Schumann's Eichendorff Liederkreui and the Genre of the Romantic Cycle
This page intentionally left blank
chumann's...
24 downloads
589 Views
18MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Schumann's Eichendorff Liederkreui and the Genre of the Romantic Cycle
This page intentionally left blank
chumann's Eichendorff Liederkreis) and the Genre of the Romantic Cycle
David Ferris
OXFORD UNIVERSITY PRESS
2000
OXFORD
UNIVERSITY PRESS
Oxford New York Athens Auckland Bangkok Bogota Buenos Aires Calcutta Cape Town Chennai Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paulo Shanghai Singapore Taipei Tokyo Toronto Warsaw
and associated companies in Berlin Ibadan Copyright © 2000 by Oxford University Press Published by Oxford University Press, Inc. 198 Madison Avenue, New York, New York 10016 Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Library of Congress Catalogmg~in-Publication Data Ferris, David, 1960Schumann's Eichendorff Liederkrew and the genre of the romantic cycle / David Ferris, p. cm. Includes bibliographical references and index. ISBN 0-19-512447-2 1. Schumann, Robert, 1810-1856. Liederkreis, op. 39. 2. Eichendorff, Joseph, Freiherr von, 1788-1857Musical settings— History and criticism. 3. Song cycles —History and criticism. I. Title. ML410.S4F472000 782.4'7'092-dc21 .
00-036740
1 3 5 7 9 8 6 4 2 Printed in the United States of America on acid~free paper
To Hannah Goodwin and Allan Keller
This page intentionally left blank
PREFACE
The present book is the culmination of a study that began with a seminar on the nineteenth-century song and piano cycle that I took as a graduate student. Our primary interest was in how nineteenth-century composers used the cycle to experiment with musical forms and, in particular, how they tried to create musical structures that expanded beyond the boundaries of a single movement. Many of the scholars whom we read at the time based their explanations of the cycle's formal organization on traditional assumptions about organic unity and primarily relied on motive and voice-leading relationships as they tried to analyze the cycle as a unified whole. When I began the seminar, I, too, assumed that this was the path that would ultimately lead us to a theory of the cycle, but by the end of the semester we had raised so many questions about this approach that we began to wonder if such a theory was either possible or desirable. In my case, these questions soon gave birth to a dissertation on Schumann's Eichendorff Liederkre'u, in which I argued that we needed to shift our focus from the -whole to the part in order to understand the Romantic cycle. Rather than try to demonstrate how the Liederkreu creates a unified structure, I explored the ways in which the individual songs are fragmentary and open-ended. As I worked on my dissertation, I sometimes worried that by the time I got to the end of it, I would have completely done away with the genre I was writing about. If all I had left was a series of individual songs, then where was the cycle? Fortunately, the members of my committee never asked that question. As I have continued to think about Schumann's cycles, I have come to see that in asking the question myself, I was missing the whole point of the Romantic aesthetic that led to the genre in the first place. I still believe that the cycle does not bind the songs into a unified whole, but I now see that the open-ended form of the individual song itself implies larger relationships that we, as listeners, performers, and analysts, must imaginatively realize as we engage with the work. It is in this sense that Schumann's cycles are Romantic fragments, "always only becoming," never fully completed. Schumann makes us aware of the potentiality of a higher unity, and it is this that gives the cycle its aesthetic and expressive power. Although he has not directly participated in the creation of the present book, I am most gratefully indebted to Prof. Allan Keiler, who taught that graduate seminar and patiently advised and shepherded me through that dissertation. He
viii
Preface
remains my greatest intellectual inspiration and my model of what a true scholar should be. I am equally grateful to my wife, Hannah Good-win, who has always been my most perceptive critic and my most demanding editor. She often knows what I am thinking better than I do, and it is only through her patience and perseverance that many of the ideas in the pages that follow have assumed a coherent form. Thanks to Rufus Hallmark, Susan Youens, Kristina Muxfeldt, Emery Snyder, Janet Schmalfeldt, Joseph Lubben, and Adrianna Ponce, who offered many suggestions and much support, and to Karol Bennett and John McDonald, who so beautifully brought the Eichendorff songs to life and shared their insights about the music and the poetry. Thanks to the students in my three seminars on Romantic song, at Rice University, the University of Houston, and Amherst College. As is so often the case, I feel as if I learned more from them than they could possibly have learned from me. Thanks to Dean Michael Hammond, for his financial support and for the strong words of encouragement that he offered me when we first met. And thanks to Maribeth Anderson Payne, Maureen Buja, and Jonathan Wiener of Oxford University Press, for their faith in the value of my project and for their -work to turn it into a published book. I would also like to offer my gratitude to a group of people who feel like old friends, even though I have never met most of them and have only spoken briefly with the others. They are the late Arthur Komar, Barbara Turchin, John Daverio, Charles Rosen, Patrick McCreless, David Neumeyer, Ruth Bingham, Reinhold Brinkmann, Charles Burkhart, Jiirgen Thym, and Jon W. Finson. The pages that follow could not have been -written -without the benefit of their collective scholarship and creativity. I have greatly enjoyed engaging and debating them, albeit one-sidedly, and I can only hope that my -work has met the high standards that they have set.
CONTENTS
PART I: THE GENRE OF THE CYCLE 1
Introduction, 3
2
Analyzing Dichterliebe, 25
3
Schlegel's Fragments and Schumann's Cycles, 59
PART I I : S C H U M A N N ' S E I C H E N D O R F F SONGS 4
Poem and Song, 91
5
Weak Openings, 121
6
Recompositional Pairings, 141
PART I I I : FROM SONGS I N T O CYCLES 7
Schumann's Process of Composing a Cycle, 171
8
The Song Cycle as a Literary Work, 195
Notes, 229 Works Cited, 259 Index, 265
This page intentionally left blank
Part I The Genre of the Cycle
This page intentionally left blank
INTRODUCTION
A:
s the title of this book suggests, it is concerned both with the genre Lof the Romantic cycle in general and with Robert Schumann's Eichendorff Liederkre.it>, opus 39, in particular. Scholarly studies that are focused on musical works, or on works of art in any field, almost always provoke the same vexing question: Is the objective to discover what it is that is engaging and unique about the single work and that sets it apart from every other artistic creation? Or is the work at hand an example, the study of which will, we hope, lead to the formulation of general principles that can apply to all of the works in a given genre, or by a given artist, or from a given time and place? Gerard Genette comments on this conundrum in the preface to his book Narrative Dutcourde: An EMay in Method, a study of both narrative method and Proust's A la recherche <)u tempj perdu, and confesses his reluctance, or perhaps his inability, to choose either of these "apparently incompatible" explanations of his approach. He cannot imagine treating "the specificity of Proustian narrative" as an example of anything other than itself. However, although his study is an analysis of this specific narrative, each of the individual elements that his analysis uncovers "lends itself to some connection, comparison, or putting into perspective." Genette fatalistically concludes: "I must therefore recognize that by seeking the specific I find the universal, and that by •wishing to put theory at the service of criticism I putcriticism, against my •will, at the service of theory."1 I must also refuse to make such a choice. On the one hand, the study of a work that belongs to a genre such as the cycle, which is so open-ended and varied and for which so few general principles have been formulated and so few adequate definitions put forth, cannot be expected to result in the creation of a model that can be applied in a straightforward way to all of its fellow members. On the other hand, I cannot pretend that my choice of subject—the Eichendorff Liederkreu — is an arbitrary one or that it was guided only by my desire to learn more about this single work, with no concern for the more general questions that are inevitably encountered in the course of such a study. Schumann's opus 39 is recognized as one of the greatest nineteenth-century song cycles, yet it lacks the attributes that have traditionally been used to define the cycle, such as a coherent narrative and an immutable order, attributes that are generally associated 3
A
The Genre of the Cycle
with organicist conceptions of musical structure. The problem already manifests itself in the circumstances of the work's composition. Schumann did not derive his text from a coherent, preexisting literary source but arranged twelve apparently unrelated poems that his fiancee, Clara Wieck, selected from Eichendorff's collected edition. And the only complete surviving set of manuscript sources provides evidence that Schumann did not begin to consider the order of the songs until after he had finished composing them. Given such a history, it is not at all surprising that scholars have found that the Eichendorff Lt£<)erkreu makes such a poor fit with the accepted model of the song cycle. They have struggled mightily, nevertheless, to squeeze it in as best they can. Over the last forty years, the scholarly reception of Schumann's Eichendorff LiederkreL) has been remarkably consistent. Commentators have sought to explain the cycle as an integrated musical -whole that is unified by a web of motivic relationships and a symmetrical arrangement of keys. They have described the text of the cycle as an ordered sequence of moods, bound together by the recurring use of landscape, time of day, and imagery, that leads up to the ecstatic fulfillment depicted in the last song. This ongoing attempt to make the Eichendorff Liederkreu fit its presumed generic model has required a certain amount of tinkering. Some elements of the work cannot be reconciled with the model and must simply be regarded as exceptional, and some aspects of the model must be altered and expanded so that the -work can fit in more easily. At the same time, scholars have replaced the organicist terminology that pervaded the earlier literature on the song cycle, but the model has essentially remained intact, and its underlying premise —the presence of some form of coherent unity — has not been questioned. The problem -with all of this is that in searching for ways to make the Eichendorff LiederkreLi into a cycle scholars have been drawn to some of the least remarkable aspects of the work and have ignored many of the most interesting. My own study of the Eichendorff Luderkreu has led me to question whether organic unity really is the most compelling model for the Romantic song cycle. The prestige and popularity that this work has enjoyed, coupled with the very obvious difficulty we have had in fitting it into the genre, have made the problems with the model especially apparent. But if we look a bit further afield and consider the entire corpus of songs that Schumann composed in his Liederja.hr, we find that the problems we encounter with the Eichendorff songs are really just the tip of the iceberg. The two Schumann cycles that have become central to the genre—Frauenliebe and Lehen 3j\dDu:hterli£be—may fit the traditional model more easily, but they, too, need to be forced in to some extent, and in the process we have ended up losing sight of some of their most interesting aspects as well. In addition, of course, a number of Schumann's cycles have proven to be at least as problematic as opus 39, and these have been either marginalized or excluded from the genre, and thus from the repertory, altogether. When we consider that we have been able to use the model of organic unity to account for only a small number of Schumann's song cycles and that it has not worked all that well even for these, we are left with the question of how this came to be the defining premise of the cycle in the first place.
Introduction
5
The Cycle as a Romantic Genre One reason that the definition of the Romantic song cycle has made such a poor fit -with the actual works it is intended to explain is that it -was not formulated until many years after those works were composed, and in the meantime the aesthetic sensibility that had motivated the creation of the genre had waned. It is considered a truism in the study of music history that theory typically lags behind practice in this way, and one of the examples that is often cited is sonata form, which was first defined in a series of composition treatises in the 1840s. But if we cannot find a formal definition of something that is called sonata form in the late eighteenth century, we can at least find detailed descriptions of how the first movements of the symphony, the sonata, the string quartet, and the concerto are typically structured, as well as an acknowledgment that these forms are related, and thus contribute to the establishment of a clearly defined family of genres. The treatises and music dictionaries of the early nineteenth century, the period during which the Romantic song cycle came into being and reached its creative peak, likewise contain no formal definitions of the terms LiedercycLiu and Liederkreid, but they also contain no descriptions of anything that remotely resembles a song cycle. And while such terms and descriptions do show up in contemporary music journals, their occasional and casual nature gives the impression that the idea of the song cycle as a meaningful genre was not very well formulated. In fact, it would be more accurate to say that the Romantic conception of genre itself is more pliable than ours and the sense in which the cycle functioned as a genre for a composer such as Schumann is quite different from the -way it functions for us today. There are both practical and aesthetic reasons that the cycle was a congenial way for Schumann to publish his piano character pieces from the 1830s and his songs from 1840—41. Since so many of these songs and pieces are extremely brief—often no more than twenty or twenty-five measures — he clearly needed to place them within larger collections for publication. And since he tended to compose in concentrated bursts, during which he would quickly write a group of such works, it was natural for him to then publish the group as a -whole. It is not surprising, given this method of composing, that there are such strong resemblances and connections among the works of each opus. Today we are quite concerned -with these connections, and we classify Schumann's publications in terms of the unity, or coherence, of the songs and pieces within them. This process of generic classification is implicitly a process of evaluation as well: we are distinguishing Schumann's "true cycles" from his "mere collections." But there is no evidence that Schumann made such distinctions himself. First of all, he is quite casual in how he designates his sets of songs, not only using Li£t)ercyclud and Liederkreid interchangeably but also, for example, changing the designation of Frauenlie.be und Leben from "Cyklus in acht Liedern" on the set of piano drafts to, simply, "Acht Lieder" on the title page of the published edition. And while there are very few records of any performances of Schumann's
6
The Genre of the Cycle
songs and character pieces in his lifetime, a point to which 1 shall return later, the records we do have consistently describe the performance of single pieces and almost never a complete cycle, suggesting that the integrity of the whole was not as inviolable in his day as -we have made it today. There is little question, by the -way, that Schumann himself acquiesced and even approved of such performances, since his wife was frequently the performer. However, the most significant evidence we have that Schumann did not use the premise of a unified whole to distinguish his cycles from his other collections is the fact that there is no evidence. Given the centrality of the cycle to his compositional output, why is it that he never explains what cyclic unity should be like or even asserts the need for it? In fact, aside from a few passing references in his correspondence, there is scarcely any explicit mention of the cycle at all in his abundant writings on the music of his day.2 Admittedly, this does not provide us with conclusive proof that our conception of the cycle differs from Schumann's. But when we consider the difficulties we have had in applying our conception to his works, it should at least make us wonder -whether we are justified in clinging to it so tenaciously. I would like to propose an alternative conception of Schumann's cycles, which can be summarized as follows: The cycle is not generically opposed to the collection but is a particular kind of collection in itself, a collection that is composed of pieces -whose forms tend to be fragmentary and whose meaning tends to be obscure. The cycle does not create an overarching unity that provides such pieces with completion and clarity but is itself discontinuous and open-ended. The context that the cycle sets up is provocative; it implies structural connections and hints at larger meanings, but it never makes them explicit or definitive. We cannot know -with any certainty how Schumann himself defined and conceived of the song cycle, and in some ways this is just as well, since it relieves us of the temptation to simply identify ourselves -with him, rather than construct our own historical understanding. But if my definition of the cycle is not necessarily Schumann's definition, it is at least consistent with the scanty evidence we do have of his viewpoint and can be used to explain all of the -works that he explicitly published as cycles. Moreover, in defining the cycle as I have, I am specifically identifying it as a Romantic genre. I believe that Schumann's cycles provide us -with especially interesting examples of how the intellectual context of literary Romanticism influenced him as a composer. Among the enduring legacies of the Schlegels and their Jena colleagues is a new understanding of the role of literature and, by extension, of all the genres of art. They believed that, in contrast with the Classical -works of the ancients, which are complete wholes, the Romantic works of the modern age are inherently fragmentary and unfinished. This aesthetic stance is highly suggestive, and -we can find much evidence that it profoundly influenced Schumann and his associates. One implication, for example, is that the role of the audience in perceiving and comprehending works of art must change dramatically. Rather than remaining a passive receptor, who simply soaks up the meaning that is inherent in the work, the reader of Romantic literature and the listener to Romantic music
Introduction
7
must become actively engaged in the creation of that meaning, which will change not only from one era to another, or even from one performance to another, but also in each perceptive act by each individual person. The cultivation first of the lyric cycle by German poets and then of the song cycle by their musical colleagues is one of the most significant manifestations of this new conception in the early nineteenth century. If it seems difficult to come up with a universal definition for the Romantic cycle that will enable us to consistently explain how such works cohere, it is not simply because different cycles have different means of coherence, as many recent scholars have assumed. Rather, in composing such works Schumann and his contemporaries were experimenting -with the very question of aesthetic coherence itself and leaving it up to their listeners, each in his or her own fashion, to realize and develop whatever unifying meaning the songs of the cycle may imply. I believe that in seeking to explain such a work by trying to impose a fixed coherence on it we not only end up misunderstanding the work, but -we also undermine the imaginative and liberating aesthetic that gave it life. The expectation that their audience would play an active role in creating aesthetic meaning inevitably led the Romantics to severely circumscribe that audience, so severely, in fact, that it sometimes disappeared altogether. This phenomenon is well documented in the case of Schumann's early piano music, where it manifested itself both in the poor sales of his publications and in the difficulties faced by Clara Wieck in trying to program her fiance's music in her public performances.3 On the few occasions when she did perform his piano character pieces during the 1830s and 1840s, it was almost always in a private setting, such as her father's home, and she invariably selected a single piece or a small group of pieces from one of Schumann's cycles. One of the few accounts that we have of a complete performance of a piano cycle during this period is in a letter that Wieck -wrote to Schumann from Berlin on November 21, 1839. She describes playing his opus 21 NoveLletten for "several music connoisseurs." It is clear from her letter that this was an unusual occurence, since she does not simply -write that she played the Novelletten, or even the cycle o^Novelletten, but that she "played all of [his] Novelletten one after the other," and she prefaces her account by excusing the "absent-minded" quality of her writing -with an appeal to her understandable exhaustion from having accomplished such a pianistic feat.4 If Schumann's character pieces in and of themselves appealed only to a very small and sophisticated audience and were perfomed only rarely, then the complete cycles that he created out of those pieces were even more private and more inaccessible. One factor that motivated Schumann's sudden turn to lieder in 1840 -was undoubtedly his desire to publish in a more marketable genre, as several scholars have recently argued.5 Song publications -were very popular throughout the first half of the nineteenth century, and in contrast with his early piano music, -which must have been unplayable by most of the amateur musicians of his day, Schumann's songs -would have been easily -within their grasp. We also know, from the contemporary song reviews that -were aimed at these amateurs, that there was at least the expectation that a cycle of songs would be sung through from beginning to end.6 However, there are even fewer known performances of
8
The Genre of the Cycle
Schumann's songs during his lifetime than there are of his character pieces, probably because the lied •was considered a private genre. This does not simply mean that it was unsuitable for performance in large public concert halls but also in the smaller private spaces that still proliferated in the early part of the nineteenth century. Two 1844 reviews from Schumann's NeueZeifockrift/ur/tf example, specifically suggest that lieder should not be performed in the salon, a setting that has become emblematic of private music making in the nineteenth century: These songs will not make a noisy triumphal procession through the salons, but in a peaceful cell they will refresh quiet receptive souls with their unadorned gracefulness, their poetic fragrance. If this collection is not well suited to a salon recital — for which a real lied, and least of all the most soulful, should not be chosen—then it is especially suitable for the shortening of lonely hours or to the stimulating conversation of a close and refined circle of poetic souls/ Both of these reviews use Romantic rhetoric to depict the ideal setting in which lieder should be sung. The image of a peaceful cell, with its unmistakable allusion to monastic life, in the first passage and the reference to a circle of poetic souls, who are sharing intimate conversation, in the second both point to a communal and almost mysteriously private conception of lieder singing. These reviews do not necessarily tell us, of course, what the people who bought the songs of Schumann and his contemporaries actually did with them, but they do give us some insight as to how he and his colleagues conceived them. If on one level Schumann intended his song publications as a practical way of selling music and making a reputation for himself as a marketable composer, on another level they take their place among his most Romantic artistic creations, which are accessible only to the most refined and poetic sensibilities.
The Cycle as a Public Genre By the time that the earliest known definition of the synonymous terms Liederkreu and Liedercycliu appeared in Arrey von Dommer's edition of H. C. Koch'd mujikalutched Lexicon in 1865, the cycle had already started to decline as a compositional genre. The question has been raised as to why a definition of the song cycle suddenly appeared at this point, nearly fifty years after the terms were first used as designations for a published song collection.8 While there is certainly no single explanation, it is probably not coincidental that the first public performances of complete song cycles by Schubert and Schumann occurred a few years before the publication of Dommer's lexicon.9 As composers became less interested in the cycle, performers became more so. It was now a public genre, and as it became accessible to a relatively large audience for the first time, they presumably wanted to understand what they were hearing. Dommer does not de-
Introduction
9
fine the cycle in terms of the Romantic conception that I have outlined in the foregoing section. One reason for this is that the aesthetic ideals that influenced its creation no longer held sway at the time he wrote his definition. But another, related reason is that the Romantic view was no longer relevant once the cycle became public. In the late nineteenth century, as in Schumann's day and as in our own, it could not be expected that a general audience would have the capability or the desire to actively engage in the construction of musical and poetic meaning in the way that the early Romantics demanded. Attending a concert is typically treated as a passive experience, and for most audience members it is not enhanced by the feeling of uncertainty. And so begins the search for a definitive definition of the song cycle, which could contain it by clearly enunciating the conventions of the genre and find a place for it by comparing and contrasting the genre with others that are more familiar. In other words, -what was needed, once the song cycle began to reach a wide public, was a definition that could tell us what to expect 'when we heard one, a definition that could be used as an aesthetic standard. Dommer's reads as follows: Liederkreu, Liedercycliu. A coherent complex of various lyric poems, Each is closed in itself, and can be outwardly distinguished from the others in terms of prosody, but all have an inner relationship to one another, because one and the same basic idea runs through all of them. The individual poems present different expressions of this idea, depicting it in manifold and often contrasting images and from various perspectives, so that the basic feeling is presented comprehensively. As far as the music is concerned, it is certainly typical for each individual poem to be through-composed. A mam melody would essentially be retained for all of the strophes (of the same poem), and only altered and turned into something else where it seems suitable or necessary. Naturally, however, the melody and the entire musical form change with each poem, and so does the key, and the individual movements are typically bound to one another through the ritornelli and transitions of the accompanying instrument. The accompaniment is essentially developed so that it portrays and paints the situation in a characteristic way, and also supplies, in regard to the expression, what the voice must leave unfinished. In comparison with the dramatic solo cantata, the Liederkreu is actually missing nothing more than the recitative and the aria form of the songs \Ge<>ange\ instead of the lied form. Otherwise one finds it is rather close to the cantata, or regards it as a middle genre between through-composed lied and cantata.10 We can divide Dommer's definition into three component parts: a description of the text, a description of the music, and a comparison with the genre of the solo cantata. The fact that Dommer begins with the text and appears at first to be defining the cycle as a poetic genre is one element that connects his definition to the critical tradition of the earlier nineteenth century, in -which the poetic text is almost always considered to be of equal or greater importance than the
10
The Genre of the Cycle
musical setting. His account of the typical song cycle text should be familiar to us in a number of respects. He emphasizes the need for coherence, and he describes the tension between the integrity of the individual poem and its relationship to the larger -whole, as -well as the balance within the -whole between variety and unity. One striking omission, at least in comparsion with more recent definitions, is any reference to a narrative that runs through the poems and determines their order. But if Dommer begins his definition by describing the cycle as a lyric genre, he ends it by comparing it to a dramatic genre, the solo cantata. He tells us that the only difference between the two is that the cycle consists of lieder, -where the cantata alternates recitative and aria, in other -words, that the one is lyrical and the other is dramatic. This is not a minor difference, as Dommer implies, but an essential distinction between the two genres. \Ve can easily imagine how such a self-contradictory comparison might lead to confusion in the later reception of the song cycle. The most curious aspect of Dommer's definition is his description of the music, which bears no relationship -whatsoever to the song cycles of Schumann, Schubert, or any of the numerous Kleinmeuterd -who composed in the genre. One •would be hard-pressed to come up with a general description that would fit all or even most of the song cycles composed in the first half of the nineteenth century, as several scholars have observed, but it is significant that Dommer resolved this difficulty by turning to one of the earliest and most anomalous cycles—Beethoven's yl« die /erne Geiiebte. Although Dommer does not mention this work by name, his description records so many of its distinctive features that there can be no question that it served as his model. Each of the songs in An die feme Geiiebte, with the exception of the last, is strophic, with only minor variations in the melodies, but Beethoven varies the accompaniments extensively from one strophe to the next, so that they take on the role of characterizing the changing situations and provide the expressivity that the strophic melodies cannot. And, of course, the use of instrumental ritornelll and transitions is one of the most celebrated techniques that Beethoven employs in unifying his cycle. But since each one of these aspects is unique to Beethoven's cycle, Dommer's definition is completely useless for anyone who wants to know -what to expect in a typical nineteenth-century song cycle. There is apparently no precedent for Dommer's tacit reliance on An die feme Geiiebte as a model for the Romantic song cycle, but it has become the most enduring aspect of his definition. And, in an odd way, -we can see -why it would be tempting to use Beethoven's cycle and to emphasize just those elements that make it unique within the genre. What An die feme Geiiebte has, -which is lacking in virtually every other Romantic cycle, is a clear sense of unity and coherence. Schumann's cycles derive much of their expressive force from setting up an irreconcilable tension between the part and the -whole. The tonal ambiguity and open-ended formal structures that characterize his songs imply that they are incomplete parts of a larger form, but the discontinuity from one song to the next prevents the cycle from becoming a unified entity. In Beethoven's cycle, there is no such tension, both because the songs are so straightforward formally and
Introduction
11
tonally and because the transitions between them help us to easily subsume them within the whole. For this reason, An die ferne Getiebte is much closer to the continuous instrumental forms that provide us with the familiar exemplars of organic unity, and as a cycle it is more accessible. But if this helps to explain why An die feme Geliebte has persisted as an implicit model for the Romantic song cycle, it also gives us some idea of -why defining the genre has been so difficult. The Cycle as a Whole I will now skip ahead about a century, to the publication of Arthur Komar's famous essay "The Music of Dichterliebe: The Whole and Its Parts."11 On the one hand, Komar's assumption that Schumann's cycle "constitutes an integrated musical -whole" reveals the influence that the model of An die ferne Geliebte still exerted, and one wonders whether Komar would have -written such an essay in the first place if he did not have this example before him. On the other hand, he is no longer interested in the very specific aspects of An die ferne Geliebte that Dommer describes and never actually mentions Beethoven's -work in his essay. By the time Komar -was -writing it, An die ferne Geliebte had become less well known among listeners and performers and his chosen cycle, Dichterliebe, had become far more popular, so much so that today we define the genre largely in terms of our understanding of this single cycle. But despite An die feme Gelizbte'j increasing unfamiliarity, we continue to regard it, perhaps unconsciously, as the prototype for Dichtertiebe and for the Romantic song cycle in general. Komar's essay marks the beginning of the modern attempt to define the Romantic cycle as a genre. Earlier studies explain how a set of Schumann's songs or piano pieces is organized as a cycle, such as Rudolph Reti's analysis of Kinderszenen and Theodor Adorno's essay on the Eichendorff Liederkreu, both of-which were published in the 1950s.12 But neither of these studies had much of an immediate impact within Schumann scholarship, and, in fact, Reti's remains virtually unknown even today. It was only -with the publication of Komar's analysis of Dichterliebe, in 1971, that a sustained scholarly interest in the Romantic cycle began, an interest that has continued in some form or other up to the present day. Since my book takes its place within this scholarship and to some extent must be understood in relation to it, I think it will be useful to briefly outline its history. We can divide the recent scholarship on Schumann's cycles into three distinct but chronologically overlapping phases. At first, the cycle was considered in theoretical terms, as analysts tried to build on Komar's effort and establish a set of defining criteria that could be demonstrated through analyses of Schumann's -works. Then attention shifted to the construction of a -wider historical context, and scholars began to study the origins and development of the song cycle by surveying the works of Schumann's lesser known predecessors and contemporaries and reading -what nineteenth-century music critics had to say about them. Finally, -within the last several years, there has been a growing interest in -what I have already described as the intellectual context of the Romantic cyle,
12
The Genre of the Cycle
by which I primarily mean the way in which the creation and development of the genre resulted from the influence of literary Romanticism. It is only in the first, analytic phase that the model of organic unity has been explicitly identified as the defining premise of the genre, and, in fact, the desire to place Schumann's cycles within broader contexts originated in large part as a reaction against the organicist viewpoint. But while a variety of alternative ideas about the cycle have been explored by recent scholars, all of them inevitably end up returning to some •way of explaining the unity of the -whole. What Komar has in mind, when he describes Dichterliebe as an integrated musical -whole, is a Schenkerian voice-leading structure that extends through the entire cycle, a coherent plan that is created by the sequence of keys among the songs, and the presence of a single overarching key that unifies them. Since his essay appeared, subsequent analysts, such as David Neumeyer, Patrick McCreless, and Peter Kaminsky, have expanded on his conception by considering other unifying elements such as motive and poetic narrative and have applied it to a growing number of Schumann's song and piano cycles.13 These scholars have attempted to construct a theory of cyclic coherence by engaging in detailed analyses of individual cycles. In the process, they have employed a variety of analytic techniques and have learned a great deal about Schumann's music. Their 'work has led to the emergence of a common set of theoretical premises and a shared analytic terminology that has been used to explain these cycles as coherent musical entities. The explicit goal, in Kaminsky's "words, is to "define the term cycle in a way that distinguishes it from the collection." He goes on to explain how that distinction should be understood: We generally think of a collection as a set of independent, closed tonal movements whose integrity would not be destroyed if they were arranged in a different order or even transposed. For a cycle, on the other hand, we assume that some sense of unity flows from a coherent tonal and formal organization. We can see, in the definitive language of this passage, that the genre of the cycle has been defined far more clearly than it was in the nineteenth century. But we can also see how rigid and un-Romantic our definition has become, especially in comparison with the flexible and open-ended formal strategies that Schumann employs in his cycles. It is understandable that musicologists began to sense that what we were defining and explaining had somehow become quite different from what Schumann was composing and to embark on a more comprehensive study of the genre as it was understood in the early nineteenth century. But how do -we shed our twentieth-century assumptions and reconsider the historical material with an open mind? As Barbara Turchin points out in the introduction to her 1981 dissertation, still one of the most important of the recent attempts to write the history of the Romantic song cycle, this is not an easy proposition. Even the preliminary step of deciding which musical -works should be considered requires us to establish "a priori, definitions, categories and criteria by -which to judge the
Introduction
13
nature and characteristics of a song cycle." Establishing the generic boundaries of the early nineteenth-century song cycle is especially problematic, as Turchin was quick to discover, because the use of designations that clearly identify works as members of the genre, such as Liederkreu and Liedercyclud, was the exception rather than the rule. Most of the song publications that appeared in the first half of the century had titles such as "Zwolf Lieder" or "Sechs Gedichte," but many of the works that were given these innocuous designations were apparently intended as song cycles nevertheless. In order to find her bearings in this unmarked terrain, Turchin came up with the ingenious solution of reading reviews of song publications from various nineteenth-century music journals in order to see which sets of songs were considered to be cycles and how these cycles -were explained and described.15 Turchin's method led her to make a number of interesting discoveries about the history of the song cycle. She found, for example, that while Beethoven's An die feme Geliebte continued to be held in high regard throughout the nineteenth century, it -was far less influential than the Friihtingdlieder and Wanderlieder cycles of Conradm Kreutzer, which were composed at about the same time. Kreutzer's cycles -were imitated by many subsequent composers and were explicitly held up as models by contemporary critics. They bear little resemblance to Beethoven's, and, significantly, Kreutzer does not employ a cyclic return or connect the songs with piano transitions, and neither of his cycles is unified by an obvious tonal structure.16 Turchin also found that in defining song publications as cycles critics relied almost exclusively on the text, rather than the music, and described a variety of ways in which the poems of a cycle could be related, including not only a narrative sequence but also mood and theme. More surprising still, critics -were rarely concerned with the question of coherence or unity, and, on the contrary, they more often evaluated cycles in terms of the variety and contrast between the poems.17 All of these conclusions point us away from a single organicist model for the Romantic song cycle and suggest that in the nineteenth century the genre was far more fluid and open-ended than it is today. Other scholars have subsequently followed up on this suggestion and have argued that the definition of the cycle as a set of songs that is organically unified is too narrow and too anachronistic. Ruth Bingham, for example, has questioned the premise that there is any meaningful distinction between a cycle and a collection at all. She believes that our insistence on this distinction results from our underlying bias toward unity as an aesthetic standard and our consequent reliance on An die feme Geliebte as the ideal model for the song cycle. Bingham suggests that if we study early song cycles in greater numbers, then "the 'cycle-versuscollection' question may ultimately prove to be irrelevant, its main function to reveal how ambiguous the boundaries between the two are." In her own definition of the song cycle, Bingham replaces the term unity with the related term coherence, which has more wide-ranging connotations and is not tainted by the historical baggage of organicism. Her intention is to acknowledge that there are a variety of ways that song cycles may cohere and to avoid privileging one over another. Similarly, in a recent chapter on the nineteenth-century song cycle
14
The Genre of the Cycle
John Daveno writes that the genre "resists definition," in part because it is so variable, both in terms of the "choice and arrangement of texts" and in terms of the musical connections that may adhere between the songs. He, too, argues that the only defining criterion is "a demonstrable measure of coherence," and he concludes that scholars need to take a more ecumenical approach and "attempt to describe the nature and quality of this coherence as it manifests itself in individual cases."19
The Limits of the Historical Model The acknowledgment of generic diversity among the song cycles of the nineteenth century, the attention to lesser known works, and the attempt to understand the cycle from a contemporary perspective are all positive developments that have contributed in significant ways to our understanding of the genre. And yet, ultimately, I am not sure that the recent historical research has changed our conception of the song cycle as much as one might think. First, although the term coherence suggests a more inclusive model than unity, the similarities between the two are far greater than the differences, and -when scholars have explained what it is that they have in mind when they describe the song cycle as coherent it usually turns out to be the same kinds of relationships that have traditionally been associated with orgamcism. Second, while the evidence that Turchin and Bmgham have found is often at odds with our modern-day assumptions about the cycle, they both rely on those same assumptions in order to make sense of the evidence and inevitably find ways to make it fit. Bingham, for example, presents the history of the nineteenth-century song cycle as a two-pronged development in which first the use of texts that are associated by a common theme is replaced by the use of texts that have a coherent narrative and then the reliance on the text as the principal means of cyclic coherence ultimately gives way altogether to "musically-constructed cycles," in which coherence depends more on the music. By using this historical schema to explain the wide variety of song cycle types that were composed in the early decades of the nineteenth century, Bingham encourages us, perhaps unwittingly, to view them as developmental stages that led up to the song cycle that is familiar to us today, by which she means an organically unified whole that is modeled on Beethoven's An die feme Geiiebte. The central figure in this culminating stage of Bingham's history is Schumann, who borrowed techniques from his instrumental piano cycles in order to create musical coherence in his song cycles and thus helped to effect an "aesthetic shift from vocal mimetic to instrumental organic paradigms" within the latter genre. In other words, Schumann moved away from the model of the song cycle as primarily a setting of a text and toward the model of a coherently unified musical structure.20 As Bingham acknowledges, her explanation of Schumann's role in this history as a composer is closely related to Turchin's explanation of his role as a critic. The song cycle reviews that appear in the Neue Zeiljchrift fiir Mtuik in the 1830s and 1840s "mark a new stage," ac-
Introduction
15
cording to Turchin, because Schumann and his colleagues distinguish the cycle in just the •way that earlier critics fail to: by recognizing that it is a musically coherent entity in which the composer uses relationships between the keys of the songs to express the relationships between the poems.21 In his analysis of Du:bterliebe, Komar argues that the cycle is an integrated whole in large part because it has a coherent tonal structure. According to Turchin, this is the same element that Schumann considers to be the most important source of cyclic coherence. It is not surprising that she discovers Schumann sharing our modern conception of the cycle, since it is this conception that provides the initial premise for her study. She explains that her intention in reading nineteenth-century song reviews is to "reveal sets of songs, recognized as such by contemporaries, that exhibit features separating them from mere collections and that, therefore, should be considered in a history of the song cycle. other words, Turchin's conviction that there is a meaningful distinction between a cycle and a collection not only determines how she evaluates the reviews that she reads but also determines which ones she selects in the first place. She cites a handful of reviews from the Neu£ Zeitjchrift that mention the keys of songs, for example, but ignores the fact that the vast majority that appeared during Schumann's tenure are primarily concerned -with the expression of the text and avoid the subject of cyclic coherence altogether. And she assumes that a reference to the key of a song had the same significance to Schumann and his colleagues that it has to us today, even though the original context often argues against this assumption. In Schumann's 1836 review of Carl Lowe's Either, em Liederkreu in Balladenform, he mentions the keys of three of the five songs and comments that the last song returns to the key of the opening. All of these references occur •within a discussion of how Lowe responds to the events that are depicted in the text and expresses the "special tone" of each of the poems. But for Turchin, the fact that Schumann mentions the keys, as well as the "tonal relationship" between the first and last songs, is "especially noteworthy," and this review becomes a central piece of evidence for her argument that Schumann considers tonal coherence to be essential to the song cycle.23 It is not so much the conception of Schumann's cycles that has changed, in the wake of the research that Turchin and Bingham have done, as it is the kind of evidence that is used to support that conception. In the 1970s and 1980s, scholars primarily relied on the analysis of individual cycles to demonstrate the importance of musical coherence in defining the genre. More recent scholars have turned to the historical record and have used Schumann's own words to make the argument for them. We find an especially clear example of this in Daverio's recent biography. Commenting on the same review of Lowe's Either that Turchin discusses, Daverio -writes: Schumann noted the various means through which the composition attained textual and musical coherence — narrative consistency, large-scale tonal logic, and motivic recall—without insisting dogmatically that the presence of all these features is a necessary condition for cyclic construction.
16
The Genre of the Cycle
Daverio then uses the three criteria that Schumann mentions in his review to determine which of the sets of songs that he composed in 1840 and ISM] "qualify as cycles" and to explain how each of these cycles coheres. But while it is true that Schumann tells the entire story of Either in the course of his review and mentions both the use of the same thematic material in two different songs as well as some of the keys, there is no indication that he intends his comments as a definition of cyclic coherence or that he is even concerned •with the song cycle as a genre. In fact, there are several reasons to believe that this is not his intention. First, it is noteworthy that Schumann does not use his review of Either as a platform from -which to generalize about the song cycle, especially when we consider that one of the hallmarks of his criticism is the fact that he is often more concerned with the state of music and with the compositional trends of his day than he is with the piece he is ostensibly reviewing. Second, Schumann never reviewed another song cycle again and never even mentioned the cycle anywhere else in his criticism. In the case of other genres, such as piano music and the symphony, we can trace the development of Schumann's ideas through a series of reviews, which provides a coherent context to support our interpretation of specific comments. The one song cycle review that we have does not offer sufficient evidence for us to come to any firm conclusions about his view of this genre. But the absence of any explicit discussion of cyclic coherence or of any attempt to explain to his contemporaries how they should create such coherence strongly suggests that this was not a significant concern for Schumann. Finally, neither Turchin nor Daverio points out that the elements that Schumann comments on in his review of Either are not at all typical of either his own song cycles or those of his contemporaries. First of all, as Lowe's unusual designation for Either—ein Lifderkreu in BaLla3enform — suggests, this -work is actually a hybrid of two genres that were normally considered to be antithetical, and the fact that the text is a dramatic narrative is an element that marks the work as a ballad and not as a song cycle.25 Second, while the use of thematic return between the songs of a cycle is not unique to Either, it is very rarely encountered, and Schumann employs it in only two instances. And so, it is conceivable that in devoting so much attention to the narrative and dramatic aspects of the cycle and in pointing out the recurrence of thematic material Schumann is simply describing the distinctive aspects of this particular piece. The Cycle as a Fragment In his recent writings about Schumann's song cycles, Daverio defines the cycle as a coherent whole, whose coherence may result from either the presence of a narrative among the texts, a tonal structure created by the relationships among the keys of the songs, the recurrence of motivic material, or some combination of these three elements. But in a chapter of an earlier book, which is concerned with Schumann's piano cycles, he presents a radically different understanding of the cycle, an understanding that brings the very notion of cyclic coherence into question. Drawing a connection between Schumann's composition and Fnedrich
Introduction
17
Schlegel's conception of the Romantic fragment, Daverio suggests that Schumann intended many of his character pieces to have an unfinished or incomplete quality and even to have moments of incomprehensibility. When Schumann juxtaposes a series of such pieces within a cycle, he creates a larger structure that is likewise fragmentary and obscure, in which it is sometimes difficult to tell where one piece leaves off and the next begins. As Daverio considers the analytic implications of this approach to musical form, he argues that we should not try to "neutralize" moments of ambiguity and incomprehensibility in Schumann's music but explain them as "constitutive aesthetic qualities." He describes Schumann's cycles as "fragment clusters" and proposes that our task, as analysts and critics, is "to investigate the principles that differentiate [them] from discrete wholes on the one hand and chaotic jumbles on the other."26 In other -words, we need to abandon altogether the search for a definition of the cycle that is based on the premise of coherence and instead strive to understand how Schumann flirts with incoherence as he experiments with fragmentary and discontinuous musical structures. With this argument, Daverio moves beyond the historical context of Schumann's fellow song composers and music critics and strives to understand his cycles in broader intellectual and cultural terms, as musical examples of the Romantic aesthetics that originated with Friedrich Schlegel and the Jena circle. We probably owe the idea that we can explain the formal experiments of Schumann's cycles in terms of Schlegel's influence on the composer and, in particular, as a manifestation of the aesthetic of the fragment to Charles Rosen, who first presented it in a lecture in the 1960s but did not publish his discussion until 1995, when it appeared in a greatly expanded form as a chapter of his book The Romantic Generation?'7 As the idea has been developed, in very different ways, by Rosen and Daverio, it has presented the possibility of a completely new approach to the nineteenth-century cycle, which can help us to appreciate more fully the imaginative, liberatory, and quintessentially Romantic nature of this music.28 But as both of these scholars have turned to a concrete consideration of Schumann's cycles, they have not followed through on the implications of this idea and have instead relied on more traditional conceptions of cyclic form, based on the notion of unified coherence. In Daverio's case, we can sense his retreat even before he has finished arguing for a radical break with earlier scholarship. Having suggested that he is about to cast off the whole enterprise of searching for a definitive explanation of cyclic coherence, he then writes that he will "propose a typology for a particular body of musical fragments, and map out the categories that animate it as a 'system' of sorts."29 The three principal terms in this sentence — typology, categories, and system — are in fact the primary concerns of Daveno's chapter, and so, as he makes his way through Schumann's piano cycles, his purpose is not so much to confront their idiosyncratic and incomprehensible elements as to categorize and define them. What Daverio means when he describes Schumann's cycles as "systems of musical fragments," for example, is that, despite their apparent heterogeneity, they still have inner coherence, which we can only understand by uncovering "the hidden connections, the
18
The Genre of the Cycle
allusive links, the network of relationships" that Schumann has created. Daverio thus adopts a characteristic strategy of organicism, in -which apparent incongruities and discontinuities are accounted for as surface phenomena that simply disguise a deeper structural unity.30 And as he turns to his final, culminating example, Schumann's Novetletten, he uses the traditional technique of motivic analysis to reveal this unity. Although Daverio's discussion of the Novelletten touches on several issues—tonal structure, the fragmentary and digressive forms of the individual pieces, and the use of musical topics such as the march and the waltz — his basic argument is that the entire work is unified by two melodic figures, commonly known by the rhetorical names "circulatio" and "lament," that recur in various guises throughout all eight pieces.31 Where Daverio is primarily interested in analyzing the ways that Schumann unifies his piano cycles, despite their fragmentary nature, Rosen focuses on the individual song and piano piece and uses the fragment as a metaphorical model for Schumann's experiments with small-scale forms. Rosen observes that "a piece that begins in the middle or does not have a proper grammatical end" is the simplest of the techniques that Schumann and other Romantic composers use to expand the limits of musical form and to call into question "the established conceptions of what a work of music ought to be."32 One example that Rosen uses to illustrate the rich explanatory potential of the fragment is "Aus meinen Thranen spriessen," the second song from Dichterliebe. In contrast with the first song in the cycle, "Im wunderschonen Monat Mai," which begins and ends on a V7 chord, and is thus an obvious example of a fragment, "Aus meinen Thranen spriessen" "appears to be a separate, closed traditional structure that satisfies all of the formal requirements, -with a well-defined melody and V7/I cadence." And yet, as Rosen observes, Schumann uses textural ambiguity to undermine the phrase structure and, in particular, the finality of the cadence. Perhaps this song "makes independent sense on paper," but it would be "nonsense" if it were performed by itself: "not merely poorer in meaning and disappointing in effect, but puzzling and even inexplicable." It is largely through this paradox that Rosen explains why the fragment has such far-reaching implications for our understanding of Schumann's cycles. They "are made up of apparently independent songs -which cannot be independently performed," and for this reason they stand apart from both Beethoven's cycle, in which none of the songs have any pretensions to independence, and Schubert's, in -which all of the songs can stand on their own.33 But it is only -when Rosen turns to Chopin's opus 28 Preluded—a publication that has a far more tenuous historical claim to being a cycle — that he makes clear what the consequences of the fragment are for the notion of cyclic unity. He describes the Preluded as "the most impressive example of a set of tiny Fragments," •which "achieves unity apparently through the simple addition of one piece to another. "He acknowledges that the modern practice of performing the opus as a complete cycle "was not thinkable during Chopin's lifetime" and even concedes that such a performance "does not allow us to fully appreciate the extraordinary individuality of the single numbers." And yet, Rosen argues, "the aesthetics of
Introduction
19
the fragment would suggest that the opposing demands of the opus as a whole and of each individual prelude are intended to coexist -without being resolved." With the Preluded, Chopm presents a "conception of unity that transcends any possible mode of presentation."34 It is not at all clear that Rosen intends his observations about Chopin's Preluded to apply to Schumann's song cycles. In the latter case, the works were given published designations that advertise their unity as a cycle, and contemporary listeners presumably did not question the appropriateness of a complete performance. However, as I have observed, there were no complete public performances of any of Schumann's cycles during his lifetime, and we actually know surprisingly little about what was considered a "possible mode of presentation" for these works. I believe that the irreconcilable tension between part and whole that Rosen describes in Chopin's Preludes has significant implications for our understanding of Schumann's cycles as •well. It is not so much a question of whether or not there is a unity or coherence that extends beyond the individual song and turns the cycle into a whole. Rather, it is a question of how the ambiguous status of the song leads us to expect and even yearn for the context of a larger whole within \vhich we can make definitive sense of it. The cycle clearly suggests such a context but does not explicitly provide it, and so we, as listeners, are left to imaginatively realize the coherence of the cycle on our own. I hasten to point out that while the comments in the preceding paragraph •were stimulated by Rosen's discussion of the aesthetics of the fragment and, at least to my mind, represent a logical extension of that discussion, they are my own thoughts and not Rosen's. One might expect that Rosen would address the issues that I have just raised in the folio-wing chapter of his book, when he turns from the fragment to the song cycle, but he does not. When it comes to defining the cycle as a genre and explaining the role that Schumann plays in the history of that genre, Rosen turns all the way back to the most traditional viewpoint. He is quite definitive about the fact that it is Beethoven's An die feme Geliebte that should be considered the first song cycle, because other contemporary sets of songs lack its "unity, cohesion, [and] power." In describing the latter works as "only loosely related groups of songs that do not even pretend to a more impressive status," Rosen cites the familiar distinction between Beethoven's song cycle—the first true cycle—and the mere collections of his contemporaries and explicitly makes the value judgment that is inherent to this distinction. With the song cycle, he writes, "a modest genre, intended largely for the unambitious amateur, becomes a major endeavor that in weight and seriousness rivals grand opera, the Baroque oratorio, or the Classical symphony."35 It would probably never occur to an early nineteenth-century musician to make such a comparison, since it ignores the very real differences that obtained between a private genre such as the song and the public genres of the symphony and the opera. By holding up the latter genres as his points of comparison for the cycle Rosen makes it clear that he is abandoning the aesthetic sensibility of the Romantics and instead relying on the evaluative criteria of the modern day to define the cycle as a genre that is -worthy of our attention.
