Harmony in Mendelssohn and Schumann [PDF]

Zitiervorschau

Harmony in Mendelssohn and Schumann

This innovative book continues David Damschroder’s radical reformulation of harmonic theory, presenting a dynamic exploration of harmony in the compositions of Mendelssohn and Schumann, two key figures of nineteenth-century classical music. This volume’s introductory chapters creatively introduce the basic tenets of the system, with reference to sound files rather than notated music examples permitting a more direct interaction between reader and music. In the Masterpieces section that follows, Damschroder presents detailed analyses of movements from piano, vocal, and chamber music, and compares his outcomes with those of other analysts, including Benedict Taylor, L. Poundie Burstein, and Peter H. Smith. Expanding upon analytical practices from the eighteenth and nineteenth centuries, and strongly influenced by Schenkerian principles, this fresh perspective offers a stark contrast to conventional harmonic analysis – in terms of how Roman numerals are deployed and how musical processes are described in words. d a v i d d a m s c h r o d e r is Professor of Music Theory at the University of Minnesota. His current research focuses on harmony in tonal music, a project that began with a careful examination of historical analytical practices, the basis for his Thinking About Harmony: Historical Perspectives on Analysis (Cambridge, 2008). He has since published studies on harmony in selected composers including Haydn, Mozart, Beethoven, Schubert, and Chopin, and he is the author of Tonal Analysis: A Schenkerian Perspective (W. W. Norton, 2017).

Harmony in Mendelssohn and Schumann david damschroder University of Minnesota

University Printing House, Cambridge CB2 8BS, United Kingdom One Liberty Plaza, 20th Floor, New York, NY 10006, USA 477 Williamstown Road, Port Melbourne, VIC 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781108418034 DOI: 10.1017/9781108284110 © David Damschroder 2018 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 2018 Printed in the United Kingdom by Clays, St Ives plc A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data Names: Damschroder, David. Title: Harmony in Mendelssohn and Schumann / David Damschroder. Description: Cambridge, United Kingdom ; New York, NY : Cambridge University Press, 2017. | Includes bibliographical references and index. Identifiers: LCCN 2017033145 | ISBN 9781108418034 (alk. paper) Subjects: LCSH: Mendelssohn-Bartholdy, Felix, 1809–1847 – Harmony. | Schumann, Robert, 1810–1856 – Harmony. | Music – 19th century – Analysis, appreciation. Classification: LCC MT90 .D35 2017 | DDC 781.2/5–dc23 LC record available at https://lccn.loc.gov/2017033145 ISBN 978-1-108-41803-4 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 i

methodological orientation: harmonic analysis through listening [1]

1 Foundational diatonic processes [3] 2 Surges

[16]

3 IV5–6 V [24] 4 I5–6 II [31] 5 Surging 6-phase chords 6 Surges with ninths

[42]

[50]

7 Colorful variants of II [60] 8 III along the path from I to V part ii masterpieces

[69]

[79]

9 Mendelssohn: Octet in E♭ Major (op. 20), movement 1 in response to Greg Vitercik and Benedict Taylor [81] 10 Mendelssohn: Song without Words in F Major (op. 85, no.1) in response to Allen Cadwallader [116] 11 Schumann: “Warum?” from Phantasiestücke (op. 12) in response to L. Poundie Burstein [127] 12 Mendelssohn: Song without Words in A♭ Major (op. 53, no.1) in response to Yosef Goldenberg [136] 13 Schumann: Three songs from Liederkreis (op. 39) in response to Charles Burkhart and David Ferris [146] v

vi

Contents

14 Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1) in response to Deborah Stein [171] 15 Schumann: Sonata in A Minor for Violin and Piano (op. 105) in response to Peter H. Smith [183] Notes [232] Select bibliography [271] Index of Mendelssohn’s works [277] Index of Schumann’s works [278] Index of names and concepts [279]

Preface

How did composers working just before or during the first half of the nineteenth century conceive of and apply harmony? My emerging answer to that question has by this point devoted due attention to the four superlative composers active in or near Vienna in the decades around 1800 (Harmony in Haydn and Mozart, Harmony in Beethoven, and Harmony in Schubert), as well as to a wondrous expatriate Polish pianist/ composer in Paris a bit further into the century (Harmony in Chopin). Now it is time to assay what Mendelssohn and Schumann were accomplishing in various German locales. Though textbook and treatise authors of the era were active in building analytical systems to make sense of the contemporary harmonic practices (generally employing Roman numerals, as I relate in Thinking About Harmony), their budding efforts have been extensively transformed in my writings, in part because I incorporate notions proposed by later thinkers (Heinrich Schenker in particular) and in part through my willingness to jettison aspects of conventional modern harmonic analysis stemming from those early efforts in favor of fresh and (I trust) improved ways of proceeding that may more fully unlock for us the processes these composers were pursuing. Though my work is intended mainly for graduate students and professional musicians, I hope that my reconfigured harmonic theory also will be introduced at the foundational level of instruction. Harmony in Beethoven offers an inviting Harmonielehre that might aptly supplement any of the standard undergraduate harmony texts, giving initiates who may have become complacent or indifferent an eye-opening exposure to a new way of thinking about the topic. As a complement to the Harmonielehre, this volume opens with Harmonic analysis through listening (chapters 1 through 8), written in a way that should be accessible to undergraduates while concurrently offering more seasoned readers plenty of rewarding content. It is curious yet true that, though music enters our consciousness through the ears, almost all of what you might draw upon to assist in developing your analytical capacities is absorbed through the eyes. I have taken the

vii

viii

Preface

unusual step of here introducing the foundations of my harmonic perspective through the act of listening to music. In contrast to passive reading about harmonic analysis, you will be invited to ponder questions that get to the heart of how a passage you listen to is conceived harmonically and realized in its details, and then to compare your responses to my suggestions. These featured excerpts may be accessed via audio examples available on this volume’s web page (www.cambridge.org .9781108418034). Rhythmic grids annotated with location symbols provide the means of identifying discrete moments within each sounding excerpt, so that both the questions and the responses can be crafted with specificity. (It is important that you peruse my proposed responses in the endnotes even when you are sure that your own are correct, since I may introduce terminology or a strategy that will be needed in future responses.) I have indicated each question’s level of challenge by the number of bullets (•, ••, or •••) to the left of its identifying number, so that you have the option of pursuing only the easier questions during a first pass through the chapters, returning to the others later. Each listening excerpt is introduced by some preliminary commentary and the vocal generation of its essential features, focusing on the harmonic relationships that will be featured in what follows. Though you might be taken aback by my request that you proceed through eight chapters of a “scholarly” book in this interactive mode, I hope that you will find it to be a transformative experience, touching on vital components of musical perception and comprehension that are difficult to access through more conventional modes of author–reader interactions. This book’s centerpiece is a Masterpieces section offering detailed analyses of compositions by Mendelssohn and Schumann, featuring the tools that I advocate for harmonic analysis. Roman numerals generally are displayed in the context of Schenkerian graphs, which provide insights regarding harmony’s hierarchical organization as well as on a range of other parameters. (During this portion of the book, some prior exposure to Schenkerian analysis is assumed, either through my Tonal Analysis: A Schenkerian Perspective, or by some other means.) As has been the case also in the earlier analytical volumes of my Harmony Project, I guide you through a direct comparison of each analysis with an interpretation by another prominent analyst (or sometimes two others), thereby deepening your perceptions regarding these works and highlighting what is at stake in the analytical process. (These alternative analyses all appear in publications that should be available at any collegiate music library. The critiques are set off from the main flow of my analyses by shading.) Consequently, my

Preface

Harmony in . . . volumes not only point a way forward for the study of nineteenth-century music but also together comprise a unique and wideranging assessment of the state of tonal analysis in English-language scholarship over the past fifty or so years. Since my Masterpieces chapters already offer substantial analytical challenges, I have not endeavored to extend the project’s purview even further to incorporate assessments of publications in other languages. (Such works occasionally have been addressed in the endnotes.) Likewise some worthy analytical publications were passed over because they focus on issues only marginally related to harmony, making the sort of comparative analysis pursued here unworkable. Though I completed my first two Cambridge books without a clear sense of what ultimately was to emerge, or even awareness that something warranting being called a Harmony Project was in the works, by now my six monographs from Cambridge together constitute a bountiful and unified body of analytical commentary on this important repertoire. I intend next to explore harmony in music after 1850, leading in due time to Debussy. I thank the University of Minnesota for granting me a sabbatical leave permitting a year of uninterrupted work on this volume and for the support of an Imagine Fund award that both covered the costs associated with the music examples and sound files and allowed me to acquire books and to visit major research libraries. I am grateful to the New York Public Library, Astor, Lenox and Tilden Foundations, for allowing me to purchase on microfilm and to make references to the Oster Collection: Papers of Heinrich Schenker. As in the earlier volumes of my Harmony Project, Peter Smucker has provided expert setting of the music examples.

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 may be indicated in melodic succession: ascending as CC. A black arrow may be used to indicate a descending-fifth relationship that is or emulates a V(7)–I succession, whereas an outline arrow may be used to indicate a succession from a chord of the augmented-sixth type; for example, C➔F–D➔G➔C; C–A♭–D⇨G➔C.

ix

x

Preface

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 numerals are presented in capital letters regardless of a chord’s quality, 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 A 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. If the chordal fifth, seventh, or ninth is altered, the analytical symbol will incorporate the 5♮ corresponding Arabic numeral, as in C Minor: II ♯ . (Arrow notation – here II➔ – offers an attractive, though less precise, alternative to the complete analytical symbol.) The bullet symbol (•) indicates an absent root. For example, B-D-F in C Major will be analyzed as V•7 (or, with less precision, as V➔). 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 that in the context ———— under discussion occur at the same moment; for example, C| F♯ B | E 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 assertion of a voice-leading phenomenon (for example, when the 6 phase of a I5–6, as in C-E-G to C-E-A, asserts the harmonic role of VI). Open parentheses 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 harmonies I and IV. When a score’s chordal spellings do not coincide with the structurally appropriate spellings (for example, the substitution of easier-to-read F♯A-C♯ for cumbersome G♭-Bº-D♭), I generally will use the structurally appropriate spellings in my examples and commentaries, often placing the enharmonic spellings within square brackets to assist readers in locating the pitches in question within the score.

Preface

I pay very close attention to hierarchies among pitches and chords. To alert readers to various hierarchical relationships I often will underline some pitch names to indicate their hierarchical prominence. For example, CB C above bass C–G–C conveys the relationship between two unfolded strands: a more prominent outer strand E>D>C, and a subordinate inner strand C>B^2, ^3>^2>^1. Because a single harmony sometimes will span multiple measures, the melodic presentation of ^3 or ^2 might not coincide with the onset of its supporting chord. For example, a gradual arpeggiation up to ^3 might be a salient feature of an initial tonic harmony’s presentation. Though a parallel period may serve as an independent musical entity, such a construction often serves as a component of a broader musical composition. The parallel period featured in audio example 1.4 constitutes the first part of a song, here preceded by a six-measure introduction. The dominant’s dissonant impact is here heightened through the incorporation of the chordal seventh and ninth. As preparation for your listening, sing the following alternation between tonic and dominant chordal arpeggiations: F♯ C♯


F>E♭ fourth, borrowing the chordal support of measures 28 and 29 (12.3b, second model), a context that fortifies Kopfton E♭ prior to the descent to ^1. Yet having already proven that he is adept at ^5 >^4>^3>^2>^1 (as in 12.3a), Mendelssohn playfully reverses course during the coda by inverting that conventional descending fifth from ^5 into an ascending fourth: E♭E>E – compensating for its earlier omission. The emblematic G♯>F♯>E third-progression of measures 36 through 40 and its reiteration during measure 42 make G♯ the only viable choice for Kopfton. By withholding the 5-phase tonic onset Schumann invites listeners to contemplate a sense of the infinite, as projected by the word Himmel (heaven) in measure 9. The text’s heaven-to-earth trajectory is suggested by the E>B>E>B>E>B bass of measures 10 through 12. In this context (weighted metrically so that the E pitches are strong and the B pitches are weak), the tonic chord alternates between its foundational 53 state and 64 unfurlings. The pitch B is not asserted as a dominant root until the phrase’s final downbeat (where the downward cascade in the bass finally subsides). The four-time presentation of similar material (most fully realized as displayed in 13.2) during A1 enhances this sense of the infinite. Even the introduction participates in the ongoing repetitive agenda: its II–V succession relates to that of measures 8 and 9. (Note how an F♯>E>D♯ third is featured in both contexts: in the soprano during the introduction, in the bass during the stanza.) In experiencing the work, one perceives how the introduction’s foundational II–V expands dynamically, enhanced at the onset by a VI➔ surge targeting II and in the continuation by the progression’s onward journey from V➔ to I, which arrives on a hypermetric downbeat.3 A proposal for how the introduction’s structure may have developed in Schumann’s mind is conveyed in 13.3. Model 1 shows the direct connection of II and V, incorporating one passing note (E). Two further enhancements emerge in Model 2: the shift from A to A♯ against F♯>E above, thereby igniting a II➔ surge; and the juxtaposition of the supertonic chord’s first-inversion and root-position

Schumann: Three songs from Liederkreis (op. 39)

configurations, thereby introducing melodic thirds A>F♯ and C♯>A♯. In Model 3 those thirds are filled in by passing notes, in coordination with the early introduction of E above. Though the three members of E Major’s tonic triad sound concurrently in this model, the context does not support their analytical interpretation as an asserted I. Finally, in Model 4 the role of the initiating high C♯ as an interior chord member temporarily sounding at the top of the texture is clarified, while the filled-in thirds now feature chromatic pitches. In fact, the lower line (from A to F♯) fails to reach its goal during the supertonic prolongation: though G♮ might have descended to F♯ prior to II➔’s resolution to V, it in fact does not do so. Consequently a more complex evolution of the surging supertonic sounds in place of the tamer choice proposed in Models 2 and 3. By this means the dominant arrival results from the resolution of three tendency notes: an upward resolution of the chordal third (A♯ to B) and downward resolutions of the chordal seventh and lowered ninth (E to D♯ and G♮ to F♯).

