97 98 99 102 114115 117 118 119 120 121 124 126° 127131 135 42 “143: 146

== ban SSS eee ee

BE = = a
eS
ethan ee 8 ee Sh
B[C>] Minor:[3———* y7* vi—+ 3 ST V——$ ——. Ve ve

10.9 Analysis of Impromptu in Eb (D. 899/2), mm. 81-146.

in lower registers twice in measures 125 and 126, first by stacking the
D>C# neighboring motion and the descent to F# on top of one another,
and then by placing them one after the other in a swift melodic descent.
D functions as a neighbor in measure 113 as well, and a G neighbor (6
in B Minor) occurs in measures 117 and 135.
(4) A stealthy return of tonic B Minor begins in measures 116. Instead of
B-D-F#, dominant-emulating D#-F#-A-C4 pushes towards an E minor
 Four Impromptus ANG

(a a
= :

E> Major: [:
Eb Minor: I 2——— ° Vi I
(= VIbb)

10.10 Analysis of Impromptu in Eb (D. 899/2), mm. 169-258.

neighboring chord, after which the B chord in its diatonic state settles
In.

The A, section, which begins in measure 169, faithfully reprises the

content of the A, section. Because of the unusual major-to-minor conver-
sion of its ending, a coda is called upon to resolve the matter. To be sure,
the matter is as yet unresolved: 3>$3>2 may cadence on | as tonic in either
Eb Major or Eb Minor. So there is still hope of ahappy outcome. The crucial
tonic chord is that of measure 258, not that of measures 239 through 248.
The structural close, which in fact does not shake off the minor inflection,
is shown in 10.10. (Icarus drowns.) Because only a small part of the B sec-
tion’s content is incorporated within the coda, Gb does not sound at the top
of the texture during the submediant chord and thus appears in parentheses
within my analysis.
The Ebb-D juxtaposition in 10.10 may be cause for concern. (Despite its
enharmonic spelling as D in the score, Ebb is in fact lowered 8, and not 1)
From my perspective, E}b stands in for diatonic Eb as a wobble. As we saw
in 10.7, the wobble may eventually resolve itself through the restoration
of Eb before the dominant’s arrival. (In that case, incidentally, Schubert
omits the D from the V’ chord.) In the coda this structure is very sketchily
reprised, and so the submediant leads much more quickly to the dominant.
Consequently the Eb is never restored. An elision has occurred: instead
of Eb>E}bb<E} followed by D, the voice leading proceeds directly from
Eb>Ebb to D.
During the coda the melody rises ever higher in conjunction with rep-
etitions of the structural descent, which morphs first into 5>1 (measures
220 Harmony in Schubert

267-270) and then, after ascending again via a quotation of Cbh>Bb<D<Eb

[10.4] in measures 275-276, into 8>1 (a transposed quotation of measure
126), which hurls downward four octaves with astonishing force.'* The dev-
astation of this close hardly leaves room for hope, and yet a few moments
later a lovely Gb Major sonority arises, inaugurating the third impromptu.

It was not Fisk’s intent to analyze the entire impromptu. However,

one needs to distinguish between Eb minor chords that are internal
to larger structures, for which an E} major close remains a possibility,
and the terminal E} minor chord that takes hold in measure 258. Fisk's
dwelling upon the Eb minor chord at the end of the A, section evades
the question of whether that event is internal to a broader initiative
(namely the G>G}>F structural line displayed in the second model of
10.7) or instead is a terminal shift of the Eb tonic from major to minor
quality.

Impromptu in Gb Major

With the third impromptu we reach the point where a careful consideration
of key relationships is warranted. (The apparent concurrent genesis of these
four impromptus in 1827 invites an attempt to discern a comprehensive
tonal plan that, if persuasive, would suggest that they were intended to be
performed as a set, despite their checkered publication history.!°) Gh Major
is of course quite distant from the C Minor that inaugurates the set. Yet
we encountered an equivalent relationship over the course of a mere eight
measures in “Die Einsiedelei” [4.26], where G Minor led to Bb Major, Bb
Major was transformed into Bb Minor, and B} Minor led to D} Major. (The
association of this pattern with the Funeral March from Beethoven's Piano
Sonata in Ab Major, op. 26, coordinates with the stories I have been telling:
the death of a soldier, Icarus plunging into the sea.)
The second impromptu’s quirky major-to-minor shift thus becomes pur-
poseful within a broader tonal agenda: it facilitates the consecutive ascent
of two minor thirds. Schubert calls upon further upper- and lower-third
relationships, which for the first three impromptus encompass the follow-
ing keys:

Eb Major G} Major Bb Minor

C Minor Eb MajorMinor Gb Major
A} Major Cb Major Eb Minor
 Four Impromptus Hp}

The first two impromptus are very similarly structured: the mediant plays a
vital role in the A section’s tonic prolongation [10.1, 10.2 (Model 1), 10.4,
10.5], whereas the submediant is a long-prolonged 6 phase of tonic at the
outset of the contrasting B section, in a progression that will eventually con-
tinue downwards another third to the supertonic’s (raised) third in the bass
before ascending to the dominant root [10.7]. (In the second impromptu
the submediant’s quality is modified through an extended wobble.) In both
cases the mediant becomes the tonic of the next impromptu.
Unlike that of the second, the third impromptu’s tonic retains its major
quality. The A sections of the ABA form are uplifting, and even the turbulent
E> Minor of the B section eventually is transformed into E} Major. Though
there is a small touch of mediant in the A section (the Bb chord of measure
6), fifth-relationships rather than third-relationships predominate: G)—>C)
followed by A>-D> is the principal route to the dominant of measure 15.
(These details will be fleshed out below.) The submediant opening of the
B section (measures 25-31) corresponds to the preceding impromptus at
that juncture. Schubert even leads downwards another third in the bass
to Cy (measures 33-39; compare with 10.7). This trajectory is a dead end,
however. No 5-6 shift to attain the supertonic (an essential harmony in the
trajectory of the earlier impromptus) or ascending step to the dominant
ensues. Instead the E> submediant returns in measure 42 and, after evolving
into a dominant-emulating chord, proceeds via a fifth-relationship (E}—>Ab
in measures 51-53), attaining the supertonic through which dominant D)
(measure 54) is secured. That supertonic begins in a diatonic, consonant
state (plainer than the supertonics of the earlier impromptus), yet later in
measure 53 it evolves into the equivalent of the supertonic in the second
impromptu.
I propose that the third-relationship of E> to Gp between the second
and third impromptus is likewise a dead end (compare with the motion
from B> to D} in “Die Einsiedelei? shown in 4.27), and that instead a
fifth-relationship — Eb to Ab - prevails. Had E> Major not been diverted
to Eb Minor during the second impromptu, the trajectory of the first two
impromptus (C Minor as III, Eb Major as V) would lead naturally to Ab
Major (the key of the fourth impromptu) as I. Given the extraordinary
events that transpire after E> Major is established, it is hardly surprising
that Ab arrives in a minor hue at the outset of the fourth impromptu. Yet
the key signature is set up for major from the outset (just as the signa-
ture in the second impromptu never succumbs to minor). The Ab tonic
chord’s C} wobble gives way to diatonic C in measure 31 of the fourth
impromptu.
DED Harmony in Schubert

“The sequence of keys in the set .. . has suggested no compelling tonal

unity or logic, or at least so various Schubert scholars [Fisk quotes
Walther Diirr in an endnote] have claimed . . . But these key signatures
tell lies” (p. 123).
From my perspective the key signatures, free of the various wobbles
that ensue, tell a persuasive tonal story. Diatonic III-V-I is a commend-
able auxiliary progression (a concept introduced on p. 21), focused on
its endpoint goal, Ab Major. Schubert occasionally juxtaposes dominant
and subtonic chords as a means of extending V-space: for example, in
1.11 (in Eb Major) subtonic D} major and dominant Bb major chords
are so related. Though typically the subtonic (as a chromaticized upper-
third chord) will precede the dominant or come between two state-
ments of the dominant, the succession of keys in the second and third
impromptus is dominant to subtonic.
These various keys and their parallels (except for the C Major parallel
that concludes the C Minor impromptu) all reside well along on the flat
side of the circle of fifths. Thus it is not surprising that some chords are
shared by two or more of the impromptus. For example, C) [B] is an impor-
tant region in three of the four impromptus. Mehrdeutigkeit (see p. 128)
was well understood by analysts during Schubert's lifetime. Fisk’s chart
(his fig. 5.1 on p. 124) continues in that tradition, showing numerous inter-
relationships.”” Perhaps the most intriguing is that the submediant key (Ab
Major) in the C Minor impromptu turns out to be the goal of the entire set.

Recalling the cadence in Eb Minor at the end of the second impromptu,

and anticipating the prominent role that E) Minor will play at the start of
the third impromptu’s B section (measure 25), the Gb>E} connection in
the opening measures of the third impromptu has a compelling unifying
effect.” The progression continues to Ab in measures 2 and 3, resulting in
a conventional [?-°-II. (In the impromptu’s B section an expanded version
of this progression will be intensified through dominant emulation within
an asserted 6-phase region: Ey>—Ab in measures 46-53.) The first phrase
concludes in a manner typical of an antecedent phrase: dominant harmony
supporting 2 (measure 4).”?
Schubert pursues an alternate approach to the dominant, via the sub-
dominant, in measures 5 through 7. The Bb chord at the end of measure 6
extends I-space, resulting in the succession Gh->C}.?3 This subdominant is
extended by its 6-phase chord, which leads effectively to the dominant and
tonic that conclude the phrase.
 Four Impromptus 223

“The opening tonic to submediant progression of the third impromptu

does more than recall the concluding harmony of the second; it also
anticipates the piece’s overall tonal structure, with its dramatic E}-minor
middle section” (p. 117).
Fisk notes only a part of the structural correlation. I do not inter-
pret the B section as being in Eb Minor. Instead, it is a fluid region that
connects the tonic (whose 6-phase chord at the outset of the B section
extends I-space) and the dominant (which is presented in first inver-
sion). In the A section the bass proceeds downwards from E} to a chord
with bass C) and root Ab (measure 3). Likewise the B section’s Eb pro-
ceeds first to a chord with bass/root C) (measure 33), and then to a chord
with bass/root Ab (measure 53). The latter continues onwards to the
dominant (measure 54), as does the progression of measures 1 through
4, As Fisk mentions elsewhere (p. 138), E} to C) is a prominent succes-
sion during the second impromptu. (See 10.7.)
“Moreover, the way in which the C)-major harmony first appears in
the third impromptu [measure 7] makes it plausible to associate it with
memories arising from the second . . . It also immediately follows E}
minor’s dominant, as if E} minor were yielding to it” (p. 117).
I do not hear a strong potential for the “possible extension of E}
minor’ (p. 117) that Fisk proposes. He is working under the assump-
tion that Bb)-D-F-A} must be interpreted as “V_/vi” (p. 116). Whereas I
acknowledge that this chord sometimes leads to E}, on other occasions it
does not. As stated above, I regard the chord as an extension of I-space,
pushing towards IV. As I will show below, Schubert provides strong
support for this interpretation in the way he leads from I to IV (in the
tonicized key of Ch Major) in measure 33.

Measures 9 through 11 likely would generate a wide range of interpreta-

tions among analysts. My reading (displayed in 10.11) is based on a percep-
tion that what occurs twice during measures 1 through 8 (an interrupted
structure in the first four measures, a somewhat modified complete struc-
ture in the next four) expands into an entity of broader scope in the next
eight measures. Root Gb (measures 1, 5, and 8) is retained as the foundation
for this continuation and recurs in the inverted chord at the end of measure
11. The C}> from the first half of measure 7 holds sway during measures 12
through 14. The dominant that begins in the second half of measure 7 and
overflows into measure 8 (where it clashes against the tonic root) occupies
all of measure 15, while the concluding tonic of measure 8 corresponds
224 Harmony in Schubert

29 aie aed 7
G> Major: 1° i Iv° Se een

10.11 Analysis of Impromptu in Gb Major (D. 899/3), mm. 1-16.

to that of measure 16. In this context measures 9 through 11 forge a path

between the root-position tonic of measure 8 and its inverted and seventh-
enhanced restatement in measure 11. Using a variety of unfurlings and
passing chords, Schubert has camouflaged the linear motion that propels
this connection.”4 An idiosyncratic ascending 5-6 sequence guides the
passage. One of its 5-phase chords (Bbb-F)-Db at the beginning of the third
cycle) has been elided. As is typical in such progressions, the third 6-phase
chord (Bb-Gb-D}) restores the pitch classes of the first 5-phase chord. (See
3.27 and the discussion on p. 47.) This expansion of tonic coordinates with
an ascending third from Bb to Db in the melody, which is answered by an
E}-to-C) descending third during the IV expansion. Only after that opera-
tion is completed does IV’s 6 phase emerge, propelling the motion to the
cadential dominant.” The 8’-’ motion above root D} during the cadential
3 is unfolded as ,, “> during measure 15.

Fisk's example 5-1 (p. 116) incorporates a detailed Roman-numeral

analysis, which is exemplary from a literalist perspective on harmony,
wherein every chord receives a label that accords exactly with its pitch
content — no imaginative additions or subtractions, no hierarchical con-
siderations. I once analyzed music in that manner as well. Now, however,
I find three principal faults in such analysis:

(1) The labels hinder comprehension of interrelationships between

similar chords. For example, I regard G}-Bb-Db in measure 8 and
G4-B)-Db-Eb in measure 9 to be very closely related (the two phases
 Four Impromptus US

of a chromaticized 5-6 shift, connected by an intervening upper-

fifth chord). In addition, Bb-D4-F-A} in measure 6 functions in the
same way as the more normative G}-Bb-D}-F) in measure 11. (Both
are evolved states of tonic, in my view, and both lead to IV.) Fisk’s
fussing (pp. 117-118) over the earlier chord’s resolution on C)
(instead of Ep) is characteristic of an unimaginative — ie., literalist -
perspective on harmony.
(2) Saturation analysis - with as many Roman numerals as chords -
bleaches out the hierarchical relationships that I find pervasive in
Schubert's music. For example, Fisk’s analysis of measures 11 and
12 as

IV/IV IV 6 VAY IV

precludes any close affiliation between the first and third chords,
even though IV (of IV) generally would proceed to V (of IV). In my
view, the chord labeled IV, is not functioning at the same level as
the chords around it. If Iwere to analyze this as a harmonic progres-
sion (which I would not: see 10.11), I would interpret that chord
as the unfurling of a cadential ° chord embellishing the dominant-
functioning chord that follows it.
(3) Not all chords participate in harmonic progressions. A sequential
progression connects the tonic chords of measures 8 and 11. In
such a prolongation none of the internal chords are harmonically
conceived. The only harmonic activity is the evolution of the tonic
— the addition of F) at the end of measure 11 to provide a stronger
propulsive force towards the subdominant.

The style of analysis that Fisk pursues is not innocuous. Such

symbols impede the comprehension of music as I think it ought to be
comprehended.”

As in the preceding two impromptus, the B section of the third also

begins with a tonicization of the principal tonic’s asserted 6-phase chord
(measure 25), and the bass likewise ultimately descends another third to
Ch (measure 33). The Eb>C) transition works similarly to the C>Ab transi-
tion of the first impromptu (measures 34-41), where a Db chord serves as
IV in Ab, evading its potential role as bII in C Minor. From Ep in the third
impromptu, F) as IV leads to G} and C} (measures 31-33). Later in both
impromptus, a restatement of this potential IV chord does behave as bIl,
226 Harmony in Schubert

thereby averting a reprise of the descending-third motion (D} major in

measures 114 and 118 of the first impromptu; Fb major in measure 46 ofthe
third impromptu).
The pensive waiting on C) in measures 35 through 39 is an unexpected
contrast to the preceding impromptus, where an inverted II proceeds
quickly to the dominant [10.7]. Thus here something seems out of kilter.
The foreboding >-°"-5 embellishment in measures 35-36 (similar to, though
less dissonant than, the embellishment in measures 83-85, 91-93, 102-104,
and 106-108 of the second impromptu) becomes less menacing in measure
37. Ultimately the progression simply stops: we reach a dead end. Measures
40 and 41 direct us back to Eb Minor, from which a different harmonic
trajectory will ensue, one that leads successfully to the D) dominant goal
of the B section.
Before proceeding to that continuation, note that in measures 33 through
35 yet another presentation — now in Cb Major - of the I>1V>°-Ve-5-I har-
monic progression ofthe A section occurs. (Compare with 10.11. There the
Bb-C} neighboring motion in the soprano, supported by I-IV, occurs as an
ascending second. In measure 33 Eb—F} occurs as a descending seventh.) The
specific form of tonic evolution that occurs in measure 33 makes this passage
especially fascinating: the Cb-Eb-G-Db chord there strongly supports my
reading of the Bb-Dk-F-A} chord in measure 7 as an evolved G} tonic.
Especially in a major key, I°~° often leads to II. The tonic’s 6 phase (Eb
minor) is prolonged (including the fruitless detour to Cb) from measure
25 through measure 46. Note especially the reprise of measure 27’s bass
Eb>Ab<Bb in measures 42-43. The Eb chord ultimately takes on dominant-
emulating characteristics that press more strongly towards II (Ab minor),
which arrives in measure 53. (Measures 47 through 51 reprise in Eb Major
the structure that was traversed in Ch Major in measures 31 through 35.)
Once the A} supertonic is secure, dominant D} duly arrives in measure 54,
completing the harmonic trajectory of the B section. Ebb>D} in measure 54
echoes Ch>Bb during the corresponding structural dominant of the second
movement (measures 165-168).

“In the G>-Major Impromptu, Eb minor persists as a troubling presence,

especially in its agitated B section” (p. 42).
Fisk's statement takes no account of the transformation of the Eb
chord during the B section. Though Schubert at first uses Eb as a link
in the connection between I and IV (Gb>E}b>C}), his revised trajectory
(Gb>Eb— Ab) makes of Eb something that was not possible in either of
 Four Impromptus 220,

the first two impromptus: namely, a chord with dominant-emulating

tendencies. As 10.7 shows, the Impromptu in C Minor’s connection
between tonic’s 6 phase and the supertonic engages roots Ab and D,
whereas the Impromptu in E}, incorporating a minor inflection, engages
roots Cb and FE. Neither A}—>D nor C)-9F is viable, because the roots
are not separated by a perfect fifth. Consequently the Impromptu in
Gb Major’s E)—A} is an emancipation from the more troubled earlier
impromptus. Fisk acknowledges the shift from E} Minor to E} Major
only much later (p. 118), where he describes it as a “moment of calm”
“A veiled self-quotation [occurs in] the opus 90 impromptus . . . The
segment in question first occurs in measures 37 and 38 of the second
impromptu. ..in whicha melodic fall from 4 to 3 in G} major is supported
by the harmonic progression V}-I, while an E>) appoggiatura in the tenor
register strains within the dominant harmony. Essentially the same music
reappears in measures 54 and 55 of the third impromptu . . . to conclude
the transition from its B section back to its opening theme” (p. 39).
Both passages are in G), and so it is not especially noteworthy that a
D)’ chord with 9-8 embellishment occurs in each. Yet indeed, given the
multiple associations that have been noted among the impromptus, this
one may be added to the mix, even if the D) chords perform contrasting
function. That of the second impromptu falls within a prolongation con-
necting root-position and first-inversion tonic chords. (I do not interpret
the succession Fisk mentions as a motion to a root-position tonic: bass
A} is here an upward-tending note that reaches its goal Bb only after a
repeat of the phrase’s initial two measures.) That of the third impromptu
is the goal of the entire B section (not part of a transition that follows the
B section, as Fisk asserts). As stated above, I relate this 9-8 resolution to
that in measures 165 through 168 of the second impromptu.

The A, section is only modestly different from its A, predecessor. The

structural close that we expect in measure 70 is elided, and the repeat of
this material, which fills measures 17 through 24 during the A, section, is
abbreviated, so that closure is achieved early, in measure 74. The bass of
the ensuing coda descends twice by step to the fourth scale degree - first
diatonic C} (measure 76), then wobbly Cb) [B}] (measure 80). Descending
fifths have pervaded the bass since Gb>Eb>C) in measures 1-3, with step-
wise descents much like the coda’s occurring in measures 31-33, 47-49,
and 51-53. The tentative relationship between local tonic C) Major and F)
minor hinted at in measures 35 and 36 is asserted as I->IV}) in the context
228 Harmony in Schubert

Model 1 Model 2
ime 78 79 80
8 9 10

See
m.

SS S—|
Gia VAD GP om Abb

Model 3 Model4
m. 78 80 81

a Ob

Gb Abu DP Gb Abs, Db

10.12 Analysis of Impromptu in Gb Major (D. 899/3), mm. 78-81.

of Gb Major during the coda, with motion from G) major to Cb minor

(measures 74-76).2” Whereas the altered II (of the “augmented sixth” type)
that occurs in measure 36 is not affiliated with the C)-Fb-Ab} that preceded
it, the same chord in the context of Gh Major serves as the IVbb chord’s
evolved 6 phase in measure 76 (Cb-Ebb-Gb to C4-Ebb-Gb-Bbb).
The second descent proceeds by whole steps - Gb>Fb>Ebb>Dbb>Cbb — in
measures 78 through 80. (With apologies, I will not be using Schubert's
spellings of convenience.) Note that the chord in the second half of measure
79 contains the same pitch classes as chords in measures 76 and 80, which
resolve to V. Why, then, does an Ab} minor chord emerge at the downbeat of
measure 80? Consider the progressions displayed in 10.12. Model 1 shows
a conventional chromaticized 5-6 motion in Gb Major, connecting I and
II. Model 2 shows a more extensively modified version that leads upwards
a half step rather than a whole step, resulting in the chord progression that
occurs during measures 78 through 80.** Now consider the familiar progres-
sion shown in Model 3. Already mildly chromatic here (thus G)-Ab—Db),
further evolution is evident in Model 4 (thus G)-Ab@Db). In a remarkable
instance of Mehrdeutigkeit, tonic Gb’s evolved 6-phase chord in Model 2 is
equivalent in sound to a chord rooted on A} in Model 4. Having discovered
this extraordinary enharmonic relationship, Schubert does with this chord
exactly what he did with the Eb minor chord in the B section: after first
proceeding one way (to Cb major), he backs up and repeats the E} chord,
proceeding another way (via E} major to Ab minor). Similarly, one resolution
of the mehrdeutig chord (as displayed in Model 2) occurs in measures 79 and
80, but then the chord returns and a different — and definitive — resolution (as
displayed in Model 4) occurs, from which the coda’s cadence arises.2°
 Four Impromptus 229

m. fest 05 107 161 169 171 201

Ec oreseee
Ay B A>

& : Oa as as : - Ta,

Ab Major: 3” rypz® vor 1

10.13 Analysis of Impromptu in Ab Major (D. 899/4), mm. 1-201.

Impromptu in Ab Major

Though the basic tonal plan of the fourth impromptu is straightforward, it

is distinguished by extended wobbles between minor and major qualities.
(See 10.13.) Long delayed and then even longer prolonged, the major tonic
(Ab Major) ultimately takes on a minor seventh, directing the progres-
sion towards the subdominant as the ABA form’s B section opens. In this
regard the fourth impromptu contrasts the descending-third motions of
the preceding impromptus at that juncture. The subdominant is introduced
as Db minor, causing renewed uncertainty regarding the quality of the
impromptu’s A} tonic. Does the A section’s Ch<C shift (over tonic root Ab)
represent a delayed arrival of a major tonic, or is a minor tonic undergoing
chordal evolution through dominant emulation targeting diatonic minor
IV? Only for a brief moment (beginning at measure 139) does radiant D}
major sound. The model’s §?fifth and f?third (marked by brackets in 10.13)
are the principal venues for linear descents.
Locally, the juxtaposition of the third impromptu’s Gb Major tonic and
the Ab minor chord that opens the fourth impromptu is a diatonic con-
nection (G}: I-II), enhanced through the subtle mehrdeutig reinterpreta-
tion of pitches from G}’s dominant ninth chord (Eb>Cb>Ab, prominently
arpeggiated in measure 83 of the third impromptu) as the tonic triad at the
opening of the fourth impromptu. Yet, as we have seen, the C-to-Eb con-
nection between the first two impromptus may be understood as III-V in
the goal key, Ab Major. Eb’s deviation into minor quality during the second
impromptu (which facilitates the third movement’s Gb) is complemented
by Ab’s emergence out of minor quality during the fourth. The minor cloud
that wafted over the set during the second impromptu does not fully dis-
sipate until well into the fourth impromptu, after the Ab Minor opening of
the A, section.
230 Harmony in Schubert

During the fourth impromptu’s initial A> Minor region, Schubert initi-
ates a local ascent in minor thirds that reprises the broad c—E}/e>—G} tonal
path of the first three impromptus. From the A> minor tonic, C> major
emerges in measure 13. Its quality shifts to minor at measure 19 (where
Schubert introduces enharmonic spellings), and we anticipate that Eb> [D]
major will sound at the downbeat of measure 25, as shown in 10.14 (Model
1). The return of Cb [B] minor at that point signals a rejection of that trajec-
tory (similar to the backtracking encountered in “Die Einsiedelei” [4.27]),
and the progression III>-°-IV-V$-> in Ab Minor, as shown in 10.14 (Model
2), concludes the section. (Observe that, after a wobble to E>) in measure
19, diatonic E} is restored with the arrival of III's 6 phase in measure 27.)

Fisk clearly delineates the tonal trajectory from A} minor through C)

major and C) minor to a potential continuation with D major without
noting the parallel between this microcosmic trajectory and the macro-
cosmic trajectory of the first three impromptus (p. 74). His assessment
that at measure 27 Schubert “returns . . . to the opening A> minor” is
a view in keeping with his literalist perspective, whereas I regard the
C} mediant as a conquest on the march to the dominant that need not
be ceded. I suggest instead that the Cb [B] root of measures 13, 15, 17,
19, 21, 23, and 25 proceeds to D> during measure 27. The intervening
A} [G#] at the downbeat of measure 27 functions as the 6 phase of an
asserted III°° (unfurled). This shift performs its characteristic role as
intermediary between two chords related by step — here III and IV.
(Compare with measure 9 of “Die Einsiedelei,’ analyzed in 4.27.)

The Ab Major tonic region that begins in measure 31 is distinguished by

an artless simplicity and uncommon repetitiveness. It is as if, after all the
travails that have preceded the set's long-awaited A} Major tonic arrival, a
time for rustic celebration is called for. The Eb<Ab fourth that transpires
in the broad melodic contour of measures 31 through 38 is the basis for
the E><F<G<Ab linear ascent stated first in measures 39 through 42 and
repeated again and again thereafter, with the simplest of harmonic accom-
paniments (I*”°)-IV-V’-I). This motion eventually whirls up into frenzy,
ascending higher and landing on IV in measures 62 and 64. Yet V7 again
directs the progression back to tonic. (This IV-V’, likewise with a 95-8
embellishment of V, will recur in a minor hue as the foundation for the
impromptu’ B section, as shown in 10.13.)
Some business remains, however, since the descent in the earlier A}
 Four Impromptus 231

Model 1

Model 2
m 1 13 19 27 28 29 30

5 dems soe!
A

, : Less

LN
|
ST y] \
Ny aN

Abd Minor: I = —_——— <3 a

10.14 Analysis of Impromptu in Ab Major (D. 899/4), mm. 1-30.

m 76 Td 78 79 80 83

5 Some ee |
: 5 os = ca eo @ eo

5 5 ° 3 g ° e = ——

ab - 2a Ps
: Z KATES
70g s .

A> Major: I" + IV Vere aol

—— 5—6 i= If

10.15 Analysis of Impromptu in Ab Major (D. 899/4), mm. 76-83.

Minor region was interrupted [10.14, Model 2]. When tonic returns in
measure 72, Schubert adopts a more serious tone to accomplish a descent to
1, modifying the melody first stated in measures 47-50 to retrace and con-
clude the linear path from 5, now in a major-mode context. (This passage’s
consequent phrase and the initial restatement of the celebratory Eb<Ab
fourth are displayed in 10.15. Note the close alliance with the passage
Zon, Harmony in Schubert

displayed in 10.14, Model 2.°° There, a 5-6 shift was employed during the
III chord within the III-IV-V progression. In this latter passage, a similar
shift is deployed instead during the IV chord.)*! Closure is achieved dually:
through the traditional descent to 1, and through ascent to 8. The registral
diversity of the various repetitions in the score extends that sense of closure
over a wide span, culminating in a sweeping descending arpeggiation from
soprano to bass in measures 99 through 104. This is, of course, the passage
that will recur (extended an additional octave in the bass) in the A, section
to end both this impromptu and the set of four impromptus.

“The D)-major climax [measure 64] sets the stage for a new, more ful-
filled variant of the melodious A}-major theme in the right hand” (p. 77).
In the sentences that precede this quotation Fisk twice asserts that D}
major has been tonicized. I instead hear

Ab Major: I°-” IV not Db Major: V*7 I

Fisk does not mention the dominant Eb chord that follows in meas-
ures 68 through 71, which brings to the passage a sense that the local
progression that occurs repeatedly in the preceding measures has been
extended. Thus measures 39 through 71 come across as:

I-IV-V-I-IV-V-I-IV-V-IL-IV-V-I1...IV...V...

This dominant is interesting also in that it closely matches the dominant

chords from the second and third impromptus that Fisk quotes on p. 41.
(In each case a 9}-8 embellishes an inverted dominant.)
Fisk's words “more fulfilled variant” (referring to measures 72ff.)
warrant fuller explication, which I have provided in 10.15. The theme not
only expands upon the modest idea of measures 47-50; it also completes
the structure initiated during the Ab Minor region [10.14, Model 2].

Contrasting the numerous major subdominant chords that occur during

the A, section and reinstating the minor hue that has often tempered major
keys within the impromptus, the B section (Trio) emerges on the minor
subdominant, Db Minor (which Schubert spells as C# Minor), beginning
in measure 107. (See 10.13.) Given that the A, section will again proceed
from Ab Minor (measure 171), the shift to minor at this earlier point is
prudent. A small a,-b-a, form transpires within the Trio. The initial a,
section follows a coniadaonal course, moving from I to V viaa dominant’
emulating supertonic in the tonicized key of Db [C#] Minor. Three models
 Four Impromptus 239

Model 1 Model2
m. 107-112 x 120 121 m. 107-112 117-118 119 120 121
“A

3 N> 3 2

Sa ee ee
a Ee aa ee
eavinar io ==" lV Ck Minor Ie" ie Vv
9
(= #VI7 )

Model3
m. 107-112 113-114 115-116 117-118 119 120 121

2
ee oe ee
ty ; 7

C# Minor: B

10.16 Analysis of Impromptu in Ab Major (D. 899/4), mm. 107-121.

of its construction are shown in 10.16. In Model 1, the essence of the

passage is displayed: a 5-6 shift on tonic employs A# rather than diatonic
Ag because the succeeding II chord will be of dominant-emulating major,
rather than diminished, quality. In Model 2, tonic’s 6 phase is modified in
two ways: Cx, a chromatic connection between the tonic and supertonic
roots, displaces C#; and ninth B sounds instead of root A#. The diminished
third between Cx and E hints that Schubert is opting for an A#>D#, rather
than an A#—D#, connection. That option generally would occur with E in
the bass. In Model 3, that preference is facilitated through a voice exchange
between E and C# during the initial tonic prolongation. Note in the score
that an elision occurs: where we expect I in first inversion (measure 115),
passing note A (which connects the expected but absent G# and B) has
already embarked upon its mission. Listeners may be intrigued by the
juxtaposition of what sounds like A major and its E dominant seventh in
measures 115 through 118, hinting at a tonicization of C# Minor’s submedi-
ant. However, that mehrdeutig potentiality is left unrealized. Model 3 docu-
ments my reading of how the passage does work, not how Schubert might
have proceeded.
234 Harmony in Schubert

ee
ae SS AR

BEAL =
or eT td “eo 2

we LE ——
SSR ee

9
Gt Dts Gt3x

C# Minor: V#

10.17 Analysis of Impromptu in Ab Major (D. 899/4), mm. 123-129.

“Because Ab major is not given from the start but is instead only
achieved, and because C# minor can be heard as a negation of Db major,
the constituent of Ab major on which the music has especially depended
for its assertion, the music of the trio suggests a denial of that achieve-
ment .. . Insofar as one musical passage can account for another, the trio
explains, at least in part, why the Ab-Major Impromptu begins as it does”
(p. 77):
I agree with Fisk that there is a strong alliance between the Ab minor
and D} minor regions. But I shy away from the chicken-and-egg ques-
tion of whether the work opens in the minor tonic because the Trio will
open in the minor subdominant (as Fisk suggests), or whether instead
the Trios minor subdominant facilitates the return of tonic in a minor
guise at the outset of the A, section.