20
The Genre of the Cycle
In fact, Rosen tacitly rejects the recent historical research that has been done on the cycle entirely and presents the works of the three renowned masters in a historical vacuum. And so, as he turns to Schumann's cycles, the only possible models that he can cite are Beethoven's and Schubert's. Perhaps it is for this reason that Rosen defines the cycle so narrowly that only two of Schumann's song publications qualify—Dichterliebe and Frauenlizbe and Leben. What distinguishes these two works from all of Schumann's others is the fact that each ends with a thematic return, just like/1/2 (tiefeme Geliebte, and, even more important, that their texts present a unified narrative, with a single protagonist, which determines the order of the songs. Rosen places Schumann's other sets of songs in a separate category, using the designation that Schumann himself gave to them—Liederkreui. Rosen defines a LiederkreiJ as "a set of related songs, often with a literary and musical structure that holds them together," but without an immutable narrative. Rosen's category of the LiederkreLf is thus less rigid than his CycLiu, but he still relies on tonal and motivic unity, both elements of the organicist paradigm, in order to define it.36 His retreat from the aesthetics of the fragment to the model of organicism is even more abrupt and more complete than Daverio's. Because he is able to explain so clearly why the fragmentary nature of the individual song has such powerful implications for our understanding of the cycle, it is all the more inexplicable that Rosen abandons this idea and returns to the narrowest possible conception of cyclic coherence.
A New Conception of the Cycle There is a distinctly Sisyphean quality to the history I have just sketched out. Each succeeding phase of scholarship has brought us further and further toward a clear understanding of the Romantic nature of the song cycle but has then rolled back at a crucial moment to more traditional models. -My impetus in writing the present study is the hope that I can manage to push the boulder just a tiny bit further and find a new perspective from which to understand the genre. My study does not really belong to any one of the phases of cycle research that I have described but draws on each of them at different points and for different reasons. The first part of my book is partly a critique of the modern-day conception of the song cycle and partly an attempt to better understand how the genre was conceived by Schumann and his contemporaries. My intent is not to suggest that -we simply abandon our current viewpoint and certainly not that we try to return to the nineteenth century but rather that we consider how far our viewpoint is from Schumann's and question how -we might further develop our conception so that it is more sensitive to the Romantic aesthetic that led to the emergence of the song cycle in the first place. In part II I present a detailed analytic study of a single work — Schumann's Eichendorff Liederkreu—that is intended to exemplify my argument. This part of the book differs from the analyses by Komar and other theorists who have been interested in defining the song cycle in that I primarily focus on the individual song and do not consider the complete cycle as a whole. I am especially concerned with the literary aspect of
Introduction
21
Schumann's cycle in this part, and I take Eichendorff's poetry as the point of departure for my analysis, treating the song in large part as a musical response to the poem that it sets. In the third and final part of my book, I address the question of how Schumann creates a cycle out of a set of songs. Although the Eichendorff Liederkreu is still my principal example, I now consider it within the context of the entire corpus of song cycles that Schumann composed between February 1840 and January 1841. I begin this part with a chapter in which I discuss the three volumes of song manuscripts that Schumann compiled during his so-called year of song and use the evidence found in the manuscripts to argue that he consistently employed a two-stage process in creating his song cycles: first composing the songs and then arranging them into a cycle. In the final chapter, I consider the question of how the text of the nineteenth-century cycle is typically organized. I argue that as Schumann turned his Eichendorff songs into a cycle, he created a coherent literary text that is a representative example of a Wanderlieder cycle. I should probably offer a few words of explanation about the second chapter of my book, in -which I discuss the analytic work that has been done on Schumann's DichterLie.be. Why do I devote an entire chapter to this cycle, given that the main subject of my book is the Eichendorff Liederkreiil And why do I focus exclusively on the work of Schenkerian analysts, which is now many years old and has already been superseded by more recent approaches? I can answer the first question by referring back to the beginning of this introduction and reiterating that one of my primary concerns is the genre within which we classify the Eichendorff Liederkreut, the Romantic song cycle, which means that in some sense I must begin -with Dichterliebe. If An die feme Geliebte is the invisible ideal by which we understand Dichterliebe, then Dichtertiebe is the very visible model by which we understand and evaluate the song cycle in general. This is as much the case for so recent a writer as Rosen, who devotes almost his entire discussion of Schumann's song cycles to this one work and practically defines the genre in terms of it, as it is for earlier writers such as Komar and McCreless. In the present study, I am trying to consider what the genre would look like if the Eichendorff Liederkreu were taken as a paradigm, rather than as a problematic and exceptional work. I am not proposing that -we replace DicbterUebe with the Eichendorff Liederkreui but that -we expand our conception of the genre so that both works—as well as Schumann's other, neglected cycles — are equally at home within it. And so my point of departure is the paradigm that we have constructed out o[ Dichterliebe, and this is not only the case in the next chapter, where I consider the cycle as an organically unified whole, but also in part III, where a reconsideration of the extensive scholarship that has been devoted to the study of the manuscript sources for Dichterliebe, as well as the interpretation of its text, becomes the starting point for my discussion of how Schumann creates a song cycle. As to -why I devote so much attention to the work that Schenkerians have done on Dichterliebe, in particular, it is because I believe that this work has been far more important and influential than many current scholars are willing to acknowledge. As I have argued already, the shift in recent scholarship on the cycle
22
The Genre of the Cycle
from organic unity to coherence has not been all that consequential, and in fact, the two concepts are often indistinguishable in practice. If we really want to move beyond the organicist paradigm, then we need to confront it directly, which means understanding the logic and the motivation behind it and explaining why it does not provide an adequate account of the song cycle. We must engage in a thoughtful critique of the various Dichterliebe analyses and especially of Komar s essay, which has never been accorded the consideration it is due. It has lately become fashionable for musicologists to simply dismiss Schenker as an outmoded formalist, whose theory is not really relevant to the more culturally and historically informed research that has come to dominate the field. I believe that such an attitude is both shortsighted and unrealistic. In the case of Schumann's song cycles I agree that one of the greatest limitations of both Komar and Schenker is the fact that they present self-contained analyses that are exclusively concerned -with musical structure and do not take into account either the poetry that Schumann was responding to or the larger context of Romanticism that influenced and animated his composition. However, I believe that one of the most interesting and important means that Schumann uses to respond to the poems that he sets is his manipulation of musical structure and his play with formal conventions, and it is in this way that the influence of Romanticism most clearly manifests itself in his music. Schenker's theory provides a powerful tool for understanding the structure of tonal music, and we should not let our frustration with its very obvious limitations prevent us from taking advantage of the compelling insights that it has to offer. The main reason that I have written a chapter about Schenkerian analysis and Dichterliebe is to explain why this approach does not seem to work very well for this piece and, by extension, for all of Schumann's cycles. But another reason is to argue that we can still learn a great a deal from the work that has been done in this direction. When I turn to my own analyses of the Eichendorff songs, in the second part of my book, I will draw freely on a variety of techniques, including Schenkerian voice-leading analysis, in order to illuminate the specific •ways in which Schumann uses musical form to respond to Eichendorff's poems. I intend these analyses as an illustration of my larger argument about Schumann's creation of his song cycles. Most of the poems that Schumann set in 1840 and 1841 are short lyrics by contemporary poets such as Heine, Eichendorff, Kerner, and Riickert. Schumann conveys the Romantic qualities of this poetry by stretching the limits of musical form and creating songs that are at one and the same time integral •wholes and fragmentary parts. He typically composed in a concentrated burst of activity over a few days or -weeks, focusing on a small group of poems by a single •writer, and then arranged the songs for publication by deciding on their order and in some cases omitting or substituting songs. In this respect, we can describe the cycle as a context that helps us make sense of the individual song, not because it is a coherently unified whole but because it is a collection of similar things. August Wilhelm Schlegel offers a compelling metaphor for Schumann's songs when he argues that a Romantic •work of art should be like a sketch.37 What Schlegel has in mind is that the sketch is more a stimulus that we must continue
Introduction
23
to develop with our imaginations than it is a completely finished work that we can view passively. Schumann composes songs that are unfinished in this same way, by creating ambiguities that remain unresolved and setting up implications that remain unfulfilled. Each of his songs is a coherent musical form, but he demands that we rely on our familiarity -with the conventions of tonal music to make explicit in our imaginations the musical meanings that he merely suggests in sound. Schumann's songs are comparable to sketches in another sense as well. In composing a series of brief pieces in a short span of time, Schumann often took up a small set of compositional ideas and worked through them in a different way in each song. So in Dichterliebe, for example, he concludes many of the songs with postludes that are so lengthy that they alter the formal structure of the song. And in several songs from the Eichendorff Liederkreuf Schumann weakens the opening tonic harmony and establishes tonal stability only gradually. We can compare each of these cycles to a series of sketches in which an artist quickly drew the same scene from different perspectives. It is sometimes easier to make sense of each sketch in such a series when we view the entire group together than it is when -we encounter a single sketch by itself. One reason for this is that the similarities that naturally arise in a set of sketches drawn in quick succession can help us to understand the pattern of stylistic elements that create meaning in each of them and thus to read what Schlegel refers to as the artist's "cipher-language of lines and forms." I believe that a similar process is at work when we listen to one of Schumann's songs within a performance of a complete cycle. The surrounding songs do not clarify the formal ambiguities or satisfy the frustrated expectations that can make the individual song so challenging to listen to. But as we hear the same kinds of ambiguities in a series of such songs, we begin to relax our expectations and to accept the uncertainties that each song creates. In proposing this new model to explain how Schumann creates song cycles I am not suggesting that we simply abandon the more commonly accepted premise of coherence altogether but that we reconsider the way in which it has typically been construed. Daverio's list of the three possible types of cyclic coherence— narrative, tonal structure, and motivic recurrence — encompasses the definitions of virtually all the scholars -who have written on the song cycle in recent decades. In all three cases, coherence is understood as something that Schumann has consciously and deliberately created, which is immanent to the cycle in a definitive form. According to this model, our role as analysts is to uncover the relationships that make the songs of the cycle cohere and explain how the cycle is a complete whole, and it is for this reason that studies that are based on the premise of coherence are largely indistinguishable from those based on organic unity. I believe that the notion of such definitive coherence in the song cycle is chimerical and that any coherence that we do perceive is more the result of the inevitable relationships and similarities that we would expect in a group of songs that set the same poet's texts and were composed at the same time. There is evidence that as Schumann arranged groups of songs into cycles he carefully considered how to emphasize the relationships among them. Schumann sometimes spent more time deciding on the order and even on the contents of a
24
The Genre of the Cycle
cycle than ke spent composing the songs in the first place. But the fact that he typically began the process of arranging the songs into a cycle after he finished composing them and, even more important, so often changed his mind as he engaged in this process makes it clear how mutable the order and contents of his song cycles are. The order of the songs in a published cycle reflects the aesthetic choices that Schumann made as he considered how to convey the various levels of poetic and musical meaning most effectively, but this does not mean that he has created a unified tonal structure or a consistent narrative discourse. On the contrary, the complete cycle is as fragmentary and open-ended as the individual songs of which it is comprised, and its ultimate coherence and meaning are recreated anew by each individual listener. Perhaps this is why the attempt to define the genre of the song cycle has been so maddening.
ANAI^ZING DICHTERLIEBE
D
\ichterliebe, which Schumann composed in just over a week at the end of May 1840 and published four years later as his opus 48, has become not only the most popular of his works but also the most widely known German Romantic song cycle, to the extent that for many people it has come to embody the genre. This is certainly true among performers and listeners, as a cursory glance at virtually any volume of the Schwann catalog will attest, but also among scholars, for whom Dichtertiebe has long exerted a powerful allure. There has been more published about Dichterliebe than about any other work of Schumann's, including the only full-length source study of any of his vocal works and the only comprehensive analysis of a song cycle by any nineteenthcentury composer. As is typical for a composition that has come to occupy such a commanding position within the musical canon, there are undoubtedly many reasons for Dicbtertiebe'^ continuing popularity. My concern, however, is not so much with how Dicbterliebe has come to be the preeminent Romantic song cycle but with how the reception of this work has come to define the genre. The assumptions and expectations that scholars have brought with them to the study of Dickterliebe have more often than not been at odds with -what they have found, and yet the same expectations and assumptions have still managed to live on and, if anything, to become even stronger after each encounter with the musical work. The story of Dicbtertiebe '<> reception, then, is in some sense the story of how -we have come to misunderstand the Romantic song cycle. And if we are to attain a new understanding of this elusive genre, then we must begin by directly addressing the problems that have beset its most famous member. Although I will consider both the compositional history of Dichterliebe as well as the interpretation of its text in part III, for the moment I am restricting my discussion to the work of three scholars, all of whom can be identified to a greater or lesser extent as Schenkerian analysts. As I explained in my introduction, I believe that the attempt to define the song cycle in terms of organicist theories of musical structure has strongly influenced the subsequent course of Schumann scholarship, but this influence has never been examined or acknowledged. The problem of organic unity in Dichteriiebe is really two separate problems, one involving the cycle as a whole and the other involving the individual 25
26
The Genre of the Cycle
song. On the largest scale, it is a daunting task to try to explain a cycle of sixteen discrete songs as a single musical structure that is organically unified. How do -we account for the lack of continuity from one song to the next or the lack of any one key to provide a tonal context? In fact, when we consider the problem objectively it seems incredible that we ever came to believe that the song cycle could ultimately be understood as an organic structure. On the smallest scale, there are several songs -within Dichtertiebe that present analytic difficulties because there is some way in which their structures seem incomplete. This is certainly not true of all of the songs, nor even of most of them, but it is true of those songs that have received the greatest scholarly attention, including, of course, the first two. We tend to regard these songs as the most typical of Dichtertiebe and of Schumann's style in general, because they are his most distinctive. But I suspect there is another reason that these particular songs have proven so fascinating to analysts. The individual song that lacks a complete structure by itself holds out the promise that it can ultimately be explained as a small part of a larger structure. And so the song cycle sets up a kind of vicious circle for the analyst: we try to solve the problem of organic unity in the song by seeking it in the cycle, and the reason we believe that we can solve the problem of organic unity in the cycle is that it is lacking in the song. As I take up the work of my three chosen analysts, I will consider both their successes and failures in explaining how Dicktertiebe, in part or in whole, is put together. My primary concern, however, is not so much with the particular analysis that each of them presents but with the theoretical basis from which the analysis is derived and the assumptions that lead each of them to take the particular approach that he does. I -will begin -with a critique of Arthur Komar's essay in -which he undertakes the ambitious project of analyzing DichterLiebe in its entirety, using Schenker's analytic theory as the basis for his attempt. Next I will turn to an analysis by Schenker himself of a single song from the cycle, an analysis that has become at least as influential as Komar's. Finally, I -will consider an article by David Neumeyer in -which he considers some of the implications of both Komar's and Schenker's approaches to Dtchterliebe and ends up getting caught in the vicious circle that I described earlier.
Komar Komar's essay "The Music of DichterLiebe: The Whole and Its Parts" has had a seminal influence on the study of Schumann's song and piano cycles, and it remains today as the most comprehensive analysis of a such a -work that has been attempted. As a part of the prestigious series of Norton Critical Scores, it is still in print, nearly thirty years after its initial publication, and it continues to be •widely read by both students and professionals. And yet the high profile that Komar's essay has enjoyed has ironically denied it the response that I suspect he himself would have -wanted for it. I believe that Komar intended his essay more as an attempt to raise questions and to stimulate debate about the cycle than as a finished theory of cyclic form. However, despite some scattered criticisms,
Analyzing Dichterliebe
27
Komar's study is largely accepted as just such a theory, and nobody has seriously responded to the problems that his analysis raises or addressed the apparent contradictions in his conception of the cycle. On the contrary, -what he so carefully qualifies as the "seven possible conclusions to -which our investigation might lead" have become reified as the seven steps to "cycle-hood."1 The basic problem that Komar raises can be restated as follows: how is it possible for a work that consists of sixteen separate pieces in a variety of keys to be understood as a single "integrated musical totality"? Komar sets out to solve the problem by expanding Schenker's theory of musical structure beyond the boundaries of a single piece or movement and applying it to an entire cycle of songs. In the course of his analysis, Komar presents a number of compelling insights about the songs of Dichterlie.be, but he ultimately fails in his attempt to demonstrate that the cycle has a unified structure that can be analyzed in organicist terms. We can identify three main reasons for this failure. First, his list of possible conclusions is really a list of criteria for cyclic coherence that is intended to be cumulative and interdependent, but different items on the list are based on incompatible theoretical premises. Second, it is ultimately impossible to demonstrate that Dichterlie.be is in a single key, yet this is the most basic requirement for organic unity in tonal music. Third, although he explicitly derives his analytic procedures from Schenker, Komar also diverges from Schenker's theory in fundamental ways and ends up rejecting those aspects that are most closely associated -with organicism.
Key Relationships Komar presents his list of possible conclusions hierarchically, with each category representing a higher degree of coherence than the preceding one. So, for example, his first conclusion -would be that "the Dichterliebe songs belong together just as much as—but no more than — any other collection of Schumann's early brief love songs," while the relationships that form the basis of his second conclusion "exceed the coincidental similarities to be found among any random collection of mid-nineteenth-century songs, and therefore they must be regarded as individual features of this particular collection." What Komar has in mind is the use of the same "thematic segments, harmonic progressions, and rhythmic figurations," in other words, stylistic and motivic relationships. In his third category, "the songs are related as in category 2, in addition to which some of the similar pitch configurations are untransposed." And fourth, adjacent songs may be paired "by elements of local continuity" (63—64). Although, as Komar notes, this is the first of his categories in -which the order of the songs becomes a determining factor, the kinds of relationships involved are the same ones that prevail in the preceding categories. He is not proposing a structural relationship, in -which two songs are wedded into a single musical entity, but a purely transitional connection that is created by motivic means. It is only with his final three categories that Komar begins to consider the entire cycle as an integrated -whole. His language becomes explicitly orgamcist
28
The Genre of the Cycle
at this point, and he reveals that his categories are not merely a series of separate conclusions that one might possibly come to about a given set of nineteenthcentury songs, as he suggests at the outset of his list, but a cumulative succession of criteria for creating cyclic unity, each of which encompasses all of the preceding categories. In fact, the interdependence of Komar's categories turns out to be a crucial element in his argument. He describes the fifth category, for example, as "a coherent key scheme, as indicated in Dichterlie.be by the frequency of descending-fifth and -third motions." But he perceptively adds a parenthetical caveat that has come back to haunt many subsequent writers who have tried to explain how the cycle •works as a musical entity: I should hesitate to place too much emphasis on key relationships alone, for one could put together an equally "coherent" song collection merely by picking any sixteen songs in the same ordered set of keys. (65) Given that this is the case, how meaningful can the succession of keys within a song cycle be? Komar answers his own objection by arguing that it is "in conjunction -with" the preceding three categories that the key scheme "represents a significant integration of compositional elements throughout a song cycle." But Komar's fifth category really amounts to a claim that DichtertLebe has some kind of meaningful tonal structure, where his earlier categories are based on motivic relationships. These two elements can clearly coexist within the same piece of music, but in theoretical terms they are incompatible, since motivic relationships do not depend upon tonality. I can most clearly demonstrate this problem in Komar's analysis by considering a single example that involves a small group of songs. Komar observes a number of associations between Song 3 and the two songs that precede it. He notes "the use of the same four pitches at the start of the vocal line, the explicit treatment of Ct as descending passing note between D and B, . . . and the use of B as appoggiatura at the end of the vocal part," all of which fits into his third category. He also observes that "the hanging B at the end of the voice part of Song 2 resolves vocally in the upbeat A of Song 3," an example of his fourth category (73). Komar's only observation that is clearly related to key succession comes in the last sentence of the analysis: Moreover, the tonicizations of Song 3 reflect the key directions both before and after Song 3: I am referring to the sequential progression that starts in A major in m. 9, proceeds to D major in m. 10, and arrives at G major in m. 12. (74) There is no attempt to relate this latter observation to the earlier ones, so it is not clear how Komar's various categories are -working "in conjunction." But I cannot see how such an observation would be significant in any case. The descending fifth relationships among the keys of Songs 2, 3, and A that Komar is calling attention to are typical of Schumann's cycles. The sequential progression that he describes in Song 3 -would be unremarkable in any song in the key of
Analyzing Dichterliebe
29
D major, since the harmonies involved are I, IV, and V in that key. If anything, this demonstrates the kind of illusory "coherence" that Komar warns of, since we should expect that in a series of songs whose keys are related by fifths, we will find harmonic progressions within the songs that mirror the key relationships among the songs. This does not mean, of course, that the succession of keys within a song cycle is insignificant or fortuitous. There is evidence that Schumann himself paid careful attention to this question. On a page in which he wrote out the order of the Myrthen songs, for example, he includes the title, the poet, and the number of pages for each song and, for the first nine songs, the keys as well. And at the bottom of the piano draft of "Es leuchtet meine Liebe," which was originally intended as the fifteenth song of Dichterliebe, Schumann reminded himself of the four songs that were to follow by writing a list of their keys.2 The point, however, is that the preference for key progressions by fifth and by third is a matter of style, not structure. It contributes to the aesthetic sense that a set of songs belongs together, and if they are performed in succession, then it helps to create a smooth transition from one to the next. But it cannot create the kind of organicist coherence that Komar argues for, and in fact, Schumann was just as careful about key succession when he published collections of unrelated songs. Komar does not really define the sixth category in his list, writing simply: "A general compositional plan embraces all of the songs of the cycle in their given order." He does specify two conditions for this category, however. It must determine the order of each song and the point at which the cycle concludes (65). What Komar apparently has in mind here is -what he later refers to as the "harmonic plan" and the "modal plan," which work together to create the "overall form" of Dichterliebe (77). It is clear from Komar's description of these plans, however, that they are really no more than elaborate patterns of key relationships, so there is no difference between this category and category 5. Furthermore, there is a crucial exception to the modal plan, as it unfolds in Dichterliebe, and an inherent inconsistency in the harmonic plan, so that neither really fulfills the basic requirements that Komar himself sets for his sixth category. Komar describes the structure of Dichterliebe in terms of an underlying tonal progression from A major, the key of the first two songs, to C(t minor, which is the key in which the last song begins. The sequence of keys initially moves through B minor, in Song 5, to C major, in Song 7', a progression that Komar describes as "analogous to a deceptive cadence." The same motion then begins again, but now from A minor, the key of Song 8, to Bl> major, in Song 12, back to B, now major, in Song 14, and finally to Ct minor in Song 16. To explain how this ascending stepwise key progression forms the basis of the entire cycle Komar invokes his harmonic and modal plans. The harmonic plan dictates that a given note is treated in successive songs first as tonic and then as dominant and is then displaced by a note one step higher, "which simultaneously or eventually becomes the new tonic."4 The modal plan dictates that the mode of each song is derived from "the given modes of the diatonic scale degrees" in the preceding song. Hence, after A major, the cycle moves to D major, not D minor, and then
30
The Genre of the Cycle
to G major, B minor, E minor, and so on (77—79). I have illustrated how the two plans operate in Dichterliebe in Example 2.1.5 By dictating the key and mode of each song within the cycle, Komar's plans are intended to satisfy the two conditions of his sixth category: explaining the order of the songs and the point at which the cycle concludes. But because the harmonic plan allows two possibilities for each successive note—it may become a new tonic either at the moment that it displaces the previous note or at a subsequent point—the plan does not really determine the position of each song. B displaces A in Song 4, but only becomes the tonic in Song 5, and Bt displaces A in Song 10, but only becomes the tonic in Song 12. However, C, Ct/B, and Dt/Ct each displace the preceding note and become the tonic in the same song. One could thus omit Song A or Songs 10 and 11 and could insert extra songs before Songs 7, 14, or 16, and in none of these cases -would the harmonic plan be disrupted. An even more serious problem for Komar is the fact that it is only through a digression in the modal plan that the last song is in Ct minor. The plan dictates that since Song 12 is in Bl> major, Song 13 should be in El. major, rather than minor. Had this been the case, then the key sequence of the cycle would continue as follows: C minor, F minor, Dl> (or Ct) major. As we shall see shortly, it is essential to Komar's ultimate argument that Dichterliebe end in Ct minor, rather than Cf major, -which means that his entire edifice hangs on this one crucial exception to his modal plan. But if it is only because of such a digression in the plan that the cycle concludes where it does, then one cannot really say that the plan has determined this conclusion. Key Unity There is a certain elegance to the form that Komar constructs out of DichterUebe'j sequence of keys, but it does not have the explanatory force that he ascribes to it, both because it lacks the "implication of inevitability," as Barbara Turchin phrases it,6 and also because it does not unify the cycle in terms of any existing theoretical model of tonal music. Komar himself tacitly acknowledges the limitations of his harmonic and modal plans by including a seventh possible conclusion to his study: "All the features of category 6 are present, and in addition, a single key governs the entire -work" (65). It is curious that this final category, presumably Komar's crowning claim, is the one criterion that has been consistently rejected by later writers, even when they are overtly adopting the rest of Komar's list as the basis for their own work. Patrick McCreless, for example, argues that the insistence on a single key "invokes an eighteenth-, rather than nineteenth-century requirement for tonal coherence." David Neumeyer and Peter Kammsky both observe that the variety of tonal structures among Schumann's cycles makes it difficult to generalize from a single -work, and for Neumeyer, the fact that at least two of them begin and end in different keys suggests that Schumann himself did not associate "key unity -with an organic or cyclic structure." Ironically, Dichterliebe itself is one of Neumeyer's counterexamples.7
Analyzing Dichterhebe
31
EXAMPLE 2.1. Komar'j harmonic and modal plaru in Dichterliebe, adapted from Dichterliebe, A Norton Critical Score, by Robert Schumann, edited by Arthur Komar. Copyright © 1971 by W. W. Norton ej Company, Inc. U
Komar is clearly aware that his attempt to demonstrate that the sixteen songs of Dichterliebe are governed by a single key embroils him in a fundamental contradiction. He observes that the outer movements of tonal instrumental works are typically in the same key, which determines the possible keys of the inner movements. And he acknowledges that the wide range of keys in DichterLiebe, "and in particular. . . the different keys of the outer songs, . . . reveal the lack of an obvious single key controlling the entire work." But Komar insists, nevertheless, that the songs of the cycle "are interdependent 'movements' governed by a single key," presumably because he understands that -without key unity, the notion of organicism in tonal music collapses (65—66). In order to prove that Dicbterliebe is "an integrated musical whole," he must show that there is a single key that governs the work, and so it is ultimately on the strength of this one claim that we can evaluate Komar's entire argument. We can break down Komar's approach to the problem of tonal unity in Dichterliebe into two basic components. Analytically, he tries to expand Schenker's theory of musical structure and apply it to a cyclical work that consists of several movements or shorter pieces. Intellectually, he argues that the key of the opening piece "controls the final events of the cycle," even though there is no piece that is actually in this key anywhere near the end. These two aspects of his argument, which are interdependent, are both based on an implicit analogy between a voice-leading progression, considered within an individual piece, and a sequence of keys, considered -within a cycle of pieces. The problem -with this analogy is that there is no continuity from one piece to another, and thus the key succession does not operate according to the syntactical rules that govern counterpoint and harmony. Because -we hear each piece within its own tonal context, a cycle cannot attain the kind of musical unity that makes a voice-leading progression audible and intelligible. Komar would have this problem even if he were analyzing a cycle that begins and ends in the same key, but Dichterliebe strains the metaphor even further, because he must somehow explain that a cycle that begins in one key and ends in another is analogous to a tonal piece of music that is in a single key. The metaphorical relationship that Komar has set up leads him to talk about keys as if they were harmonies and tonics as if they were actually pitches within the scale of some other, more all-encompassing key. He explains the conclusion of the cycle by simply stating that "in tonal pieces, the final chord is usually I; in
32
The Genre of the Cycle
Dichterliebe it is III." ^Ve are not really to understand Cf as a tonic, then, but as the root of a harmony within the key of A major, albeit a harmony that is tonicized. Komar acknowledges that the presence of Et in the postlude of the final song "tends to deny locally the structural priority of A major." But he insists that "there is no actual doubt about the subsidiary relationship of Ct to A" because, among other reasons, "the final approach" to the key of Ct is not from Bt, but from Bll, which is first the tonic and then the dominant in the two preceding songs. Komar can thus argue that the scale of A major "controls" the end of the cycle, because the tonics of the last three songs — B, E, and Ct—are all members of that scale (80). But let us consider for a moment how such a theoretical conceit relates to the -way in •which we, as listeners, actually perceive the songs o^Dichtertiebe. Each time a song ends in one key and the next begins in a different key, we receive a number of aural cues that encourage us to listen in terms of a completely new tonal context, which has no relationship to the key of the previous song. Each song ends with a more or less conclusive cadence that is followed by a brief silence. When the next song begins, the change to a new key is accompanied by the introduction of entirely new thematic material and usually by dramatic changes in other parameters, such as texture, tempo, rhythmic figuration, meter, range, dynamics, and mode. By the time we get to the end of Dichterliebe, the last song we have heard in A major is Song 2, and the last song in which A was even the tonic is Song 8. After that there are eight more songs in eight different keys. At this point, A major has lost any aural significance as a tonal context, and so it is meaningless to describe Ct as the root of a III chord in that key. Schumann unambiguously establishes Ct as the tonic of the final song, and the cycle clearly ends in the key of Ct major. The analytic example that Komar uses to illustrate the overall form o^DickterLiehe, which is shown in Example 2.2, further demonstrates the conceptual problems that are inherent in equating key succession and voice leading. Komar adopts certain elements of Schenker's analytic notation, but his graph differs from a Schenkerian Urdatz in several essential respects, and the form that it represents cannot be reconciled with Schenker's orgamcist conception of musical structure. Komar's graph contains only one voice, which is not clearly derived from either the upper or lower voice of an Ur
Analyzing Dichterliebe
33
EXAMPLE 2.2. Komar'j "overall form,"from Dichterliebe, A Norton Critical Score, by Robert Schumann, edited by Arthur Komar. Copyright © 1971 by W. W. Norton t3 Company, Inc. Uded by permiuion ofW.W. Norton e3 Company, Inc.
turn to the root of the tonic chord in order to achieve closure. Because Komar's structure consists of a single ascending voice, it lacks both unity and closure, and, in fact, it does not even represent the unfolding of a single triad. Komar's prose description reveals even more clearly the crucial differences between his form and a Schenkenan background structure. Komar's terminology also appears to be derived from Schenker, but again there are divergences that are left unexplained. He -writes that "the arpeggiated interval A—Cf is filled in by the passing-note B, with an interruption of the motion from B to Ct between Songs 5 and 14" (77). In referring to the progression from A to Ct as an arpeggiation, Komar is apparently using this traditional term of music theory in Schenker's peculiar and innovative sense. In other words, Komar is describing not merely a melodic pattern on the surface but a deeper structure that is underlying a longer passage of music, in this case the entire cycle. For Schenker, an arpeggiation is a prolongation of a single harmony, even if one of the pitches is supported by a different, subsidiary chord on the foreground. In the case of Komar's graph, this would mean that the Ct chord with which the cycle ends is derived as a supporting harmony for the third of an A major chord that originates on a more background level. If we were to follow this through to its logical conclusion, then we -would have to construct an Ur
54
The Genre of the Cycle
motion is at first left incomplete, and there is a return to 3, which is followed by a complete descent to 1. With the interruption, Schenker's analogy has strayed far from its putative theoretical source, since there is obviously no model for such a progression within traditional counterpoint. The more likely model is the parallel period, which is perhaps the fundamental formal construct of the Classical style. The interruption enables Schenker to argue that this structure is the basis not only of the individual theme within a sonata form, but of the entire form as well, so that the part is understood as a replication of the whole.10 As with the uninterrupted passing motion, the interruption is a musical structure that is coherently unified, not only because the eventual descent to 1 represents the closure of a final tonic cadence but also because the analytical return to 3 typically corresponds to an actual return of the opening thematic material. When Komar describes the structure of Dichterlie.be as a passing motion that is interrupted, these terms lose their explanatory force, because there is no longer an analogy either with strict counterpoint or with Classical phrase structure. Since there is only one voice in Komar's example, there is no contrapuntal context within which we can understand B as a dissonance. And since he is representing a succession of keys, there is no tonal context, either. If Komar is arguing that as the tonic of a song that is in B minor B is metaphorically a dissonance in relation to the key of the opening song, A major, then it is not clear why Ct, as the tonic of a song in Ct minor, should be any less of a dissonance. Similarly, when Komar writes that the passing motion is interrupted after Song 5, the construct that he has in mind bears no relation to a parallel period or to Schenker's interruption. There is not a return of either the opening tonic or of the opening thematic material; rather, it is the return of B itself as a tonic pitch in Song 14 that ends the interruption and signals a resumption of the basic motion of Komar's form, as we can see in Example 2.3. Komar's interruption is really nothing more than a delay in the ascending stepwise sequence of tonics from A to B and then to Ct. In fact, I suspect that one reason Komar invokes the term is to justify the fact that half of the songs in the cycle appear to have no role in •what he refers to as its overall form. It is curious that this example, as well as Komar's description of it as an interrupted passing motion, cannot be reconciled with his subsequent graph of the harmonic and modal plans (see Example 2.2), which is intended to explain the role of the songs between 5 and 14 and does so far more convincingly. The Flrdt Five Song,) In a footnote to his analysis of the first song of Dichterlie.be,, Komar observes that his "application of Schenker's theories is not entirely orthodox" (67, fn. 3). While there is, of course, nothing wrong -with this lack of orthodoxy in and of itself, it does begin to impede our understanding of Komar's analysis when he is not explicit about how his use of Schenkerian terminology and analytic notation diverges from Schenker's. This is especially problematic since Komar intends to show that Dichteriiebe is an "integrated musical whole" and yet ends up jettison-
Analyzing Dichterliebe
35
EXAMPLE 2.3. Komar'/t "interruption,"from Dichterliebe, A Norton Critical Score, by Robert Schumann, edited by Arthur Komar. Copyright © 1971 by W. W. Norton ej Company, Inc. U
ing just those elements of Schenker's theory that are crucial to his orgamcist conception of musical structure. One 'way to consider the problem is to observe that where Schenker's prototypical musical form unfolds from background to foreground, the form that Komar describes moves temporally through the cycle like a left-to-right narrative. This distinction is perhaps not so clear in practice as it is in theory, since -we often end up reading and comprehending Schenker's analyses from left to right as well. But they are in principle conceived as a whole, and Schenker insists that it is from the background structure that all of the individual events of the foreground are derived and not vice versa. In applying this conception to Dicbterliebe, we would expect Komar to begin with the form of the whole cycle and then proceed to his discussion of each of the songs, thus demonstrating how the total structure manifests itself in the individual parts. But Komar does not do this, and in fact, the organization of his essay is one its most peculiar aspects. He divides his analysis into three parts, as follows: I. The Beginning of the Cycle: Songs 1-5 II. The Overall Form of the Cycle III. The Remainder of the Cycle: Songs 6—16 Komar never explains his rationale for organizing his essay in this way, but he does provide a clue at the end of his introduction, when he describes the first five songs as "a unified subsection" of the cycle and writes that in turning to the form of the entire cycle he will be "drawing upon the musical implications" of these songs (66). Komar apparently discovered several compelling relationships among Songs 1—5, and it was by extrapolating from his analysis of these relationships that he came up with his idea about the overall form of Dichterliebe. He implicitly corroborates this when he describes the form as "an outgrowth of the inital tonal events of the cycle" (77). But by arguing that Dichterliebe is a unified whole in a single key and dressing his analysis up in Schenkenan garb Komar must continually struggle to constrain the flow of his ideas within the systematic straitjacket of organicism. In the process, he obscures many of his most convincing insights. As -we might expect, it is in his discussion of the first five songs that Komar presents his most extensive analyses, as well as his most suggestive and interesting ideas. He offers several reasons that these songs create a coherent grouping,
36
The Genre of the Cycle
which involve both stylistic and structural elements. He observes, for example, that suspensions and appoggiaturas appear prominently throughout the songs, that they are all at a relatively soft dynamic level, and that their melodies all begin in the same part of the vocal register and emphasize the third between b1 and d2. Komar also notes that the coherence of these songs is highlighted by the contrast of Song 6, which begins with the first forte marking in the cycle and whose vocal melody begins with e1, a note that is outside the vocal range of all but one of the first five songs. But for Komar, "the most remarkable individual quality of Song 6 is the contrapuntal independence of the piano part," -which is another way in -which this song contrasts with those that precede it (76, 81). With this final comment, Komar has put his finger on one of the most significant elements that bind the first five songs otDichtertiebe into a group. It is not that the piano parts of these songs lack contrapuntal interest, as he is quick to point out, but there is clearly something about them that makes them feel as if they are incomplete, more like sketches than full-fledged accompaniments. This is partly a question of texture, as Komar implies. Consider, for example, the blurring between accompaniment and melody in Song 1, the incredibly high and narrow range of the piano part at the opening of Songs 2 and 3, and the unsettling lack of motion in the first measure of Song 4's accompaniment. But an equally important reason for this effect has to do with harmonic structure. Song 1 and Song 5 both begin -without a tonic triad, for example, and in Songs 2, 3, and 4 there is an unusual emphasis on the subdominant, which replaces the dominant at the opening of the former two songs and in the postlude of the latter. After these five songs, the straightforward tonic harmony that begins Song 6 —with its forte dynamic and its bass root doubled at the octave below —creates the impression of powerful stability. Although Komar comments on some of the unusual structural aspects of these songs -within his individual analyses, he does not mention them as a reason for hearing Songs 1—5 as a. subsection of the cycle. This is a surprising omission, especially since he does emphasize the significant contrapuntal relationships among the notes B, C|, and D in virtually all of the songs, relationships that play an important role in creating the structural irregularities in each of them. Once again, Komar seems to be caught between his analytic intuitions, -which are quite compelling, and his desire to fit these intuitions into an organicist model of musical structure. The conflict between these contradictory impulses is especially clear in his description of Songs 1—5 as a continuous narrative, in which the harmonic support for the pitch B gradually increases: B is initially presented as a dissonant seventh in a Ct major chord, and forms a major second with the tonic, A, of Song 1. In Song 2, B still forms a major second with the tonic, but mainly receives consonant support in E major triads. In Song 3, B forms a consonance with the tonic D, but remains outside the tonic triad. In Song 4, B is a triadic third, and becomes the tonic in Song 5. (76)
Analyzing Dichterliebe
37
Komar claims that "while the coherence of this subsection depends upon the ordered set of keys of Songs 1 —5, it does not follow that I regard any one note as the tonic of the five songs taken as a whole" (66). But if there is no single tonal context within which we can perceive B's gradual transformation throughout the songs, then we cannot trace a coherent path from the B that is an unresolved seventh in Song 1 to the B that is the tonic in Song 5. Considered in terms of Schenker's theory, we cannot derive the group of five songs from a single background structure, and so we have no basis for analyzing contrapuntal and harmonic progressions that extend from one song to another. It is at this point that the confusion between a structure that unfolds from back to front and a narrative that moves from left to right becomes problematic. Komar describes the form of Songs 1—5 in narrative terms, but what he has in mind requires a coherent background structure in order to make sense. Komar s discussion of Song 3 provides a good example of this problem. In Song 2, B forms a dissonance with the tonic but receives consonant harmonic support as a member of an E major triad. In Song 4, B forms a third with the tonic, becoming a member of the tonic triad. Komar describes Song 3 as an intermediate stage between these two songs, in which B already forms a consonance -with the tonic pitch but is still outside the triad. If we consider the theoretical context more completely, however, it becomes clear that the harmonic support for B actually weakens between Song 2 and Song 3. In the former song B is 2 and receives dominant support, and in the latter song it is 6 and receives support from the structurally -weaker subdominant harmony. That 6 forms a consonance with the tonic pitch is not relevant to the dissonant role that it plays within the major scale. In fact, one of the most striking aspects of Song 3 is that Schumann continues to feature B prominently on several different levels of structure, despite its dissonant position within the key of the song, and it is the way in which he features this pitch that creates a link with the surrounding songs. At the beginning of Song 3, the emphasis on b1 in the vocal melody leads Schumann to harmonize the opening phrase with a plagal progression and delay the arrival of the dominant until the end of the phrase, the same harmonic strategy that he employs at the beginning of Song 2. And in the B section of the song, he transforms the tonic triad into a V7/IV and prolongs both b1 and the supporting subdominant. There is a similar subdominant prolongation in the second phrase of Song A, and Schumann substitutes V^IV lor the tonic triad at the recapitulation of Song 2. Komar represents all of these progressions in his analytic graphs and mentions most of them in his discussions of the songs, but he evidently does not consider them to be unifying elements. As far as he is concerned, the emphasis on G major in Song 3 is not important because it is the subdominant harmony but because G major is the key of the following song. And he finds the corresponding emphasis on the pitch b1 unremarkable in and of itself but worthy of mention simply because it is also emphasized in the preceding and following songs. Komar is trying to construct an ongoing musical narrative through a series of songs in a variety of keys, so he ends up focusing exclusively on pitch
38
The Genre of the Cycle
class and ignoring tonal function. But while Schumann's emphasis on the same pitch class in each of these songs may be audible, we primarily hear this pitch within the tonal context of the individual song. The songs are not related because b1 becomes increasingly consonant throughout them but because this pitch is implicated in unusual aspects of the tonal structure in each one and several of these aspects are very similar. Komar's observation that the first five songs of DichterLiebe have an especially close relationship that sets them apart from the rest of the cycle is one of his most compelling insights. But rather than describing this relationship as an organic structure or as a continuous narrative, it would be more accurate to say that Schumann creates an association between the songs because he experiments •with similar formal ideas in each of them. The Schenkerian techniques that Komar uses in his analyses can illuminate many of these ideas, but only if we explicitly detach these techniques from the organicist premise that they were originally intended to serve. Komar's failure to do this is perhaps the greatest weakness of his essay. Subsequent scholars who have studied Schumann's cycles have generally taken one of two attitudes toward Komar's work. They have either accepted his organicist premise unquestioningly, perhaps with the qualification that a song cycle need not be in a single key, or dismissed him outright. The unfortunate result of this is that his study has not had the positive impact on the field that it could have had. His idea of comprehensively analyzing a complete song cycle was a bold and experimental stroke, -which no one else has attempted since. And in constructing his argument about DichterLiebe he engaged in a careful and thought-provoking consideration of the applicability of the organicist model to the song cycle. This is Komar's most important contribution to our understanding of the genre.