Example 13.3 Genesis of Schumann: “Mondnacht” from Liederkreis (op. 39, no. 5), mm. 1–3.

Before considering Burkhart’s harmonic perspective in detail, some attention to metrical issues is in order. My 13.2 displays eight measures of content, which in my view constitute two four-bar

149

150

Harmony in Mendelssohn and Schumann

hypermeasures, notably with a tonic arrival on a hypermetric downbeat (401). Comparing that reading with the diagram Burkhart presents on page 148, it is apparent that our views conflict: whereas I display 36–43, he indicates the partitioning 37–44. (Likewise with the preceding three sets of measure numbers; the two sets that follow will be assessed later.) In my reading the piano’s C♯ just after the onset of measure 6 helps shape a chord, prolonged for nearly two measures, that inaugurates the first phrase. Bass E♯ from 71 “belongs” at 61, as in fact occurs eight measures later – at 141, the onset of the second phrase. Burkhart instead explicitly assigns measure 14 to the first phrase, declaring it to be an “anticipation of the start of the second phrase” in measure 15 (p. 149). The divergence in our metrical readings helps explain how two seasoned Schenkerian analysts could propose such wildly different harmonic perspectives regarding this work. Given his partitioning, Burkhart may with some justification propose that the C♯ 65 chord of measure 14 is an offshoot of the “B major chord begun in measure 13” (p. 149), whereas in the context of my partitioning it is a fresh beginning that has nothing to do with the preceding B dominant. Instead its B, its implied G♯ (compare with measure 29), and its raised E suggest a derivation from the E-G♯-B tonic triad. In my score I have placed vertical lines just after the first sixteenth-rest or -note of measures 1, 3, and 6, thereby partitioning the introduction and first phrase into four components that achieve progressively higher levels of complexity:

I6 (= VI➔)

II II II

V V V V7

I

...

From this perspective, the V chord that concludes the first three trajectories and the VI➔ chord (Burkhart’s C♯ 65 chord) that initiates the fourth are unrelated. (The next trajectory, corresponding to the second phrase, begins on – rather than just after – the downbeat in measure 14.) Beyond this, Burkhart appears not to have taken into account the fact that the phrase’s evolution persists right through its fourth statement, beginning in measure 36. My ear focuses on soprano F♯ during

Schumann: Three songs from Liederkreis (op. 39)

measures 1–2 and 3–4 and F♯>E during measures 8–11, 16–19, and 30– 33. Given that the work is in E Major, those pitches seem curiously fragmentary. Consequently I am delighted by the high G♯ that emerges in the piano during measure 36. Now the F♯>E second may be understood within the broader context of a G♯>F♯>E third (measures 36–41). (That is why my 13.2 presents an analysis of the fourth phrase, rather than one of the earlier ones, as Burkhart does.) Alas, Schumann’s fillingin of the GE ♯ third does not coordinate well with Burkhart’s conception of a broad dominant prolongation or his reading of the Kopfton as ^5 . My contrasting tonic prolongation and ^3 Kopfton project a structure much more in line with normative tonal conventions. If my reading is to prevail, Burkhart would need to retract a number of potent words that appear at the onset of his essay: “drastic,” “strange,” “less natural,” “subverting,” and “unusual.” He applies these words to Schumann’s composition. I instead would apply them to Burkhart’s analysis. My revisions to the part of Burkhart’s ex. 2 that corresponds to my A1 would focus on the harmonic impact of measures 10–12 (tonic, not dominant) and the continuation from the II of measure 8 (to the V of measure 9, not to that of measure 13). Whereas I endorse the succession from GE ♯ to FD♯♯ leading up to measure 13, I regard the G♯ as a pitch prolonged over seven of the phrase’s eight measures, explicitly presented (in two registers) during the fourth phrase, measures 36 through 42. Burkhart’s interpretation of the introduction corresponds for the most part to what I display in 13.3. My most urgent request for revision would be to remove the ^5 that annotates measure 5’s soprano B in his exx. 5b and 5c. Instead F♯ is the passage’s most prominent melodic pitch. A corresponding F♯ will be incorporated within a G♯>F♯>E third-progression once the A1 phrase has sufficiently evolved. Whereas the arrow in measures 1/3 of his example 5c correctly displays the high C♯ as a note that belongs below F♯, the B of measures 3/5 likewise should be interpreted as interior to F♯ (prolonged by means of the F♯>E>D♯ third-progression).

“Mondnacht”: stanza 3 and coda (measures 44–68) Returning to 13.1, one may observe how the set-up for continuation established as early as measure 13 ultimately is carried out. The restoration of Kopfton ^3 from above (as B>A>G♯) inaugurates the A2 section. The critical

151

152

Harmony in Mendelssohn and Schumann

factor thereafter is that the progression must proceed beyond the dominant (supporting ^2) to a tonic that coordinates with the continuing downward melodic trajectory to goal ^1. Because Schumann is working within what at first will appear to be a four-bar hypermeasure (measures 56 through 59, ultimately extended to 61), the sounding of ^2 shifts from the downbeat of the fourth measure (as in measure 43) to the last beat of the third measure (58), thus making the fourth downbeat available for ^1. Despite the melody’s striving for an on-time arrival, the bass B>A>G♯ (a reprise of the descending third that reinstated ^3 at the onset of A2) during measures 58 and 59 complicates matters, resulting in an expansion of the final tonic from one to three measures. (This structure, reminiscent of final cadences by J. S. Bach, is displayed in 13.4. The A neighbor of G♯ in measure 60 echoes that of the Kopfton in measure 51.) Example 13.4 Analysis of Schumann: “Mondnacht” from Liederkreis (op. 39, no. 5), mm. 47–61.

The ascending bass trajectory from I to V may be divided into a second plus a fourth in either order. The two progressions I5–6 II V and I IV5–6 V both get the job done, and both deploy (in E Major) an F♯-A-C♯ chord immediately preceding V. Whereas the former route is pursued during A1 (with, as 13.2 shows, the surge of an asserted VI➔ serving as I6), Schumann shifts to the latter route during A2 (where, as 13.4 shows, a surge transpires as I➔ targets IV). By this means the latter part of the tonic prolongation (commencing in measure 53) still corresponds to what happened earlier, while concurrently the section offers some novelty, especially welcome after the uncommonly repetitive A1. Despite this change in harmonic course, again the progression lacks an initial consonant tonic chord (E-G♯-B): a surge already is underway at the onset of the tonic restoration

Schumann: Three songs from Liederkreis (op. 39)

(with D♮ at 471 serving as I➔’s minor seventh). Because IV replaces II as the principal intermediary between I and V, the middleground G♯>F♯>E third-progression of A1 is replaced by an upper-neighbor embellishment of ^3, since A – and not F♯ – is a member of IV. That G♯G♯ trajectory is covered by a coordinating BB motion, so that the pitch A is not overemphasized (in both outer voices) at the onset of IV. The brief coda, which begins just after the tonic resolution at 611, projects echoes of Kopfton G♯, along with reiterations of the B>A>G♯ line that reinstated G♯ at the onset of A2. My reading of the hypermetric units within A2 likewise contrasts that presented in Burkhart’s diagram on page 148. Through comparison with measure 10 and its replicates during A1, measure 56 should be interpreted as a hypermetric downbeat (with the unit extending through measure 61, resulting from the addition of two measures to accommodate the embellishment of the final tonic). Acknowledging that the bass F♯>E>D♯ of measures 8 and 9 expands to three measures (53 through 55) would support the proposal that measures 51 through 55 form an expanded five-bar hypermeasure. In the reading that is emerging, both hypermetric downbeats serve as venues for the introduction of an important harmony (IV in measure 51, I in measure 56). Extending backwards, the preceding four measures (47–50) would form a hypermeasure, again confirmed by the harmonic analysis, which proposes that exactly those four measures correspond to the prolongation of A2’s restored (and surging) tonic. Consequently measures 44 through 47 might be interpreted as a threemeasure extension of A1’s concluding hypermeasure, allowing the V attained at 431 some space to unfold, with the bass fifth B>A>G♯>F♯ leading to tonic root E on the next hypermetric downbeat. Thus I suggest that my partitioning 40–46 I V8–7–––

47–50 51–55 56–61 I–––––––––––––––––––––––––––– (= I––––––––––––––––––––––––V7 I) (= I––––––––➔ IV5–6 V7 I)

should replace Burkhart’s 37–44

45–52

53–60

A comparison of Burkhart’s ex. 2 and my 13.4 reveals our contrasting conceptions. For me, root E at 471 is one of the composition’s five most

153

154

Harmony in Mendelssohn and Schumann

significant bass pitches (warranting inclusion within my sparse 13.1). The chordal dissonance at that point relates to events closer to the musical surface; foundationally the chord represents a consonant E-G♯-B tonic triad.4 Burkhart instead treats the E chord as a local embellishment of an A that falls within a B>A>G>F♯ fourth connecting the root of V and a (II) that comes between that preceding and a following V chord. (Note the broad B-to-B slur in the bass of Burkhart’s exx. 2a and 2c.) Though my interpretation likewise relates the V harmony that concludes each phrase within A1 and the V harmony of measure 58, it is a relationship of an entirely different sort: as the dominant attained in each half of an interruption-generated binary form. Within that perspective, the latter dominant serves as the internal component of an E–B–E bass arpeggiation, as conveyed in 13.1 and fleshed out in 13.4. As stated above, the IV of measure 51, targeted by the tonic’s surge, emerges on a hypermetric downbeat. A descent in thirds transpires in the bass thereafter: A > G♯> F♯ > E > D♯. Burkhart’s reading focuses on a dominant-prolonging B-to-F♯ fourth (slurred in his exx. 2a and 2c), whereas mine focuses on this A-to-D♯ fifth, which coordinates with the conventional pursuit of a IV5–6 V harmonic succession, whose V in this instance is built above the chordal third D♯, rather than above root B. The fact that both Burkhart and I propose A (which he labels as ^4 in his examples 2a and 2c and I display as a prolongation of the A neighbor (N) introduced at 511) as the principal melodic pitch for measure 53 does not bring our two views closer, since our readings of the voice leading are contrasting. Whereas he displays the B of measure 52 as a passing note between C♯ and A, I interpret it as a local upper neighbor of A, below a prolonged C♯ that persists until the descent to B at 551.

“Schöne Fremde”: stanza 1 (measures 1–7) As in “Mondnacht,” Schumann uses interruption to shape a binary form in “Schöne Fremde”: each of the first two stanzas proceeds from I to V (thus resulting in two iterations of the A1 section), followed by the third stanza (in which the text responds to the question posed during the second stanza), which proceeds along a somewhat different route (featuring IV rather than II➔ as the principal intermediary between I and V) to reach a tonic close. Compare this structure (13.5) with that of “Mondnacht” (13.1).

Schumann: Three songs from Liederkreis (op. 39) Example 13.5 Analysis of Schumann: “Schöne Fremde” from Liederkreis (op. 39, no. 6), mm. 1–24.

The B Major tonic root emerges at 41, bringing to a close a circular trajectory (D♯ G♯➔ C♯ F♯➔ B) that evokes the movement of treetops, likened by the text to the gods making their rounds. (See 13.6a.) Though beginning a composition on the tonic’s upper-third chord is not common, the second pass through A1 (to be explored below) offers a clarifying modification, while Kopfton ^3 (= D♯) is suitably introduced at 13. Two preliminary approaches to the dominant (deploying a descending parallel progression of 63 chords, as shown in 13.6b) transpire during the span from 42 through 61. Then a more definitive dominant arrival is attained by way of II➔ between 62 and 71. Whereas the trajectory of 13.6b descends into the texture’s interior without engaging the higher strands emanating from the tonic’s D♯ (the Kopfton) and B, Schumann’s alternative trajectory accounts for both: the D♯>C♯>B third-progression of measures 1 through 4 is complemented by a C♯>B>A♯ third-progression over the course of II➔ V. The alliance between D♯ and C♯ is reinforced by the reinstatement of D♯ during 62. Observe how 13.6b’s stepwise fourths B>A♯>G♯>F♯ and D♯>C♯>B>A♯ both recur in 13.6a, now exuding a hierarchical shape – as B>A♯>G♯, F♯ (in the tenor register) and as D♯, C♯>B>A♯ (in the soprano register) – due to the assertion of II➔ between I and V.

155

156

Harmony in Mendelssohn and Schumann Example 13.6 Schumann: “Schöne Fremde” from Liederkreis (op. 39, no. 6) (a) analysis 6 of mm. 1–7; (b) parallel 3 chords in measures 4 and 5. (a)

(b)

Though Burkhart applies a I numeral to the B-D♯-F♯ chord at 41, in his conception it serves as what would be referred to in my terminology as an unfurled 64 embellishment of a preceding dominant. (Note the numbers 4 and 6 above the I numeral in his ex. 8b.) I instead grant this B-D♯F♯ the status of a foundational tonic (the onset of the Ursatz). My argument against Burkhart’s reading will sidestep measure 4 itself, focusing instead on what seems to me a misguided attempt to read measures 1 through 3 as a prolonged V. (Observe how only F♯, A♯, and E and only Roman numeral V appear during that span in his ex. 8b.) Burkhart develops a context for an A♯>G♯>F♯ third spanning measures 1 through 3. (This third appears under a slur, annotated “3rd ↆ,” in his ex. 8b.) That conception cuts off the crucial arrival point at 41, which is an essential component of my reading, as follows: m.