The Trio’s b section hardly wavers from the dominant. In 10.17 the
essence of the structure for measures 123 through 130 (which is repeated
in measures 131 through 138 with slight variants) in a normalized register
is shown. A similar construction transpires during the b section of “Erster
Verlust” [4.38].
Though the Trio’ a, section reprises the opening measures of the a,
section (with a temporary C# major inflection), its full harmonic trajec-
tory contrasts that of the earlier section, in that it closes on the C# tonic.
The progression’ evolution is charted in 10.18. In Model 1, a chromatic
glide connects I# and II>*. (In such progressions the spelling of the inter-
nal chord(s) is of no consequence - Cx-Ex-Gx would be equally viable,
though more cumbersome.) This model’s supertonic proceeds as II>V.
 Four Impromptus IS

m. BS si (ISB) Ts es si m. ISOS ISIN GS53) els ee SoS,

C7MiMinor, eC ) WsSetees
Ve=> 1 C#Minor; Te ( ) m2 Vox?
Seal
1

Model3
m. 139-144 145-146 147-148 149-150 151 151-152 153-154 155 156 157
Nn

& ZRN> aS

te is es 43 a 3 = = 7)
e So 4 # 3 +e. 3. 7 53
ee
C# Minor:
, I#
;Sa this
ake
pe Sse

10.18 Analysis of Impromptu in Ab Major (D. 899/4), mm. 139-157.

The opening tonic’s major quality is part of a trajectory of major chords

leading to the major supertonic. In Model 2, II->V has evolved into II™V,
and the bass pitch of this modified II is now the traditional one: the sixth
scale degree, A. Finally, in Model 3 we see just how far Schubert persisted
in pursuing the same progression as in the a, section. (Compare with 10.16,
Model 3, measures 107-118.) As late as measure 150, it seems that we are
headed directly towards a D# supertonic chord. Yet at the last moment
(in fact in collision with the arrival of bass D4), Ey ascends to Ef, thereby
restoring the pitch that initiated the progression in the soprano. In this
model the II chord has evolved further: its final form lacks the root and
incorporates a ninth. Observe also that in this progression there are two
unrealized mehrdeutig potentialities: not only A major and its dominant
seventh in measures 147-150 (a repetition of the same potentiality we
encountered in the a, section), but also D major and its dominant seventh
in measures 151-154.
The remainder of the impromptu follows the plan shown in 10.13. The
background descent from 5 to 1, which accords with the structure dis-
played in 10.15, concludes in measure 250. No further wobbles impede the
straightforward celebration of goal Ab Major that concludes not only the
impromptu, but also this superbly integrated set of four impromptus.
11 “Auf dem Flusse” from Winterreise (D. 911, no. 7)
with David Lewin

David Lewin’s multifaceted article on “Auf dem Flusse” from Winterreise

offers a bold and provocative perspective on Schubert's transformation of a
poetic text by Wilhelm Miiller into a musical composition.! A sophisticated
analysis of the work’s pitch content informs his discussion. Our contrast-
ing readings have led us to divergent conclusions regarding how Schubert
interpreted Miiller’s text, a topic I will address after presenting a careful
assessment of the work’s tonal structure. Lewin also cites and addresses a
graph of the lied by Schenker (FC, fig. 40, ex. 2). A more detailed graph is
available for study in Schenker’s Nachlass.’
Lewin’s table 1 offers an account of how the text is traversed over
the course of the lied’s seventy-four measures. A few revisions are in
order. First, the distribution of the text’s five stanzas over what Lewin
regards as eight strophes becomes vague after strophe 5. It is analytically
pertinent to state decisively that stanza 5 is spread between strophes
5 and 6 (that is, measures 41 through 54) and then repeated in full in
strophes 7 and 8 (measures 54 through 70). This information can be
gleaned from the table's “text” column, but neither the “stanza number”
nor the “key” column would lead one to suspect that the end of strophe
6 (which I would place at the downbeat of measure 54) incorporates
a powerful cadence on tonic E Minor. Second, I quibble over Lewin’s
calculation of strophe durations. The musical entity that begins on the
upbeat to measure 48 extends into measure 54; and that which begins
on the upbeat to measure 62 continues into measure 70. Though perhaps
the term “strophe” is not appropriate for what occurs here, certainly the
number of measures occupied by these entities is seven and nine — not
six and eight, as Lewin states in his “duration” column. I regard measures
54 and 70 as structural overlaps, being concurrently the conclusion of
one initiative and the beginning of the next. Lewin instead acknowledges
only their roles as beginnings. Finally, I will attempt to demonstrate
below that strophe 6 is not composed in the key of G# Minor.
I propose a different structural representation of the work in 11.1. Two
essential features of this analytical perspective, suggested by the contrasting
borders in my diagram, are missing from Lewin’s conception:
 “Auf dem Flusse” from Winterreise 237

Prelude mm. 1-4 Establishes ae E Minor

Stanza 1 mm. 5-13 ja)

a E Minor Focus on
river scene

Stanza 2 mm. 14-22 Same content as E Minor Focus on

stanza | river scene

Stanza 3 . ee Focus on
love inscription
Stanza 4 ; Same content as Focus on
stanza 3 love inscription

Interlude mm. 38-40 Extension of V, Reversion to E Minor

concluding in silence

Stanza 5 mm, 41-54 trate E Minor Focus on

protagonist/
completing content river affinity
initiated in stanzas 1
and 2

Stanza 5 mm. 54-70 E Minor Focus on

: (repeat) merging elements of tee

: stanzas 3 and 4 and first ee
: statement of stanza 5

Postlude mm. 70-74 Extension of I E Minor

11.1 Structural diagram for “Auf dem Flusse” from Winterreise (D. 911/7).

(1) The first statement of stanza 5 completes the structure initiated in

stanzas 1 and 2; and
(2) The second statement of stanza 5 not only reprises the structure of that
stanza’s first statement, but also incorporates structural elements from
stanzas 3 and 4.

Stanzas 1 and 2 and the initial statement of stanza 5

Music’s most basic form-defining device is the antecedent/consequent

relationship. A musical idea is first stated incompletely, generally leading
to the dominant (half cadence). Then the material is restated and recon-
figured so as to proceed through the dominant to the tonic (full cadence).
238 Harmony in Schubert

Model 1 Model 2
m.

fi —— i = & i—8
10—9
E Minor: 18-7 IV# Vi-—# BMinorl'= sas" Ve Ne

11.2 Analysis of “Auf dem Flusse” from Winterreise (D. 911/7), mm. 5-13.

In the context of a narrative, the antecedent portion may correspond to an

inconclusive state (here, the protagonist’s taking an initial view of a desolate
wintry river landscape), while the consequent portion may represent reso-
lution (the protagonist coming to understand that this landscape has much
in common with his own mental state).
3-3
An intriguing realization of |~,,occurs in measures 5 through 13 (stanza
1). (Model 2 of 11.2, which will be discussed in detail below, displays its
structure.) Despite the emphasis on D#-F#-A# in measure 12, dominant
B major emerges as the culminating element of the stanza. As in the final
measures of “Erster Verlust” [4.34], the piano completes this progression
without vocal participation, as if the protagonist would prefer to remain
outside the bounds of convention. With similar textual content stanza 2
reprises this structure, and thus the dominant chord of measure 22 intensi-
fies the sense of irresolution, which for now is allowed to dangle as the lied
enters the domain of E Major (to be addressed below). Only at measure 41
is the initial content brought back into play, in tandem with text that picks
up the thread from stanzas 1 and 2. Listeners should expect — this time - a
resolution on tonic E Minor: interrupted progressions generally will even-
tually attain their goals. Such a resolution does occur in measure 54, as
displayed in 11.3. Whereas the first two stanzas each traverse the path from
I to V# in nine measures (thus eighteen measures in all), in stanza 5 that
structure is extended to goal I in fourteen measures.
The two models of 11.2 juxtapose contrasting chordal support for the
arrival of2(F#) in E Minor: D-F#-A, or D#-F4-A#. Model 1 shows a context
like the one we explored in 4.3. There is a fork in the road: though IV may
proceed directly to V, it often instead pursues its innate tendency to descend
 “Auf dem Flusse” from Winterreise 239

m. 41 46 47 48 49 ou 58 54
50 51
(3 2) t

é abal ale
LS
’
———
ats.
CPN
:
W
ped
SN
Sea ane

~ SS
i 10 —9 9# 7 8 7t - 8 7
E Minor: I oeaaG
5 ——os
I esx WES 64 5x i 6—> I
bf #

11.3 Analysis of “Auf dem Flusse” from Winterreise (D. 911/7), mm. 41-54.

by a perfect fifth (here enhanced through the presence of major third C4),
consequently entering V-space at V’s upper-third chord. When root B
arrives, diatonic D is transformed into chromatic D#. (Compare with 1.8
and 1.11.) Model 2 displays a more daring alternative, which Schubert
employs in the first two stanzas: its A¢—>D# bass succession falls outside of E
Minor’s diatonic tonal space. I interpret the A#” chord as an evolved super-
tonic. The broad F#<B (II<V) root succession functions similarly to the D<G
(V<I) succession in 1.30 (or like I<IV in 1.5). In this case we can discern
a parallel tonal universe shaped by the upper-third chords: with E<F#<B
serving as the foundation, Gf<A#<D# floats a third higher (or sixth lower).
Whereas the G# and D# chords either extend or introduce spaces defined by
their roots a third lower, no II root sounds in the vicinity of bass Ad.
My assertion that the structure of stanza 5 (measures 41 through 54) is
derived from and completes that of stanza 1 is fleshed out in 11.3. The D#
minor chord of measure 47 occurs in a tonal location more often occupied
by diatonic D major. (See 11.2, Model 1.) Yet despite its chromatic modifi-
cation it nevertheless retains some of the innate potentiality of the subtonic
to proceed to the mediant (here raised subtonic D# to raised mediant G#).
Though Schubert offers some tantalizing signals, that outcome does not
occur. As with the D subtonic, the composer may either ascend a fourth to
the mediant (completing a circular progression, E-A#-D#-G#); or, descend
a third to the dominant (as a chromatic variant of the dominant’s upper-
third chord). The D# chord embraces two features of dominant emulation:
its quality is adjusted to major (D#-Fx-A# in measures 49 and 51), and it
undergoes two rounds of 3-°-? embellishment similar to that which extends
240 Harmony in Schubert

the B dominant chord in measure 13. (Note in the score the distinctive
4-3 suspension during the return of the ;in both contexts.) The § chord is
unfurled, so that G# and then B reside in the bass (measures 48 and 50),
instead of D#. This is the passage that Lewin interprets in the key of Gt
Minor. In his defense, the bass arpeggiations of measures 41-42 and 43-44
occur during a tonic prolongation (in E Minor), whereas those of measures
48-49 and 50-51 do not (in Lewin’s alleged G# Minor). To interpret the
passage in that way misses the point of root D#’s prolongation from measure
47 through 51 and continuation exactly as in the opening stanza: down a
third to dominant root B (now with the protagonist’s participation — he has
made up his mind what he thinks about the scene before him). The unfurl-
ing of the $on Dé is echoed by a similar unfurling of the {on B in measure
53 (here to §position only, not also to :position).° In stanza 1 (measure 13),
that ° occurs over prolonged bass B, with the inner-voice melodic descent
B>A>G in sixteenth notes leading into the ¢. That very descending third,
in augmented time values, is transferred to the bass of measures 52 and 53.
The vocal melody, both in stanzas 1 and 2 and in stanza 5, generally
pursues an ascending course among the inner voices of the chordal struc-
tures displayed in 11.2 (Model 2) and 11.3. Keeping in mind that the tenor
voice sounds an octave lower than notated, only occasionally does the vocal
line attain the soprano position shown in the examples: F# in measure 46,
and the span from Gb to E in measures 53 and 54.
Schenker’s graphs interpret the span from measure 5’s E through measure
53’s Gy as a large-scale ascending arpeggiation of a tenth to Kopfton 3. Lewin
rejects that reading, as do I. Whereas Schenker hears the tonic of measure
41 leading to an inverted tonic in measure 53, I suggest that dominant root
B arrives already in measure 52. In my view, Lewin misses the unfurling of
D# (= BS) in measures 48 and 50, while Schenker misses the unfurling of
B§ in measure 53. Both passages evolve out of measure 13, the foundation
for my interpretation. In all three cases, the ° arrives during or at the peak
of a crescendo.
Clearly my reading is not compatible with the notion of a “secret E-major
deep structure” (p. 51) that Lewin entertains in tandem with his interpre-
tation of bass (root?) G# within his strophe 6. His naming of G#-B-D#
(measures 48 and 50) as the “dominant parallel” — a term borrowed from
Hugo Riemann - of E Minor’s dominant B-D#-F# (measure 42) highlights
the deep divide between us regarding the structure of measures 41 through
54. Whereas he connects those chords with a slur in his fig. 1, in my reading
the B-D#-F# chord of measures 42 and 44 plays only a local role. (It does not
appear in 11.3.) Lewin’s assertion that bass G# in measures 48 and 50 serves
 “Auf dem Flusse” from Winterreise 241

as a chordal root - even as a tonic — is a notion that should be distinguished

from the equally problematic notion that the G# chord is compellingly
related to the earlier B chord.
Several details of Lewin’s reductions warrant review as well. In an appen-
dix to the article, he himself expresses some concern about his omission of
bass G# (measure 11) from his fig. 2.4 In any event, his reading is consistent,
for G# is omitted in measures 20 and 46 as well. (I regard G# as structurally
deeper than the preceding A# — which he includes in his reduction — in each
of these contexts.) His fig. 5, the most reduced “picture” of the lied that he
presents, seems to me bizarrely misleading, for the sparse content and rig-
orously positioned bar lines make it appear that stanzas 1 and 2 (the figure’s
first “measure”) end ona tonic chord and that the first statement of stanza 5
(the figure’s third “measure”) ends on a dominant chord. In my 11.2 (Model
2) and 11.3, I suggest the exact opposite: that a central tenet of Schubert’s
organization of the lied is that the opening stanzas do not achieve closure,
whereas the fifth stanza does. Schenker’s graph may have misled Lewin,
since there the background dominant does not occur until measure 53. In
my reading, the first two components of the Schenkerian 3 2 I (3 2) 1 inter-
ruption structure (as in FC, fig. 21b) appear already by measure 13 and are
repeated both in E Minor and in E Major through measure 40. The post-
interruption return of tonic occurs at measure 41. At that point Schenker
has not yet attained his Kopfton 3.

Stanzas 3 and 4 and the second statement of stanza 5

The cadence on the tonic in measure 54 coincides with the final word of the
text (“schwillt”). Listeners finally feel a sense of resolution, which Schubert
had withheld until this point. Yet since each of the first four stanzas pro-
ceeded from tonic to dominant, in coordination with a melodic descent to
2; perhaps a single cadence is inadequate. The structure of stanzas 3 and
4 contrasts that of the preceding stanzas. In that the statement of stanza 5
that transpires in measures 41 through 54 extends the structure of stanzas 1
and 2 (and in itself nearly matches their combined duration), at measure 54
one may well sense that the full content of the first half of the lied (stanzas
1 and 2 plus stanzas 3 and 4) has been only half resolved. Schubert appar-
ently understood that to conclude at measure 54 would be inadequate. His
repetition of stanza 5, with altered musical content and an extension to
seventeen measures, provides a counterweight to balance stanzas 3 and 4.
Schubert avoids one potential correlation: he does not again call upon E
Major, the key of stanzas 3 and 4. The lied must not end on a hopeful note.
242 Harmony in Schubert

PONOE le ne | renee
Cana reee an
(= VI?)
7
(ye)
—_

11.4 Analysis of “Auf dem Flusse” from Winterreise (D. 911/7), mm. 23-30.

Yet the course he pursues in measures 54 through 62 reprises, in a minor-

key context, several distinctive features of those earlier stanzas, while also
maintaining a close alliance with the first statement of stanza 5. The magic
he works stems from his transposition of the content of measures 43 and 44
(E<B) up a third in measures 56 and 57 (G<D). Whereas the former leads
to a variant of the dominant’s upper-third chord (D# minor in measure 47),
the latter reaches a third higher, to the supertonic (F# minor in measure
60). That chord and the Bi, chord that follows reprise the F# and B’ chords
of stanza 3 (measure 25). These chords are shown in their original context
in 11.4. Within stanzas 1 and 2, as well as the first statement of stanza 5, the
tonic chord does not return once the progression is well under way (until
the final cadence in the first statement of stanza 5), whereas in stanzas 3 and
4 and the second statement of stanza 5 it does (in measures 26-28, 34-36,
and 62). In stanzas 3 and 4 the reinstated tonic takes on a minor seventh,
a dominant-emulating feature that presses towards IV (measures 28 and
36), which then continues to V (via a [V°~¢ in which the 6-phase chord has
evolved into A$-C#-E-Gk). The tonic of measure 62 is extended instead via
one of its chromatic upper-third chords, G-Bb-D, which coordinates with
the lied’s only presentation of a harmonically supported high 3 in the vocal
line. (See 11.5. Recall that the vocal G in measure 53 is not a harmonic
pitch, but instead a neighbor to F#, as displayed in 11.3.) In this intense final
phase of the lied, the path first traversed in measures 5 through 13 gives
way to a less chromatic structure. Now D, rather than D#, precedes domi-
nant root B. (Compare with the D major chord in 11.2, Model 1.) Tonic
 “Auf dem Flusse” from Winterreise 243

m. 62 64 65 66 70

E Minor: 13, 5 Vi—?# I

11.5 Analysis of “Auf dem Flusse” from Winterreise (D. 911/7), mm. 62-70.

E transfers authority to its upper-third chord, G (measure 64); that chord

leads to its dominant, D (measure 65); that dominant transfers authority
to the chord of which it is the upper-third chord, B (measure 66). These
unfoldings of minor thirds in 11.5 correspond to the unfoldings of major
thirds in 11.3.
Lewin omits the supertonic root F# of measure 25 in his fig. 2, thereby
concealing what I regard as a key feature in the correlation between stanza
3 and the second statement of stanza 5. In addition, he and I assess the
hierarchical arrangement of the progression’s chords differently: he regards
root B in measure 24 as beginning a dominant prolongation that extends
through measure 25, while I interpret the bass descending sixth E>G# (with
internal B) in measures 23 and 24 as an expansion of tonic, a chromaticized
I>-°. That reading is supported by the recurrence of the G#-B-E# chord of
measure 24 as C#-E#-G#-B in measure 59.
Though Lewin does not pursue a rigorous Schenkerian analysis of the
lied, he does probe Schenker’s reading in some detail, in particular assess-
ing Schenker’s placement of Kopfton 3. It is unfortunate that the vocal high
G of measure 64 does not appear in Lewin’s fig. 2. Granted, most readers
not well versed in the Schenkerian perspective (as distinct from Schenker’s
analyses of “Auf dem Flusse,’ with which I am in disagreement) may
wonder why I fuss over a sixteenth-note G squeezed into the last moment
of this measure. Yet only here, the sixth time that a stanza has been set in
this work, the melody always pressing ever higher beneath the surface, does
the vocal line finally attain this high 33 supported by tonic’s upper-third
chord. G (3), present since the first chord of the introduction and the initial
melodic pitch in my structural models for all the stanzas, finally sounds — at
the last possible moment, with but one-fourth of one beat to spare - within
244 Harmony in Schubert

I-space in the upper vocal register before the final descent to 1! Even in the
postlude this upward striving does not cease: E<G<B in the upper piano
line echoes one last time the vocal melody of measures 5 and 6, while the
bass ascends a full octave, E<G<B<E.

Miiller’s, Schubert’s, and Lewin’s stories

Miiller’s questions concerning a protagonist/river affinity in the final stanza

may be rhetorical, or they may be answered either in agreement or in nega-
tion. The text does not assert definitively whether the protagonist's heart
is or ever again will be teeming with life, like the water's flow beneath the
icy surface. Yet I find it difficult to accept that Schubert had anything other
than an affirmative answer in mind when he set Miiller’s poem.° Lewin, in
contrast, takes the negative view: “the poet’s heart is frozen solid forever”
(p. 57). Were I to film the scene, I would take medium-range shots for
stanzas 1 and 2, showing a solitary figure assaying a frozen river landscape.
We do not know what he is thinking, but the scene is forlorn, and he seems
lost in contemplation. For stanzas 3 and 4, I would zoom in, so that what
the character we spy carves into the ice can be made out. We learn that
he has suffered a loss. He is careful and precise, as if he were inscribing a
tombstone. Backing off again for stanza 5, we see the silent, immobile figure
contemplating the inscription - the grave. Though stillness prevails, the
soundtrack (Schubert’s lied) reveals that this man is not indifferent to his
loss. Under his glacial demeanor a torrent still flows.
The music’s deep melodic structure represents the river's icy crust:
Stanza 1 Stanza 2 Stanza 3 Stanza 4 Stanza 5 Stanza 5 (repeat)

G>FR G>FH Gt>FR G#>FH G>FHE>E G>FH>E

Under this surface the vocal melody rises (“schwillt”, the final word of
the poem) ever higher from its initial low E, ultimately merging with the
surface and completing the structural descent (G>F#>E in measures 64
through 70):
Stanza 1 Stanza2 Stanza3 Stanza4 Stanza 5 Stanza 5 (repeat)

Fifth (to B) Octave (to E) Ninth (to F#) Tenth (to G)

Just as the river's frozen surface conceals a vigorous flow underneath, so
also the scene of an isolated figure staring downwards in silence conceals
the tumult within.
12 | Piano Sonata in Bb Major (D. 960), movement 1
with Richard L. Cohn

As analysts we depend mightily upon inherited strategies and techniques,

components of which can be traced back to the time when the music they
address was conceived. Such tools require able practitioners, of course.
Even a virtuoso analyst may be challenged by an especially complex or
novel work. When no existing tools seem adequate, new strategies may
emerge. In his study of the first movement from Schubert’s Piano Sonata
in Bb Major,' Richard L. Cohn cites various authors who, employing a wide
range of conventional strategies, offer an assortment of analytical insights.
In his view they all come up short, and so he develops his own “instruments
for gazing at tonality.’ Though his work antedates Harmony in Schubert, I
suspect that Cohn would be no more satisfied with the tools that I employ
than with those that he cast off during his own study of Schubert.” Thus I
shall defend my project through careful comparison of what his and my
tools reveal.

The exposition’s tonic region and transition (measures 1 through 79)

The exposition’s tonal plan engages the space between tonic Bb and domi-
nant F. A transition, triggered by the diminished seventh chord of measure
45 and continuing through measure 79, connects these two regions. In
outline the exposition pursues the following course:
m. 1 45 74 78 80

Bb Major: I () Is V
Cohn dissociates himself from such analytical representations. His state-
ment, “if a Schenkerian model were to label harmonies, it would simply
measure distances between roots along a diatonic spectrum,” is followed
by the advisement, “to do so here would be problematic” (p. 220). Cohn
elects not to do so, and in fact his analysis pointedly contradicts the reading
presented above. In his quest to “liberate the triads from their commitment
to a particular diatonic context” (p. 220) he reads measures 72 through
80 (displayed in his example 2) in a way that makes Schubert's harmonic
practice seem less reliant upon eighteenth-century conventions than |
246 Harmony in Schubert

Cohn’s analysis

ee 80
70-71 73 74 74-75
eae = =o =e Her S — = = -

Bb+ E+ Aas ie

My analysis
m 1-39 72 73 74 78 80
(2) 2
ee
6
ee az
is
a
cae
Z

e 8 ye
Bb Major: I° ig ; A
9b
Gavi i )

Bb+ G+ C+ F+

12.1 Analyses of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 72-80,
preceded by tonic (mm. 70-71 or mm. 1-39).

think it actually is. Our views are juxtaposed in 12.1. (To facilitate compari-
son I have appended Cohn’s neutral letter notation for my reading of the
progression below my Roman numerals.) I assert that the transition (the
beginning of which is for now marked by a hairpin symbol) leads eventu-
ally to a diminished seventh chord (measures 72 and 73) that functions as
a chromaticized 6 phase of the initial Bb tonic. (Diagonal lines mark a voice
exchange - ;) to 5’ - in my analysis.) My reading of root G is incompat-
ible with Cohn’s interpretation of bass Ey in the middle of measure 73 as a
root. In my view that Ek serves as a passing note between chord members
F and D.’ (Cohn’s reduction is misleading. No Ey sounds against bass D or
against bass C in measures 73 and 74 of Schubert’s score.) Because the fol-
lowing C chord’s fifth G will be embellished by sixth A (in a straightforward
3 construction that Cohn likewise does not acknowledge), the G chord’s
ninth Ab is respelled as G# in measure 73 to direct that strand upwards to
neighbor A.” Cohn posits roots A and then F for a passage that I regard as
the extended prolongation of C (with embellishing $). If B}-C—F consti-
tutes the essence of the exposition’s tonal plan, as I contend, then Cohn’s
 Piano Sonata in Bb Major 247

instruments for gazing at tonality have led him very far astray: not only is
the linchpin C chord misconstrued, but in addition the F chord’s arrival
is misplaced by six measures (measure 80, not measure 74).° Though I
am myself vulnerable to the accusation of interpreting the passage based
on pre-formed notions about how sonata expositions are “supposed” to
go, it appears to me that in adopting his novel perspective and in concur-
rently discarding the accumulated wisdom of scale-step theory, Cohn is
vulnerable to an equally serious accusation: that his groupings of pitches
into chords and chords into progressions sometimes are at odds with what
Schubert actually wrote.
The capacity for tonic’s 6 phase to incorporate chromaticism and added
dissonance (often fostering dominant emulation directed towards II) has
been a persistent theme in Harmony in Schubert, starting with 1.1. Thus
my proposed relationship between tonic Bb and measures 72-73 should
not be surprising. (The intervening measures will be explored below.)
Yet Schubert's deployment of a broad range of 6-phase chords to initiate
various harmonic motions within the exposition is remarkable. It appears
that a central concern of his agenda was to touch upon all three chromatic
variants of tonic’s 6-phase chord within both the exposition and the recapit-
ulation. Cohn arranges these and the remaining chords in an eye-catching
geometrical design (his fig. 2) that makes my catalogue of tonic’s diatonic
and chromatic 6-phase chords [2.17] appear downright dowdy. Yet it -
unlike Cohn’s figure, which dissociates the G and Gb [F#] 6-phase chords
— seems almost a blueprint for Schubert's gradually more adventuresome
initiatives within the exposition.
In 12.2 all four of 2.17’s two-chord 5-6 successions transposed to Bb
Major are presented, juxtaposed with reductions of passages from the
sonata. In the diatonic version Schubert elects to employ two neighbor-
ing notes concurrently — not only the model's 6, but also a 4 — in a tonic-
prolonging initiative (measures 1 through 6). The G of measure 5 was
already touched upon in measure 4, and it will recur - ominously trans-
formed to G} - during the dominant prolongation of measures 7 through
9. In Chromatic Variant #1, By works in the service of dominant emulation,
pushing strongly towards the supertonic. That is exactly how Schubert
employs the device in measure 14. Observe that the 6 phase is unfurled,
so that the 6 (G) sounds in the bass. Schubert leads into this pitch via
the descending line B}>A}>G. The ensuing II chord leads directly to V’,
followed by a cadence on I in measure 18.’
That firmly established Bb tonic serves as the foundation for the next
5-6 motion, this time corresponding to Chromatic Variant #2. Observe
248 Harmony in Schubert

Diatonic
m. 1-4 5 6

Gr ———4 io Se
ire — ee
Bb Major: 13 peal yaa

Chromatic Variant #1
m. 10-13 14 i)

6 —s e | = ~ a

= a ; ——— a)

a S
Bb Major: 13 2 I
(= VIf)

Chromatic Variant #2
m. 1-18 19 20-21 34

o =e 6 g ee Bo
CE PES eee

(? b S bf te =

ae
;
# 95
Bb Major: 13; \ air ree ee
(- bVI*) *
Chromatic Variant #3

ee See
m. 39 46 48
ye 28 he

l ee

. oe “y —

Sie

' 5———6
Bb Major: I}; 2

12.2 Diatonic and chromatic configurations of I5-° and examples from the Piano
Sonata in Bb Major (D. 960), mvmt. 1, mm. 1-48.
 Piano Sonata in Bs Major 249

that a descending third recurs in the bass of measures 18 through 20 (now

By>A>>Gz, filled in by chromatic passing notes), suggesting that Schubert
regarded the G and G) chords as variants of the same entity. (Cohn, in con-
trast, regards G+ and G— within B) Major as “Subdominant” in character,
and G>+ and Gp~ as “Tonic” in character. See his fig. 4.) Not only is the Gb-
B>-D> chord further removed from tonic By Major than was G-Bs-D, but it
is prolonged for fourteen measures. Though its dominant-emulating poten-
tial is directed towards the lowered supertonic (C)), Schubert proceeds to
another easily attainable supertonic alternative, that with raised third (E:)
and lowered fifth (G>). As does the lowered supertonic, that chord targets
V, whose arrival in measure 36 (with y embellishment) coincides with the
return of the opening theme. The tonal and thematic parameters do not
realign until measure 39,° at which point Schubert restates the 5-6 initiative
in the version of Chromatic Variant #1.
An alternative analysis for measures 20 through 39 warrants considera-
tion. My preferred reading, which has informed my comments above, cor-
responds to 12.3, Hypothesis 1. A broad traversal of the B>-F-B> “sacred
triangle” (Schenker’s term) transpires during the movement’ first thirty-
nine measures, which are divided into three distinct regions, separated by
bar lines in the example. From this perspective the middle region’s soprano
D> is a chromatic passing note leading to C in measure 38. (That C is locally
embellished by neighbor D:.) The alternative analysis (Hypothesis 2) is
linear in character, not harmonic. Tonic’s 6 phase (in chromatic variant
B>-D>-G») reverts to B>-D-F in measures 36 and 37, assisted by Es.’ That
restored tonic, in ?position, coincides with the return of the opening theme.
In this perspective the D> of the middle section is a wobbly note. There are
good reasons why one might prefer either Hypothesis 1 or Hypothesis 2.
Thus I acknowledge them both.
The cadence that we expect at measure 45 is elided. A diminished
seventh substitute leads us into the tonic region’s final and most remote 5-6
initiative, corresponding to Chromatic Variant #3. (See 12.2.) Observe how
bass Ab [G2] (measures 46 and 47) again leads the progression downwards
to G> [F2]. This 6 phase is so remote that continuation to II in By Major
becomes doubtful. How will Schubert proceed from that point?
In 12.4 my analysis from 12.1 is tidied up. (The voice exchange and
the 6 embellishment are omitted, and registers are normalized.) The only
significant addition is tonic’s upper-third chord (D minor) at measure 68.
It is displayed as an intermediary between the 5 and 6 phases of I°*. This
model represents a highly conventional tonal structure. Yet how Schubert
traverses the span between the tonic and its upper-third chord is anything
250 Harmony in Schubert

Hypothesis 1
21 34 37 38 39
m. 2
3

eer
xe =
:
——— 4
cP

= SS Es:
i
:
Ob
5
5 5)
6>b WeEe as
Ss
Ves I
13:
Bb Major:
bVI’) :
Hypothesis 2
3 3 3

==
e a 7.
See:
Oo Do

Bb Major: 13, :: :

m. Dh Soto ih TN AES sy. TG US 21 22 23 24 25 34 37 38 39 4041 42 >

3 3 3
N N |
ba 5
- rz tb i

"3 D 4 ae eer He = =: a Pe
G: s = = i —! See
e SS

b az
ay;
Bee = 2 : = = = A

Sa se hee Sea EF ai

12.3 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 1-45.

but conventional. Consider 12.5, which begins with the passage quoted in
12.2 (Chromatic Variant #3). The descent of amajor third (B} to Gb) is now
followed by another descent of a major third (G})/F# to D). Consequently the
bass (with an implied E displayed within parentheses) traverses a segment
of the whole-tone scale.'° (Compare with 3.10.) Whereas the bass descent
from 8 to 3 normally would require five steps (three whole steps and two half
steps in a major key), here four whole steps accomplish the task. Eliminating
the diatonic bias of standard pitch notation by employing numerical nota-
tion (with C = 0), the descent could be displayed as 10>8>6>4>2.!!
Though my model might make this turn of events seem inevitable,
Schubert's writing in this passage is filled with misgivings.!* Consider the
progression of roots. B}-D}’—>G}[F#] will continue, in an obstinate descent
by major thirds, with A’->D. A-C#-E (without seventh) first appears in
measure 52, but Schubert scurries back to F#-A-C# within two beats.
Similar ploys prevail until measure 58, when A major is finally secured.
 Piano Sonata in Bb Major

m. 1-39 68 2B 74-79 80

ae ae
ee

cP J
Bb Major: 1° d & nd V
9
Ce)

12.4 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 1-80.

m. 1-39 47 48 58-67 68 70 is

ay YY fa get

bepei=2—be = ———
7 i ee

Bb Major: 1° | alieee slic

€ VI?)
Insteadof

|
se ees ae ae
SS
Bb Major: 18 z S
© VI4)

12.5 Analysis of the Piano Sonata in B} Major (D. 960), mvmt. 1, mm. 1-68, with
continuation to m. 73.

The next initiative — the addition of its minor seventh, Gk — leads to further
qualms. Potential A’>D successions in measures 60 and 64 are averted.
Only from measure 67 into measure 68 is that last hurdle on the path to
D minor overcome. Schubert’s descent of two major thirds from tonic
(interpreted free from diatonic bias as 10>6>2) leads to Bb Major's third -
not lowered fourth — scale degree."
AS) Harmony in Schubert

Cohn analyzes the opening tonic region and the transition to the domi-
nant region in his example 2, the latter part of which I have assessed above
in the commentary to 12.1. In the early measures his pursuit of what I label
Hypothesis 2 in 12.3 is reasonable, a choice that I also considered. Whereas
I interpret measures 35 and 36 as a harmonically functional succession
from altered supertonic to embellished dominant, within a progression
that regains tonic only in measure 39, his reading of measure 36 as Bb+
coordinates with the thematic placement and acknowledges that a harmony
may in fact be represented by a chord in second inversion. (Compare with
measures 6-7 and 15.)
Despite our disagreement regarding where Bb+ returns, I agree with
Cohn’s chord choices between that point and measure 68: B}+ F#—- A+ D-.
However, I suggest that Db+ [C#+] (measures 45 through 47, which Cohn
displays within parentheses but does not label) should be included as well.
(See 12.5.) This is the fourth time that Schubert has traversed the Bb>Ab>G())
melodic path in the bass. (Though bass pitch Ab [G#] sounds thirteen times
during measures 46 and 47, and in the end realizes a descending register
transfer, Cohn refrains from acknowledging it as a connector between Bb
and Gb [F#]: neither Bb nor Ab appears as a bass note in his example 2.)'°
Yet root D} plays exactly the same role here as will root A in measures 58
through 67. As explained above, to descend a second major third from Bb is
a daring and provocative act, transpiring outside the confines of normative
modulo 7 tonal space. Schubert's hemming and hawing adds considerably
to the duration of the A chord, as compared with the earlier D} chord. Yet
that durational distinction should not result in one chord being perceived at
a level that warrants analysis, and the other being relegated to a parenthetical
aside.'° Both could be analyzed; or, both could be placed within parentheses,
thereby emphasizing the deeper circular descent: Bb+ Gb—/F#— D—.!7
Cohn proposes a full-fledged return of a tonic Bb major chord in measure
70, while I suggest instead a linear connection between tonic’s upper-third
chord (D minor in measure 68) and its chromaticized 6 phase (G major in
measures 72 and 73). (See 12.4 and 12.5.) This G chord points (as G>)
towards the II, that I described above as the “linchpin” between the tonic
and dominant regions.