Schenker The problems that Komar encounters in trying to apply Schenker's theory to the whole of Dichterlie.be are certainly not surprising, given the fact that this theory was never intended to explain musical structure beyond the boundaries of a single movement or short piece. We might expect that Schenkerian analyses of the individual songs of the cycle would be far less problematic, since the small musical forms that -we typically encounter in the nineteenth-century lied repertory are generally regarded as the analyst's bread and butter. In fact, Schenker himself seems to have had a special attraction to the DichterLiebe songs. Although, in general, he used relatively few examples from Schumann's music in his published work, Derfreie Satz includes complete analyses of Songs 2 and 4 and a partial analysis of Song I.11 Schenker also left behind an unfinished analysis of Song 3, which Komar has published in his Norton Critical Score of DichterLiebe. Of course, the biggest reason that -we have almost unconsciously come to regard the songs of DichterLiebe as apt subjects for Schenkerian analysis is the fact that Allen Forte chose the analysis of Song 2, "Aus memen Thranen spriessen," as the primary example for his 1959 article in which he introduced Schenker to many
Analyzing Dichterhebe
39
American theorists for the first time.12 Two decades later, as Joseph Kerman mused about how we might 'get out from under" Schenker's influence, he turned to the same analysis to demonstrate the limitations of his technique and, in fact, of the entire discipline of musical analysis. Kerman describes Schumann's song as "a familiar, standard German-masterpiece-rype example," the sort of piece for which conventional "analytical methods traditionally work best," and writes of Schenker's analysis of the song that it "bids fair to attain exemplary status."13 And yet if we ignore the historical baggage that Schenker's analysis now carries with it and regard it simply as an attempt to explain Schumann's song •within the framework of his mature theory, then it soon becomes clear that it is not at all a paradigmatic example of Schenker's analytic technique. In fact, given the problems that Schenker encounters m trying to analyze the song, as -well as the improvisatory and ad hoc nature of his solutions, one can only wonder why Forte chose it in the first place. Kerman's description of the analysis as exemplary is a fair characterization nevertheless. Even before Forte turned it into his model Schenkerian analysis, it stood out as one of the most comprehensive in Der freie Satz, and Schenker himself does not suggest that there is anything unusually problematic about it. He presents it, with almost no commentary, as one of his two examples of an interruption. It is also true, as Kerman claims, that Schenker's problems "with the song appear to result directly from the constraints that his system sets up. And it is hard to argue with Kerman's observation that Schenker instinctively tried to reduce the song's "ambiguities out of existence."14 Still, I believe that Kerman's attempt to use Schenker's analysis of "Aus meinen Thranen spriessen" in order to criticize his theory is mistaken. In fact, at the risk of seeming perverse, I would argue that Schenker's difficulties in analyzing the song are not so much a result of the problems with his theory as they are a reflection of the problems -with the song. I do not mean to suggest that there is anything wrong with Schumann's song in the sense of aesthetic or compositional weaknesses. On the contrary, despite its brevity and its seeming simplicity, it is one of Schumann's most beautiful and complex creations. What I have in mind is that a large part of the song's complexity results from Schumann's play with the formal conventions and the models of musical structure that he inherited from the Classical style. Schenker's Urdatz is intended to represent these same structural models analytically, and so his attempt to apply the Unatz to Schumann's song can reveal some of its most interesting aspects. But since much of the interest results from the fact that the song does not exactly fit the model, it is often precisely at those points where Schenker is least successful in analyzing the song that we find the most illuminating insights.
Schenker'j
Difficulties
I will consider two problematic aspects of Schenker's graph, reproduced as Example 2.4, both of which have been called into question by other -writers. The first is his analysis of the ct in measure 1 as the Kopfton, or head note, which then returns at the analogous spot in measure 13. Schenker's placement of the head
40
The Genre of the Cycle
EXAMPLE 2.4. Schenker'j analy
note is the one thing that Forte takes issue with in his discussion of the analysis, suggesting that the return is actually delayed until measure 14. Komar then takes this suggestion one step further and argues that the initial head note should be moved back to measure 2, in order to retain the apparent parallelism between the two passages.15 Kerman dismisses these disagreements, arguing that they are irrelevant to "the way people actually hear, experience, or respond" to the song, and instead calls attention to a second difficulty in Schenker's analysis, which concerns what Kerman considers to be the song's "most striking feature." This is the unusual cadence that occurs in measures 4, 8, and 17, which from the vocal melody appears to be a half-cadence but in the accompaniment turns out to be authentic. As Kerman observes, Schenker's problem is that his analytic notation is not intended to represent such subtle ambiguities, and so ultimately he must choose in each case -which part to follow. According to Kerman's reading of Schenker's graph, the first two cadences are analyzed as half-cadences and the final one as an authentic cadence, a "disappointingly conventional interpretation," which shows no "appreciation of the whole extremely original and suggestive situation."16 If-we consider the graph more sympathetically, however, we see that Schenker does appreciate Schumann's ambiguity and struggles mightily against the constraints of his notation in order to express it. I believe that both of these problems—the questions of where to place the head note and how to analyze the cadences—ultimately result from the lack of correspondence between the vocal line and the piano accompaniment. What has
T
41
been most frequently observed about the relationship between voice and piano in this song is the fact that the voice stops on the cadential dominant and lets the piano resolve the cadence. This is what Kerman considers to be the song's most striking feature, and it is also what Schenker himself observes about the song in his early Harmonielehre, published in 1906: The vocal part first brings the half-close, whereupon the piano follows immediately with the full close. . . . It is true that basically this example represents a perfect full close; by interrupting at the dominant, however, the vocal part succeeds, at least for the duration of the fermata, in most conVincingly simulating the effect of a half-close.17
By the time Schenker returned to the song twenty-five years later, his system of analysis had developed in such a way that he could not capture the ambiguity of Schumann's effect as plainly as he does in this passage. However, by focusing solely on this one moment we miss the fact that the conflict between the melody and the accompaniment does not suddenly appear in the cadential measure but extends throughout the phrase. And it is largely by means of this conflict that Schumann takes a simple ternary form and creates such profound ambiguities. The vocal melody of the A section, considered alone, consists of a fourmeasure phrase that is divided into two equal halves, which balance and complement each other. The first half prolongs 3, which is minimally embellished by its upper neighbor on the first beat of the second measure. In the second half of the phrase, the principal pitch is 2, -which likewise begins and ends the segment but is embellished by a double neighboring motion that circles around it. The addition of the piano accompaniment to the vocal line destroys the symmetry of the phrase. Schumann ties the initial upbeat over into the folio-wing measure, so that there is no impulse on the first downbeat. He also begins with a bare third, in the treble range, and presents a complete tonic triad only on the second beat of the second measure, so that the tonic harmony gradually grows in strength. Furthermore, since the triad in the second measure is approached by a plagal progression, it is weaker than the tonic triad that comes on the second beat of the third measure, which completes an authentic cadence and thus concludes the phrase. This means, first of all, that there is no break between the second and third measures, as there is in the vocal melody and, second, that the fourth measure is merely an echo of the third measure. So where the vocal line presents a four-measure phrase that is divided 2+2, the accompaniment presents a continuous three-measure phrase, -with an extra measure tacked on. This discrepancy in the internal phrase structure of the two parts complicates the analysis of the third measure. In terms of the vocal melody, the a1 in the second half of the measure is a lower neighbor, which resolves to b1 on the following downbeat. Once we add the accompaniment, a becomes the note of resolution, and the following b simply begins the cadential descent a second time. For this reason, the account that -we usually read of this cadence does not give the whole story. The weakness
42
The Genre of the Cycle
of the tonic resolution in the second half of the fourth measure results not only from the fact that the vocal melody has already ended but even more so from the fact that the piano has cadenced once already and done so more forcefully. The opening of the phrase is just as ambiguous as its close. The first measure of the vocal phrase is strong, and the second is weak, despite the appoggiatura. But the lack of a rhythmic impulse in the piano part on the first downbeat and the fact that the texture begins so thinly and gradually increases as we move into the second measure make it impossible to hear the first measure of the accompaniment as strong. However, we cannot hear it as an upbeat measure either, since it contains the initial tonic harmony and is thus the true beginning of the phrase. When the phrase returns, in measure 13, Schumann recomposes the accompaniment in such a way that he reverses these parameters. Rhythmically and texturally, he now treats the first measure as strong by placing an impulse on the downbeat and thickening the piano texture to five voices, but he also transforms the initial tonic into a V7/IV, so that in terms of the harmony it does become an upbeat measure. One effect of this recomposition is a blurring of the moment of recapitulation. The melodic return comes in measure 13, where -we expect it to, but the harmonic return is delayed until the second beat of measure 14. Another effect is that it retrospectively intensifies the ambiguity of the opening by making the status of the first measure of the phrase even less clear. It is no small wonder that analysts have disagreed about the placement of the head note in this song. It may be true, as Kerman argues, that these disagreements seem to take us into the realm of abstract theoretical debate, but if we are willing to interpret them as symbolic statements about the song, then they carry a number of revealing implications about how we hear and respond to the music. Does the song begin with a strong measure, as the vocal melody suggests, or is the initial tonic delayed until the second measure? And does the return of the phrase come in measure 13, as expected, despite the inner voice G^ or is it also delayed until the following measure? Are these two moments essentially the same in any case, or does their apparent parallelism conceal a more significant difference between them? All of these questions relate directly to the listener's experience of the song, and the divergent analyses of Schenker, Forte, and Komar can be read as different answers to each of them. The existence of the debate in and of itself does not point to a flaw in Schenker's theory, as Kerman suggests, but to a fundamental ambiguity in Schumann's seemingly simple song. By forcing us to make a choice about the placement of the head note, Schenker's analytic system provides a coherent framework within which "we can grapple with this ambiguity and achieve a deeper understanding of how the song works. But if the process of Schenkerian analysis can be very illuminating in this particular instance, the graph that results from this process can only limit and distort our understanding of the song. The ultimate decision of precisely where to place the head note is an impossible one, because there is no single moment when the opening tonic prolongation actually begins or when this indefinite beginning returns, and this is obviously what Schumann is after. Looking at Schenker's graph of the song, I would say that if anything is "disappointingly conventional," it is his
Analyzing Dichterliebe43
43
solution to this problem. By placing the head note on the downbeat of the first measure, which is where the opening tonic harmony would typically come, he ignores the way in -which texture and rhythm conflict with the expected harmonic structure. And -while he acknowledges the substitution of V7/IV at the return in measure 13, retaining the G^ on both the foreground and middleground levels of his graph, he still labels the harmony as a tonic that supports the return of the head note and thus implies that the return is completely normal anyway. In this case, it seems that Schenker is merely folio-wing his analytic instincts, -which dictate, first of all, that typically it is the initial statement of a sonority within a given prolongation that is more structurally significant and, second, that passages of music that appear to be parallel on the surface are generally analyzed in the same -way. Schumann',) Cadences Schenker s analysis of the cadence at the end of the A phrase is more interesting. Although he is again constrained by his symbolic notation, he is far more creative in how he uses it, with results that are inconsistent in terms of his system but manage to convey much of the song's complexity. In this case, Schenker has no scruples about analyzing the same cadence in completely different ways each time it occurs, -which becomes most obvious when we compare the opening phrase and its immediate repetition. Here we have virtually the same musical phrase, heard twice in succession, and yet the first time it ends, in measure 4, Schenker analyzes an authentic cadence, where the second time, in measure 8, he analyzes a half-cadence.18 The discrepancy is all the more striking because the latter cadence is all that Schenker chooses to include from the second phrase. We could explain this apparent inconsistency as an inevitable result of Schenker's theory and observe that if he -wants to analyze the song as an example of an interruption, he must treat the cadence differently in measure 8, since the end of the A section necessarily has to function as the beginning of a dominant prolongation. But -whatever Schenker's motivation may be, his analysis reveals something quite peculiar about Schumann's cadence. Because of the conflict between the vocal melody and the piano accompaniment in the cadential measure, it can be heard as either a half-cadence or an authentic cadence, depending upon the context. Since in measure 4 the cadence is followed by a repetition of the phrase, we are more likely to supply in our imaginations whatever is missing in the incomplete resolution and hear it as an authentic cadence. In fact, there is an almost symbiotic relationship between the end of the first phrase and the opening of the second, in that the two tonic harmonies, each weakened in its own way, strengthen each other through their close proximity. In measure 8 it is quite a different story, since the tonic chord, which is heard in the accompaniment alone, is sandwiched between two nearly identical dominant chords, both of •which support 2 in the vocal melody and the second of-which continues through the following measure. In both cases, then, Schenker's analysis is true to the way that -we hear the cadence.
44
The Genre of the Cycle
A comparison between Schenker's analyses of the cadence in measure A and the cadence in measure 17 is even more telling, since in this case his theory would dictate parallel analyses, so the discrepancies can only be the result of his conscious effort to engage those aspects of the song that his analytic system would just as soon leave untouched. The difficult question here is how to analyze the third and fourth measures of the phrase. In his foreground graph, Schenker represents the echo in measure 4 clearly enough, but he is ambiguous about its status. The bass slur from the initial a1 to the a in measure 3 and the corresponding Roman numerals below the first three measures suggest that Schenker hears the phrase much as I do—it is three measures long, -with a fourth measure tacked on. But this apparent analysis is contradicted by the fact that the upper-voice descent into measure 3 consists purely of black note heads, while in the subsequent descent into measure 4 both cj2 and a1 have flagged stems, suggesting that it is in this measure that the definitive cadence comes, with measure 3 functioning as an internal phrase expansion. This latter interpretation is confirmed in the middleground and background graphs, -which retain the second descent. Schenker may well have found himself caught between his desire to represent a striking and idiosyncratic aspect of the piece and his need to maintain one of the most crucial theoretical premises of his system. Schenkerian analyses are left-branching structures, as Allan Keiler has observed, which means that, in general, the closer to the end of the piece that a prolongational event occurs, the greater its structural significance within the analysis. This -would dictate that Schenker analyze the cadence in measure 4 as a higher level event than the cadence in measure 3, which is what he does, even though -we perceive the earlier cadence as the stronger of the two. But even if this aspect of Schenker's theory leads to an incorrect analysis of this passage, that does not mean that it is arbitrary. On the contrary, as Keiler points out, this is one of Schenker's most significant insights "about the way in -which pieces move toward completion as a function of time; that is, the sense in -which degrees of closure increase until the final cadence of the piece." From the perspective of the song, -we could observe that by cadencing in the third measure of the phrase and turning the fourth measure into an echo Schumann subverts a basic principle of tonal phrase structure, and this is one of the reasons that the phrase feels unbalanced and even disturbing. That Schumann was consciously aware of the peculiarity of his phrase is apparent from the care that he took to ensure that the pianist will do what the music implies. He places accents on the second and third downbeats but not on the fourth, which instead has a pianudimo marking. And, of course, the fermata on the cadential dominant makes the tonic resolution fall at least slightly out of time. Here again, we can learn a great deal about Schumann's song by retracing Schenker's steps, and in this case the contradictory analysis in his foreground graph has left behind a revealing piece of evidence. Although we must ultimately interpret his analysis of this phrase as another case in which he distorts the music by normalizing Schumann's idiosyncrasies, Schenker's graph does call attention to the -way in which
Analyzing Dichterliebe
45
Schumann derives his phrase from a more prototypical model, in which the definitive tonic does not come until the fourth measure. When we consider Schenker's usual desire to analyze parallel passages in similar ways, we can only wonder at his treatment of the song's final cadence. As I have noted, the form that Schenker posits for the song dictates that the cadences in measures 4 and 17 should function in the same way and, if anything, he would be more inclined to normalize the latter cadence, which is the highest level event in the song and brings final closure. But Schenker does not analyze the ending in this way at all. In this case, the phrase is presented unambiguously as three measures long, with the definitive cadence in measure 15. The closure of Schenker's Urjatz thus conies two measures before the song ends and m the middle of the final vocal phrase, as a comparison between his foreground graph and his middle- and background graphs makes clear. What then is the status of the cadence in measures 16—17, in terms of the hierarchical structure of the song? Since this cadence first appears on the foreground level of Schenker's analysis, it appears to be of less structural significance than the cadence in measure 15, -which is part of the background structure, and even than the cadence in measure 4, which is first derived on a middleground level. If -we compare these latter two cadences as they appear in the foreground graph, we can see how carefully Schenker uses note heads, stems, and beams to represent their respective levels of derivation. His notation of the cadence in measures 16—17, however, is entirely inconsistent -with the fact that it is first derived on the foreground of his analysis. By representing the final tonic harmony -with quarter notes, Schenker implies that it is on a higher level of structure than the pitches that are first derived on his middleground graph, since these appear on the foreground as eighth notes and filled-in note heads. Even more inexplicable is his use of half notes for the dominant harmony in measure 16, -which implies that this harmony is on a higher level than the tonic to which it resolves and is even derived earlier than the harmonies that support the Urdatz. Ultimately, there is no -way to reconcile these contradictions and to interpret Schenker's representation of the final cadence as a coherent part of his graph. In this regard it is the least successful aspect of the analysis, since the consistency and thus the meaningfulness of his notation have broken down completely. But at the same time, this part of his graph is his most ambitious attempt to express the ambiguity of Schumann's cadence, and if -we read Schenker's analysis more intuitively, then -we can see that he does capture, to a surprising extent, the complex way in -which we hear the ending of the song. This is the most disturbing of the three presentations of the cadence, both because we have already heard it in two different contexts, which have led us to interpret it in contradictory ways, and also because it is the end, and so our desire for unambiguous closure is especially acute. And what is so disturbing about it is precisely the fact that we cannot determine the status of the dominant harmony in measure 16, in relation either to the immediately preceding cadence or even to the tonic harmony that tries to resolve it. We can perhaps read the breakdown in Schenker's symbolic
46
The Genre of the Cycle
notation as a representation, in and of itself, of Schumann's subversion of the seemingly straightforward syntax of his cadence. Scbenker't) ImprovutatwnalAnalyjM I would like to return for a moment to the question of why, in this particular instance, Schenker presents different analyses for two phrases that appear to be completely parallel, both in terms of their surface structure and in terms of their function within the larger form of the song. I do not believe that he is claiming that the two phrases are in fact fundamentally different, but that he is trying to work out the most effective means by which he can use his analytic notation to express the unusual structure of Schumann's cadence. In the first instance of the cadence, in measure A, Schenker reproduces the fermata that appears over b1 in the vocal melody. Since this is not a symbol that represents anything about the structural significance of a given pitch within Schenker's analytic system and, in fact, it is not a normal part of his symbolic notation at all, he can use it to express the peculiar emphasis that this pitch receives within the song but still make it clear that in terms of structural hierarchy it is subordinate to the a to which it resolves. His representation of the same pitch in measure 16 is far more visually striking, both because the note has now come out from under the beam of the Urjatz and also because it is one of only a handful of open note heads in the graph. Of course, another reason that Schenker's representation of this pitch in measure 16 is so striking is precisely that the half note is such a significant part of his notation and is a symbol that he typically reserves to represent the highest level of structure •within a piece. While the use of this symbol in measure 16 thus leads to the problematic inconsistencies that I have already enumerated, it also goes much further than the fermata in implying that the emphasis on b is not merely a matter of inflection or surface expression but a basic structural issue that calls into question the hierarchical relationships within the song. What I am suggesting, in contrasting the two cadences in Schenker's foreground graph, is that we might read the graph not as a coherent and internally consistent analysis but as a record of Schenker's improvisatory process as he struggles simultaneously to understand the most difficult and interesting aspect of Schumann's song and to represent his understanding in a manner that is visually effective. I have been focusing most of my discussion of Schenker's analysis on his foreground graph alone, but I will conclude by considering the relationship of this graph and the level of structure that it represents to the middle- and background levels. This relationship is an aspect of Schenker's theory that is frequently misunderstood, and at the same time it is of crucial importance if we want to evaluate the usefulness of Schenker's analytic system in helping us to better understand the songs of Schumann's cycles. The -way in which Schenker lays out his graphic analyses, at least in his later works, and the way in which he uses his analytic notation to represent the content of the various levels of structure can easily lead to the implicit assumption that he considers the back-
Analyzing Dichterhebe
47
ground to be of greater aesthetic or compositional significance than the later levels and, in particular, than the events that are first derived on the foreground. Kerman, for example, criticizes Schenker's analysis of the cadences that I have been discussing by arguing that their unusual aspects are completely ignored in his middleground and background graphs. He then writes: The Urdatz confuses the issue, for in bars A and 8 the cadences lack status because they are regarded simply as details of prolongation, along with many others, and in bars 16—17 they are trivialized because true closure is conceived as happening a bar earlier.21 By suggesting that the first two cadences "lack status" and the third is "trivialized" in Schenker's analysis, Kerman makes the explicit claim that by representing the cadences only as foreground events and not as part of the Ur<>atz Schenker is offering a value judgment about their significance to the piece. But, as Kerman himself acknowledges, the Ur^atz is intended to represent the underlying structural prototype from which all tonal compositions are derived, and so it is axiomatic that Schenker's background graphs will leave out the most interesting elements of virtually any piece he analyzes. Perhaps it would be more useful, when comparing the background and foreground levels in Schenker's system, to discard such value-laden dichotomies as "deep versus surface" and "significant versus insignificant" and instead describe the background as the more universal or abstract level and the foreground as the more individual or concrete. In other words, the background represents a prototype or model of tonal structure and the foreground represents the particular realization of that model that occurs in a given piece. The question, then, is not -why Schenker omits the most striking aspects of Schumann's cadences from his background graph but -why he has such difficulty representing them in his foreground graph. And to answer this question, we must again consider Schenker's organicist conception of musical structure. One of the central tenets of Schenker's theory is that the complete content of a piece of tonal music can be derived from the Unatz. The implication of this premise for a layered analysis is that each successive level is derived from the preceding level and all of the levels are ultimately derived from the background. Schenker's analysis of the final cadence in Schumann's song is incomprehensible, -within the logic of his system, because there is no possible way in which it can be derived from the background level. The problem is not, as Kerman implies, that Schenker is exclusively concerned with harmonic and contrapuntal syntax and thus ignores features of the song that involve other parameters, such as texture, rhythm, and dynamics. We can easily find ancillary notational symbols to call attention to such features, as Schenker does with the fermata in the fourth measure of his foreground graph, and we can supplement a voice-leading analysis -with prose description or with other types of analytic graphs. The reason Schenker has such intractable difficulties -with Schumann's song is precisely because its harmonic syntax is its most striking feature. By creating parametrical conflicts between the harmony, on the one hand, and the phrase rhythm, the
48
The Genre of the Cycle
melody, and the texture, on the other, Schumann undermines the conventions of tonal syntax and creates a great deal of ambiguity as to where the formal and prolongational boundaries occur within the song. Ultimately, as -with Komar's analysis of the whole of Dichterlu.be, Schenker's analysis of the second song of the cycle is not entirely successful, in the sense that neither of them manages to demonstrate what he sets out to. Considering the matter from Schenker's perspective, we might simply observe that Schumann's songs do not provide the best examples with which to demonstrate how his theory -works. It may well be the case, as Kerman argues, that because this song is a "standard German-masterpiece-type example," its failure is indicative of a more widespread problem with Schenker's theory. But even if this is true, it is Schenker's problem and not ours, and I do not think that it is a reason to reject his analytic method altogether, as Kerman suggests. If we insist that the crucial question is whether Schenker is ultimately right or wrong, then we force upon ourselves a false and irrelevant choice. Whatever the failings of his theory may be, there is no question that Schenker has offered many valuable insights about how tonal music works, and he has irrevocably changed the -way we listen. If we •want to come to a greater understanding of Schumann's song cycles, we clearly need to move beyond -what Schenker has done with "Aus meinen Thranen spriessen" and find a broader analytic approach that is not so focused on trying to demonstrate that Schumann's songs adhere to the organicist model. But if we exclude any hint of Schenkerian thinking from such an approach, on principle, then we are placing restrictions on our thinking that are far more arbitrary than Schenker's Urjatz is.
Neumeyer Having considered the question of organicism in Dicbteriiebe from both ends — the whole of the cycle and the part of the song — I -will now turn to the work of a theorist who brings these two perspectives into juxtaposition. David Neumeyer, in his article "Organic Structure and the Song Cycle: Another Look at Schumann's Dichteriiebe," begins by addressing the same problem that Komar does: how is it possible to analyze DichterLiebe as a complete and organically unified structure? But Neumeyer ends up trying to solve a very different problem, which is that the first song, considered by itself, cannot be analyzed as an organic -whole. Because of this, the analytic portion of his article does not follow logically from the theoretical frame-work that he sets up in the first half, and so -we can most fruitfully read the article as two discrete arguments that have subsequently been interwoven. In one argument, Neumeyer establishes the parameters of musical organicism, as it has been understood since the early nineteenth century, and demonstrates why this metaphorical construct does not fit Schumann's cycles. Neumeyer then proposes expanding the concept so that it can be used to explain the cycle. In his second argument, he suggests that -we can analyze individual songs and piano pieces from within Schumann's cycles that appear to be incomplete structurally by pairing them with the folio-wing
Analyzing Dichterliebe
49
song or piece, using the slow-fast pairs found in the Baroque sonata da chi&ia. as a model. A New Approach to OrganicLtm Neumeyer observes that key unity has been the cornerstone of the organicist metaphor from its origins in the early nineteenth century. To begin with, the specific meaning of key unity was not very clearly defined, beyond the fact that a piece must begin and end in the same key. Schenker then clarified the concept by adding the necessity for melodic as well as harmonic closure and also the stipulation that in analyzing a tonal -work harmony and counterpoint must be the prevalent and overriding structural factors. Since at least two of Schumann's cycles do not meet even the original, modest requirement for key unity, Neumeyer is rightly skeptical that there -would be much value in trying to apply Schenker's more elaborate theoretical constraints to the cycle. However, Neumeyer does not conclude from this that we should abandon the aesthetic criterion of organic unity but that we need to find some -way to expand Schenker's system. Neumeyer proposes that in the case of the song cycle, we should no longer consider harmony and voice leading alone to be the source of organic unity but include the narrative and dramatic aspects of the text as well. As Neumeyer puts it, we must treat Schenker's Urjatz and the unity of the narrative as "co-equal structural determinants" in order to demonstrate that the song cycle is an organic structure. I, too, believe that if we really -want to come to a comprehensive understanding of Schumann's song cycles, then we need to consider the text as well as the music, but I wonder if it is really possible that by integrating these two criteria we can solve the specific theoretical problem that concerns Neumeyer. He points out that in the case of Dichterliebe, his primary example, -we cannot explain either the text or the music, considered on their own, in terms of organic unity. Given that this is the case, it is hard to imagine how a theoretical system could be devised that would provide an organicist explanation for both together. In one sense, Neumeyer's problem is very similar to the one that we have already seen Komar come up against, when he tried to use motivic relationships among the songs o£ Dicbterliebe as evidence for a form that is based on Schenker's model. In theoretical terms, poetry is just as incompatible with the Ur
99
50
The Genre of the Cycle
and that, if anything, "the words actually impede . . . enjoyment of the -whole cycle."23 Surely we can stake out some middle ground and find an approach to Schumann's songs that can help us to explain the relationship between text and music -without insisting on their mutual dependence. While I agree with Neumeyer's implication that an understanding of the words is essential to a complete apprehension of song, I have also, like Komar, had the experience of enjoying the music of a song perfectly -well without understanding what was being sung. A -well-composed song is an aesthetic response to the poem that is being set, which creates a symbiotic relationship between music and text, but at the same time it is meaningful as a purely musical utterance.24 Neumeyer's claim that the text of Dichterlie.be raises some of the same problems with the metaphor of organicism that the music does suggests to me that Schumann's play with musical structure, both in the individual song and in the complete cycle, may be one -way that he responds to the poetry he is setting. There are many moments when an unusual aspect of the text inspires him to come up with an innovative compositional device, but I believe that in every instance we can understand the musical gestures that Schumann creates on their own terms and need not go back to the poetic text for clarification. When he turns from theoretical speculation to a concrete analytic example, Neumeyer makes an abrupt and unexplained shift from a discussion of the song cycle as a whole to a consideration of a small part, -which happens to be one of the most unusual pairs of songs that Schumann composed. Neumeyer argues that the first two songs cfiDichterLie.be are a good example to try out his integrated approach, essentially because the first song is structurally incomplete and so he can avoid the "presently unanswerable question" of how to create large-scale structural connections between separate songs, each of which already consists of a complete structure (97—98). But this is precisely the question that he must address it he wants to find some way to expand Schenker's notion of organic structure to the entire cycle. By choosing a song that truly has an inconclusive ending Neumeyer limits the implications of his example to a small group of Schumann's songs and piano pieces and reveals very little about how -we might analyze entire cycles as integrated wholes. However, within this more limited sphere, Neumeyer raises questions that are of great interest. And while there are relatively few songs within Schumann's cycles that are paired -with their neighbors by virtue of an incomplete close, there are many others that are involved in subtler and more complex kinds of pairing relationships. It is significant that Neumeyer faces substantial difficulties even when dealing with what appears to be a relatively straightforward example. The subject of pairings comes up quite early in Neumeyer's discussion, as a rare instance in -which Schenker is willing to consider organic structure beyond the boundaries of a single movement. The example that Neumeyer cites is Handel's Suite #2 in F Major for Harpsichord (HWV 427), -which is a typical fourmovement sonata da cbieda. Schenker observes that since the first and third movements lack a true Urtinie, the "suite does not consist of four movements, but only two, each preceded by an introduction. 25Neumeyer interprets this comment as
Analyzing Dichterliebe
51
a suggestion that we connect each of "the incomplete movements" to the complete one that follows and construct a single Unatz for the pair (94). In fact, however, it is not exactly clear what Schenker has in mind, especially since he does not provide an analytic example. Certainly it is difficult to imagine how he would analyze the third and fourth movements as a single unified structure, since, as Neumeyer observes, they are in different keys. In referring to this example Neumeyer makes an insightful connection, since its resemblance to certain of Schumann's song pairings is so close that he must have been using this Baroque formal convention as a model. There are five songs within Schumann's cycles whose endings are derived from the Phrygian cadence, and in three cases Schumann calls attention to the Baroque origin of the gesture by combining it with a number of other archaic elements.26 As in the typical Janata da chie
52
The Genre of the Cycle
EXAMPLE 2.5. Schenker'ti analyjLi of "Im wunderMbonen MonatMai."Heinricb Schenker "Derfreie Satz" © 1935 by Universal Edition A G., Wien.
minor is actually subsumed •within an A major arpeggiation, and Ejt is merely a chromatic inflection, which resolves upward as it moves in parallel tenths with the bass. Schenker characteristically emphasizes the unifying effect of the voice leading and explains the entire passage as a prolongation of a single underlying chord. There is no tonal ambiguity for him because the Ctt triad does not have a harmonic function and thus does not imply a harmonic resolution. Komar also analyzes the Ctt harmony as a contrapuntal chord that is subsidiary to the tonic A major triad. But where Schenker emphasizes the continuity from one harmony to the other, Komar calls attention to the break between them and creates a voice-leading graph that hovers at the very margins of musical coherence, shown in Example 2.6. If one plays Komar's graph at the piano, it does not sound as if it could be the basis of a piece of tonal music. One aspect of his analysis that is especially striking is the direct voice-leading relationship between the Ett of the C| harmony and the E*l of the A major harmony, a progression that both defies the rules of strict counterpoint and also destroys our sense of harmonic syntax. It is also significant that Komar retains the pitch B, the seventh of the Cf harmony, in what is intended as a background graph. In fact, Komar takes note of this peculiarity of his graph, observing not only that "B receives dissonant harmonic support" but also that it is "accompanied by Ett, the chromatic neighbor to Ft" (69). Why does Komar describe Et as a neighbor to Ft even though it appears in his graph as a neighbor to E^l? What he apparently has in mind is the implied dominant function of the harmony that supports Et, which is an issue that he addresses at the very beginning of his discussion of the song: "Song 1 is justly famous for the ambiguity of its key. The opening four measures suggest Ft minor, but no Ft triad appears, either here or elsewhere in the song" (66). It is to express this tonal ambiguity that Komar treats the Et as he does and includes the B as part of the Ct harmony. Unlike Schenker, who professes not to hear the harmonic implication of the opening at all, Komar considers it to be of primary importance in explaining the structure of the song. So perhaps we should not read his graph as a literal representation of Schumann's voice leading but as an attempt to use Schenker's analytic notation to convey the fact that Schumann never realizes the implication that he sets up in the piano introduction and thus disrupts the structural unity of the song. Neumeyer's solution to the problem is pairing the song -with the one that fol-
Analyzing Dichterliebe
53
EXAMPLE 2.6. Komar's background graph of "Im wunderschonen Monat Mai, "from Dichtei A Norton Critical Score, by Robert Schumann, edited by Arthur Komar. Copyright © 1971 by W. W. Norton eJ Company, Inc. U
lows, but it is clear both from his analytic graph and from his discussion of it that he never definitively resolved the question of the song's tonality in his own mind. His analysis of Song 1 and the first two measures of Song 2 is Example 2.7. Neumeyer notes that by analyzing "the song pair 1/2" in A major he is able to preserve the tonal ambiguity of the first song, whereas if he graphed this song by itself he would be forced to choose a single key. He acknowledges that there must still be a "bias toward one key orientation" in his analysis of Song 1 and then explains his choice in a passage that makes his ambivalence palpable: I have favored ft somewhat (through its dominant, of course), because I find that treating the tonal emphasis of the piano prelude and postlude frame as more significant than the internal move to A is more satisfying than the reverse, which would make the close simply gratuitous — or worse, not just "open," but inexplicable. Still, if obliged to do so, I would think of the first song, taken by itself, as in both ft and A—an indefinite harmonic relation of the third. (It has, however, been the argument here that the first song should not be taken by itself.) (103—4) This understandable unwillingness to choose between one key and another helps to explain a glaring contradiction in Neumeyer's graph. He identifies the key of Song 1 as A major but insists it is the Cf V7 chord that is being prolonged, going so far as to explicitly reject Komar's analysis, in which this harmony frames the A major triad. This leaves Neumeyer in the untenable situation of analyzing the entire song as a prolongation of a harmony that is not a Stu/e within the key, as is clear from the empty brackets below the lower level of his graph. Neumeyer tries to solve the problem by arguing that "the harmonic-contrapuntal structure is only a middleground feature in the song pair" (103). But if Neumeyer's graph is intended to represent a middleground level, what possible background could it be derived from? Ultimately, we are still left with the problem of explaining how Schumann's tonal progression from the dominant of an implied Ft minor to the tonic of A major can be explained in terms of a coherently unified structure, but now the problem has been pushed back into the opening measures of Song 2.
54
The Genre of the Cycle
EXAMPLE 2.7. Neumeyer'ff analyduf of the opening pair of Jong<> from Dichterliebe, © 1982 by Society for the Study o/Mutiif Theory, reprinted from Music Theory Spectrum v. 4, pp. 92-105, by permuitiuin.
It is true that the bare third that begins Song 2 creates a voice-leading association -with the V7 that ends Song 1, as Neumeyer's graph shows. If we were to hear the second song by itself, then we -would interpret the interval as the root and third of an A major triad. But when it follows Song 1, we could momentarily hear the interval as the third and fifth of an Fit minor triad, and this interpretation receives apparent confirmation when the left hand reaches ft1 on the second beat of measure 1. Almost immediately, however, we realize that our interpretation is wrong, that ft1 is part of a passing motion from a1 to d1, the roots, respectively, of a tonic and subdominant chord, and that the opening two measures are a prolongation of a tonic A major chord. That this latter interpretation is, in fact, the right one is corroborated by the authentic cadence in measure 3, and by the time the opening third returns on the upbeat to measure 5 there is no possible doubt that this interval is part of an A major triad. In other words, what Neumeyer presents as a graphic representation of the underlying structure of the song pair is, in fact, a phenomenological account of how one hears the transition from one song to another when they are sung and played in succession. There is no question that Schumann has created a connection between the end of one song and the beginning of the next, but this connection is transitional and not structural, and so it is misleading to represent it within a Schenkerian voice-leading graph. \Vhen discussing Komar's analysis of the first five songs in the cycle, I argued that what he was describing did not involve a unified background structure but could be described more accurately as an associational relationship, in which
Analyzing Dichterliebe
55
Schumann explores similar formal and tonal issues in each song. Neumeyer is concerned with a different kind of relationship, which we might describe as a transitional pairing, but both analysts really come up against the same basic problem. Whatever connections there may be from one song to another, they are still separate pieces, each -with its own thematic content, expressive affect, and tonal structure. Text and MLMLC What of the second part of Neumeyer s thesis, that we can integrate music and text and come up with an organically unified structure on a higher level? To respond to this claim, we must consider two different but interrelated issues. One is simply the question of whether or not it makes any more sense to regard the text of the pair in organicist terms than it does to regard the music in this -way. And the other is the more general problem of whether the text of a song can determine how we analyze the music. Neumeyer argues that the song pair from Dicbtertiebe that he has chosen demonstrates how text expression can strongly affect our interpretation of musical structure, "including the determination of some of its controlling elements." The two examples that he has in mind are the use of a V chord as the concluding harmony in Song 1 and the fact that the head note is delayed until the second measure in Song 2 (103). Neumeyer describes the first song as a prologue to the cycle, in which the poet confesses his love. The concluding seventh chord "leaves this prologue open for the action to begin in earnest," -while at the same time it "preserves a faint sense of irony or of doubt." He hears the portamento descending line in the first measure of Song 2 as a depiction of the tears that are mentioned in the opening line of its text, a musical image that both resolves and continues the "question" of Song 1. As the key of A major is more securely established, the "feeling unit of questions and tears . . . 'opens outward' into the images of hope: flowers, nightingales, the chance that love be requited" (102—3). Regardless of whether one agrees with his particular interpretation of the songs, Neumeyer is certainly right that Schumann's decision to create such an unusual transition from one song to another was one way in which he responded to Heine's text. But to go from this statement to the claim that this response determines the controlling elements of Schumann's musical structure is to make a huge leap. Neumeyer himself offers a convincing explanation, on exclusively musical grounds, of the concluding V7 harmony in Song 1 when he observes that it is derived from a Phrygian cadence. And as I have demonstrated in my discussion of Schenker's analysis of Song 2, there are also compelling musical reasons to analyze the head note of this song in measure 2. The texts that Schumann was setting may have inspired him to compose these two passages in the way that he did, and we will certainly come to a deeper understanding of the songs if we consider how the music relates to the poetry, but we do not need the texts in order to create a coherent analysis of the music, because the music is completely coherent on its own terms. Neumeyer's argument about music and text is also problematic because it is
56
The Genre of the Cycle
based on the premise that these two elements have a relationship that is essentially complementary, with each providing what the other lacks. But it would be more accurate to assume that they are analogous, since in composing a song Schumann is providing a musical analogue to the original poem. He may change the poem in the process of setting it by a variety of more or less subtle means — altering its formal proportions, repeating words or lines, or even destroying prosodic elements such as its meter or rhyme scheme — but we tend to understand and evaluate such transformations in terms of his attempt to interpret the poem and ultimately to capture its sense in music. For this reason, song analysts who focus on the relationship between text and music usually emphasize the parallels between them, and Neumeyer is no exception. His interpretation of the text of Song 1 as a prologue to the cycle, for example, is clearly based on his analysis of the song as an introduction to the following song. It is only in Song 2 that the "action" of the text begins, and it is also in this latter song that the harmoniccontrapuntal structure—the musical action of the pair—is initiated. Similarly, his observation that both texts contain "images of opening" and these images take on a more hopeful cast in the second song is also based on the musical connection between the songs, since their transformation into "images of hope" corresponds to the increasing stability of A major as Song 2 unfolds. According to Neumeyer's thesis, what he is describing should essentially break down into two different relationships between the two opening songs of Dickterliebe, one that is based on the text and the other on the music, neither of which creates an organically unified structure in itself. The underlying argument of his article is that it is only by integrating these two relationships that we can explain a pairing such as this in organicist terms. But the pairing he describes is really a single relationship, -with both poetic and musical aspects, and just as he presents the music as an organically unified structure, in the form of a Schenkerlan voice-leading graph, so, too, he describes the texts of the two songs as a unified and continuous narrative, in which, to borrow from the metaphor both of Heine and of the organicists, the budding images of the first poem blossom out in the second, and the doubt and uncertainty are resolved. Ultimately, I find Neumeyer's interpretation of the two poems as a unified pair to be just as contrived as his analysis of the songs as a single musical entity. Some of his observations about the connections between the texts are compelling, but they are two separate poems, which Schumann sets as two separate songs. There is no continuity between them in terms of stanzaic structure, rhyme scheme, or even narrative voice, since in the first poem the narrator tells about the beginning of his love, where in the second he directly addresses his beloved. It is perhaps to explain this lack of continuity that Neumeyer describes the first poem as a prologue, with the story beginning in the second. But the problem with this explanation is that Heine adopts a narrative tone in the first poem and immediately starts to tell his story: "In the lovely month of May, as all the buds burst forth, it was then that love rose up in my heart." By the end of the poem we get no further than the narrator's confession to his beloved, and then the story is interrupted by the second poem. Heine creates an effect not of unifying clarity but of disturbing discontinuity.
Analyzing Dichterliebe
57
The question that Neumeyer hears in the first poem is an implied one, which depends to some extent on the metaphor that Heine's narrator sets up: Will the love that is budding within his heart bloom into a happy union? Will his longing be fulfilled? The second poem does not directly address this question, although an answer is perhaps implied by Heine's use of the same imagery. As the buds in the first poem become flowers in the second and the singing birds become a nightingale chorus, there is a disconcerting shift in the metaphorical relationship between these images and the narrator's love. Now the flowers bloom from his tears and the nightingales' song grows out of his sighs, images that do not evoke the bliss of requited love but the increasing pain of frustrated desire. In the opening poem the narrator implies that the natural process of growth as spring leads to summer will inspire an analogous process in the development of his love, as his first pangs lead to a full-blown love affair. But as we move into the second poem, the process is thwarted, and the expected feelings of hope and fulfillment are replaced by pain and doubt. At the same time, the normal and expected direction of the metaphor — in which the growth of the natural world inspires the growth of human feeling — is reversed. There is something false and unnatural in the notion that the narrator's tears and sighs can cause the flowers to bloom and the nightingales to sing. In this •way, Heine's poem quite literally subverts the metaphor of orgamcism. As Schumann transforms the pair of poems into a pair of songs, he finds ways to convey the discontinuity and the resulting frustration of our expectations in terms of music. Although the piano introduction of Song 1 implies a resolution to Ft minor, the implication is left unfulfilled as the singer enters. He describes the beginning of spring and the harmony clearly establishes the tonality of A major, but as he turns to his supposedly analogous feelings of love the melody rises, first to 4 as the accompaniment tonicizes B minor, and then to 6 as it tonicizes D major. The vocal melody ends on the latter pitch, which resolves in the accompaniment to a chromatic ett , and the song is back to the dominant of Ft minor. It is on this harmony, at the end of the second strophe, that it is ultimately left hanging. Although there appears to be a moment of resolution as the second song begins, it turns out to be fleeting, and Song 2 turns emphatically to A major, thus thwarting the implication of Ft minor in the same way that the first song does. And while in Song 2 the dominant of the latter key does not return at the end, we have seen that there is a more complex way in which Schumann denies resolution even as he presents an apparently straightforward authentic cadence. So the musical relationship between the two songs is analogous to the relationship between their texts, and in neither case can this relationship be expressed m terms of organic unity. The three analysts whom I have considered in this chapter are each concerned with the question of organic unity in Schumann's Dichterliebe at a particular level of musical structure. Komar tries to demonstrate that the entire sixteen-song cycle is "an integrated musical whole," which can be analyzed as a single piece in terms of Schenker's theory, despite the fact that each song is a discrete musical form and virtually all of them are in different keys. Schenker himself tries to
58
The Genre of the Cycle
analyze just a single song from the cycle as a complete and unified structure, in order to illustrate one of the central ideas of his theory, but the surprisingly stubborn difficulties that he encounters suggest that Schumann was no more interested in creating a musical form that is organically unified within a single song than he -was among a cycle of songs. I have juxtaposed these two very different analyses in order to show that as Komar and Schenker approach the song cycle from opposing directions, they set up the vicious circle that I alluded to at the outset of the chapter. Neumeyer, my third chosen analyst, finds himself stuck in this circle, shuttling between the incomplete song and the discontinuous cycle, in the hopes that each will solve the analytic problem that the other creates. One of my purposes in -writing this book is to propose that we open the circle up. In composing his songs, Schumann plays with the conventions of musical structure and questions the aesthetic necessity for closure within small-scale forms. Schumann's songs thus lend themselves to inclusion within cycles, and it is natural for us to look to the cycle to provide, on a larger scale, the closure that is often lacking in the individual song. But it is impossible to find 'what we seek in the cycle, for the reasons that I have discussed in the present chapter. I do not see why we should consider this to be an analytic problem. On the contrary, it is a solution, a key to understanding the meaning and significance of Schumann's cycles, and a hint that we should turn to the aesthetics of Romanticism—the artistic and intellectual movement that influenced Schumann so deeply and with •which he explicitly identified himself— in order to find a more fruitful approach to the genre. In the second part of my book, I will use the example of Schumann's Eichendorff Liederkreii to demonstrate such an approach. But first, in the folio-wing chapter, I will establish a coherent framework for my chosen example by exploring the intellectual origins of the nineteenth-century song cycle. My intent is not to -write a history of the genre but to try to place it within the context of Romantic thought by considering how Schumann and his contemporaries were influenced by the early Romantic -writers and, in particular, by the ideas of Friedrich and August Wilhelm Schlegel. Obviously, these ideas are no more prescriptive or authoritative for our understanding of the cycle than the organicist model that has so strongly dominated recent scholarship. But they can serve as a kind of corrective to that model by reminding us of how far our historical moment is from Schumann's and by encouraging us to question our assumptions and expand our vision.