1

2

A♯

G♯

D♯

B♯

5

3

4

G♯

F♯

F♯

C♯

A♯

B

5

5

Schumann: Three songs from Liederkreis (op. 39)

The analytical tug-of-war should be between the two endpoint chords: D♯ minor and B major. If the former prevails, then the upper A♯>F♯ third-progression is the “leading” line (a filling-in of the D♯ triad’s upper third), whereas if the latter prevails, the lower D♯>B third-progression prevails (a filling-in of the B triad’s lower third). I contend that the latter chord overpowers the former (the song is in B Major, after all), and thus the A♯>F♯ span above simply “follows” the trajectory forged by the lower line in upper fifths. In this context the A♯-to-F♯ third is not a meaningful interval at a deep level. Burkhart takes an opposing view: “phrase 1 rapidly traverses the third a♯1–f♯1 (measures 1–3), leaving, as it were, nowhere to go except back to a♯1, a move finally made at the phrase’s end (measure 7)” (p. 159). Thus his conception seems wrong to me on two counts: first, his assumption that A♯>F♯ must lead (rather than follow); and second, that A♯>F♯ must be interpreted as a third-to-root span of the F♯ dominant, rather than as a fifth-to-third span of the D♯ mediant. In addition, through comparison with my 13.6a it would seem that he is focusing on events of the chordal interior, neglecting the prominent D♯>C♯ trajectory above, which, though a straightforward and often encountered foundational structure for the A1 section of a composition, conflicts with his dominant-prolongation conjecture. The parenthesized bass F♯ (representing the dominant root) at the onset of Burkhart’s ex. 8c suggests that the dominant is the beginning of something, whereas for me it serves instead as the end of something. As also was the case in “Mondnacht,” multiple dominant-root arrival points dot the tonal landscape. The first two are followed by a new beginning. Though less intensely in “Schöne Fremde” than in “Mondnacht” (which even begins with a dominant “arrival” followed by a new beginning in measure 1), repeated approaches to the dominant are presented, with arrivals at 71 (adumbrated in measures 5 and 6), 151, and 231 (which, as the principal dominant within A2, resolves to I). The fact that the dominant is still in force at 74 (where the first word of the second stanza emerges) and during measures 16|17 and 18 (the onset of the third stanza) does not make the dominant the structural starting point of those latter stanzas. It appears that Burkhart is using those fuzzy boundaries to justify what seems to me a bizarre imagining of dominant harmony at the song’s onset.

157

158

Harmony in Mendelssohn and Schumann

“Schöne Fremde”: stanza 2 (measures 7|8–16) In “Mondnacht” a B>A>G♯ third-progression in E Major facilitates the restoration of Kopfton ^3 for the A2 section. (See 13.1.) In “Schöne Fremde” an equivalent line (F♯>E>D♯ in B Major) is deployed in the bass of measures 7 and 8 to inaugurate the second, somewhat altered presentation of A1. One might regard that D♯ as the actual sounding of the Kopfton (in a novel register); or, one might add a D♯ parenthetically in the register where it first occurred during measure 1 (and where ^2 will emerge during measure 18). The graph presented in 13.7 is left in a state of incompletion at that point, allowing you to gravitate toward whichever course you prefer. Example 13.7 Analysis of Schumann: “Schöne Fremde” from Liederkreis (op. 39, no. 6), mm. 8–18.

One may note with delight that two features from the first presentation of A1 are integrated in an unexpected and beautiful way during this second presentation. I propose that a principal reason why the Kopfton is not dynamically presented in the obligatory register in the vicinity of measure 8 is that Schumann was intent upon projecting the interior B>A♯>G♯>F♯ fourth-progression (featured in 13.6b) now in a tonic context. (The initiating chord at 81–2 is a broadening of that at 42. The descending fourth that initially takes place over four beats now fills over six measures and, notably, is followed immediately by a restoration of a four-beat presentation, from 142 through 151.) To support that span, he expands the five-chord circular progression of measures 1 through 4 into a full eight-chord traversal: from tonic B through tonic B (as B➔ E A♯➔ D♯ G♯➔ C♯ F♯➔ B). Because the pattern, as introduced in the song’s opening measures, requires every

Schumann: Three songs from Liederkreis (op. 39)

other chord to be surging, the initial B tonic possesses a dissonant minor seventh, A♮. Though again there is a danger that the stepwise descent to F♯ might leave the upper strands in the lurch, Schumann is careful to reinstate B (at 142) after the fourth-progression has run its course. B’s successor A♯ and Kopfton D♯’s successor C♯ belatedly emerge during measure 18, after A2 commences. (The dominant harmony of A1 persists beyond the boundary between A1 and A2, with the tonic harmony postponed until measure 19 and its root, B, in the bass until measure 20.) During measure 14 (and also measure 15) the II➔ harmony, spelled as C♯-E♯-G♯-B during measure 6, sounds in the more highly evolved state E♯-G♯-B-D♮. This supertonic occurs above a retained tonic root and fifth, FB ♯ . Thus I and II➔ collide, as indicated by the bracket in 13.7. Though the Urlinie’s D♯>C♯ is not conveyed here with the same prominence as during the first presentation of A1, now that line features a chromatic element in its foreground realization, as D♯>D♮>C♯ (discernible in the score just above Middle C between 142 and 151). Though I hear a tonic restoration at 81, I acknowledge that some listeners might reasonably propose a broad descent by step from dominant root F♯ at 71 through E (83), D♯ (103), and C♯ (123) to tonic root B at 141. On p. 163, Burkhart describes what transpires here as “prolonging the opening V chord by means of a complete journey through the diatonic circle of fifths.” His “complete journey” is from F♯ to F♯ though the “diatonic” key is B. In my 13.7 the “complete journey” coincides with the key: B to B. His melodic “falling fifth” – C♯ to F♯ – conflicts with the motivic fourth – B to F♯ – that my reading highlights. My broken tie from B at 81 to B at 142 is not viable in the context of his dominantprolongation reading. Again Roman numeral I emerges in the midst of Burkhart’s broader V prolongation (at 141 in his ex. 8b). I instead interpret V as the end of something (the close of the first statement of A1, with interruption at ^2), followed by a new beginning on I. Though he labels the C♯-E♯-G♯-B chord of measure 6 as II♯, the similarly spelled and correspondingly deployed E♯-G♯-B-D♮ chord of measure 14 is acknowledged by noteheads in his graphs and by Arabic ♯7 and ♮6 (ex. 8b), but not by a Roman numeral. Is the surging supertonic somehow suppressed during this second statement of A1?

159

160

Harmony in Mendelssohn and Schumann

“Schöne Fremde”: stanza 3 and coda (measures 16|17–30) Whereas the local tonic prolongations during the two halves of “Mondnacht” shift from a deployment of I5–6 II to that of I➔ IV, the principal progressions (incorporating the structurally deepest dominants) in “Schöne Fremde” shift from the use of II➔ during A1 to IV5–6 during A2. Consequently it is not surprising that the tonic restoration of measures 19 and 20 is not a consonant I but instead a surging I➔. (That is, the tonic’s pre-surge stage has been elided. Compare with “Mondnacht,” measures 47–50.) Again the Kopfton is not strongly articulated at the onset. (Note its placement within parentheses during 13.8.) Despite IV’s deployment, its 6-phase evolution is such that the same highly evolved surging II➔ chord that emerged during measure 14 recurs at 224 (E♯-G♯-B-D♮). (The text’s evocation of happiness is underscored by the fact that I➔ is surging from the onset, IV is of major quality, and IV’s 6-phase II evolves into surging II➔.) This time the dominant that follows resolves to I, resulting in a PAC. The coda that follows is notable for the recurring sounding of a D♯ that echoes the earlier Kopfton. Even the bass D♯ of measure 8, proposed above as a low-register representative of the Kopfton, is acknowledged by the coda’s Kopfton reminiscence at 293.

Example 13.8 Analysis of Schumann: “Schöne Fremde” from Liederkreis (op. 39, no. 6), mm. 20–24.

Schumann: Three songs from Liederkreis (op. 39)

Given how my perspective regarding the tonic has conflicted with Burkhart’s in the two statements of A1, we do not need to belabor the fact that, in his ex. 8b, I endorse the Roman numeral I at measure 19 but reject the broad slur that connects the V numerals of measures 17 and 23. In a point of agreement, we both propose the onset of that tonic as a consonant chord prior to the emergence of seventh A♮, even if in the score (measure 19) that initial consonant state is elided. As a Schenkerian analyst, I would be very concerned if I had shaped my reading of a work using Kopfton ^5 but then encountered a coda that emphasizes ^3 (at 243, 263, 283, and 293, all omitted from Burkhart’s graphs). I wonder if this bothered him at all – if he at any point considered how an analysis with Kopfton ^3 might transpire. Nor am I comfortable with the path of descent from ^5 that he proposes. In my reading of the work, there is no direct voice-leading connection between F♯ and E, as his ex. 8a proposes. The score conveys quite clearly that F♯ instead ascends to G♯ (in two registers). My (D♯)E thirds in 13.8 assert that two distinct strands are at play, with the lower strand ((D♯)B>A♯ line, at whose conclusion C♯ is reinstated at 233, resembles the upper-staff structure of 63 through 72, displayed in 13.6a. Beyond that, I struggle with consistencies in Burkhart’s interpretation of measures 21 and 22. In my view measure 21 projects IV, targeted by the preceding surging tonic, while measure 22 projects IV’s 6-phase chord, here asserted as II and evolving into a surging II➔ (C♯-E-G♯ into E♯-G♯B-D♮). At some points Burkhart seems to concur with this reading: the half-note bass C♯ in his ex. 8b suggests IV5–6, even if IV is displayed without further annotation. Jumping to ex. 8a, I discover the “5 –(6)” that might have been included beside IV in ex. 8b. (The parentheses around the 6 probably are intended to acknowledge that the 6-phase chord has been unfurled (my term), thus preventing a literal sounding of sixth C♯ above bass E.) Yet at other points it seems Burkhart is proposing a contrasting reading: just after the dotted bar line in his ex. 9, one encounters a provocative slur covering a stepwise fourth descending from E (with upward stem) to B (with downward stem). Likewise

161

162

Harmony in Mendelssohn and Schumann

Burkhart’s ex. 8c displays the annotation 5 – 6 – ♮7 during these measures, thereby not allowing the 6-phase chord to be established as the 5-phase chord’s successor, but instead showing a local chromatic passing note E♯ emerging in the bass against what might be interpreted as IV8–♮7 (with the filled-in E-to-B fourth, G♯, and minor seventh D♮). If 5–6 in fact is what Burkhart intends, then the E>D♯>C♯>B slur of ex. 9 is untenable, since by that point C♯ would outrank B hierarchically (that is, B would be a passing note connecting C♯ and the dominant’s A♯ that follows).

“Zwielicht”: introduction and stanzas 1 and 2 (measures 1–23) Twilight (Zwielicht) is a period of indistinctness, when the paucity of atmospheric light makes visual perception challenging. One may need to take a second or third look before discerning what visual field lies before one’s eyes. Schumann makes a sonic correlation with that visual phenomenon by only gradually allowing the song’s foundational structure to reach its full luster. Though what transpires during measures 1 and 3 from the introduction could reasonably be interpreted as conveying chords rooted on tonic E, that tentative perception is not solidified until later: by bass E during measure 8 from the first stanza, and by bass E during measures 16 and 18 from the second stanza. Given that foundation, the binary structure displayed in 13.9 ultimately is fleshed out over the course of the song. Its similarities to 13.1 and 13.5 are obvious. Example 13.9 Analysis of Schumann: “Zwielicht” from Liederkreis (op. 39, no. 10), mm. 8–41.

Schumann: Three songs from Liederkreis (op. 39)

The focused E>B>E bass of measures 16 through 18, against the melody’s GF♯>E third-progression, here G>F♯>F♮ substitutes, so that minor ninth F♮ sounds within the surging tonic chord (where concurrently G♯ displaces G en route to upper neighbor A). The D (rather than D♯) deployed during the interior embellishing chord anticipates the D of that I➔. The surge’s dissonant focus helps the listener to perceive GE (measures 13 and 21), and D♯B and A♯C>A filled in by passing notes) reaches IV’s root at 51, 6-phase F♯ has emerged above it. From this point a fresh problem takes hold (problematic in comparison with the well-behaved structure projected during the first two stanzas that follow). We noted above how the melody’s unfolded AF♯ third is followed by two additional thirds – GE and FD♯♯ – resulting in the sounding of ^2 exactly

Schumann: Three songs from Liederkreis (op. 39)

where it “should” sound: in the vocal melody above dominant root B at the HC (measures 15 and 23). Yet in the introduction the G and F♯ instead take up positions in the texture’s interior (during measure 6). Though the V♯ at 63 may be accepted as resolving the preceding II⇨, one needs to forgive both the absence of soprano F♯ and the tardiness of bass B. It appears that Schumann has used these inherent defects to motivate a quick reiteration. That dominant is followed immediately by a renewed thrust from the tonic to the subdominant (I➔ IV), and upon the arrival of IV’s 6 phase, upper neighbor A is restored in the soprano (now an octave lower), after which the G and F♯ resound within the melody. Schumann adds one further creative wrinkle as this reiteration proceeds: the subdominant 6-phase chord substitutes chromatic F♮ for diatonic F♯ during 72. A further nuance of Schumann’s writing is documented in 13.11. The left model shows the evolution of IV’s 6-phase chord into a supersurge. The D♯ from the end of measure 12 is not displayed, since that pitch serves as a local neighbor to chord member E (which sounds after but not before the D♯, acknowledged by the parentheses around E in the model). Four of II⇨’s chord members sound during the first beat of measure 13. In the right model, D♯ is projected more prominently, though its chord is displayed using filled-in noteheads to indicate a subordinate layer of structure. The sonority at that point has long been recognized as what is called a common-tone diminished seventh chord.6 Locally it resolves into the C-EG triad that follows. Yet that vignette occurs as such only because the A♯ that normally would follow directly after A (as on the lower staff of the left model) is withheld temporarily. In such a context, the C major triad plays no independent harmonic role, but instead projects the fifth, seventh, and ninth of the emerging II⇨ harmony. The measure numbers indicate that this idiosyncratic realization of a conventional supertonic evolution takes place both during the introduction and during the fourth stanza. Example 13.11 Schumann: “Zwielicht” from Liederkreis (op. 39, no. 10), relationship of mm. 4–6 and of mm. 36–40 to mm. 11–15.