The exposition’s dominant region (measures 80-1 16)

The dominant chord that arrives at measure 80 [12.4] is still intact at

measure 116, just before the transition back to the beginning of the exposi-
tion or to the development. Though the dominant key, F Major, is tonicized
 Piano Sonata in Bb Major 253

(80 eS 84 85)
m. —-86 88. 469 90 91
A

2 GP

bot.

Papee
=
is
“Sigilei
ze

F ( J) By ec
12.6 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 86-91 (with
reference to mm. 80-85).

throughout this region, Schubert manages to employ the initial tonic’s

pitches (Bb, D, and F) frequently, either as a ° embellishment of F-A-C
(measures 81 and 87, the dominant-key equivalent of the © in measure
5), or as a IV chord within a progression in F Major (measures 85, 97,
106-107, and 111-114, as well as a special case, measure 91, which will be
discussed below). Eb, tonic F’s chromatic minor seventh, leads into the D of
the B}-D-F chord. In fact, F>Eb>D>C becomes a pervasive melodic feature
in this region: for example, in the interior of the texture in measures 80
through 82, and at the top of the texture in measures 82 through 85 (with
chromatic passing note Es connecting F and Eb).
The progression of measures 80 through 85 serves as the foundation for
Schubert’s continuation in the upcoming measures. ‘The large-scale root
progression is from F through Bb to C, as shown in 12.6. In a harmonic
context (as here) it corresponds to I-IV-V. The I chord takes on strong
dominant-emulating tendencies as A-C-E}-Gb (the G) spelled as F#),
directed towards IV.'* Between the consonant tonic and this dissonant
enhancement, Schubert employs tonic’s upper-fifth chord, which supports
chromatic passing note Ek (measure 83). (At the downbeat of measure 83
soprano F, which arises from the interior of the texture, is functioning as a
suspension into the passing Ex.)
When this progression recurs, with an expanded register, in measures 86
through 91, most listeners reasonably will interpret it initially as another
traversal of the tonic-to-dominant span. Yet as the progression continues
one comes to realize that now F-C serves as the beginning of the linear
initiative outlined in 12.7. The F—C span is followed by similar spans con-
necting E}—B} and then (D})-Ab. This Ab chord reverts to F (with added
254 Harmony in Schubert

m, 86-90 91 92 93 x 95-96 96

2
b = ;
g Da
ie" ee Sle, eos
es
Ses

gz a bo ee
we, es Z DG 2

Bb Major: V Sangean
(= FMajor: If=( er )

12.7 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 86-96.

dissonance). In the actual composition the content gradually recedes in

each successive statement, according to the following model:
Tonic Upper Enhanced Subdominant Dominant
Fifth Tonic 2
mm. 88-91 F & F Bb
mm. 92-93 (Eb) (Bb) Eb Ab Bb
mm.94-95 (Db) (Ab) (Db) Grae Ab
mm. 95-96 modified reiteration of | Ab

The chromatic Ab major chord makes the motion to tonic’s diatonic upper-
third chord earlier in the exposition (measure 68, as displayed in 12.4)
seem plain in comparison. Yet this “lowered” upper-third chord corre-
sponds to a very common Schubertian initiative: the subtonic-dominant
succession (see 1.11) is here transferred from the context of the dominant
to the context of a dominant-emulating tonic. Thus F—Bb is expanded into
E<Ab>F—>Bb. (The upper-third chord provides consonant support for the
introduction of the parent chord’s seventh, Eb.) IV duly arrives in measure
97, followed by V and I in the following two measures.
A new progression in F Major begins in measure 99. The content of
measure 100 is very closely allied with that of measures 10 through 15,
displayed in 12.2 (Chromatic Variant #1). The progression cadences on F in
measure 101. Schubert does with this progression exactly what he did with
the I-IV-V progression earlier in the F Major region: he uses it to gener-
ate successive cycles in a sequence. This time the sequence is ascending,
traversing the span I<IV. We have already explored its structure in chapter
2. (See 2.9 and 2.10.) Then, mimicking the VI in measure 101, closure is
attained in measure 109. The progression is repeated (without the sequence
connecting I and IV), cadencing again in measure 116.
 Piano Sonata in Bb Major AS

Cohn’ treatment of the exposition’s dominant region is cursory:

The final period of the exposition contains transient tonicizations of other triads
in the region ofFmajor: the implied Db major at mm. 94-96 (suggesting displace-
ments of C + D} and A — Ab), and the rhetorically pronounced but fleeting A
minor at m. 105 (F > E). (p. 225)

In several ways this statement might mislead readers. First, Cohn makes it
seem that Dp and A are the only tonicizations. I would think that, accord-
ing to his notion of tonicization (which I do not share), the E} major chord
of measure 92 (already highly evolved, as G-Bb-Db-F}[Es]), the G minor
chord of measure 103, the Ab minor chord of measure 104, and the B}
major chord of measure 106 would qualify as well, and that in the context
of F Major the remote Ab minor chord ought to be singled out for explora-
tion, rather than the diatonic A minor chord.'? Second, he relates both the
D> and the A chords directly to the F major chord. In my view, D} comes
about instead through an obstinate deployment of the sequential initiative:
F descends by whole step to E>, and then E} descends by whole step to Db.
This D} chord (which in any event does not sound in the composition) is
internal to the sequence. The more significant relationship is that between
F major in measure 88 and its upper-third chord, Ab major, in measure 95.
The A minor chord likewise is internal to a broader sequential initiative.
It is one of only two diatonic chords between tonic F and subdominant Bb
in 2.10, which also displays five chromatic chords, none of which Cohn
mentions.

Development (measures 118-215)

We have observed that ever more chromatic deployments of I?° serve as a

unifying and invigorating force within the exposition. I-° in the tonicized
dominant key, F Major, is similarly a crucial element of the development.
The C# minor chord introduced during measure 117 bis and prolonged
through measure 127 is an enharmonically respelled 6 phase of F-A-C,
presented in its third chromatic variant (F-A-C to Fb-Ab-Dpb). Without
returning to F-A-C, Schubert transforms this chord first into the second
chromatic variant (F-A}-D} in measure 149) and ultimately into its diatonic
state (F-A-D in measure 171).”° These nodal points are connected using
contrasting realizations of a common construct: the descending circle of
fifths.
The first circular departure from C# minor is aborted after just a few
chords. Though Schubert proceeds from C# to F# and then B in measures
256 Harmony in Schubert

139 140 145 146 148 149

m. LEZ. 131

} ibe ba a jl
ee ee

2 be Do a Oo be 40 92 bo

Db Gb CG F Bb Eb Ab’ Db

12.8 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 117-149.

118 through 125, that initiative quickly unravels: B leads back to F# and
a restored C# in measures 125 through 127.7! The second attempt is a
success. It is initiated by a division of the first descending fifth into two
thirds [12.8].22 In order for this circular progression to lead back to root D>
at its eighth chord, one of the melodic fifths must be imperfect. Schubert
takes the hit at C)—E, at a point where the diatonic pitch collection of goal
chord D} major begins to take hold. D}’s fortissimo arrival in measure 149
is followed by a full measure of bass D) stated alone and piano.
D} minor to Db major is a common tonal motion; Db major to D minor
(measure 171) is not. Schubert separates the latter chords using an exten-
sive progression organized again as a descending circle of fifths. This time,
no diminished fifth is deployed to shorten the route. As 12.9 shows, despite
the D} major chord of measures 149 through 155, Schubert leads first to D}
minor's diatonic upper-third chord, Fh-Ab-Cb, in measure 158. That chord
is enharmonically a major second - or, major ninth - above goal chord D
minor. The seven whole steps within the major ninth guide the circle of
fifths on its obstinate downward course. Observe that D-F#-A appears in
measure 161, and thus all three of the 6 phase’s chromatic variants sound
- in order of decreasing chromaticism (contrasting the increasing chro-
maticism pursued during the exposition) - between measures 117 and 161,
before the diatonic 6 phase, D-F-A, arrives in measure 171. Granted, 12.9
does not convey the extraordinary richness of Schubert's writing during
this passage. Some of the chords are emphasized as goals of dominant
emulation, a common occurrence during a descending circle of fifths,
while others are stealthily presented in the midst of further voice-leading
activity, such as the chord of measure 160, which fulfills the role of A minor
within the circle despite an added sixth, F. This tonal trajectory is pursued
in tandem with a gradual crescendo from pianissimo to fortissimo. Like the
repeated bass D} in measure 150, the completion of this circle is marked by
a repeated D-F-A chord in measure 173.
From this perspective a large portion of the development’s content is
 Piano Sonata in Bb Major IS

me L7, 149 158 160 161 162 163 164 165 166 167 168 169 170 171

rae
Cue a
5 Po-
eae —
he eae
T
a A Z + brea
ge she assrhe Z Hg

Be 5B po Fhe 4 (es =
a o O = fe fo be = Zs
ho oe
Fb/Ek TROD Yrsee (Cine KC at Bb Eb Ab Db/Ch Fe Bh Ek A D

12.9 Analysis of the Piano Sonata in B} Major (D. 960), mvmt. 1, mm. 117-171.

devoted to negotiating the chromatic spaces among the various versions of

dominant F major’s 6 phase, ultimately cleansing that chord of chromatic
accretions. The story outlined by the various 6-phase chords of the exposi-
tion and development correlates astonishingly well with probable elements of
Schubert's biography. Though we do not know with whom Schubert engaged
in sexual relations, we do know that such activity led to a serious health crisis
and consequent lifestyle adjustments. Could the exposition’s dalliances with
ever more chromatic versions of Bb major’s 6 phase correspond to Schubert’s
gradually more dangerous personal encounters?** Could the development’s
transition from the most chromatic of dominant F major’s 6-phase variants
to the sober diatonic D-F-A represent his tumultuous crisis and gradual
transition back to a life of restraint? Certainly any Schubert enthusiast must
wonder what was on his mind during his final years, and how those thoughts
are reflected in his compositions. It is not difficult to read measures 118/119
through 130, with their pervasive minor, diminished seventh, and bII chords
and the collapse of the circle of fifths progression, as a musical portrayal of
the somber period when Schubert was adjusting to the implications of his
illness, before he began to pull himself together again.
After the spectacular circular progressions that transpire earlier in
the development section, the prolongation of D minor in measures 173
through 200 seems especially reserved. Even though various melodic
motives from earlier in the development recur, the soft dynamic and tonal
restriction create an ascetic mood. Three possible continuations from D
minor are auditioned in turn. First there is the potential for dominant F's
6-phase chord to revert directly to F major. In measures 174 and 175 bass
D, the 6 of F>-®, descends to C#, a step towards the reinstatement of 5 (C).
At first Schubert returns to D, but then he tries again and by measure 177
has attained C, only to restore D yet again. The same gambit is played out
in the soprano in measures 180 through 184. A second possible continua-
tion (the one that ultimately will prevail in measure 201) is continuation to
G, the conventional successor of F°~*. In measure 189 a reiteration of bass
D>Db>C proceeds another half step to By. For now the G? chord that By
258 Harmony in Schubert

99 100 101 171/201 202 203

m.

Bb Major: V* 5
V)
es 9
(=F Major: P-° WD VP Pe Tlganva
(= VI ) ‘
12.10 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 99-203.

supports leads via A back to D in the following measure (thus I-IV4-V#-I

in the tonicized key of D Minor). A third possible continuation is the
descent of another third (regarding the development section to this point
as having traversed the third F>D) to Bb in measure 193. Could this be a
muted restoration of the movement's tonic? Or is D Minor being prolonged
by an excursion to its lower-third chord and back (D°*-°~”)? Whatever that
eerie apparition is, ultimately it is rescinded, for D-F-A is restored once
again in measure 199. After these tentative explorations of possible ways
out, Schubert chooses the most normative of them, F°°-G, now continuing
with C and EF, as displayed in 12.10. Concurrently he resolves some unfin-
ished business from the exposition: the melodic filling-in of fifth C>F
prolonging dominant F major. As that initiative concludes in measure 203,
another initiative is set in motion: the destabilization of F major, from toni-
cized key to revitalized dominant in Bb Major. Dissonant E}, which eventu-
ally rises to the top of the texture, heralds the D (3) of the tonic Bb chord
that inaugurates the recapitulation.
Cohn and I concur regarding the large-scale chord progression that
extends from the end of the exposition through the beginning of the recap-
itulation.” Yet we disagree concerning the relative weighting of various
chords within that expanse. The essence of our divergent views is apparent
even in our charts of relatedness. In my perspective D minor, D major, D}
major, and D) minor are all potential 6-phase extensions of dominant F
major, with zero through three chromatic adjustments. (I say “potential”
because there may be other motivations for their occurrence, as in D minor’s
role as upper third of tonic B} in measure 68, as shown in 12.4, mirrored
 Piano Sonata in Bb Major My)

by Bb major’s role as lower third of D minor in measures 193 through 198.)

In Cohn’s perspective, D+ and C#- reside within the “Dominant” region,
while D+ and D- reside within the “Tonic” region. (See his fig. 4.) Thus
there is no provision for atonement, as I suggest Schubert pursues through
his gradual conversion from deviant D}-F}-A} to diatonic D-F-A.”°
In 12.10 a deep connection is proposed between the F major chords that
end the exposition and end the development, despite the extraordinary
and extensive attention that Schubert devotes to F major’s various 6-phase
chords in the interim (indicated by the hairpin symbol in the example).
The example's fleeting, dominant-emulating D chord of measure 100 and
the prolonged D chord of measures 171 through 201 (attained only after an
extended interaction with its chromatic variants) are both intermediaries
between tonicized F and its supertonic G. In Cohn’s view the extraordi-
nary duration and distinctive context of the latter D region elevate it to a
superior hierarchical rank: “D minor is marked by a tonicization so thun-
derous that it siphons off the rhetorical weight conventionally bestowed
on the return of Bb major at the opening of the recapitulation” (p. 226). As
noted above, D minor falls within his “Tonic” region, and so he regards
the transition from dominant to tonic to be inaugurated by this D region.
That requires a corresponding demotion of the restored F chord (with
minor seventh) in measures 203-215, which Cohn describes as “an opiated
and spineless retransitional dominant” (p. 226).’’ I instead regard it as the
dignified embracing of stability and a rightful course after insalubrious
transgressions — not a stupor, as Cohn suggests, but instead the supplicant
reaffirmation of how things ought to be, in the hope that the further course
of the movement (and of the composer's life?) will not be adversely affected
by the wayward path that was recently traversed. The reduced volume and
texture and the use of chordal inversion suggest a seriously ill protagonist
who is not confident that he will be able to step back onto the stage for a
robust recapitulation. Yet he is contrite, and it is clear that he intends to try
his best.28 Several bouts of dissonant, chromatic Gb (measures 206-207,
212, and 214) remind one of the exposition’s devastating course. Will a
recurrence of that trajectory be averted during the recapitulation? Or is our
protagonist (Schubert himself) doomed?

Recapitulation (measures 216-357)

The principal imperative within a recapitulation is of course to adjust the

exposition’s tonal plan so that the work ends squarely in the tonic key.
Alterations within the transition generally suffice in redirecting the tonal
260 Harmony in Schubert

path, so that the exposition’s dominant region can be presented in a tonic

transposition. Schubert's alterations extend a bit beyond that, however. The
first significant variance might be unintended: though both bass F in meas-
ures 6 and 7 and bass B} in measures 221 and 222 will work in this context,
I see no good reason why the two spots do not coordinate exactly. Perhaps
Schubert was working very quickly, writing the latter measures from
memory rather than by consulting the earlier passage, and consequently he
did not replicate it exactly.
The changes within measures 235 through 258 have a more obvious
motivation. Since the upcoming transition (measures 264 through 298)
will be altered in accordance with the revised tonal plan of the recapitula-
tion, the potent Chromatic Variant #3 of tonic’s 6 phase will not recur at
that point. Schubert doubles up, as it were, presenting both Chromatic
Variant #2 (Bb-Db-G}b) and Chromatic Variant #3 (Bbb-Db-Gb [A-C#-F#])
in direct succession. Just as that latter chord led to A major in its earlier
presentation [12.5, measures 48 through 67], A major again succeeds it
in measure 242. Once its seventh joins in at measure 253, the A chord’s
tonal implications arouse intense curiosity. Comparing A’ in measures
253 and 254 with the parallel spot in measures 34 and 35, we may suspect
that A-C#-E-G represents A-C-E-Fx, and that the resolution might be to
a G# chord. Comparing it with the A” chord in measures 58 through 67,
we alternatively may suspect that the resolution will be to a D chord, as
in measure 68. (See 12.5.) Schubert pursues a third and quite unexpected
continuation instead: as happens on occasion, here A’ behaves as Bb’s
dominant (F?—).”? This resolution matches one that we encountered in
“Auflésung” [1.30]. In this context the exposition quandary regarding
whether tonic or dominant harmony (with ¢ embellishment) supports
the opening measures of the thematic return is no longer an issue, for the
chord of measures 255 and 256 is in >position and clearly functions as the
restored tonic.
The transition within the recapitulation, like that of the exposition, opens
with a diminished seventh chord that occurs in place of a cadence on tonic.
(Compare measures 45 and 264.) Schubert’s task here is to lead his transi-
tion not to dominant F, as in the exposition, but instead to tonic Bb. The
necessary adjustments have been accomplished by measure 267: B minor
occurs where F# minor resided during the exposition (measures 48 and
following). The path during the exposition was simply

m. 1-39 45-47 48

Bis Db? =SGh (FHI

 Piano Sonata in Bb Major 261

m. 255-256 287 291-292 293-298 299

:7 si é

avalon eye et Ome Byt= ie e

12.11 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1, mm. 255-299,

(See 12.5.) During the recapitulation Schubert recomposes this passage so

as to continue an additional descending fifth (or ascending fourth), as
m. 255 264-265 266 267

Bb Db? — Gb’ [FR] 7-> B

(Db’ has evolved into F-A}-C)-Ebb [F-A)-B4-D].) From that point onwards
the two transitions proceed in exactly the same manner. In the exposition
G} [F#] leads downward a major third to D minor, tonic’s diatonic upper-
third chord, and the structure displayed in 12.4 ensues from there. In the
recapitulation B leads downward a major third to G minor, tonic’s diatonic
lower-third chord, and the structure displayed in 12.11 ensues from there.
Since D minor and G minor are a perfect fifth apart, so also are the goals
of these two progressions: dominant F major (measure 80) versus tonic Bb
major (measure 299).
As 12.10 suggests, the exposition ends without a full closure on its F
major chord. (The graph’s beamed melodic descent has progressed only
so far as C>B}>A by the exposition’s close in measure 116.) That line
continues to F during the development. In the recapitulation a similar
situation prevails in that tonic’s fifth F>Bb has been traversed only so far as
F>Eb>D (measure 334) by the time the material from the exposition has
run its course. Schubert extends this region, completing the descent with
D>C>B} in measures 335 through 340 (repeated in measures 340 through
345). These are the concluding noteheads in my model for the entire move-
ment, 12.12.°° The sunniest version of tonic’s 6 phase, Chromatic Variant
#1, participates in this progression (in measures 337 and 342), attained via
yet another traversal of the B)>Ab>G melodic third. A final rumble of G)
262 Harmony in Schubert

74 80 204 216 293 299

m. D

Nee dy) 4
A Aw AN

3 5

Exposition Development Recapitulation

Bb Major: I Gs Veenaak Aneel

12.12 Analysis of the Piano Sonata in Bb Major (D. 960), mvmt. 1.

(measure 352, derived from measure 8 and reminiscent of measures 120,

212, and 214) is held in check; it fails to deflect the transcendent ending. In
Schubert’s own life such transcendence was achieved only in his art. Within
months of penning these notes he expired.
The gap between Cohn’s and my perspectives widens during the recap-
itulation. For example, he injects references to the subdominant harmony
— even to components of a “standard Schubertian subdominant recapitula-
tion” (pp. 227-228) - that I reject, considering my reading of the transition’s
structure, displayed in 12.11. The G minor chord of measure 287 is for him
a “proxy” for the subdominant (p. 228), whereas for me it is a 6-phase chord
performing a conventional role: connecting the tonic and the supertonic.
The admittedly extraordinary A’—Bb succession of measures 254 and 255
might demonstrate how a chord may resolve “‘deceptively’,” as Cohn sug-
gests (p. 229). Yet his very use of that term (cryptically placed within quo-
tation marks) makes me wonder whether he subscribes to the notion that
A+ represents dominant F+ as wholeheartedly as do I. If A+ represents the
dominant in B} Major, what is deceptive about its succession to Bb+?
Cohn asserts that the harmonic content of this music is “not adequately
reconcilable to diatonic tonality” (p. 231), a conjecture that his instruments
for gazing at tonality in Schubert may seem to support. Yet those instru-
ments impose powerful associations that seem to me to be contradicted
by Schubert's writing within this movement. For example, Cohn’s matrices
impose the notions that Bb and D chords are related in function and that
G}b and G chords are not. I have argued against both of those assumptions
in my analysis: that in the context of the development, D is dominant
F's lower-third chord, and not tonic Bb’s upper-third chord; and that the
exposition’s G} chords are intensified fulfillments of the function that was
earlier performed by G. At the same time those instruments make it hard to
 Piano Sonata in Bb Major 263

access the accumulated wisdom concerning normative harmonic practice

that seems to me as valid in this movement as in any other from the period.
For example, the chords that constitute the I?-°-IIk-V progression in B}
Major are bizarrely distributed among the regions of Cohn’s matrix.*! That
progression is for me the quintessential path between tonic and dominant.
Schubert does not disappoint: tonic’s 6-phase chord and the major super-
tonic shape the exposition’s transition [12.1]. Though Cohn could have
hit upon G+ C+ using his instruments, he is defiantly proceeding without
access to the perspective that nurtures my analytical decision-making.
Is it any wonder, then, that his reading is “not adequately reconcilable”
to mine? Though many practitioners of the scale-step perspective have
not succeeded in extending its reach to successfully accommodate highly
chromatic music (a point on which Cohn and I agree), our responses are
diametrically opposed: he has abandoned the methodology, whereas I am
endeavoring to regenerate it.
Epilogue

Schubert once said to me: People say that my compositions are not free of errors in
harmony etc. If only they would show me where the errors are!
Joseph Lanz’

Though Schubert would not have had every measure he composed in

mind when confiding his frustration to Lanz, we may assume that his
own harmonic practice made sense to him. He should not be accused of
willfully sabotaging the coherence of his works, though clearly he was
aware of the bafflement occasionally experienced by listeners. Schubert
aimed high, higher even than the connoisseurs among his contemporar-
ies could keep up with. As a result we are the fortunate recipients of a
wondrous corpus of music that has retained its challenge and fascination
to this day.
Modern analysts who delve into Schubert's music likewise admit to
an occasional feeling of bafflement: words such as “riddling” (Richard
Kramer), “lost” (Robert Hatten), and “nonsensical” (Lawrence Kramer)
occasionally appear in their discourse. The causes of bafflement might
include: (1) Schubert wrote illogically in a passage; (2) Schubert sought
to depict the notion of “incoherence” within a passage; (3) the passage
was incorrectly notated or printed; (4) the available analytical tools are
not suited to the music at hand; or (5) the analyst is not making full and
appropriate use of the available analytical tools. I have directed my ener-
gies towards alleviating the fourth and fifth of these causes (sometimes
revealing how instances that might seem to result from the first or second
cause may actually fall within one of these latter realms), both by refin-
ing the Stufentheorie perspective as warranted by what I have observed in
Schubert's music and by using this enhanced perspective as carefully and
as creatively as possible. Whereas academic culture tends to value novelty,
I relish the continuity between my approach and the practices of music
analysts from earlier generations, tracing back to Schubert’s time. Much
of what I have presented revives notions with extensive histories. Granted,
my amalgam is unique and includes some novel touches. Yet most of the
raw material for my perspective was readily available, though perhaps
 Epilogue 265

neglected. Even Schenker seems neglected nowadays, as fresh agendas vie

for attention within the realm of music theory.
Formulating my analyses in conjunction with samples of various alterna-
tive approaches has been a rewarding and instructive experience for me,
and I trust also a useful framework for readers to come to terms with my
perspective and how it differs from analysis as currently practiced. I salute
my fellow authors. They have engaged with challenging music in sophisti-
cated ways, and though I have disagreed with them on matters both broad
and minute, I commend the sincerity and conviction with which they have
pursued their agendas. I am less sympathetic towards authors who attempt
to deal with Schubert’s music without coming to terms with how it works as
music—as pitches sounding together and in succession. To be sure, creating
insightful music analyses is a challenging occupation, one that requires an
enduring commitment of time and energy. Those who forgo the perspective
that such analysis offers are handicapped in dealing with the music in any
other manner, be they performers, scholars, or listeners.
Whereas in Thinking About Harmony | focused on how a wide range
of musicians with an analytical or pedagogical bent formulated ways of
coming to terms with the music of their time, in Harmony in Schubert
my focus has been unidirectional: the master composer Schubert himself.
It is his mind, his ears, his musical sensibilities that have claimed my
attention, much more than those of the musicians in his milieu. He chal-
lenged those around him to perk up their ears to his astonishing creations
and to evolve in their musical thinking along the lines that he set forth.
That challenge still resonates. Harmony in Schubert is both my personal
response to that challenge and a guide for other practitioners who take
it seriously. Granted, adopting the analytical procedures that I have
espoused in Harmony in Schubert means forsaking others. Readers must
decide for themselves whether such an extensive overhaul is warranted
in their own musical lives. For me, the transition has been invigorating,
even revelatory, opening new regions of personal musical perception and
transforming both my teaching (of undergraduate as well as graduate
students) and my performance. Though there is no shortage of additional
works by Schubert that call out for careful analysis, what I have included
within the covers of Harmony in Schubert likely will serve as a substantial
initial helping for most readers. I am eager for the ideas that I have devel-
oped to be taken up and tested by others. In the meantime I intend to
investigate how the strategies introduced here fare in other realms, both
prior to and following Schubert. I do not doubt that my decision to focus
on a single composer - and Schubert in particular - was beneficial to the
266 Harmony in Schubert

sort of study I had in mind to pursue. I am equally convinced that it is

now time for me to broaden the scope of my undertaking. I am looking
forward to engagement with composers from Haydn through Chopin in
the next phase of this study.
Notes and references