SCHLEGEL'S FRAGMENTS AND SCHUMANN'S CYCLES
T!
organicist model of musical form is ultimately inadequate for explaining Schumann's song cycles. This failure is ironic, given that the metaphorical use of organic unity as a standard of aesthetic value is one of the greatest legacies of Romanticism and is, in fact, a standard that was embraced enthusiastically by Schumann himself in his musical criticism. It is perhaps for this latter reason that theorists and musicologists continue to operate on the premise that this is the most appropriate model for the song cycle, despite the difficulties that have been encountered in applying it. There is some question, however, as to whether the self-representations of Romantic artists such as Schumann should be privileged in this -way. In recent years, scholars of literature and music have described Romanticism as an ideology that continues to dominate and shape the criticism and analysis of Romantic -works of art and have argued that the tenacity of this ideology has resulted in a lack of critical distance.1 We should be especially reluctant to take at face value the Romantics' adoption of the metaphor of organic unity. Thomas McFarland has argued, for example, that Romanticism can in fact be defined as a sense of incompleteness, fragmentation, and ruin—what he refers to as "diasparaction." He describes the doctrine of organic form as a desperate and ultimately unsuccessful "attempt to abridge the power of the diasparact" and suggests that it was the increasing awareness that their -world and their lives -were fragmented that led the Romantics to imagine and yearn for organic -wholeness in their art. McFarland observes that organicist metaphors involve the relationship between the whole and its parts. He argues that even if the whole is accorded theoretical priority, it is more a hypothetical construct than an objective reality, and so the theory of organic form paradoxically becomes a confirmation of the fragmentary.3 McFarland's argument points to an inherent contradiction in the Romantic metaphor of organicism. Leonard B. Meyer, for example, observes that the Romantic view of aesthetics leads to "a curious, almost paradoxical, dichotomy between the creative act and the aesthetic object." The process of artistic creativity should be spontaneous, uncontrived, and largely unconscious, yet the product that results from this process is evaluated in terms of its adherence to an inevitable inner logic.4 We can restate this paradox in more general terms by ob59
60
The Genre of the Cycle
serving that where the specific doctrine of organic unity prizes immutability and perfection, the broader philosophy of organicism turns imperfection and change into positive values. Morse Peckham has argued that the acceptance of imperfection and change is a natural consequence of organicism, because it is a "philosophy of becoming, not a philosophy of being." The Romantics replaced the traditional metaphor of the universe as "a perfect machine" with a new metaphor, in which the universe is a living organism. Because it is alive, it grows and changes, and because it changes, it is inherently imperfect.5 In applying this worldview to themselves and to their art, the Romantics emphasize gradual and open-ended processes of growth and development. The ultimate goal of such processes is the "emergent self-realization" of one's potentiality and originality, but as Meyer points out, the consummation of this goal is not only impossible; it is undesirable, because it would preclude the further possibility of transcendence, which is -what the Romantics find attractive about the notion of becoming in the first place.6 If-we use the Romantics' standard of organic unity to evaluate their own works of art, then we paradoxically run the risk of miscontruing their aesthetic intent. Meyer, like JVtcFarland, explains the contradictions of organicism in terms of the gap between objective reality and hypothetical ideal. He believes that the underlying structural stability that the Romantics emphasize in their theory of organic form can be seen as an attempt to compensate for the increasingly rapid rate of cultural change in the world around them/ Similarly, as theorists and critics of music were turning to orgamcist models of musical structure in the early nineteenth century, composers were finding coherence and unity to be increasingly problematic, and it seems reasonable to conclude that the ideal posited in theory was a response to the crisis in practice. Such a conclusion is corroborated by the fact that Romantic critics, including Schumann, find the most perfect examples of organic form not in the music of their contemporaries but in that of their Classical predecessors, in particular that of Beethoven. Unity became a problem for Romantic composers because their ideology led them to reject formal and generic conventions and to value originality and diversity, as Meyer has observed. He argues that they use such compositional strategies as the disguising of conventional syntactic patterns and the weakening of closure to manifest their ideological values in their music and to evade the requirements of organic unity.8 Thus the translation of the Romantic doctrine of organicism into musical terms has ultimately led theorists of nineteenth-century music in a very different direction than it did the composers themselves. It is not surprising that the application of orgamcist theories of musical form to the most distinctive genres of Romanticism has been fraught with difficulties. A greater understanding of the formal organization of the song cycle requires a more self-conscious and critical attitude toward the ideological context within which it originated. We must carefully examine the contradictions and ambiguities inherent in the orgamcist metaphor and acknowledge the complexities of determining which aspects of this metaphor are relevant both to Schumann's criticism and to his composition. To this end, I will make a heuristic dis-
SchlegeL'j Fragments ant) Schumann'^ Cycled
61
tinction between two types of organicist metaphors, -which I refer to as dynamic organicum and formal organicuim. The former term was coined by Peckham, who acknowledges that it is somewhat redundant, since "organicism includes dynamism," but justifies its use in order to highlight the fact that Romanticism is a philosophy that is predicated on a belief in infinite growth and, in particular, "to emphasize the importance of imperfection and change" in that philosophy.9 Peckham's term is useful precisely because these latter values are incompatible with the metaphor of organicism as it is typically applied to art. The emphasis on organic unity — in which all of the elements of a-work of art are completely integrated and interdependent — leads to the belief that the work must be in a state of perfection, in which the alteration of even a single detail -would destroy the integrity of the whole. It is this model of unity that I have in mind when I use the term formal orgamcutm. The basic argument of this chapter is that dynamic organicism is a more useful metaphor than formal organicism for explaining and evaluating Romantic art generally and the Romantic song cycle in particular. It is with the -writings of Friedrich and August Wilhelm Schlegel that we find the most ambitious and influential attempt to use what I am referring to as dynamic organicism to explain the concepts of genre, creativity, and aesthetic meaning, and their ideas had an indirect but profound influence on Schumann and his contemporaries. There is no evidence that Schumann ever read the writings of the Schlegels. But whether or not he knew their work firsthand, he was undoubtedly familiar with their ideas through his reading of Jean Paul Richter and E.T.A. Hoffmann.10 We should hardly expect that the literary theory sketched out in Jena would have an unmediated influence on a composer and music critic in any case. As Meyer convincingly argues, the disparate strands of thought that make up a composite ideology such as Romanticism must undergo a process of popularization, in •which they are removed from the specific theoretical contexts in -which they originate and then "re-worked through metaphor, model, and analogue," before they can be transmitted from one discipline to another.11 The tendency of the Jena Romantics to universalize what are essentially literary and philosophical ideas, as -well as their unwillingness or inability to provide clear definitions of their terms and a systematic exposition of their theory, make it tempting to seek out direct pattern matches between the Schlegels' writings and the subsequent Romantic movement in music. But even if the members of the Jena circle do not always acknowledge the specific context of their ideas, we ignore the boundaries that this context sets up at our peril. The present study is not a history of ideas, and it is not my intent to argue that Schumann's music can be explained in terms of the Schlegels' theory. I do believe, however, that Schumann shares certain assumptions about the nature of art -with the Schlegels and that this common ground can be explained, at least in part, by his re-working of a number of significant ideas that were first formulated in Jena. Although Schumann left behind a relatively wide-ranging record of his ideas about music, he did not -write a theory of musical aesthetics, and many of his underlying premises remain unstated.12 As a result, much of the recent scholarship on his song
62
The Genre of the Cycle
and piano cycles has been hampered by misconceptions about his attitudes and assumptions. I believe that a consideration of the influence of the Romantics on Schumann's thought can help to clear up some of the problems that scholars have encountered and lead us toward a more fruitful understanding of these compositions.
Fragments and Sketches The Aesthetic of the Fragment It is in Friedrich Schlegel's writings about the fragment that he explains how the ideas that I am referring to as dynamic organicism are relevant to aesthetics. The fragment is one of the central notions of Romantic literary theory, as it -was formulated by the members of the Jena circle at the turn of the nineteenth century, and yet it is a concept that was never clearly defined. In fact, the Jena Romantics are ambiguous as to where the fragment even fits into literature. Is it a form? A genre? Does it apply to specific works? Or is it a description of the entire body of literature as a (fragmentary) whole? The intentional imprecision with which Friedrich Schlegel uses the term has made it a problematic concept in the recent scholarship of Romanticism. The problem is further complicated by the fact that Schlegel presents his conception of the fragment not as a systematic exposition or theory but in an appropriately fragmentary form. Most of his significant ideas on the subject appear in a collection of aphoristic writings titled Fragmente, which was written collectively and published anonymously in 1798 in the Jena circle s journal, Athenaeum.1^ The Fragmente — along with two other such collections by Schlegel and one by Novalis — belong to the genre of literary criticism known as the fragment, which descends from the French genre of the aphorism, or maxim.14 These critical writings are the only works published in Jena that -were known as fragments. But it is clear that for Schlegel and his colleagues, the fragment was relevant to literature itself and not just to criticism, even if the nature of its application was never specified. The idea that literature is in some sense fragmentary is closely related to the Romantic view of: nature as an open-ended process of continuous and gradual growth. Just as the individual must always be striving toward, and never achieving, self-realization, Romantic poetry is also in a perpetual "state of becoming." In the most famous of the Athenaeum Fragmente, 116, Friedrich Schlegel distinguishes the Romantic style in these terms: "Other styles of poetry are finished, and can now be completely analyzed. The romantic style of poetry is still in a state of becoming; in fact this is its true essence, that it is always only becoming and can never be perfect."15 Schlegel is apparently contrasting Romantic poetry with all other styles of poetry in this passage, but the most important distinction in his writings is between the Romantic and the Classical. Where the former is eternally evolving toward a perfection that it can never reach and is thus infinite, the latter is a perfect and
Schlegel'j Fragments and Schumann '<> Cycled
63
complete poetry but one that has run through the natural cycle of birth, growth, and decay and is now over -with.16 Schlegel's dichotomy between Romantic and Classical thus reflects the basic contradiction in the doctrine of organicism. Where the Romantic style is characterized by continuing growth and change, the Classical style exhibits completeness, order, and unity. To describe the perfect state of Classical poetry, Schlegel turns to the metaphor of formal organicism: "All of the classical poems of the ancients cohere inseparably, form an organic whole, are rightly considered to be just One Poem, the only one in which the art of poetry itself appears perfectly."17 Schlegel frequently associates the Classical with ancient Greece and the Romantic -with the modern age. In one of the Kritidche Fragments, for example, which were published in the journal Lyceum in 1797, he writes: "In the ancients one sees the perfected letter of all poetry: in the moderns one anticipates its spirit in becoming."18 But there are moments in Schlegel's writings when such a historical correspondence no longer makes sense. He concludes the Athenaeum Fragment 116, for example, by universalizing the Romantic. He claims that it is the only style of poetry {Dichtarf) that is not just a style but, so to speak, the art of poetry (DLchtkurut) itself, "for in a certain sense all poetry is or should be romantic."19 It is perhaps more fruitful, then, to consider Schlegel's conception of the Classical and the Romantic not as two historical periods but as two contrary aesthetic principles, one of which corresponds to what I have called formal organicism and the other to dynamic organicism. Thus far we have seen how Schlegel contrasts the wholeness of Classical poetry with the incomplete and dynamic state of the Romantic style. But the metaphor of the fragment, like that of organic form, is most commonly applied to the individual work, and although Schlegel never states explicitly that a work of art is a fragment, he does suggest that a fragment is like a work of art: "A fragment, like a small work of art, must be entirely isolated from the surrounding world and complete in itself like a hedgehog."20 This description of the fragment is characteristically paradoxical, for how can something that is complete in itself be a fragment? Maurice Blanchot argues that with this statement Schlegel alters and inevitably weakens the concept of the fragment, in that "he aligns it with the aphorism, -with the closure of a perfect sentence."21 But Charles Rosen calls attention to the image that Schlegel chooses to illustrate his point. When threatened, the hedgehog "rolls itself into a ball," but even in this shape it has a form that is 'Well defined and yet blurred at the edges," and "it projects into the universe precisely by the way it cuts itself off," thus denying its own closure. The hedgehog represents the peculiar quality of the Romantic fragment: it is at once complete and fragmentary, and the intentionality of its fragmentary state is its most essential characteristic. In this -way it differs from the ruin, the fragment of antiquity, which once had a complete form that has now been broken and could, theoretically, be reconstructed.23 As Schlegel writes: "Many works of the ancients have become fragments. Many works of the moderns are such at the moment of their genesis."24
64
The Genre of the Cycle
In comparing the Romantic work of art to the fragment, Schlegel suggests that it is inherently a work in progress, just as the Romantic style is in a perpetual "state of becoming." He does not conceive of literature as a body of inert and finished texts but as an ongoing process of creation and interpretation, in which the form of the -work itself is not fixed. He expresses the dynamic quality of the work of art in a metaphor that exploits the multiple connotations of the German word Bildung, which can refer to the formation of a man-made structure but also to the formation of a cultured and educated person.25 "A-work is formed [gebiJ2lei\," •writes Schlegel, 'When it is sharply bounded all over, but within its boundaries it is boundless and inexhaustible; -when it is completely true to itself, the same all over, and yet exalted above itself." It must, like a young Englishman, embark on le grand tour, traveling through the "continents of humanity, not to smooth the corners of its individuality, but to expand its vision and to give its spirit more freedom and inner versatility and thus more independence and self-sufficiency."26 Schlegel here elaborates upon the paradox of the hedgehog. Although a work of art, like a person, is discrete and self-sufficient, its very individuality depends upon its limitless potential for growth, upon its ability to point beyond itself and to take on an ever-expanding set of meanings. The gaps and openings within the work are what create its sense of completeness as a -work of art. In his description of the BiGung of the work, Schlegel alludes to the role of the reader in defining and shaping a literary text. A work that is fragmentary will only reveal its full meaning to those sensitive enough to receive it. He writes that a "classical text," by which he means not an ancient text but a text that will become a classic,27 "must never be completely comprehensible. But those who are cultured [gebildei} and cultivate themselves \jich bi[den\ must always be willing to learn more from it."28 By virtue of its essentially fragmentary state, in other words, the comprehensibility of such a text depends upon one's "ability to immediately and at one and the same time idealize and realize objects, to complete them and in part carry them through within oneself."29 Schlegel's description of the fragment as "isolated from the surrounding world and complete in itself" makes it clear that it is not a broken-off piece. The completion of the fragmentary -work takes place through the process of understanding and interpretation and not through its reconstitution within a larger whole. It is true, as Philippe Lacoue-Labarthe and Jean-Luc Nancy observe, that the Jena Romantics never published a single fragment but only collections of fragments. Yet these collections are characterized by the heterogeneity of their contents and the lack of any systematic development, so that, rather than clarify or complete the meaning of the individual fragment, the context provided by the collection emphasizes its fragmentary state. Lacoue-Labarthe and Nancy describe the whole of the fragment collection as a "plural totality": it "does not make up a whole (in, say, a mathematical mode) but replicates the whole, the fragmentary itself, in each fragment." Applied to the work of art, the plural totality suggests an aesthetic that is incompatible with formal organicism. In the latter case, the whole is conceived as prior to its constituent parts, which are necessarily subordinate to its unifying shape. However varied and complex the ele-
Schlegei'<) Fragments ant) Schumann j Cycles
65
ments that make up an organic -whole, they must be completely integrated, and the implications of each part must be fulfilled by its relation to the other parts. In the fragment collection, however, the part cannot be integrated because of its very nature, -which is inherently fragmentary. The relationships among parts and between each part and the whole are thus more flexible than in the case of organic form, and the condition that no part can be altered or removed -without destroying the integrity of the whole cannot obtain. The consequence of this unusual relationship between part and whole is that the collection, like the individual fragments of -which it is composed, is openended, and it, too, depends upon the reader's imagination in order to become completely comprehensible. Rather than a closed form in which structural connections are made explicit, the fragment collection presents a potential structure in -which implicit relationships are left to the reader to be realized. For this reason, the fragment collection is a useful model for the Romantic cycle. Morse Peckham, in language that is reminiscent of Schlegel, describes the cycle as follows: The cycle forces the reader to be creative. The author offers the reader the material for a structure, for developing an understanding, . . . but he does not complete that structure. That is the task, the opportunity, and the pleasure of the individual reader. And in thus becoming creative himself the reader transcends his own cultural limitations of comprehension and intellectual perception.31 The specific work that Peckham is referring to in this passage is Sir Walter Scott's cycle of Waverley novels, but his comments clearly have much -wider significance and, in fact, he subsequently applies them to Heine's poetic cycle Lyruche<> Intermezzo and to Schumann's Dichteriiebe.^ The sense of incompleteness and potentiality that one perceives in Schumann's cycles largely results from the nature of their individual parts. One of the most characteristic features of his songs and character pieces is the way in which he challenges the comprehension of the listener through his play -with formal convention and his use of incomplete musical structures. There are also moments in Schumann's own criticism that reveal a kinship with Schlegel's aesthetic of the fragment. In an 1839 review, for example, Schumann -writes of Chopin's opus 28 Prelude as follows: I would describe the Preludes as strange. I confess that I imagined them differently, like his etudes in the grand style. Almost the contrary; they are sketches, the beginnings of etudes, or if you will, rums, individual eagle wings, all colorfully and wildly mixed up.33 The last sentence of this passage is very close to Schlegel's description of modern poetry, in his Krit'wche Fragmente, as a "multitude of poetical sketches, studies, fragments, trends, ruins and materials," as John Daverio has observed.34 Schumann suggests that, because of their unfinished quality, Chopin's Preluded surprise and even baffle him. And his description of the whole collection—"all
66
The Genre of the Cycle
colorfully and •wildly mixed up" — derives directly from his description of the individual prelude. Schumann's review of Stephen Heller's opus 16 Etiidu, which appeared two years later, is even more explicit: Other, thrifty composers would have built entire concertos and sonatas from many of the basic ideas of these etudes, but our composer prefers only to suggest and to briefly stimulate; his prevailing humor wishes it so, and the silhouette is also welcome.35 In using the silhouette (Schattenrifi) as a metaphor for Heller's Etude
Schlegel'j Fragment*) and Schumann 'd Cycles
67
and establishes an explicit connection between the sketch and the formal idea of the cycle in an essay titled "Uber Zeichnungen zu Gedichten," which was published in the Athenaeum in 1799. August Wilhelm Schlegel's essay is concerned with the creation of a new genre that involves "the interaction of two arts, in harmony and without servitude."37 What he has in mind is the union of poetry and drawing, a more sophisticated equivalent to the then popular practice of accompanying published plays or novels with engraved illustrations. He cites as the model for his proposed genre a series of outline drawings by the English sculptor John Flaxman, which were intended to accompany works by Dante, Homer, and Aeschylus. But he implies that another source of inspiration is the collaboration of poet and composer in vocal music, and Kristina Muxfeldt has suggested that Schlegel's ideas subsequently "found their truest realization" not in the visual arts but in the Romantic song.38 Schlegel points out that it is difficult to illustrate a truly Romantic work of literature, which he refers to as a Roman, because what is often of the greatest significance is also the least visible in the external appearance of everyday reality, in other words, in the images that the visual artist usually represents: A Roman could be excellent and not contain any suitable moment for artistic depiction. It would, however, reveal no particular depth if everything could be made visible. What is just the most significant thing can often be the least evident in external appearance. In charming symbols, the Roman enables us to guess the tenderest secrets of life, which can never be completely pronounced. Poetry here familiarly nestles up against reality and breathes a higher soul into it. It is no longer mere reality, but it should still appear to be so. There is no bridge across which the artist can go from his field to the center of such a poetry, and so he should be content to creep around the edges.39 In suggesting that the poet transforms and exalts reality in the Roman, Schlegel suggests that the spatial and visual aspect of the Romantic portrayal of reality conceals a higher vision of poetic imagination, "-which can never be completely pronounced" but only hinted at. There is thus no question of the visual artist filling in what the poet leaves implied, and Schlegel makes it clear that this is not the intent of his new genre. Rather, the poet and the artist should each translate the other's imaginative vision into his own aesthetic medium: "The visual artist -would give us a new organ to feel the poet, who in turn would translate into his high dialect the charming cipher-language of lines and forms."40 For this reason, according to Schlegel, the sketch is preferable to the painting for illustrating Romantic poetry: The more the visual art stays with its first lightly sketched hints, the more it will act analogously to poetry. Its marks will almost become hieroglyphics, like those of the poet; the imagination will be required to complete and continue to develop according to the received stimulus, rather than becom-
68
The Genre of the Cycle ing captive to the accomodatmg satisfaction of the completely worked out painting. . . . Just as the words of the poet are really incantations of life and beauty, •whose secret power cannot be recognized in their components, so a successful outline is like real magic, so few and sensitive strokes in which can live so much soul.41
Schlegel, like his brother Friedrich, describes artistic creation as an ongoing process. His emphasis on the active involvement of the imagination—not only the artist's but the viewer's as well — helps to explain why the sketch is such an attractive metaphor for the Jena Romantics. The completed painting satisfies the viewer's imagination, turning it into a passive receptor, but the sketch acts as a stimulus that he or she must "continue to develop." The work of art can thus bypass the external world altogether and remain completely in the realm of the imagination. Schlegel points out that the kind of work he is describing cannot be understood by everyone: "Of course, one must already have used his imagination like a painter and have seen many complete artworks in order to successfully read this language."42 He again echoes his brother in concluding that the comprehensibility of such an art depends upon a viewer who has the sensitivity and cultivation to understand it. Another problem that is inherent in comprehending a drawing intended to accompany a literary work is that its meaning is dependent not only upon the viewer's imagination and artistic knowledge but also upon the -work that it is illustrating. Such a drawing is thus not "self-sufficient in its expressiveness," which, Schlegel acknowledges, is a valid requirement for a -work of art. But he argues that there are precedents for extending special license to the artist. First of all, he or she should be free to allude to circumstances that are presumed to be common knowledge: "Such a circle of myths or legends is thus to be considered as the common poem of a people or time, the knowledge of which is expected of each individual." A second exception arises in the case of a "cycle of paintings," since "in the cyclic form, scenes may occur that only receive their full meaning through what precedes or follows." Schlegel then argues that, by analogy with these two cases, the same license should extend even further, "where one art uses only a part of its means to ally itself with another."43 Schlegel attributes his comment on cyclic form to an anonymous art critic, thus implying a preexisting common usage for the term, at least in reference to painting. In the present case, however, Schlegel is using the notion of the cycle to make a more general point about aesthetic meaning, and there is no reason to assume that cyclic form must be limited to the visual arts. Yet his reference to this form is casual and relatively isolated, and so one hesitates to read too much into it.44 For the moment, I will simply observe that Schlegel hints at a logical connection between the possibility that a work's expressive meaning may be incomplete in some sense and the placement of that work -within a cycle. His emphasis on the individual part, rather than the whole, is the most provocative aspect of his description, as Muxfeldt points out: "It is not just that a work might gam in meaning by association -with something that lies outside its boundaries, but that
Scklegel'ff Fragments and Schumann d Cycled
69
the work Lt
Romanticism and Genre Genre, Displaced In comparing the fragment to a -work of art, Friedrich Schlegel reveals some of the principles of his aesthetic philosophy but tells us virtually nothing concrete about the work itself. He does not provide a definition or even a description of what the Romantic work of art should be like, and if we -were to set out in search of such a work armed only with his instructions, we could conceivably return •with almost anything. The reason for this is that Schlegel's is not so much a descriptive or even a prescriptive theory as it is a kind of literary utopianism. Despite his characterization of Romantic poetry as perpetually becoming and never reaching perfection, Schlegel clearly has some ideal vision of -where it will end up. Just as the poems of the ancients form a single organically unified poem, he believes that the perfect literature of Romanticism shall be a single "infinite book," a new Bible.47 In Schlegel's idealized history of literature, Romantic poetry will eventually achieve a higher unity than that of Classical poetry. One important consequence of Schlegel's theory—which had a profound influence on the history of nineteenth-century music — is that Romantic literature does away -with the concept of genre altogether. At the end of literature, as Schlegel envisions it, all of the separate poetic genres will be reunited and poetry
70
The Genre of the Cycle
•will come into contact with philosophy and rhetoric.48 In its current transitional "state of becoming," Romantic poetry has already rendered the Classical genres "ridiculous in their strict purity," but as yet there is no perfect unitary genre to take their place, so we are left -with a confusion of genres.49 In part, this confusion results from the emphasis that the Romantics place on originality and uniqueness. In his literary notebooks Schlegel suggests that "every poem is its own genre," and in one of the Athenaeum Fragmente he writes that mongrel or deviant genres (Abarten) of poetry have value for Romanticism, even if they are "eccentric and monstrous," as long as they are original. Paradoxically, the unique, eccentric, and original are worthwhile because they provide "materials and preliminary exercises for universality."61 Schlegel uses a metaphor from astronomy to explain how the current confusion of genres has come about and -why it is a necessary stage in the development of literature. He compares the traditional conception of what he calls the poetical •world system (poetidchen Welfoyjtem) to the way the universe was conceived before Copernicus: The common divisions of poetry are only a dead framework for a limited horizon. What one can produce or what is now current is the resting Earth in the center. In the universe of poetry, however, nothing rests, everything grows and changes and moves harmoniously; and even the comets have unalterable laws of motion. But until the course of these celestial bodies is calculated, their return predetermined, the true -world system of poetry cannot be revealed.52 Schlegel makes it clear that while the poetry of his day appears to be in a state of confusion and perpetual change, there is, in fact, an underlying system that unifies the poetic universe and governs its evolution, and this system 'will ultimately become known. Elsewhere Schlegel conceives of this system as a unitary work of art, which he refers to as the Roman. As Hans Eichner has persuasively argued, Schlegel's theory of Romantic poetry is essentially a theory of the Roman, and these two terms are used as virtual synonyms in much of his -writing.53 While the Roman is the only literary genre that Schlegel describes concretely and exhaustively, it has ironically proven to be one of the most confusing aspects of his theory of literature. The reason for this is that his use of the term overlaps with the more limited sense that is far more common and refers to a genre that includes Goethe's WdkelmMeuiter, Schlegel's own Lucinde, and Novalis's Heinrich von Ofterdingen and Die Lehrlinge zu Sau, as -well as many of Jean Paul Richter's books. In reference to these works, Roman is normally translated as novel, but -while Schlegel includes them as examples of Romane, there are many other novels that he does not include.54 However, Schlegel's genre is broader than that of the novel, in that it includs works of Shakespeare, Dante, and Ariosto. Furthermore, since the Roman is the ideal book that is the goal of all literature, there is a sense in which it is not really a genre at all but a displacement of genre. Because of the particular role that the Roman plays in Schlegel's idealized history, he is least interested in those aspects of his own and his contemporaries' works that connect them to the larger tradition of the novel and most interested
Schlegel'j Fragments ant) Schumann '<> Cyc/ej
71
in the idiosyncratic elements that set them apart as a unique body of literature. For Schlegel, in other words, the Roman is not necessarily a work of prose that has a coherent narrative, but it is a work that includes within itself a variety of literary genres and forms, that has a structure that is broken or discontinuous, and whose subject matter moves beyond concrete reality and into the realms of imagination, fantasy, and spirituality. In the "Brief iiber den Roman," which was published as part of his Gedprdch fiber die Poedie in 1800, Schlegel questions the assumption that the narrative genre of Classicism—the epic—is the genre to which the Raman is most closely related. He maintains "that a song can as -well be romantic as a story" and that the Roman is a "mixture of narrative, song and other forms."55 Schlegel also argues that the Roman does not become a unified whole through "the dramatic coherence of its story" but rather "through the connection of the entire composition on a higher level of unity than that of the letter, which it often should and does disregard, through the bond of ideas, through a spiritual central point."56 In short, the Roman is the fragmentary genre par excellence, an exemplar of dynamic organicism. Its form is heterogeneous, more a plural totality than a coherently unified whole. Its meaning is not revealed simply by grasping its narrative but by understanding the more elusive higher ideas toward which it points. The Roman is also fragmentary in a larger sense, in that it is a genre that is still in a state of becoming and is yet to realize its full potential. It can be seen as an emblem of the problematic status of genre in Romanticism generally, since it calls into question the distinctions upon which the definition of the literary genres is based. Schumann and the Problem of Genre A generation after Schlegel -wrote about the confused state of genre in the literature of his day, Schumann described a similar confusion in the music of his. Throughout his decade as the editor of the Neue Zeit
72
The Genre of the Cycle
paring it to what came before or what followed. Linda Correll Roesner, on the one hand, contrasts it with two statements that Schumann published four years earlier and concludes: Schumann's writings of the mid to late 1830s show that as a critic he became increasingly convinced that the sonata had run its course. Whereas in 1835 he was speaking optimistically of the sonata's embodying the ideals of total unification and the "organic" fusing of form and spirit, by 1839 he considered the genre to be at a standstill.68 Rosen, on the other hand, argues that although sonata form was not a very natural outlet for the major composers of Schumann's generation, "they could not do without its prestige." He compares Schumann's 1839 article to his statements from the early 1840s and concludes that while Schumann initially believed that the form was dead, he eventually acknowledged the necessity of bringing it back to life: After having proclaimed that the sonata was dead and that one could not go on repeating the same forms forever, Schumann, only a few years later, lamented that there were too many composers who could write short pieces — nocturnes, songs, and so forth; what was needed was a composer of sonatas, symphonies, and quartets—that is, of sonata forms. These incompatible interpretations of Schumann's 1839 article point to an ambivalence in the text itself. For -while it is true that in advocating a turn to something new Schumann seems accepting and even enthusiastic about the death of the sonata, the article as a whole projects a tone that is resigned and •wistful. He -writes of sonatas that "it is always a pleasure, among the colorful confusion of fashionable portraits and caricatures, to once again encounter one of those honorable faces -which -were once the order of the day, but are now exceptional." Schumann observes that the composers who were writing the most sonatas at this time tended to be young and unknown, -while the older composers, "who grew up when the sonata was still in its prime," wrote the least. The unfortunate result -was that most of the sonatas being -written at the time were "to be regarded only as a kind of specimen, as studies in form; they are hardly born of a po-werful inner impulse. The older composers must have their reasons for not composing any more, which we leave it to others to guess." When -we consider the complete passage, then, it becomes clear that Schumann is not arguing that it is just as well the sonata is dying but that for a variety of reasons it has lost its vital essence and, while this is to be regretted, it is probably inevitable. His review certainly does not support the claim, which Rosen makes, that Schumann started out as "an implacable enemy" of sonata form. And if one goes back earlier, such a claim becomes even more implausible. In an 1835 review of sonatas by Mendelssohn and Schubert, for example,
Scklege['<> Fragments ant) Schumann',/ Cycles
75
Schumann writes that, to the Davidsbiindler, the sonata is the "most cherished artistic genre of piano music." ^Vhile Schumann clearly regrets the sonata's moribund state, however, he does not advocate a return to the Classical style but a turn to "that which is new." And what most seems to trouble him about the decline of the sonata, as Rosen surmises, is that no other instrumental form enjoys such great prestige. In the 1839 article, in fact, Schumann speculates that the reason that younger and unknown composers write so many sonatas is because "there is no worthier form through which they can introduce themselves and favorably impress the important critics."63 And his frequent advice to the most promising of his contemporaries to move beyond the smaller forms and compose in the larger ones is undoubtedly based on the same premise. Schumann's view of genre is largely analogous to Schlegel's. Schumann believed that the Classical style had been brought to completion and that the Classical genres — those based on the sonata cycle — had run their course. But since nothing had yet appeared to replace the sonata, the music of Schumann's time, much like the literature of Schlegel's, was transitional, and the Romantic genres were in a state of confusion. And although he lamented this confusion, Schumann, again like Schlegel, valued the emphasis on originality that was partly responsible for it and was attracted to the potential for self-realization and growth that he saw in the smaller genres with which Romantic composers -were so preoccupied. In his review of Stephen Heller's opus 16 Etudes, for example, Schumann complains that Heller "squanders his riches on such small forms" but also praises him for his individuality. Schumann writes that in addition to the "fingerwork" that -would be expected in such etudes, one finds more, "namely character pieces in colorful succession."64 Schumann's enthusiasm for using an ostensibly pedagogical genre to compose music that is expressive in intent is typically Romantic. And in describing Heller's works as "character pieces" Schumann points to the conflict between this desire for personal expressivity and the concept of genre itself. The character piece calls into question the very possibility of genre, because it is, paradoxically, a genre that is defined by the unique individuality of each of its members. The essential element of "the characteristic, " which gives it its name, defies the possibility of any formal or even functional conventions.65 The extent to which the personal and unique nature of the individual character piece threatens the clear definition of the genre of which it is a part is demonstrated by the wide range of titles that are attached to such pieces, as well as the nature of many of those titles—fantasy, impromptu, prelude, caprice, romance, reminiscence, song without words. To argue that each of these titles represents its own genre only compounds the problem, for then one is left with an open-ended series of closely related genres and no systematic means of distinguishing one from the other. Schumann's practice in the Neue Zeit
74
The Genre of the Cycle
"Konzerte fur Pianoforte," "Symphonien fur Orchester," and so forth. It is not surprising that a genre so strongly predicated on personal expression was unable to replace the sonata as the most prestigious type of musical composition in the early nineteenth century. By its very nature, the character piece cannot hope to aspire to monumentahty and universality, both of which would be essential to the achievement of the ultimate Romantic genre. But if Schumann recognized that the future of musical Romanticism could not be found in the composition of small forms, he saw in the potential for self-realization provided by such pieces the pathway toward that future. He expresses this in his review of Heller's Etudes by referring to them as "beautiful seeds" that "give claim to more beautiful hopes," adopting a metaphor that Friedrich Schlegel and Novalis both use to describe Romantic poetry's "state of becoming."66 Schumann's 1841 critique of Chopin is based on the same premise. But where in Heller's case Schumann sees the promise of greater self-realization despite the fact that Heller has thus far restricted himself to small forms, in Chopin he sees a "considerable power of originality" that has enabled him to create an "abundance of new forms, "yet he worries that Chopin -will not progress any further than he already has. Schumann writes that "Chopin could now publish everything anonymously, one would still recognize him. In this lie praise and blame together, the former for his talent and the latter for his aspiration." Although his music is "always new and inventive externally, in the shape of his pieces and his use of special instrumental effects, internally he remains the same." For this reason, Schumann feels that Chopin -will never realize his full potential and will not have the influence of -which he is capable. Originality and individuality are essential talents for a Romantic composer, but they must be placed in the service of continuing growth and change. Only in this way will Romantic music evolve to its more perfect future state.67 As with Schlegel's ideal of Romantic literature, we are not given any explicit sense of ho-w Schumann envisions this state. But one crucial difference is that, for Schumann, the Classical style remains within recent memory, and its genres, however sickly they may be, constitute an ongoing tradition and not, as is the case for Schlegel, a remnant of a distant historical era. So while the current confusion of genres exemplified by the character piece cannot lead back to the Classical sonata but must lead ahead to some more perfect and more Romantic genre, the sonata cannot simply be buried and forgotten. As Rosen argues, "The old aesthetic of the hierarchy of genres still ruled almost unchallenged, although Schumann's own work, along -with that of Chopin and Liszt, had already shaken it to its foundations."68 The Genre of the Cycle Schumann's compositional oeuvre suggests that one solution to the problem of genre is to continue writing in the Classical genres but to transform them into something more Romantic. Schumann's sonata form, much like the Romantic character piece, is relatively unrestricted by formal conventions, and each move-
Schkgel'tf Fragments and Schumann'^ Cycled
75
ment that he composes in the form tends to be uniquely different from every other. Roesner and Joel Lester, each of whom have recently come to the defense of Schumann's sonata form movements, both describe his approach as a reinterpretation of the sonata that is far more flexible than that of his Classical predecessors and that enables him to tailor the form to the individual character of each piece. It is not surprising, given such an approach, that the reception of Schumann's sonatas has been problematic. Lester is right to argue that Schumann's large-scale instrumental works "need not be set against Beethoven's and then criticized for failing to do what Beethoven did," but by explicitly identifying his works as members of the same genres in which Beethoven composed Schumann himself invites such criticism.69 In this context we can view the cycle as an alternative, and ultimately more successful solution, to the problem of fashioning a Romantic genre that can replace the Classical sonata. Rather than treating the sonata as a model and Romanticizing it, Schumann succeeds in ignoring the Classical genres and instead uses the small form as the basis of a large-scale •work. \Vith the single exception of Beethoven's An di£ ferru Getiebte, there are no cycles by any of the Classical masters with which to compare Schumann's, and one could argue, in fact, that Schumann's relationship to the cycle is analogous to Haydn's and Mozart's relationship to the Classical genres. Just as the two great Classical composers took typical eighteenth-century genres such as the piano sonata and the string quartet and raised them to a level of sophistication and refinement such that their works became the models for later generations, so Schumann took a genre that had been developing from the beginning of the nineteenth century and brought it to a new height, creating a body of classic -works. It is a genre, however, that embodies the confusion and imperfection of Schlegel's Romantic style. The genres of the Classical style have a clear sense of definition that stems from their adherence to a particular set of conventions. Because the genre of the cycle is based upon the character piece and the song, such conventions are impossible, resulting in problems of generic definition that can be illustrated by contrasting Schumann's cycles -with one of the Classical genres. In each of Mozart's Viennese-style string quartets, for example, the demands of the genre determine the external shape of the whole work to a great extent—dictating the number and order of the movements, the form of each one, and the relationships among them in terms of tempo and key. These generic conventions, which are further strengthened by the place of Mozart's quartets within the larger family of sonata-cycle genres, enable us to determine the boundaries of the genre and to explain any exceptions we might encounter. A Classical string quartet must have four movements, of which the first must be in sonata form and the third must be a minuet and trio. There is a limited range of possible forms for the second movement, but it must be the slowest in terms of tempo, and it is the only one that is not in the same key as the other movements. The Classical string quartet is defined as much by these conventions as by its scoring. There are no analogous conventions in the case of Schumann's cycles, because the shape of each one is unique. How, for example, would we determine
76
The Genre of the Cycle
the typical number of movements in a cycle? And how do the tempo and form of a given movement relate to its position? For that matter, what forms could be considered typical? One does not usually encounter such questions in the literature on Schumann's cycles, because it is generally agreed that they are irrelevant to the definition of the genre. But the absence of such conventions makes it impossible to use the order of the movements as a defining element as well, despite the efforts of recent scholars, since there is no way to determine what that order should be. The pattern of key relationships between the movements of a cycle is cited even more frequently in defining the genre, as we have seen, but there is really no generic convention in this case, either. Again, a comparison with Mozart's Classical quartets may be useful. Given the key of the first movement of a Mozart quartet, we know that the remaining movements—with the significant exception of the slow movement—will be in the same key. And while the key of the slow movement is not absolutely fixed in relation to the key of the first, the latter does determine the range of options that are possible, as •well as the likelihood of any one of them actually occurring. The slow movement in any sonata-cycle work is typically in a key that is diatonically related to the key of the first movement, and the subdominant relationship is the most common. In Mozart's quartets, this relationship occurs in nine of the sixteen works. Of the remaining quartets, the most anomalous examples are those in which there is no change of key at all for the slow movement but merely a change of mode (K. 168 and K. 173). And those in which the slow movement is in the dominant key (K. 170, K. 575, and K. 590) are also atypical, since this relationship is considered to be in the "opposite direction" from the subdominant.72 Given the key of the first movement in a Schumann cycle, there is no -way to predict that of the second or of any of the others. It is no more typical for a majority of the movements to be in the same key, as in the Novelletten, than it is for virtually every movement to be in a different key, as in Diehterliebe. There is not even a convention that governs the relationship between the key of the first movement and that -with -which the cycle ends. As I have argued in chapter 2, the preference for diatonic key successions by third or fifth is a matter of style, and since such successions are typical of many early nineteenth-century works, they play no role in distinguishing the cycle as a genre. Even those scholars who claim that the succession of keys between movements is the most significant defining element in Schumann's cycles acknowledge that this parameter is not governed by convention. Barbara Turchm writes, for example: Even a cursory overview of the 1840 song cycles reveals considerable diversity in their tonal designs. While some confine themselves to a small range of keys, others traverse a far broader realm. While some cycles are tonally closed, returning at the end to the key with which they began, others remain tonally open.73 As with his sonata form movements, the individual takes precedence over the generic in Schumann's cycles. But where in the former case there is at least
Schlegel'j Fragments arid Schumann '<) Cycled
77
a connection to a Classical tradition that provides a model for comparison (however much this comparison has negatively impacted the reception of Schumann's compositions), the cycle is a truly Romantic genre, since the lack of virtually any generic conventions suggests that, to paraphrase Schlegel, every cycle is its own genre. The cycle, like the character piece upon which it is based, is in a state of confusion and transition and is ripe for further growth and development.