165

166

Harmony in Mendelssohn and Schumann

In broad terms Ferris and I share the same conception regarding the initial stages of the song’s projection: “The song is based on a musical phrase that is not clearly defined when we first hear it, gradually comes into focus as it is varied in each succeeding stanza, and ultimately turns out to be something quite different than what we had first imagined” (p. 140). Though I will offer a contrasting interpretation of the “ultimately turns out” phase (regarding events of the third and fourth stanzas as aberrational rather than as culminating), we agree on the path toward increasing clarity through the second stanza. Ferris suggests that the introduction lacks a “coherent harmonic progression until the cadential dominant at the very end” (p. 143). His block-chord reduction of the musical content (ex. 2a on p. 145) is annotated by question marks in place of Roman numerals until the V at 73. Though I agree that how the A chord of measure 4 is attained is open to question, my ear more quickly establishes E Minor as a tonal center, through the A^4 above the system, whereas an arrow is deployed to convey that the high F♯ is actually an “alto” pitch temporarily transferred to the top of the texture. The dominant arrival point in measures 2 and 4 comes across as a dissonant chord whose resolution is stymied. (Augmented fourth EB ♯ persists in a suspenseful state but does not resolve.) Over the short term, Schumann gets around this dilemma by traversing a contrasting path once the voice enters (as will be explored below). Yet he eventually reaches the same V♯ 37 stymie point – in measures 13 and 15 and again in measures 24 and 26. After the former, the same solution as occurred at the juncture between the introduction and the first stanza is applied again. After the latter, a contrasting path (in the cycle’s next song) is substituted in order to get beyond the impasse. Stein’s essay on musical ambiguity begins with an unintended analytical ambiguity: the very first Roman numeral annotating the score provided in her ex. 7.1 reads as ii6 (measure 1, beat 2), whereas the textual commentary addressing that passage instead indicates iv6 (page 78). Looking for how recurrences of the chord in the same context are analyzed later in the work is of no help in resolving the discrepancy, since those chords are not annotated by a Roman numeral in her presentation.4 Despite this confusion, I suggest that Stein intends for the D-F♯-B chord of measure 1 to be understood as iv6 in F♯ Minor, not ii6 in A Major (a key that arrives later in her analysis).5 If that is the case, her reading is in accord with mine in 14.1, taking into account the inherent differences in our styles of analytical notation. My imaginative interactions with the score contrast Stein’s literalist approach. For me, measure 1’s first beat is an exceedingly important venue for analytical endeavor. By definition a suspension is a pitch that once was meaningful in a way that is thwarted by its prolongation into a new context. For me the concurrent sounding of A♯ and C♯ where they clearly do not fit as chord members triggers an analytical backtracking to reconstruct a context (prior to the first downbeat) in which they might have thrived.6 In contrast, Stein’s first Roman numeral (intended as iv) does not appear until the suspensions have about run their course, on beat 2 of measure 1, rather than at the point where the subdominant harmony’s assertion is initiated, at beat 1.

Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1)

The first stanza through the arrival of the subdominant (measures 4|5–10) An appealing way for a composer to correlate a song’s introduction and its first stanza is to shape the former as a miniature version of the latter, deploying the same foundational harmonic progression in both contexts. Note how the IV of the introduction’s measures 1 and 3 holds forth in the first stanza’s measure 10, and how the V♯37 of measures 2 and 4 recurs in measures 13 and 15. Unless something emerges later to challenge the assertion, let us proceed under the hypothesis that a I–IV–V♯37 harmonic progression in F♯ Minor fulfills such a double duty. Perhaps because the initial tonic is only hinted at rather than robustly asserted, Schumann does not backtrack to that tonic at the dividing point between the introduction and the first stanza. The root progression works as follows, with the ellipses conveying crucially that each B chord should be understood as following after an initial F♯ tonic: Introduction F♯

➔

B

C♯

... B

C♯

... B ... B

E➔ E➔

(before the composition begins through measure 2) (measures 3–4)

Stanza 1 A A

(measures 5–6) (measures 7–8)

After twice proceeding directly to dominant C♯ during the introduction, Schumann supports the opening vocal line by proceeding instead to the mediant (F♯ Minor’s diatonic A-C♯-E).7 That path is explored in 14.2.

Example 14.2 Analysis of Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1), mm. 4|5–8.

175

176

Harmony in Mendelssohn and Schumann

A composer often will deploy several individual chords to prolong one foundational harmony. For example, in the tonicization of, say, a VI harmony, a local progression from the root of VI to its dominant (perhaps incorporating some other chords along the way) and back will serve to dynamically convey the sense of VI within the broader harmonic trajectory. Third-relationships sometimes take on a role similar to fifth-relationships in such initiatives, especially during the nineteenth century. To prolong an F♯ minor tonic harmony, one may proceed from an F♯ minor chord to an A major chord (perhaps incorporating some other chords along the way) and then back to F♯. The progression in 14.2 shows the tonic-to-mediant phase of such a prolongational strategy, wherein the circle of fifths serves as the means of locomotion. The surging F♯ chord is a local event within the circle’s initiating F♯-to-B fifth. It does not convey the key of F♯ Major. (All of the chords in measures 5 through 8 are diatonic in F♯ Minor.) The broader context in which that F♯-to-A trajectory transpires is shown in 14.3. Contrasting the introduction’s bass D at 11, the subdominant of measure 10 emerges in its root position, with bass B, and so Schumann is able to deploy the strongest possible bass motion during the I-to-IV succession: root F♯ to root B. This is accomplished in the context of a surging tonic: after mediant A-C♯-E, the restored tonic of 92 surges from its onset, replacing diatonic A with A♯ and retaining the E introduced by the mediant. This is a vigorous and potent harmonic succession. No longer must the analyst excuse tonic chord members for their truancy, since F♯, A♯, C♯, and E all have reported for duty.8 Example 14.3 Analysis of Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1), mm. 4|5–10.

Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1)

In my interpretation of the song’s harmonic content, the introduction’s I ➔ IV V♯37 progression expands to encompass the entire first stanza. Given the luxuriant framework afforded by that time span, Schumann undertakes a creative double deployment of a B chord. Clearly B-D-F♯ in measures 5 and 7 sounds in a context that contrasts that of the B-D-F♯ in measure 10. Which of these chords fulfills the same role as the B chord of measures 1 and 3, as the principal internal point between the tonic and the dominant? The only viable answer, I think, is the latter B chord, in measure 10. If my proposal of a foundational F♯-A-C♯ tonic chord in F♯ Minor as the predecessor of the first stanza’s content is rejected, then my analysis falls apart completely. But if one posits such a primordial chord, then its means of prolongation and evolution through measure 9 may be conveyed concisely as follows: F♯ F♯

A A A♯

C♯ C♯ C♯

E E

(prior to the stanza) (measure 6, repeated in measure 8) (measure 9)

From this perspective the A-C♯-E chord does not warrant a III label indicating an asserted scale-step along the path to V♯37 , but instead falls entirely within the orbit of the F♯ tonic chord.9 Stein assigns a more prominent role to the A chord, not only shifting her analysis into the key of A Major, but also proposing that the crucial B-D-F♯ chord of measure 10 functions as ii in A Major. My analysis interprets the chordal hierarchy in a much different way, with the F♯ chord prevailing through measure 9, where, in a typical harmonic evolution, it surges just prior to the progression’s conquest of IV. Though Stein shows iv as one of two alternative readings for the chord of measure 10, her analysis goes blank in measure 9. To me that measure’s F♯➔ chord is the culmination of the stanza’s broad tonic prolongation. I interpret the foreground scurrying up to the mediant chord that transpires during measures 5 and 6 as a circular – not a harmonic – progression, warranting letters indicating roots (as in 14.2) rather than Roman numerals (as in Stein’s ex. 7.1). In this delightful moment a witty Schumann allows a B chord like those encountered during the introduction to veer off in a different direction. Though Stein binds small clusters of chords via Roman numerals in her analysis, how the goal A chord correlates with the prior F♯ Minor focus is left unexplored.

177

178

Harmony in Mendelssohn and Schumann

The broad progression of the first stanza (measures 5–15) In the spirit of exploring how the first stanza expands upon the structure that occurs during the introduction, it is reasonable to expect that the IV harmony attained at 101 will undergo some sort of prolongation before the dominant’s arrival, which in fact does not occur until 131. Whereas there are many ways to prolong a IV harmony, in the context of this particular song one option is especially attractive and appropriate: to again proceed to the upper-third chord and back. Schumann does everything possible to emphasize the relationship between those two chords in his song. A connective chord at 91 (shown using filled-in noteheads in 14.3) leads into an F♯➔ tonic chord that targets subdominant B at 101. Then the procedure is repeated a third higher: another connective chord at 111 (again shown using filled-in noteheads, now in 14.4) leads into an A➔ chord that targets the subdominant’s upperthird chord on D at 121. Example 14.4 Analysis of Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1), mm. 10–12.

By this point in the song we have grown accustomed to hearing exact repetitions of content (for example, with measures 3 and 4 repeating measures 1 and 2). Schumann departs from that practice in the connection between the subdominant and the dominant. Whereas the return from upper-third D back to B is entrusted to a mere sixteenth-note B sounded in the bass at the end of measures 12, measure 14 offers a potent confirmation that this return actually has occurred. That upper-third D chord is not reinstated during measure 14. (Both the tonic’s and the subdominant’s upper-third chords are displayed within large parentheses in 14.5.) By this point Schumann has begun to concern himself with another important initiative: making measures 14 and 15 sound like measures 3 and 4, so that the second stanza may commence in a context that exactly matches that of the first.

Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1) Example 14.5 Analysis of Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1), mm. 6–15.

Another notable feature of the expansion concerns the vocal melody. In 14.1 observe how the foundational melodic line ^5 >^4 is complemented by the emergence of a high F♯ (embellished by G♯) in the soprano register at the end of measure 1 (and measure 3). The first stanza’s expansion of IV likewise proceeds melodically upward to F♯ (now with a G♮ embellishment in measure 12, “corrected” to G♯ by the piano later in the measure to correlate exactly with measures 1 and 3). Given the premise that the harmonic progressions of the introduction and the first stanza are foundationally the same, encountering this distinctive trait at both locations instills a confidence that the analysis is attuned to Schumann’s conception. In that all the song’s harmonic content has sounded by measure 15, we may now speculate regarding how Schumann’s chordal progression conveys the sentiments of Heine’s poem. Three factors – two unequivocally positive and one still to be resolved – seem to me central: Winter A loveless state

gives way to gives way to

Spring Love

Reciprocation

is as yet

Uncertain

Though many commentators have noted that the song contains no F♯ minor chord, if Schumann assigns winter and a loveless state to that foundational harmony, then its absence is to be expected: winter has yielded to spring; the former loveless state has given way to a burgeoning love. The imagery of blooming buds and singing birds (at some height above the earth’s surface) both convey a rise above the dormant state. In harmony there are two principal ways to bring a minor triad to life: its parallel major

179

180

Harmony in Mendelssohn and Schumann

(as in F♯-A-C♯ to F♯-A♯-C♯, perhaps with added seventh E), and its relative major (as in F♯-A-C♯ to A-C♯-E). I propose that the song’s three foundational triads each undergo one or both of these transformations: Wintry, loveless chords

Springtime, love-impregnated chords

F♯ minor B minor C♯ minor

F♯ major (with minor seventh)

A major D major

C♯ major (with minor seventh)

The song captures a moment when the love has been revealed though not yet reciprocated. What sort of harmonic progression might best convey this blend of yearning, hope, suspense, and uncertainty? I suggest there is no more suitable choice than to proceed from the tonic through the subdominant to a dominant seventh (V ♯37 ). The lack of resolution in no way indicates a negative outcome. It is simply that our snapshot of the scene occurs at that delicate moment when reciprocation is as yet uncertain. Schumann’s deployment of IV’s upper-third chord matches what transpired in the context of I (now lacking the surge, of course, since IV cannot surge to V♯). The pitches may be displayed as follows: B B

D D D

F♯ F♯ F♯

A

(measure 10) (measure 12) (end of measure 12, confirmed in measure 14)

Sixteenth-note B at the end of measure 12 (repeated at the end of measure 14 in a more overtly subdominant context) reinstates the subdominant root.10 Though Stein has placed a iv numeral below measure 10 (as one of two options) in her ex. 7.1, the reinstatement of iv after the excursion to its upper-third chord is noted neither in measure 12 nor in measure 14. Thus neither of the two foundational harmonic successions that I display in 14.5 (I to IV, deploying a surge; and IV to V♯37 ) has a counterpart in Stein’s analysis of the first stanza. Whereas her summary of the stanza emphasizes a “strong arrival on A major followed by the return to V7 of F♯ minor” (p. 78), I instead would emphasize how an A major chord (not key) helps to prolong the F♯ tonic and how the dominant’s arrival proceeds from IV, as it did during the introduction.11

Schumann: “Im wunderschönen Monat Mai” from Dichterliebe (op. 48, no. 1)

Given my reading of how the harmonic progression conveys the meaning of the poem, I do not sense the “pain of lost love” or “the poet’s ultimate lack of fulfillment” (p. 78) that Stein proposes. Granted, during the early stages of any love relationship one might harbor selfdoubts and worries about rejection. Yet the transformation of all three of the progression’s principal chords through parallel and/or relative relationships into chords of major quality encourages interpretation in an optimistic light. As the song ends, the season is spring and the poet is in love, awaiting the loved one’s response – bracing for the possibility of 12 a No, but full of hope for a Yes.

The second stanza (measures 15|16–26) There is not much to say analytically about the second (final) stanza. The open-ended content of the first stanza is left intact during its repetition. Consequently there is no conventional cadence. The melodic C♯>B second (indicated by ^5 followed by ^4 in 14.5) leaves a dangling dissonance sounding at the movement’s close. This is, of course, an exact parallel to the impasse at the end of the introduction. Schumann responds in a similar way: the mediant, now established as the key of A Major, again comes to the rescue in the next song (“Aus meinen Tränen sprießen”), where a C♯>B>A third like that of 14.2 is traversed multiple times during that song’s opening eight measures (supported by a harmonization that both begins and ends on an A major tonic chord), thereby resolving the impasse. In retrospect, we come to realize that the cycle’s first song is a fragment – a beginning whose ending (if one occurs at all) does not fall within the song’s boundaries. I have resisted expanding this chapter into a broader study of Dichterliebe as a whole.13 What I state above regarding the C♯>B>A third from 14.2 migrating to the second song is tantalizing, as is a look at how that song’s middle section relates to the tonic chord: A A

C♯ C♯ C♯

E E♯ E

G♯ G♮

(measures 0|1 through 8) (measure 12) (end of measures 12 and measure 13)

181

182

Harmony in Mendelssohn and Schumann

Though this reading contradicts one of the most studied analyses in Schenker’s Free Composition (fig. 22b),14 it so clearly relates to how I propose both I and IV are expanded in “Im wunderschönen Monat Mai” that my analytical claims are bolstered significantly. I close leaving these loose ends dangling, awaiting further analytical contextualization, just as Schumann closed his first song with loose ends.15 In retrospect, perhaps Stein’s selection of one song from a cycle was not an ideal way to generate a demonstration analysis intended for undergraduate students.