1 Harmonic progression

il Analysts do not all agree regarding which pitches are diatonic in the minor
keys, particularly the seventh scale degree (the subtonic versus the leading
tone). Thus the list of keys in which, for example, B is diatonic will not be the
same for everyone: some lists will include C# Minor, others instead C Minor. I
regard the so-called “natural minor,” the pitches designated by the minor key
signature, as the diatonic state.
When a diatonic interpretation remains elusive, such chords often become
friendly foreigners: “French,” “German,” “Italian,” “Neapolitan.” (The diatonic
» « » «

interpretation of the “Neapolitan” chord as the subdominant of a minor key’s

submediant has few adherents nowadays. See TAH, 6.13 and 8.6.)
The battle between a chromatic interpretation within one key versus a dia-
tonic interpretation within multiple keys has been raging since the eighteenth
century. In TAH, pp. 149-155, I explore various antecedents to my non-
modulatory stance.
Though I began deploying this symbol thinking that I was forging a new path,
Tan Quinn later reminded me that Walter Piston used the degree circle (e.g., V§)
for the same purpose. Perhaps a distant memory was an influence here: Piston’s
Harmony (New York: Norton, 3rd edn, 1962) was the course textbook for my
foundational theoretical studies in 1972-1974.
A frequent exception to this generalization is that a chordal ninth may descend
to the octave (root) during the prolongation, rather than at the arrival of the
succeeding chord. This accords with Kirnberger’s distinctions among a chord’s
elements: the triadic foundation, the “essential” seventh (which resolves in
coordination with the succession to the next chord), and the “incidental” ninth
(which may resolve within its chord). See TAH, pp. 37-39.
The employment of two separate sets of figures, one for the chord as presented,
another for its idealized root-position state, derives ultimately from the first
analyst who used Roman numerals to track the progression of roots, Georg
Joseph Vogler. See TAH, 1.4, which dates from ca. 1778.
In omitting an indication of inversion, I follow the lead of Johann Christian
Lobe. See TAH, 2.7 and 2.8.
Tassume that all analysis will incorporate a visual component, so that the chord
symbols will appear in view of the readers or auditors who are following the
268 Notes and references to pages 6-10

analytical presentation. In written or spoken commentary such chords may

be referred to imprecisely using locutions such as “evolved II” or “II” (pro-
nounced “two nine”), assuming that the visual representation will be consulted
for a full indication of the chord’s constitution.
I process the eight-measure phrase in four two-measure chunks, conforming to
Schubert’s slurring. Thus bass G arrives at the proper moment, while the other
pitches of measure 61 usurp space that rightfully belongs to II’s E and B.
10 Though presented differently, my observation parallels Rameau’s notion con-
cerning roots related by ascending second: “whenever it is permissible to
have the fundamental bass ascend a tone or a semitone, the progression of
a [descending] third and [an ascending] fourth is always implied.” See J.-P.
Rameau, Traité de ’harmonie (Paris: Ballard, 1722); trans. P. Gossett as Treatise
on Harmony (New York: Dover, 1971), p. 234.
11 Kirnberger analyzes a D#-F#-(A)-Cy chord in an example that I cite and anno-
tate in TAH, 1.14a. He and | agree that its root is B.
1 Even a substantial monograph on Schubert’s harmonic practice cannot treat
such a rich topic comprehensively. I focus especially on close analysis of
selected scores by Schubert, on how my procedures extend the analytical prac-
tices of the nineteenth and early twentieth centuries, and on how my readings
differ from those of other modern analysts. Another very worthwhile (and
immense) inquiry that I largely neglect is how Schubert may have been influ-
enced by the compositions of others. For example, this extraordinary collision
in Schubert’s Sonata in A Minor is almost identical to what occurs in Mozart’s
String Quintet in C Major (K. 515), mvmt. 1, m. 15.
13 Another unifying force is subtly expressed in the relationship between soprano
and bass in 1.1. Soprano D<D#<F#>E (measures 60 through 62) is echoed by
bass G<G#<B>A (measures 61 through 66).
14 In his Distant Cycles: Schubert and the Conceiving of Song (1994), Richard
Kramer cites several instances of augmented dominant chords, suggesting that
“the configuration ... was very much in Schubert’s ear at this time. Still rare, in
1820, as a harmony with its own credentials, it had become something of a code
in the Schlegel settings” (p. 211). Comparing measure 34 in his example 9.3 (a
passage from “Nachthymne,” D. 687) and the corresponding passage in his
example 9.6 (a linear analysis of that passage), one observes that Kramer inter-
prets F4-A#-Cx [F#-Bb-D] as representative of an elided diatonic F#-A#-Cd.
(The F4-B-D# chord that follows in measure 35 is not a resolution, however, but
instead a connector between F# chords.)
1S Compare with Lobe’s treatment of B-D-F-A, which may proceed to either a C
chord or an E chord, in TAH, 1.15. Also compare Schubert’s E-G#-B-D chord
with an identical chord employed by Wagner in a passage from Lohengrin and
by Louis and Thuille in their Harmonielehre, quoted and discussed in TAH,
pp. 224-229. A late example by Schubert of the same approach to IV occurs in
“Liebesbotschaft” from Schwanengesang (D. 957/1): in the tonicized key of B
 Notes and references to pages 10-14 269

Major, the progression leads from B-D#-F# through D#-Fx-(A#)-C# to E-G#-B

in measures 43 through 45.
16 Because tonic’s diatonic upper-third chord (here E-G-B) often undergoes
chordal evolution, this imprecise locution generally will be used in conjunction
with visual analytical notation (such as that in 1.5, m. 249) indicating the exact
constitution of the chord.
17 In his Unfinished Music (2008), Richard Kramer describes the Trio from
Schubert’s unfinished Piano Sonata in F Minor (D. 625) as “riddling” (p. 327).
The bass begins with a stepwise descent from the dominant root to the A Major
tonic: E>D>C#>B>A. After an upward diversion creating a half cadence at the
double bar, the descent continues through G4 to F#, continuing as F#—>B. Thus
far, the deeper structure is normative: I>° II#, with dominant emulation occur-
ring during tonic’s 6 phase. Yet instead of a direct connection between II# and
V, Schubert inserts Af—D#, landing on II#’s chromaticized upper-third chord.
I propose that the B<D#[Eb]<E connection between II# and V is the equivalent
of the C<E<F connection between |and IV in “Hagars Klage” [1.5 through 1.7].
Major supertonic B-D#-F# is extended by its upper-third chord D$-Fx-A# [Eb-G-
B}] before dominant E-G#-B arrives. Because the a, section of the a,-b-a, form
begins on the dominant (as does the a, section), the b section proceeds only so
far as this D# chord, which measured from tonic A indeed may seem “riddling.”
I suggest that it should be measured instead from B. The upper-third chord’s Fx
connects II#’s F# and V’s G# (embellished by A). Thus I disagree with Kramer’s
assertion that the D# chord “fails to connect with the return of the opening
phrase” (p. 327).
18 Some readers may note a similarity between my formulation and that of Hugo
Riemann (1849-1919). While acknowledging that our analytical perspectives
share certain features, I do not regard Riemann as a strong influence on my
thinking. This chapter’s references to Vogler, Kirnberger, Lobe, the Louis and
Thuille team, and Schenker provide a better notion of how my analytical prac-
tice has developed. I advocate the scale-step perspective (Stufentheorie), though
with a sparser application of Roman numerals than is typical. Riemann advo-
cated a functional approach (Funktionstheorie). (These terms are introduced in
TAH, chapter 1.)
19 See TArs ppsi7—19.
20 The chromatic alteration of lower- and upper-third chords, a feature of many
analyses throughout this book, will be explored more fully in chapter 2, where
I will introduce analytical terminology to describe the variants.
21 Other variants will be introduced in 4.18 and 11.2 (Model 2).
22 I assess the subtonic/dominant interaction in detail in “Schenker, Schubert,
and the Subtonic Chord,” in A Music-Theoretical Matrix: Essays in Honor of
Allen Forte (Part I), ed. David Carson Berry, Gamut 3/1 (2010).
Wi) Compare with Schenker’s analytical practice in FC, figs. 111, 113, and 114,
where V’s upper-third chord often is supplied with no analytical label, or with
270 Notes and references to pages 14-18

the numeral VII placed within parentheses. His deployment of parentheses is

frustratingly inconsistent.
24 An exceptional case will be explored in 1.30.
US, Thus I concur with Schenker’s endorsement of an “ex post facto” recognition of
harmonic function. See his Harmonielehre (1906/1954), p. 254.
26 Readers interested in this topic may wish to consult David Lewin’s “Music
Theory, Phenomenology, and Modes of Perception” (1985-1986). Compare
his commentary on measures 10 through 15 of “Morgengruf” from Die schone
Miillerin (D. 795/8) with two dominant prolongations that I discuss in TAH:
first 5.2b (in which the bass ascends), and then 2.23b (derived from Berlioz, in
which the bass descends). Though Schubert’s chromaticism in “Morgengruf”
does not match the intensity of Berlioz’s, I propose that the same strategy is at
work - namely, the simultaneous filling-in of intervals from the V’ chord: D>B,
B>G, and G>(F).
27 Another tension exists between what analysis proposes one should be able
to comprehend aurally and what listeners actually perceive. One could test a
random sampling of undergraduate music majors in a laboratory to collect
sophisticated data on their perceptions. I suspect that if we had similar data
on Schubert’s perceptions, we would be astonished by the gap. I propose that,
even if unachievable, we should aspire to rise to Schubert’s level and gauge our
analytical discourse accordingly.
28 Edward T. Cone explores this notion in “Schubert’s Promissory Note: An
Exercise in Musical Hermeneutics” (1981-1982).
2S For example, in Hugo Riemann’s system D-F-A in C Major would be labeled
Sp, the subdominant’s parallel chord.
30 The II chord is assessed in chapter 4. An accidental to the left of a Roman
numeral applies to the chord’s root.
31 In that this upper-fifth chord (m. 729), the evolved 6-phase chord (m. 730), and
the II chord to which it resolves (m. 731) are all transformed in equivalent ways,
this progression is realized as three consecutive diminished seventh chords
descending by half steps. Upper-fifth chords will be explored in greater detail
later in this chapter.
sy My use of the term chord, instead of harmony, reinforces the perspective that
various simultaneities in music function as connectors between harmonies,
rather than as harmonies in their own right. Though such simultaneities may
function harmonically in other contexts, the transfer of their harmonic labels to
non-harmonic contexts is not justified. Granted, many analysts would provide
Roman-numeral labels for measure 10 in 1.13, measures 727 and 729 in 1.14, and
the A-C#-F# chord in 1.15. By opposing that practice and imposing ahigher stand-
ard for admission into the harmonic realm, that realm becomes less congested and
the progression of bona fide harmonies generally is easier to comprehend.
38 My analysis of the G# chord initially may frustrate advocates of applied domi-
nants. Its relationship to its context remains that of a 6-phase chord, even if it
 Notes and references to pages 19-26 AY

is treated to extensive chromatic modification, as documented in the analyti-

cal notation. When a 6-phase chord is asserted, the 6 is transformed from an
anticipation into a harmonic root. (Consequently a Roman numeral is placed
within parentheses below the Arabic 6.) Yet it remains a 6-phase in relation to
the initiating 5-phase chord and thus is interpreted as VI (in relation to I) rather
than as V (in the context of the key of the succeeding II). The arrow notation
introduced earlier in this chapter (here #VI—) may be employed as a substitute
for the retired “V/II” symbol. Meticulously formulated analytical symbols need
not be replicated in written or spoken commentary. I might say “a dominant-
emulating 6-phase chord,” but never “sharp-six seven five-sharp three-sharp”!
34 Concerning the antipodal relationship, see TAH, pp. 207 and 224-229.
By converting Schubert's flats in measures 73 and 74 into their sharp equiva-
lents, the score’s bizarre bass A>Ab>G}>E is transformed into normative
A>G#>F#>E. If, as proposed, tonic’s 6 phase is taking shape starting in measure
72, then Gé and F# are the logical connectors within the span A>E. While the
outer voices fill in the E<G and A>E intervals of the A’ chord, the inner voice
undertakes C<C3.
36 See Burstein, “Unraveling Schenker’s Concept of the Auxiliary Cadence”
(2005).
37 In any tonic prolongation Schubert is likely to traverse one or more intervals
of the tonic chord. Here C>B)>Ab serves as the principal soprano descent (C
to incomplete neighbor Db, passing through C down to Bb, then Ab), leading to
a perfect authentic cadence in measure 16. (Because the vocal line substitutes
a high Eb for 2, I have circled these pitches in the keyboard melody.) Another
descent, an A>>C sixth, works in coordination with that third. That line is split
between registers both below and above that of the principal line. Ab sounds as
the keyboard melody’s first pitch, whereas Gp and F (which I have placed within
boxes in the score) are the melody’s highest pitches. Fb restores the lower register
in measure 15, followed by Eb and Db. The concluding C (notated in the treble
clef) coincides with the upper strand’s closing Ab. Observe that both of these
tonic-expanding lines recur in measures 16 and 17, this time with the Ab>C sixth
presented lower in the texture and without a registral shift. The two descending
lines that pursue their courses together in this excerpt appear one after the other
in 1.19: a subsidiary G>B sixth followed by the principal B>G third.
38 Schubert unfolds the ted',dominant embellishment of measure 43 as © sae
Compare with FC, fig. 43.b4, and TAH, 2.24 (especially the commentary to
2.24b).
59 Though the upper-fifth notion was championed by Schenker, there were
nineteenth-century antecedents. See FC, fig. 22b, m. 2, and fig. 56c, m. 7; and
see TAH, pp. 126-131.
40 IV’s 6-phase chord (A-C-E) likewise is extended via its upper-fifth chord (m.
4), promoting its assertion as II.
4] Parallel progressions will be explored in chapter 2.
MP) Notes and references to pages 28-33

42 Compare with 3.32, Model 3.

43 Others may hear things differently. In his “Third Relation and Dominant in
Late 18th- and Early 19th-Century Music” (1980), Harald Krebs applies slurs
that conform to my Hypothesis 1b. Compare Schubert’s writing here with
similar passages in “Liebesrauch” (D. 179), mm. 11 through 14, and “Beim
Winde” (D. 669), mm. 69/70 through 77.
44 This notion will be explored in greater detail in chapter 3.
45 Whereas I have recommended above that potential harmonic interpretations
that are not confirmed by the progression’s continuation should not be the
focus of an analysis, here Schubert is making a vigorous point of juxtaposing
two chords of dominant character, each vying for ascendancy. Not to consider
the F#7 chord as a dominant is to miss the point of the passage.
46 Unlike a chromatic passing note (e.g., A<A#<B), a wobbly note is preceded and
followed by the same note (A<A#>A). (The initial note may be either diatonic
or chromatic, but it must be the same both before and after the wobbly note.)
Unlike a chromatic neighboring note (e.g., A<Bb>A), a wobbly note employs
the same note name and notehead as the preceding and following notes. The
three-note progression (A<A#>A) may be referred to as a wobble. My use ofthe
term wobbly note here differs from that in TAH, p. 203.
47 In “The Functional Extravagance of Chromatic Chords” (1986), Charles J.
Smith makes similar observations concerning the role of what appears to be
a dominant seventh chord built on the leading tone. His analytical symbols
grant the leading tone the status of a root, whereas my symbols interpret
the leading tone as the chordal third above an absent root. Smith further
develops this perspective in “The Love of Fundamentals is the Root of All
Evil: Alternatives to Harmonic Fundamentalism,” in A Composition as a
Problem V: Proceedings of the Fifth International Conference on Music Theory/
Tallinn, September 28-30, 2006, ed. M. Humal (Tallinn: Eesti Muusika-ja
Teatriakadeemia, 2008), pp. 5-24. See also Poundie Burstein’s “Surprising
Returns: The VII# in Beethoven’s Op. 18 No. 3, and its Antecedents in Haydn”
(1998). Burstein quotes examples of Haydn proceeding directly from a major
triad on the leading tone (with or without minor seventh) to the tonic. (See
especially his example 6.)
48 Compare with “Auf dem Strom” [1.10], measures 7 through 9, during which
the dominant’s upper-third chord is embellished in a manner comparable to
measures 11 and 12 of “Auflésung.” In “Auf dem Strom,” the F# and A that
serve as foundation for the Fx and A¥ wobbly notes occur only at the close of
the wobble.
49 Though Schubert seeks to create a parallel between measures 1-2 and 3-4,
he wisely rejects V in ° position for the second half of measures 3 and 4, even
though bass G<B and D<F# would match exactly. The problem comes when
adding passing notes in measure 4: Deb-r4 Would present questionable parallel
fifths. The voice-leading connection between bass B in measures 1 and 2 and
 Notes and references to pages 34-37 He:

bass A in measures 3 and 4 further justifies the dominant’s presentation in 3

position.
Although placing °below bass B and ?below bass G in measure 7 is an analytical
option consistent with how such figures often appeared in scores of especially
the eighteenth century, I will rarely do so. To prevent a clutter of symbols I gen-
erally restrict the analytical notation to the Roman-numeral row (in which the
Arabic numbers assess intervals above the root, not the bass) and discuss any
noteworthy unfurlings in written commentary. I make an exception in measure
72 of 1.34.
Compare the absent C and B here and the absent C# in 1.15, measure 20. A
conventional descent for a similar passage occurs near the beginning of “Am
Feierabend”: 5 to 1 in measures 20 through 24. The unexpected high E in
measure 72 adds a note of urgency.
An A-D succession might instead imply the onset of a circular progression
by descending fifths. In fact, Robert Schumann appears to have borrowed the
content of “Ungeduld,” measures 1 through 6, to inaugurate a circular progres-
sion in the fourth movement of his Piano Sonata in F# Minor, op. 11 (1835).
See TAH, 3.9 and 3.10. Schenker’s pupil Felix-Eberhard von Cube offers an
analysis ofthe lied in The Book of the Musical Artwork (1988), p. 321. Our read-
ings of the chordal hierarchy during the introduction are contrasting: whereas
I regard the dominant of measure 3 as residing within an expansion of the A
tonic (which upon its return in measure 5 incorporates a minor seventh in the
service of dominant emulation), he prolongs that dominant through measure
8.
I suggest that the kinship in sound between measure 1 and the final chord of
measure 5 trumps the kinship of metrical placement between B—E (measures
2-3) and A-+D (measures 5-6). The occasional non-alliance of deep structure
and surface parallelisms is a topic of considerable importance to analysis, as we
saw in the discussion of “Peculiar juxtapositions,” above.
4 IV>* in A Major engages the same pitch classes as I>-° in D Major. Thus this
passage may be compared with 1.13, measures 9-11. Note in particular the cor-
respondence between F#-A2-Cz-E (over D pedal) in measure 7 of “Ungeduld”
and the AZ-C2-E-G chord that comes between the 5 and 6 phases of I in 1.13
(likewise over D pedal!).
Qn In this interpretation the A chord of measure 9 would count as both a 5-phase
and a 6-phase chord, in a common j teal ag of the sequence in which A>
C#5'D5-6 E5-6 F#. This topic is explored
B>* C2>* D> E** F# is presented as [A>
A”
in chapter 2.
Compare the dominant-prolonging progression from the downbeat of measure
12 through the downbeat of measure 13 in “Ungeduld” and the submediant-
prolonging progression from measure 72 through 75 in “Die Allmacht” [1.18,
Model 3]. The C# chord warrants careful consideration by performers. Two
contrasting approaches might be pursued: (1) through tools such as dynamics
274 Notes and references to pages 41-51

and rubato, making measure 12 come across as much like measure 10 as pos-
sible, with an unexpected continuation at the downbeat of measure 13; or
(2) binding the E chords on the downbeats of measures 12 and 13 via a more
aggressive drive through (rather than to) the C# chord.

2 Linear progression

1 For a nineteenth-century presentation of this notion, see TAH, 7.14c. Chapter

3 of TAH offers numerous examples of parallel progressions.
The vocal melody emphasizes unfoldings of thirds: F>D during the tonic pro-
longation of measures 22-24, Eb>C during the IV*° of measures 24 through 27,
and Bb<D during the cadential ° of measure 27 (with D dramatically delayed,
arriving when we expect to hear the 3resolution). Only the A of the dominant’s
‘ third is missing in the vocal melody’s projection of the inner strand, a third
below the principal descent from 5 to 1.
A I®-7*IV succession occurs in this movement’s trio section. Schubert’s juxta-
position of these passages underscores my assertion that the same chords may
sometimes be used in a harmonic context, and at other times in a sequential
context.
Further examples of this strategy and a more detailed exploration of this
sequence type are presented in my “Schubert, Chromaticism, and the Ascending
5-6 Sequence” (2006).
A linear chord connects tonic’s 5 and 6 phases, as also occurs in 1.13. Such
chords may themselves emulate dominants or supertonics (F#-B in 1.13;
AD in 2.9).
Late in the nineteenth century, Anatole Loquin proposed changes in music
terminology intended to minimize the diatonic bias. See TAH, pp. 208-209 and
264. See also Catherine Nolan’s “Combinatorial Space in Nineteenth- and Early
Twentieth-Century Music Theory” (2003), which explores Loquin’s ideas in
detail, as well as those of other pioneers in modulo 12 thinking, including Carl
Friedrich Weitzmann and Heinrich Vincent.
At a few locations within the full-octave traversal of the sequence, dominant
emulation occurs naturally: for example, E° to F° in C Major and D* to E}> in C
Minor.
Perhaps the most unusual aspect of this interpretation is the relationship of the
exposition’s dominant region (starting at measure 32) to the broader ascending
trajectory of the sequential initiative. A structurally deep motion such as D-A
here often occurs during a major-key sonata exposition. Since the next point
along the ascending sequential trajectory that transpires during the develop-
ment section is F Major, I regard the exposition’s dominant A as subservient
to the development’s 5-phase chords (starting with F major in measure 111).
(Compare with how I interpret the D major chord prolonged during measures
127 through 136 of the Trio from the Piano Sonata in D Major, analyzed later in
 Notes and references to pages 52-56 YS:

this chapter. In that case tonic’s upper-fifth chord is subservient to the broader
flow of a descending circle of fifths.) Measures 1 through 111 of 2.12 correspond
to the “stretched” model at the bottom of 2.11, transposed up a major second.
An enharmonic adjustment (Fb major = E major) was required at one point
in 2.9 to accommodate the notational system. In 2.12 such adjustments are
employed at four separate points. (The alternative spellings are juxtaposed
in the example, using square brackets.) The first two, in measures 124 and
140, cancel one another out. That is, D + (m3 x 3) + (M2 x 2) is presented as
D<Fy<Ab<B<C#<Eb instead of as D<Fk<Ab<Cb<D}<Eb, with Eb attained in
either case. The latter phase, Eb + (m2 x 3), is akin to 2.9. Schubert proceeds
E><Ek<F4<F# rather than the unwieldy Eb<F)<Gbb<Abbb. (These problems
would disappear in a modulo 12 notational system, where the ascent could
be displayed as 3<4<5<6.) F#’s 6 phase, F#-A-D (which fortunately is not pre-
sented as Abbb-Cbbb-Fbbb!), restores the D Major tonic.
10 For an early illustration, see TAH, 3.8.
ll An exemplary presentation of the alternative view — that “a necessary part of the
feeling of magical relocation is our ability to hear an enharmonic reinterpreta-
tion of function” — is presented by Frank Samarotto in “Intimate Immensity in
Schubert’s The Shepherd on the Rock” (2006), pp. 212-217.
12 In “Schubert and Equal Division of the Octave” (2007), David Beach seconds
this view: “the syntax of harmonic progression at this level is non-functional; it
becomes functional only at the macro-level, corresponding to the terminals of
the... progression” (p. 251).
12 Compare this passage with Schubert’s Symphony No. 6 in C Major (D. 589),
mvmt. 4, mm. 467-471.
14 Compare with TAH, 4.5a and 4.5b.
15 Compare with 1.9, where B major and G major chords are likewise juxtaposed.
16 This progression will be explored in greater detail in chapter 3.
7) The numbers 1, 2, and 3 in 2.17 are conveniently both ordinal and cardinal: the
first, second, and third chromatic variants are chords with one, two, and three
chromatically modified pitches, respectively. In his Chromatic Transformations
in Nineteenth-Century Music (2002), pp. 10-11, David Kopp organizes a triad’s
four 6-phase chords within three categories: relative (my diatonic), chromatic
(my Chromatic Variants #1 and #2), and disjunct (my Chromatic Variant #3).
Whereas in this chapter segment I focus especially on Schubert’s deployment
of the disjunct type, Kopp’s emphasis is on those of the chromatic type, which
incorporate at least one common tone. His statement that chords of the dis-
junct type “begin to appear with some regularity in later nineteenth-century
music” might better read “. . . in Schubert’s late music.” (Kopp’s perspective
will be explored in chapter 9, below.) In this context, consider also the notion of
hexatonic poles presented by Richard Cohn in “Uncanny Resemblances: Tonal
Signification in the Freudian Age” (2004). (Cohn’s perspective will be explored
in chapter 12, below.)
276 Notes and references to pages 60-65

18 The wobble of the root Ab is displayed as 15-bb-b. Here the initial flat symbol
refers to the diatonic state (A}), which is lowered a half step to Abb and then
restored to its diatonic state.
19 Xavier Hascher offers a similar listing of upper-third chords in Schubert, la
forme sonate et son évolution (1996), ex. 9. His presentation of the lower-third
chords is not so systematic, though his exx. 110, 112, 113, and 115 display most
of the possibilities.
20 The graphic representation of an interruption is problematic, and will be con-
sidered in greater detail later, in the context of a work that deviates from the
norm (pp. 89-91). If could work in three dimensions, I would display the tonic
chord of measure 163 behind that of measure 119, and the dominant chord of
measure 179 behind that of measure 159, to convey more clearly the sense of
statement/reiteration. The descent achieves closure only in the reiteration. Yet
its initial elements are fully articulated in the initial statement. Thus the path
through the 2 of measure 159 ultimately reaches its goal at measure 180.
21 Compare the F# of measure 149, which leads to the Fk of a dominant-emulating
G’ chord in G Major, and the F# in 1.8, Model 2, which leads to the Fk of adomi-
nant G’ chord in C Major. Both F¥s occur in the context of an upper-third chord.
22 In their Counterpoint in Composition (1969), pp. 442-444, Felix Salzer and Carl
Schachter offer a contrasting analysis of the Trio, in which this D major chord
represents the arrival of background V, which they prolong through measure
162. Compare my interpretation of dominant D major in measure 127 here
with that of dominant A major beginning in measure 32 of 2.12. In both cases
these chords, which may seem to reside deep within the structure, instead rein-
force the preceding tonic, from which a broader structural initiative extending
into the succeeding formal region emanates.
13) In our discussion of circular progressions the traversal of an octave via four
ascending minor thirds was proposed. Here Schubert extends half that far, for
a bit of ascending circle-of-thirds motion within a descending circle-of-fifths
motion. The descending diminished fifth C>F# is represented enharmonically
by an ascending diminished fifth C<Gb, which Schubert fills in as C<E)<G}
(which in modulo 12 would appear as 0<3<6).
24 Compare the dynamic markings in these instances of chromatic upper-third
chords with those in the examples of chromatic lower-third chords in 2.16,
2.18, and 2.19. In “Common-Tone Tonality in Romantic Opera” (2008),
William Rothstein assesses third-relations in operas by Rossini, whose influ-
ence is discernible both in Schubert’s stylistic development and in his aspira-
tion to win success as an opera composer.

3 Common prolongations and successions

1 The melody of the sonata’s first movement ends with a B>A suspension
and resolution over tonic root A. Schubert creates a unifying bond between
 Notes and references to pages 65-72 DY

movements by repeating that descent as the second movement’s initiating

gesture. B's neighboring role persists throughout 3.1. Coincidentally, an upper
neighbor embellishes tonic’s fifth in the quintet as well: compare measure 1 of
3.1 and measure 6 of3.2.
Certainly not all analysts will agree regarding exactly where melodic embel-
lishment ends and bona fide harmonic progression begins. I propose that V7
asserts itself as a harmony in measures 6 and 7 of this phrase for several reasons:
the leap from tonic root D to dominant root A, the restatement in first inver-
sion, its role in the phrase’s cadence. In contrast, the same pitches in measures
3 and 4 lack that impact. Though the chord displaces the tonic’s root and third,
it does not lead the phrase into its post-tonic phase. My notion of assertion
(introduced in chapter 1) enriches the analytical process: instead of mechani-
cally placing a Roman numeral under every chord, I propose that analysts
pause to ponder whether the role of each pitch combination is aptly interpreted
as harmonic.
ies) Compare the I-space of measures 32 and 33 (in its dominant-emulation phase)
and the V-space in 1.8, Model 2. Also compare with 2.23, Model 2.
Compare this model with the first half of Schenker’s analysis of a Bach chorale
in FC, fig. 22a.
Schubert here precedes the initial tonic }with the sequentially appropriate $.
WN
Nn The wobble of the root Eb is displayed as 1) — bb — b. Here the initial flat symbol
refers to the diatonic state (E>), which is lowered a half step to Ebb and then
restored to its diatonic state.
In “Chernomor to Kashchei” (1985, pp. 84-85), Richard Taruskin comments
that “quite the most prophetic element in the passage is the use of passing tones
in the bass to connect the harmonic roots. These create what is surely one of the
very earliest seriously intended whole-tone scales in the history of European
music.”
Schenker’s point is that an ascending stepwise line can be achieved without
any ascending steps. In this case the procedure facilitates the motion from one
perfect fifth (£) to another ([*) without actually ascending by parallel fifths. See
FC, fig. 41. Peter H. Smith offers a contrasting reading of the passage in “Outer-
Voice Conflicts: Their Analytical Challenges and Artistic Consequences”
(2000). Whereas I regard bass B in measure 7 as a passing note connecting the
dominant root and third, Smith perceives a “deceptive progression to VI” (p.
9). Schubert’s legato slur (connecting bass pitches B and D) is retained as an
analytical slur in Smith’s example 4b. I would begin the analytical slur on the
preceding A. Compare with TAH, 5.2b.
See TAH, pp. 134-136.
10 I am claiming that Bb, which Schubert spells as if it were a wobble, functions
instead as a neighbor. There seems to be an unwritten rule that a composer
should try whenever possible not to employ two different pitches with the same
letter name concurrently. For ease of reading composers tend in such contexts
278 Notes and references to pages 72-80

to misspell a pitch. Schubert does occasionally violate this rule: for example, in
measure 34 and its replicates in the second movement of his Piano Sonata in A
Minor (D. 784), where E’s upper neighbor F4 and G’s lower neighbor F# sound
together. That movement will be explored in chapter 8.
iil See TAR pp 26-132.
12 That is, in D-F#-A to D-F#-B, the 6 phase is represented by D#-Fh-A-C4.
13 In the Trio from the Piano Sonata in D Major [2.21], tonic’s upper-third chord
arrives where tonic is expected (measure 163). This is not a backward step, but
instead an atypically gradual emergence of I-space. What many analysts call a
“deceptive progression” sometimes leads from V to a chord such as VI, which
may serve as a substitute for I (it is the tonic’s 6-phase chord), though in some
cases (e.g., 3.1, measures 6-7) what initially may seem like a resolution turns
out instead to foster the dominant’s prolongation.
14 In “Schubert, Chromaticism, and the Ascending 5-6 Sequence” (2006), I
explore creative variants of the ascending 5-6 sequence in detail.
15 See Matthew Riley, “The ‘Harmonic Major’ Mode,” Music Analysis 23 (2004),
pp. 1-26, for a discussion of such variants.
16 The chord at the downbeat of measure 25 embellishes the Ab’ that arrives in the
second half of that measure. The Ab’ chord is in 3position, and so the preceding
chord appears not in §position, a conventional configuration for a lead-in to a
dominant-emulating chord, but instead is unfurled as a $.
17 See TAH, 3.8 and 3.9.
18 Compare measures 414-417 with 3.2, measures 1-4. Within the local tonic
expansion of the movement’s opening measures, C-Eb-F#-A performs an embel-
lishing role — a combination of neighboring and wobbly notes above the tonic
root. In the latter passage its harmonic potential (as an altered II) is unleashed.
iS This statement will generally hold true when Bb is absorbed as a vital compo-
nent within the chord. Other contexts would proceed differently, for example
a gradual transition within I>-®, as I®-”-°, in which the seventh serves as a local
passing note; or, the functioning of Bb as A%, in which case the root would shift
to F# for an F#>B succession.
20 Passing note C connects the melody’s Db and Bb. This dissonant C is made
locally consonant at the end of measure 12 through the support of IV’s upper-
fifth chord. Compare with 1.24.
21 In measure 3 the F minor chord has greater structural weight due to the bass
leap that follows: Ab>G>F>Db. In that case the harmonic progression is [5°
Il. The dominant arrives at the downbeat of the fourth measure. In the second
phrase [3.20], IV serves as the intermediary between I and V, thereby induc-
ing the tonic to pull instead in that direction, via I8-”. The dominant arrives
at the phrase’s fifth downbeat. Kofi Agawu assesses both phrases in his Music
as Discourse (2009), pp. 30-35. Though he notes that melody pitch C sounds
three times in measure 3 and four times in measure 11, that alteration does not
persuade him to interpret the F-Ab-C chord supporting the third C differently
 Notes and references to pages 82-87 279

the second time. In contrast, I propose that during the first phrase Schubert
proceeds to that chord (as tonic’s 6 phase, en route to II), whereas he proceeds
through it (to the point where tonic’s dominant-emulating 7) arrives, en route
to IV) during the second phrase. I also would recommend a different reading
of the first phrase. Agawu himself advises one “to be mindful of the origin and
destination of every event” (p. 33). Pursuant to that “golden rule,” one should
note that the first phrase’s melody is devoted largely to an ascending filling-in of
the tonic’s ¢, third followed by a descending filling-in of the dominant’s ® fifth:

m. 1 2 3 4 5 6 W 8

melody: Ab Bb Cy iD Bb AbG F<F Eb

AMjoe its ll ===
aa i
(= Eb Major: |
Bares Vie I)