The Cycle in the Nineteenth Century The considerable difficulties that have plagued our attempt to define the cycle as a genre and, in particular, to distinguish it from a collection of pieces are at least partly the result of our failure to recognize that Schumann and his contemporaries, like the Jena Romantics, had a far more flexible conception of genre than we do today. The central role of the song and the character piece in the early nineteenth century is indicative of the extent to which composers were becoming more interested in the unique individuality of each work and less concerned with the strict adherence to generic conventions. When a group of such pieces were put together for publication, they might be related to one another in a variety of ways. The use of the term cycle implied the closest possible kind of relationship among a set of short pieces in the nineteenth century, just as it does today, but where we are now likely to see a difference in kind between cyclic relationships and those that may be found in other published collections, to a nineteenth-century composer or critic the difference was one of degree. In terms of Fnedrich Schlegel's theory of aesthetics, the cycle is an example of the Romantic style. The pieces of which it is composed tend to have an unfinished and open-ended quality that results both from their extreme brevity and from the way in "which they destabilize conventional patterns of musical structure. The formal organization of the cycle is similarly open-ended. It depends more upon implied relationships between its constituent pieces than it does upon concrete structural connections, and it allows for a wide degree of variability in terms of the order and number of those pieces. And because the cycle is not clearly distinguished from other collections of songs and character pieces, it is a transitional genre that presents unlimited potential for variety and change. Perhaps the clearest evidence that this is how the cycle was conceived in the nineteenth century can be found in the song reviews that were published in contemporary music journals. But these reviews are problematic as historical documents for a variety of reasons and, "with the notable exception of Turchin, have been largely ignored by recent scholars. They are •written for a general audience and not for professional musicians, and so they tend to be imprecise and superficial. And since nineteenth-century critics are usually not very explicit about their aesthetic premises, it is easy for us to misinterpret their reviews by reading them from our own perspective and to miss the most valuable evidence they have to offer. Daverio observes, for example, that contemporary critics do not provide us -with much help in determining -which collections of songs should be considered cycles, because they typically limit their reviews to "song-by-song descrip-
78
The Genre of the Cycle
tion. 74 But this observation is quite significant in itself. If nineteenth-century critics are more interested in the individual song than they are in the larger whole, then it is unlikely that they would define the cycle as an integrated and unified structure that can be clearly distinguished from other song collections. Turchin presents a comprehensive survey of the song reviews that were published in the first half of the nineteenth century and uses the material she finds in them to make a number of interesting observations about the Romantic song cycle. The reviews that Turchin cites reveal that a wide range of criteria were used to characterize the cycle and that there was very little consistency and no clear development over the course of the period in terms of how the cycle was defined. Because of the assumptions that she herself brings to her primary material, however, she imparts a misleading sense of historical coherence. Turchin, like virtually all modern Schumann scholars, believes that the cycle can be distinguished from other published collections by the presence of a structural coherence that binds the songs into a whole and that the source of this coherence lies both in motivic relationships between the songs and in a logical arrangement of closely related keys. She claims that Schumann and his colleagues at the Neue, Zeitdchrift fiir Miuik were the first critics to recognize that the song cycle is a coherent musical entity and that the most important means of musical coherence that they describe is a pattern of close key relationships among the songs/5 Turchin thus sees Schumann's circle as a progressive and influential force precisely because she believes that they were the first to define the cycle as modern-day scholars do. In fact, however, the Neue Zeitjchrift critics refer to the key relationships within song cycles very rarely, and there is no evidence that Schumann and his colleagues considered them to be a source of structural coherence or even a particularly significant criterion in defining the cycle as a genre. It we consider the song reviews that appeared in the Neue Zeitdchrift during Schumann's tenure as a whole, they suggest that his view of the cycle -was consistent with that of his contemporaries and thus radically different from ours. Elective Affinities One of the most explicit and provocative nineteenth-century descriptions of the song cycle is found in Oswald Lorenz's review of Ferdinand Miller's Sechd Lieder von F. Riickert, opus 18, which appeared in the Neue Zeitjchrift in 1841. Lorenz, one of the regular song critics at the journal, was also one of Schumann's closest associates and, in fact, took over as editor on two occasions — once when Schumann took an extended trip to Vienna during the winter of 1838—39 and again after Schumann resigned in 1844. I quote Lorenz's review in its entirety: The poems are all taken from Liebe
Schlegel'j Fragments and Schumann'^ Cycles
79
technical treatment, an elective affinity that is of a more sensitive and spiritual nature reveals itself in the conception of the poems. It is scarcely hinted at in the wording, and through a few delicate strokes, so that the particular position of this or that song does not create an isolated, rounded-off whole, but rather the member of a chain: thus through a sketchhke form, as in that of the third song, or through an inconclusive harmonic cadence, as in the second. The first song could be considered as an independent entity, but this could not be justified for the others. It is precisely in this loose coherence, however, in this exposition only lightly hinted at, as in the meaningful selection of the poems, that the mature taste of the cultured artist reveals itself. Likewise, the skillful hand of the intelligent and well-trained composer can be recognized in the complete technical finish, declamation, harmony, and the like; in the ease and sureness of the work; and in the tactful handling of seasoning and ornament. Considered from a purely human viewpoint, the songs are so peaceful and cheerful in nature, so light and friendly in their expression of the phases of emotional life, that they can count on securing many warm friends. Lorenz refers to the connections among Miller's songs in a variety of ways, all of which point to a form that is flexible and open-ended. He describes the cycle as a thread and as a chain, images that suggest a series of songs in -which the number is variable and there is no necessary beginning or end point. It has a coherence that is spiritual and is conveyed by light hints and delicate strokes. It is interesting that for Lorenz the source of this coherence is found in the sketchlike and inconclusive nature of the individual song and not in the overarching unity of the cycle. In fact, there is no mention anywhere in Lorenz s review of the kind of all-encompassing structure that is typically evoked in modern scholarship on the cycle. The more exotic concept of "elective affinity," which alludes to the title of Goethe's 1809 Roman Die Wahlcerwandtjchaften, suggests even more strongly that Lorenz does not consider the structure of the cycle to be predetermined and unchanging. The term was apparently first used by a Swedish chemist, in the late eighteenth century, to explain the fact that chemical elements that are opposites — such as a base and an acid — often have an attraction, or affinity, to each other, but that the bond between them can be broken when a third element is introduced, thus forming a new combination/8 One of Goethe's characters offers the following illustration: What we call limestone, for example, is a more or less pure chalky soil, bound closely with a volatile acid that is familiar to us in its gaseous form. If one places a piece of such stone into dilute sulphuric acid, the acid engulfs the lime and appears with it as gypsum, while the volatile gaseous acid escapes. Here is a separation from which a new combination is produced, and one now believes that it is justifiable to use the term elective affinity, because it really seems as if one relationship is preferred over the other, as if one is chosen over the other.79
80
The Genre of the Cycle
In using the term elective affinity to describe the songs of Miller's cycle, Lorenz seems to have in mind several different aspects of this concept. He suggests that because the words of the individual song consist only of "scarce hints" and because its musical structure is fragmentary or "sketchlike," it has a tendency to pair with another. He alludes to the pliant quality of such pairing relationships by observing that they are present in the cycle, even though Hiller has taken Riickert's poems out of their original context and rearranged them. And Lorenz refers to the tendency for opposites to attract when he writes that the elective affinity in Hiller's cycle reveals itself, "despite the diverse technical treatment" of the poems, by which he is presumably referring to the contrasting musical settings that Hiller composed for them. In the notion of elective affinity, Lorenz found a single metaphor that could embrace many of the qualities that he and his contemporaries used to characterize the song cycle: the selection and arrangement of the poetic texts, the degree to which the songs contrast -with one another, the extent to which the individual songs are incomplete in and of themselves, and the fact that this fragmentary quality implies coherent relationships among them. His metaphor suggests that the cycle results more from the tendency of one song to pair with another than it does from a unified structure that binds the songs together. In general, the issue of musical coherence or unity among a set of songs is rarely addressed by nineteenth-century critics, but in the few cases where it does come up it is almost always associated with the question of whether the individual song would be viable outside the cycle. We read in the Neue Zeifochrift of Georg Vierling's 1852 Cyciiu arabucher Dichtung, for example, that "this cycle is composed of 5 songs, of-which none can separate itself from the others. They express a coherent image . . ."80 Similarly, a critic for the Leipzig Allgemeine miuikaLitche Zzitung writes of Carl Lowe's Edther that although the cycle consists of five songs, "none of them can or should exist by itself, be sung separately, or be satisfying by itself. They are one (or should be), like the scenes or acts of a drama."81 In a review of the same composer's Gregor auf Jem Stein, however, Lorenz describes the work as "a cycle of 5 songs, each of which is in itself musically closed, but which . . . together form a whole."82 All of these reviewers define cyclic coherence in terms of the individual song. Either the cycle at hand is a whole despite the fact that each song is complete in itself or it is a whole because each cannot stand alone. The coherence of the cycle is not imposed by the structure of the -whole but grows out of the implications of the part. We should not be surprised, given such a conception, that the borderline between cycles and other collections of songs is not very well defined. The -way in which composers designated their song cycles and critics referred to them further demonstrates this lack of definition. That two of the critics I have cited explicitly refer to the sets of songs they are reviewing as cycles is exceptional. The absence of this term in Lorenz's review of Hiller's opus 18, a -work that was published as Sechd Lieder von F. Riickert, is more typical of nineteenth-century practice. Most published song collections -were designated simply as Lieder, Gedichte, or Ge^dnge. This does not necessarily signify that they -were
Schlegel'
81
considered mere collections, however, since in referring to a given publication critics use these neutral designations interchangeably with characteristic titles such as Liederkranz, Liedenpiei, Lie'derroman, and, more commonly, Liederkreut and Ltfdercykliu. The latter two terms, which are the most familiar to us today, have been synonymous throughout their history. We currently make no distinction between these terms in English, for example, translating both of them as tiong cycle. The first definition of either of them, which appeared in 1865, is a joint definition of both. And in the earliest published instance in -which the term Cycliu is used to refer to a set of songs, an 1818 review of Conradin Kreutzer's Nean Wanderlieder von Ublarid, the same work is referred to as a Liederkreid a fe'w sentences later.83 The lack of a consistent terminology to refer to the song cycle suggests that the various terms that were used did not refer to a set of subgenres that were clearly organized and distinguished from one another but reflected a general confusion as to ho-w the cycle should be characterized and defined. And the fact that we cannot tell from the title of a work whether it is a cycle or merely a collection suggests that this was not a significant distinction in the nineteenth century. The Role of Key Relationship<> The early nineteenth-century reception of Beethoven's An die feme GeLiebte, a •work that does stand apart from other song collections because of its structural integrity, provides further evidence that organic unity was generally not associated with the cycle. While critics universally praised Beethoven's work and even singled out the unity of the songs as one of its most distinctive traits, An die feme Geliebte was regarded as an anomaly and had virtually no influence on the development of the genre. The tone of the earliest review, -which appeared in the Leipzig Allgeineine miuikalLicbe Zeitung in 1817, reflects the almost mythical stature that Beethoven had already attained during his lifetime. It begins as follows: With joy, and with thanks to the master who granted them to him, the reviewer announces these lieder, which are closely connected both poetically and musically. Certainly, all readers who have become familiar with them will agree with him when he counts them as the most beautiful that we have received in many years. They will hardly contradict him when he asserts that, particularly in regard to imagination and feeling, they belong among the most beautiful that we possess at all. He cannot refrain from giving an account of the most essential aspects, even though Beethoven's name, and this general declaration, should make it unnecessary. The reviewer goes on to enumerate some of the ways in -which Beethoven creates close connections among the songs, mentioning the transitions between each of them and spelling out their key relationships. After describing the first song, for example, he -writes: "Now follows a short and sudden interlude, forming a transition to the second, somewhat cheerful lied, that is as effective as it is graceful and pleasant." And he -writes that the "second lied turns directly into the
82
The Genre of the Cycle
third (G major to A flat major)." The critic also points out that the last song is a varied return of the opening: The short turn to C minor [at the end of the fifth lied] makes the return, in the last lied, of the key, as well as the mood and loveliness of the first lied, all the more magnificent. It is truly splendid that, toward the end, even the poet lets the first lied itself again become discernible, only condensed, ending with a free and heartfelt close. Thus, at the same time, the whole work concludes as a true IAeder-Kreu>, completely satisfying and according to plan.84 The fact that this critic singles out those aspects of An die feme Geliebte that twentieth-century commentators have found significant and even suggests that the recapitulation at the end makes the work "a true Lieder-^/frw" may lead us to assume that Beethoven's work enjoyed the same exemplary status in the nineteenth century that it does today. A series of further editions and several more reviews in the major music journals of Berlin and Vienna during the 1820s and 1830s attest to its continuing popularity. And, following the lead of the first review, subsequent critics continue to call particular attention to the unity of the songs.85 Nevertheless, other song composers did not follow in Beethoven's footsteps, and only rarely did they imitate the techniques that he used to unify his cycle. It is also striking that, in reviewing other song cycles, nineteenth-century music critics did not complain about the absence of these techniques or suggest that Beethoven's cycle should be held up as a model. On the one hand, it is still generally acknowledged that the use of modulatory piano transitions to create continuity from one song to the next makes An i)ie. ferne Geliebte unique among nineteenth-century song cycles. On the other hand, scholars have used its tightly knit tonal structure as a model for subsequent cycles and as a defining criterion for the genre. The problem is that these two aspects of Beethoven's cycle are inextricably intertwined. Because An die ferne Geliebte is a single continuous movement, rather than a series of discrete songs, one can hear the keys of the inner songs as extensive excursions within the tonal framework provided by the outer ones. In a more typical cycle, each song is a separate musical entity that has no structural continuity with the song that follows, no matter how inconclusive, tentative, or ambiguous its tonality may be. For this reason, we cannot hear the tonal events of two songs that are in different keys -within the same frame of reference. There are other composers, including Schumann, who evidently exercised care in choosing keys within their song cycles. But there is a crucial difference between the significance of these choices and those that Beethoven made in An Die ferne Geliebte. In the latter case, the choice of keys becomes a matter of structural necessity, comparable to the modulatory scheme in a large-scale instrumental movement. For other nineteenthcentury cycle composers, the choice of keys is a matter of aesthetic preference. This is -why there are no rules that govern the succession of keys between the movements of a cycle and why it is impossible to use key relationships to distinguish a cycle from any other set of songs or piano pieces. When we read a reference to key relationships in a nineteenth-century song
SchkgeL'j Fragments and Schumann'^ Cycled
83
review, we are likely to assume that it implicitly alludes to musical structure, but this is not necessarily •what the critic intends. In his lengthy review of Carl Lowe's Edther, for example, which appeared in the Neue Zeit
84
The Genre of the Cycle
Lowe's cycle creates a logical pattern, but it primarily results form the fact that each song in the cycle is succeeded by one in a closely related key—either the subdommant, relative minor, or parallel major. We cannot assume that such a pattern of keys -would carry the same structural implications for Lorenz as it does for a twentieth-century scholar. It is not the order of the keys that he finds significant but merely the fact that they are closely related, and it is striking that he does not even identify them. We find another reference to the tonal relationships among the songs of a cycle in an anonymous review of Schumann's own Heine Liederkreid, opus 24, •which appeared in the V\e~nn&Allgemeinerm.u<4ikatL]cherAnzeiger in 18-40: As demanded by the contents of the words, a charming connecting thread also draws itself through the whole, including the chosen key-cycle which, in the analogous succession D major, B minor, B major, E minor, E major, A major, D minor and major gives evidence of the secret nexus of their elective affinity.90 In this case, the complete succession of keys is laid out for us and the use of terms such as "connecting thread" and "key-cycle" implies that their order is of some significance to the critic. But even in this review -we are given no indication that key succession plays a role in organizing the cycle. Does the reviewer simply feel that the close relationships between the keys of consecutive songs contribute in a general way to the cohesion of the cycle? Or is he making the more specific claim that the succession creates a coherent structure out of the songs? There is nothing in the text of the review to suggest that he intends the latter reading. I will cite one further passage to support my contention that the sequence of keys within a published collection of songs did not carry any structural implications for nineteenth-century musicians. It does not come from a published review but from a letter, of March 14, 1841, from Felix Mendelssohn to Ignaz Moscheles. Mendelssohn writes that he has received ten of Moscheles's songs and has selected six of them to "put together a book" for publication. He lists the songs he has chosen, in the order they will appear, and then apologizes about the resulting key sequence: The keys certainly ensue most madly: F major and B major, all in a muddle; but I have always found that no one thanks you for the most beautiful succession of keys. It is rather a certain change from slow to fast and from solemn to cheerful that is really required. So please excuse the key fricassee. Mendelssohn makes it clear that the contrast between successive songs is more important to the public, and presumably to the critics, than the sequence of keys. His apologetic tone implies that he himself "thought the sequence of keys to be an important consideration when arranging a group of songs for publication, " as Turchm observes.92 However, Mendelssohn is willing to have the
Schlegel'j Fragments ant) Schumann '
85
songs published in this order despite the fricassee, so he apparently does not consider key succession to be that important — it is a matter of preference and not of necessity. But it is equally significant that this set of songs was clearly not intended as a cycle, since the composer did not even select and arrange them himself. That Mendelssohn is concerned about the succession of keys nevertheless suggests that this consideration was relevant for any published song collection and not just for a cycle. Rather than demonstrate that key relationships provided a means for differentiating the cycle from other collections in the nineteenth century, his letter shows that they were not considered to be a relevant criterion. The Romantic Cycle As one reads nineteenth-century song reviews, it is often impossible to tell whether the work at hand is a song cycle or simply a published collection. Reviewers typically treat each song in a publication separately, and -where reference is made to the group of songs as a whole, it is often in terms that could just as easily apply to a collection as to a cycle, such as the themes of the poems or the general stylistic character of the music. This is the case for Schumann and his colleagues as much as it is for other critics. It is true that, with the exception of the An die. feme Geliebte reviewers, they are virtually the only early nineteenthcentury critics who describe the key relationships among the songs of a cycle. But the handful of reviews in which we find such descriptions are hardly representative, and in general, references to the keys of songs in the pages of the Neue Zeifochrift are rare, explicit statements about key relationships rarer still. For the most part, the Neue Zeitdcbrift critics are interested in the same issues as their contemporaries—the selection and arrangement of texts, the appropriateness and tastefulness of the musical settings, and, perhaps most important, the extent to •which the composer is able to convey the essence of the text in musical terms. The fact that nineteenth-century critics rarely describe song cycles as integrated •wholes does not necessarily mean that they never considered them in these terms, of course, but it does suggest that the question of cyclic unity -was of far less interest to them than it is to modern-day scholars and that the distinction between a cycle and a collection may not have been as clear-cut as we believe it to be. Friedrich Schlegel's dichotomy between two aesthetic principles — Romantic and Classical — provides a useful means for explaining the difference between the nineteenth-century conception of the song cycle that I have been outlining and the viewpoint that is more common today. On the one hand, in arguing that the cycle is differentiated from other collections of songs and piano pieces because it has an integrated and unified musical structure, which is based on a coherent key scheme and a network of motivic connections, most modern-day scholars describe a type of complete and perfectly closed work that Schlegel would label as Classical. To the nineteenth-century critic, on the other hand, the cycle is typical of what Schlegel describes as the Romantic style. Because it orig-
86
The Genre of the Cycle
mates as an expansion of the song or the character piece, in which personal expression is emphasized and formal conventions are lacking, the cycle is not a pure genre but is in a state of flux. It is not clearly distinguished from the more typical collections of songs and piano pieces that were published in the early nineteenth century but emerges gradually from them and implies an endlessly growing variety of formal schemes within which such pieces can be organized. The fact that such formal schemes tend to be open-ended and mutable also characterizes the cycle as a Romantic work, and this again stems from the kinds of short pieces of •which it is composed. They are like fragments — discrete movements that nevertheless tend to have incomplete structures and to hint at further potential meanings—and they form a whole that is discontinuous and heterogeneous. To illustrate the wide gulf between these two viewpoints, I shall return once more to Schumann's Dichterlie.be and consider the first published commentary about it, an anonymous review that appeared in the Neue ZeitMbrift in 1845, shortly after Schumann sold the journal to Franz Brendel. In the previous chapter I explained how this work has come to be the single most celebrated example of the modern, organicist conception of the cycle. The 1845 review reveals that its initial reception was quite different and, in particular, that the question of whether or not "it constitutes an integrated musical whole" was not even considered. I quote the review in its entirety: These songs have a completely characteristic aura; the accompaniment indulges in graceful figures and surprising turns; the melodies are intimate and fresh. As we can refer the reader to the essay in these pages: "Robert Schumann with Consideration of Felix Mendelssohn-Bartholdy," in which Schumann's general characteristics are more closely examined, we skip to the individual songs and emphasize those that please us the most: Nr. 1 "Im wunderschonen Monat Mai." The accompaniment appears to introduce Ft minor, the vocal line turns surprisingly to A major and strangely closes on a D major chord, while the accompaniment ends -with Ct, as the dominant of Ft minor. — Nr. 5. "Ich will meine Seele tauschen," especially beautiful; a peculiar magic radiates over the whole. The opening is extraordinary— (the song is in B minor) the vocal line begins alone on an upbeat with b—the accompaniment nestles under it with the two arpeggiated seventh-chords cf-e-g-b, then ft-at-cS-e.—Nr. 7 "Ich grolle mcht und wenn das Herz auch bricht," especially characteristic. — Nr. 8 "Und wiifiten's die Blumen" — the accompaniment, like murmuring evening wind; the vocal melody very expressive. Of all the settings of these words that we know, this appears to us to be the best. —Nr. 15 "Aus alten Marchen winkt es hervor" moves in the favorite realm of the composer, in the realm of the ghostly and fantastic, in the land of yearning, in which he is especially happy. — Nr. 16 "Die alten bosen Lieder, die laftt uns jetzt begraben" is composed with much humor. —
Schtegel'j Fragment) and Schumann'*! Cycles
87
It is just too bad that many songs of this cycle hurry over us like fleeting butterflies and leave us no time to delight in the beautiful, iridescent colors that are so suitably con anwre?^ This review is remarkable on two counts. First of all, the anonymous critic who wrote it demonstrates a sophistication and sensitivity about technical analytic issues that is rarely found in nineteenth-century music journalism. Second, given his level of theoretical understanding and his willingness to express it in this review, it is telling that he does not mention those elements of Dichterliebe that have made it the subject of so much interest today. By choosing to present his review in a typical nineteenth-century format — a general comment praising the style of the songs, short descriptions of a few of them, and another general comment that is more critical — the author does not even allow himself the opportunity for a discussion of the sequence of keys or of any structural relationships that may exist among the songs. He ignores the cyclic return at the end of Song 16, which is certainly one of the most striking aspects of the -work and -which sets it apart from almost all of its contemporaries. And he describes the inconclusive ending of Song 1 but does not remark on its implied connection to the opening of Song 2. There are, of course, twentieth-century commentaries on individual songs from Dickterliebe in which there is no attempt to relate them to the rest of the cycle. One such instance, -which I discussed in the last chapter, is Schenker's pair of analyses, in Derfreie Satz, of the two opening songs. Schenker treats each of them as an independent piece of music and, like our nineteenth-century critic, makes no mention of their relationship to each other, even placing the two analyses in different parts of his book. More recently, Rosen began a chapter on the Romantic fragment with a lengthy discussion of the first song alone. But both of these cases are explicitly presented as commentaries on a song and not on the cycle. Rosen eventually does turn to the relationship between the end of Song 1 and the beginning of Song 2, and when he comes back to Dickterlie.be in a later chapter that is concerned with the song cycle he presents many of the familiar themes from the recent literature, such as key relationships, narrative, and thematic return. He may be willing to separate out a single song from the cycle and discuss it on its own, but he considers this to be an altogether different matter from discussing the cycle. For a critic writing in 1845, commentaries on the individual songs, in and of themselves, constitute a discussion of the cycle, even when he is faced -with a -work as exceptional as Dichterliebe. In the one case there is an underlying assumption that a cycle is a coherently unified structure; in the other there is a contrary assumption, just as strongly ingrained, that a cycle is primarily a series of independent songs. My point in contrasting these two conceptions of Dichterliebe is certainly not to suggest that -we simply replace our present-day assumptions with those that were prevalent in Schumann's day, a task that would be as impossible as it is undesirable. I argued at the outset of the present chapter that -we should not privilege Romantic views of Romantic music above all others, and this is as much the
88
The Genre of the Cycle
case when they diverge from our own as it is when they appear to agree. But we need to recognize that a mid-mneteenth-century critic is operating under a different set of assumptions than a late twentieth-century scholar, and so we should be careful in interpreting contemporary comments as evidence for our modernday viewpoint. In the case of the Romantic cycle, as in many cases, an awareness of such differences can help us to question our own assumptions and to come to a greater understanding of the genre and of the particular works that make it up. One implication of the foregoing discussion is that to the extent that Schumann •was trying to create a unified structure in Dicbteriie.be he was composing a work that was atypical. It is thus misguided to generalize about the nineteenth-century song cycle on the basis of Dichterliebe and to treat other Schumann cycles, such as the Eichendorff Liederkreu and the Kerner Liederreihe, as problematic or marginal examples of the genre. A second implication is that formal organicism is not the only possible model for explaining the cycle and it is not necessarily the most appropriate model. Given the difficulties we have faced in making even Dichterliebe fit this model, perhaps it is time to search for an alternative. My attempt to reconstruct the intellectual context within -which Schumann composed his cycles is intended to help stimulate such a search.
PartII Schumann's Eichendorff Songs
This page intentionally left blank
POEWAND SONG
T T
he nineteenth-century song cycle is not primarily "an integrated
musical whole," to use Arthur Komar's formulation, but a collection of parts. If we want to understand the ways in which these parts interrelate and cohere, then it is more fruitful to begin from the individual part and work toward the whole than it is to go the other way around. What we need, in order to do this, is some model that can replace, or at least radically modify, the metaphor of formal organicism. And this is where the ideas of the Schlegels, as refashioned in the writings of Schumann and his colleagues, can be of great usefulness. For they suggest that the relationship between the cycle and the model of organicism is not one of conformance but one of opposition. The part calls into question the possibility of the whole because of the way in which it subverts the conventions and expectations of musical form. But at the same time, this very questioning taps into our innate desire for closure and thus creates the need for a larger whole within which the individual piece can be placed. I believe that it is this creative tension, perhaps more than anything else, that defines the nineteenthcentury cycle. And, in true Romantic fashion, it is a tension that can never be resolved but is simply replicated on ever more encompassing levels. Analysts who assume that the cycle is defined by the presence of a comprehensive structure that integrates the individual parts into a whole will inevitably neglect the tension between openness and closure that I am describing. And yet the peculiar nature of the part is hinted at in many of the recent writings on Schumann's cycles, and I suspect that it is the very difficulty of making these works conform to the orgamcist model that makes them so attractive to theorists in the first place. This is clearly the impetus behind David Neumeyer's study of the opening pair from Dichteriiebe, and it is also why Patrick McCreless decided to write about the Eichendorff Liederkreui. Even Komar begins his essay on Dichteriiebe by posing the question of whether this cycle should be regarded more as a whole or as a series of discrete parts. And while his answer is never in doubt, he does suggest at one point in his book that the incomplete nature of the part is one factor that makes Dichteriiebe the most impressive of Schumann's cycles. He writes that the songs "are so brief and quiet — so unsubstantial—that they emerge quite clearly as dependent parts of an inclusive whole." In particular, 91
92
Schumann '<> Eichendorff Songd
Komar points to "the absence of the prevailing tonic triad at the beginning" of several of the songs and "the harmonically inconclusive endings" of some of the others.1 In the 1845 review o^DLchteriizbe that I quoted at the end of the last chapter, there are only two songs that are discussed in any detail: "Im wunderschonen Monat Mai" and "Ich -will meine Seele tauschen." These are both songs that have been analyzed in the twentieth century as well, and we have singled them out for the same reasons that this nineteenth-century critic did.2 What interests him, and us, is the fact that Schumann frustrates the listener's expectations and purposefully undermines the strength of the tonic. In "Ich will meine Seele tauschen," the tension and uncertainty are momentary—the b1 that is heard as an unharmonized upbeat becomes the dissonant seventh of the opening harmony as the piano enters, but it is subsequently revealed as the tonic of the song. In "Im wunderschonen Monat Mai," however, the tonal ambiguity is never resolved and, as our reviewer points out, each thwarted expectation leads to another that is frustrated in turn. By focusing on these aspects of Schumann's songs -we may be able to use nineteenth-century ideas to supplement our twentieth-century analytic techniques and find an approach to the cycle that is more consistent with Schumann's own conception. I believe that many of Schumann's songs and piano pieces function as "dependent parts" and that this is why it was natural for him to publish them as cycles, but I do not think that the cycle acts as "an inclusive whole," which provides the closure they may lack on their own. As Schumann arranged his twelve Eichendorfl settings into the work that he ultimately published as the opus 39 LuderkreL), he was guided by his aesthetic sensibility and his instinct for formal organization. I consider this work to be one of the great artistic creations of the nineteenth century, and I believe that the order in which Schumann arranged the songs contributes in important -ways to its greatness. But even if the order of the Eichendorff songs is aesthetically pleasing, that does not mean that it is immutable. I certainly do not intend by this statement to question Schumann's arrangement or, worse yet, to propose that singers and pianists should feel free to rearrange the songs of the cycle as they see fit. I simply mean that Schumann was not bound by aesthetic necessity, musical logic, or generic convention to place the songs in the order that he did. This can be demonstrated by historical and analytic evidence and also by the evidence of the manuscript source, which suggests that Schumann only began to consider the arrangement of the songs after he had finished composing them. The work that he ended up with is a whole, but it is a Romantic whole, a whole that is open-ended and fragmentary and that is characterized more by potentiality and implication than it is by a sense of closure. In a process that was apparently typical of Schumann's cyclic composition, he wrote the drafts of the Eichendorff songs in rapid succession, at one point writing out four songs in three days.3 The spontaneity of Schumann's compositional process is reflected in the almost improvisational nature of the songs, many of which feel unfinished and open-ended. And the sketchlike quality of the
Poem, and Song
93
individual song is enhanced and made more comprehensible by the exploration of similar compositional ideas in the others that surround it in the cycle. I believe that many of these musical ideas are directly inspired by Schumann's engagement with Eichendorffs poems. There is an inherent tension between Eichendorff's folklike style, which is characterized by a prevailingly regular prosody and by simple stanzaic forms, and his typically Romantic use of imagery and theme, through which he attempts to express the ineffable. In his settings of these poems, Schumann translates this tension into musical terms. He transforms Eichendorffs stanzaic structures into simple ternary and binary forms and tends to retain Eichendorffs lineation, typically setting each line of text with a two-measure phrase grouping. Only rarely does Schumann create irregularities by expanding phrases or repeating words. His vocal melodies thus exhibit a regularity in terms of both phrase rhythm and large-scale form that mirrors the folklike simplicity of Eichendorffs poetic style.4 But Eichendorff often works against the predictable regularity of his prosody by avoiding concreteness in his poems, relying more on implied suggestion than on explicit statement. Schumann, likewise, uses the simple formal schemas of his songs as points of departure, from which he creates musical structures that defy conventional expectations, as they question the compositional necessity of closure and continuity. Many of the Eichendorff songs have weak openings, in which Schumann deliberately obscures the initial tonic harmony, and open endings, in -which the closure of the final cadence is compromised. Recapitulations are disguised, the boundaries between sections obscured, and in some cases the identity of the tonic key is not clear for much of the song. It is through the exploration of such compositional ideas that Schumann finds the most powerful means of expressing the Romantic poetry that he set in his early song cycles, and it is for this reason that much of my discussion of the Eichendorff songs will be concerned -with musical structure.
Schumann's Setting of "In der Fremde" "In der France" or "Der fro he Wanderjmann"? In choosing to begin my discussion with "In der Fremde," I immediately come up against the question of -whether or not there is a definitive version of the Eichendorff Liederkreu and, if so, which version it is. Although "In der Fremde" was one of the first of the Eichendorff songs that Schumann composed in May 1840 and -was originally intended as the opening song of the cycle, he did not include it when he published the songs in 1842. In this first edition of his opus 39, Schumann substituted another Eichendorff song, "Der frohe Wandersmann," and then, when he republished the cycle in 1850, he changed his mind again and went back to "In der Fremde." It is the second edition that has come down to us today as Schumann's Eichendorff Liederkreu, and "Der frohe Wandersmann" has not only lost its place within the cycle but has also virtually disappeared from the lied repertory. Schumann scholars who have written about the cycle have gen-
94
Schumann j Eickendorff
Stmgj
erally concurred with the historical preference of performers and audiences, either ignoring the song or dismissing it as clearly inferior to its replacement. But Jon W. Finson has recently come to the defense of "Der frohe Wandersmann" and has suggested that "performers might occasionally program the original version" of the Eichendorff Lifderkreu or even take advantage of the technology of the compact disc to allow listeners to choose between the two songs.5 I think Finson's suggestion is a good one, and in fact, his implication that the contents of the opus need not be definitively fixed fits quite nicely with the nineteenthcentury view of the cycle that I sketched in the previous chapter. In the process of defending "Der frohe ^Wandersmann," however, Finson inevitably casts doubt on Schumann's decision to replace it with "In der Fremde" and suggests that the cycle has not only been weakened by the change but has also become less Romantic and more conventional. He acknowledges the incongruity of beginning a cycle of such mystery and subtlety with the simple and folklike "Der frohe Wandersmann," which he describes as a "rhythmically square and forthright march" supporting a "cheerful melody in an uncomplicated way." But he goes on to argue that this very incongruity enhances the Romantic irony that Schumann creates through the juxtaposition of various texts •within the cycle. Finson also claims that in beginning opus 39 with "Der frohe ^Vandersmann" Schumann identifies the Kichendorff Liederkreu as a narrative within the nineteenth-century tradition of the Wanderlieder cycle and argues that in its latter version, with "In der Fremde" as the opening song, this narrative has eluded many writers.6 But, as Barbara Turchin observes, in replacing "Der frohe Wandersmann" with "In der Fremde" Schumann did not abandon the theme of the wanderer but simply replaced "the cheerful journeyman" of the former song with the "lonely traveler" of the latter.7 And as the narrator of "In der Fremde" describes her feeling of homelessness and expresses her yearning to be united with nature, she introduces a number of themes that are central to the development of the cycle. I will return to the question of what kind of narrative Schumann creates in the Eichendorff Liederkreu in chapter 8. For the moment, I will just point out that my decision to begin my discussion of the Eichendorff songs with "In der Fremde" and leave out "Der frohe ^Wandersmann" does not imply that I consider the second edition of the cycle to be the definitive version or that I have any objections to the performance of the first. It is simply that because Schumann's setting is so straightforward, "Der frohe Wandersmann" is not the most apt example for my purposes. The text does not display the tension between folklike simplicity and Romantic elusiveness that characterizes most of the Eichendorff poems that Clara Wieck chose for Schumann, and so it is not surprising that the complex relationship between vocal line and accompaniment, the tonal instability, and the formal peculiarities that are so pervasive in the cycle are absent from his setting of this particular poem. "In der Fremde," however, exhibits a number of unusual characteristics that are typical of the Eichendorff songs and that help to mark them as members of the same cycle. And "In der Fremde" is uniquely suited to serve as a starting
Poem and Song
95
point for my discussion, because of two circumstances that provide evidence about the process by •which Schumann turned the poem into a song. First, "In der Fremde" is the only song in the cycle in •which Schumann significantly alters the form of Eichendorff's poem, repeating phrases of text, rearranging line breaks, and essentially destroying the meter and the rhyme scheme in the poem's second half. By considering how his changes affect our understanding of the poem and also how they relate to the structure of the song, we can ultimately gain insight into how Schumann uses musical form to express the Romantic themes in Eichendorff's poetry. As he alters this particular poem, Schumann leaves behind traces of his creative process, which can tell us something about his general approach to Eichendorff's poetry. Second, Schumann heavily revised the piano draft of "In der Fremde," rewriting the accompaniment of one phrase and recomposing another phrase in its entirety. Although he revised the drafts of several other Eichendorff songs, "In der Fremde" is the only one in which the revisions are so extensive that they create a completely new formal and tonal structure. By reconstructing the path from Schumann's original conception of the song to his final version, -we find further evidence of his evolving interpretation of the text and his search for an appropriate musical setting for Eichendorff's poem. From Poem to Song "In der Fremde," which is printed as a single octave in Eichendorff's edition of his collected poems, is the only text in the opus 39 Liederkreu that does not have a stanzaic form, and it is perhaps for this reason that Schumann was inspired to dramatically alter its poetic structure. Despite its appearance on the page, however, several factors imply a division into two quatrains, and the resulting tension between a continuous form and a division into strophes is one that Schumann exploits and intensifies as he turns the poem into a song. Eichendorff's poem reads as follows: Aus der Heimath hinter den Blitzen roth Da kommen die Wolken her, Aber Vater und Mutter sind lange todt, Es kennt mich dort Keiner mehr. Wie bald, wie bald kommt die stille Zeit, Da ruhe ich auch, and fiber mir Rauschet die schone Waldeinsamkeit Und Keiner mehr kennt mich auch hier.
From home beyond the red lightning The clouds approach, But father and mother are long dead, Nobody knows me there anymore. How soon, how soon will come the peaceful time, When I also rest, and over me Rustles the beautiful solitude of the woods And nobody knows me here anymore, either.8
One factor that implies a new strophe after the fourth line is the fact that this line concludes -with the only period before the end of the poem and, since the
96
Schumann '
fourth line is echoed by the last line, there is the suggestion of a. refrain form. A second factor is that the meter breaks down after the fourth line. The first four lines have two levels of metric regularity: lines of four feet alternate with lines of three feet; and "within each line, anapestic feet alternate with iambic feet, with the sole exception of an anapestic substitution in line 3. A second substitution at the beginning of line 5, however, creates two consecutive iambs, and with the next two lines there is no trace left of the alternating feet. These three lines each have four feet, so the higher level of regularity is destroyed as well. This breakdown in the meter creates a continuousness that is further enhanced by the enjarnbment between lines 6 and 7, -which are bound together not only grammatically but also metrically. The succession of a pyrrhic foot at the end of the first of these lines and a trochee at the beginning of the next creates the feeling that the final accented syllable of line 6 has been delayed until the beginning of line 7. The final refrain line has three feet, restoring the alternation of four-foot and threefoot lines, and although it ends with an anapestic substitution, it initially restores the underlying iambic/anapestic alternation as well, thus framing the preceding lines and providing the poem with formal closure. The poetic structure of "In der Fremde" corresponds to its thematic content. In the first, metrically regular group of four lines, the heroine of the poem is reflecting on the loneliness of her current situation.9 She is a stranger in a strange land, her parents are dead, and her home is now an unfamiliar and forbidding place. Her vulnerability is made all the more clear to her by the approach of a threatening storm. The consecutive iambs at the beginning of the fifth line signal her impatience as she turns longingly toward the future. Her thoughts transform from a sober assessment of an unhappy reality to a dreamy anticipation of the peacefulness of her own death, which will lead to a union with nature. The formal continuousness of lines 5—7 reflects the dreamlike quality of her thoughts. But what about the last line? I have thus far described this line as a refrain, and in this sense it is simply pointing out the parallelism between the situations described in each half of the poem: in both cases, nobody knows her. But if we consider just the context of the poem's second half, line 8 feels like an uncomfortable intrusion. Lines 5—7 strongly evoke the peaceful solitude of death, an image that is completed by the word "Waldeinsamkeit" at the end of line 7. The addition of another line breaks the spell. And the line itself recalls the heroine's loneliness and reminds her that she is still in the present, where this peaceful solitude has not yet come. So one dichotomy, or ambiguity, in Eichendorff's poem concerns the way in which it closes and the possible functions of the last line. The narrator, in foreseeing her own death, invokes a kind of ultimate closure in the penultimate line. In one sense, the last line of the poem is a refrain and thus enhances this closure. But there is another sense in which this line is an interruption that negates the closure of "Waldeinsamkeit" and returns the narrator to reality. The desired closure of death cannot be achieved; it is the anxious loneliness of the present that is closing her in. As he set "In der Fremde" as a song, Schumann exaggerated certain elements of its formal structure to such an extent that he practically created a new
Poem and Song
97
poem. Table 4.1 compares the form of the original poem, as published by Eichendorff, with the new form that Schumann creates in his musical setting. In the first four lines, he preserves the strong metrical regularity by setting the first couplet as a four-measure period and then essentially repeating the same music for the second couplet. He also sets up a two-bar hypermeter and a proportion of one hypermeasure of music for each line of poetry. Beginning with the fifth line and continuing through line 7, Schumann slows down the rhythm and repeats phrases of text, so that three lines of poetry are set by twelve measures of music and the proportion of music to poetry is doubled. He still retains the twobar hypermeter but creates extra hypermeasures by repeating half-lines of the poem, and at the same time he ignores the line breaks, so that three equal lines of poetry become six lines of varying lengths, arranged in two groups of three. In terms of the poem's formal structure, we can say that Schumann is exaggerating the irregularity and continuousness of these lines.10 Schumann also makes explicit the implied break between lines 5—7 and line 8. He destroys the line break after "mir" by burying it in the middle of a phrase and thus obscures the rhyme between the end of line 6 and the end of line 8. The latter line stands alone from the preceding three and at the same time strongly recalls the opening pair of couplets, since "hier" creates a half-rhyme with "her" and "mehr." Schumann further strengthens this connection by altering the line so that it ends -with an internal half-rhyme. And just as this final line restores the
TABLE 4.1 Schumann's Transformation of Eichendorff's "In der Fremde" Eichendorff
Schumann (each line of text = a 2-bar hypermeasure) A (1-measure intro. + 8 measures)
Aus der Heimath hinter den Blitzen roth Da kommen die Wolken her, Aber Vater und Mutter sind lange todt, Ks kennt mich dort Keiner mehr.
Aus der Heimath hinter den Blitzen roth Da kommen die Wolken her. Aber Vater und Mutter sind lange todt, Ks kennt mich dort Keiner mehr. B (12 measures)
Wie bald, wie bald kommt die stille Zeit, Da ruhe ich auch, und iiber mir Rauschet die schone Walkdemsamkeit
Wie bald, ach, wie bald kommt die stille Zeit, Da ruhe ich auch, Da ruhe ich auch Und iiber mir rauscht die schone Waldeinsamkeit, Die schone Waldeinsamkeit, A' (4 measures + 3-measure postlude)
Und Keiner mehr kennt mich auch hier.
Und Keiner kennt mich mehr hier, Und Keiner kennt mich mehr hier.
98
Schumann '
initial metrical regularity in the poem, in the song it restores the original proportion of music to text—-a. two-bar hypermeasure for each line of poetry—and the more regular pairing of hypermeasures with which the song began. By repeating the line and adding a three-measure postlude, Schumann creates a seven-measure section of music that essentially balances the nine-measure A section, which sets the opening four lines. In Schumann's setting, then, the final line, which echoes the last line of the first quatrain, stands in for a return of the entire quatrain. We can sum all of this up by saying that Schumann has taken a double quatrain poetic form and transformed it into a ternary ABA' musical form. But rather than repeating the first quatrain for the A', as we might expect (and as Schumann does in "Intermezzo," the following song in the cycle), he breaks up the poem into three irregular sections: four lines for the A, three lines for the B, and one line for the A'. One result of this transformation is that Schumann does not treat the last line as a refrain. He concludes his setting of the poem not by strengthening the sense of closure evoked by the penultimate line but by returning the narrator to the reality of the first four lines. And Schumann's alteration of the line seems to corroborate this. By removing the word "auch," he diminishes the comparative sense of the line and thus deemphasizes the paralellism between the two halves of the poem. However, there are -ways in -which Schumann's song responds to and even intensifies the ambiguity of Eichendorff's poem. I have described Schumann's setting of the last line as the A' section of a ternary form, because of the thematic content and character of the vocal line, the affect of the music, and the proportions of the song. In terms of the tonal and contrapuntal structure, however, this section functions as a coda, since the final cadence comes at the end of the seventh line of the text. And while the placement of this cadence creates a kind of formal imbalance in the song and demands a coda to strengthen our sense of closure, the coda that Schumann has composed is rather ambivalent and continues to develop and open the song up, even as it provides seven more measures of tonic harmony. Schumann thus beautifully represents the way in which Eichendorff's protagonist longs for the peaceful solitude of the future as the loneliness and alienation of the present close in on her. He also creates a perfect opening for his cycle: a song that is at once complete in itself and yet rich in implication.