15

Schumann: Sonata in A Minor for Violin and Piano (op. 105) in response to peter h. smith

Peter H. Smith has been one of the most productive of all tonal analysts during the past few decades. His numerous publications and lectures reveal a special devotion to music by Schubert, Schumann, and Brahms. His route to music theory was via study of the viola at Juilliard. This has resulted in a special emphasis on music for stringed instruments in his analytical writings. Smith explores Schumann’s opus 105 in three publications: an article from 2009, a book chapter from 2011, and another article from 2013.1 Because I dart back and forth among these sources during my commentary, I indicate the publication year before each page or example number cited. Smith proceeds from technical analysis to a consideration of broader issues, especially tonal pairing and the TMS complex (where TMS refers to the tonic, mediant, and submediant). Because my technical analysis contrasts his in many ways – with disagreements not only on some of the finer points but also regarding basic parameters such as form, tonicization, and Kopfton – in many cases my perspective does not affirm the premises upon which his broader assertions are made. My Schumann composes more within the mainstream channels of early- to mid-nineteenth-century tonal practice than does Smith’s Schumann. This chapter invites readers to take a stand on a range of the work’s compositional details. Just how wide a gap separates Schumann’s practice from the music he knew? Do we need fresh analytical notions to deal with his practice? Or can his innovations be accommodated within the analytical framework that has been developed for his immediate predecessors?

Movement 1 exposition: P (measures 1–27) A sonata exposition’s P is charged with introducing the tonic key, generally by traversing the harmonic path from tonic root to dominant root and then back to tonic root in the bass. That arpeggiation usually will support a middleground melodic third- or fifth-progression descending by step from the Kopfton (almost always ^3 or ^5 ) to the tonic pitch, resulting in a PAC. 183

184

Harmony in Mendelssohn and Schumann

The opus 105 first-movement exposition’s P is built from four interconnected, varied traversals of that path, creating points of articulation at measures 5, 9, 16, and 27. (Locate these goals in 15.1, noting that each corresponds to the endpoint of a C>B>A third-progression.) Only the final iteration deploys both root-position dominant and root-position closing tonic chords, as befits its comparative structural depth. Schumann prevents a sense of full closure at the first three articulation points: a C covers the melodic A of measure 5, and the preceding dominant is inverted; and the melodic A at both measures 9 and 16 is supported by an inverted tonic. (The final traversal’s majestic dominant expansion, whose truncation within 15.1 is indicated by a hairpin symbol, will be addressed separately below.) Example 15.1 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1, mm. 1–27.

Though the violin’s first pitch – a Middle C – fulfills the Kopfton role for the time being (note the B and A that follow in that register over the course of measures 1 through 5), the higher pitches that soon emerge arpeggiate the tonic chord in an upward trajectory, ultimately resulting in a registral shift: E (preceded by an appoggiatura F) in measure 1 begins the process, while the piano’s A inaugurating the theme’s second iteration at 61 continues it. Though a higher C is not attained prior to measure 6’s upper-neighbor D, clearly by that point the octave above Middle C has taken over as the principal register

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

for the main melodic initiatives. (The Cs in measures 11–12 and 16 confirm Kopfton C’s ensconcement in that register as P continues.) The violin’s juxtaposition of the tonic triad’s C and E during measure 1 amounts to an inauguration of two distinct melodic strands, which descend together during measures 1 through 5: C>B>A and E>D♯>D>C.2 As is fitting for this early point in the work, the bass does not jump dynamically among harmonic roots but instead shifts by half steps: A>G♯D♯>D>C♯). Now, with that theme deployed in the bass, a AD ♯ diminished fifth resolves to the dominant’s GE ♯ third (measure 19), fulfilling D♯’s ascending tendency. Over the course of an eight-measure expansion, the dominant’s dissonant seventh (D) emerges in the chordal interior and resolves to the tonic chord’s C at 271, as shown in 15.1. Before that D takes hold the dominant undergoes a dynamic expansion that may be broken down analytically into three parts. In 15.2a a descent from B to A in the melody sounds in a context that prevents the A from coming across as the broad fifth-progression’s goal ^1. Instead it connects the preceding B and an interior G♯ that emerges in conjunction with a reinstatement of the dominant harmony during measure 26. (Thus the C>B>A third within each of the earlier tonic prolongations is complemented by a B>A>G♯ third in the context of the dominant.) The intervening F-A-C and D-F-A chords (at 251 and 261, respectively) embellish the dominant. Another phase of the dominant’s prolongation takes an altogether different approach. Whereas 15.2b shows the transfer of an inner-strand G♯ to the top of the texture (above the B>A that appeared on the beam in 15.2a), 15.2c clarifies how that transfer is carried out (the open noteheads) and how those dominant-chord pitches are embellished by upper neighbor embellishments (the filled-in noteheads). A third distinct component of the dominant prolongation is the local D-F-B♭ chord (shown within parentheses in 15.2c) that intervenes before all the dominant’s pitches have fallen back into place. The D that sounds during this stunning and unexpected chord is retained once the dominant is restored, serving as its seventh (the D of measure 24 in 15.2a, not that of measure 26 in 15.1). Though this interpretation projects the D-F-B♭ chord of measure 23 and the D-F-A chord of measure 26 as embellishments of an already established dominant, these conspicuous events do to some extent have the effect of backtracking within the harmonic progression – of a fresh approach to the dominant. Thus by the time this dominant resolves to the tonic (measure 27) it has been paired with three different predecessors: II➔ (measures 17 and 18), ♭II (measure 23), and IV (measure 26).

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

Example 15.2 Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1 (a) analysis of mm. 19–26; (b) foundational voice leading of measures 19–25; (c) detailed voice leading of measures 19–25.

In my presentation above, I segment P into four regions based on the tonic arrival points at measures 5, 9, 16, and 27. My comments on Smith’s analysis (as conveyed principally in his ex. 2009/4) will be organized within those boundaries as well. Measures 1–5: Smith regards the sforzando F of measure 1 as the onset of a foundational neighbor-note motive, with a resolution to E at 42. I instead interpret it as a local neighbor of E within measure 1 (with F>E matched immediately by E>D♯ on the following downbeat and later by D>C at 51). Though that F does not appear in my graph, I would grant that its introduction at the onset readies the sensitive ear for

187

188

Harmony in Mendelssohn and Schumann

similar deployments – for example, in measures 12 (bass) and 25 (violin) – later.5 These contrasting interpretations have resulted in 7

conflicting symbols in our graphs. Whereas I deploy ♯ 5 (which would ♯3

translate into

♯6 ♯4 2

were I to acknowledge the chordal inversion) beside my

II numeral at measure 3, Smith instead proposes 4þ 2 (where his + here is the equivalent of my ♯), implying diatonic 6. Though indeed Schumann remains noncommittal for a while, during the second half of measure 3 an F♯ (♯6 above bass A) sounds thrice. I cannot endorse Smith’s reading of F>E in light of such an obvious conflict.6 The first upper-staff notehead in Smith’s graph is a Middle C, bearing a downward stem. What transpires in that register during the chords that he labels as 4þ 2 and as 6, neither of which accommodates a C? Are we to hear an upward leap of an augmented second to the downward-stemmed D♯? My graph shows a descent from C to B (delayed until 32 in the score due to the suspension of C) in conjunction with II➔’s arrival. That B persists during V➔, with the emergence of A coordinating with the resolution to I. (Both B and A sound at the ends of measure-long descending violin trajectories, following their less pronounced introduction in the piano part.) Though Smith displays an unstemmed B an octave higher at measure 4 and an unstemmed A in the register of the initial C at the tonic resolution (measure 5), he does not connect the dots. That A’s upward transfer for the piano presentation of the opening theme (see the arrow in 15.1) secures the register in which C>B>A will occur repeatedly during the remainder of P. Though the initial C>B>A is set below other content, it is important to acknowledge it here, since it will play a significant role in the higher register as P continues. Measures 5–9: I am in wholehearted agreement with Smith’s beamed line C>B>A during this second iteration of the local I–V♯–I harmonic progression (which is acknowledged via Roman numerals only in its third and fourth traversals in his graph). Yet our treatments of the line differ in two respects. First, I regard it as a reiteration of a C>B>A line already introduced during measures 1 through 5, rather than as the second half of a line emanating from E. Second, I follow Schumann’s explicit projection of the line an octave higher than where Smith displays it. Regarding the latter, Smith’s note 2009/14 clarifies that he has

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

placed “the piano’s ‘covering’ line . . . down an octave” (p. 2009/82), an intervention that I suggest should be rescinded. Given that the pitch B is articulated four times in that upper register (above a persistent E) during measures 7 and 8, I feel justified in interpreting the melody’s E>C>A (above another E) during measure 9 as the introduction of goal A via an arpeggiation from its upper fifth. (Compare with BA at the cadence of measures 26 and 27.) In Smith’s view that A instead is an internal point within an E>C>A>E octave arpeggiation. Because IV – rather than II➔ – now serves as the preparation for V♯ within the harmonic progression, F>E is no longer problematic as a neighboring embellishment. The question thus becomes whether EE (projected by the violin) or C>B>A (projected by the piano) is the defining melodic line. Over the broader course of P, my graph displays focused C>B>A third-progressions consistently, whereas Smith displays static neighbor-embellished E prolongations with equal consistency. It seems to me that if one perceives a PAC (which I regard as normative for P) transpiring at measure 27, then a descent to A is the more compelling reading here. Measures 9–16: Though Smith and I both acknowledge an F major chord at measure 12, for me it serves as a conventional 6-phase expansion of the tonic (A-C-E to A-C-F, here unfurled into F-A-C), whereas for him it is the principal harmony between i6 and V. (Keep in mind that Arabic 6 is being used in contrasting ways in his and my graphs.) I hear bass D (measure 14) as the principal intermediary between the tonic’s C (measure 9) and the dominant’s E (measure 15). Whereas a similar D supported IV during measure 6, now it supports II7, whose seventh, A, is delayed in the melody until the downbeat of measure 15. An unfolding symbol connects bass D and F in both of our graphs. Yet whereas he adds a flagged stem descending from the F (and refrains from labeling the D chord as ii°), I have deployed such a stem from D (and incorporate II7 within my harmonic analysis).7 Smith’s omission of a broad C>B>A melodic line stems in part from his reading of the linear trajectory during measures 9 through 12. Note that a B♭ sounds in the piano part before C emerges (measure 11). This has led me to display a double application of the reaching-over technique in my graph, so that from the pitch A at measure 9 first B♭ and then C emerge, resulting in what is (for me) a reinstatement of the Kopfton.8 The tonic’s CA third (which may be interpreted as soprano and alto

189

190

Harmony in Mendelssohn and Schumann

strands) then proceeds to

B A

during II7 (with the melody’s F>D>B

arpeggiation echoing the E>C>A of measure 9), BG ♯ during V♯, and finally a unison A (surmounted by a fresh C for the next iteration of the cycle) at I. Smith instead explicitly shows the B arriving from below (via a slur connecting A and B in his graph), and he leaves the dominant’s leading tone (upward-stemmed G♯) dangling, despite the piano’s A at 161. Measures 16–27: The three pitches of my C>B>A thirdprogression are all present in Smith’s graph: C and B both sport upward stems (measures 16 and 17), though B’s stem is reversed to downward at the onset of V in measure 19; and A is present above tonic root A in measure 27, though presented unstemmed as an appendage of the E above. Smith again does not connect the dots, proposing “a precipitous plunge to ^1, rather than a generic stepwise descent to closure” (p. 2009/54, emphasis added). In that the visual projection of such a third-progression certainly would trigger questions regarding whether perhaps ^3 (instead of ^5 ) serves as the Kopfton, its omission in Smith’s presentation makes for a more consistent (though in my view faulty) interpretation. Yet if F>E is again to play the principal role, then the problem noted above recurs: his 65 annotation at measure 17 corresponds to a spot that 6

in the score instead would be figured as ♯53 . My sensibility again rebels against the proposal to prolong a neighbor F against a sounding F♯. I display this B-D♯-F♯-A chord as one of the four principal harmonic events of P (as presented in the second row of Roman numerals in 15.1). ♭

Not so for Smith, where the II and VI that precede reiterations of this dominant are not complemented by some sort of harmonic label between the long-prolonged initial tonic and the onset of the dominant. I likewise am curious why VI rather than IV is displayed during measures 25 and 26. Though in my reading F-A-C appears in the same places as in Smith’s (measures 12 and 25), I do not deploy a VI numeral at either location. My proposed B>A>G♯ third-progression over the course of the dominant prolongation (as shown in 15.2a) likewise has no counterpart in Smith’s reading. A downwards-stemmed B is displayed in the alto register in his graph at measure 19 and an upward-stemmed A with flag appears an octave higher at measure 25. But the G♯ that completes the

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

third-progression during measure 26 is missing, despite the sharp in the figured bass indicating its presence in the chord.