Observe how the broad I>* II V progression of the entire phrase recurs within
the tonicization of dominant Eb Major. One of Agawu’s models (p. 34) shows
the ascent from A> to C, but the dominant-prolonging Bb>Eb fifth is not
acknowledged. He instead displays an ascending third, C<D4<E}, which in
my view is a chunk of a coordinating ascending line traversing the dominant’s
», fourth. Either way, his chopping of fourteen measures of music into eleven
“Units” (p. 34) conceals the “origin and destination” of several events, because
these events participate in trajectories that extend beyond the limited purview
of any of his individual Units.
22 Harmonically conventional, though melodically adventuresome! The struc-
tural line initiated by 3 in measure 2 continues to neighbor Bb on the last beat
of measure 3 (at the end of A<C and D>Bb unfoldings). Measure 4 “should”
support an A>G passing motion to 2, followed by F (1). Neither A nor G is
found there. (The A sounds an octave lower, however.) Compare with 1.34,
measure 72, where the melody likewise focuses on the sixth from 5 down to the
leading tone instead of traversing the stepwise path to i.
3, See TAH, 6.18 through 6.21.
24 I address this work in greater detail (including an assessment of how the text
influences Schubert's musical invention) in TAH, pp. 79-84. As was also
the case in 2.9 and 2.12, our seven-note notation system cannot effectively
accommodate an obstinate sequential ascent proceeding in half steps (which
in my view is functioning within a chromatic modulo 12 environment). Only
when one employs enharmonic respellings does the goal chord conform to
the broader tonal environment. (Otherwise Abbbb-Cbbbb-Ebbbb would appear
instead of F-Ab-C in 3.28.) The diatonic pillars I and IV are the only harmoni-
cally asserted chords within this sequential span.
25 Just as a 5-6 shift often facilitates the stepwise root motions I<II and IV<V,
here that technique is applied to V<VI in measures 48 and 49. G#-B-D-F repre-
sents the dominant’s 6 phase. Compare with TAH, 4.7.
280 Notes and references to pages 88-91

26 The downbeat of measure 2 contains a collision of chords similar to that at the

downbeat of measure 61 in the Piano Sonata in A Minor [1.1]. Here F#, Af, and
C# represent the belated arrival of tonic’s chromaticized 6 phase, while D and
shared tone F# assert the arrival of II.
27 Compare with TAH, 5.8a and 5.8b, where a descending glide occurs on the
upper staff and an ascending (diatonic) stepwise motion occurs on the lower
staff.
28 To my ears, measure 86 presents two independent initiatives, each pursued
vigorously. On one hand - the left hand — the connection is purely linear, with
a conventional glide of diatonic §chords linking the pitches ofII and V. On the
other hand - the right — a IIy~-# evolution, in the service of dominant emulation,
occurs. Though I do not, others might reasonably interpret D#-F#-A-C# at the
end of measure 86 as an evolved II (with absent root, raised third, and added
seventh and ninth).
29 Compare with FC, fig. 23.
30 My interpretation of A} as neighbor of the tonic chord’s G (as the 6 within a
I>-* expansion) is reinforced by the eventual behavior of Gb, which at first
seems to serve as the A} chord’s minor seventh. Though G) does not sound
during measure 20, it would seem that it resolves to G in measure 21 and thus
might better be spelled as F# to reflect its role as a neighbor. The Ab’ chord ulti-
mately functions as what nowadays is often called a “common-tone (or embel-
lishing) augmented sixth chord,” built from the retained root and third of the
C minor tonic, plus upper and lower half-step neighbors ofthe fifth. Common-
tone resolutions were discussed by analysts of the nineteenth century: Swoboda
addresses them in his Harmonielehre (1828), mentioned in TAH, p. 313, n. 14;
while Weitzmann addresses them in his Der verminderte Septimen-Akkord
(1854), sampled in TAH, 7.22a. Ab-C-E}-F# and A-C-Eb-F# are slightly differ-
ent formulations of the same procedure. We encountered the latter in 3.2.
31 As in the circle of fifths from Schubert’s String Quintet in C Major [3.19],
the initial progression warms up with an expansion of tonic before the fifths-
motion begins in earnest: C-G-C—F ...
32 Whereas Diabelli allows C-E-G to sound for two measures before adding
seventh Bb (measures 21 through 23), Schubert waits only one beat in measure
21 before adding Bb. Performers need to distinguish the central importance of
tonic C’s return from the local chromatic and dissonant modifications. (This
statement pertains to measure 8 as well as to measure 21.) Consequently I disa-
gree with Brian Newbould’s interpretation in “Cornered in the Middle Eight”
(2003, p. 110): “Schubert’s procedure is to re-characterize his C, the leading-
tone of D flat, as the dominant of F.” Instead, 3.37 proposes a tonic focus
extending from measure 1 through measure 24. Newbould’s F tonic (measures
25 and 26) resides within a C>F-Bb—-Eb circle-of-fifths connection of the
tonic and mediant. The dominant-emulating modifications to the C chord
within measures 21 through 23 are of local significance only.
 Notes and references to pages 91-99 281

33 This topic was introduced in the context of the Trio from the Piano Sonata in D
Major, pp. 60-64. See especially n. 20 on p. 276.
34 Ernst Oster, who translated and edited FC, adds an important footnote to
his translation of Schenker’s explanation of “Division by Interruption in the
Case of 3.” He asserts that Schenker intended a background structure like that
displayed in my 3.36 even when his notation does not appear in that format.
(Another notational style, which Oster suggests is of earlier genesis, displays
interruption using two independent beams apparently at the same structural
level: 3-2 followed by 3-2-1.) See FC, p- 37, ne 7.
ee) Compare Schubert's deployment of tonic’s lower-third chord here with
Rossini’s deployment of tonic’s upper-third chord in a number from Guillaume
Tell that I discuss in TAH, pp. 213-219. In both cases 3 is prolonged.

4 Chords on 5II, on III, and from the parallel key

1 Recall our discussion of the Trio in G Major from the Piano Sonata in D Major
(chapter 2), where the quality of the Fs chord in the circular progression G-C-
F2-B is adjusted to major [2.24].
Raymond Monelle offers a fascinating account of how music represents horses
and hunters in The Sense of Music (2000), ch. 3. This passage from “Der Jager”
appears as his fig. 3.9.
In previous endnotes I have noted how the diatonic bias of music notation
produces inelegant results in the context of intensely chromatic music, such
as a 5-6 sequence ascending obstinately by half steps. Roman numerals for
chromatic chords or tonicized regions are likewise inelegant. For example, the
lowered mediant in C Major might be displayed as III”. Though it would not
be difficult to develop an analytical practice using modulo 12 Roman numer-
als (I through XII, wherein C to E> would be I-IV and C to E would be I-V),
that system would have liabilities as well. (For example, in a seven-numeral
system tonic’s asserted “lower-third chord” is always a kind of VI. In a twelve-
numeral system, it might be IX or it might be X, a state of affairs that would
complicate considerations of chordal function. I explore this issue further in
chapter 12, n. 17.) Given my rather stringent standards for bestowing har-
monic labels, my analyses are less saturated with Roman numerals ~ especially
chromatic Roman numerals - than are those of most other scale-step analysts.
Though I do not find such symbols to be particularly onerous, I do lament their
inelegance.
This passage develops a strategy that Schubert employs in less elaborate for-
mulations as well. For example, in the opening movement of his Piano Sonata
in A Minor (D. 537), tonic A Minor and mediant C Major are connected via
their respective dominants. Dominant root E in measure 7 supports a chord
structure that extends to the ninth (E-G?-B-D-F). Then a straightforward
passing motion in the bass (E<F#<G) establishes root G. The persisting chord
282 Notes and references to pages 103-113

components B, D, and F join G (taking on Ab as well) in targeting the mediant,

which arrives in measure 11.
nn The lowered supertonic chord is addressed in detail later
in this chapter.
6 To do full justice to a reading in B Minor, we perhaps should back up to
measure 18, regarding the D# of measures 22 through 24 as a chromatic modi-
fication of the minor tonic (in the service of dominant emulation), thus Binet
In this perspective the D# is an internal voice-leading event within the circle of
fifths leading to III.
A composition in B Minor will often end in B Major (“Picardy third”). Here,
most unconventionally, the direction of the modal shift is reversed: from B
Major to B Minor.
Schenker explores the bass descent of a third to the dominant — sometimes
employing and sometimes not employing a Roman numeral for the upper-
third chord - in FC, figs. 111 and 113. His fig. 114, ex. 5e, shows a passage by
Chopin in which a minor chord a minor third above the dominant root initiates
V-space, as in the Moment Musical. (That possibility could be added to the set
of models presented in 1.8.) See also my “Schubert, Schenker, and the Subtonic
Chord” (2010).
See, for example, Kirnberger’s and Jelensperger’s charts of relatedness in TAH,
pp. 210-211 and 8.11.
In TAH, pp. 198-209, I explore a wide range of eighteenth- and nineteenth-
century analytical responses to the bII chord.
1 See FC, fig. 74.
12 Compare with the sequence/glide shift in “Aus Heliopolis Il” (D. 754), dis-
cussed above [3.28] and in TAH, pp. 81-84. The embellishment of a sequential
progression via upper-fifth chords is demonstrated in TAH, 5.11, where E-G-
A-C# (measure 2) serves as upper fifth to the sequence’s third chord, D-F-A.
In Schubert’s progression the upper-fifth chords are more highly evolved: for
example, G>-Bbb-C in measure 40 in place of Gb-Ab-C-(Eb).
13) See also 4.15, measure 20.
14 That is to say, an analyst may well be uncertain whether it would be better
to regard the chord of measures 4 and 5 as the confluence of melodic embel-
lishments within a tonic prolongation (compare with 3.1), or instead as a Vi
harmony clashing with root G. For me, the vocal F#-D-C-A arpeggiation of
V’ in measure 5 tips the balance in favor of a harmonic reading.
15 Compare with a formulation by Simon Sechter (with whom Schubert inter-
acted briefly on theoretical matters in 1828) discussed in TAH, pp. 200-203.
16 I discuss both of these works in “Liszt’s Composition Lessons from Beethoven
(Florence, 1838-39): ‘Il penseroso’” (1990). Schenker’s analysis of Beethoven’s
Funeral March appears in FC as fig. 40, ex. 6.
7 In Die schéne Miillerin we first encounter a happy-go-lucky miller lad effusively
expressing his wanderlust in the key of Bb Major. Over the course of the cycle
he meets and unsuccessfully woos the beautiful miller maid. Schubert sets the
 Notes and references to pages 113-126 283

twentieth (final) song, by which point the lad has drowned himself, in Bb’s
antipodal key, E Major.
18 Compare my distinction between two consecutive Bb chords here with
Schenker’s distinction between two consecutive G chords in a sonata move-
ment by Mozart (FC, fig. 47, ex. 1). In both cases, only the second of the two
chords participates in the deeper structure.
19 The tonal trajectory backs away from Db only so far as Bb. The ground from G
to Bb is no longer in dispute. The G major chord of measure 9 is hierarchically
dependent upon III. It serves as the mediant’s chromaticized 6 phase, not as a
restoration of the background tonic.
20 From a voice-leading perspective, F is a downward-tending dissonance, and
in fact it descends to E in the keyboard part at the downbeat of measure 17.
The dominant’s seventh, G, reaches over and displaces E in the vocal line.
Concerning reaching-over (Schenker’s Ubergreifen), see p. 72.
21 Schenker displays several prototypes for this interruption structure in FC, fig.
23s
Pe; Lawrence Kramer offers a thoughtful assessment of the song and its text
(though without a detailed account of how Schubert’s harmony responds to
the poetic content) in “Performance and Social Meaning in the Lied: Schubert’s
Erster Verlust” (1994). Concerning the notion of Hope, he writes: “The young
man cherishes not only his wound but also a fantasy of repetition, a return of
the first love in the person of a second, that he knows, or says he knows, to be
impossible” (p. 20). Kramer’s exploration of this song continues in his Franz
Schubert: Sexuality, Subjectivity, Song (1998), pp. 9-26.
25) A conventional first-inversion dominant seventh falls into place on the next
beat.
24 This sort of diminished-seventh resolution was acknowledged by several
nineteenth-century authors, including Swoboda and Weitzmann. See TAH, pp.
190-197 and 313.
25 In “Pitch-class Motive in the Songs of Franz Schubert” (1992), Steven Laitz
offers an intriguing alternative analysis, interpreting the Eb chord of measure 5
as an internal element within a [,xP” voice exchange encompassing measures 4
through 6. I instead relate bass F>D}<E} in measures 3 through 5 to bass F>E}
in measures 1 and 2 and thus regard Eb in measure 5 as structurally deeper than
the preceding Db. Only upon its third iteration, in measure 9, does Eb finally
succeed in realizing its potential to progress to Ab.
26 Kramer calls this passage a “chain of sobs” (1994, p. 19).
27 This construct often serves as the foundation for a dominant prolongation: for
example, in the third movement of Mozart’s Piano Sonata in Bb Major, K. 281,
mm. 60-63.
28 Compare with the similar situation in a passage from a sonata by Clementi
presented in TAH, 5.7. This parenthetical G corresponds to background 2 in
both Jonathan Dunsby’s and my readings. See his Making Words Sing (2004),
284 Notes and references to pages 128-137

pp. 125-132, where a perceptive commentary on the song touches upon issues
not covered here.
29\ SeevFAH, pps 155=160.
30 Compare with the Scherzo from Schubert’s “Death and the Maiden” String
Quartet (D. 810), where D Minor is firmly established in the first phrase before
material like that which opens the Landler (only twenty Deutsch numbers
earlier) appears to initiate the second phrase (mm. 9ff.).
31 These pitches also sound a sixth lower, as they did in measures 1 through 3.
32 Compare with FC, fig. 38 and fig. 40, ex. 8.

5 “Ganymed” (D. 544) with Lawrence Kramer and Suzannah Clark

1 See especially Maynard Solomon, “Franz Schubert and the Peacocks of
Benvenuto Cellini,” 19th-Century Music 12 (1989), pp. 193-206; and a forum
on the topic (edited by Lawrence Kramer) in 19th-Century Music 17 (1993),
pp. 3-101. The tradition of characterizing Schubert’s music as feminine, par-
ticularly when juxtaposed with Beethoven’s masculine output, began with
Robert Schumann, who nevertheless played a central role in raising Schubert’s
stature in the eyes of his contemporaries. Schubert’s Mddchencharakter is
exhaustively explored in Scott Messing’s Schubert in the European Imagination
(2006-2007).
iV) I recuse myself from assessing the contributions of Susan McClary, who has
responded provocatively to this question. She and I were colleagues at the
University of Minnesota, where she served officially as my mentor until I
attained tenure. Likewise analyses by my former students Walter Everett and
Steven Rings, who have written engagingly on Schubert, are not addressed in
Harmony in Schubert. Selected contributions by these three scholars are listed
in the bibliography, below.
3 “The Schubert Lied: Romantic Form and Romantic Consciousness” (1986);
“The Ganymede Complex: Schubert’s Songs and the Homoerotic Imagination,”
in Franz Schubert: Sexuality, Subjectivity, Song (1998), pp. 93-128.
“Schubert, Theory and Analysis” (2002).
a Ch—F) is expressed first in measures 49 and 50, though at that point the Cb [B]
chord occurs in second inversion, with minor ninth, and without root, while
the Fb [E] chord occurs in first inversion. A stronger arrival on F) occurs at
measure 56.
6 Initially one likely would regard Ab Major as the lied’s tonic, in which case Fb
major would be the second chromatic variant of Ab’s lower-third chord. On the
other hand, by measure 56 one might instead regard the opening A} major as
the first chromatic variant of Fh Major’s upper-third chord.
7 Compare with my commentary to 4.8.
8 This procedure is similar to the “seismic composition” by Verdi in Luisa Miller
that I explore in TAH, pp. 230-237.
 Notes and references to pages 138-145 285

9 In TAH, 1.16, I explore some early analytical practices in which II is interpreted

as a V chord with absent root and third. In “Ganymed,” the first half of measure
17 offers an interesting test case. Is Eb-Bb-D} a V’ chord, or instead a II chord
over dominant pedal?
10 One could make the case that the E} chord of measure 7, beat 2, serves as the
dominant chord that I place within parentheses, rather than as support for a
passing Bb during a three-beat tonic prolongation. In that reading (which is
inspired by FC, fig. 43, ex. b4), the melodic C>A} is unfolded on the strong
beats, interweaved with the unfolding of B}>E} on the weak beats. My resist-
ance to this conception involves both the maintenance of a leisurely pace for
the harmonic shifts for as long as possible and the motivic parallel between the
Db>C>B) third of measure 6 and the C>Bb>Ab third of measure 7.
ge) Schubert’s repetition of the final lines of the poem extends the alternation.
Goethe’s text is complete as of measure 94.
12 Since Clark makes a point of finding similarities between Schubert’s “Ganymed”
eee

and an earlier setting of the poem by J. F. Reichardt, it should be noted that

Reichardt likewise prolongs the dominant in exactly this manner in measures
10 through 12 ofhis song.
13 The appealing symmetry on display in his bass (graph B) over the course of
the entire lied - Ab<C) |A>F - is in my view untenable, because the A chord
of measures 64/66, which Kramer interprets as iii in F Major and, locally, as I
(then i) in A Major, is internal to a tonicization of Fb [E] Major. (Compare with
my 5.6.)
14 My understanding of Clark’s view is based on the assumption of a typographi-
cal error on p. 236. The passage reads as follows: “Just as Ab major became Cb,
so F# major [measure 46] will briefly become D major, though it will be the
F# that is retained in the bass and that will move to a G# [measure 50], itself
supporting a first inversion harmony.” The only way I can make sense of this
sentence is if “D major” is changed to “B major [measure 48].” As stated above,
I interpret that succession as I-IV, not as VI. P#-A#-E-Gb is an evolved,
dominant-emulating tonic. Alternatively, might Clark instead have misread the
D¢ on beat 4 of measure 49 as a D4, in which case F#-C-Ab-D4 would be her D
major chord?
15 Other words that Kramer invokes to describe Schubert’s harmonic practice in
“Ganymed” include “impressionistically conceived,” “floats unpredictably,”
“arbitrary,” “a little bizarre,” “paradoxical,” “phantasmagorical,” and “wander-
» « » « 2 «

ing” (1986, pp. 226-228). Given my disagreement with much of his analysis, I
cannot endorse this assessment. His 1998 book presents a similar perspective
on the lied using more sober language: “The apparently free-flowing stream
of tonalities does follow a logic which is, or evolves into, that of the higher
eros. From the lower-erotic standpoint of nature . .. that logic can appear only
as illogic; it has to be heard transcendentally to become intelligible” (1998,
p. 122).
286 Notes and references to pages 145-150

16 A characteristic Schubertian structure might continue from 5.1 with an evolved

supertonic chord (F}-D4-Ab-Ch) proceeding to a cadential $ (bass E>, soprano
Ck), from which a suitable close in Ab Major could easily be accomplished.
J. F. Reichardt’s “Ganymed,” another setting of Goethe’s text, likewise traverses
a descending major third (measures | through 19), at which point an evolved
supertonic arrives to direct the progression to V and I. Then the descending
major third recurs in conjunction with Ganymede’s ascent, and this sub-
mediant (or usurping tonic?) chord is prolonged through the end of the song.
Though the local progressions contrast those of Schubert’s setting, the third-
relationships are similar.
Wi It appears that Clark does regard A as a tonic, however, since she describes the
preceding E chord as a “dominant” (p. 236). My IIV in E is for her VI in A.
The figure for bass B in Clark’s graph (measure 63) lacks a natural to the left of
the number 3. This matches Kramer’s omission of a natural sign to the left of
the D notehead in measure 63 of his graph.
US) Compare with FC, fig. 19a.
20 It is worth underscoring that Kramer, Clark, and I are in agreement that the
first chord of measure 75 functions as the tonic in F Major. In “Hinauf strebt’s:
Song Study with Carl Schachter” (2006, pp. 192-194), Timothy L. Jackson sug-
gests instead that “the F chord of bars 75ff. does not yet function as the tonic,
but rather as a passing chord (V7 of IV) inserted between A} (bar[s] 1ff.) and Bb
(bar 79).” He reads the structure for the entire song as bIII - IV - V - I, with T’s
arrival occurring “definitive[ly]” in measure 116. Unfortunately his handwrit-
ten Schenkerian graph (p. 194) is so reduced and so poorly scanned that it is in
many places illegible.

6 Quintet in A Major (“Trout,” D. 667), movement 1, with David Beach

1 This variant of a sonata movement’s structure is discussed in Hepokoski and

Darcy, Elements of Sonata Theory (2006), pp. 262-268. Readers familiar with
the authors’ terminology and abbreviations will benefit from consulting their
exploration of the topic, while anyone might appreciate their extensive list
of works that fall within this category. Their central point is that “although
infrequent, recapitulations starting on IV turn up consistently enough in the
eighteenth century that we consider it a lower-level default option within the
genre, not a deformation” (p. 264).
“Schubert’s Experiments with Sonata Form: Formal-Tonal Design versus
Underlying Structure” (1993).
Schubert pursues a similar cyclic strategy in the Impromptu in C Minor (D.
899/1), explored in chapter 10. There each cycle proceeds from I to Vi.
That is, measures 25-26, and also measures 38-39, are the end of something,
not the beginning of something. Measures 27 and 40 are hypermetric down-
beats, as is measure 3.
 Notes and references to pages 150-156 287

> A similar construction occurs in a passage from Beethoven’s “Ghost” Trio (op.
70, no. 1) that Robert Hatten addresses in his Interpreting Musical Gestures,
Topics, and Tropes (2004), pp. 21-26. Hatten goes to some length in assessing
the passage from what he purports to be the perspectives of Heinrich Schenker,
Leonard B. Meyer, Rudolph Réti, and E. T. A. Hoffmann without exhausting
the potential readings. I question the assertion that “an unstable German aug-
mented sixth . . . resolves in contrary motion by half steps to a cadential AP:
23). Though I understand that this moment is a significant breakthrough and
that some of the built-up tension is released, neither the 6 nor the 4 of the §has
yet resolved. From a structural perspective the resolution is not complete until
the following 3chord.
Schubert’s spelling of such chords is variable. Often when a cadential ¢follows,
the pitch here spelled as Ck is instead spelled as B# to accommodate the ascend-
ing motion to the dominant fifth’s upper neighbor.
N Continuing the point raised in n. 6 above, now ninth Eb’s ascending motion to
the neighbor E in measure 55 (interpreting that measure as the unfurling of a c
chord embellishing G’) results in its spelling as D#.
Compare the use of E# en route to IV’s F# in measure 58 with earlier instances
in measures 34-35 and 47-49.
Suzannah Clark assesses both this movement of the “Trout” and Beach’s
analysis of it in “From Nature to Logic in Schubert’s Instrumental Music”
(1997), pp. 29-32 and 156-160. Whereas I interpret the D chord of measures
48 through 54 as functioning within an extended prolongation of A Major
[6.1, Model 2], she regards the D chord of measure 50 as the theme’s endpoint,
followed by a transition (p. 156). Her ex. 5.5 (p. 157) positions this D chord
between A in measure 25 and E in measure 64, a reading that I reject. Despite
our differences Beach and I agree that A Major is prolonged through measure
62.
10 Note that this passage is omitted from the recapitulation (at mm. 285ff.), where
it would have occurred in tonic A Major. The passage corresponding to the
introductory A>Fs (measures 1-18) is likewise omitted.
11 Compare with TAH, 5.2 and 5.3.
12 Compare this C>Ab connection with the Ab>Fb connection in “Ganymed”
(5.1 and 5.3]. Note especially the similar but not identical treatment of Ab’s
chromatic upper-third chord (C}) in “Ganymed” (measures 31-40) and C’s
chromatic upper-third chord (Eb) in the “Trout” (measures 164-170).
13 This is yet another case in which the diatonic bias of music notation makes
Schubert’s writing seem stranger than it is. (Compare with my comments on
5-6 sequences ascending by half steps in chapter 2.) Here Schubert divides
the octave into 4 + 4 + 4 half steps. In a modulo 12 system (with C = 0), the
bass of 6.3, measures 64 through 248, would appear as 4>0>8>4. Notating
such a progression in the context of our seven-note diatonic system, the
concluding Ab>F)/E “third” brings on superfluous concerns about an
288 Notes and references to pages 156-157

“enharmonic seam,” a notion that will be discussed in greater detail in

chapter 12, n. 15.
14 Since Beach clearly favors his Model a over Model b in his example 8 (pp.
11-13), and since his Model a is closer to the model I present in my 6.3, I limit
further consideration of his Model b to n. 16.
WS) Compare with Schenker’s analysis of the opening theme from Beethoven's
Symphony No. 8 (FC, fig. 150). Measures 3 and 4 present a diminished fifth
whose Bb we expect will descend by step. Schenker does not accept the A of
measures 7 and 8 as the true successor. That moment is not connected to his
broader progression. A reiterated Bb instead leads through the A of measure 11
to G and then F. In similar fashion one might ultimately reject the bass D that
opens Schubert’s recapitulation as the true successor of the development’s C
and C#. Like the B} in the Beethoven symphony, C and C# are reiterated (meas-
ures 245-246) and then followed by D and E to complete the ascending motion
that stems from the Ab in measure 189.
16 Suzannah Clark (1997, ex. 5.5, p. 157) likewise jumps directly from Ck (the
curiously dangling last pitch in her graph of the exposition) in measure 106 to
D in measure 210, offering no account of the context in which D emerges. Her
interpretation reduces ultimately to

m. 1 64 210 249

I Vv Vial

This corresponds to Beach’s second hypothesis (his ex. 8, Model b, p. 12). Were
I working from Clark’s perspective (that is, without my contrasting interpre-
tation of measure 210’s D as displayed in 6.5), I would invoke the notion of
“peculiar juxtaposition,” introduced in chapter 1. Though

m. 1 63 64 210 248 249

i |B 8 Dwr
eee cA

might seem to give more structural weight to D than to the second E, ultimately
the pull of tonality causes me to regard D as subordinate to the E chords that
surround it. (Compare with my preference for Hypothesis 2 over Hypothesis
1 in 1.26.) Though Clark includes fourteen chords in her ex. 5.5, neither the B
chord nor the second E chord appears. In my 6.3, which includes only seven
chords, both are present.
Another hypothesis emerges if we remove the parentheses that Beach places
around the A chord within the development. We then would have
 Notes and references to pages 157-159 289

which invites the proposal that tonic is reinstated early — at the end of the
development — rather than late, and that a tonic-confirming A~D-E-A pro-
gression is already in progress at the outset of the recapitulation. That this pro-
spective tonic possesses dominant-emulating characteristics does not prevent
it from functioning as a tonic. That is a common occurrence preceding iV. I
resist this hypothesis because the structural logic of the preceding E>C>A}
progression seems better fulfilled by an E chord [6.3] than by an A chord. Yet
the fact that both Beach and I entertain alternative readings in print is indica-
tive of how nebulous this particular region is.

ys Symphony in B Minor (“Unfinished,” D. 759), movement 1, with

Richard Taruskin
1 The Oxford History of Western Music, vol. 3: The Nineteenth Century (2005).
Chapter 34 (“The Music Trance”) is devoted to Schubert. In the 2010 paperback
edition, chapter 34 is renumbered as chapter 2.
Ibid., pp. 107-113.
WN
Ww Though the same opportunities for harmonic progression are available regard-
less of which pitch class functions as the tonic, the twenty-four keys were differ-
entiated to some extent through usage, and perhaps also through the subtleties
of non-equal temperaments. In Distant Cycles: Schubert and the Conceiving of
Song (1994), Richard Kramer suggests that “B minor... is a key freighted with
its own tropes and allusions, and nowhere more so than in Schubert’s music,
where it seems to call forth a language of its own, profound and at the edge of
despair” (p. 13).
wo My model is similar to FC, fig. 109, ex. e4, though without Schenker’s asser-
tion that II is the principal intermediary between tonic and dominant.
Schenker’s student Felix Salzer, in Structural Hearing: Tonal Coherence in
Music (1952), ex. 497, likewise does not read a II chord into the introduc-
tion’s melody. However, he does not correlate the b-G-F# contour of the
introduction and the tonal plan of the exposition, as I do. Compare his
analysis of measures 105 through 110, which he interprets as an expansion
of the tonic, with my 7.9, which interprets measures 107 through 109 as the
emergence of a structurally deep dominant to conclude the exposition, to be
followed by a return of the tonic for the repeat of the exposition or the begin-
ning of the development. Dominant F# is here expressed in a manner similar
to a dominant-emulating Ab chord from Schubert’s youthful “Erinnerungen”
(3.19, measure 25]. In the lied, the A} chord is a link within a circle of
descending fifths; in the symphony, the F# chord concludes the exposition’s
tonal trajectory.
On Descending tetrachords such as B~A-G-F# have “traditional semantic asso-
ciations with lament,” explored by Su Yin Mak in “Schubert’s Allusions to the
Descending Tetrachord” (2007), pp. 163-179. Taruskin likewise describes this
290 Notes and references to pages 161-169

tetrachord as “an emblem of lament” (p. 109), a term he borrows from Ellen
Rosand’s study of Monteverdi.
Compare with FC, fig. 56, ex. 2f. Though I agree with Schenker that C# passes
between B and D, I would not describe its chordal support as “consonant” (as
he does on p. 62).
I regard B-D-F# in measures 34 and 35 as an unfurled F#%, as if the sequence
were attempting to continue beyond goal F#},, and not as the arrival of tonic.
Those motions are offset by the persistence of root F#. The FB (Vio) suc-
cession in measures 37 and 38 falls outside the domain of the sequence.
In doing so I am attempting to avert the sort of criticism that I hurled upon
Gottfried Weber in TAH, pp. 147-148.
Compare Schubert’s connection here of G Major and C Minor with the tonal
relationship between the symphony’s two movements: B Minor and E Major.
10 See Schenker’s samples of unfoldings in FC, fig. 43. Though none of his models
depicts two consecutive fourths or fifths, he shows various instances in which a
succession may proceed either through alternating the pitches of two harmonic
intervals, or instead through presenting one strand followed by the other.
(Examples fl and f2 offer the clearest comparison.) See also the commentary to
TAH, 2.24b.
In their Elements of Sonata Theory (2006), Hepokoski and Darcy interpret the
broad trajectory of the exposition as “i- VI” (p. 317). In contrast, I regard the tonal
path of both the introductory theme and the exposition proper to be I>-°-V#. (The
correction of older published scores to incorporate Schubert’s pedal B sounding
against C# in the strings in measure 109 — and also in measure 327 — does not alter
this basic tonal trajectory. Instead it strengthens my case that measures 1 through
113 are a large-scale projection of tonic, to be followed by the subdominant.)
Compare with my concern, articulated above, that the dominant chord within
the G Major region [7.7, measure 79] will be undervalued. Duration, chromatic
enhancement, and thematic engagement (or their absence) do not always cor-
relate with structural depth. Here I propose that an inverted, pianissimo, three-
measure dominant (measures 107-109) resides deeper within the structure than
does the preceding sixty-five-measure expansion of the submediant.
12 The approach to the background V# in the development has much in common
with the approach to the middleground V# within the opening tonic region of
the exposition. D-F¥-A (heralded by C#-E-G-Bb) in 7.3 is tonic’s upper-third
chord, asserted as the mediant linking I and V#. D-F#-A (likewise heralded by
C#-E-G-Bhb) in 7.11, Model 2, is a preliminary phase of the dominant arrival.
3! As is typical with obstinate progressions, the linear path takes no heed of the
diatonic context that prevails outside its realm. Yet the composer must employ
notation that was designed to favor diatonic progressions (seven noteheads
per octave). Consequently some enharmonic adjustment is necessary. In a
twelve-note chromatic system (with C = 0) the glide of Model 1 would appear
as 6<8<10<0<2<4<6. I contend that only the 6 plays a role within the broader
 Notes and references to pages 170-181 291

diatonic environment, as II# in tonicized E Minor (as displayed, with further

alterations, in 7.12).
14 We saw in “Aus Heliopolis II” [3.28] how an ascending 5-6 sequence morphs
into a glide. Now, in the Symphony in B Minor, a glide morphs into a sequence.
15 I explore an example of seismic composition by Verdi in TAH, pp. 230-237.
16 Two melodic lines approach 1 concurrently: whereas the clarinets and first
violins ascend F&<A#<B, the flute line descends D>C#>B. Compare with the
cadence in measures 35 through 38 and with 7.4, which posits a middleground
3-(2)-1 descent, with C# absent from the melodic line.