An Analysis of "In der Fremde" Narrative and Lyricum
Schumann sets the first half of "In der Fremde" as if it -were the beginning of a narrative poem, rather than a lyric. The accompaniment consists of a simple arpeggiated figure, which evokes a plucked string instrument such as a bardic harp, and the harmonic language is largely restricted to alternating tonic and dominant chords. The unusually austere vocal line is characterized by frequent repeated pitches and a recurring rhythmic pattern, which create the effect of a
Poem and Song
99
rhythmicized declamation of the poetic text rather than a sung melody. The range is essentially limited to the minor third between fit1 and a1, with a brief motion to the upper neighboring tone b1 at the end of each phrase. And this upper neighbor is introduced by a pair of grace notes that forces a slight break in the melodic line and forms an embellishing formula that marks the phrase endings, another device that alludes to the tradition of narrative recitation. One final factor that contributes to the narrative style of Schumann's setting, which I have already noted, is the regularity of the phrase rhythm. Schumann uses virtually the same melody for the first and second couplets and concludes each of them -with a perfect authentic cadence. He thus creates a paratactic phrase structure that is typical of simple narrative genres such as the ballad. The one expressive change that Schumann makes in the second phrase is to vary the accompaniment. Rather than moving to V in the second measure of the phrase, as he does at the opening of the song, Schumann delays its arrival by introducing a descending third span in the tenor that prolongs the tonic harmony for three measures, turning it into a I6 chord in measure 8. One result of this variation is that it creates a series of passing diminished-seventh chords in measures 6 and 7, adding some variety to the otherwise constricted harmonic palette and poignantly highlighting the narrator's reference to her parents' deaths. A second result is that Schumann creates conflicting implications in the vocal melody, which inevitably set up tensions that are left unresolved. The counterpoint implies a voice exchange between the outer voices, -which would lead the vocal melody to descend to fjj1 at the beginning of measure 8, as shown in level a of Example 4.1. But there is an even stronger implication to return to a1 at this point, as Schumann does, both because it is the third measure of the phrase and thus too early to bring melodic closure and also because of the parallelism with the previous phrase. By returning to a1, however, Schumann creates some ambiguity as to the function of the harmony that supports this pitch, as I have shown in level b of Example 4.1. If there were a voice exchange, we would clearly hear the I6 chord as a prolongation of the tonic, but as it stands, this harmony could just as easily be heard as a substitute for a 164, which is what the upper voices of the accompaniment suggest. It is, once again, the parallelism •with the first phrase that militates against the latter interpretation, since on the downbeat of measure 4, the analogous spot in the first phrase, there is a root position tonic harmony. The embellishing neighboring tone formula in the second half of the measure is another element that discourages us from hearing a1 as an 11 appoggiatura, since it obscures the resolution to gl . It is only in the B section of the song that Schumann introduces a greater degree of lyricism and continuousness as he responds to the breakdown in the poetic meter after the fourth line of Eichendorff's poem. Each of the two phrases in this section is expanded from four measures to six, and together they create a period that modulates to B minor. As the phrase structure becomes more continuous, the vocal line becomes more lyrical. Because Schumann breaks up the sixth and seventh lines of the poem, each two-measure grouping contains fewer syllables of text, and the rhythm slows down from quarter- and eighth-note mo-
100
Schumann '
EXAMPLE 4.1. Analysed of measures 6—9, showing hypothetical implications.
tion to halves and quarters. This tends to make the vocal melody more legato, an effect that Schumann explicitly indicates in the right hand of the piano in measures 10—15. The melody also sheds its restricted scope and exclusively stepwise motion, first outlining its range thus far, by leaping between ft1 and b1, as the third phrase begins in measure 10, and then unfolding outward, reaching e , the melodic high point, in measure 13. The right hand of the piano, which suddenly becomes quite active after its virtual absence in the first nine measures of the song, complements the increasing expansiveness of the vocal melody. The pianist continues the singer's upward leap in measure 10 from ftt1 to b1 with a second leap from b1 to f(t2, introducing a new motive that plays a crucial role in altering the tone and affect of the song. This motive contrasts sharply with the motivic shape that underlies the vocal melody in each of the first two phrases. Its larger melodic contour and its -weak metrical placement — beginning on the second beat of the measure and avoiding downbeats altogether — contribute to the expansive feeling of the third phrase and evoke the protagonist's yearning to converge with the natural world that surrounds her. As Schumann opens up the registral space of the song, he also stretches out the phrase rhythm, turning an underlying four-measure phrase into six measures through an internal phrase expansion. \Ve can see this expansion most clearly by
Poem and Song
101
considering the half-line "da ruhe ich auch," which is first set by the two-measure grouping that begins -with the upbeat to measure 12 and is then repeated, beginning with the upbeat to measure 14, as shown in level c of Example 4.2. The initial measures of each of these groupings, setting the words "ruhe ich," are almost identical, with the exception of some changes in the accompaniment, most significantly the addition of G^l in measure 14. By omitting the first of these twomeasure groupings and connecting the end of measure 11 to the beginning of measure 14, I have reconstructed the hypothetical four-measure phrase shown in level a of Example 4.2, which is the source of this third phrase. My analysis of this hypothetical phrase, in level b, shows that it remains coherent in terms of harmonic syntax and voice leading. It also preserves the basic harmonic function of Schumann's phrase — the transformation of an Ft minor triad into a V7 chord that tonicizes B minor.12 In terms of the voice-leading structure, the internal expansion in measures 12—13 is motivated by the opening up of the higher register in measure 10. The leap up to e creates a connection between ft in measure 10 and d in measure 16, as can be seen in my analysis of the actual six-measure phrase in level d. One way to describe the transformation that takes place at this point in the song, then, is to say that the expansion of registral space as the phrase begins leads to an expansion m rhythmic time as it continues. Schumann also uses the extra two-measure grouping to insert an implied progression to A major, before the far more definitive motion to B minor at the end of the phrase. There is an underlying root progression that moves by descending fifth, which I have shown on the added staff of level d. Schumann omits the bass voice, however, and uses a pattern of parallel sixths between the tenor and soprano to control the motion to measure 13. The descent from e2 in the vocal melody then leads to a repetition of the contrapuntal pattern in measures 14—15, but with a crucial chromatic alteration that steers the phrase to its new harmonic goal. The pattern thus enables Schumann to change the tonal direction m the last two measures of the phrase without disrupting its continuity or coherence. One of the most striking aspects of the third phrase is the increasing emphasis on b , which had been introduced m the A section as an upper neighbor. In measures 10—12, it becomes the central point of the vocal melody, and it is through this pitch, rather than the members of the tonic triad, that Schumann effects the registral transfer from ft1 to ft2. But it is only in the second half of the phrase, as Schumann transforms the tonic chord into the dominant of B minor, that the function of b1 changes from a dissonance to a melodic goal, a goal that is never really attained. The fourth phrase does begin by prolonging a B minor triad, in measure 16, but by the time the melody finally descends to b1 in measure 19 it is not a resolution but a part of a continuous passing motion to ft1. We can hear in the relationship between the two phrases of the B section a reflection of the lines of text that they set. In the third phrase, as the protagonist expresses her impatient desire for the peace of death, the harmony turns toward the dominant of B minor. This desire leads to her description of the image of "Waldemsamkeit" as the dominant resolves at the beginning of the fourth phrase. But like
a. Hypothetical four-measure basic phrase.
b. Voice-leading analysis of basic phrase.
EXAMPLE 4.2. Measured 10-15.
102
c.
d.
103
104
Schumann'*! Eichendorff Song
the resolution to B minor, the fulfillment that she imagines proves to be elusive, and the return to Ft minor, as the words "die scheme Waldeinsamkeit" are repeated in measures 20—21, introduces a tone of resignation and makes explicit the protagonist's recognition that she still cannot rest. Schumann also brings out the relationship between line 7 of the poem and lines 3—4 by making the fourth phrase of the song, in measures 16—21, an expanded and transposed recomposition of the second phrase, in measures 6—9. Where in the earlier lines the narrator refers to the deaths of her parents and her consequent feeling of homelessness, in the seventh line she looks ahead to her own death, which will enable her soul to return to its home in nature. Schumann responds to the more peaceful and dreamlike state that is evoked in the seventh line by transposing the phrase to the subdominant key and making the vocal melody slower and more lyrical. At the same time, however, he heightens the tensions that are implicit in the earlier phrase, in large part by placing the same music -within a new context. As I have shown earlier, in Example 4.2d, the harmonic progression from i to V7/iv in the third phrase of "In der Fremde" is rhythmically displaced by two contrapuntal patterns. In Example 4.3, I have normalized the harmonic rhythm of measures 10—18 and reduced the progression to its underlying three-voice framework.13 This example reveals that the harmonic rhythm in the third phrase does not correspond to the two-bar grouping that is set up by the hypermeter but creates a conflicting three-bar grouping, -which I have represented by reducing the six measures of the phrase to two measures in 3 meter. What most clearly defines this grouping is the descending third in the upper register, which begins with the ft2 in measure 10, continues with the e2 in measure 13, and concludes with the d2 in measure 16, at the beginning of the next phrase. Example 4.3 also shows that the apparent motion to A major in measure 13 is simply part of the contrapuntal displacement of the progression to V7/iv. The climactic arrival of e2 does not feel like a tonic resolution in A major, in any case, due to the fact that it is approached by an upward leap and lacks support from a root position chord. As -we can see in levels a and b of my example, this pitch is actually the seventh of an Ft V7, •which unfolds down-ward to the leading tone. The unfolding from e2 to at1 in measures 13—15 implies a second unfolding from d2 to b1, which would provide their expected resolutions, and the harmonic rhythm that has been set up implies that this unfolding will occur in measures 16—18. For this reason, our desire to hear the vocal melody descend is much stronger in measure 18 than it is at the analogous point in measure 8. However, the elimination of the neighboring tone formula in the second half of the measure turns the melodic line into an echo of the appoggiatura figure in measure 17 and thus encourages us more strongly to hear the harmony on the downbeat of measure 18 as a substitute for an embellishing ,. In this way, Schumann intensifies the ambiguity and tension that are implied in the second phrase. The succession of appoggiatura figures in measures 17 and 18 also affects the way we hear the following measure, in -which b1, supported by a root position B minor chord and prepared by its dominant, is nevertheless presented
Poem ant) Song
105
EXAMPLE 4.3. Rhythmic normalization of meadur&i 10-18.
as if it were yet another appoggiatura, seemingly resolving on the second half of the measure to the seventh of a chromatic diminished-seventh chord. There is no sense of cadential resolution in measure 19 but a continuous motion into the following two-measure grouping, -which concludes the phrase with a perfect authentic cadence, not in B minor but in the tonic key of Ft minor. As Schumann recomposes the music of "In der Fremde's" second phrase in the fourth phrase, he expresses the complex psychological implications of Eichendorff's text -with great sensitivity. When the protagonist refers to her parents' deaths, Schumann heightens the melancholy mood of the song and introduces a subtle tension. As this tension pulls ever so slightly against the predictable regularity that is set up at the beginning of both poem and song, it already hints at the more expansive musical gestures that follow. It is as if the
106
Schumann '<> Eicheiidorff Song
thought of death and the memory of her parents (and by extension of her home as well) are enough to give the heroine a glimpse of the freedom and the spiritual homecoming that she anticipates from her own death. Then, as her thoughts turn more explicitly toward that future time, the regularity breaks down completely, creating the momentary feeling that she really is free. But at the same time, the tension intensifies, the gap between future dream and current reality is too great, and the only way that it can be resolved is through a return to the stifling world of the opening. Yet this return, which is symbolized by the Fit minor cadence in measure 21, does not provide a sufficient resolution, either. Open Endings
Considered from the perspective of tonal structure, the harmonic and contrapuntal closure that is provided by the cadence in measure 21 of "In der Fremde" is insufficient because it comes too early. There is an imbalance between the enormous emphasis that Schumann places on the subdommant in the B section of the song and the abruptness -with which he returns to the tonic, and so the final cadence cannot entirely resolve the tension that has been created by the thwarted move to B minor. The cadence is also too early in terms of the formal proportions of "In der Fremde," since it comes only three-quarters of the way into the song, before the singer has gotten to the last line of the text. And because Schumann already reaches the final cadence at the end of the B section, he undercuts the strength of the return that follows, -which is more an afterthought than a moment of resolution. One way to describe the unusual formal structure of "In der Fremde" is to say that the recapitulation has been elided into the coda. In one sense, Schumann's setting of the last line is a synecdochic return of the A section of the song. Not only does the poetic refrain stand in for the entire opening quatrain, but the cadential neighboring tone figure in the vocal melody stands in for the entire melodic phrase as well. At the same time, however, the last seven measures of the song are completely static in terms of both melody and harmony, and in this sense they do not really function as a recapitulation but simply reinforce the final cadence. The ambiguous function of the closing section of Schumann's song reflects the ambiguous way in which the final line concludes Eichendorff's poem — on the one hand framing the dreamlike image that is central to the second half, while on the other hand interrupting the dream and returning the heroine to her unhappy reality. But Schumann's ending also adds a further level of complexity, making explicit an element that Eichendorff merely hints at. For even if we read the last line of the poem as the heroine's return to her loneliness, it is clear that her anticipation of future peace has in itself transformed the present, turning it into a transitory stage in a spiritual and psychological quest, rather than a terminal state of despair and hopelessness. Schumann expresses this transformation through a technique that I refer to as an open ending, in which there is closure at the end of the song, but its conclusiveness is compromised in some -way. The open ending is one of the most important means by which Schumann creates songs and piano pieces that are like
Poem and Song
107
Schlegel's fragments: musical forms that are complete and viable, yet unfinished, that point beyond themselves and imply a connection to the larger world that is created by the ongoing cycle of which they are a part. This kind of ending is quite different from those songs that lack harmonic and melodic closure and are thus simply inconclusive, such as "Im wunderschonen Monat Mai." Although the latter song has received a great deal of attention from Schumann scholars, the inconclusive ending is a technique that he uses quite rarely, perhaps because it is a conventionalized gesture that provides him with a relatively restricted means of expression. As I have noted in chapter 2, "Im wunderschonen Monat Mai" ends -with a Phrygian cadence, a closing formula that is characteristic of Baroque slow movements. Schumann's use of this cadence in the early nineteenth century is an example of the sort of musical archaism that Romantic composers often employ to create an exotic effect. With the open ending, however, he experiments with the conventional expectations of musical form, creating a uniquely Romantic compositional device that has much greater expressive potential. Where the Phrygian cadence is an unambiguous musical gesture that always has the same underlying contrapuntal and harmonic form, the open ending takes on myriad guises and creates its effect through a variety of musical elements. For this reason, it is more wide-ranging and nuanced in its musical meaning, but it is also harder to classify, and this may be why it has been largely ignored by scholars. If there is one feature that most clearly defines the variety of Schumann's open endings as members of the same category, it is the disconcerting simultaneity of resolution and tension.15 In the case of "In der Fremde," Schumann reinforces the tonic resolution of the final cadence, in measure 21, with a coda in which 1 is prolonged in the vocal melody with tonic support, but he presents the tonic harmony as if it were a dominant and treats 1 as if it were 5. The final vocal phrase, shown in Example 4.4, is a repeated two-measure melodic gesture that embellishes fit1 with a heavily emphasized upper neighbor that is chromatically lowered to g^f, thus implying a resolution from 6 to 5 in B minor, rather than 2 to 1 in Fit minor. The harmonic progression that supports this neighboring motion does not begin with an Ft minor triad but with an Ft V7 chord that moves to an F| major triad by way of a dissonant neighboring harmony. The latter chord is actually a variant of a German sixth chord, which Schumann weakens in several -ways: he inverts the augmented sixth to a diminished third, between Et and Gt he places the E( in an inner voice and puts an Ft pedal in the bass; and he unfolds the chord in such a way that G'l and Et appear as part of a linear double neighboring motion, rather than a verticahzed interval. Schumann's unusual treatment of this neighboring chord enables him to subtly suggest the possibility that the Ft harmony that it resolves to has a dominant function, without creating the powerful tension toward a B minor resolution that a more conventional German sixth chord would. The effect that Schumann creates in the coda of "In der Fremde" is characteristic of the open ending. What is crucial is not so much the implication that Ft could resolve to B minor but rather the suggestion that it may not be the tonic.
108
Schumann '<> Eichendorff
Songj
EXAMPLE 4.4. Motivic analysis of measured 22—28.
Even as the Gh disappears in the course of the piano postlude, in measures 26— 28, mitigating the implication that Ft is a dominant, Schumann still does not go out of his way to reestablish that it is a tonic. The right-hand melody presents a typical closing gesture in the last two measures, leaping upward from 5 to 1, but the Ft pedal continues in the bass, so there is no cadential harmonic motion, and the harmony is still a major chord, which has functioned 'within the song as V, rather than I. Schumann contributes to the effect of the open ending in "In der Fremde" through a process of motivic transformation, shown in Example 4.4, that synoptically recapitulates the melodic path from the beginning of the song to the B section by gradually turning the underlying shape of the opening vocal phrase into a version of the piano motive from measure 10. The two elements that create the closest connection between the vocal gesture in measures 22—23 and the opening melody are the characteristic grace note figure and the melodic contour —a stepwise arch between ft1 and b1. As the singer repeats this line in measures 24— 25, the grace note figure is eliminated, and the stepwise motion to b'is replaced by a leap to d2. The latter phrase is then echoed in condensed form in the right hand of the piano in measure 25, and Schumann shifts its metrical placement so that, like the piano motive from measure 10, it begins on the second beat of the measure. In the last three measures Schumann first transposes the melodic figure down a perfect fourth, so that it outlines 5 and 1, and then strips away the embellishing eighth notes to reveal the underlying rhythmic shape of the original piano motive. A significant element in this motivic process is Schumann's treatment of the pitches d and b as he expands the vocal melody in measure 24. These are the two notes that become melodic goals in the B section, but within the context of the coda they are dissonances that are left unresolved. As the vocal melody ends, the singer reminds us that she still yearns for the closure of Wal3eLrwainkeit. In general, -we can understand Schumann's pervasive use of the open ending in the songs of his early cycles as a response to the poems that he set, since
Poem and Song
109
it enables him to translate into musical terms the themes of transcendence and ineffability that are so ubiquitous in the literature of the German Romantics. As the example of "In der Fremde" demonstrates, the open ending is a complex process in -which he employs particular musical techniques to express specific thematic elements of the text. It is a question not merely of how the song ends but of the -way in -which the ending responds to the implications that have been set up in the course of the song. In "In der Fremde" this is perhaps most obvious in the case of the motivic transformation that takes place in the coda, since the very notion of motive depends upon the recurrence and development of melodic material. But it is no less true of the unusual -way in which Schumann presents the tonic harmony in the coda. While the mysterious and elusive quality of this harmony is primarily the result of Schumann's manipulation of generic musical conventions, the powerful 'way in which it expresses the heroine's psychological transformation depends upon associations that Schumann sets up between the imagery of the text and the harmonic structure of the entire song.
The Two Versions of "In der Fremde" Schumann d Evolving Interpretation of the Poem
The piano draft of "In der Fremde," which is preserved in the Deutsche Staatsbibliothek in Berlin, reveals that while the ending of the song was part of Schumann's original conception, the thwarted modulation to B minor in the middle section -was not.16 Schumann thus composed the ending before he had even conceived of the context that would ultimately provide its expressive power. More generally, we can see from the draft that the beautiful and complex way in -which he expresses Eichendorff's text emerged only gradually. While some elements of Schumann's final version -were already in place when he first sketched the song, his most striking ideas were not, and as a -whole, the original "In der Fremde" is an entirely different and far less inspired song. As far as -we can tell from the manuscript material for Schumann's songs, this situation is highly unusual. The revisions that -we find in the other Eichendorff drafts, for example, primarily alter details or at most single phrases, but in general the conception of a song does not change between the first draft and the published version. The draft of "In der Fremde" thus affords a unique opportunity to observe Schumann's struggle as he tries to find an appropriate setting for Eichendorff's text. At some point after he first composed the draft, Schumann crossed out virtually the entire piano accompaniment in measures 6—9 and wrote a revised accompaniment, which represents the final version of these measures, directly on the draft. He also began to revise the accompaniment in the second half of measure 13, by crossing out notes and again -writing the new version over them. But by the time he got to measure 15, the revisions were becoming so extensive that he decided to simply cross out four measures of music and paste an insert over the draft. On the insert, Schumann expanded these four measures to six, and the result is a version that is very close to, but not identical with, measures 15—20 of
110
Schumann J Eickendorff
Sonffit
the published song. After completing the Berlin draft, Schumann revised the rhythm of the vocal line and the text setting in measures 16—23, and these final revisions are documented by a later draft, mostly in Clara Schumann's hand, that was evidently an engraver's copy and is now housed in the Bayensche Staatsbibliothek in Munich.17 But these final revisions alter details of the song and do not change the basic conception, so in fact there are essentially two versions of Schumann's "In der Fremde," and both of these can be reconstructed from the Berlin piano draft.18 The piano draft of "In der Fremde" has been discussed and at least partially transcribed by several twentieth-century scholars, yet the nature and significance of Schumann's revisions remain shrouded in confusion. The first scholar to study the piano draft was Viktor Ernst Wolff, in a 1914 dissertation that is concerned with the Berlin notebooks. But at the time that Wolff saw the draft, Schumann's insert hid the earliest version of what was to become measures 15 20. Wliat \Volff describes as a transcription of the "first version" of measures 15 28 is thus a conflation of the second version of measures 15—20, written on the insert, and the original version of measures 21—28. Sixty years later, in 197-4, Herwig Knaus published a facsimile of the Eichendorft drafts, making them widely available for the first time, but by this time the insert had fallen off and had been repasted onto a later page of the notebook. Knaus thus revealed the original version of the song, but he apparently never saw the insert and so was unaware of the second version of measures 15—20. The first scholar to publish transcriptions of both of the earlier versions of this passage was Barbara Turchin, in her 1981 dissertation and in a 1985 article, but she evidently did not realize that what she describes as "a separate, half-page fragment" had, in fact, been pasted onto the draft, and this led her to inaccurately transcribe the manuscript material and misrepresent the stages of composition that it records. It •was Rufus Hallmark who discovered a "telltale glue stain" on the separate fragment of "In der Fremde" that "matches a stain on the cancelled measures of the second page of the draft" and thus figured out that the fragment had been displaced. Hallmark accurately transcribed the original version of "In der Fremde" in its entirety in 1978, several years before Turchin's dissertation, but his transcription has still not been published, and so the actual history of the song's composition continues to be misunderstood.19 There is no way to determine the order of the two major revisions that Schumann made to "In der Fremde," but the close similarities between the second and fourth phrases in the final version of the song suggest that they are related. Hallmark writes that "one revision triggered the other in a chain reaction" and describes the process as a way of "tightening the musical coherence of the composition."20 While essentially accurate, this description perhaps understates the case, since in revising the fourth phrase Schumann not only created a more coherent connection with the second phrase but also drastically altered the tonal and formal structure of the song and in the process created a far more profound and sensitive setting of Eichendorff's poem. Table 4.2 compares the two versions in terms of formal and harmonic structure.
Poem and Song
111
TABLE 4.2 The Two Versions of "In der Fremde" Original Version Lined of Text
Formal Section.
No. of Bard
1-2 3-4
A A' B
A A
5-(6) (6)-7
8
A" CODA
6 A 4 (+3)
Final Version Formal Section
Principal Chordd
i-V-i i-V-i i-III i-V-i I
A A'
B
A'VCODA
No. of Ba/v
Principal Chord*
A A 6 6 4 (+3)
i-V-i i-V-i i-V7/iv iv-V-i I
We can see from this table that the way in -which Schumann rearranges Eichendorff's poem is virtually the same in both versions, but that the peculiar form that he eventually found for "In der Fremde," with an oversize B section that brings the harmonic structure to its conclusion before the A section has a chance to return, had not been conceived when the song was originally composed. At first, he placed the recapitulation at the beginning of the fourth phrase, so that the last seven measures of the song functioned as a clear-cut coda. In terms of the placement, proportions, and function of each formal section, Schumann's initial conception of "In der Fremde" adheres closely to the conventional expectations of ternary form. And since Schumann's play with formal convention becomes one of the primary means through which he expresses the complexities and ambiguities of Eichendorff's text, it seems likely that his understanding of the poem was still growing and developing as he revised the song. A second element that is crucial to Schumann's ultimate conception of "In der Fremde," yet virtually absent in the initial version, is the use of contrasting narrative and lyrical styles to differentiate between the narrator's description of her current situation and her dream of future transcendence. We can see in Example 4.5, which presents Hallmark's transcription of the second half of the third phrase and the complete fourth phrase from the original version of the song, that while the third phrase already exhibits the increased lyricism, the more active accompaniment, and the expansions of registral space and phrase rhythm that characterize the final version, it ends with a perfect authentic cadence in A major and then modulates abruptly back to Fit minor for the recapitulation. The contrasting B section is thus a self-contained six-measure unit, which does not disrupt the essentially paratactic formal structure of the song, and the balladic style that Schumann evokes in the opening phrases pervades the original version in its entirety. This, in a sense, is the basic problem with "In der Fremde" in its original guise. The simplicity of the texture, melody, harmony, and form practically makes it a folk song, and the contrasting B section does not really manage to turn it into an art song but rather introduces an incongruous element. From another perspective, we could say that what Schumann begins in the third phrase is cut off before it has a chance to develop, and thus the phrase is
EXAMPLE 4.5. Rufud Hallmark '<> transcription of measured 13-19 of the Berlin piano draft. From RufiM Hallmark, "Schumann i Reviiioiw of 'In der Freinde,' op. 39, no. 1, "presented at the Conference of Nineteenth-Century Mtuic, Nottingham, 1978. Reprinted by permitMion of the author.
Poem and Song
113
not coherently integrated into the song. In the first four measures of the phrase, which remain the same in both versions, Schumann creates an open-ended musical gesture that implies further expansiveness. As we have seen, the character of this gesture results not only from the disjunct melody, the increasing range, and the lyrical style of the vocal line but also from the fact that the underlying harmonic progression to A major in measures 12—13 is weakened by the pattern of descending parallel sixths between the outer voices. In the final version, Schumann repeats the pattern, beginning in the second half of measure 13, but changes the harmonic direction so that the phrase leads to the dominant of B minor. In the original version, he also begins to repeat the pattern of descending sixths in measure 13, but in the second half of measure 1-4 he interrupts it by adding the bass line that was implied in measures 12—13, creating a perfect authentic cadence in A major. We can compare the two versions of the phrase by observing that the music of measures 12—13 implies two different continuations, and where in the final version Schumann realizes the contrapuntal implications and ignores the harmonic implications, in the original version it is the other way around. But in strengthening the progression to A major as the phrase ends, Schumann creates too much closure at a moment when the music demands a further opening up. And he makes matters even worse by leading into the following phrase with an abrupt modulation back to Ft minor. The problem with Schumann's original third phrase results from the interrelation of the harmony, the counterpoint, and the melody, and so in revising it Schumann had to consider all three aspects simultaneously. His decision to complete the pattern of descending sixths that begins in measure 13 is just as significant as his decision to change the cadential harmony from a tonic chord in A major to a dominant in B minor. And while it is possible that it was the latter revision that necessitated the chromatic alteration of the cadential melodic pitch, the converse is equally plausible. Regardless of how it is harmonized, the original melody in measures 10—15 feels imbalanced, because it begins by gradually expanding and opening up but then contracts abruptly and resolves too quickly. In purely melodic terms, the alteration of a1 to at1 on the downbeat of measure 15 provides an elegant solution to the problem, preserving the melodic contour of the phrase yet giving it an open-ended quality. It is thus impossible to single out any one musical parameter as the motivating force behind Schumann's revision. All three play a role in creating the problem, and so he needed to fix all three in order to solve it. Another element that is introduced in the original version of the third phrase and then disappears inexplicably is the covering voice that begins with ft2 in the accompaniment in measure 10 and continues -with e2 in the vocal line in measure 13. Schumann's awareness that this voice is left hanging may account for one of the least successful moments recorded in the piano draft. In the fourth phrase of the song, Schumann recapitulates the opening phrase but varies the vocal melody by transposing the embellishing cadential formula in the second half of measure 18 up a minor third. Rather than ascending to b and then descending by step to ft1 at the cadence, as he does in the first two phrases of the song, Schumann
114
Schumann'if Eichenclorff Song
ascends to d2, comes partway down the scale, and then leaps to the cadential pitch. He must have recognized how awkward this is to sing, and his decision to revise it, by restoring the melodic contour of the opening, is one of his earliest revisions, "which he made directly on the draft before pasting in the insert. But although Schumann quickly fixed this spot, we are left -wondering how such a gifted melodist could conceive of this vocal phrase in the first place. Perhaps Schumann's aesthetic instinct told him that he needed to return to the higher register he had opened up in measures 10—13, not only to fulfill the structural implications of the descending line that is projected by the pitches ft2 and e2 but also to accurately express the psychological sense of the text. The sudden appearance of the covering voice in the third phrase is motivated by the narrator's expression of her longing for future peace, and so this voice should continue as she begins to imagine what it is she desires. If the motion up to d2 in Schumann's initial version of the fourth phrase was an attempt to establish such a registral connection, however, then it not only resulted in an awkward melodic line but also failed to achieve his expressive purpose. Coming -within a third incarnation of the opening melody and, moreover, as a variant of the most formulaic gesture of that melody, the ascent to d2 does not manage to evoke an expressive link with the contrasting lyrical melody of the preceding phrase. As we witness Schumann's initial struggle to accommodate his expressive response to the text to the requirements of the simple and repetitive form that he chose for the song, -we can appreciate all the more how inspired his ultimate solution is. For in his final version of the song the fourth phrase is still in some sense a return of the opening music, but it has been recomposed in the lyrical style of the third phrase and transposed to the subdominant, so that we do not hear it as a return but as a continuation and a further opening up. The arrival of d2 now creates an audible connection with the e2 of the previous phrase. The reason for this is not simply that the arrival comes at the beginning of the phrase, rather than the end, or that the singer continues to reiterate the note for two and a half measures rather than a single beat. More important, the emphasis on d2 has been motivated by several different yet interrelated musical processes—the covering voice that is initiated by ft2, the unfolding of the melody from e2 to at1, and the harmonic progression to the dominant of B minor — all of-which are set into motion within the third phrase. In the original version of "In der Fremde" Schumann responds to specific lines in Eichendorff's poem through the particular phrases he sets them with, but in revising the song he began to respond to the poem as a whole and use the form of the entire song to express its meaning. Each individual moment gains greater expressive force because one moment builds on another, resulting in a setting that captures the complexities of Eichendorff's poem. I have demonstrated one example of this process by explaining how the revision of the fourth phrase of the song enabled Schumann to realize the expressive implications of the third phrase. We can see a second example in the changing relationship between the fourth phrase and the second, as Schumann revised each of them in the piano draft. As shown in Example 4.6, Hallmark's transcription of the first version of
Poem am) Song
115
EXAMPLE 4.6. Ru/iM Hallmark '<> transcription of measures 6—9 of the Berlin piano draft. From Rufus Hallmark, "Schumann'
measures 6—9, Schumann's revision of the second phrase is far less drastic than that of the fourth. The vocal line remains the same in both versions, and -while he altered the accompaniment extensively, he left the underlying harmonic structure unchanged. What is different about the first version of this phrase is that Schumann makes explicit the conflicts and ambiguities that he only hints at in the final version. In the published song, as -we have seen, Schumann uses a descending third span in the tenor voice to delay the arrival of the dominant until the second half of measure 8. We find the same descending third in the first version of the phrase, but it is more strongly emphasized. The initial note of the span, Ctt, does not just appear in an inner voice on the downbeat of measure 6 but is reiterated on the downbeat of measure 7, now as part of the bass line. Because of the Ct in the bass, the dominant arrival is not delayed, and this changes the -way -we hear the phrase. In the later version, the unrealized implication of a voice exchange in measures 6—8 is one factor that encourages us to hear the I6 chord on the downbeat of measure 8 as part of a tonic prolongation, as I illustrated in Example A. la. But since -we hear this chord -within the context of a dominant prolon gation in the first version of the passage, it functions as an unambiguous substitute for an embellishing *. However, the step-wise descending bass line, -which
116
Schumann '
Songj
creates a pair of parallel sixths between the outer voices in measure 7, strengthens the implication that the melody should descend to ft)1 on the downbeat of measure 8, an implication that conflicts -with the harmonic function of the supporting chord. So, although there is no implied voice exchange in the first version of the second phrase, there is an even stronger conflict between the harmonic and contrapuntal implications. In both versions of the phrase Schumann uses these conflicting implications to express the heroine's feelings about her parents' deaths, but in the later version the conflict is more suggestive than explicit, so he can more easily manipulate it when the same music returns in the fourth phrase. By simultaneously increasing the lyricism and heightening the tension as the heroine imagines the peaceful solitude that she cannot attain, Schumann creates a compelling relationship between this image and her memory of her parents. Reading Schumann '<> Motivations
As is always the case with manuscript material that documents the compositional history of a musical -work, the piano draft of "In der Fremde" tantalizes us with a glimpse of Schumann's creative process, but ultimately it does not reveal the secret of his creative genius. We can see that his first conception of the song was fundamentally different from his final conception, we can document the specific changes that he made as he went from one conception to the other, we can explain how these changes affect the song, and we can even postulate the kinds of compositional choices that Schumann found himself faced with and speculate as to why he made the particular decisions that he did. As Leonard B. Meyer has observed in a more general context, however, the sketches and drafts that composers leave behind may enable us "to describe their compositional procedures" but do not allow us "to explain their compositional thinking." The actual mental processes that are involved in composing a work can probably never be known, not by scholars and not even by the composers themselves.21 The draft of "In der Fremde" enables us to reconstruct several stages of revision, but because these consist of changes that involve a variety of interrelated musical parameters, we cannot attribute them to any single motivating factor, and in reading previous literature that attempts to do so we learn more about the assumptions of the scholars who wrote about the draft than we do about Schumann's thinking as he revised the song. Barbara Turchin, for example, argues that the changes in measure 14, which appear on the piano draft itself, predate the changes in the following measures, which appear on the insert. This leads her to transcribe an intermediate version of the passage, the beginning of which is Example 4.7. Turchin compares this putative version to Schumann's initial conception by observing that where in the latter case "A major is approached by its major dominant," it is now "approached much more obliquely by its minor dominant seventh in third inversion." Turchin's description exemplifies her single-minded focus on the tonal structure of the song, which she considers to be "the primary source of Schumann's dissatisfaction"
Poem and Song
117
EXAMPLE 4.7. Barbara Turchin'^ transcription of the "second" version ofmeasures 14-15, © 1985 by The Regents of the University of California, reprinted from 19th Century Miulic v. 8, n. 3, Issue: Spring 1985, pp. 231—44, by permission.
with his original version.22 But in altering the Gts in the accompaniment in measure 14 to Gfe, as shown in Turchm's transcription, Schumann had already transformed the harmony in that measure from V in A major to iv in B minor, which means that he had already decided to change the harmonic direction of the phrase. And in terms of counterpoint, an element that Turchin ignores, her intermediate version defies musical logic. It is inconceivable that Schumann -would decide to continue the pattern of descending sixths through the end of measure 14 yet not bring the pattern to completion on the downbeat of measure 15. An analysis of these measures of the song, as they appear in the manuscript, is just as conclusive as the paleographic evidence discovered by Hallmark in proving that the separate fragment that contains the revised version of measures 15—20 had originally been pasted onto the piano draft and that Schumann recomposed the entire passage at one time. Turchin's hypothesis that Schumann was dissatisfied with the tonal structure of the original version of "In der Fremde" undoubtedly results from her belief that the first song generally plays a crucial role in Schumann's cycles and that in the case of "In der Fremde" this role is manifested in the correspondence between the "tonal plan" of the song, in its final form, and that of the Eichendorff LiederkreLt as a whole.23 It is true that Schumann's decision to eliminate the A major cadence at the end of the third phrase and replace it with a more expansive modulation to B minor is an important aspect of his revision. But as we have seen, the revision of this passage involves the melody and the counterpoint as well, and so we cannot single out the choice of key as the one factor that motivated Schumann to revise it. In fact, one could argue that the problem with the original version is not so much where Schumann is modulating but how he is doing it. Reinhold Brinkmann speculates that Schumann's difficulties with the song "are connected to the metric-rhythmic complications and the expressive character of lines 6-7 of Eichendorff's poem." In particular, Brinkmann points to the
118
Schumann j Eichendorff Song<>
medial caesura in line 6 and the enjambment with the following line, which together preclude setting the two lines with a symmetrical phrase structure. When Schumann first composed the song, he set the -words "iiber mir rauschet" with a two-measure grouping in measures 16—17 and the words "die schone Waldeinsamkeit" with another two-measure grouping in measures 18—19, as can be seen in Example 4.5. In the second version of this passage, which appears on the insert that Schumann pasted over the draft, he transposed it to B minor but retained the same 2+2 grouping. It was only after Clara Wieck had written out the fair copy of the song that he revised the internal structure of the vocal phrase, notating this final revision directly on the fair copy. Schumann contracted the first grouping and expanded the second, so that the four measures divide asymmetrically as 1+3, shifting the emphasis from the word "iiber" to the word "schone." It is this version of the phrase that appears in the published song.24 When we compare the alteration of the text underlay in this phrase to the other changes that Schumann made to the song, it appears relatively minor in scope, and in fact, he was evidently ready to publish "In der Fremde" with the original 2+2 grouping. And yet Brmkmann implies that we can look to this last-minute change to discover the motivating factor behind Schumann's decision to revise the song. It is true that Schumann's most significant changes occur in the passage of the song that sets lines 6 and 7. But it does not seem that any of them -were the direct result of his problems with Eichendorff's prosody. Even the added repetition of "die schone Waldeinsamkeit" in measures 20—21, as well as the altered grouping of the vocal melody in the preceding four measures—the two changes that are most closely bound up with the prosody—were apparently motivated more by expressive and musical considerations than by the demands of the poetic meter. And Brinkmann himself acknowledges that Schumann had already altered the poem quite significantly in his initial version by repeating "da ruhe ich auch," so that the cadence in measure 15 comes in the middle of the sixth line.26 But if we take a step back from the particular changes that Schumann made to the song and consider his revision in more general terms, we can see that Brinkmann's basic supposition is correct and that the irregularities of Eichendorff's poem were apparently an underlying source of Schumann's problems with the song. I have described the original version of the song as an unsuccessful compromise between strophic and ternary forms. Schumann was undoubtedly inspired to turn to the former by the folklike simplicity of Eichendorff's opening lines. And it is clear that the irregular prosody in the second half of the poem prevented Schumann from maintaining a regular phrase structure and a completely paratactic musical form and led him to expand the third phrase to six measures. I am not suggesting that he might have considered setting the poem •with a form that was strictly strophic. He -would clearly need to create some contrast in the song and expand the range of the vocal line for musical reasons, regardless of the poem's prosody. The breakdown of the meter in Eichendorff's poem does not simply necessitate musical variation, however, but specifically
Poem and Song
119
calls for some degree of irregularity in the phrase structure. And one of the central problems with Schumann's original setting is that the irregularity he introduces—a phrase expansion in the third phrase —implies lurther expansion and further development, and this was clearly a factor in Schumann's decision to revise the song. At the same time, the revisions themselves reveal his growing awareness of the correspondence between the prosody of the poem and its thematic content. As Schumann created an increasingly irregular musical form for "In der Fremde," he not only responded more sensitively to the implications of the poetic structure but also found a way to coherently express the meaning of the text. In this respect, the entire process of revision was inspired by the unusual aspects of Eichendorff's poem. I have made "In der Fremde" serve as a model for my analyses of the other songs from the Eichendorff Liederkrew because the history of its composition and the form in which Schumann ultimately cast it reveal important aspects of his approach to Eichendorff's poetry that in other cases remain hidden below the surface. In "Friihlingsnacht" and "Intermezzo," which are the subjects of my next chapter, and in "Mondnacht," which I discuss in chapter 6, the overwhelming regularity of the phrase rhythm, the simplicity of the form, and the faithfulness of the text setting leave us with the deceptive impression that Schumann's intention is simply to present Eichendorff's poems as straightforwardly as possible. But just as the texts themselves clothe a profoundly expressive world of spiritual exploration in naively simple poetic forms, so Schumann's settings transcend the regularity of their outward forms and plumb this world to its depths. Behind the apparent conventionality of these settings lies the same complexity and freedom that we have found in "In der Fremde."
This page intentionally left blank
WEAXIOPENINGS
O;
kne way to explain the concept of closure in music and literature is "to say that there is a metaphorical frame that separates the complex structure of the artwork from the chaotic flow of everyday life. Barbara Herrnstein Smith argues that because our experiences are so often "fragmentary, interrupted, fortuitously connected, and determined by causes beyond our agency or comprehension," we feel compelled to create "structures that are highly organized, separated as if by an implicit frame from a background of relative disorder or randomness."1 In the case of a musical performance, as Edward T. Cone observes, this frame is created by the silence that precedes and follows. But Cone describes a variety of techniques that Romantic composers use to blur the boundary between the musical -work and the surrounding silence, creating the impression that the music has begun before we hear the first notes and will continue after we hear the last ones, and it is in this category that Schumann's experiments with song forms belong.2 The Romantic impulse to disguise or overflow the musical frame must be differentiated from the attempts of some twentieth-century composers to get rid of it altogether and thus to negate the viability of the musical work as a closed structure. Both procedures can be interpreted as responses to an increasing sense of chaos and puposelessness, •which calls into question the possibility that art can provide the stability and coherence that are felt to be lacking in life. But where the elimination of the frame in more modern music expresses resignation or perhaps acceptance, the weakening of closure in Romantic music is associated with the wistful yearning for transcendence, the hope that •what appears to be fragmentary in the material world will turn out to be whole in the world of the spirit. In a weak opening, which is one of Schumann's most important techniques for obscuring the closural frame in his musical forms, the tonic is not very strongly defined as the song begins. This may result from the absence of an authentic cadence or the lack of any harmonic motion at all. Sometimes the tonic chord is not stated explicitly but is implied by its dominant, and the first cadence we hear may tonicize the latter harmony. At its most extreme, the weak opening results in a sense of tonal ambiguity. As with the open ending, the weak opening encompasses a •wide variety of compositional techniques, a surprising number of 121
122
Schumann 'd Eichendorff Songd
which are found within the Eichendorff songs. Not every song in the cycle begins with a weak opening and Schumann employs similar techniques in numerous other pieces, but there is no other work in which he explores the possibilities of the weak opening so comprehensively, and it is this formal idea, as much as anything else, that gives the Eichendorff Liederkreu its individual stamp. In analyzing musical form, we usually consider closure to be a function of how a piece ends and not how it begins. But form itself is a spatial construct that is used only metaphorically to refer to temporal arts such as music and literature, and to say that a spatial form has closure clearly does not imply that it concludes, as Smith has observed. When we use the word closure to refer to "visually perceived forms," we mean that they are "spatial structures which exhibit relatively clear, coherent, and continuous shape." And -when we extend the term to a work of music or literature, it still implies a form that "is experienced as integral: coherent, complete, and stable." It is not so much the ending itself that determines whether or not a musical form has this coherence and stability but the extent to •which the ending fulfills certain expectations that have been created both by the formal design of the work and by the conventions of its style and genre. In this sense, we could describe the ending as simply one moment — albeit the most important moment—when our desire for closure can potentially be satisfied. The opening, however, is the moment when these expectations are first aroused, and they can be exploited in a variety of -ways as the musical form unfolds. Schumann often frustrates our expectations, and in this way he creates song forms that are fragmentary in the sense that they do not fulfill their own implications. One reason that the weak opening has so much expressive potential is that it is an end-accented structure, by which I mean that the rhythmic impulse with which it ends is stronger than the impulse with which it begins. Such structures tend to replicate hierarchically, and so they lead us to expect a series of increasingly stronger impulses as successively larger formal units conclude, with the most powerful climax coming near the end of the piece. In orchestral music, Romantic composers often satisfy this expectation by resorting to what Leonard B. Meyer calls a "statistical climax," in which the listener is overwhelmed by a massive increase in texture, dynamics, and range. Because these types of climaxes are created through the use of secondary parameters, they do not resolve to a clearly articulated moment of closure but lead to a gradual "abatement" or "cessation," in other -words, to an open ending. Meyer argues that the sheer magnitude of the statistical climax enables Romantic composers to express the sublime^— "the transcendent aspect of experience" — and he contrasts this with "the intimacy of the small piano piece and the modest lied."4 But even if, in absolute terms, Schumann cannot create as much sound with a voice and a piano as he can -with an orchestra, he can still produce a statistical climax on a smaller scale. In any case, the texts of Schumann's Eichendorff songs are not so much concerned with the expression of transcendence in and of itself but with the unfulfilled yearning for transcendence, and so the denial of an expected climax is sometimes more powerful than the satisfaction of being overwhelmed -with sound. The theme of transcendence is perhaps the one common element in the other-
Weak Openings
123
wise contrasting texts of "Friihlingsnacht" and "Intermezzo," the two songs that I will consider in the present chapter. Schumann's musical settings of these texts are also quite different from each other, but they both derive from the same simple underlying form: a ternary ABA', preceded by a one-measure introduction and followed by a five-measure coda. Each section of the form consists of a single eight-measure phrase that breaks down into four-measure and two-measure groupings. The A section begins with a weakly defined tonic harmony and ends with an authentic cadence that tonicizes the dominant. The A' section is then varied such that the tonic becomes stronger over the course of the phrase, and rather than cadencing on the dominant, the section concludes with the first (and only) perfect authentic cadence in the tonic key. Schumann manipulates this schema very differently in each of the songs. In "Intermezzo," on the one hand, he repeatedly prepares and then frustrates a strong tonic arrival, undermining even the final cadence, so that the song feels open-ended despite having harmonic closure. In "Friihlingsnacht," on the other hand, there is a strong tonic arrival relatively early in the form, but it comes at the beginning of the B section and thus turns the entire A section into an extended upbeat. And Schumann does create a powerful climax near the end of "Friihlingsnacht," but there is a slight disjunction between the climax and the final cadence, and this compromises the closure of the song. As I consider these two songs in juxtaposition, I will explain how Schumann uses the formal idea of the weak opening to respond to a particular poetic theme and convey some sense of the subtlety and diversity that this idea provides him with.
"Friihlingsnacht" Conflicting Patterns
"Friihlingsnacht" begins with a root position tonic triad on the downbeat of the first measure, but its key remains ambiguous throughout the A section of the song (measures 1—9), primarily because of Schumann's pervasive use of chromaticism. All twelve chromatic pitches appear within the first six measures, and the vocal phrase begins with a nondiatomc sequence (measures 2—5), so that the first two melodic gestures are based on the GK minor scale and the Alt minor scale, respectively. Schumann accompanies this half of the phrase with a chromatically rising line that tonicizes each of these pitches in turn. The two gestures that comprise the second half of the phrase (measures 6—9) are both diatonic within the tonic key of Ft major, but Schumann harmonizes the first of them by tonicizing D( minor and the second by tonicizing Ct major, so that within the course of this relatively short opening phrase we rapidly move through four different keys, none of which turns out to be the tonic of the song. But even though this chromaticism creates the impression of tonal ambiguity as the song begins, it is actually embellishing a straightforward voice-leading structure, which is illustrated in Example 5.1.