Movement 1 exposition: TR ⇨ FS (measures 27–64) Once A Minor’s root-position tonic chord is secured cadentially at 271, thereby concluding P, Schumann shifts from a tonic-establishing to a tonic-departing agenda. In an exposition that juxtaposes A Minor (the key of the movement as a whole) and a C Major tonicization (the venue for the upcoming Fortspinnung – “spinning-out” – that concludes the exposition) in the context of Kopfton ^3, the latter key often will be prolonged by means of a fifth-progression descending from an inner-strand G transferred to the top of the texture, as is conveyed in the foundational model of 15.3. Schumann deploys a segment of the descending circle of fifths to link those two more stable regions. The A chord of P’s cadence is expanded by means of an A–E–A bass arpeggiation (during measures 27 through 34) that supports two instances of reaching-over in the melody, so that E and then G sound at the top of the texture, as shown in 15.4. The latter pitch emerges against the raising of the A tonic’s third to C♯, resulting in a surge: A➔ targeting D. Though D’s seventh, C, is introduced during measure 35, that chord’s minor quality is retained during the succession to a G-rooted chord, which surges from its onset. The expected resolution (following the 7–10 precedent of A➔ D, as marked in the graph) would be G➔ to a rootposition C chord for the onset of the Fortspinnung coinciding with the mediant tonicization. The first sign that what lies ahead will involve many disparities between expectation and compositional realization emerges with that initial C chord: though we expect F>E in the soprano (below a G) and G>C in the bass, that resolution’s E and C trade places in Schumann’s rendering. (Compare the score and 15.4, which projects the normative resolution pitches within parentheses.) Repetitions of the circular progression’s chords (now with a surging D➔) continue through measure 43, finally attaining a C-rooted chord with the expected soprano E and bass C at 432. Yet while resolving one issue, Schumann introduces another: a B♭ sounds along with C, E, and G in this tonic chord. Consequently there is no point in the vicinity of the C Major tonicization’s onset where a root-position, consonant C major chord sounds.

191

192

Harmony in Mendelssohn and Schumann

Example 15.3 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1, foundational model for how an exposition in A Minor might be structured.

Example 15.4 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1, mm. 27–43.

Given the circular progression’s reliance upon surges during its reiterations (with D➔ followed by G➔ leading toward C), adding B♭ to C-E-G at 432 might come across initially as a C➔ surge targeting F, putting into question C’s role as the transition’s goal. Schumann here deftly navigates a shift of context: C’s seventh B♭ ultimately functions instead as 6-phase A’s ninth (with A➔ represented by C♯-E-G-B♭). We may excuse C♯’s tardiness (coinciding with the arrival of bass F at 441 rather than sounding before the bar line, as occurs during 452), since it facilitates Schumann’s ruse. Colliding with the arrival of II’s bass F, this surging VI➔ definitively terminates the transition’s circular progression and ultimately helps confirm C Major as the prevailing key for the remainder of the exposition, as shown in 15.5.

Example 15.5 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1, mm. 38–63.

194

Harmony in Mendelssohn and Schumann

Schumann’s writing during the exposition’s Fortspinnung (FS) is extraordinary because it seems intent upon preventing a full realization of the expected outcome, embracing numerous partial or contorted iterations of a compositional initiative that in a more routine deployment would offer a straightforward means of tonicization. This initiative’s principal component is displayed unadorned in 15.3: a stepwise descent from G down to C. The harmonization of that line might be expected to incorporate C Major’s II, IV, or V7 as support for F; and certainly the harmonization of the concluding D>C would be V(7)–I, both in root position. Consider how Schumann’s fertile imagination and secure mastery of tonality conspire to thwart this scenario repeatedly, as conveyed below. Once all these iterations have transpired, the listener confronts the discomforting fact that the exposition has run its course without projecting an EEC. Measures 42–47: By taking into account the transition’s C chord during 421 (which itself hearkens back to that of 381), one may perceive how a high G sets the stage for a stepwise descent, so that the violin’s E>D during measure 44 comes across as part of an interior strand. What has happened to the G? Schumann goes so far as to prevent the piano G an octave lower from descending to F during measure 44, though this prohibition is lifted in the reiteration two measures later. The graph offered in 15.5 shows an E>D second appearing below a broad set of open parentheses that convey how the G starting point for a potential descending fifth-progression has been neglected. Measures 47–49: Taking a somewhat broader view, one may regard the E>D second discussed above as an event of the chordal interior, preceding a soprano descent, with the G reinstated at 472 serving as the starting point for a fifth-progression occurring in the piano. This rendering is inconclusive because the goal tonic chord is presented in 64 position and because a G is retained at the top of the texture. Measures 49–51: The preceding descent is here reiterated, with a different irregularity at its close: a tonic 6-phase chord (C-E-A, unfurled), rather than 5-phase C-E-G in second inversion. Measures 51–55: Instead of starting again on the tonic, the A minor chord (the tonic’s 6-phase chord) reached at 511 proceeds to II. Schumann accomplishes this through two descending-third successions in both outer voices, resulting in the introduction of the fifth-progression’s F an octave lower than expected. As 15.5 shows, that F is transferred up an octave in

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

coordination with the arrival of V7, after which the descent to C proceeds without a hitch. Once again, though, the goal C is supported not by the C-E-G tonic harmony, but by its 6-phase surrogate. Measures 55–57: At 552 Schumann backtracks all the way to the C-E-GB♭ chord introduced at 432. Whereas there and at 452 the melody focused on the interior strand E>D, now G>F is projected, though the thread is lost in measure 57. Thus we have experienced yet another failure in the quest for closure. Measures 57–59: Schumann’s backtracking at this point is indicated in 15.5 by the two asterisks placed between the graph’s staves. (We now proceed from the chord marked by the second asterisk, with a surging A➔ chord sporting root A in the bass.) Both the harmonic progression (I(5)–6 II V I) and the melodic descent (G>F>E>D>C) are exactly what we would want in order to achieve the long-awaited closure. Yet success remains elusive once more, since the concluding tonic chord sounds in its 63 position.9 Measures 59–63: As a last-ditch effort to get all the factors to work together in achieving a PAC, Schumann revises the harmonic progression, substituting I IV5–6 for I5–6 II. Alas, the dominant that follows after IV once again misses its tonic goal, with the tonic’s 6-phase chord sounding yet again at 611 and at 631. Though Schumann of course could have constructed the desired cadence with no problem, he now bows out – as if done in by his unruly progressions – leaving behind a panorama of failures. The A minor chord lingers precariously for two measures and then is fully embraced for the exposition’s repeat and, the second time, for moving onward to the development. Though its analytical notation contrasts mine, Smith’s Example 2009/9 conveys a conception of the transition that is very similar to that on display in 15.4. By consulting the score, one might become convinced that the transfer of the piano melody from 272 through 291 down a fourth for the violin line from 292 through 311 would warrant placing the onset of Smith’s local V at 292 rather than at 311. (That is, the C-E-A chord at 301 serves as the subdominant within a tonicization of dominant E, rather than as a reiteration of the preceding A tonic.) We both are coming to terms with the broad progression of four roots, A–D–G–C, which I interpret as an indivisible circular progression connecting A Minor’s I

195

196

Harmony in Mendelssohn and Schumann

and III Stufen (with III tonicized over the course of FS).10 Smith instead interprets A–D in one key and D–G–C in the other, with D serving as a pivot chord. In his ex. 2009/6, the completion of the motion to C is qualified by the bold placement of the symbol ♮VII (in A Minor) under the G chord, which is then prolonged for most of the section; whereas in ex. 2009/10B (= ex. 2011/9.9), the C chord does not even appear after the G chord. Our disagreement regarding the C chord at 381 is of great consequence. Though Smith grants the chord a I label in ex. 2009/9, this I is conveyed as subordinate to a prolonged V, a view that is carried over into his FS graph (ex. 2009/6), where the G chord’s relative prominence is conveyed through the impressive length of its stem (over twice as long as that of the C tonic’s bass E). I instead hear G➔ fully resolving to C – despite the C chord’s inversion at 381 and 421, and despite the added B♭ at 432. The A-C-E chords that transpire during FS all relate convincingly to a C tonic, just as the F-A-C chords during P relate at some level to an A tonic. My conviction wavers only at measure 63. Given my contrasting perspective, I would recommend a few alterations in the details of Smith’s ex. 2009/6. Most urgently, I would want to see a wholehearted embracing of C as a tonic Stufe in C Major at 432, 452, and 552.11 I also would propose C as a melodic goal (descending from D) at 491, at 511, and at 551. Roman numeral IV should be removed as a harmonic label for measure 51 (in that that chord is internal to a descending-thirds trajectory between A and D) but added for 601 and 621 (where a G held over from the preceding tonic harmony prevents chord member F, and with it A in the bass, from sounding on the downbeat). Finally, I would refrain from using Schenker’s interruption symbol at measure 56, instead interpreting the backtracking and reiteration as merely a local rhetorical flourish. Whereas Smith rejects his ex. 2009/10A, which (except for its projection of E as the Kopfton) corresponds closely to my 15.3, and instead endorses his ex. 2009/10B, which proposes that G rather than C is the principal bass pitch between the As at the endpoints of the exposition, I propose that Schumann in fact pursues the “conventional tonal structure” of 2009/10A. For me, the local quirks in his writing can be (and ought to be) fully accommodated within that framework, as my analysis demonstrates.

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

Movement 1 development (measures 65–ca. 115) The exposition’s failure to attain a PAC in C Major has seriously compromised the viability of the common minor-key tonal plan of proceeding from I to III during the exposition, followed by a continuing upward course to V♯ during the development. The A minor chord of measure 63 not only reintroduces the tonic for the repeat of the exposition, but also serves as the departure point for the chordal trajectory that will occur within the development. We are back on square one, with no inkling of what will happen next. Because so many chords appear between the end of the exposition and the first hints of P’s return for the recapitulation, it is challenging for listeners (and analysts) to get a clear sense of what motivates Schumann’s writing during this movement’s development section. My approach, whose outcome is presented in 15.6, has been to apply hierarchical thinking sensitively yet adamantly, so that a more manageable number of chords may be assayed. Before entering into the details of how these foundational moments emerge out of their rich chordal context, one might attempt to discern what factors could have motivated their incorporation within the progression. If logical relationships – in keeping with compositional procedures of the time – are noted, a confidence that one has hit upon a viable means of coming to terms with this development section may emerge. Analytical endeavors then might focus on assaying how well this representation corresponds to what Schumann actually wrote. Example 15.6 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1, mm. 65–119.

197

198

Harmony in Mendelssohn and Schumann

The trajectory displayed in 15.6 may be interpreted as a broad and roundabout traversal of a descending circle of fifths, incorporating all of A Minor’s seven diatonic pitch classes as roots, arranged in the following order: A

D

G

C

F

B

E

A

This scenario may have developed in Schumann’s mind as an outgrowth of the thematic content from measures 1 through 5, where a revised interpretation of the pitch F in measure 1 (corresponding to what was discussed above regarding the pitch D in measure 6) would generate the F B E A segment of this circle. Each pair of adjacent chords that may be enhanced by a surge in fact sports such an evolution at least once during the development section (thus A➔ D, D➔ G, and so on, with only F and B not so connected). The circle proceeds through its fourth chord during measures 65 through 75; a backtracking to the second of the circle’s chords ensues through measure 81; a temporary period of stalling involving a hovering around that second chord (incorporating its upper- and lower-third chords) persists through measure 99; and finally the circular progression begins to move forward again, reaching the tonic in measure 119. Our exploration of how the model of 15.6 corresponds to the composition’s surface details will consider each of these four regions in turn. Measures 65–75: Just as the work proceeds from the exposition’s closing measures by looping back to P in the context of the repeat sign, the initial measures of P likewise prevail as one moves beyond that sign into measure 68. At a local level, a harmonic progression like the one that opens the exposition occurs: m.

A Minor:

65

66

68

69

I

II➔

V➔

I

From a somewhat broader perspective the circular progression that will guide the entire development section is inaugurated by the surge that emerges between the tonic harmonies at that progression’s endpoints: m.

65

A

69

) ➔

(

The circle’s next component – a surging D chord – sounds for only one measure (71), a terseness that will be compensated for later in the development. Then G arrives. This G chord’s B♭ is a wobbly lowering of diatonic B♮, a factor that Schumann offsets with some effort. Reprising an initiative also deployed during measures 9 through 12, a local sequential progression now connects the broader circular progression’s two versions of the G chord, as follows: m.

72

G5———– G (

♭6

73

A

♭5–——–6

74

B♮5–——–6 )➔

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

Whereas normally an ascending 5–6 sequence’s sixth chord is built from the same pitch classes as its first chord, here that relationship is tempered by a B♭to-B♮ inflection that allows the G chord to target the C chord of measure 75 as a surge. In that this C chord likewise is introduced with minor quality, an eventual shift to E♮ might be expected. In this case, however, Schumann instead rescinds most of the progress made thus far, as we shall see below. Measures 75–81: Given the setbacks encountered during the exposition attempts to arrive at a PAC in C Major, it hardly seems appropriate that a circle of fifths would proceed through all of its stations without resistance. Though Schumann might have relished the superlative drive that a circular progression offers, he likewise could enjoy setting to work on it and making it yield to his wishes. Though the C minor chord of 751 is transformed into a surge (at bass B♭ during 761, where E-G-B♭-D♭ is an evolved state of C➔), the chords that transpire over the next few measures are of local impact, not components of the broad circular progression we have been mapping out. Schumann is drawing upon a property particular to circular progressions. Since a closed circle contains seven distinct descending fifths (one of them diminished), one may proceed forward six fifths to backtrack by one fifth. In this case Schumann incorporates some of the flats characteristic of C Minor and G Minor, as follows: C ➔F

B♭ ➔ E♭

A

D➔

G

Once G is reached in this manner, another backtracking by a fifth occurs during measures 78 through 81, as: G ➔C

F ➔ B♭

E

A➔

D

Though at first it may seem that a third backtracking (which would lead to the initiating A chord) gets underway soon after this D chord arrives, it turns out that that trajectory goes elsewhere, as will be explored below. Thus the backtracking may be regarded as having ceased with the arrival of the D chord at 811. Measures 81–99: In the midst of a circular progression in which every internal component is related to adjacent chords by both ascending and descending fifths, Schumann pauses to replicate such relationships on a smaller scale, with ascending and descending thirds. Whereas the earlier local deployments of circles of fifths that began after measures 75 and 78 led to a chord rooted a fifth higher, such a circle leads instead up a third in its deployment after measure 81, as follows: D ➔G

C➔

F

Then borrowing the local harmonic progression that landed on its sixth scale degree near the end of the exposition (starting at 592), descents by a

199

200

Harmony in Mendelssohn and Schumann

third occur twice: from F to D, and then from D to B♭. Consequently both upper- and lower-third chords reinforce the D chord attained at measure 81. From lower-third B♭ the restoration of D is accomplished by again deploying a segment of an ascending 5–6 sequence (with the bass now chromatically enlivened), as follows: m.