8 Piano Sonata in A Minor (D. 784), movement 2, with Robert Hatten

1 Hatten and I were colleagues at the University of Michigan School of Music

during the early 1980s. At the time we considered publishing side-by-side
analyses of a chosen movement as a means of comparing our contrasting
methodologies. We did not bring that idea to fruition at that time. This chapter,
which complements one of Hatten’s recent analyses, is a belated fulfillment of
that intention.
2 Hatten’s analysis of the second movement of Schubert’s Sonata in A Minor
(D. 784), which I will assess, appears on pp. 194-198.
Ww Once we have fully assessed its context within the phrase, it will be apparent

that C<F<A<C is best understood as a C<F<A sixth followed by an A<C third.

This issue will be addressed in n. 8, below.
4 The D} chord of measure 21 seems a more natural successor of the F Major
region because the Dp pitch has already sounded several times by that point.
nn Though he acknowledges that “one cannot know Schubert's intent in this
regard” (p. 196), Hatten repeatedly refers to it as the “serpent” motive in his
ensuing discussion.
6 Hatten acknowledges that “Schubert composed the A minor sonata, D. 784,
in 1823, perhaps in response to a foreboding of death from having contracted
syphilis” (p. 187). However, he does not interpret Schubert’s motive specifically
in that context. In addition to the reference to Satan, he describes it as “fateful,”
“an otherworldly reminder of something less desirable,” and “if not a spe-
cifically evil connotation, at least a reminder of the lurking force of the tragic
behind the beauty of the surface” (pp. 195-196).
7 Hatten explores the “abrupt release on an upbeat eighth note” in measures 2
and 4 of the first movement, part of a gesture that he describes as “a weight that
is too heavy to bear and must be constantly shrugged off with a sigh” (p. 187).
8 The relationship of A in measure 1 to G and F in the following two measures,
as well as the emergence of A an octave higher in measure 6, suggest that the
opening arpeggiation (C<F<A<C, with neighbor D before the upper C) should
be regarded not as a C<C octave, but instead as a C<A sixth followed by its
inversion A<C. Thus the initial inner-voice C is raised an octave. Though 5 at
292 Notes and references to pages 183-184

first might seem a reasonable Kopfton for the structural melody, 3 eventually
emerges as the better choice. I acknowledge A’s crucial structural role in my
analysis of measure 1, and recommend that performers interpret this A with its
functions as starting point for both the A>G>F downward motion and the A<A
coupling in mind. For example, the use of rubato on A to delay the arrival ofD
could reinforce the emphasis created by the release of the accompanying chord
that sounded at the outset, assuring that the opening arpeggiation comes across
as motion to A, not through A. Schubert’s slur in measure 1 merely requests a
legato performance. It should not be mistaken for an analytical slur such as one
would find in a Schenkerian graph.
Though I would like to report that Schubert always spells chords in a con-
sistent and exemplary fashion, the evidence often contradicts such an asser-
tion. For example, compare the first chords of measures 7 and 12. Though
configured differently, their roles are identical. Yet in the first case Schubert
employs C# and in the second D). Perhaps in measure 7 he was attempting
to suggest the potential resolution shown in 8.2, which ultimately he did not
pursue. Yet perhaps he simply did not make the effort to ensure that his spell-
ings were consistent and meaningful. Likewise, if Db (not C#) is employed
in measure 12, then Gb (not F#) should have been employed in measure 11.
Another instance in which his spelling is inconsistent is in a chord such as
that containing C# in measure 9. In its context C# represents root C’s minor
ninth, and in fact it is often spelled as Db. Sometimes - as here — Schubert
instead spells it as C#, when resolution pitch C is embellished by upper neigh-
bor D.
10 Continuing the discussion of Schubert’s enharmonic (mis)spellings from n. 9
above, observe that A-C#-E-G-B} in F Major (measures 6 and 7) is recast as
E-B-D-F-Ab in C Major (measures 36 and 37) rather than as E-G#-B-D-F. I
attribute this inconsistency to indifference rather than to some deeper implica-
tion that music analysts might endeavor to illuminate. Often in my analysis of
intensely chromatic passages by Schubert, the chordal spellings that I deem
best from an analytical perspective do not match those in the score. Though I
endeavor not to be hasty in overriding Schubert’s spellings, I am willing to do
so without fretting about it.
11] Hatten and I interpret measures 9 and 10 of the opening movement in mark-
edly different ways. Whereas he proposes a “iv-i progression” (p. 187), and
thus a transfer of root back and forth between D and A during measures 9
through 16, I regard A as a prolonged root throughout this passage. In my
perspective D is not a root, but instead a dissonance against root A. C<D>C
and E<F>E above it are neighboring motions, akin to the E>D#<E neighbor-
ing motion that opens the movement. Though Hatten’s interpretation accords
with a prominent nineteenth-century perspective (compare with an example
from 1830 by Jelensperger, presented in TAH, 5.4), analyzing this passage as
iv-i is akin to treating a cadential °-3 as I-V, a practice that I reject.
 Notes and references to pages 184-197 293

12 Taking into account measures 45 and 48 as well, this gesture embellishes all
three pitches of the F tonic triad at some point during the Andante.
13 My caption for 8.4 harbors a contradiction, in that one reading of the conclud-
ing passage (from By to G)) places chords in a different hierarchy from the
other (F¥ G3). In this case the restored bass F forms a long-range connection
that is not inherent in the descending-thirds pattern. I consider a similar con-
tradiction within the ascending 5-6 sequence in “Schubert, Chromaticism, and
the Ascending 5-6 Sequence” (2006), examples 3f and 7, where C> D> E>
F transcends the hegemony of 5-phase chords and expresses a prolongation of
the C Major tonic from C° through E* before continuation to IV. Compare also
with 3.26, above.
14 Compare with the G<A><A<B) passage in 2.10 and with examples 8 and 12 in
my “Schubert, Chromaticism, and the Ascending 5-6 Sequence” (2006).
15 Compare 8.8 with FC, figs. 21b, 23a, 76.2, and 154.5a. Schenker’s notation for
the display of interruption is inconsistent.
16 Numerous names for this form, which occurs frequently in slow movements,
have been proposed. Schenker calls it “four-part form,” whereas Hepokoski
and Darcy call ita “Type 1 sonata.” See their Elements of Sonata Theory (2006),
pp. 345-352, for an introduction to this form and for their review of how other
analysts have named and interpreted it.
17 Similarly, Adolf Bernhard Marx suggested that a passage by Beethoven that I
analyze in TAH “loses its way.” (See TAH, pp. 134 and 306, n. 19.)
18 Compare this movement's final two measures with the final two measures in
the Impromptu in C Minor (D. 899/1), which I likewise interpret as a death
knell. (See chapter 10.)

9 “Die junge Nonne” (D. 828), with David Kopp

1 Chromatic Transformations in Nineteenth-Century Music (2002). “Die junge

Nonne” is addressed on pp. 200-201, 208-209, and 254-263.
Compare with 2.12 (measures 141-144), 3.28, and 8.7, Model 1, above;
with TAH, 3.16b and 3.17; and with examples 8 and 12 in my “Schubert,
Chromaticism, and the Ascending 5-6 Sequence” (2006).
A glimpse into the sorts of decisions a composer must make comes from a
careful assessment of the octave Cs at the end of beat 2 in measure 61. A closer
alliance with measures 1, 8, 35, etc., could have been achieved if Schubert had
instead written B>s. He chose instead to underscore the relationship with the
vocal melody of measure 60, where C occurs in the corresponding location.
Though Kopp acknowledges that C>A>F “arpeggiates the tonic triad” (p. 209),
he does not associate this arpeggiation with any others within the work.
Kopp analyzes the succession from B>-D-F to C-F-A (measures 59-60) as iVee
and that from C-F-A to C-E-G-B»> (measure 60) as another D~'. In contrast, in
the text he refers to the progression of measures 59 through 61 asa “IV -V -I
294 Notes and references to pages 197-204

cadence” (p. 209). Which pitch functions as the root during C-F-A (the caden-
tial 6) - F or C? He seems to want to have it both ways.
Given the perspective enunciated on his p. 209 (n. 7), if Kopp would acknow-
ledge this G, I assume he would modify his analysis of measure 60 as well.
In an embellishing context, these pitches are often described as a “common-
tone augmented sixth chord.” This notion is demonstrated in terms of a similar
(though not identical) chord in TAH, 7.22 (Weitzmann), and also is mentioned
by Swoboda (TAH, p. 313, n. 14: G#-B-D-F to Ab-C(b)-Eb). We will consider
this resolution again in chapter 12.
Compare with the close of Schubert’s “Das Grab” (analyzed in TAH, pp.
338-343), which likewise hovers above | at its close. In “Die junge Nonne,”
however, the keyboard provides a conventional descent to i at both measures
61 and 83, which the singer complements with an ascent to 8.
I interpret the Eb that sounds with this Gb chord as an anticipation of the fol-
lowing 6-phase chord’s sixth.
10 A diminished fifth root succession between VI and II occurs naturally in the
minor mode. Compare with measure 27 of the Piano Sonata in D Major, mvmt.
2, where B}-D-F to G#-B}-D-F constitutes a diminished fifth motion from }II to
altered V? in A Major. (See 4.24, m. 27.)

10 Four Impromptus (D. 899), with Charles Fisk

l For example, from a harmonic perspective measures 139 through 158 of the
Impromptu in Ab Major offer some of the most extraordinary writing in the
set. Fisk does not address that passage.
My interpretation of this passage as a theme and variation is more assertive
than Fisk’s descriptions as “eight opening phrases” (p. 125) and as four periods
(p. 126). Through measure 33, listeners might reasonably expect that the
impromptu will continue as a set of variations. (The Impromptu in Bb Major,
D. 935/3, is so constructed.)
I interpret the F-G#-Bk-D chord on beat 2 of measure 18 as dominant (G)-
Bk-D-F-Ab. Compare with the spelling of this chord in measure 177. Ab is a
downward-tending ninth, yet since II rather than I follows, the voice leading
proceeds as Ab<Ag rather than Ab>G. Consequently Schubert spelled the note
as G#.
Another potentially confusing alteration during the variation is the more
aggressive unfurling of the § chord os Whereas these pitches occur in both °
and {positions in measures 8 and 16 (where I regard G-G-G-E} as a chord that
implies a C as well), they attain 3position in measures 20 and 28. These chords
do not represent the tonic, however, but instead are embellishments of the
dominant.
Mehrdeutigkeit is explored in TAH, pp. 155-161. In measures 34 and 38,
the pitch Ab might be understood as the unfurled 6 of 15-5, without its
 Notes and references to pages 205-210 295

assertion as a root. Such uncertainty seems to be what Schubert intends here:

the wafting back and forth between G and Ab (against C and E}) leaves the
hierarchical relationship in doubt: is Ab a neighbor to G, or has G transferred
authority to Ab? Clearly the latter ultimately is achieved (A) Major is tonicized
through measure 82), yet the point at which that tonicization begins cannot be
pinned down precisely. We want to hear the G in measure 39 as a replication
of that in measure 35 (thereby again resisting Ab’s challenge), and yet in its
broader context it is hard not to hear this G as a neighbor between the preced-
ing and following Abs, enhancing a succession from I to IV in Ab Major.
B><E», which supported the soprano ascent from Eb through F to G in meas-
ures 26 and 27, translates into C)<Gb>C} in measures 48 through 51, support-
ing the soprano descent from E} through D to C). Note the less well-defined
though perceptible G> chord that precedes the first Ch in measure 48 (with
Db-Fp-B» representing G}’).
See my caution against equating tonicization and structural depth in my cri-
tique of an analysis by Gottfried Weber in TAH, pp. 147-148 (including n. 13
on p. 308).
(oe) Could my story actually have been Schubert’s story? My inability to provide

concrete evidence to support my narrative does not hold my imagination in

check. We do know from some sketch markings that Beethoven was thinking
about the vault scene from Romeo and Juliet during the composition of the
Adagio from his String Quartet in F Major (op. 18, no. 1). In Unfinished Music
(2008, pp. 171-186), Richard Kramer offers an assessment of the evidence in
that instance and ponders its implications.
Compare the root succession Ab-D here with Bb-E in 4.24, measure 24. In the
impromptu, the strategy connects 6 and 2 ina minor key, whereas in the sonata
it connects 52 and 5 in a major key. In both cases the roots are separated by a
diminished fifth.
10 Compare with 3.32, Model 3, where D-F-Bb comes between D-F-A and D-F-
Bg. Just as no )VII harmony comes between II and V there, no IV harmony
comes between VI and the evolved II in the impromptu. Instead, a short
segment of the circle of descending thirds connects the two harmonic chords.
11 A similar disagreement emerged between Taruskin and me regarding the large-
scale tonal plan of the first-movement exposition in the B Minor Symphony,
which also includes an extensive tonicization of the submediant. See chapter 7.
12 The progression from measure 185, beat 4, through measure 193 is very similar
to that of measures 726 through 733 of the String Quartet in D Minor, mvmt. 4
[1.14]. The following chart shows the correspondences:

Impromptu Quartet Interpretation

m. 185: C-E4-G m. 726: D-F#-A Major tonic
m. 185: Bb-D-F (pedal G) m. 727: Ck-E-G Passing chord in contrary
motion
296 Notes and references to pages 210-212

m. 186: A-C#-E-G m. 728: B-D#-F# Tonic’s 6 phase (evolved)

m. 188: G#[A}b]-B-D-F m. 729: At#-C#-E-G 6-phase chord’s upper-
fifth chord (highly
evolved)
= m. 730: D#-F#-A-Ck Restored tonic 6-phase
chord (further evolved;
quartet only)
m. 190: F#-A-C-Eb m. 731: G#-B-D-F¥ — supertonic (highly
evolved)
m. 191: G-B-D-F m. 733: A-C#-E-G Dominant seventh
m. 192: C-E-(G) — Tonic resolution
(impromptu only)

Bass Ab perhaps ought to have been spelled as G#. Schubert may have elected
the Ab spelling to avoid using Fx for G#’s lower neighbor at the downbeat of
measure 189. The concurrent presence of F and Fx at that spot, though justified
in theory, likely would have been disconcerting to many of Schubert's con-
temporaries. More substantively, comparison with the quartet reveals that the
impromptu’s harmonic progression lacks an important element. Bass G at the
end of measure 189 might well have supported the chord G-A-C#-E, restoring
tonic’s evolved 6 phase, were it not for the slow rate of harmonic motion in this
vicinity. Though one might complain that there is a syntactic error at this point,
certainly this region of the impromptu, which harbors other irregularities, is
where such an event might most reasonably occur. (I suggest above that these
irregularities convey a soldier’s delirious state shortly before his death.)
13 Chopin would deploy a similar contrast between the third and fourth move-
ments of his Sonata in Bb Minor (“Funeral March,” op. 35) a decade later.
14 Though a descending E}>E} octave does not quite reach fruition in measure 24,
that deficit will be repaid amply in measures 276-279.
15 Frank Samarotto offers detailed Schenkerian graphs of portions of the
Impromptu in E} Major as example 7 in his “Determinism, Prediction, and
Inevitability in Brahms’s Rhapsody in Eb major, op. 119, no. 4” (2007), pp.
69-100. Though overall his analysis is commendable, in my view his treatment
of register in measures 1 through 24 is insufficiently nuanced. The first cycle
fully engages both the low and middle treble-clef registers (as displayed in the
first “measure” of my 10.3). Samarotto instead stems the G in both the middle
and lower registers, the F only in the middle register, and the Eb only in the
lower register. Whereas I position the initial engagement of the higher register
during the second cycle, Samarotto regards that cycle as a mere “rep[eat]” of
the first cycle at an increased dynamic level. Likewise I position the departure
from the higher register (and reinstatement of the lower register) during the
third cycle, which he labels perfunctorily as “m[ain] d[{roite] 8” sopra.” The
passage’s drama stems from the registral shift in measures 15 and 16. Schubert
 Notes and references to pages 212-222 207

cannot simply walk away from that tantalizing event. He must either confirm
the upper register or relinquish it during a third cycle. Readers also might wish
to consult Steven Rings’s “Perspectives on Tonality and Transformation in
Schubert’s Impromptu in E}, D. 899, no. 2” (2007), in which Schenkerian and
neo-Riemannian approaches are integrated. Rings proposes 5 as the impromp-
tu’s Kopfton.
16 This issue will require further discussion later, when Eb} (spelled as D) in
measure 253 and leading tone D in measure 255 occur in close proximity.
We Likewise the G><Eb sixth of 10.8, Model 1, eventually ascends to Gb (as shown
in 10.4).
18 Samarotto (2007) offers a contrasting perspective on the background descent.
His example 7f (p. 83) and my 10.10 differ principally in that I regard measure
258 as the first of numerous cadences resulting from descending lines (with
the F of the dominant in measures 255/257 forming the link between a pro-
longed G) and goal Eb). Samarotto proposes instead an ascending arpeggiation
that continues through measure 267 [2687]. Rings (2007) likewise postpones
closure until measure 270. In my view the Bb of measures 267-268 resides
within a 3<5<8 ascending arpeggiation (whose components serve as the start-
ing points for multiple descents to 1) forming a secondary, post-structural-
close approach to the tonic pitch. (Compare with measures 60-61 and 82-83 of
“Die junge Nonne,” discussed in chapter 9.)
19 Only the first two impromptus were published - separately - in 1827. The
third and fourth were not available until 1857, at which point Schubert’s tonal
scheme was annihilated through the transposition of the third impromptu up a
half step to G Major.
20 The term “associative harmony” is sometimes used to refer to such relation-
ships. In “Harmonic Cross-Reference and the Dialectic of Articulation and
Continuity in Sonata Expositions of Schubert and Brahms” (2006), Peter H.
Smith offers an overview. Given my more historically based perspective, I prefer
to retain the noun Mehrdeutigkeit and adjective mehrdeutig when addressing
how the same chord performs contrasting functions within a composition.
21 Third-relations are a central concern in Murray Dineen’s Schoenbergian analy-
sis of this impromptu (based on sketches by Patricia Carpenter), presented in
“Tonal Problem, Carpenter Narrative, and Carpenter Motive in Schubert’s
Impromptu, Op. 90, No. 3” (2005).
Deciding on the precise hierarchical relationship among the chords in meas-
ures 3 and 4 proves to be challenging. Two competing hypotheses could be
supported: that the tonic returns at the downbeat of measure 4, followed by
the goal V; or, that the dominant arrives in the second half of measure 3 and is
prolonged through measure 4. My preference for the latter reading is based on
the register of bass pitch D} (twice) and comparison with the positioning of V
in the consequent phrase’s third measure. Compare with 4.38, Model 1, where
a dominant-prolonging §-4-3-2 line occurs in the soprano. In the impromptu
298 Notes and references to pages 222-240

an equivalent descent begins in the bass but is transferred to the soprano during
a prolongation of 3.
23 Consequently the Eb chords in the two phrases play contrasting roles. In
measure 2, E} is an anticipation of the II chord’s fifth (thus P-° II). In measure
6, Eb is a neighboring note (thus I**>-IV). In place of the diatonic return of
the 5-phase chord with added seventh (Gb-Bb-Db-F) Schubert substitutes the
highly evolved Bb-D4-F-A}, which in this context is dominant-emulating in
the direction of C), not Eb. (Compare with 1.5.)
24 Some editions of this impromptu show misleading and incorrect slurring
for this passage, impeding comprehension. A single slur should connect bass
pitches Fh and D} in measure 11, as also in measure 19. Only that slurring
(leading to a performance that proceeds through, not to, the chord on Eb) will
confirm the analysis that I propose in 10.11.
aD Compare with 2.23 (Model 2), where the arrival of the incomplete neighbor (C
in m. 157) coincides with the onset of IV’s 6 phase.
26 In his “Schubert’s Self-Elegies” (2008), James William Sobaskie offers yet
another detailed analysis of this impromptu, like me incorporating some
Schenkerian thinking. Readers may find it instructive to compare our contrast-
ing views, especially of measures 4, 6, 9-11, 14, 31, 35, 42, 53, and 74-82.
Di, The chord at the downbeat of measure 75, though reminiscent of that in the
second half of measure 35, here functions within a two-measure G}b*-” expan-
sion: the bass and inner-voice Ebbs are passing notes, while inner-voice Ch is a
neighboring note.
28 I interpret the Cb [Bk] in the first half of measure 79 as a neighboring note
linking Bb and Bb} [Ag], and not as a chordal root.
29 Compare with how Schubert deploys Ch-E-G-(A#) in the closing four measures
of the Quintet in C Major (D. 956), mvmt. 2.
30 Observe that in measures 13 through 25 tonic’s chromaticized upper-third
chord is asserted as III, whereas in measure 76 its diatonic counterpart is not.
31 The direct succession from IV to V with no 5-6 shift on IV and no cadential ¢
embellishment of V risks parallel fifths (here *? to ®?). Compare with measures
39-42, where Schubert’s descent in the bass (Db>E}) prevents that outcome.
On the other hand, such fifths (twelfths) are rather pronounced in measures
48-49 and its replicates, one reason why I described the passage as “rustic.”

1 — “Auf dem Flusse” from Winterreise (D. 911, no. 7, with David Lewin

“Auf dem Flusse: Image and Background in a Schubert Song” (1982-1983).

The Oster Collection: Papers of Heinrich Schenker, File 22/22, housed at
the Performing Arts Division of the New York Public Library. (Available on
microfilm.)
ee) Compare measures 53 and 54, which extend VI, with measures 11 and 12 of
“Die bése Farbe” from Die schéne Miillerin, which extend II->V. (See 4.14.)
 Notes and references to pages 241-246 299

In both cases a passing chord (absent from the graphs) comes between the 4
unfurling and the normative °embellishment of the dominant V or dominant-
emulating II chord.
I address the relationship between bass A¢ and G# (measures 9 through 11) in
the context of a treatise from Schubert’s era in TAH, pp. 219-223.
The fact that the object of one’s love rejects that proposition does not in itself
quell the love. (That is too rational a notion.) For a certain period a rejection
will make matters worse, because powerful emotions formerly directed con-
structively are still being generated with no receptacle to receive them. There
can be wishful thinking of reconciliation (the setting of stanzas 3 and 4 in E
Major hints at this), but such thoughts often merely stir up the emotions, which
Schubert represents through the ever higher vocal surges in the two statements
of stanza 5.

12 Piano Sonata in B} Major (D. 960), movement 1, with Richard L. Cohn

“As Wonderful as Star Clusters: Instruments for Gazing at Tonality in

Schubert” (1998-1999). Both Cohn and I are challenged by commenta-
tors who claim that Schubert’s writing is “arbitrary” or “irrational” (p. 214).
Whereas my response develops out of traditional notions of tonal relations,
Cohn proposes instead “to position triadic harmonies in relation to neither
a diatonic system nor a tonal center, but rather to other triadic harmonies on
the basis of the number of pitch-classes that they share, and more generally on
the efficiency of the voice leading between them” (p. 214). Michael Siciliano
offers a sympathetic assessment of Cohn’s perspective in his “Neo-Riemannian
Transformations and the Harmony of Franz Schubert” (2002), pp. 79-109.
Cohn continues the development of the ideas introduced in his “Star Clusters”
article in “Uncanny Resemblances: Tonal Signification in the Freudian Age”
(2004).
Cohn cites but does not engage significantly with the work of two Schenkerian
analysts, Felix Salzer and Harald Krebs (p. 220, n.). Salzer’s graphs, published
in Structural Hearing (1962), vol. 2, exx. 337 and 394-400, treat only meas-
ures 321-329 and 1-44, respectively. Krebs’s analysis, in “Third Relation and
Dominant in Late 18th- and Early 19th-Century Music” (1980), figs. II.7 and
11.35, is extensive though not exhaustive. Xavier Hascher’s Schenker-inspired
Schubert, la forme sonate et son évolution (1996) had just been published when
Cohn’s article went to press. Hascher’s exx. 108 and 200 — eleven graphs in
all — offer a detailed reading that diverges from the analysis I present here in
numerous ways.
N. 13, below, offers a justification for this reading.
Cohn does place C-F-A-C and C-E-G-C chords adjacent to one another in his
example 2 (my 12.1). However, his only chord label is F+, and the C-E-G-C
chord is notated using filled-in noteheads placed within parentheses. Thus
300 Notes and references to pages 246-249

clearly he does not interpret the F and A as resolving downward by step to E

and G.
nn Compare the “misspelled” G# here with the “correct” spelling F4 (which would
descend to E, though with detour to F#) in the Symphony in D Major [2.6 and
2.7, m. 52], where the context of II-V matches that of VI-II in the Sonata in Bb
Major. Another “misspelling” occurs in “Gefror’ne Thranen” from Winterreise
(D. 911, no. 3). In the connection of chords on Ab and Db in measures 33
through 36, the D} chord is embellished by a §. Consequently Schubert's dis-
sonant, dominant-emulating expansion of Ab-C-Eb is spelled not as C-Eb-Gb—
Bbb, but instead as C-Eb-Gb-Ab.
While affirming the F+ analysis for measure 74 in his commentary, Cohn
acknowledges the significance of measure 80: “the dominant [F Major],
which arrives in m. 74 and is stabilized in m. 80” (p. 222). In Tonal Pitch
Space (2001, pp. 114-115), Pred Lerdahl likewise omits the linchpin C chord.
He interprets the tonal plan of the exposition as “a pair of major-third
regional motions”: first from tonic Bb Major to F# Minor (measure 48), then
from A Major (measure 58) to F Major (measure 80). My interpretation
of the F# and A regions will be presented in 12.5. In Die Raumvorstellung
in Schuberts Harmonik (1995, pp. 50-97), Balz Trimpy proposes a more
extensive web of third-relationships throughout the movement, to some
extent irreconcilable with my reading. See especially his Beispiel 39, which
posits a Terzbeziehung (third-relationship) between G} Major in measure 20
and D Minor in measure 188, and Beispiel 45, which relates the opening Bb
Major tonic to F# Minor in measure 48 (as does Lerdahl), and the F Major
dominant (measure 80) to C# Minor within the development (measure 118).
These latter relationships will be acknowledged in my analysis: the “lower-
third” chords of the tonic and the dominant both appear in their third
chromatic variants.
It is uncommon for a II chord to appear in § position. Perhaps Schubert was
motivated by a desire to create a linear connection to the dominant root in the
bass (G>G}>F), thereby expanding upon (and defusing the sinister character
of) the Gb>F second of measures 8 and 9. On the other hand, the progression
may be a conscious or subconscious echo of a passage in the same key from
Haydn’s Piano Sonata in E> Major (Hoboken XVI: 49, mvmt. 1, mm. 32-36).
Gb will displace bass G for the “augmented sixth” version of II featured in
measure 34 (also shown in 12.2).
Compare with the non-tonic return of the opening theme in the Trio from the
third movement of the Piano Sonata in D Major [2.21].
The embellishing deployment of a chord such as Bb-D}-E4-Gb is catalogued in
most standard harmony textbooks as a “common-tone augmented sixth chord”
or “embellishing augmented sixth chord.” Compare this usage with that in “Die
junge Nonne,” measures 69 through 72 (chapter 9), and with “Thr Bild” from
Schwanengesang (D. 957/9), measures 23 and 24.
 Notes and references to page 250 301

10 Compare 2.23, Model 2, and 12.4. In both cases an inner-voice line descends
chromatically from 8 to 67 (G>FH>F4 vs. Bb>A>Ab), and in both cases the
middle pitch of that descent is supported by tonic’s upper-third chord. In the
Sonata in D Major, 57 arrives above the tonic root, and thus the chord takes
on a dominant-emulating character, pushing towards IV. In the Sonata in Bb
Major, 7 arrives above an active bass (Bk) that instigates a shift to tonic’s 6
phase (that is, B)-D-F to Bb-D-G, the latter represented by By}-D-F-A}). In the
Sonata in D Major, the connection between tonic and its upper-third chord is
accomplished via a descending circle of fifths (G-C-F#-B, displayed in 2.24).
In the Sonata in Bb Major, that connection is accomplished via a descending
circle of major thirds (B)-G}/F#-D, displayed in 12.5).
ib — I acknowledge that my matter-of-fact acceptance of the arrival point in
measure 68 as tonic’s upper-third chord (not fretting over its potential inter-
pretation as an augmented fifth below tonic, with root Ebb) may disappoint
Cohn and others, who are intrigued by such conflicts between what seem
to be local harmonic initiatives and the broader trajectory of the exposition
(clearly Bb to F, not to Gbb, which would result in the absence of Schubert’s
enharmonic respelling: see Cohn’s n. 23 on p. 220). I do not sense that Schubert
finds any conflict between the chord of measure 73 (which is another point of
contention between Cohn and me, as noted above) and a conventional tonic
6 phase, here evolved from diatonic G-Bb-D into dominant-emulating B4-D-
F-Ab. Consequently I interpret the progression by descending “thirds” here as
occurring within a modulo 12 tonal space in which the diatonic framework of
Bb Major, instilled into the listener’s consciousness at the outset, is temporar-
ily suspended. I make bold distinctions between harmonic and non-harmonic
chordal progressions. Circular, parallel, and sequential progressions may be
pursued either heeding or not heeding the diatonic constraints of the prevail-
ing key. Either way, only the endpoints of such trajectories — not the various
internal chords - play a role within the deeper structure.
12 The transition is considerably expanded from a draft version, a facsimile of
which has been published in Drei grofve Sonaten fiir das Pianoforte: D 958, D
959, D 960 (Friihe Fassungen), ed. Ernst Hilmar (Tutzing: Hans Schneider,
1987). The original conception, which proceeded without many misgivings,
corresponds to measures 48-51, 52/53 (the content appearing within a single
measure), 54-56, 58, 59 (67), 68-71, 73-75, and 78 leading to the F Major tonic
of measure 80, from which point the draft and final version correspond more
closely. Some analysts interpret the F# Minor passage as the middle key within
a three-key exposition. For example, see James Webster’s “Schubert's Sonata
Forms and Brahms’s First Maturity” (1978-1980), p. 22, and Deborah Kessler’s
“Schubert’s Late Three-Key Expositions” (1996), chapter 5. I instead regard F#
Minor as a step along a broader path: Bb>Gb/F#>D [12.5]. The fact that F# is
tonicized but D is not does not mean that F# must be construed as hierarchi-
cally deeper. Instead the D chord is a representative of the tonic (namely, its
302 Notes and references to pages 251-252

upper-third chord), reviving the I from the opening measures just prior to the
onset of its 6 phase (Bb-D-G, represented by Bk-D-F-Ab in measure 73).
Schubert’s writing during this passage is extraordinarily subtle. At measure
59 we are uncertain whether bass Bb functions as a neighbor (thus confirming
root A), or whether instead the A at the end of measure 59 is a passing note
between Bb and G (thus accomplishing a shift to F#?). Gy-C#-E at the downbeat
of measure 60 could represent either Gk-A-C#-E or Gh-A#-C#-E. Only at beat 3
of measure 60, with soprano Ad, is the latter confirmed. The same gambit plays
out again in measures 63 and 64. Only at measure 67 does G4-C#-E turn out to
represent Gk-A-C#-E. Exactly this issue is at stake again in measure 73, where
the bass descends F>Ek>D. As stated above, I view the Ey as a passing note
between F and D, and thus my interpretation of measures 73 into 74 is GC,
with the resolution from Ab [G#] (G’s ninth) into G (C’s fifth) delayed until
measure 78. Cohn hears a resolution of F into E during measure 73 and thus
interprets the harmonic succession as EA. (See 12.1.) The crux of the matter
is whether D-G#-B at the end of measure 73 stands for D-F-G#-B (a misspelling
of G®) or for D-Es-G#-B.
14 Compare with the similarly provocative ascent of two consecutive minor thirds
in “Die Einsiedelei” [4.26]. Schubert ascends from G to Bb (not daring), and
then from Bb to D} (very daring). Upon attaining D}, he quickly scurries back
to Bb, and the progression continues as if Db had never sounded.
LS Both my 12.2 (Chromatic Variant #3) and 12.5 propose that despite the local
dominant chord of measures 42 through 44, the broader connection is between
Bb-D-F and F-Ab-Cb-Ebb [F-Ab-B}-D], the latter of which settles into D}-F-
Ab-Cb in measure 47. (The situation is not easily comprehended from the
score because Schubert employs spellings of convenience and allegedly passes
through the “enharmonic seam” - Cohn’s term - in measure 46. Cohn men-
tions having consulted the facsimile that I cite in n. 12, above (p. 219). In that
version Schubert returns to the B} tonic before descending to Gb/F#, deploying
the bass that appears in measures 19 and 20 of the published version (expanded
with an extra measure of Bb/C} trill before the chromatic descent). In the final
version the tonic resolution of the local dominant of measures 42 through 44
has been elided. By positioning the D}’ chord in 4position, Schubert retains the
gist of his initial B}>Ab>Gb descent.
16 Cohn suggests that his example 2 might seem “Schenkerian in its general
features,” though uncharacteristically displaying inattentiveness to register (I
sometimes normalize registers as well, as in 12.4) and an “insufficiently inter-
preted” “prolongational structure” (p. 220). I would add: insufficient atten-
tion to hierarchical levels. A Schenkerian graph would not display the D)+/
C#+ chord of measure 47 and the A+ chord of measure 58 in such divergent
manners.
7) Cohn justifiably criticizes analysts who would concoct a Roman-numeral label
for the G}/F# chord. (He proposes bvib, though such analysts probably would
 Notes and references to pages 253-255 303

use >VI> or >vi instead.) He suggests that “a diatonic characterization . . . pays

the price of revealing the descriptive clumsiness of a diatonic model of this
chord,” a sign of that methodology’s “explanatory inadequacy” (p. 218). In
Cohn’s system F— is By+’s “polar key,” maintaining tonic functionality. (See
his fig. 4.) In mine it is the “Chromatic Variant #3” of tonic’s 6-phase chord,
which may be used within a prolongation of the 5-phase chord (as in 2.18) or,
as here, may lead away from it, in this case as an interior element of a circular
progression. The interior region of such a progression is a prime location for
a chromatic efflorescence, because the construct is linear in nature, not har-
monic. As mentioned above, G—, G+, G+, and G)- are assigned variously to a
Subdominant or a Tonic role in Cohn’s system. In my system they are progres-
sively more extreme realizations of a single phenomenon that in some contexts
(that is, when not asserted) may warrant no harmonic interpretation. I propose
that one should allow the linear progression to run its course, assessing where
within the tonal landscape the progression has led at its conclusion. Here that
goal is diatonic: tonic’s upper-third chord. Such trajectories often look more
bizarre than they actually are, because music notation is biased in favor of the
seven diatonic pitch classes of a key. (Any key can be favored, but every key will
have seven favored elements.) By>G>/F#>D in Bp Major is disconcerting. Were
one to employ a modulo 12 numerical notation system (with C = 0), the expan-
sion of 10>2 into 10>6>2 would seem routine, barely worthy of comment. I
contend that Schubert has entered that twelve-element realm but has been
compelled to display his conception using the conventional seven-note nota-
tional system. In my view all the fuss over the “enharmonic seam” stems from
this notational inelegance and from a dogged persistence of musicians to
process linear chordal progressions as if they were harmonically conceived.
18 Schubert’s F# spelling and the reiterated soprano E>>D clearly favor a reading
of the chord as tonic’s dominant-emulating 6 phase - (D)-F#-A-C-Eb. The
resolution to a Bp major chord in measure 85 encourages one to reevaluate the
situation, and to decide retrospectively that Fé in fact represents Gp and that
soprano E} did not resolve to D in measure 84 after all. (One could interpret
that D as an anticipation of the D that arrives in measure 85.)
19 Perhaps Cohn has highlighted Dp Major and A Minor in anticipation to future
passages within the movement: he imputes a relationship with Dp Minor and A
Major regions that occur during the development section (pp. 225-226).
20 I regard the tonal elaboration of both the exposition’s By major harmony and
the development’s F major harmony to engage the full range of their 6-phase
chords (or, in the case of g, a surrogate): g, G, Gb, g> [ff] and (in reverse order)
db {c#], Db, D, d, respectively. Though his fig. 2 inhibits such a broad correlation
(unlike my less imposing 2.17), Cohn articulates a part of this perception in
his statement: “the framework adopted here encourages us to see C# minor as
standing in the same relation to the dominant as F# minor stands to the tonic”
(p. 218).
304 Notes and references to pages 256-259