124
Schumann d Eichendorff Song<>
EXAMPLE 5.1. MSKHegroutid voice-leading analydui of metuur&i 1—9,
My analysis shows that the phrase is an example of a Schenkerian third divider. The bass motion from I to V is arpeggiated by the iii chord, and the arpeggiation is filled in further by a passing motion between I and iii and an embellishing fifth motion that leads to V. The soprano begins by prolonging 3 through an ascending line in parallel tenths with the bass. Although the melody moves into an inner voice at the cadence, the descent from 3 to 2 is strongly implied by the unresolved at1 in measure 8 (doubled at the octave above in the accompaniment). If-we -were to derive a later structural level from this graph by filling in the chromatic pitches, it would be clear that they all serve, in one way or another, to elaborate and prolong the underlying motion from I to V. One reason that this prototypical voice-leading structure sounds so ambiguous is that Schumann creates a conflict between the harmony and the phrase rhythm. The initial tonic comes on the downbeat of the first measure, but as is clear from the vocal melody, this entire measure is an upbeat, and the phrase begins in measure 2. The tonic is thus in a weak hypermetrical position, and the phrase is end-accented on several hierarchical levels. The first two-measure grouping moves from the chromatic passing harmony of measure 2 to the consonant and diatonic Gl minor triad of measure 3. This latter chord is itself a passing harmony, however, that resolves at the end of the following grouping, in measure 5, to the mediant harmony, giving the fourth measure of the phrase an even stronger accent than the second measure. But the eighth measure of the phrase, measure 9, is accented more strongly still, because it is at this point that the harmonic motion concludes with the arrival of the dominant, which is tonicized through an imperfect authentic cadence. The -weak hypermetrical position of the opening tonic chord leads us to expect that it will return at a stronger point in the phrase, but with each successive accent the harmony moves further from the tonic. Since the strongest harmonic motion comes with the progression to the dominant at the cadence, the A section is both end-accented and open-ended. This conflict between harmony and phrase rhythm manifests itself in the opening phrase through the interaction between two patterns—a contrapuntal pattern that controls the harmonic motion and a melodic pattern that is created by the motion of the vocal line, as I have illustrated in Example 5.2a. Schumann moves from the tonic harmony in measure 1 to the mediant chord in measure 5
Weak Openingj
125
a.
b.
EXAMPLE 5.2. Analyjut of the nulodic and contrapuntal pattern*).
through a. three-voice contrapuntal pattern in the accompaniment, -which consists of a series of parallel tenths between the outer voices and a 5—6 alternation between the two lower voices, -with the latter generating the chromatic passing harmonies. As the mediant chord concludes this pattern in measure 5, it simultaneously initiates a descending variant of the same pattern, which moves twice as quickly, so that -where the ascending tenths come every other measure, the descending tenths come every measure. The descending pattern breaks off in measure 6 after the second tenth, between gf and b2, but when the same music is recapitulated in the A' section, beginning in measure 18, the pattern is completed, and we can see in the two lower staves of Example 5.2b that in its complete form it returns to the tonic in the sixth measure of the phrase, measure 24. The reason that the contrapuntal pattern is disrupted in the opening A section is that it ultimately comes into conflict -with the melodic pattern of the vocal line. The vocal phrase in this section is an example of what Meyer calls a "complementary melody," in which the second half of the melodic line inverts the first
126
Schumann 'd Eichendorff Song<>
half. In the present case, the inversion is not exact, since the pair of rising sixths in measures 2—5 is answered by a pair of falling fourths in measures 6—9, as shown in the uppermost staff of Example 5.2a. The complementary relationship is clearly audible, nevertheless, because both halves of the phrase have essentially disjunct motion and similar rhythmic relationships, and both are sequential. This melodic pattern corresponds to the way that the phrase is grouped, since each unit coincides with a two-measure grouping, and in the first half of the phrase it corresponds to the contrapuntal pattern as well, with each rising sixth creating a link between the two upper voices. But as the two patterns reverse direction in measure 5, the acceleration of the descending tenths causes them to fall out of sync. The contrapuntal pattern implies a return to at1 in measure 7, but the complementary melody implies a motion into an inner voice, with at1 coming at the beginning of the following two-measure grouping in measure 8.6 The vocal melody continues the latter pattern, descending to fit1 in measure 7, and the bass, as if following the singer, substitutes dtt for its expected ft. But as I have shown in Example 5.2a, djt actually initiates a cadential ii—V—I progression in Ct major, and when the vocal line finally reaches at1 in measure 8 the harmonic progression turns this pitch into a suspension, which Schumann leaves unresolved as the vocal line leaps back into an inner voice at the cadence. Climactic Momenta It is only in measure 10, at the beginning of the B section, that the tonic returns and provides consonant harmonic support for at1. Unlike the upbeat measure with which the A section begins, measure 10 is a strong measure, and so, where the opening phrase is end-accented, in the B section the strongest rhythmic impulse comes at the beginning. In fact, the sense of stability and resolution in measure 10 is so powerful that we hear it as a structural downbeat, which turns the entire A section into an extended anacrusis. Schumann uses several means to emphasize this arrival. The vocal melody is static for the first time in the song, and in contrast with the opening of the A section, -where the harmony immediately begins to move away from the tonic through a chromatic progression, Schumann uses a dominant neighboring chord to prolong the tonic for three measures. Equally significant is the way in -which the vocal line and the uppermost voice of the accompaniment converge at this point in the song. In the first phrase, the principal melodic line is shared between these two voices, but Schumann divides it up in such a -way that they do not create a unified utterance. The complete line, which I have reconstructed in Example 5.3, is created by the alternation between the step-wise descents of the right hand and the arpeggiated ascents of the vocal melody, which are presented in conflicting rhythms and in different registers. As the B section begins in measure 10, the nervous triplets in the upper voice of the piano slow down to straight sixteenths, and Schumann introduces a new motive that embellishes the vocal line's motion from at1 to b1, and does so in the same register/ When this motive repeats in measures 12—13, the vocal line
Weak Openings
127
EXAMPLE 5.3. Reconstruction of the composite melodic line in measures 1—5.
picks it up, and, for the first time in the song, singer and pianist are moving together in unison. Despite the strength of this tonic arrival, however, the feeling of resolution that it creates turns out to be transitory. One reason for this is that the stability with which the phrase begins quickly dissipates in the second half, as the dominant harmony unfolds through a series of chromatically rising parallel tenths, and the upper voice of the accompaniment once more diverges from the vocal line. A second reason is that, in terms of the form of the song, the arrival comes too late. The conventional expectations of ternary form dictate that the tonic be firmly established before the A section ends, leaving the B section to begin •with the dominant or an even less stable chord. In the case of this particular song, what -we hear in measure 10 is what we expected three measures earlier, where it would have concluded the contrapuntal pattern that is derailed in the course of the opening phrase. In fact, •when the contrapuntal pattern is repeated in the final A' section, it is completed in measure 24 by a varied return of the music of measure 10, including the accompaniment motive, and one of the reasons that this arrival feels so much stronger than the earlier one is because we finally hear this music at the expected moment within the form. The tonic arrival in measure 24 leads to the climax of the song, which in turn resolves to the first and only perfect authentic cadence, as the final vocal phrase concludes. The downward sweep of the right hand as the piano crescendos in measure 24, the powerful emphasis on the dominant in the following measure, and the slowing of the rhythm from triplets to straight sixteenths just before the cadence resolves all work together to create a strong closural gesture. And yet, for all of its rhetorical finality, the ending of "Fruhlingsnacht" is strangely unsatisfying, as if the enormous restless energy that has built up throughout the song can barely be contained. We can explain this feeling, analytically, by considering the conflict between the melodic pattern of the vocal line and the contrapuntal pattern of the accompaniment, -which, in its earlier incarnation, contributed to the open ending of the first phrase. We have seen, in that instance, how the two patterns fall out of sync after the melodic reversal in the fourth measure of the phrase, so that the continuation of one of the patterns inevitably causes the disruption of the other. In the final phrase, shown in Example 5.2b, the vocal melody descends by step to a| in measure 24, as dictated by the contrapuntal pattern, rather than moving into an inner voice, as the complementary schema implies. Although the pattern concludes at this point, the descending melodic line from 5 in measure 22 to 3 in measure 24 implies a further continuation to 1 at the cadence in measure 26, a melodic progression that would normally provide a strong sense of closure. But Schumann cannot conclude in this -way, because the pair of as-
128
Schumann i Eichendorff Song
cending leaps in measures 19—22 implies a complementary pair of descending leaps, and while he can get away with substituting a linear descent for the first melodic grouping, in measures 23—24, a line that descends entirely by step would create a weak and unconvincing ending. Schumann's solution to the problem, which I have illustrated in the upper staff of Example 5.2b, is to return to ct2, the melodic peak of the vocal line, as the final grouping begins in measure 25, and then conclude the melody with a leap of a perfect fifth to 1. There is, of course, scarcely any trace of the complementary relationship between the two halves of the phrase and none at all of the sequential relationship between the two groupings of the second half. But the motion up to ct2 creates a melodic climax as the final cadence arrives, and the subsequent leap downward from the climactic pitch compensates for the stepwise descent of the preceding grouping and provides the entire melodic phrase with a sense of balance. If we consider the climactic moment in measure 25 from another perspective, -we can see that it results from the -way in which Schumann varies the music of measure 10 -when it returns in measure 24. In measure 10, the motive in the right hand of the accompaniment begins from the singer's at1. Because of the contrapuntal pattern that controls the voice leading in the A' section, however, when the motive returns in measure 24 it must begin from ct2, in order to complete the 5—6—5 alternation, and leap up to at (preceded by its upper neighbor b2) to complete the series of parallel tenths. And so the right hand stays above the vocal line, rather than converging •with it, as it had in its earlier appearance. But as the piano line descends from b2, the vocal line joins it, and, as if swept up by the pianist's excitement, the singer leaps into a slightly higher register. The immediate result of this upward leap is the melodic climax just as the dominant arrives. But the ultimate result is that the melodic closure of the cadence in measure 26 is weakened, since the leap up to ct2 disrupts the stepwise descent toward the tonic that has been under way since measure 22 and substitutes a less forceful motion from 5 to 1. This, in turn, creates parallel octaves between the outer voices and thus weakens contrapuntal closure as well. It is perhaps for this reason that Schumann substitutes a V7/IV for the cadential tonic harmony, which only gradually resolves to a consonant triad. The counterpoint at the cadence, •with all of the voices leaping downward, is too abrupt to provide a sufficient resolution, and by drawing out the cadential harmony for three measures Schumann is able to soften it and, in a sense, to correct it. "Fruhlingsnacht," like "In der Fremde," concludes with an open ending, but Schumann creates this ending in a different •way and uses it for a different expressive purpose. In the case of "In der Fremde," there is an unequivocal final cadence that resolves both the accompaniment and the vocal melody at the same moment. This cadence does not provide sufficient closure for the song because it comes too early within the form, and its resolution is undermined by the subsequent events of the coda. In "Friihlingsnacht" the final cadence comes where we expect it, at the end of the final vocal phrase and of the third section of a conventional ternary form. But now Schumann compromises the moment of the cadence itself by -weakening the melodic and contrapuntal resolution and delaying
Weak Opening*
129
the harmonic resolution until after the melody and the bass line have concluded and after the singer has stopped singing. The peculiar quality of this ending derives in large part from the implications that Schumann sets up at the opening. In my analysis of the song, I have focused in particular on the irreconcilable conflict between the melodic pattern of the vocal line and the contrapuntal pattern of the accompaniment, and I have observed that this conflict necessitates an open ending because there is no possible way to conclude both patterns. More generally, I would describe the formal process that Schumann uses in "Friihlingsnacht" as an example of how the weak opening typically works. The lack of a strong and clearly defined tonic harmony at the beginning of the form and thus of an initial point of stability leads to the desire for a powerful climactic moment followed by a clearly articulated close. This desire grows stronger as the form unfolds until it can no longer be satisfied through conventional closural formulas, and so even if Schumann ends with a melodic and harmonic resolution, he must still leave the ending open in one way or another. Lottt and Reborn in the Night
Schumann uses the weak opening in "Friihlingsnacht" to create a sense of kinetic energy that drives us forward toward climactic moments that dissipate rather than resolve. It is a song that overwhelms us with its uncontrollable excitement and disorients us even as it rushes onward in a single, clearly directed stream of notes. In this way, Schumann beautifully captures the psychological world depicted in Eichendorff's poem, in -which the sensations of the spring night elevate the narrator to a state of ecstasy that dissolves the boundary between the self and the surrounding -world. The poem presents a version of a familiar German Romantic topos—the epiphany, or Augenbiick — a paradoxically fleeting moment in which one transcends the boundaries of time and space and catches a glimpse of eternity. For Eichendorff, as for many Romantic writers, such moments almost always occur during the night, a time -when the activity of the everyday world ceases and we enter a dream-world that enables us to come into contact with our souls. Eichendorff also depicts the night as a time of rebirth and renewal, since it always leads to a new dawn and thus to a reawakening, and this is why in "Friihlingsnacht" he associates it -with the season of rebirth, spring, literally welding the two into a single -word in the title. Ueber'n Garten durch die Liifte Hort' ich Wandervogel ziehn, Das bedeutet Fruhlingsdilfte, Unten fangt's schon an zu bliihn.
Above the garden through the skies I heard the migrating birds flying, Which signifies the smells of spring, Below it is already beginning to bloom.
Jauchzen mocht' ich, mochte weinen, 1st mir's doch, als konnt's nicht sein! Alte Wunder wieder scheinen Mit der Mondesglanz herein.
I would like to rejoice, I would like to cry, It is to me as if it could not be! Ancient miracles again appear With the shining of the moon.
130
Schumann'^ Eichendorff
Songj
Und der Mond, die Sterne sagen's, Und in Traumen rauscht's der Hain, Und die Nachtigallen schlagen's: Sie ist Deine, sie ist Dem!
And the moon, the stars say it, And in dreams the grove murmurs it, And the nightingales sing it: She is yours, she is yours!8
Schumann not only retains Eichendorff's simple stanzaic structure in his setting of the poem, turning each four-line strophe into an eight-measure phrase, but also mirrors its emotional contour through the -way in which he constructs his ternary form. The opening of the poem, like that of the song, is full of anticipatory motion, as the narrator senses the coming of spring. Eichendorff emphasizes the encompassing and even overwhelming quality of this sensation by locating it both "ueber" (in the skies) and "unten" (in the earth) and by describing it as a synesthetic experience, in which the sounds of the birds cause the narrator to smell the blossoming of spring flowers. At the beginning of the second strophe, the narrator passes from his perception of the surrounding landscape to his emotional response, a feeling so powerful that he does not know whether to laugh or cry. He alludes to an ecstatic state in which he loses his sense of himself and becomes open to the surrounding world. His description of this feeling suggests a confluence of opposites -within that corresponds to the oneness that he perceives around him. This is the first climactic moment in the poem, -which Eichendorff emphasizes by inverting the syntax of the fifth line and placing the verb "jauchzen" at the beginning. He thus starts the strophe with a strong initial accent that results from the convergence of metrical, syntactic, and semantic elements. It is at the setting of this -word, the moment -when Eichendorff transforms the confused rush of impressions into a clear and powerful feeling, that Schumann places his structural downbeat, suddenly clarifying the tonal ambiguity of the preceding phrase and bringing the vocal line and the accompaniment into phase. But for Eichendorff, as for Schumann, the moment is fleeting, and in the second half of the strophe the narrator's awareness again moves outward to the surrounding night. In contrast to the descriptive opening lines, the middle strophe does not refer to anything concrete, and the lack of detail emphasizes the ephemeral quality of the landscape. Eichendorff characterizes whatever it is that the moonlight reveals as the reappearance of ancient miracles, suggesting that the narrator is not experiencing material reality but has entered the -world of dream and spirit, which is accessible in the night but -will vanish when day returns and the dullness of the physical -world again becomes visible. The third strophe of the poem, like the third phrase of Schumann's song, recapitulates the opening, as Eichendorff once more describes the elements of the landscape. But now the narrator perceives this landscape exclusively through his sense of hearing, as it appears to attain consciousness and communicate to him in a single, unified voice. Because of its anaphoric structure, the entire strophe is clearly directed toward the final line, in -which the message of the night is stated. This moment, the climax of the poem, is in many ways analogous to the
Weak Openings
131
musical climax with which Schumann sets it. Eichendorff also uses rhetorical means to create a powerful closing gesture but simultaneously undermines its effect. In thematic terms, the climactic force of the last line derives from the narrator's sensation that the boundary between himself and the natural world has disappeared completely and he has merged into the landscape. For when "we consider -what it is that the night is saying, we realize that it is giving utterance to the narrator's own desire, and so the voice that he hears surrounding him actually comes from within. At the same time, however, the final line disappoints the expectation that Eichendorff has aroused from the beginning of the poem. We anticipate that the voice of nature -will express a transcendent and revelatory truth and not simply tell the narrator's fortune or confirm what he already feels. But in fact, it is inevitable that this expectation, much like the weak opening of Schumann's song, will remain unfulfilled, since the true message of the night is ineffable. The ending of "Friihlingsnacht" is among the most misunderstood moments in the Eichendorff Liederkreui. Schumann scholars have typically explained it as the fulfillment of the protagonist's desire to be united with his beloved, even though there is nothing at all within the text of the poem that suggests such an interpretation. On the contrary, it is clear from the mysterious and dreamlike quality of the scene and from the depiction of the narrator's solitary communion with nature that he is alone throughout the poem and that the words of the final line are to be read more as his imaginative transformation of the surrounding landscape than as a description of external reality. The reference to his beloved almost appears as an unmotivated intrusion, which is only justified by the traditional association between spring and love and by the analogy between the narrator's spiritual union •with nature, on the one hand, and his imagined physical union with his beloved, on the other.10 The significance of the line thus depends upon the beloved's absence, since Eichendorff's point is that the narrator is able to transcend their physical separation through the agency of the spring night. It is not love but transcendence that is the mam theme of Eichendorff's poem, and it is the treatment of this theme that Schumann responds to through his use of a weak opening.
"Intermezzo" Dem Bildmfi wundersehg Hab' ich im Herzensgrund, Das sieht so fnsch und frohlich Mich an zu jeder Stund'.
Your blessed image I have in the bottom of my heart, It looks so brightly and happily At me through every hour.
Mem Herz still in sich singet Em altes, schones Lied, Das in die Luft sich schwinget Und zu dir eilig zieht.
My heart sings quietly to itself An old, beautiful song, Which soars into the air And quickly flies to you.11
132
Schumann's Eichendorff
Songj
Transcendent Love The theme of transcendent love is more explicitly the focus of the poem "Intermezzo," but in this case the object of that love is left completely ambiguous. Most Schumann scholars have assumed that the poem is addressed to the narrator's distant beloved, but the Eichendorff scholar Lawrence Radner reads it as a religious poem that is directed to the Virgin Mary.12 The latter interpretation is somewhat truer to the poem, given the description of the image as "wunderselig" in the first line and the allusion to prayer in the first half of the second strophe. But since the beloved is completely absent, her identity is, by necessity, irrelevant to our understanding of the poem, and in fact, as in "Friihlingsnacht," it is the absence itself that is the most important element. Eichendorff describes a mode of communication that is both -wordless and silent and creates a secret inner bond that transcends time and space. The narrator's beloved communicates to him through her image, which he sees gazing at him within his heart, and it is his heart that responds, by singing a silent song that takes flight and bridges the implied distance between them. Eichendorff embodies the theme of ineffability in the poem itself by omitting any reference to the setting or the characters and by leaving unstated the content of the communication. "Intermezzo" thus provides an excellent example of the -way in which a Romantic poem can be like a sketch, in the sense that August Wilhelm Schlegel suggests.13 Eichendorff's poem is not concerned with the visible reality of the external world but -with the higher reality of the soul, which is beyond the grasp of concrete language. Schlegel argues that an appropriate illustration for such a literary work is not a "completely worked out painting," in which the visual artist tries to add what the poet has left unstated, but a sketch that is analogous to the poem and provides a stimulus that can only be completed in the viewer's imagination. Schlegel's injunction is all the more appropriate in the case of a composer who is setting such a text to music, since the latter art form is far more successful in hinting at the -world of the spirit than it is in depicting everyday life. Schumann sets Eichendorff's poem with a song that is virtually a musical sketch, in which he establishes the tonality, the harmonic structure, and the phrase rhythm with "few and sensitive strokes," as Schlegel puts it, and leaves a great deal to the listener's imagination. Rather than leading up to a moment of epiphany, as in "Fruhlingsnacht," Eichendorff describes a timeless state of transcendence in "Intermezzo," so Schumann cannot track the emotional contour of the poem, as he does in the former song. His setting of "Intermezzo" does not depend on a formal correspondence between lines of text and phrases of music but on a more general response to the affect and meaning of the poem, and so it does not matter that he rearranges the stanzaic structure by returning to the first strophe for the recapitulation of the song. In fact, the only point when Schumann's music corresponds in a meaningful way to the specific lines of poetry that he is setting is in the second strophe, -which becomes the B section of the song. As the singer describes the silent song that begins deep in his heart and then soars into the air
Weak Openingd
133
and flies quickly to his beloved, he traverses a major ninth, beginning from the lowest note in the melody, e1, on the word "Herz," and reaching the highest note, fit2, on the word "Luft." The accompaniment adds to the effect with its modulatory harmonic progression and its continuously syncopated rhythm, and Schumann's instruction to both performers — "nach und nach schneller und schneller" — is clearly intended to create a literal depiction of the quickly flying song. But this image provides the only opportunity in the poem for such explicit tone painting, and for the most part, Schumann responds to Eichendorff's text through the way in which he creates musical form. Parallelism and Continuaiunedti
The A section of Schumann's "Intermezzo" is an eight-measure phrase (preceded by a one-measure upbeat) that is organized as two successive periods, the first of which, in measures 2—5, establishes A major as the tonic and the second of which, in measures 6—9, tonicizes E major. Schumann uses several means to create a sense of continuity and increasing intensity throughout the phrase, so that, as in "Friihlingsnacht," it is end-accented, with the initial prolongation of the tonic weaker than the subsequent tonicization of the dominant. Because its structure is both parallel and end-accented, the opening sets up a pattern in •which a relatively weak cadential motion is followed by one that concludes more strongly. We expect the same pattern to repeat in the recapitulation, but now with the stronger cadence providing tonic closure. The weak opening of "Intermezzo" thus implies a series of increasingly strong cadences, •with the most conclusive cadence coming at the end of the song. As in "Friihlingsnacht," however, our desire for a climactic resolution becomes so strong that the conventional cadential formula -with which Schumann tries to end the song is rendered anticlimactic, and he is forced to undermine the cadence and open up the ending. If the ambiguity and instability of "Friihlingsnacht s" opening depend upon excess — of rhythmic motion, texture, and chromaticism—then it is just the opposite with the opening of "Intermezzo," as here it is a question of •what Schumann leaves out. The lack of a bass note in the first measure, for example, makes it impossible to determine the meter or the tonality as the song begins. Schumann quickly clarifies both by placing an A on the downbeat of measure 2, but he continues to syncopate the inner voices, creating a gentle, pulsating tension that persists for almost the entire song. And he uses the most minimal of means to establish the key of A major, avoiding the leading tone and the tonic in the melody and preventing any motion in the bass line through the use of a tonic pedal. The vocal line has a very narrow range in measures 2—5, essentially restricting itself to the fourth between b1 and e2, and a periodic structure that is •weakly defined. A double neighboring note motion to 2 gently establishes a tension at the end of the first two-measure grouping, which is partially resolved with a similar motion to 3 in the second. The second grouping complements the first by inverting its melodic contour and replacing its pattern of appoggiaturas with a series of strong downbeats, as shown in Example 5.-4.
13-4
Schumann'<> Eicbendorff Song<<
EXAMPLE 5.4. Vocal melody and it
2-5.
The at1 is the only chromatic pitch in the melody, but because of its strategic placement at the end of measure 2 it weakens the periodicity by implying a tonicization of b1 at the end of the first grouping and thus a continuous harmonic progression throughout the four-measure period, as I have indicated with the Roman numerals in Example 5.4. Schumann does not explicitly present this progression in the accompaniment but hints at it through the dissonant counterpoint of the inner voices, -which unfold above the bass pedal. The counterpoint, which is illustrated in Example 5.5, results in two further chromatic pitches, the g^ in measure 2, -which acts as a passing tone, and the dK1 in measure 3, a lower neighbor to e1. The dtt1 helps to turn the harmonization of B into a V7/V, which prevents us from hearing a caesura altogether. The gll creates more complex impli cations, -which are ultimately worked out in the course of the recapitulation. Schumann typically uses this chromatic inflection to transform a tonic harmony into a V7/IV. In the present case, he is adding the gll to a tonic triad, but the har monic function of this pitch is complicated by the at1 in the vocal line, which creates an augmented ninth above it. We thus hear gll as a displacement of the following ft that turns the expected V7/ii in the second half of the measure into a diminished-seventh chord. I have suggested two ways in which Schumann's accompaniment contributes both to the continuousness of "Intermezzo's" opening period and to the instability of its initial tonic. First, there is no bass motion, so we do not hear the accompaniment as a clearly defined harmonic progression but as counterpoint that implies such a progression. Second, Schumann uses chromaticism and dissonance within the counterpoint to transform the implied progression into a series of consecutive seventh chords -whose resolutions are elided. A third element of the accompaniment that also weakens the grouping structure of the vocal line is the countermelody that begins in the right hand in the second half of measure 3. We initially perceive the rising fifth from b1 to ft 2 as an inverted echo of the opening vocal gesture, but as the right hand continues with another fifth from e2 to b2 in the second half of the following measure it becomes clear that -we are hearing an independent melodic line. In measure 5 this countermelody undermines the cadential motion that is suggested by the vocal line. The major sixth between d2 and b2 that is suspended into the downbeat verticalizes the double neighboring motion that resolves to 3 in the vocal melody, delaying the resolution in the accompaniment until the second eighth of the measure. Furthermore, the b2 completes an upward octave transfer and thus opens up a higher register and reaches a climactic point just as the vocal melody concludes. So even as the right hand resolves to cl3 in measure 5, it implies continuing motion rather than closure.
Weak Openings
135
EXAMPLE 5.5. Voice-leading anaLyju of measured 2-9.
As we move into the second half of the phrase, Schumann intensifies all three elements of continuity as he moves to the E major cadence in measure 9. First, the increasing melodic activity in the right hand culminates with the arrival of the new dominant in measure 7. The beginning of the sixteenth-note figure in that measure overlaps with the ending of one vocal grouping and is immediately imitated by the beginning of the next, resulting in a measure and a half of continuous sixteenth-note motion that helps to propel us toward the cadence. Second, there is more chromaticism in the second half of the phrase, but with the exception of the many Dts, none of the chromatic pitches have any harmonic significance, and so they do not create the conflicting tonal implications that the few chromatic notes in the opening measures do. Instead, they all function as ascending passing tones, which intensify the forward motion of the phrase. Finally, Schumann displaces the pedal on A upward into the tenor voice in measure 5, so the bass line disappears altogether, along with any sense of downbeat in the accompaniment. The continuous syncopated motion is another element that drives us forward, and the absence of the bass until the very end of the phrase makes its return at the cadence all the more conclusive, even though Schumann still avoids the downbeat in the cadential measure. If we consider the structure of the entire eight-measure vocal phrase, we can see two contradictory tendencies. On the one hand, there is a straightforward parallelism between its two halves, since the four-measure grouping in measures 6—9, which is reproduced as Example 5.7a, takes the form of a period that is closely analogous to the opening grouping. The melodic contours of the antecedent and consequent are again inversions of each other, and the complementary rhythmic patterns closely imitate those of measures 2—5. On the other hand, the entire eight-measure phrase is strongly end-accented for at least two reasons. First, the antecedent—consequent relationship is more clearly defined in the second period, and so this period establishes the key of E major more strongly than the first period does A major. As shown in Example 5.5, Schumann uses two registral displacements to expand the range of the melody, which
136
Schumann's Eicheiidorff Songj
outlines a complete E major scale, and he now emphasizes 7 and 1, rather than 2 and 3. Second, he creates a sense of continuity across the entire phrase by connecting the end points of each two-measure grouping within an ascending stepwise line that consists of the upper tetrachord of the E major scale. Schumann emphasizes this rising line by beginning each successive vocal grouping with the same pitch that the preceding one ends on. The A section of "Intermezzo," like that of "Friihlingsnacht," begins by weakly establishing the tonic and ends by tonicizing the dominant, and so it is both end-accented and open-ended. In "Friihlingsnacht," Schumann simply delays the return of the tonic until the beginning of the B section, so that the A section essentially becomes an introduction and the B section begins from a point of stability. "Intermezzo" has a B section that is more typical for a ternary form, in that it has a restless and unsettled affect and, harmonically, moves even further from the tonic. As a result, Schumann must alter the opening of the song when it returns, in order to create a stronger arrival that •will resolve the tension he has built up throughout the first two sections of the song. The Recapitulation At the recapitulation of "Intermezzo," Schumann creates the strongest arrival thus far in the song and clarifies certain ambiguities of the opening, but he does so in unexpected ways. The return of the bass note A, now doubled at the octave below, on the downbeat of measure 18, is set into relief because of several aspects of the musical texture. Schumann has syncopated all of the voices of the accompaniment since the downbeat of measure 5 and has avoided the bass register since the downbeat of measure 9. And as the overall range of both the accompaniment and the vocal line has ascended higher and higher throughout the B section of the song, there has been a continuous accelerando, which ends abruptly with the ritardando in measure 17. But as Schumann goes out of his way to create a clear textural and rhythmic arrival in measure 18 and thus to give the impression that the fragile stability of the opening has been strengthened at its return, he simultaneously undermines that stability altogether in harmonic terms, as he expands, in both register and duration, the presence of the passing g1! that first appears in measure 2, turning the tonic harmony into a V7/IV.14 This chromatic pitch returns in the second half of measure 17, before the moment of recapitulation, and it now begins in the one-line octave as part of the alto voice, transfers down to the tenor in the first half of measure 18, and finally becomes part of the new bass line on the second beat of that measure, as I have shown in Example 5.6. The bass line helps to strengthen and clarify G^'s contrapuntal function as a passing tone by giving it a more prominent placement -within the texture and moving forward its resolution, Fit, from the beginning of the second measure of the phrase to the end of the first measure. If we consider the matter from the other end and compare the opening of the song to measure 18, we can now see that one reason the harmonic progres-
Weak Openings
137
EXAMPLE 5.6. Analyju of measure.) 18-21.
sion becomes elided in measure 2 is that Schumann delays the passing motion from a to fi As we hear this descending third in its proper metrical position at the recapitulation, we realize that it creates a series of parallel tenths with the vocal melody and fills in the motion between the root of the initial tonic harmony and the root of the Ft V7 chord.15 By clarifying the rhythm and the counterpoint in this way Schumann paradoxically calls attention to the lack of a consonant tonic triad at the recapitulation, which greatly intensifies the sense of harmonic ambiguity that we felt at the opening of the song. The gl| in measure 2 is brief enough to become swallowed up by the following dissonant harmony, and the expected implication of a V7/IV is immediately suppressed. But in measures 17 and 18 this chromatic pitch is too persistent and too prominent to be ignored. It forces us to interpret the opening harmony as a V7 rather than a I and makes the passing motion to FI feel like an unsuccessful attempt to deflect the desired resolution to IV. The effect of the added bass line at the recapitulation is to exaggerate the end-accented shape of the opening. Although in harmonic terms the opening tonic prolongation begins much more weakly and even ambiguously when it returns in measure 18, it also ends much more forcefully, -with an actual cadence in measure 21, which creates the first tonic arrival in the song. The harmonic progression that leads to this cadence still consists of a series of continuous seventh chords that do not resolve until the cadential tonic triad, but now we can actually hear the fifth motion in the bass, which in the opening phrase is merely implied by the counterpoint of the inner voices. The bass line also helps us to hear the rising fifths in the right hand as a coherent part of the texture that strengthens, rather than -weakens, the sense of large-scale meter. Because these fifths occur on the second beat of each measure, in the opening phrase they produce a series of cross-accents and create the impression of an independent melodic line. At the recapitulation, the rhythm of the bass line in measure 18 establishes a pattern that helps us to hear the fifths in the following two measures as part of a composite accompaniment figure that begins in the bass on the downbeat of each measure, as we can see in Example 5.6.
138
Schumann 'j Eichendorff Songd
Schumann's transformation of the second half of the phrase, following the cadence in measure 21, is more predictable. Where in measures 6—9 he uses increasingly chromatic and complex counterpoint to move away from the tonic, in measures 22—25 he stabilizes and reinforces the tonic arrival with a strongly directed harmonic progression that sets up another, even more conclusive cadence. In measure 5, Schumann introduces an et1 in the right hand of the accompaniment, an ascending chromatic passing tone that intensifies the motion toward the following ft1 in both the accompaniment and the vocal line. The emphasis on this latter pitch initiates the modulation to E major in two ways, which are illustrated in Example 5.5. Melodically it is transferred up to a higher octave and then descends to the new leading tone in measure 7, and harmonically it is the root of a ii6 chord that leads to the new V7. At the analogous place in the recapitulation, in measure 21, Schumann again introduces a chromatic passing tone that leads to fit, but in this case it is a second repetition of the gll, now restored to its original registral and metrical position. And the ft in measure 22 that resolves the passing tone appears in a new harmonic context, as the third of a IV chord in A major, a chord that not only realizes the twice-thwarted harmonic implication of the g1! but also signals closure by initiating a prototypical cadential progression. The vocal line in measures 22—25 is more closely modeled on the analogous portion of the opening phrase than the accompaniment is, but here, too, Schumann varies the melody to create a more stable and conclusive cadence. The two passages are compared in levels a and b of Example 5.7. Once more we have a four-measure period with an antecedent that ends on 7 and a consequent that resolves to 1 but now, of course, in the tonic key. Schumann inverts the melodic contour, replacing the initial descent with an ascent, and this also changes the direction of the registral transfer at the beginning of the second measure. The leading tone in measure 23 is now in the same register as the cadential tonic two measures later, and this creates a stronger feeling of resolution. Another, more significant variation involves the rhythmic shape of the consequent. In measure 8, the three-note upbeat figure at the beginning of the consequent is delayed, so that there is no impulse on the downbeat, and the first strong beat of the grouping comes in the middle of the measure. In measures 22—25, Schumann moves the upbeat ahead to the end of measure 23 and provides a more conclusive and more conventional rhythmic pattern of three successive strong beats in the final two measures of the phrase. He uses this pattern to create a formulaic 3—2—1 descent into the cadence, -which he harmonizes with an equally formulaic V^j—j—I cadential progression, as the texture solidifies into pure homophony for the first time and the syncopation, which has persisted in at least some of the voices throughout the song, finally relaxes. In other words, Schumann fulfills our expectation for closure at the end of "Intermezzo" -with the most familiar and conclusive closing gesture that he could find. But as in "Friihlingsnacht," the use of such conventional closing material paradoxically prevents him from providing the expected resolution as the final vocal phrase ends. If he -were to bring the formula to its typical conclusion, then
Weak Openings
139
a.
b. EXAMPLE 5.7. Comparison of two melodic gejtureJ.
its very predictability would create not a climax but an anticlimax. Instead, Schumann interrupts the phrase by turning the cadential measure into an overlap, in which the pianist begins the coda as the singer concludes. Schumann recomposes the opening measures of the song, but now the vocal melody becomes a fugato for the piano, and the brief moment of solidity vanishes as suddenly as it appeared. As the homophonic texture dissolves into imitative polyphony, the syncopations in the left hand return, as well as both the chromaticism and the tonic pedal, the latter now gradually displaced upward through two octaves. Schumann also returns to the higher register that the piano had touched upon in the earlier statements of the opening phrase, with a varied inversion of the theme in measures 26—27 that answers the descent to b1 by climbing up to b2. As this chromatic melodic fragment drifts up out of the contrapuntal texture, it evokes the lover's quiet song, soaring into the air. It is an open-ended musical gesture, and it is followed by a cadence that is as metrically ambiguous as the song's beginning. It is as if the music has gradually returned to the timeless state in -which it began, a state that frees the narrator and his beloved to connect with each other, unbound by the limits of physical reality. Despite their contrasting affect and imagery, "Intermezzo" and "Friihlingsnacht" are both concerned with the transcendence of the physical -world and the loss of identity that is engendered in the quest for spiritual union. Schumann's settings of these poems contrast quite sharply as -well, but he uses the same compositional idea in each of them to express their common theme. I have juxtaposed my analyses of the two songs in order to show how the recurrence of a formal strategy, such as the -weak opening, can be as significant a means of coherence -within a song cycle as the use of closely related keys and motivic relationships. To some extent, we hear relationships among the Eichendorff songs simply because Schumann is experimenting with the same aspects of musical form in each of them. But in part, the coherence of the cycle results from the nature of the -weak opening itself. In each of the two songs that I have discussed in this chapter, Schumann leads the listener to expect a climactic resolution at the end, because the opening is unstable and even ambiguous. We have seen that, in
140
Schumann',! Eicheridorff Songj
terms of harmony and voice leading, the syntactic possibilities that are available to Schumann do not enable him to fulfill the expectation that he arouses, since the very predictability of tonality's cadential formulas would turn them into moments of anticlimax. The weak opening thus leads, almost inevitably, to an open ending, and this particular use of musical structure becomes the very means by •which Schumann expresses the yearning for spiritual transcendence that is the central theme of the poems. The lack of formal closure in each song encourages us to hear coherent relationships between them, since it implies that the individual song is not a discrete piece but a part of a larger whole.
RECelftPOSITIONAL PAIRINGS
I;
n the preceding chapter, we saw that the use of the same formal strat: egy in different songs is an important means of cyclic coherence in Schumann's Eichendorff Liederkreu. In some cases, Schumann takes this idea a step further, using a single underlying structure as the basis for a pair of songs. "Mondnacht" and "Schone Fremde," for example, are both related in a general way to "Intermezzo" and "Fruhlingsnacht," because they, too, have •weak openings. But the former two songs are even more closely related to each other, because Schumann realizes the weak opening in each of them through the same structure, a variant of bar form. This relationship, which I shall refer to as a recompositional pairing, differs in degree but not in kind from the relationships that obtain among the Eichendorff songs generally. In fact, we have already seen many of the techniques that Schumann uses to pair "Mondnacht" and "Schone Fremde" in the other songs that I have discussed. Among the most significant of these are the following: (1) the vocal line and the right hand of the piano accompaniment share a composite melodic structure, but they each present the melody in different registers and conflict in terms of their phrase rhythm; (2) the melodic gestures tend to be open-ended and fragmentary, moving in unexpected directions and ending on unresolved pitches; (3) there is more emphasis on the dominant harmony than on the tonic for much of the song, and Schumann unexpectedly substitutes V7/IV for I at crucial points in the musical form; and (-4) internal formal boundaries are blurred and large-scale repetitions are disguised. I explained in the last chapter that while Schumann uses these basic tools to create weak openings in both "Intermezzo" and "Fruhlingsnacht," the particular means through which he weakens the initial tonic is different in each song, and he exploits the implications of the opening to varying ends. In the case of "Mondnacht" and "Schone Fremde," however, Schumann uses the same means in each song to create a bar form with a -weak opening. The Aufgejang, or A section, consists of a dominant prolongation, in which Schumann avoids a definitive tonic arrival through a sequential pattern that arrives at a root position tonic in a weak hypermetrical position. On a more background level, this tonic functions as a * appoggiatura harmony that resolves to a stronger dominant. Schumann delays moving toward a climactic tonic resolution until the third strophe of the 141
142
Schumann '<> E/ichendorff
Song<>
song, which is the Abguang, or B section, of the bar form, and he uses the same harmonic progression in each case to reach the final tonic. The recompositional pairing is essentially a variation technique. It does not derive from the theme with variations, in which the basic melody always remains audible, no matter how elaborate the variations become, but from the Baroque aria with variations, in which the shared material is not a recognizable theme but a bass line or harmonic progression, which may have a more subliminal presence throughout the piece. When Schumann composed "variation cycles" such as the opus 5 Impromptus <>ur une Romance de Clara Wieck and the opus 13 Etuded ttymphoniquej, it was this Baroque type of variation procedure that he favored. The former piece is based on both a melody and a bass line, and in the latter the harmonization of the theme plays as much of a role in the variations as the theme itself.1 Schumann implicitly acknowledges his debt to Baroque variation when he writes, in a critical survey of piano variations composed by his contemporaries, that the composer who "thought up the first variations" was Bach. He then takes a swipe at the virtuosic style of theme and variations that was popular in the early nineteenth century: There certainly is no genre of our art that has promoted more bungling in our day—or yet will. One can scarcely conceive of the wretchedness that has blossomed in this soil, or of the shameless vulgarity. Before, at least, there were good boring German themes, but now one must hear the most hackneyed Italian ones, in five or six watery decompositions, gulped down one after another.2 I am not the first scholar to argue that Schumann's cycles are related to variations. In fact, the earliest attempt to demonstrate that a cycle can be analyzed as a coherently unified structure, Rudolph Reti's 1951 chapter on Kiru)er
Recompojitional Pairings
1-43
Turchin also argues that Schumann uses motivic relationships to create pairings between adjacent songs. She writes that "Mondnacht" and "Schone Fremde" are "harmonically and melodically interlinked," for example, and describes "Auf einer Burg" and the second "In der Fremde" as "melodic twins." Several other scholars have pointed to the same pairing relationships, without necessarily making explicit the analogy with variations. But discussions of such pairings invariably go no further than the observation that the opening melody or the foreground harmonic progression is the same in each song. In this chapter, I will argue that these motivic similarities are but a single manifestation of a structural relationship between the songs, with which Schumann responds to an analogous relationship between the poems. As in the preceding chapter, I -will consider a range of musical parameters, including voice leading, texture, phrase rhythm, and tonal structure. Before I turn to my discussion of "Mondnacht" and "Schone Fremde," I would like to make one further observation about the history of the pairing in Schumann scholarship. Although the term appears in the literature with some frequency, it has never been clearly defined, and this has led to some confusion as to how the process actually works within Schumann's cycles. In fact, we need to consider two separate categories of pairing relationships. What I am calling a recompositional pairing is based on the use of an analogous structure in two songs. We saw in chapter 2 that Schumann pairs the first two songs of Dichterliebe in a very different way: he leaves an implication unfulfilled at the end of "Im wunderschonen Monat Mai" by concluding with a Ct V 7 chord and then momentarily resolves this implication with the F$ minor triad in the first measure of "Aus meinen Thranen spriessen." In the case of a recompositional pairing, the relationship exists on the most background level of structure, but it is not sequential, and the pairing does not necessarily depend on the order in which we hear the songs. In the latter case, which I shall refer to as a transitional pairing, the relationship does depend on the order of the songs, but it is strictly a transitional connection that exists only on the foreground. Scholars have tended to conflate these two kinds of relationships and have assumed that if one of them is present, then both are. We have already seen, for example, that in his analysis of the DichterLiebe pairing David Neumeyer assumes that the transitional connection between the songs is of structural significance and tries to construct a single background structure for the pair.5 Turchin runs into the opposite problem in her discussion of the pairing between "Mondnacht" and "Schone Fremde." She assumes that this recompositional pairing is transitional as well and that both kinds of relationships are structural in nature. She observes that in "Mondnacht" the dominant harmony, B major, "takes on greater importance than the tonic itself." She then argues that it is only in "Schone Fremde" that the significance of B major is "fully realized," since in this song it "finally achieves the status of tonic key.6 But the sequential progression that Turchin describes cannot have any meaningful significance unless -we hear the two songs as a continuous musical entity. Since "Mondnacht" and "Schone Fremde" are separate songs in different keys,
144
Schumann 'j Eichendorff Song
there is no possible frame of reference that will enable us to identify the dominant of the first song with the tonic of the second, even though they happen to be the same chord. No matter how much Schumann emphasizes B major in "Mondnacht," it is unambiguously heard as a dominant harmony that eventually resolves to E major. When B major subsequently functions as a tonic in "Schone Fremde," we hear it within a new and unrelated tonal context. One way to avoid the kinds of problems that Turchin and Neumeyer find themselves faced with is to recognize that there is a distinction between transitional and recompositional pairings and that the two kinds of relationships rarely occur together in the same pair of songs.