88

B♭5

91

92 6

B♮

C

5

95

96 6

C♯

D5

Measures 99–119: The D minor chord of measure 96 returns with surge characteristics during measure 99, the endpoint of a local tonicizing I II➔ V I progression with minor dominant.12 The G minor chord of measure 100 is prolonged in a similar manner, with its own tonicizing I II➔ V➔ I progression (with major dominant) concluding with a surging G➔. Given this exemplary commitment to moving forward along the circular path that was inaugurated at the start of the development section, the momentum continues with C in measure 104, which surges at its onset.13 In that Schumann will inaugurate a broadly paced version of the P theme above this C root in measure 110 to launch the recapitulation, it is fitting that the C chord is treated to an expansion, as displayed in 15.7. That example’s open noteheads convey the foundational succession from C-E-G-B♭ to F-A-C. Observe how Schumann heightens the passage’s intensity by persistently preventing a recurrence of a pure C-E-G-B♭ sonority: at least one ascending diatonic or chromatic passing note (the graph’s filled-in noteheads) infiltrates every sonority between the C➔ of measure 104 and the F of measure 111. The F chord is intensified through its presentation in 53 position during measure 115, thereby confirming that the melody’s F introduced at measure 111 is transpiring in a chordal context that contrasts its deployment during measure 1, where the introduction of the A-C-E tonic chord was of paramount importance. Example 15.7 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1, mm. 104–111.

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

Among these internal chords, that of measure 110 is the most enigmatic. Schumann bridges the divide between the C and F chords with a C-E-A sonority that might be interpreted as the C-E third of the prolonged C➔ with an anticipation A (the third of goal F-A-C), or instead as the A-C third of the upcoming F-A-C with an E suspended from C-E-G-B♭. (These alternative perspectives are juxtaposed in 15.7, where open oval noteheads correspond to C➔ and open diamond noteheads correspond to F.) Yet that is not all! Given the thematic content initiated by the violin in measure 110, there is also a resonance with the A-C-E tonic chord of measure 1. My reading proposes that an A tonic is not asserted at this point, but instead that the context (dominated by the extended C➔ surge) molds the theme to its own agenda, which the development’s overriding circular trajectory suggests would inevitably proceed as C➔ F. The circular progression then concludes with the B, E, and A sonorities that were deployed in the theme’s harmonization during measures 2 through 5. Consequently there is no tidy dividing line between the development and the recapitulation. The circle-of-fifths initiative concludes on the recapitulation equivalent of measure 5’s tonic during measure 119. There is no tonic equivalent of measure 1.

Smith’s presentation passes over a large chunk of the development section. After the A Minor tonic harmony at the end of the exposition (measure 65, displayed at the right edge of his ex. 2009/6 with an opennotehead A in the bass), we regain contact with the progression at the CE-G-B♭ chord of measure 104 (labeled as III at the left edge of his Example 2009/11). Despite the fact that III possesses a minor seventh (acknowledged by a ♭7 symbol annotating his graph), his middleground harmonic analysis shows the chord proceeding (via a 5–6 shift) to II♯. This all seems dubious to me. Whereas Smith would reject my reading of the exposition (prior to the reinstatement of tonic A during its final measures) as I ( ) III, I reject his reading of the development as I ( ) III. The fifth-relationships projected in my 15.6 correspond to a procedure that Schumann would have known and experimented with at the keyboard from an early age. In that context the C➔ chord has a fifth-related predecessor (G➔) and a fifth-related successor (F). Schumann here integrates the P theme’s II V➔ I within the circle, modifying the chordal trajectory to

201

202

Harmony in Mendelssohn and Schumann

accommodate root F (thus making the neighboring note F of measure 1 a stable chord member), so that a seamless continuation of the circular progression may be achieved as P’s thematic material emerges. There is no rationale for picking and choosing among the fifth-related chords (resulting in Smith’s III-to-II♯ succession, reinforced by a C–B slur in the bass of his ex. 2009/11), nor is an E [Minor] tonicization (between measures 110 and 118) warranted. Measure 119 should be displayed as the definitive return of the A Minor tonic (corresponding to measure 5), and not as a subordinate i (subservient to a prolonged V), as Smith conveys in his ex. 2009/11.14 My 15.7 organizes the pitch content between measures 104 and 111 into three categories: members of the circle’s surging C-E-G-B♭ chord, members of C➔’s F-A-C successor, and connective pitches. At various moments along that trajectory, chords of no structural significance are formed as members of the C➔ chord and one or more connective pitches coincide. Smith interprets some of these incidental chords as “motivic sonorities” (p. 2009/63), labeled using the letters F and A (which he proposes bring to mind the other players of his TMS complex), including one A chord that he interprets as iv in E [Minor] (ex. 2009/11).15 In my view, he has at this point lost sight of the essence.

Movement 1 recapitulation and coda (measures ca. 115–209) Because most of what is presented during a typical recapitulation will have been explored analytically in the context of the exposition, recapitulations generally receive limited attention in analyses. Here we focus on three specific questions: one pertinent to all recapitulations, one to all recapitulations within minor-key sonata movements, and one to this particular recapitulation. A graph of the entire movement, presented in 15.8, should be consulted as required. (This graph accommodates the unusual development section – which essentially prolongs the tonic – by displaying it as a parenthetical passage between the tonic harmonies at the end of the exposition and beginning of the recapitulation. Kopfton ^3 is maintained through to the recapitulation’s TR.)

Example 15.8 Analysis of Schumann: Sonata in A Minor for Violin and Piano (op. 105), mvmt. 1.

204

Harmony in Mendelssohn and Schumann

How does the recapitulation’s TR differ from the exposition’s, in order to accommodate the fact that what follows will take place in A Major rather than in C Major? The exposition TR’s A ➔ D G➔ C circular progression (15.4) is ignited by a surging A➔ chord during measure 34. At the equivalent spot during the recapitulation TR (measure 148) what begins to take shape as another A➔ surge is transformed into an F♯➔ surge. As a result, the former continuation is transposed down a minor third, to F♯➔ B E➔ A.16 The E chord here is asserted as a structural dominant between the A Minor and A Major portions of the recapitulation. The former supports the background ^3 prolonged since the movement’s onset; the latter supports middleground fifth-progressions whose goal is background ^1. Consequently the ^2 of TR’s dominant harmony is displayed in 15.8 as a component of the Ursatz. Given that the exposition’s juxtaposition of A Minor and C Major regions is converted into a juxtaposition of A Minor and A Major regions during the recapitulation, does the recapitulation fully embrace the new key? Or is A Minor still a force to be reckoned with even after A Major has been established? Because the material presented during the recapitulation’s A Major FS is a transposition of the exposition’s C Major material, A’s major modal characteristics are firmly supported – until measure 177. From that point onward, the piece seems torn between these contrasting pitch collections. Though the ESC sounds as an A major chord (measure 195, replicated at least once every two measures through measure 205), the F♮ and C♮ of A Minor infiltrate the local chordal progressions as the coda unfolds. Schumann’s last word on the subject is devastating: the tonic shifts to minor quality in measure 206, the supertonic that follows is of diminished quality (indicative of A Minor), and the final tonic is a chord of minor quality. The shift to A Major has been decisively revoked. The A minor chord at the end of the C Major tonicization during the exposition is easily accommodated, since the movement is in the key of A Minor. How is this unusual detail handled in the context of the recapitulation’s A Major region? The F♯-A-C♯ chord of measure 175 indeed is worrisome. How will Schumann get beyond the impasse? The unexpected shift to F-A-C at measures 177 through 179 calls listeners to attention: certainly something unusual is brewing. This chord hearkens back to the F-A-C chord of measure 25. Though they occur in different parts of their respective sections (near the end of P versus near the end of FS), these chords share the characteristic of preceding a strong cadence on an A-

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

rooted tonic: minor during the exposition, major during the recapitulation. (This discussion is predicated upon the fact that the recapitulation does eventually achieve a cadence, the ESC, whereas the exposition lacks an EEC.) It was noted earlier how the exposition P (and of course the recapitulation P as well) juxtaposes three distinct predecessors of the dominant leading up to P’s cadence on tonic A: II➔, ♭II, and IV. By adding D♯ to F-A-C during measure 180, Schumann’s already impressive collection grows to include II⇨ (in the formulation referred to in conventional harmonic analysis as a German augmented sixth). The distinctive augmented-sixth interval DF ♯ alternates with its FD ♯ inversion over the course of the measures preceding the dominant restoration in measure 189, followed by the tonic resolution (the ESC) at measure 195. In our considerations of the recapitulation and coda, Smith and I both focus on the reintroduction of A Minor beginning with the F-A-C chord of measure 177 in a context that has projected A Major since measure 157. The main difference in our perspectives is that I regard this F to reside within the recapitulation (whose close resembles the exposition and recapitulation P’s cadence because the exposition FS does not offer a suitable cadential model), whereas for Smith the F of measure 177 instead inaugurates the coda (p. 2009/63).

Movement 2 refrain and transition (measures 0|1–15) Local upper neighbor F in an accented metrical position embellishes the fifth of the A-C-E tonic triad that opens the sonata’s first movement. Local lower neighbors E and G in an accented metrical position embellish the root and third of the F-A-C tonic triad that opens the second movement, which is in rondo form. (See 15.9.) In both movements the violin melody proceeds swiftly from Kopfton ^3 to its upper third (CE and AG second of the descending fourth is filled in chromatically at the end of measure 2, and a D serves as a neighbor to C (with a concurrent shift up an octave) in 15.10b. (In the models, the foundational pitches of 15.10a are displayed using open noteheads, while the various accretions appear as filled-in noteheads.) Though one might regard this new sonority as mere connective voice leading, what happens later encourages one to interpret it as an incipient formulation of a surging II➔ – that is, as (B♮)-D-F-A♭ above dominant pedal C. Though under normal circumstances positing the assertion of a G➔ chord when neither a G nor a B♮ sounds would be regarded as excessively imaginative, in this context it might be justified, in that the imagined B♮ emerges in sound at the equivalent location just one measure later (as the alignment of 15.10b and 15.10c helps clarify). Two other novelties enhance model 10c: the cessation of the C pedal point, and the introduction of the pitch D, projecting the tonic’s 6-phase chord, between I and II➔. A different mix of chords emerges in 15.10d. Whereas the C pedal point is reinstated and the tonic 6-phase chord does not intervene, B♮ sounds against pedal C for the II➔ chord, in which A (rather than A♭) now serves as the ninth. The next variant (shown in 15.10e) begins in measure 7. Reprising the ritardando and fermata to remind us that, yes, the broader progression is still on hold, this variant is much more elaborate than any of its predecessors. Not only is the tonic’s 5–6 shift restored, but also it is doted on by the rhetorically emphasized A➔ embellishing chord that targets the 6-phase D

Schumann: Sonata in A Minor for Violin and Piano (op. 105)

chord.18 Both the 6-phase chord and the supertonic that follows are presented in two stages: first diatonic and then surging. Though this surge-rich region is accommodated within the harmonic progression (as conveyed by the Roman numerals annotating the model), listeners also might note a correlation between this trajectory and the circular progression of the first movement’s development (15.6, measures 65 through 104). The expected cadence at 101 is delayed by a second pass through this variant’s trajectory, now with the diatonic onsets of VI and II suppressed and with II➔ presented in root position at 112. The PAC finally is realized at 121.19 All that has transpired constitutes one phrase with multiple internal backtrackings (offering opportunities for creative modifications that Schumann has exploited admirably) rather than multiple phrases. What may seem at first to initiate a repeat of the refrain’s contents (starting at the upbeat to measure 13) ends up serving as a transition. The dominant reached at 141 resolves without much ado to an F Minor tonic chord to initiate the movement’s first episode, to be discussed below. Both the violin melody and the piano bass descend a second over the movement’s first bar line: A>G above F>E. Schumann’s slurring calls attention to the 10–10 relationship. These lines immediately snap back: a third 10 ( AF ) completes the gesture, with leading tone E resolving to F in the bass. Whereas the violin’s G is a local neighbor between two As, the piano melody’s A>G in the middle of measure 1 is a more substantial structural element. G’s arrival coincides with the onset of II➔, and though G is transferred to the tenor register at 31 it eventually is restored to prominence (most notably just prior to the arrival of the thirdprogression’s goal F at 121). The dominant’s extended prolongation (which I describe as reverie above) makes the normalcy of this basic structure (as projected in 15.9) somewhat challenging to appreciate. As an orientation, one might perform from the beginning of the movement through the fermata in measure 3, and then pick up with the dominant at the end of measure 11 (noting how the violin’s GG there complements the initial AA), so that the PAC at 121 arrives normatively on the phrase’s fourth downbeat.20 Smith’s Example 2009/12 displays the opening F major chord only as IV in C Major. (Earlier, in ex. 2009/5 – reprinted as ex. 2011/9.3 – the option of a tonic interpretation was acknowledged for the F chord at the onset, but not for that during 11.) The possibility of the most normative

209

210

Harmony in Mendelssohn and Schumann

of all structural openings – tonic harmony with Kopfton ^3 – is thereby removed from consideration. (Smith describes the refrain, which he concludes at measure 8, as “tonally unstable and harmonically open” (p. 2009/64). My 15.9, which conveys the refrain as a twelve-measure entity, is tonally stable and harmonically closed.) I suggest that readers endeavor to hear the opening three chords as a projection of I supporting ^3, pondering how performers might convey that interpretation effectively, as well as how the bass G that follows later in measure 1 might be projected as the fulcrum between roots F and C. Though ending ex. 2009/12 with the III♯ chord of measure 8 might seem justified by Schumann’s placing a fermata on that chord in his score, I suggest that this chord actually is in the middle of things – that the AB–B>A. (The initial D actually “belongs” in the preceding measure, where it would have served as the II harmony’s seventh. Its delay until the following downbeat, where it clashes with the onset of V7, converts it from an unaccented into an

241

242

Notes to pages 38–39

14. 15.

16.

17.

18.