21 Compare with measures 176 through 192 of the Symphony in B Minor, mvmt.
1, discussed above in chapter 7 (pp. 170-171).
Mp) The second third, Bbb>Gb, is traversed through its own subordinate descending-
fifths motion: Bb)-Eb—>Ab-Db—G} in measures 131 through 139.
Dp) In a reading that I acknowledge to be pure speculation, measures 1 through 7
might represent Schubert’s virginal state of innocence, with trilled bass G} in
measure 8 the first rumbling of his sexual urges. Schubert initially sublimates
those urges (G>G}>F in measures 15 and 16), but soon he launches into an inti-
mate encounter with gusto (the acceptance of G} as a tonicized root in measure
20). That episode gradually intensifies, leading to an (orgasmic?) climax at the
forte in measure 35, coordinating with an exultant restatement of the opening
theme. The ensuing encounter with a less salutary character (G} [F#] Minor)
beginning in measure 48 marks a pivotal moment both in the movement and in
Schubert’s personal life. Though he hopes he has escaped the danger by finding
his way to the dominant key, F Major, Gb rises to the surface again during the
exposition’s first ending and, worse, the feared life-threatening disease strikes
at the beginning of the development. Etc.
Alternatively, one might rely upon Donald Francis Tovey’s description: “The
whole movement . . . runs a course not unusual in Schubert’s large designs;
opening with a sublime theme of the utmost calmness and breadth; descend-
ing, by means of a good though abrupt dramatic stroke, from the sublime to
the picturesque, and then drifting from the picturesque through prettiness to
a garrulous frivolity. But then comes meditation.” Etc. The trilled G} repre-
sents “distant thunder.” Tovey’s commentary appears in his “Franz Schubert”
(1927), pp. 103-104.
For a narrative perspective less decorous than Tovey’s — and less provoca-
tive than mine - see Byron Almén’s A Theory of Musical Narrative (2008), pp.
139-161.
24 Schubert effectively realizes a mehrdeutig shift as he explores potential continu-
ations during this region: Bb-E4-G corresponds to the fifth, seventh, and ninth
of a chord rooted on A in m. 190; whereas in m. 202 these pitches represent the
third, fifth, and seventh of a chord rooted on C. A>D extends the D Minor
tonicization, whereas CF is the pathway to the long-delayed restoration of
the F dominant (m. 203).
aS Cohn also confirms the content of my 12.9: he describes that passage as “a
chromatic circle of fifths so distended that it cycles back to its origin and over-
shoots it by two stations” (p. 226). Though I regard the “overshooting” to be a
consequence of starting the circle on D}’s upper-third chord and he regards it
to be an event related to the circle’s closure, we are in agreement concerning
what transpires between those points.
26 James M. Baker describes D Minor as “a ‘correction’ of Db, to Db,, a substitu-
tion of diatonic for chromatic scale-step 3.” See his “Skirting the Structural
Tonic: A Romantic Impulse in Schubert’s Later Works” (2007), pp. 233-234.
 Notes and references to pages 259-264 305

Zi Cohn adds that the “tentative character” of the retransitional dominant “calls
into question its ‘true’ role as dominant” (p. 227). (In that I do not concur with
his reading of the broad harmonic progression within the exposition, I cannot
endorse his further comments concerning symmetry on pp. 226 and 227.)
28 Nicholas Marston offers a similar description in “Schubert’s Homecoming”
(2000): “By the end of its 13-bar span, the dominant of bars 203-15 seems to
be looking not so much homeward as straight ahead into a void. Rather than
shoring up and affirming whatever force the initial dominant might possess,
this music sounds more like a leaching away of the limited power to reach
home possessed by that particular, attenuated harmony in the first place.
Epiphany is not to be encountered here” (p. 252). “In the silence which consti-
tutes most of bar 215 can be heard an immense effort of will, a husbanding of
largely spent resources in the face of the awful need to begin again. But only by
beginning again will epiphany be granted” (p. 255).
29 Here Cohn and I concur, for in his fig. 4 both F+ and A+ reside within his
“Dominant” region (along with F-, Db+, C#-, and A-). See the arrow (with
inadvertently missing tip) labeled 2c in his fig. 9.
30 C arrives in measure 338 and is prolonged through measure 339. The preceding
D is covered by an ascending arpeggiation (D<F<Ab) leading to the Bb chord’s
minor seventh, which arrives at the top of the texture during tonic’s 6 phase
and resolves to G. A descending arpeggiation (G>Eb>C) leads to the principal
line’s C, which is prolonged by a descent to inner-voice A (measure 339) before
resolution to Bb.
31 Cohn classifies the G+ and G- chords (my 6 phase of tonic in Bb Major) as
“Subdominant” chords. That is hard to reconcile with their placement in a
hierarchically dependent position between the tonic and the supertonic in
12.1, 12.2, 12.3, 12.4, and 12.11, or the placement of equivalent D+ and D-
chords in tonicized F Major in 12.10. Using my tools, these chords are natural
outgrowths of tonic. They do not register in any of Cohn’s models of harmonic
motion. II fares equally poorly: C+ and C- appear in a region on Cohn’s matri-
ces for which no function name is provided. Cohn enters that region only once
(for an E+ chord in his fig. 5).

Epilogue
1 “Schubert sagte einst zu mir: Die Leute sagen, daB meine Komposizionen nicht
rein von Fehlern der Harmonie etc. sind. Wenn man mir nur sagen méchte wo
die Fehler sind!” From a reminiscence of 1857, transcribed in Rita Steblin and
Frederick Stocken’s “Studying with Sechter: Newly Discovered Reminiscences
about Schubert by His Forgotten Friend, the Composer Joseph Lanz,” Music
and Letters 88 (2007), p. 262.
List of references to music examples

1.1 8, 9, 12, 31, 85, 247, 268 Pel, 276

1.5 TDs MB» XS; 0), SD, ISP, BOX); 2.21 181, 196, 247, 278, 300
298 Papa) DT, PX, AI
1.6 269 2.24 281, 301
ow 269 3.1 Wy 22MOs:
1.8 WD, sk, AS) DUO AMS XS? Se Aes OURS:
1.9 275 3.19 280, 289
1.10 272, 3.20 278
1.11 PRs DMDYY) 3.26 293
1.12 18 3.28 YD; Mev), MDE
1.13 270, 273, 274 3:29 C2)|
1.14 IQ, BOS 3.31 88
1.15 205270; 273 3.32 Qe,
1.18 273 3.36 281
1.19 Key PM 3.37 280
1.24 278 4.1 oe)
1.27 73 4.3 96
1.30 270 4.5 100
1.34 ote),
278 AIS 4.8 99, 284
1.36 82 4.14 298
2.1 44 4.18 269
2.6 300 4.20 108
Mell 51, 300 4.24 294
MS) DOA DIAS Dy 219) 4.26 220, 302
2.10 254, 293 4.27 230
Za 198, 275 4.34 238
2.12 DES HOSP
Ds D3} 4.38 234
2.14 81 Sal 287
2.16 71-72, 276 5.3 287
2.17 UES 8.7 293
2.18 276 11.2 269
Select bibliography

Agawu, V. K., “Schubert's Harmony Revisited: The Songs ‘Du liebst mich nicht’
and “Dass Sie hier gewesen?’ Journal of Musicological Research 9 (1989-90),
pp. 23-42
“Theory and Practice in the Analysis of the Nineteenth-Century Lied” Music
Analysis 11 (1992), pp. 3-36
“Perspectives on Schubert's Songs,” Music Analysis 16 (1997), pp. 107-122
Music as Discourse: Semiotic Adventures in Romantic Music, Oxford University
Press, 2009
Agmon, E., “Music and Text in Schubert Songs: The Role of Enharmonic
Equivalence,” Israel Studies in Musicology 4 (1987), pp. 49-58
Almén, B., A Theory of Musical Narrative, Bloomington: Indiana University Press,
2008
Baker, J., “Skirting the Structural Tonic: A Romantic Impulse in Schubert’s Later
Works,’ in Le style instrumental de Schubert: Sources, analyse, évolution, ed. X.
Hascher, Paris: Publications de la Sorbonne, 2007, pp. 215-248
Bass, R., “From Gretchen to Tristan: The Changing Role of Harmonic Sequences
in the Nineteenth Century, 19th-Century Music 19 (1995-1996), pp.
263-285
Beach, D., “Schubert's Experiments with Sonata Form: Formal-Tonal Design versus
Underlying Structure,” Music Theory Spectrum 15 (1993), pp. 1-18
“Harmony and Linear Progression in Schubert's Music,” Journal of Music Theory
38 (1994), pp. 1-20
“An Analysis of Schubert’s ‘Der Neugierige’: A Tribute to Greta Kraus,” Canadian
University Music Review 19 (1998), pp. 69-80
“Modal Mixture and Schubert’s Harmonic Practice,’ Journal of Music Theory 42
(1998), pp. 73-100
“Schubert and Equal Division of the Octave: A Study of the First Movement
of the G-Major Quartet D887,” in Le style instrumental de Schubert: Sources,
analyse, évolution, ed. X. Hascher, Paris: Publications de la Sorbonne, 2007, pp.
249-260
Blume, J., “Analysen als Beispiele musiktheoretischer Probleme: Auf der Suche nach
der angemessenen Beschreibung chromatischer Harmonik in romantischer
Musik, Musiktheorie 4 (1989), pp. 37-51
Bockholdt, R., “Die Kunst, heim zu finden: Uber Schliisse und Anschliisse in
Schuberts Instrumentalmusik,” Musiktheorie 13 (1998), pp. 145-156
308 Select bibliography

Bretherton, D., “In Search of Schubert’s Doppelganger,’ Musical Times 144/1884

(2003), pp. 45-50
“The Poetics of Schubert's Song-Forms,’ Ph.D. thesis, University of Oxford,
2007
Bribitzer-Stull, M., “The Ab - C - E Complex: The Origin and Function of
Chromatic Third Collections in Nineteenth-Century Music,” Music Theory
Spectrum 28 (2006), pp. 167-190
Brower, C., “A Cognitive Theory of Musical Meaning,” Journal of Music Theory 44
(2000), pp. 323-379
Brown, A. P., The Symphonic Repertoire, vol. 2: The First Golden Age of the Viennese
Symphony: Haydn, Mozart, Beethoven, and Schubert, Bloomington: Indiana
University Press, 2002
Brown, M., “The Diatonic and the Chromatic in Schenker’s Theory of Harmonic
Relations, Journal of Music Theory 30 (1986), pp. 1-34
Burnham, S., “Schubert and the Sound of Memory,’ Musical Quarterly 84 (2000),
pp. 655-663
Burstein, L. P., “Lyricism, Structure, and Gender in Schubert's G Major String
Quartet,’ Musical Quarterly 81 (1997), pp. 51-63
“Surprising Returns: The VII# in Beethoven's Op. 18 No. 3, and Its Antecedents
in Haydn,” Music Analysis 17 (1998), pp. 295-312
“Devil’s Castles and Schubert’s Strange Tonic Allusions,’ Theory and Practice 27
(2002), pp. 69-84
“Unraveling Schenker’s Concept of the Auxiliary Cadence, Music Theory
Spectrum 27 (2005), pp. 159-186
Clark, S., “From Nature to Logic in Schubert’s Instrumental Music,” PhD diss.,
Princeton University, 1997
“Terzverwandtschaft in der “Unvollendeten’ von Schubert und der ‘Waldstein-
Sonate’ von Beethoven,’ trans. T. Aigner, Schubert durch die Brille 20 (1998),
pp. 123-130; response by H. Federhofer (as “Terz- und Quintverwandtschaft
in der Harmonik Schuberts: Eine Replik”) in Schubert durch die Brille 20
(1998), pp. 157-161
“Schubert, Theory and Analysis,’ Music Analysis 21 (2002), pp. 209-243
Code, D. L., “Listening for Schubert’s Doppelgangers,” Music Theory Online 1/4
(July 1995)
Cohn, R., “Maximally Smooth Cycles, Hexatonic Systems, and the Analysis of Late
Romantic Triadic Progressions,” Music Analysis 15 (1996), pp. 9-40
“As Wonderful as Star Clusters: Instruments for Gazing at Tonality in Schubert,
19th-Century Music 22 (1998-1999), pp. 213-232; response by C. Fisk and
reply by R. Cohn in 19th-Century Music 23 (1999-2000), pp. 301-304
“Uncanny Resemblances: Tonal Signification in the Freudian Age? Journal of the
American Musicological Society 57 (2004), pp. 285-324
Cone, E. T., “Schubert’s Promissory Note: An Exercise in Musical Hermeneutics”
19th-Century Music 5 (1981-1982), pp. 233-41; reprinted with revisions in
 Select bibliography 309

Schubert: Critical and Analytical Studies, ed. W. Frisch, Lincoln: University of

Nebraska Press, 1986, pp. 13-30
Cube, F.-E. von, The Book of the Musical Artwork: An Interpretation of the Musical
Theory of Heinrich Schenker, trans. D. Neumeyer, G. R. Boyd, and S. Harris,
Lewiston, N.Y.: Edwin Mellon Press, 1988
Damschroder, D., “Liszt's Composition Lessons from Beethoven (Florence, 1838-
39): ‘Il penseroso ’» Journal ofthe American Liszt Society 28 (1990), pp. 3-19
“Schubert, Chromaticism, and the Ascending 5-6 Sequence,’ Journal of Music
Theory 50 (2006), pp. 253-275
Thinking About Harmony: Historical Perspectives on Analysis, Cambridge
University Press, 2008
“Schenker, Schubert, and the Subtonic Chord, in A Music-Theoretical Matrix:
Essays in Honor ofAllen Forte (Part II), ed. David Carson Berry, Gamut 3/1
(2010)
Denny, T. A., “Directional Tonality in Schubert’s Lieder? in Franz Schubert — Der
Fortschrittliche? Analysen - Perspektiven - Fakten, ed. E. W. Partsch, Tutzing:
Hans Schneider, 1989, pp. 37-53
Dineen, M., “Tonal Problem, Carpenter Narrative, and Carpenter Motive in
Schubert's Impromptu, Op. 90, No. 3,” Theory and Practice 30 (2005), pp.
97-120
Dittrich, M. A., “Der ‘Ohren zerreissende Fortschrit?: Zur neuen Harmonik in
Schuberts Liedern,” in Franz Schubert - Der Fortschrittliche? Analysen -
Perspektiven - Fakten, ed. E. W. Partsch, Tutzing: Hans Schneider, 1989, pp.
55-72
Harmonik und Sprachvertonung in Schuberts Liedern, Hamburg: Karl Dieter
Wagner, 1991
Drabkin, W., “Schubert, Schenker and the Art of Setting German Poetry,’ Eighteenth-
Century Music 5 (2008), pp. 209-236
Dunsby, J., Making Words Sing: Nineteenth- and Twentieth-Century Song, Cambridge
University Press, 2004
Diirr, W,, “‘Ausweichungen ohne Sinn, Ordnung und Zweck’: Zur Tonart und Tonalitat
bei Schubert,” in Franz Schubert - Der Fortschrittliche? Analysen — Perspektiven —
Fakten, ed. E. W. Partsch, Tutzing: Hans Schneider, 1989, pp. 73-103
Everett, W. T., “A Schenkerian View of Text-Painting in Schubert's Song Cycle
‘Winterreise, PhD diss., University of Michigan, 1988
“Grief in Winterreise: A Schenkerian Perspective,” Music Analysis 9 (1990), pp.
157-175
Federhofer, H., “Terzverwandte Akkorde und ihre Funktion in der Harmonik
Franz Schuberts,” in Schubert-Kongref Wien 1978: Bericht, ed. O. Brusatti,
Graz: Akademische Druck- und Verlagsanstalt, 1979, pp. 61-70
Fieldman, H., “Schubert’s Sonata Form: The Case of the C-Minor Quartettsatz,’ in
Le style instrumental de Schubert: Sources, analyse, évolution, ed. X. Hascher,
Paris: Publications de la Sorbonne, 2007, pp. 83-92
310 Select bibliography

Fisk, C., “What Schubert’s Last Sonata Might Hold,” in Music and Meaning, ed. J.
Robinson, Ithaca, N.Y.: Cornell University Press, 1997, pp. 179-200
Returning Cycles: Contexts for the Interpretation of Schubert's Impromptus and
Last Sonatas, Berkeley: University of California Press, 2001
Gerhard, A., “Entfernte Harmonien und fehlende Gegenwart: Anmerkungen zu
Franz Schuberts Goethe-Lied ‘An die Entfernte’ von 1822,” Musiktheorie
13 (1998), pp. 123-30; response by C. Kithn and reply by A. Gerhard in
Musiktheorie 13 (1998), pp. 263-265
Girton, I., “Promises Fulfilled: Neighbour Notes in Some Late Works of Schubert,”
in Le style instrumental de Schubert: Sources, analyse, évolution, ed. X. Hascher,
Paris: Publications de la Sorbonne, 2007, pp. 93-100
Graubart, M., “The Frustrated Supertonic and the Pathetic Added Sixth,’
International Journal of Musicology 8 (1999), pp. 451-461
Haas, B., Die neue Tonalitat von Schubert bis Webern: Horen und Analysieren nach
Albert Simon, Wilhelmshaven: Florian Noetzel, 2004
Hascher, X., Schubert, la forme sonate et son évolution, Bern: Peter Lang, 1996
“Eine ‘traumhafte barcarola funebre: Fragmente zu einer Deutung des langsamen
Satzes des Streichquintetts D 956; in Schubert und das Biedermeier: Beitrage
zur Musik des friihen 19. Jahrhunderts: Festschrift fiir Walther Diirr zum 70.
Geburtstag, ed. M. Kube, W. Aderhold, and W. Litschauer, Kassel: Barenreiter,
2002, pp. 127-138
Symbole et fantasme dans Adagio du Quintette a cordes de Schubert, Paris:
LHarmattan, 2005
“Sur les pas du ‘Wanderer’: Pour une cartographie de lerrance schubertienne,” in
Le style instrumental de Schubert: Sources, analyse, évolution, ed. X. Hascher,
Paris: Publications de la Sorbonne, 2007, pp. 181-204
In dunklen Traéumen’: Schubert's Heine-Lieder through the Psychoanalytical
«ec

Prism,’ Nineteenth-Century Music Review 5 (2008), pp. 43-70

Hatten, R. S., Musical Meaning in Beethoven: Markedness, Correlation, and
Interpretation, Bloomington: Indiana University Press, 1994
Interpreting Musical Gestures, Topics, and Tropes: Mozart, Beethoven, Schubert,
Bloomington: Indiana University Press, 2004
Head, M., “Schubert, Kramer, and Musical Meaning,” Music and Letters 83 (2002),
pp. 426-440
Hepokoski, J., and Darcy, W., Elements of Sonata Theory: Norms, Types, and Defor-
mations in the Late-Eighteenth-Century Sonata, Oxford University Press, 2006
Hughes, M., Moss, L., and Schachter, C., “Analysis Symposium: Moments musicals,
Op. 94, Franz Schubert,” Journal of Music Theory 12 (1968), pp. 184-239
Jackson, T. L., “Schubert’s Revisions of Der Jiingling und der Tod, D 545a-b, and
Meeresstille, D 216a-b,” Musical Quarterly 75 (1991), pp. 336-361
“Hinauf strebt’s: Song Study with Carl Schachter” in Structure and Meaning in
Tonal Music: Festschrift in Honor of Carl Schachter, ed. L. P. Burstein and
D. Gagné, Hillsdale, N.Y.: Pendragon Press, 2006, pp. 191-202
 Select bibliography 311

Kerman, J., “A Romantic Detail in Schubert’s Schwanengesang” Musical Quarterly

48 (1962), pp. 36-49; reprinted with revisions in Schubert: Critical and
Analytical Studies, ed. W. Frisch, Lincoln: University of Nebraska Press, 1986,
pp. 48-64
Kessler, D., “Schubert's Late Three-Key Expositions: Influence, Design, and
Structure,’ PhD diss., City University of New York, 1996
“Motive and Motivation in Schubert's Three-Key Expositions” in Structure
and Meaning in Tonal Music: Festschrift in Honor of Carl Schachter, ed. L. P.
Burstein and D. Gagné, Hillsdale, N.Y.: Pendragon Press, 2006, pp. 259-276
Kinderman, W., “Schubert's Tragic Perspective? in Schubert: Critical and
Analytical Studies, ed. W. Frisch, Lincoln: University of Nebraska Press,
1986, pp. 65-83
Kopp, D., Chromatic Transformations in Nineteenth-Century Music, Cambridge
University Press, 2002
Kramer, L., “The Mirror of Tonality: Transitional Features of Nineteenth-Century
Harmony, 19th-Century Music 4 (1980-1981), pp. 191-208
“The Schubert Lied: Romantic Form and Romantic Consciousness, in Schubert:
Critical and Analytical Studies, ed. W. Frisch, Lincoln: University of Nebraska
Press, 1986, pp. 200-236
“Decadence and Desire: The Wilhelm Meister Songs of Wolf and Schubert; 19th-
Century Music 10 (1986-1987), pp. 229-242
“Performance and Social Meaning in the Lied: Schubert's Erster Verlust? Current
Musicology 56 (1994), pp. 5-23
Franz Schubert: Sexuality, Subjectivity, Song, Cambridge University Press,
1998
Kramer, R. A., Distant Cycles: Schubert and the Conceiving of Song, University of
Chicago Press, 1994
Unfinished Music, Oxford University Press, 2008
Krautwurst, F., “Zu Schuberts Harmonik,;’ in Franz Schubert - Der Fortschrittliche?
Analysen — Perspektiven - Fakten, ed. E. W. Partsch, Tutzing: Hans Schneider,
1989, pp. 171-183
Krebs, H. M., “Third Relation and Dominant in Late 18th- and Early 19th-Century
Music; PhD diss., Yale University, 1980
“Alternatives to Monotonality in Early Nineteenth-Century Music,’ Journal of
Music Theory 25 (1981), pp. 1-16
“The Background Level in Some Tonally Deviating Works of Franz Schubert,” In
Theory Only 8/8 (1985), pp. 5-18
“Techniques of Unification in Tonally Deviating Works,’ Canadian University
Music Review 10 (1990), pp. 55-70
“Tonart und Text in Schuberts Liedern mit abweichenden Schliissen,’ Archiv fiir
Musikwissenschaft 47 (1990), pp. 264-271
“Some Early Examples of Tonal Pairings: Schubert's ‘Meeres Stille’ and “Der
Wanderer,” in The Second Practice of Nineteenth-Century Tonality, ed. W.
oly Select bibliography

Kinderman and H. Krebs, Lincoln: University of Nebraska Press, 1996, pp.

7/333).
“Wandern und Heimkehr: Zentrifugale und zentripetale Tendenzen in Schuberts
friihen Liedern? Musiktheorie 13 (1998), pp. 111-122
Krims, A., “Music Theory as Productivity,’ Canadian University Music Review
20 (2000), pp. 16-30; response by W. J. Renwick and reply by A. Krims in
Canadian University Music Review 20 (2000), pp. 31-44
Kurth, E., Romantische Harmonik und ihre Krise in Wagner's ‘Tristan’, 2nd edn,
Berlin: M. Hesse, 1920
Kurth, R., “Music and Poetry, A Wilderness of Doubles: Heine —- Nietzsche -
Schubert - Derrida? 19th-Century Music 21 (1997-1998), pp. 3-37
Laitz, S. G., “Pitch-Class Motive in the Songs of Franz Schubert: The Submediant
Complex,’ PhD diss., University of Rochester, 1992
“The Submediant C: Its Musical and Poetic Roles in Schubert's Songs,’ Theory
and Practice 21 (1996), pp. 123-165
Lambert, J. S., “Franz Schubert and the Sea of Eternity,’ Journal of Musicology 21
(2004), pp. 241-266
Leblanc, M., “Franz Schubert: Un pas vers latonalité,” Canadian University Music
Review 9 (1989), pp. 84-115
Lerdahl, E, Tonal Pitch Space, Oxford University Press, 2001
Levenson, I. M., “Motivic-Harmonic Transfer in the Late Works of Schubert:
Chromaticism in Large and Small Spans,” PhD diss., Yale University, 1981
“Smooth Moves: Schubert and Theories of Modulation in the Nineteenth
Century,’ In Theory Only 7/5-6 (1984), pp. 35-52
Lewin, D., “Auf dem Flusse: Image and Background in a Schubert Song,’ 19th-
Century Music 6 (1982-1983), pp. 47-59; reprinted in Schubert: Critical and
Analytical Studies, ed. W. Frisch, Lincoln: University of Nebraska Press, 1986,
pp. 126-152; reprinted in D. Lewin, Studies in Music with Text, New York:
Oxford University Press, 2006, pp. 109-134
“Music Theory, Phenomenology, and Modes of Perception,’ Music Perception 3
(1985-1986), pp. 327-392
McClary, S., “Pitches, Expression, Ideology: An Exercise in Mediation, Enclitic 7
(1983), pp. 76-86
“Constructions of Subjectivity in Schubert’s Music,’ in Queering the Pitch: The
New Gay and Lesbian Musicology, ed. P. Brett, E. Wood, and G. C. Thomas,
New York: Routledge, 1994, pp. 205-233
McCreless, P. P., “Schenker and Chromatic Tonicization: A Reappraisal} in Schenker
Studies, ed. H. Siegel, New York: Cambridge University Press, 1990, pp.
125-145
“An Evolutionary Perspective on Nineteenth-Century Semitonal Relations” in
The Second Practice of Nineteenth-Century Tonality, ed. W. Kinderman and H.
Krebs, Lincoln: University of Nebraska Press, 1996, pp. 87-113
Mak, S., “Schubert's Allusions to the Descending Tetrachord? in Le style instrumental
 Select bibliography a3}

de Schubert: Sources, analyse, évolution, ed. X. Hascher, Paris: Publications de

la Sorbonne, 2007, pp. 163-179
Marston, N., “Schubert's Homecoming; Journal of the Royal Musical Association
125 (2000), pp. 248-270
Messing, S., Schubert in the European Imagination, 2 vols., University of Rochester
Press, 2006-2007
Monelle, R., The Sense of Music: Semiotic Essays, Princeton University Press, 2000
Morgan, R. P., “Dissonant Prolongation: Theoretical and Compositional Precedents)”
Journal of Music Theory 20 (1976), pp. 49-91
Muxfeldt, K., “Schubert, Platen, and the Myth of Narcissus,’ Journal of the American
Musicological Society 49 (1996), pp. 480-527; response by V. S. Ramalingam in
JAMS 50 (1997), pp. 530-536
Newbould, B., “A Schubert Palindrome,’ 19th-Century Music 15 (1991-1992), pp.
207-214
Schubert and the Symphony, Surbiton: Toccata Press, 1992
“Cornered in the Middle Eight: Dance Miniaturism vis-d-vis Sonata, in Schubert
the Progressive: History, Performance Practice, Analysis, ed. B. Newbould,
Aldershot: Ashgate, 2003, pp. 107-116
Nolan, C., “Combinatorial Space in Nineteenth- and Early Twentieth-Century
Music Theory,” Music Theory Spectrum 25 (2003), pp. 205-241
Perry, J., “Ihe Wanderer’s Many Returns: Schubert’s Variations Reconsidered,’
Journal of Musicology 19 (2002), pp. 374-416
“Schubert's ‘Aus Goethe's Faust; D. 126: The Scena as Fragment,” Indiana Theory
Review 26 (2005), pp. 105-121
Proctor, G., “Technical Bases of Nineteenth-Century Chromatic Tonality: A Study
in Chromaticism,” PhD diss., Princeton University, 1978
Raab, M., “Phrygische Wendung und Mediantik: ‘Der Wanderer’ und die ‘Fantasie’,”
Musiktheorie 13 (1998), pp. 131-143
Rast, N., “‘Schéne Welt, wo bist du?’: Motive and Form in Schubert's A Minor
String Quartet? in Schubert the Progressive: History, Performance Practice,
Analysis, ed. B. Newbould, Aldershot: Ashgate, 2003, pp. 81-88
Rings, S., “Tonality and Transformation,” PhD diss., Yale University, 2006
“Perspectives on Tonality and Transformation in Schubert’s Impromptu in Eb,
D. 899, no. 2? Journal of Schenkerian Studies 2 (2007), pp. 33-63
Rosen, C., Sonata Forms, New York: W. W. Norton, 1980
The Frontiers of Meaning: Three Informal Lectures on Music, New York: Hill and
Wang, 1994
The Romantic Generation, Cambridge, Mass.: Harvard University Press, 1995
“Schubert and the Example of Mozart,’ in Schubert the Progressive: History,
Performance Practice, Analysis, ed. B. Newbould, Aldershot: Ashgate, 2003, pp.
1-20
Rothstein, W., “Common-Tone Tonality in Romantic Opera: An Introduction,’
Music Theory Online 14/1 (March 2008)
314 Select bibliography