19.

accented passing note. The following C♯ and A are chord members of V7, while B connects those pitches as a passing note.) Four thirds are involved in projecting these two strands: A♭B♭, C>A♭, and B♭>G. While the melody’s initial A♭B♭ third results in the immediate shift to IV’s 6 phase (thus IV5–6). The dominant follows at X. Though only pitch classes E♭ and G participate in its projection, no successors to the recently stated D♭>B♭ third have sounded, and so listeners should mentally merge them with the E♭ and G (which might be conveyed analytically using parentheses as V(7)). Finally, the tonic is restored during R2 (thus I). Note that at that goal point the melody’s A♭B♭ thirds have been followed only by the upper-strand restoration of C. The full flowering of the C>A♭ third transpires during the passage from R2 to R3, coordinating with the onset of further chordal activity that will lead the broader harmonic progression from I to the phrase’s goal at T2. (See question 4.) Because the chords at R1 and U1 are almost identical, the connective chord at Q1 features neighboring notes: G and D♭ embellish the tonic’s A♭ and C, respectively, while E♭ is maintained as a common tone. In contrast, the chord at U2 is situated higher than that at U1, and so the connective chord at Q2 features passing notes. Three concurrent ascending stepwise motions occur: that between root and third (A♭E♮, a diminished fifth. That dissonant interval, here unfolded in the downward direction, is resolved by the upward unfolding of a minor third (FD♭ third that coordinates with the FA ♭  AF ♭ voice exchange that surrounds it. 12. The cadential 64 chord at U1 generally will be followed by a ♮ 35 resolution (often with the dominant’s seventh emerging also at that point). That resolution here transpires at U4. The chord at U2 serves as an embellishing chord, incorporating the upper neighbors of all three of the cadential 64 ’s members (F, A♭, and D above dominant root G). 13. A PAC appears to be imminent, with U4’s V♮ 37 poised to resolve to I with a C residing at both edges of the texture. (Though Mendelssohn does provide a tonic chord with a C in the soprano, his bass note on the following downbeat is G, not C.) 14. The B♭>A>G third during R1 projects a third from the tonic harmony, which transpires at that point. The A♭>G>F♯ third from S2 through U2 harbors a wobbly note – A♭ – that reverts to A♮ at U2 for the projection of V♯. 15. In a diatonic context, A-C-E♭ (= II) often is deployed between I and V♯ in a minor-key context. In a common deployment of chromaticism, that chord’s A may be lowered to A♭ (as a wobbly note), thereby changing the supertonic’s quality from diminished to major. This chord often will be represented by the Roman numeral ♭II in analyses. 16. The tonic harmony’s G and B♭ are embellished by neighbors F♯, A, and C, which sound along with a prolonged D to create the embellishing chord at T. At the conclusion of T, the outer voices project a

F♯ C

compound augmented

fourth. Whereas the bass descends as expected to B♭ at R2, the F♯’s resolution to G is accomplished within the following chord’s interior, since the melody instead leaps up to D.

Notes to pages 68–72

17. Though the bass descent from root D at U2 to C at U3 is a conventional means of dominant intensification (added minor seventh, in

♯4 2

position), concur-

rently the soprano F♯ of U2 is embellished by a lower neighbor, E♮. Consequently the full sounding of the dominant

♯4 2

occurs only during the

beat’s final eighth note. 18. The B♭>A>G third of R1 is presented in retrograde, as GC>A tonic arpeggiation in the melody. 8. The progression proceeds from the tonic to the mediant (which shares two pitch classes with the tonic) between R4 and R5. The progression is led toward that mediant through a surging embellishing chord – G-B-D-F – at S2. Its diminished fifth, BF , resolves to the mediant triad’s lower third, CE . 9. The D is supported by II7 during T2, followed by a cadential 64 inaugurating the 7

dominant at U3 and then V 5 at U4. Because the embellishing chord preceding ♯3

the mediant is built above bass B and the II7 chord is presented in

6 5

position,

the bass ascends AC♭>B♭ line from neighbor C helps differentiate Kopfton B♭ from the tonic chord’s other arpeggiated pitches, as does the sforzando C>B♭ stated multiple times in the bass, beginning in measure 10. 6. See my Thinking About Harmony, pp. 155–161. 7. This particular evolved state of the supertonic will be deployed again later in P, at 274 and at 314. 8. Without the “harmonic reduction” caption, I might be able to convince myself that the example shows merely a series of initiation points – paths from E♭ to Fm, as emphasized by the first three brackets below the staff and described as “a series of overlapping harmonic expansions up a step of ever increasing scale, which become progressively more firmly established” (p. 61) – with no intention to convey a broader continuity. Taylor’s analysis transpires in the context of an attempt to demonstrate the following proposition: “Both thematically and harmonically, this opening movement reveals a continual process of growth out from its opening phrase that indeed justifies the analogy with the aesthetic ideal of organic growth and unity claimed by several commentators” (p. 60). As my chapter unfolds,

253

254

Notes to pages 88–94

9. 10.

11.

12.

13.

14.

15.

16.

17.

readers will encounter evidence of “organic growth and unity” not so much in disjointed juxtapositions of E♭ and Fm as in diverse manifestations of I5–6 II V I. None of these accents are included in the score excerpts as printed in Taylor’s exx. 2.2 and 2.3a–b. The vi of fig. 1 certainly is a typographical error. It should appear as VI, as in figures 3 and 6, confirmed also by the words “major submediant” in his commentary (p. 73). Vitercik is not consistent in his application of this protocol. The E♮-G-B♭-D♭ chord of measures 23 and 24 is labeled as VI in figures 3 and 6. Yet C-E♮-G-B♭D♭ in measure 41 is labeled as V/ii in fig. 4 – and then as VI in fig. 6! Readers concerned that my style of Roman numeral analysis is more cumbersome than Vitercik’s are welcome to use my shorthand symbol VI→ instead of the four-tiered entity displayed in 9.2a. Though I appreciate my system’s capacity for precision (keep in mind that several different evolutions of the submediant could be labeled as VI→ but have contrasting appearances when my Arabic numerals, accidentals, and bullet symbol are deployed), others may find my arrow notation a convenient and intuitive alternative. My conception of harmony as a constantly churning, vibrant process involving frequent shifts in a chord’s configuration (generally resulting in a more dynamic thrust toward its successor, as in II’s eventual surge targeting V) makes Vitercik’s word “sitting” seem especially inappropriate. Whereas my harmonic analysis features a conventional E♭–B♭–E♭ bass arpeggiation (beamed in 9.2a), Taylor’s slur from C to E♭ in his ex. 2.1 (where I read the C as corresponding to that of both measures 25 and 31) is matched by a slur from C to G (for the E♭ chord’s inversion) during measures 31 through 33 of ex. 2.5 Thus my description above of his slurring in the vicinity of measures 28 and 29 as “adequate” may be wishful thinking on my part. If pressed, he might clarify that his three small slurs break up what foundationally would be a broad slur from the C of measure 25 to the E♭ of measure 29. Though many analysts espouse the ubiquitous symbols Fr+6 and Ger+6, one may reasonably ask whether the chords they represent have roots. The vital presence of A♮⇨ and C⇨ within the circular progression of 9.3c offers a strong incentive to answer that question in the affirmative. In both cases the augmented sixth interval is formed by the pitches a major third and a diminished fifth above the root (which is omitted in the “German” version of the chord). I might hope that, if pressed, Vitercik would name C as the root of the chord in measure 58. In his chapter he follows most other analysts in merely applying the “French sixth” nickname (p. 76). (See also note 29, below.) In his commentary Vitercik alludes to “the abrupt arrival in G minor in m. 52” (p. 75) – the key of G Minor, whose dominant root D emerges at 521. Even if I

Notes to pages 94–99

18. 19.

20.

21.

were to concur that the passage is “in” G Minor, I would not interpret either the G-B♮-D or G-B♭-D chord of measure 55 as an asserted tonic. G Minor’s V♯ holds sway for the entire duration of measures 52 through 56. My 9.4c also omits several foreground chords that systematically traverse an upward path between the accounted-for F and C: F E♭→ A♭ G→ C. Another infelicity in Vitercik’s Roman-numeral deployment should be noted as well: his use of the symbol VI/ii for the D-F♯-A chord of measure 52. Appropriately, the VI numeral is capital because the D chord is of major quality. But VI in the key of ii (F Minor) is D♭-F-A♭, not D-F♯-A. In that either D♭ or D♮ may be deployed as a root in this context, the analytical notation must be able to differentiate between them. (Compare with my 9.11, measures 179–180.) As much as possible, conventional harmonic analysis neutralizes the impact of chromatic chords. For example, chromatic D-F♯-A-C in C Major is made to appear as diatonic V7 in G Major through the symbol V7/V. Alas, that tactic will not work for D-F♯-A♭-C, since that chord’s pitches are not all diatonic in any one key. What can be done to make such intense chromaticism palatable? Nicknames! But in doing so, information about D’s role within the chord’s construction and the chord’s location within tonal space are sacrificed. Though this lamentable situation is not Taylor’s fault, of course, it helps to explain how he could fail to come to terms with the chords at the ends of measures 51 and 58 within his conception. Mendelssohn’s Instrumental Music by Erez Rapoport (Hillsdale, NY: Pendragon Press, 2012) is a book that all Mendelssohn enthusiasts should know. I have elected not to interact with it in a substantial way in this volume because, as editor of Pendragon’s Harmonologia series (in which it appears as volume 18), I was deeply involved in its production. I will honor his achievement by simply mentioning our contrasting, equally viable treatments of one specific chord from the Octet’s first movement: D♭-E♮G-B♭ at 1161. I take an overtly harmonic approach in 9.6, interpreting the chord as a manifestation of II→. He instead leaves an open area between I6 (his symbol for the tonic in first inversion) in measure 115 and V at 1163 in his ex. 1.9. His approach thus is more linear than mine: a formation consisting of multiple appoggiaturas delaying the dominant’s arrival by two beats. Janet Schmalfeldt offers a middle path in her book, In the Process of Becoming: Analytic and Philosophical Perspectives on Form in Early Nineteenth-Century Music (Oxford University Press, 2011). While numerals akin to Rapoport’s, she fills in the her ex. 7.13 shows I6 and V 4
7 3

space he left between them with the annotation (°7). 22. Vitercik’s brief comment regarding the G major submediant – “strongly biased toward its own minor submediant, C minor” (p. 82) – does not make sense to me. I think the word subdominant was intended, instead of submediant, in

255

256

Notes to pages 100–108

23.

24. 25.

26.

27.

28.

29.

30.

reference to the G-C-E♭ chord at 781 and 821. In my view, Mendelssohn is merely upholding the B♭ Major tonicization through the deployment of its diatonic E♭, rather than G Major’s E♮. In any event, the chord is a local 64 embellishment of G-B♮-D. I suggest that it does not warrant being described as a subdominant, which would imply C’s assertion as a root. Though the placement of parentheses around V/vi appears intended to imply a temporary diversion from a broad V extending from measure 90 to the purported I in measure 96 – “a momentary shift from V of B♭ to V of G minor” (p. 82) – it is in fact the V/vi that resolves instead of the F dominant. While rejecting Vitercik’s proposal that the B♭ tonic emerges at measure 96, I would advise against merging the dominant of measures 96–101 with that which sounded in measure 90. That earlier dominant’s tendencies are used up with soprano E♭’s descent to D (supported by the B♭ tonic’s 6-phase rather than its 5-phase chord). The latter chord is referred to as “V 65 of IV” in Vitercik’s commentary (p. 85). My thought that perhaps V/c was intended to convey “V of c” followed by “c” was shot down when I noted the juxtaposition of V/f and f, V/g and g, etc., earlier in the same diagram. The precise wording of Taylor’s remark at this point exemplifies the wide gap between our harmonic conceptions: “This G major then moves, via C minor, to a temporary A♭” (p. 61). Whereas my G→ C conception proposes that the former chord is dependent upon the latter, Taylor appears to hear a GF>E♭ of measures 277–278 (repeated in measures 279–280) concludes the linear initiative, not acknowledged in her reading. From my perspective, her “CODA” and PAC markings at 2761 should be moved to 2801.

10 Mendelssohn: Song without Words in F Major (op. 85, no. 1) 1. “Form and Tonal Process: The Design of Different Structural Levels,” in Trends in Schenkerian Research, ed. A. Cadwallader, New York: Schirmer Books, 1990, pp. 1–21. 2. The initial impetus for the A chord certainly has something to do with symmetry: during stanza 1 the trajectory proceeds down a third from F to D (measures 5 and 6), whereas during stanza 2 the equivalent measures (13 and 14) instead lead up a third from F to A. Once there, however, Mendelssohn was free to use the A sonority however he pleased. My reading ultimately connects the F chord of measure 13 and the D chord of measure 19 (thus again F5–6), more akin to what happened during the first stanza than one might have expected at first, given the potency of the intervening A chord’s presentation. 3. Given the persistence of intervallic unfoldings in the melody (C>F in measures 3 and 11, ED♯ in movement 1, F>E in movement 3). As will become apparent later, I read both Fs in the same way: as neighbors resolving immediately (F>E). 6. A context for the slurred F>E>D♯ that Smith displays twice during measures 1 through 3 will emerge near the end of the movement – see measures 179 and 180. (That deployment begins against root F rather than against root A, thus reversing the hierarchical relationship between F and E.) As will become clear as my analysis unfolds, Schumann deploys a most potent chord (II⇨) at that decisive moment as a means of exceeding the impact of three other approaches to the dominant in earlier measures. Smith’s graph suggests that we encounter this highly charged “supersurging” chord already in measures 2 and 3! 7. Though I suspect that Smith intends for the Arabic 6 he has placed below bass D in measure 14 to indicate that the D-F-B chord is in 63 position, it is possible (given the conspicuous connection from the F-A-C chord of measure 12) that he instead wants to convey that the D is the 6-phase pitch within VI5–6, prior to B’s arrival.

265

266

Notes to pages 189–200 8. Though Schumann’s projection of the linear trajectory is not quite complete – D’s 6-phase pitch is displayed within parentheses in 15.1 – the sense that a linear trajectory is being traversed should nevertheless be apparent. As such, I question Smith’s breaking up of the C