Salzer, E, “Die Sonatenform bei Franz Schubert,” Studien zur Musikwissenschaft 15

(1928), pp. 86-125
Structural Hearing: Tonal Coherence in Music, 2 vols., New York: Boni, 1952;
reprint edn, New York: Dover, 1962
Salzer, E, and Schachter, C., Counterpoint in Composition: The Study of Voice
Leading, New York: McGraw-Hill, 1969
Samarotto, E, “Intimate Immensity in Schubert’s The Shepherd on the Rock,’ in
Structure and Meaning in Tonal Music: Festschrift in Honor of Carl Schachter,
ed. L. P. Burstein and D. Gagné, Hillsdale, N.Y.: Pendragon Press, 2006, pp.
203-223
“Determinism, Prediction, and Inevitability in Brahms’s Rhapsody in Eb major,
op. 119, no. 4? Theory and Practice 32 (2007), pp. 69-99
Schachter, C., “Motive and Text in Four Schubert Songs,” in Aspects of Schenkerian
Theory, ed. D. Beach, New Haven: Yale University Press, 1983, pp. 61-76;
reprinted in C. Schachter, Unfoldings: Essays in Schenkerian Theory and
Analysis, ed. J. Straus, Oxford University Press, 1999, pp. 209-220
Schenker, H., Harmonielehre: Neue musikalische Theorien und Phantasien I,
Stuttgart: Cotta, 1906; abridged trans. E. M. Borgese, ed. O. Jonas as Harmony,
University of Chicago Press, 1954
Der Tonwille: Flugblatter zum Zeugnis unwandelbarer Gesetze der Tonkunst,
einer neuen Jugend dargebracht, 2 vols., Vienna: A. J. Gutmann, Leipzig:
EF. Hofmeister, 1921-1924; ed. W. Drabkin, trans. I. Bent et al. as Der Tonwille:
Pamphlets in Witness of the Immutable Laws of Music, Offered to a New
Generation of Youth, Oxford University Press, 2004-2005
Das Meisterwerk in der Musik: Ein Jahrbuch, 3 vols., Munich: Drei Masken,
1925-1930, ed. W. Drabkin, trans. I. Bent et al. as The Masterwork in Music, 3
vols., Cambridge University Press, 1994-1997
Der freie Satz: Neue musikalische Theorien und Phantasien IIT, Vienna: Universal,
1935; 2nd rev. edn, ed. O. Jonas, Vienna: Universal, 1956; trans. and ed.
E. Oster as Free Composition, New York and London: Longman, 1979; reprint
trans. Stuyvesant, N.Y.: Pendragon Press, 2001
Seidel, E., Die Enharmonik in den harmonischen GroBformen Franz Schuberts,
PhD diss., Johann Wolfgang Goethe-Universitat, Frankfurt am Main,
1963
“Ein chromatisches Harmonisierungs-Modell in Schuberts Winterreise” Archiv
fiir Musikwissenschaft 26 (1969), pp. 285-296
Siciliano, M., “Neo-Riemannian Transformations and the Harmony of Franz
Schubert,” PhD diss., University of Chicago, 2002
“Two Neo-Riemannian Analyses,’ College Music Symposium 45 (2005), pp.
81-107
Smith, C. J., “The Functional Extravagance of Chromatic Chords? Music Theory
Spectrum 8 (1986), pp. 94-139; response by D. Beach and reply by C. Smith in
Music Theory Spectrum 9 (1987), pp. 173-194
 Select bibliography SUS

Smith, P. H., “Outer-Voice Conflicts: Their Analytical Challenges and Artistic

Consequences,” Journal of Music Theory 44 (2000), pp. 1-43
“Harmonic Cross-Reference and the Dialectic of Articulation and Continuity
in Sonata Expositions of Schubert and Brahms,’ Journal of Music Theory 50
(2006), pp. 143-179
Sobaskie, J., “A Balance Struck: Gesture, Form, and Drama in Schubert’s E-flat-
Major Piano Trio,’ in Le style instrumental de Schubert: Sources, analyse, évo-
lution, ed. X. Hascher, Paris: Publications de la Sorbonne, 2007, pp. 117-148
“Schubert's Self-Elegies,” Nineteenth-Century Music Review 5 (2008), pp. 71-105
Stein, D., “Schubert's ‘Die Liebe hat gelogen’: The Deception of Mode and Mixture,’
Journal of Musicological Research 9 (1989-1990), pp. 109-131
“Schubert's “Erlkénig’: Motivic Parallelism and Motivic Transformation,” 19th-
Century Music 13 (1989-1990), pp. 145-158
Suurpaa, L., “The Path from Tonic to Dominant in the Second Movement of
Schubert's String Quintet and in Chopin's Fourth Ballade? Journal of Music
Theory 44 (2000), pp. 451-485
Swinden, K., “When Functions Collide: Aspects of Plural Function in Chromatic
Music,’ Music Theory Spectrum 27 (2005), pp. 249-282
Taruskin, R., “Chernomor to Kashchei: Harmonic Sorcery; or Stravinsky’s ‘Angle?
Journal of the American Musicological Society 38 (1985), 72-142
The Oxford History of Western Music, vol. 3: The Nineteenth Century, Oxford
University Press, 2005
Telesco, P., “Forward-Looking Retrospection: Enharmonicism in the Classical Era,”
Journal of Musicology 19 (2002), pp. 332-373
Temperley, N., “Schubert and Beethoven's Eight-Six Chord,” 19th-Century Music 5
(1981-1982), pp. 142-154
Tovey, D. E, “Franz Schubert, in The Heritage of Music, ed. H. J. Foss, London:
Oxford University Press, 1927, pp. 82-122; reprint edn, Freeport, N.Y.: Books
for Libraries Press, 1969
“Tonality, Music and Letters 9 (1928), pp. 341-363; reprinted as “Tonality in
Schubert; in The Main Stream of Music and Other Essays, Oxford University
Press, 1949
Triimpy, B., Die Raumvorstellung in Schuberts Harmonik, Winterthur: Amadeus,
1995
Waldbauer, I. FE. “Recurrent Harmonic Patterns in the First Movement of Schubert's
Sonata in A Major, D. 959? 19th-Century Music 12 (1988-1989), pp. 64-73
Webster, J., “Schubert’s Sonata Forms and Brahms’s First Maturity,’ 19th-Century
Music 2-3 (1978-1980), pp. 18-35, 52-71
Wintle, C., “Franz Schubert, ‘Ihr Bild’ (1828): A Response to Schenker’s Essay
in Der Tonwille, Vol. 1; Music Analysis 19 (2000), pp. 10-28 (plus trans. by
R. Pascall, pp. 3-9)
Index of Schubert’s works

“Die Allmacht” (D. 852) 19-21, 273 “Gesang der Geister tiber den Wassern”
“Am Feierabend” from Die schéne Miillerin (D. 714) 99-100
(D. 795/5) 34-35, 85-86, 273 “Das Grab” (D. 330) 294
“Am Flusse” (D. 160) 116-121
“An die Nachtigall” (D. 497) 21-22 “Hagars Klage” (D. 5) 8-10, 22, 23, 30, 83,
“An Emma’ (third version, D. 113) 66-67 152, 269
“Auf dem Flusse” from Winterreise (D. “Das Heimweh” (D. 851) 55-58, 71-72
911/7) 236-244, 298-299
“Auf dem Strom” (D. 943) 11-12, 272 “Thr Bild” from Schwanengesang (D.
“Auflésung” (D. 807) 30-32, 272 957/9) 300
“Aus Heliopolis W” (D. 754) 54-55, 83-84, Impromptus in C Minor, Eb, Gb Major, and Ab
282, 291 Major (D. 899) 201-235, 286, 293,
294-298
“Beim Winde” (D. 669) 272 Impromptu in B} Major (D. 935/3) 294
“Die Berge” (D. 634) 97-99
“Die bése Farbe” from Die schéne Miillerin “Der Jager” from Die schone Miillerin
(D. 795/17) 102-104, 298 (D. 795/14) 92-93, 99, 281
“Die junge Nonne” (D. 828) 191-200,
Characteristic March No. 2 in C Major for 293-294, 300
Four-Hand Piano (D. 968b [886])
52-54 Landler in D Major from Zwolf Landler
(D. 790/1) 15-18
“Das war ich” (D. 174) 28-29, 73-74 Landler in G# Minor from Zwélf Landler
Divertissement a la Hongroise for Four-Hand (D. 790/6) 126-131, 284
Piano (D. 818), mvmt. 3 55 “Der Leidende” (D. 432a) 18, 20
“Die Einsiedelei” (D. 337) 113-116, 220, 230, “Die Liebe hat gelogen” (D. 751) 81-82
302 “Liebesbotschaft” from Schwanengesang (D.
957/1) 268
“Erinnerungen” (D. 98) 77-78, 289 “Liebesrauch” (D. 179) 72, 272
“Erlkonig” (D. 328) 207, 214 “Lied der Mignon” (D. 877/3) 58-59
“Erster Verlust” (D. 226) 121-126, 234, 238,
283 Mass No. 2 in G Major (D. 167), Credo 34
“Die Erwartung” (D. 159) 44-45 Mass No. 6 in Eb Major (D. 950), Sanctus
70-71
Fantasie in C Major (“Wanderer,” D. 760), Menuet in C Major (D. 995/1) 68-69
mvmt. 4 41-44 Moment Musical in Ab Major (D. 780/2)
“Fulle der Liebe” (D. 854) 58 104-107, 282
Moment Musical in Ab Major (D. 780/6)
“Ganymed” (D. 544) 135-148, 284-286, 287 67
“Gefror’ne Thranen” from Winterreise Moment Musical in C# Minor (D.
(D. 911/3) 300 780/4) 93-95, 96-97
“Das Geheimniss” (D. 250) 20-21, 23 “Morgengruf” from Die schéne Miillerin
“Geist der Liebe” (D. 414) 33 (D. 795/8) 270
 Index of Schubert’s works Sly

“Der Miller und der Bach” from Die schéne “Schwertlied” (D. 170) 11-12
Miillerin (D. 795/19) 108-110 String Quartet No. 14 in D Minor (D. 810),
mvmt. 1 83-85, 99-100; mvmt. 3 284;
“Nachthymne” (D. 687) 41-42, 268 mvmit. 4 16-18, 295-296
String Quintet in C Major (D. 956), mymt.
Piano Sonata in A Major (D. 664), mvmt. 1 65-66, 78-80, 83, 280; mvmt. 2 298
1 73-74; mvmt. 2 65-66, 72 “Suleika I” (D. 720) 75-77
Piano Sonata in A Major (D. 959), mymt. Symphony No. 1 in D Major (D. 82), mymt.
3 26-28 IS
Piano Sonata in A Minor (D. 537), mvmt. Symphony No. 3 in D Major (D. 200), mvmt.
1 281; mvmt.3 3-8, 8, 9, 85, 268, 280 3 44-47, 300
Piano Sonata in A Minor (D. 784), mvmt. Symphony No. 4 in C Minor (“Tragic,”
2 179-190, 278, 291-293 D. 417), mvmt. 1 48
Piano Sonata in A Minor (D. 845), mvmt. Symphony No. 6 in C Major (D. 589), mvmt.
1 72-73, 86-87 4 275
Piano Sonata in Bb Major (D. 960), mvmt. Symphony No. 8 in B Minor (“Unfinished,”
1 48-49, 245-263, 299-305 D. 759), mvmt. 1 159-178, 289-291, 295,
Piano Sonata in C Minor (D. 958), mvmt. 304
1 12-13, 15, 24-26; mvmt. 2 80-81; Symphony No. 9 (7) in C Major (“The Great,”
mvmt.4 108 1D. 944), mvmt. 3 95, 100
Piano Sonata in D Major (D. 850), mvmt.
1 50-52; mvmt. 2 87-88, 110-113, 294; “Todtengraber-Weise” (D. 869) 68-70
mvmt. 3 60-64, 181, 195-196, 274-275, “Tranenregen” from Die schone Miillerin
278, 281, 300, 301 (D. 795/10) 100-101
Piano Sonata in F Minor (D. 625), Trio
269 “Ungeduld” from Die schone Miillerin
(D. 795/7) 36-40, 82, 273
Quintet in A Major (“Trout,” D. 667), mvmt.
1 149-158, 286-289 Variation on a Waltz by Diabelli
(D. 718) 89-91
Rondo in B Minor for Violin and Piano “Von Ida” (D. 228) 92-93
(D. 895) 29-31
Walz in D Major from Zwanzig Walzer for
“Der Sanger” (D. 149) 83-84 Piano (D. 146/20) 23-24
“Schlaflied” (D. 527) 82 “Der Wanderer” (D. 649) 32-33
Die schéne Miillerin (D. 795) 147, 282-283 Winterreise (D.911) 201
Index of names and concepts

absent root 5, 29, 30, 37, 43, 66, 87-88, 92, bullet symbol xi, 5
284, 285 Burstein, L. P. 271, 272
Agawu, K. V. 278-279
Almén, B. 304 cadential § 34, 167, 174, 197, 224-225, 274,
antecedent/consequent 103, 117-118, 148, 286, 287, 298
161, 175-176,191, 222, 237 Carpenter, P. 297
antipode 19, 64, 65, 113, 115, 157, 220, 271, Chopin, F. 266, 282
283 Piano Sonata in Bb Minor (“Funeral
applied dominant 3, 5, 7, 10, 11, 196, 223, 225, March,” op. 35) 296
270, 271 chordal evolution 5, 10, 16-17, 36, 43, 123,
arrow symbols (> and >) 7 126, 207, 221, 225, 239, 242, 268, 269,
ascending 5-6 sequence 36, 44-52, 68, 75-77, 270, 280, 285, 286, 296, 298
83-84, 90, 99-101, 100-101, 108, 140, chordal glide see parallel progression
148, 155-157, 161-162, 192, 193, 198, chordal ninth 5, 6, 10, 23, 29-30, 37, 43, 92,
224-225, 254, 274, 278, 281, 282, 290, 99, 112, 150, 169, 196, 207, 229, 235, 246,
291298) 267, 268, 280, 281, 284, 287, 292, 294
assertion 6, 7, 15, 59, 65, 71, 73, 78, 80, 108, chordal seventh 5, 18, 29, 37, 43, 44, 71, 73,
WSS PMS) MISS NTS OES OPPs PS) 77, 80, 99, 104, 112, 120, 154, 166, 195,
230, 271, 277, 280, 290, 294, 298 207, 224, 229, 253-254,
280, 283
associative harmony 297 chordal unfurling see unfurling
augmented dominant 10 chromatic variant 56, 59-60, 64, 70, 105, 150,
augmented sixth chords 5, 74, 78-80, 95, 112, 200, 222, 239, 242, 247-250, 254, 255,
115, 169-170, 193, 200, 228, 235, 249, 256, 258-259, 260-261, 269, 275, 276,
267, 274, 278, 287, 300 280, 283, 284, 287, 303
auxiliary progression 21, 222 chromaticism 3, 6, 7, 11, 15, 18, 19, 20, 23, 26,
29, 47, 48, 50-51, 63, 68, 70, 72, 75, 80,
background 203 81, 85, 88, 90, 95, 96, 99, 100, 102, 103,
Baker, J. M. 304 106, 107, 110, 121, 124, 128, 140, 143,
Beach,D. 149-158, 275 150, 161, 168, 174, 191, 200, 216, 225,
Beethoven, L. von 188 228, 233, 234, 239, 242, 243, 246-247,
Piano Sonata in Ab Major (op. 26) 113, 220, 249, 252, 253, 254, 255, 257, 267, 270,
282 272, 280, 281, 282, 290, 292, 298, 301,
String Quartet in C Major (op. 59, no. 3) 302, 304
293 circle of fifths 14, 21, 52, 63-64, 73-74, 76-79,
String Quartet in F Major (op. 18, no. 1) 91, 92-95, 100, 103, 104, 109, 126, 129,
295 136137, 144, 151, 155-156, 170-171, 176,
Symphony No. 8 in F Major (op. 93) 214, 222, 239, 255-257, 273, 275, 276,
288 280, 281, 282, 289, 301, 304
Trio for Violin, Cello and Piano in D Major circle of thirds 52-56, 70-71, 81, 86-87, 88,
(op. 70, no. 1) 287 115, 176, 184, 250-252, 258, 276, 295,
Variations on a Waltz by Diabelli (op. 301, 302
120) 89 Clark, S. 135, 141-148, 285-286, 287, 288
Berlioz, H. 270 Clementi, M. 283
 Index of names and concepts a19

Cohn, R. 245-263, 275, 299-305 Everett, W. 284

collision 6, 11, 186, 235 exposition 73
common-tone augmented sixth chord 280,
294, 300 figured bass 5, 8, 32
common-tone diminished seventh chord 139, Fisk, C. 201-235, 294-298
182 Funktionstheorie 269
Cone, E1270
Craigher de Jachelutta, J. N. 191 Ganymede 135-136, 139-141, 146-147
Cube, F.-E. von 273 glide see parallel progression
cyclic repetition 149-150, 172, 201, 203, 207, Goethe, J. W. von 116, 121, 135, 136, 285
208, 209, 210, 211, 214, 216, 286, 296-297
Hascher, X. 276, 299
Daedalus 215 Hatten, R. S. 179-190, 264, 287, 291-293
Darcy, W. 286, 290, 293 Haydn, F. J. 266, 272
Death 180, 184, 189-190 Piano Sonata in Eb Major (Hob. XVI: 49),
death knell 190, 210 mymt. 1 300
deceptive progression 10, 262, 277, 278 Hepokoski, J. 286, 290, 293
development section 73-74 hexatonic poles 275
Diabelli, A. Hoftmann, E. T. A. 287
Waltz in C Major 89-91, 280 homosexuality 135, 147
diatonic sequence 48 hypermeter 26, 286
Dineen, M. 297
dominant emulation 4, 7, 9, 15-17, 18, 20, 21, Icarus 215, 220
30, 36, 41, 48, 50-51, 52, 56, 63, 68, 71, idiosyncratic sequence 48, 224
73, 75, 77-78, 80, 83, 84, 85, 90-91, 95, illness (Schubert’s) 187-188, 189-190, 259,
OOM 1027 103 104 LOZ SLO; MLS OS I275 262, 291, 304
128, 138, 145, 148, 151, 162, 166, 170, incidental dissonance 267
193, 195, 196-197, 200, 214, 218, 221, interruption 60-62, 90-91, 159, 161, 166, 175,
222, 226-227, 229, 232, 233, 239, 247, 202, 204, 237-238, 240, 276, 281, 293
249, 253-254, 259, 260-261, 269, 271,
273, 274, 276, 277, 278, 279, 280, 282, Jackson, T. L. 286
285, 289, 298, 299, 300, 301, 303 Jelensperger, D. 282, 292
dominant parallel 240
Dunsby, J. 283 Kessler, D. 301
Kirnberger, J. P. 267, 268, 269, 282
elision 29, 31, 43, 52, 82, 83-84, 109, 204, 219, Kopp, D. 191-200, 275, 293-294
224, 227, 233, 249, 268, 302 Kramer, L. 135, 141-148, 264, 283, 284,
embellishment 8, 31-32, 33, 55, 56, 58, 59, 64, 285-286
65, 67, 70, 104, 109, 113-115, 128, Kramer, R. 264, 268, 269, 289, 295
137-138, 142, 150-151, 159-160, 166, Krebs, H. 272, 299
167, 168, 177, 178, 179-181, 183-184,
192, 194, 198, 204, 216-219, 226, 227, Laitz, S. 283
230, 232, 246, 249, 252, 253, 260, 272, Lanz, J. 264
277, 278, 282, 292, 294, 298, 299, 300 Lerdahl, F. 300
enharmonic equivalence 21, 41, 44, 49-50, Lewin, D. 236-244, 270, 298-299
52-53, 55-56, 70, 71, 96-97, 105, 136, linear progression 41-64
146, 151, 156, 160, 168-170, 185, 188- Liszt, F.
190, 194, 198, 199, 200, 216, 219, 228, “Tl penseroso” from Anneées de pélerinage:
230, 246, 255, 256, 275, 276, 278, 279, Italie 113, 282
287-288, 292, 294, 296, 297, 300, 301, 303 Lobe, J. C. 267, 268, 269
equal subdivisions of the octave 55-56, 71, Loquin, A. 274
154-158, 275, 287, 301 Louis, R. 268, 269
essential dissonance 267 lower-fifth chord 33
320 Index of names and concepts

lower-third chord 6, 10, 56-60, 65-71, 90-91, pivot-chord modulation 193

95, 143, 150, 153-155, 188-190, 258, 261, Pyrker, L. 56
262, 269, 276, 281, 281, 284, 300
lowered supertonic see bII Quinn, I. 267

Mac, S. Y. 289 Rameau, J.-P. 268

McClary, S. 284 registral shift 7, 8, 62, 68-69, 88, 90, 120, 168,
Marston, N. 305 175, 211, 212, 2445 2535296

Marx, A. B. 293 Reichardt, J. F.

medial caesura 184 “Ganymed” 285, 286
Mehrdeutigkeit 14-15, 128, 182-183, Réti, R., 287
188-190, 204, 222, 228, 229, 233, 235, retrograde 162, 171, 196
294, 297, 304 Riemann, H. 269, 270
Messing, S. 284 Riley, M. 278
Meyer, L. B. 287 Rings, S. 284, 297
modal mixture see parallel keys Roman numeral 4-5, 6, 10, 14, 26, 53, 59, 65,
modulo 7 vs. modulo 12 50, 53, 55-56, 70-71, 71, 142, 144, 177, 183, 191, 193-194,
250-252, 274, 275, 276, 279, 281, 287, 224-226, 246, 267, 269, 270, 271, 273,
290, 301, 303 DT, 2ole 282,302
Monelle, R. 281 Rosand, E. 289
Mozart, W. A. 175, 283 Rossini, G. 153, 175, 276
Piano Sonata in Bb Major (K. 281) 283 Guillaume Tell 281
String Quintet in C Major (K. 515), mvmt. Rothstein, W. 276
1 268
Miller, W. 102, 236, 244 Salzer, F. 276, 289, 299
multiple meaning see Mehrdeutigkeit Samarotto, F. 275, 296, 297
Satan 180, 291
natural minor 267 Schachter, C. 276
Neapolitan sixth see bII Schenker, H. 236, 240-241, 243, 249, 265, 269,
neo-Riemannian analysis 297 270; DIN, 27S) 27 52 Oily BOLNL OS, LOU5
Newbould, B. 280 288, 289, 290, 293, 298, 299
Nolan, C. 274 Schenkerian analysis viii, xii, 60-62, 72, 91,
141, 148, 186, 196, 202, 204, 240-241,
obstinate progression 48, 83-84, 192, 193, 243, 245, 276, 281, 285, 286, 291-292,
250; 259; 2797281, 290 294, 296-297, 298, 302
omnibus 72, 164, 176 Schoenbergian analysis 297
open parentheses 26, 59, 246 Schumann, R. 284
Oster, E. 281, 298 Piano Sonata in F# Minor, op. 11, mvmt. 4
overlap 162, 236 273
Sechter, S. 282
parallel fifths 6, 11, 72, 126, 172, 272, 277, 298 seismic composition 137, 173-174, 284, 291
parallel keys 56, 75, 89, 96, 99, 101-106, 113, sexual activity (Schubert’s) 257, 304
142, 195, 198, 200, 210-211, 212, 213, Shakespeare 295
Qa, 2122052225 2275 229237
—2385 Siciliano, M. 299
256 Smith, C. J. 272
parallel progression 41-44, 50, 52, 53, 83-84, Smith, P. H. 277, 297
88, 92, 99, 108, 168-170, 177, 192, 234, Sobaskie, J. W. 298
280, 282, 290, 291 Solomon, M. 284
parenthetical passage 29, 37, 55-56, 183 Steblin, R. 305
passing chord 23, 29, 224 Stocken, F. 305
peculiar juxtapositions 26-31 Stufentheorie 4, 191, 247, 263, 264, 269
Picardy third 16, 282 subdominant recapitulation 149, 262, 286
Piston, W. 267 Swoboda, A. 280, 283, 294
 Index of names and concepts Syl

Taruskin, R. 159, 174-178, 277, 289-291, 295 voice leading 7, 18, 68, 85-86, 164, 175, 176,
Thuille, L. 268, 269 272, 283, 294
tonicization 3, 12, 14, 41, 96, 102, 104, 120,
123, 154, 161, 162, 170, 184, 199, 200, Wagner, R.
2055, 213; 2325233, 252253, 255; Lohengrin 268
258-259, 279, 285, 290, 295, 304 Weber, G. 290, 295
Tovey, D. F. 304 Webster, J. 301
Trumpy, B. 300 Weitzmann, C. F. 274, 280, 283, 294
twelve-note chromatic space see modulo 7 vs. whole-tone scale 71, 250, 256, 277
modulo 12 wobbly note 29, 31, 54, 56, 58, 64, 65, 66, 70,
80, 81, 83, 95, 106, 108, 143, 150, 152,
Ubergreifen 72, 277, 283 1545173; 196; 197; 2005215; 216,219,
unfolding 19, 23, 62, 166, 202, 243, 271, 274, 221-222, 227, 229, 230, 249, 272, 276,
279, 285, 290 DHT p27o
unfurling 31-35, 47, 48, 55, 84, 102, 109, 140,
159, 166, 174, 192, 216, 224-225, 230, Zeus 135-136, 140-141, 147
240, 247, 273, 278, 287, 290, 294, 299
upper-fifth chord 17, 24-26, 63, 73, 78, 96-99, 5-6 shift 6, 9, 15-16, 18, 20, 21, 23, 26, 36, 41,
108, 113, 124, 130, 137-138, 141, 225, 44, 45, 48, 50-52, 56, 59, 62, 65, 70, 81,
D59;210, 2/1, 279; 21852082, 290 83, 85, 90, 93, 99, 102, 107, 109-110, 124,
upper-third chord 10-11, 12, 14, 26, 56-60, 131, 140; 51-1535 156-1575 59}
63-64, 65-71, 80, 82-83, 85, 87-88, 161-162, 164, 165, 171, 180, 182, 195,
90-91, 95, 113-115, 137-138, 143, 151, 197, 200, 202, 205, 216, 217, 221,
152-154, 160, 188-190, 195-197, 202, 222-223, 224, 225, 226, 228, 230, 232,
208, 210, 213, 222, 239, 242-243, 249, 233, 242, 243, 246-250, 252, 255-259,
252, 254-256, 261, 262, 269, 272, 276, 260, 262-263, 269, 270, 271, 273, 274,
278, 281, 282, 284, 287, 290, 298, 301, 278, 279, 280, 283, 293, 294, 296, 298,
302, 303, 304 301, 302, 303
I-space 64, 67, 95, 129, 131, 195, 202,
Verdi, G. 222-223, 244, 277, 278
Luisa Miller 284 IV-space 195
Vincent, H. J. 274 V-space 11, 12-14, 28, 80, 95, 105, 123, 222,
Vogler, G. J. 267, 269 LID; Diy LOL
voice exchange 29, 67, 68-69, 88, 143, 246, bIT 15, 103, 106-113, 121, 204-205, 225, 249,
249, 283 257, 267, 270, 282, 294
 OS a Seen bo 4
| (OD) Ot at i &

A ; i 0a Ah SOY Sk
1 SA fewtguey/ itp ma Ae
AT ot¥prenthe.) Jz, pape
we i setw : ~ eh OVE eat

ee ee
hae ee j=3 ") pee ieVé an ee Pr ”"
| OY ATI! eee wror-ehrlp th eipere ney atten
PD ABO & 06 10 AS orig Baw a ee >
oe ere ei ey am _
” Sit aeat Bt Pai 0% Ae _ ies, 28 get TS ET &
= © | be, Oe tt Aa ATH ee eaats
R ~ a we a
ake 0} Litulll 22 A Th Ai
— Abel OEM Ae mel Ol SATS
ATE ARS
. ri Sen el
> ae ai i, di utieei GS Gai fi :
(ATLAS BAW) 0d 28 at a So de
~ mess

Oe hee Pe eo a ee ee
bri 7
- ne 5 MES A }
nc, (ALL Site i ed eT -— Aba Ta b ;
ABRAM I OREae (iE 9 -ETHR j08,
ity STL A)4 RL ity, AIA PRT ALLA
(Ge) =) Wi Aes efett or wise, ee |
IL Ra ANS 145
ee ee ee
ME wi ie imei pA Lol aS
1 @) ye Div ite F D
olen galjly .t63Oe eepl gee i Ke
aren
a tineinn Melber Wet 4a in Se.
Veto a se ae “a. i _ -_
7 Ahh 25) aD) AY At wt Bl dopey 6+9 Ooi i
3a. 15 26S 550, 1 ie, 18. 6
~ Sdn Ai A AU EN obi Sine bid AS 05 Crone.
PAs) ow, GR OL Fa 29
SE S= Sr), TE
TA te eee
PUreae=s 4 7 hel) ks ae
tha’ @
Ficeed 7 A ;
ee ' 7 7 Pr) = :

emai, S204

See S18 20 * ere |

magl pe—aee) 4) i ear |? o

SA VA (16. 10.) ‘TrLIG2 Sea.
et OS oa]

Terre pgm 1 ta Uk
a is Th, 24
Pet earn oi
ands: Jay
Oe
a p
DAVID DAMSCHRODER is Professor of
Music Theory at the University of Minnesota
School of Music, where he teaches a range of
courses devoted to the analysis of eighteenth- and
nineteenth-century music, the repertoire that also
serves as the focus for his scholarly writings and
performance activities on fortepiano. His previous
books include Thinking About Harmony: Historical
Perspectives on Analysis, Music Theory from Zarlino
to Schenker, Listen and Sing, and Foundations of
Music and Musicianship.

Jacket illustrations: top — analysis of Quintet in

A Major (“Trout,” D. 667), mvmt. 1, mm. 64-84;
middle — analysis of Symphony No. 8 in-B Minor
(“Unfinished,” D. 759), mvmt. 1, mm. 125-169;
bottom — analysis of “Auf dem Flusse” from
- Winterreise (D. 911/7), mm. 41-54.

Jacket designed by Hart McLeod Ltd_

PRINTED IN THE Uf
 HARMONYIN
SCHUBERT
One of Western music’s great harmonists, Franz Schubert created a
wondrous and treasured body of music that has retained its fascination
to this day. His innovative harmonic practice has been a topic of
lively discussion among analysts for generations. Harmony in Schubert
presents a fresh approach, yielding insightful readings of a large
and varied range of excerpts, as well as readings of fifteen complete
movements spanning Schubert’s chamber, choral, orchestral, piano,
and vocal output. Damschroder reformulates the apparatus for Roman-
numeral harmonic analysis, integrating his own speculations with
various strands of historical analytical thought, including Schenkerian
principles and historical perspectives. In addition, he juxtaposes his
readings of complete movements by Schubert with discussions of how
they have been interpreted by other Schubertian analysts. The book sets
a new direction for the future of music analysis, proposing innovative
improvements on existing methodologies.

m © 125 129130 131 132 133 134 145 146 153 154 158 161 162 4