Songs of the Night "Mondnacht" and "Schone Fremde" are both, like "Friihlmgsnacht," songs of the night that depict Romantic moments of epiphany. These moments are characterized by the narrator's sensation that all of nature has attained consciousness and that he is merging with the nocturnal world that surrounds him. He transcends the physical boundaries of time and space and the psychological boundary of his identity, and there is some ambiguity as to whether the conscious agent who is experiencing the events depicted in these poems is the human protagonist or the natural setting. Eichendorff creates this ambiguity in the final strophe of "Friihlingsnacht" through the protagonist's description of the entire natural world expressing his own inner desire, as we have seen. In "Mondnacht" and "Schone Fremde," Eichendorff personifies the natural -world throughout the poem and at the same time delays any explicit first-person reference to the narrator until near the end, so that the reader experiences genuine confusion as to the narrative point of view. "Mondnacht" Es war, als hatt' der Himmel Die Erde still gekufit, Dafi sie im Bliithen-Schimmer Von ihm nun traumen mufit'.
It was as if the heaven Had quietly kissed the earth, So that she, in the shimmer of blossoms, Must dream only of him.
Die Luft ging durch die Felder, Die Aehren wogten sacht, Es rauschten leis die Walder, So sternklar war die Nacht.
The breeze went through the fields, The ears of corn swayed gently, The woods rustled softly, The night was starry and clear.
Und meine Seele spannte Weit ihre Fliigel aus, Flog durch die stillen Lande, Als floge sie nach Haus.
And my soul spread Its wings wide, Flew through the quiet land, As if it were flying home.7
Recompottitumal Pairings
145
The metaphorical image of the heaven kissing the earth, with which "Mondnacht" begins, is apparently the product of someone's imagination, but Eichendorff presents it in an objective tone that betrays no hint of the lyric consciousness that conceived of it.8 At the same time, the personification of the earth as a dreamer implies that the rest of the poem is a description of the earth's dream and thus a product of the earth's consciousness.9 Eichendorff never actually uses the word ich, and the only indication that there is a narrator at all is the reference to "meine Seele" at the beginning of the third strophe. Even at this moment, Eichendorff leaves the point of view somewhat ambiguous. The second strophe, which, like the first, is presented objectively, consists of a paratactic succession of images, and since the third strophe begins with the conjunction "und," we have the impression that the image of the protagonist's soul spreading its wings is the next in the series. Although it now becomes clear that the narrative voice is the protagonist's, we are still unsure whether the scene that he narrates is something he is remembering himself or part of the earth's dream. And the sense that there is no real separation between his consciousness and the force of nature that brought his consciousness to life is reinforced by the image presented in the third strophe, in which his soul moves outward and upward toward the infinitude that surrounds him. Eichendorff uses several means to suggest that the scene depicted in "Mondnacht" transcends time and space. First, as in "Fruhlingsnacht," there is an implied correspondence between the diurnal and seasonal cycles of nature, but where in the latter poem the mythical qualities of the nocturnal world are specifically associated with the birth and renewal of springtime, in "Mondnacht" Eichendorff alludes to a different season in each stanza, implying that the entire seasonal cycle is unfolding within a single night.10 This collapsing of one temporal cycle into another adds to the dreamlike quality of the scene and creates the sense that the natural passage of time has been superseded. Second, the initial image of the heaven kissing the earth represents both the convergence of the eternal and the transient as -well as the spatial union of all of nature. And this image is complemented by the final one, in which the protagonist is represented synecdochically by his soul, the aspect of his self that exists beyond time and space. There is thus a downward and inward movement over the course of the poem, from the heaven to the earth and finally to the individual. But as soon as the sequence of images arrives at the narrator's soul, it expands outward again. The movement that Eichendorff depicts is illusory in any case, since it is initiated by the heavenly kiss, and as the kiss reaches over the earth and into the human soul it suspends all movement by stopping the passage of time. The use of the subjunctive at the beginning and the end of "Mondnacht" is another important means through -which Eichendorff expresses its timeless and unreal quality.11 The opening image is presented not as a literal occurrence but as a metaphor, yet it is a metaphor that has no associational context, since Eichendorff does not reveal the actual scene to which it refers. The grammatical logic of the turn from the subjunctive to the indicative, as the second strophe begins, suggests that the string of images presented in that strophe will fill in de-
146
Schumann '<) Eickendorff
Songj
tads and provide the setting that is absent in the first strophe. But the images themselves reveal a very different logic. The heaven kissing the earth brings to mind a sunset, and the reference to the blossoms' shimmer in the third line of the poem suggests the springtime. In the scene that is depicted in the middle strophe, we have continued along the cycles of nature, bringing us into the middle of the night and the late summer. It is for this reason that we do not understand these images as an elaboration of the opening scene but as a continuation—a depiction of the earth's dream. When the subjunctive tense returns at the end of the third strophe, it now appears to refer to a particular image: the protagonist's soul flies through the quiet landscape cut if it -were flying home. In fact, however, the image of the soul's flight is as elusive and unreal as the feeling of spiritual transcendence that is metaphorically invoked in the last line. The effect of the "as if" is to further break down the sense of time by connecting the past memory within which the poem unfolds with the future expectation of death's transcendence. "Schone Fremde " Es rauschen die Wipfel und schauern, Als machten zu dieser Stund' Um die halbversunkenen Mauern Die alten Gotter die Rund'.
The treetops rustle and shudder, As if at this hour Around the half-sunken walls The ancient gods made their rounds.
Hier hmter den Myrthenbaumen In heimlich dammernder Pracht, Was sprichst du wirr wie in Traumen
Here beneath the myrtle trees In secret twilit splendor, What do you say, confusedly as if in dreams To me, fantastic night?
Zu mir, phantastische Nacht?
Es funkeln auf mich alle Sterne All of the stars twinkle at me Mit gliihendem Liebesblick, With a glowing look of love, Es redet trunken die Feme The distance speaks drunkenly Wie von kiinftigem, grofiem Gliick! — As if of great future happiness! — 1 2 Eichendorff placed "Mondnacht" and "Schone Fremde" in separate sections of his collected poems, and there is no evidence that he conceived of them as a related pair.13 They use very different images and create contrasting affects, but thematically the two poems have so much in common that they can be read as different versions of the same nocturnal moment. In "Schone Fremde," as in "Mondnacht," time and space converge and nature attains consciousness and communicates to the protagonist in a dreamlike state, breaking down the boundary that separates his conscious and physical self from the surrounding -world. Eichendorff uses the imagery of the poem to present these themes more explicitly than in "Mondnacht," yet, paradoxically, "Schone Fremde" is, if anything, even more mysterious and suggestive. The personification of the natural world
Recompoditwnal Pairings
147
is never actually depicted in "Mondnacht" but is alluded to through the metaphorical description of the earth as a dreamer. In "Schone Fremde," however, the narrator enters the poem in the second half of the second strophe by directly addressing the night and asking what it is that the night is saying to him. In the next strophe, he describes the twinkling of the stars as a "glowing look of love" and hears the distance speak to him. But as the sense that the night is conscious becomes more concrete in "Schone Fremde," the significance of what it is expressing becomes further shrouded in mystery. The night speaks "confusedly, as if in dreams," and the protagonist does not know what it is saying. In the folio-wing strophe, he describes the speech that he hears in the distance as "drunken." These epithets suggest that what the narrator hears is not coherent speech but the expression of the ineffable. The imagery of "Schone Fremde," like that of "Mondnacht," appears to unfold through time and space in a single unified arc, which in reality is just a representation of the protagonist's feeling that he is converging with nature. Eichendorff begins the poem with an image that concretely depicts the simultaneity of past and present — the ruins of an old castle — and describes the feeling that the "ancient gods" are making their rounds "at this hour," thus representing again the union of the eternal and the transient/In the succeeding strophes, the images move outward and upward, from the half-sunken -walls of the castle, to the surrounding myrtle trees, and finally to the stars. And as the images move outward through space, they also move forward in time, from the ruins of the past to the vivid immediacy of the present scene and into the future happiness of -which "the distance speaks." In fact, however, the entire poem depicts the same unchanging scene and occupies but a single moment. It is only the perspective of the narrator that expands, as his consciousness grows and becomes one with the surrounding night. And the references to the past and the future, -which Eichendorff is careful to place in the conditional tense, merely represent his transcendent glimpse of eternity. If we consider "Schone Fremde" in relation to "Mondnacht," we can see that as Eichendorff's imagery becomes more concrete and his language less ambiguous, the gap between the narrator's experience of his physical reality and his growing spiritual awareness increases, and the significance of the moment he is describing becomes more elusive. In "Schone Fremde," Eichendorff establishes a connection between the themes of nocturnal epiphany and ineffabiliry. The night not only dreams, it speaks, but -what it has to say cannot be put into words. In this way, the tension that is implicit in "Mondnacht" becomes exaggerated. Schumann brings out this relationship between the two poems by juxtaposing them as the fifth and sixth songs in his opus 39 Liederkreid and by recomposing the same unusual structure in each of them: a variant of bar form in which the repeated A section consists entirely of a dominant prolongation and the B section moves to a tonic resolution. In this way, he is able to respond, through musical form, to the tension between the floating stasis of transcendence and the ecstatic motion toward convergence that is at the heart of these poems.
148
Schumann 'j Eichendorff Song<<
"Mondnacht" Transcendence
Just as Eichendorff's poem "Mondnacht" affects a naive folklike tone that masks a dimension of profound spirituality, so, too, does Schumann's song present a deceptively simple surface that conceals underlying rhythmic, harmonic, and contrapuntal complexities. The song is based on a relatively small amount of musical material that is repeated several times, with little or no variation, and so its external form is quite clear.14 The entire A section, which consists of the piano introduction and the first strophe, unfolds over an implied dominant pedal. And since these twenty-two measures are repeated almost exactly for the second strophe, there is no real sense of harmonic motion until after the third strophe begins, roughly two-thirds of the -way into the song. The only motion that occurs in the first two strophes results from the voice leading, which creates a series of subsidiary chords within the prevailing B major harmony. Despite the enormous emphasis that B major receives, however, there is never any ambiguity about its function as the dominant of E major.15 More than anything else, it is Schumann's ability to suspend so much of the song on the dominant harmony that makes "Mondnacht" such a quintessentially Romantic work. There is a constant two-bar hypermeter throughout "Mondnacht," and most scholars who have written about the song describe the phrase structure as almost completely regular on every hierarchical level. They typically break the song down into three sixteen-measure strophes, each of which is composed of two eight-measure phrases. And yet if we compare various discussions of the song, we find that there is disagreement about even so basic a question as where the phrases begin and end.16 In fact, the apparent symmetry of the form is illusory, and as we shall see, one of the ways that Schumann creates the mysterious floating quality of "Mondnacht" is through the use of expansions and contractions in the phrase rhythm that subtly work against the regularity of the duple hypermeter. Another apparently straightforward element in "Mondnacht" is its songlike texture, -which is unusually conventional for Schumann. Unlike "Intermezzo," for example, in -which the right hand of the piano adds increasing complications that begin to take on a life of their own, the right hand in "Mondnacht" presents an unchanging texture of repeated pitches that is unmistakably accompanimental in character, leaving all of the melodic activity to the vocal line. But the simplicity of the texture — in -which the differing functions of the singer and the pianist are clearly defined and the singer's dominant role remains unchallenged — is completely at odds -with the way in which the musical material is actually disposed. The entire voice-leading structure is contained within the piano accompaniment, and virtually all of the pitches of the vocal melody are derived through fragmentary doublings of contrapuntal voices that are presented in the accompaniment in their complete form. The relationship between the vocal line and the accompaniment is thus more complex than it appears, and as in "Friihlingsnacht," Schumann exploits the implications that are created by this relationship as the song moves toward its conclusion.
Recompojitional Pairings
149
"Mondnacht" is one of only three songs in the opus 39 Litderkrei) that begin with a piano introduction. The most obvious function of this introduction, as we hear the opening of the song, is to create the mysterious feeling of infinite space converging that is so central to Eichendorff's poem. Texture and register are as important as harmony and counterpoint in this passage, and in fact, Schumann clarifies the identity of the opening chord largely through its setting. The collection of pitches in the first measure adds up to a V9 chord, but the downward arpeggiation in the right hand outlines an Ft minor triad, and the df2 on the last sixteenth, which would be a crucial chord tone m a dominant harmony, is presented as an upper neighbor. Just as important, Schumann uses the great registral gulf between the arpeggiation and the Bl in the bass to make it clear that these two entities do not create a single harmony but rather a ii chord that is unfolding over a dominant pedal.17 As the harmony progresses to V on the downbeat of measure 3, Schumann abandons the extreme registers that he began with, moving both hands to the middle of the keyboard as he thickens the texture. There is a sense of convergence even as the phrase avoids harmonic resolution, thus reflecting in the music of the introduction the poem's essence. It is in keeping with this expressive purpose that the introduction does not frame the first strophe with a stable harmony but presents a varied and foreshortened version of its contrapuntal structure and provides an explicit statement of the dominant chord that underlies the entire opening period.18 The first vocal phrase begins in measure 7, and in both the vocal line and the accompaniment measure 6 is an upbeat measure. The chromatic chord on the downbeat of measure 7 is not really the opening harmony of the phrase, however, but a product of the voice leading that is derived on a foreground level. In harmonic terms, the phrase appears to begin with the cadential dominant chord in measure 5, which is also the moment that the keyboard texture changes into an accompaniment. The boundary between the end of the introduction and the beginning of the strophe is thus blurred, and this is one means by which Schumann creates the feeling that the song is emerging out of the infinite. The piano texture maintains this otherworldly feeling as the strophe continues. The right hand has rapidly repeated pitches, but the actual melodic motion that these pitches produce is extremely slow, creating a sense of static expectancy. And Schumann uses register to evoke an ethereal atmosphere, avoiding the bass range until the cadence in measure 13 and placing the lowest voice all the way up in the one-line octave. In fact, Schumann just barely touches upon the bass note that serves as the basis for the entire structure of the strophe, placing it on the very first downbeat of the introduction in the contra octave and then as a grace note an octave higher at the cadence in measure 5. But since there is no bass line throughout the period that follows, there is an implied continuation of B1 as an underlying dominant pedal. As I have illustrated in Example 6. la, the accompaniment consists of three contrapuntal voices, each of which derives from and prolongs the dominant harmony in measure 5. Because the motion of these voices continues to the cadence in measure 13, -we hear the two vocal gestures as part of a single unbroken phrase. The b1 in the right hand splits off into two voices, one of which
150
Schumann '<> Eichendorff
Songj
EXAMPLE 6.1. Voice-leading analyjui of mecuLired 5—13.
descends to ft1, and the other of which prolongs b1 through its upper neighbor ct2. The df 1 in the left hand is prolonged through its upper third, ft1, which appears in measure 8, descends by step to e1 in measure 10, and then returns to dt1, now in the right hand, at the cadence. It is through the elaboration of this latter third span that the harmonic progression of the phrase arises. Moving from level a to level b in my example, -we can see how the harmonic progression expands as we get closer to the foreground and, in particular, how the function of the E major triad in measures 10—12 changes. On one level of structure, it is a ^ appoggiatura chord generated by the passing e1 in the left hand that resolves to V at the cadence. On a more foreground level, Schumann embellishes ft1 and e1 -with lower neighbors, creating a contrapuntal pattern that breaks up the parallel fifths between the outer voices. On this level, we hear the E major triad as a consonant tonic harmony that resolves the g in measure 6. Schumann emphasizes the ambiguous nature of this chord through its rhythmic placement. On the one hand, the e1 on the downbeat of measure 10 begins a four-measure grouping that consists of a descending bass arpeggiation, in which
Recompoditional Pairings
151
the tonic note keeps resolving downward to the dominant through three octaves. On the other hand, the same e1 completes the sequential pattern that begins in measure 7 and thus concludes a four-measure grouping. We thus hear measure 10 as an internal overlap that contracts the phrase from eight measures to seven. But the elongated upbeat that precedes the phrase, in measure 6, compensates for the contraction, and this gives the phrase its regular length. Where the piano accompaniment in the first strophe is basically grouped in four-measure units, the vocal melody is grouped in three-measure units, each of which sets a three-foot line of poetry. And yet, paradoxically, the melody is more regular than the accompaniment, despite the fact that its grouping structure appears to be at odds with the duple hypermeter of the song. The reason for this is that Schumann treats Eichendorff's prosody as an example of what is known as Langzeilenverd, in which there is a silent fourth foot implied at the end of each line.19 Schumann retains this metrical pattern in the vocal melody of his song by placing a silent measure after each three-measure grouping, extending it to a more regular length of four measures. The unstable opening gesture, much like that of "Intermezzo," establishes the tonic by suggestion rather than assertion. It remains on 6 for two measures before quickly rising to 2 through a tetrachord that tonicizes the latter pitch. In fact, the only reason that we hear the gesture in the key of E major is because the leap down to 5 at the very end of the grouping, on the third beat of measure 9, implies the tonic pitch as the next in a series of descending fifths: Ct—Ft—B— E. This implication, which carries us over the silent downbeat of measure 10, is then realized by the rhythmically free sequential continuation from b1 to e2 as the consequent grouping begins. But the overall contour of the antecedent, an incomplete neighboring motion from 6 to 5, implies further descent, and its limited range implies an expansion to a complete octave, so -we expect that ultimately the melody will descend to e1 as well. Although this implication is also realized within the consequent, it does not come in the cadential measure, where it would resolve the melody, but on the last beat of the penultimate measure. This enables the vocal line to return to 5 at the cadence, giving the melody an open-ended shape that implies a continuation into the folio-wing silent measure. If we look at the melody from a different perspective, we can explain its unusual shape by considering its relationship to the contrapuntal structure of the accompaniment. In Example 6.2, I have represented the vocal melody on the upper staff and two of the accompaniment voices on the lower staff. We can see that the vocal line, like that of "Friihlingsnacht," consists of a polyphonic melody that links the two voices. In the latter song, there is a strict metrical regularity at every level of structure, which inevitably causes the different patterns that control the motion of the melody and that of the accompaniment to come into conflict. In "Mondnacht," however, it is only on the most foreground level that there is any sense of meter at all. At higher levels, the structure is essentially static, and what little contrapuntal and harmonic motion there is does not correspond to a metrical pattern. The vocal melody does not establish a regular pattern of its own, either, but simply follows the path of the accompaniment.
152
Schumann'^ Eichendorff' Song<)
EXAMPLE 6.2. Relationship between vocal melody and countrapuntal structure in nuajiiru 6-13.
The initial dominant harmony of the strophe comes before the singer enters, so the melody, again like that of "Friihlingsnacht," begins in the middle of the contrapuntal structure, doubling the neighboring tone cS2 in the uppermost voic of the right hand. It is the extremely slow motion of this voice that causes the vocal line to sit on the neighboring note for so long and then to keep circling back to the note of resolution, b1, all the way into the cadential measure. The ascents to ft 2 and e2 create a link between this voice and the first two pitches of the descending third span in the left hand, displaced upward in register and delayed by a measure. The downward arpeggiation in measures 11 and 12 can be understood as an octave transfer that brings the vocal line into the same register as the accompaniment. And the fact that the melody is doubling the third span explains the early arrival of e1 in measure 12, since this pitch is functioning contrapuntally as a passing tone that resolves to di1 in measure 13. The resolution comes only in the accompaniment, however, and not in the vocal melody, and this divergence between the two parts ultimately creates a conflict between them, not in terms of the hypermeter, -which remains undisturbed in both, but in terms of the phrase rhythm. Where the upward leap to 5 in the vocal melody creates the sense that the phrase continues into measure \4, the dominant arrival in the accompaniment creates a half-cadence in measure 13. In the piano part, measure 14 is analogous to measure 6, an elongated upbeat to the following phrase, which Schumann emphasizes by anticipating eji1 in the left hand. In both parts, then, the phrase rhythm is organized in terms of four-measure groupings, but the groupings do not coincide. The melody of the strophe is completely regular, with the groupings consisting of measures 7—10 and 11 — 1-4. The accompaniment begins -with the same grouping but treats measure 10 as an overlap, so that its second grouping consists of measures 10—13. Convergence
In the foregoing section, I have explained how, in the A section of "Mondnacht," Schumann evokes in music the still, ethereal atmosphere of Eichendorff's poem. As in "Intermezzo" and "Friihlingsnacht," Schumann expresses the affect of this song through the use of a weak opening, a technique that sets up particular formal expectations. The end-accented structure of "Mondnacht's" A section is not
Recompojitianal Pairings
153
so clearly pronounced as in the weak openings of the former songs, since it consists of an open-ended fragment that is repeated, rather than a series of increasingly strong arrivals, and it ends with a half-cadence, rather than a tonicization of the dominant. We can more accurately describe the way in •which this weak opening works by observing that the static quality of the A section arouses the expectation of more forceful motion in the B section and that because of the lack of any cadential resolution -we cannot help but expect a more conclusive cadence at the end of the song. Schumann intensifies the expectation that his opening arouses by setting "Mondnacht" as a bar form, rather than a ternary ABA form. The second strophe, in measures 23—43, does not present contrasting material or increase either the motion or stability but simply continues to repeat the same static phrase, which is sung four consecutive times. It is not until the beginning of the third strophe — the moment in the poem when the narrator first enters and explicitly describes his feeling of epiphany — that Schumann starts to realize the implications he has set up with his weak opening. Although this strophe contrasts sharply with the first two, it does so primarily by transforming material that we have already heard and presents surprisingly little that is new. In the first half, in measures 45—52, the accompaniment includes a transposed paraphrase of the introduction, and while the vocal melody is different from that of the A section, it is modeled on the opening phrase in terms of both rhythm and melodic contour. The second half of the strophe, in measures 53—61, consists entirely of a varied return of the opening phrase. We now hear this phrase as a consequent to the preceding phrase, as Charles Rosen points out, and it is through this changed context that Schumann clarifies its function. When the phrase first appears, at the beginning of the song, it feels like the middle of something that is ongoing, with no clear beginning or end. At its return halfway through the third strophe, the preceding phrase provides it with a stable beginning point and its ending is varied in such a way that there is some degree of closure. At the same time, in keeping with the conditional ending of the poem, Schumann uses several means to weaken the stability of the strophe and create an open ending. The strongest element of dynamic contrast as the third strophe begins is its clear and regular movement. A sudden surge of momentum already comes at the end of the cadential measure of the previous phrase, measure 43, as Schumann introduces sixteenth notes in the left hand for the first time in the song and brings the melodic line in that hand into the small octave, where it becomes a real bass line. As shown in Example 6.3, there is a progression of descending fifths in the bass, which begins with the cadential B in measure 43 and continues to the Gt at the vocal cadence in measure 59. And since each succeeding fifth arrives in the third measure of a grouping, the phrase rhythm of the accompaniment is now coordinated with that of the vocal line. Where in the A section the melody appears to grow out of the contrapuntal structure of the inner voices, in the B section it has a harmonic foundation that gives it a clear sense of shape and direction. Each time the opening vocal phrase begins during the first two strophes, for example, the lowest voice of the accompaniment has a chromatic ejt1.
154
Schumann i Eichendorff Songd
EXAMPLE 6.3. AnaLyjui of mecuurej 40-59.
which, as we have seen, is a contrapuntal elaboration of the underlying dominant that delays the arrival of ft1, the fifth of the chord, until the second measure of the phrase. When the phrase returns in measure 53, however, not only does the ft1 arrive at the same time, but it is doubled at the octave below by the bass line. The f I in the bass is part of a stepwise passing motion that fills in the descending fifth from a, at the end of the preceding phrase, to df1, in measure 55, as I have shown in Example 6.3. But it also becomes the root of a ii chord that has a clearly defined harmonic function. The ei1 has been relegated to a middle voice by the expanding piano texture and moved back to the previous measure, -where it becomes part of a chromatic passing harmony that connects the two phrases.21 The vocal melody itself has a coherent shape in the third strophe that corresponds to the harmonic shape of its accompaniment. There are now two complementary phrases that balance each other as antecedent and consequent. The antecedent begins in measure 45, with 5, -which initiates two contrapuntal lines, represented with the beams in Example 6.3. One line continues in the first grouping with a descent to gtt . This descent implies a continuation down to e , but it is interrupted in the next grouping by the second melodic strand, an upward motion to cl1 that provides the starting point of the 6—5 neighboring tone figure. The neighboring tone resolves in the first half of the consequent, and the descending line is picked up again -with the gt1 in measure 58. In the A section of the song, as we have seen, the descent from gt1 to e1 comes in the penultimate measure of the phrase because the latter note is not a resolution but a passing tone. At the end of the third strophe, Schumann uses a hemiola to delay the arrival of e1 until the cadential measure, so that it is coordinated -with the arrival of
Recompotfitional Pairings
155
the next descending fifth in the bass. And so the 6—5 neighboring figure and the motion from fit2 to e2, which create such a static quality when we first hear this phrase, are now framed by the descent from 5 to 1, giving the entire period a sense of balance, of clearly directed motion, and of resolution. We have thus far seen how Schumann realizes some of the implications of "Mondnacht's" weak opening by turning the initial vocal phrase into a consequent -when it returns at the end of the song and thus providing a new and clearer context for it. But even as he builds up a sense of momentum, provides a stable starting point for the opening, and moves toward a tonic resolution in the Abgejang of his form, Schumann still maintains the transcendent quality that has hovered over the first two strophes and creates an ending that is at once satisfying and yet suggestively open-ended. In part this results from the unusual nature of the antecedent phrase with which Schumann begins the third strophe. For even though this one phrase provides the only contrasting music in the entire song, Schumann somehow manages to make it feel like a moment of return as well, conflating formal functions that are normally opposed. The vocal gesture with which the strophe begins in measure 45 is a more stable version of the gesture that we have already heard in measures 7—9. As I have shown in Example 6.4, it has the same rhythmic shape, and we hear its melodic contour as a variation in which the upper neighbor, ctt2, is replaced by its preparation, b1, and the initial ascent is replaced by a descent. The underlying harmony is still V and the right hand of the piano still has a dominant pedal on top, which, as in the vocal line, is no longer encumbered by its upper neighbor. The bass line in measures 45—47 is really no more than a varied expansion of the descending third that appears in the left hand in measures 7—9. And, of course, the fact that the accompaniment presents a transposed paraphrase of the introduction only adds to the sense that we are hearing a return of the opening. This vaguely recapitulatory moment feels unsettling for several reasons. First, it is the only moment of contrast and dynamic motion after forty-three measures of static repetition, and its familiarity makes us wonder if we are really hearing anything new at all. Second, although this phrase sounds familiar, -we have never actually heard it before, and so it is difficult to discern what it is that has returned. Third, despite the regular and coordinated harmonic motion of the
a.
b. EXAMPLE 6.4. Comparison of two melodic gestured.
156
Schumann '
accompaniment in this phrase, there is still no sense of harmonic stability or resolution. As the phrase begins, the harmony is poised on a V7 chord in second inversion, and as this chord finally moves on in measure 47, Schumann substitutes V7/IV for the expected tonic resolution, an effect that he emphasizes by setting this chord with a transposed fragment of the atmospheric piano introduction. Even more significant is the fact that the series of descending fifths in the third strophe is an open-ended bass progression, and Schumann uses it to create a harmonic spiral in which the end of the phrase returns to the harmonies with which the progression begins, but now in first inversion, as shown in Example 6.3. Despite the apparent coordination between melody and accompaniment, then, they are really at odds, since the melody has a closed shape and the accompaniment implies further continuation. And since in the cadential measure Schumann again substitutes V7/IV for the tonic, he does delay the resolution in the accompaniment until after the phrase has ended and the singer has finished singing, just as in "Friihlingsnacht." In the latter song, the ultimate resolution takes place over a tonic pedal, as we saw in the last chapter, creating an authentic cadence whose tonic is drawn out over three measures. In "Mondnacht," however, the deflection of the cadential tonic in measure 59 leads to a root position IV chord, transforming what would have been an authentic cadence into a weaker plagal cadence. In both songs, the effect of the open ending results as much from the disjunction between the vocal melody and the accompaniment as it does from the weakening of the harmonic progression. Schumann sets up a rhythmic conflict between the two parts at the beginning of each song and then resolves the conflict in the third strophe, as he builds up momentum toward the final cadence. The fact that the singer and the pianist then fail to cadence together diffuses the energy and undermines the resolution. Again as in "Friihlingsnacht," Schumann uses the open ending in "Mondnacht" to express the narrator's elusive feeling that he is converging 'with the surrounding night. Eichendorff represents this in the poem through the image of the soul spreading its wings and flying through the quiet landscape. In contrast -with the gentle motion of the wind in the second strophe, which is presented through a series of images that are each contained within a single line, the movement of the human soul is clearly directed toward its goal and fills up the third strophe in its entirety. Each line in the second strophe is a self-contained grammatical unit, but as Eichendorff returns to the theme of transcendence with which the poem begins, he reflects the expansiveness of the theme by adopting the more continuous syntax of the first strophe, in -which there is an enjambment in the middle of each couplet. Schumann evokes the feeling of convergence as the third strophe begins by bringing the vocal melody into sync with its accompaniment and responds to its continuousness by increasing the sense of forward motion and turning the entire strophe into a single unbroken phrase. Eichendorff's turn to the subjunctive tense at the very end of the poem reveals that the narrator's moment of epiphany is transitory, however, and his allusion to the soul's return home makes it clear that, -within his earthly life, the narrator cannot achieve the
Recompojitional Pairings
157
convergence for which he yearns. By ending his song with an open ending Schu22 mann captures the complex emotions of the strophe with great sensitivity.
"Schone Fremde" Confusion in the Night By pairing "Schone Fremde" with "Mondnacht" in the Eichendorff Liederkreui, Schumann presents two complementary reactions to the nocturnal moment of epiphany that both poems depict. Where "Mondnacht" is suffused with a sense of inner peacefulness, "Schone Fremde" vibrates with ecstatic expectancy. The different affect in the latter song is in part the result of stylistic elements, such as the increase in harmonic rhythm and the continuous syncopations in the accompaniment, as Barbara Turchin has observed.23 But the contrast also results from the different way that Schumann realizes the underlying bar form that the two songs share. In "Mondnacht" almost the entire form is built up from the same small amount of material, -which is repeated with little or no variation, but in "Schone Fremde" Schumann completely disguises the repetition of the A section for the second strophe of the song, which contributes in large measure to the character of confused excitement. And where the A section of "Mondnacht" is suspended on a dominant harmony that is first stated before the singer enters and is still sounding at the half-cadence with 'which it concludes, "Schone Fremde" begins with a chord that has no clear harmonic function and the tonality remains ambiguous for most of the opening phrase. When Schumann finally clarifies the tonality at the cadence, he does so by tonicizing Fit major, the dominant chord, and it is not until we are in the middle of the third strophe, the Abgejang of the form, that we realize that the song is, in fact, in the key of B major. If the music of "Mondnacht" gradually emerges out of the infinite, then "Schone Fremde" begins abruptly in the middle of something that is already ongoing. And yet it is interesting that in terms of melodic contour, rhythm, and harmonization the opening vocal phrase is almost identical in both songs, as I have illustrated in Example 6.5 by transposing the melody of "Mondnacht" to B major.24 The different effect that this phrase produces in each song is largely the result of the different context within -which -we hear it. In "Mondnacht," as -we have seen, the phrase is preceded by a piano introduction that prolongs the dominant harmony and the cadential dominant is left hanging. "Schone Fremde" begins with a single introductory measure that contains a seemingly inexplicable harmony — DJ minor—and the vocal phrase ends with a two-measure expansion that tonicizes the dominant. There is thus a sense of motion, but it is a motion with no clear beginning. Another significant difference is that the vocal melody in "Mondnacht" emerges out of an accompanimental texture, but in "Schone Fremde" it begins as a sequential continuation of the motive that appears in the right hand of the piano in the upbeat measure, and this motive provides an apparent starting point for the sequence of descending fifths upon which the vocal melody is based.
158
Schumann 'd Eichendarff
Songj
a.
b. EXAMPLE 6.5. Companion of the opening vocal phrcuej in "Mondnacht" and "Schone Fremde."
In "Schone Fremde," as in "Mondnacht," the sequence is rhythmically free, but in this case it is because Schumann distorts the meter by delaying each unit of the sequence to a successively later part of the measure, as I have illustrated in Example 6.6. On level a of the example I have represented the basic sequence that is implied by the opening measure and on level b the rhythmic variation of that sequence that is implied by the composite melody of the vocal part and the accompaniment in the first two and a half measures. We can see that the delay of ct2 from the middle of measure 2 to the downbeat of measure 3 alters but does not destroy the sequential pattern. Level c, which is a reduction of the composite melody as it actually appears in the song, reveals that the sequential motion is broken only when Schumann replaces the expected leap to b1 in measure 3 with a return to ct2. When the leap from ft 1 to b1 finally comes in the first half of measure 4, it now implies a cadential close in Ft major, which Schumann realizes by filling in the leap with a stepwise descent that brings us back to ft 1 on the downbeat of measure 5. If we once more compare this phrase with the openended melodic phrase of "Mondnacht," we can see that there is a decisive difference in the penultimate measure, which in this case creates a strong sense of melodic closure through the descending line. Schumann's harmonization of the phrase calls the tonality into question, however, since the harmonic motion from measure 4 to measure 5, very much like the analogous point in "Mondnacht," suggests a tonic * that resolves to V in B major, rather than a plagal progression in Ft major. There is a conflict between the tonal implications of the melody and of the harmony, and it is in part to resolve this conflict that Schumann extends the phrase for two more measures, concluding with an unambiguous authentic cadence in Ff major in measure 7. Of course, another reason that Schumann extends the first phrase is because he needs more music for the second half of the strophe. The first two lines of text are each set with the typical two-measure grouping that we find in most of the Eichendorff songs, and if he were to continue in this way, Schumann would need to double the length of the phrase, adding another four measures. Interestingly, he adds only two more measures, and so he must squeeze in the two lines of poetry by abruptly increasing the rhythmic activity of the vocal line. There are other moments in the cycle -where Schumann expands a musical
Recompojitional Pairings
159
a.
b.
c. EXAMPLE 6.6. AnalydL> of the nulodic sequence in meiuurej 1-5.
grouping so that a line of text is set by more than two measures, but this is the single instance in which the grouping contracts and the setting of the text must speed up. One recent commentator has criticized this moment of the song and complained of "the uneasy feeling that the plethora of syllables in these measures necessitated the 'poco rit.' in measure 5, and that the overall rhythmic effect of the phrase just escapes being cumbersome."25 I find, on the contrary, that the crowding of the text beautifully captures the feeling of barely controlled excitement that is at the heart of Eichendorff's poem.26 It also enables Schumann to exaggerate the contrast between the contraction at the end of this strophe and the expansion at the beginning of the next, -which results not only from the slower and more lyrical declamation but also from the -wider range, the larger melodic contour, and the slower harmonic rhythm. In fact, the opening of the second strophe turns out to be an expanded version of the same sequence that begins the first strophe, starting a minor third higher -with the fifth between cjp and ft2. We again hear the accompaniment's statement of the basic motive in the upbeat measure, but as the vocal melody enters, Schumann augments the rhythm of the pattern even more drastically, by delaying the upward fifth for a full measure, so that each unit of the sequence occupies two measures rather than one. When -we reach the third unit of the sequence, in measure 11, returning to the point where the first strophe begins, Schumann contracts the pattern, so that it is once more moving by measure, as it does -when we first hear it at the opening of the song. In Example 6.7, I have juxtaposed the vocal melody of the strophe with the sequential pattern upon -which it elaborates, which is partially presented in the accompaniment. We now realize that "Schone Fremde" really does begin in medias res, and we understand why it is that the harmonic function of the opening chord and the tonality of the opening strophe are so ambiguous. The Dt minor chord in measure 1 is not the starting point of a sequence but the midpoint, and the entire first
160
Schumann '<> Eichendorff Song
EXAMPLE 6.7. Vocal melody in measures 8—13, with ihi sequential model.
strophe is just a fragment of the complete pattern, -which begins and ends with an Ft major harmony, and is clearly in this key. The fact that the first strophe begins in the middle of the pattern also explains why Schumann has such difficulty squeezing in all of the text. He has elided the first half of the phrase, which in its entirety consists of eight measures, and when he presents the complete phrase in the second strophe he can easily maintain the proportion of two measures of music for each line of poetry without resorting to a phrase expansion.27 If we consider the counterpoint of these first two strophes, we can see that the sequence is a polyphonic melody that connects the upper two voices of a three-voice 5^6—5 pattern, which I have represented in level a of Example 6.8. The pattern consists of a stepwise descending line in the soprano that fills in the fifth between ct2 and ft1 over the course of four and a half measures, as the middle voice creates a series of suspensions and the bass moves down sequentially from ft to B.28 Schumann concludes the pattern by extending the ft1 in the upper voice, as the bass leaps down to Ft and the middle voice continues its descent to at, so that in its complete form it prolongs an Ft major harmony. The basic pattern, as I have represented it in level a, never actually appears in the song, because Schumann alters it in both strophes through rhythmic expansions and the addition of chromatic inflections. In levels b and c of Example 6.8, I have represented the rhythmic alterations of the pattern, as it appears in each strophe, through the Arabic numbers above the upper staff, which correspond to the numbered measures of the basic pattern in levela. If we compare level a with level b, which is the first strophe of the song, it becomes clear that the delayed arrival of b in measure 4 results from an augmentation of the pattern. ^Ve can see on level c that Schumann retains this same augmentation in the second strophe, in measures 13 and 14, and also expands the first two and a half measures of the pattern in a more unusual way. He first approaches the e and the dt in the bass by step, as in the basic form of the pattern, and then repeats the approach to each of these notes in the following measure, now replacing the step-
a.
b.
c
EXAMPLE 6.8. Contrapuntal analyju of the first two strophes.
162
Schumann j Eichendorff Songj
wise motion with an ascending fourth. The second approach to dt, the note with which Schumann first picks up the pattern at the beginning of the song, is elaborately embellished, as I have shown in the added measures below level c. The At on the downbeat is transferred to gf a seventh above, and the arrival of dt is delayed until the last sixteenth of the measure by an imitation, in diminution, of the vocal line's descending third. The bass line in measures 11 and 12 thus becomes a continuous motion from the At on the downbeat of measure 11 to the At on the downbeat of measure 13, and Schumann completely obscures the recapitulation of the opening strophe, which begins in measure 11. The disorientation that this moment causes in the listener corresponds to the sensation that Eichendorff's narrator describes as the night appears to speak to him "confusedly, as if in dreams." We have a vague sense that we know where we are and that it is someplace familiar, but we cannot be sure why, since it is almost impossible to tell that -we are actually hearing the return of the song's opening.29 As the strophe ends, Schumann replaces the curious cadential measures of the first phrase with a clear progression to Ft major. The B major harmony in measure 14 does not become a Jj chord, so there is no implication that the Ft major chord in the next measure is a dominant, and the bass now descends determinedly to Ft on the cadential downbeat. Since he has reached the end of the poetic strophe at the cadence, Schumann can replace the breathless, almost unmeasured rush of notes that we hear in measures 6 and 7 with a single ft1, which is held out for more than a measure. Perhaps more important, the cadential Ft major chord in measure 15 is a return to the opening harmony of the phrase, and the ft1 in the vocal melody concludes an octave descent that begins with the upward arpeggiation in measure 7. But at the same time, there is something slightly unsettling about this moment of resolution. Despite the initial harmonic ambivalence at the end of the first strophe, we do finally get a Ct V 7 chord in measure 6 that resolves to Ft major through descending fifth motion in the bass. In measure 14 this harmony is replaced by a mysterious-sounding diminished-seventh chord, -whose progression to Ft major is more evocative than resolute, and the plagal motion from B to Ft in the bass does not really provide closure but implies a further descent to Bl an octave below. Ineffable Clarity Throughout the first two strophes of "Schone Fremde," the static prolongation of an Ft major harmony, on the background level, is disguised by the rapid and erratic harmonic rhythm of the foreground. Beginning in measure 15, Schumann pauses on the cadential Ft major chord for four measures, overlapping the end of the second strophe and the beginning of the third. This moment is reminiscent of the A section of "Mondnacht"; the surface rhythm moves rapidly, but there are repeated pitches in most of the voices, and so there is no actual motion. As in the earlier song, Schumann changes direction and introduces the first true harmonic progression as the third strophe begins, in this case by slowing down the harmonic rhythm to a steady rate of one harmony for each two-measure
Recompoditional Pawingj
163
grouping. He establishes the transition from the end of the second strophe by the most minimal of means, a chromatic descent from ft1 through ejt1 to e1, -which turns the Ft major harmony into a V7, and somewhat unexpectedly directs the song toward B major, -which is finally established as the tonic at the very end of the third strophe. I have illustrated the harmonic structure of the strophe in Example 6.9, a durational reduction in -which each two-measure grouping is reduced to a single measure in common time. If we compare the two levels of this example, we can see that the continuation of Fit in the bass in measure 19 is a suspension that delays the arrival of Bl until after the change of harmony. One effect of this suspension is to prevent us from hearing measure 19 as the closural bass motion that is implied by the cadence in measure 15. A second effect is that the bass descent to Bl is now coordinated with the climactic melodic motion to ft2, and in fact, this climax directly results from a voice exchange between the bass and the melody, as is clear from level b. The vocal melody of this strophe, again as in "Mondnacht, " contrasts with the earlier two strophes in that it has a clear shape to it, -with two balanced halves that create a periodic phrase. We can see the overall contour of the melody in level b of my example: in the first four measures a gradual arpeggiated motion from ft 1 to ft2 and in the second half a swift descent from gt2 to b1. By the time ft 2 arrives, in measure 20, the harmony has already changed, and so we do not hear this arrival as the completion of an arpeggiation of an Ft harmony. In fact, the upward motion in these four measures occurs in
EXAMPLE 6.9. Durationalreduction of ineajurej 17—24.
164
Schumann '<) Eichendorff Songtt
two stages. In the first two-measure grouping, Schumann builds up tension by sitting on ft1 for an entire measure and then releases the tension by quickly ascending to at1 and clt2. In the second grouping, he begins from b1, a pitch that is outside the chord, and leaps up to fl2. The melody ascends through sequential fifth motion, reinforcing the harmonic fifth motion, and helps to build up momentum toward the coming resolution at the cadence in measure 24. The melody and the harmony in measures 17—20 create a conventional progression that is often used in tonal music to set up a closing gesture. And this is, in fact, how Schumann uses it, resolving the climax in the second half of the phrase with an equally typical cadential formula, which creates a powerful and satisfying conclusion. But it is precisely the conventionality of the gesture and the rhetorical finality with which it concludes the song that make the ending of "Schone Fremde" stand out within the cycle. None of the other Eichendorff songs ends so conclusively, and there are only three or four others in which the final cadence is not undermined in one -way or another. Furthermore, "Schone Fremde" is the only song with a weak opening that does not have an open ending as well. We have seen in other cases how the use of such an opening, in and of itself, forces Schumann to open up the ending, by creating implications that are so powerful that they render conventional closing formulas anticlimactic. How is it, then, that in this song Schumann is able to create a satisfying close with such a formula? I believe that the answer lies, at least in part, in the extreme nature of its opening. In each of the other songs that I have discussed in these two chapters, Schumann has weakened the tonic but has not left us in any doubt as to its identity, at least not for very long. But in "Schone Fremde" we have no idea that B major is the tonic until we get to the third strophe, and it is the final cadence that establishes the tonic at the same moment that it provides closure for the song. In other words, because we have not been expecting a strong authentic cadence in B major throughout the song, it is possible for the cadence to create a climax when it does come. For all intents and purposes, "Schone Fremde" is a song that begins, confusedly, in one key and ends, far more clearly, in another, and it is for this reason that Schumann does not have to undermine the conclusiveness of its ending. I •would like to conclude my discussion of "JVLondnacht" and "Schone Fremde" by returning to the idea that Schumann creates two different responses to the moment of epiphany that their texts depict, as he recomposes the same underlying structure in each of them. In the former song, as we have seen, he evokes the affect of ethereal stillness by suspending the harmony on the dominant throughout the Aufg&tang of the form and constructing the vocal melody out of a pair of open-ended gestures that are sung four times in succession. The use of elongated upbeats and overlaps in the accompaniment and of conflicts between the accompaniment and the vocal line prevents us from feeling any forward motion, but these asymmetries are obscured by the superficial regularity of the phrase rhythm, and so they do not disturb the peaceful mood of the song. When, in the AbgMiing, Schumann finally begins to move toward resolution through a regular harmonic progression, he still maintains a sense of timelessness by turning the
RecompoditionalPairingj
165
one moment of contrast into a disguised return of the opening and preventing the final cadence from achieving true closure. And so even as the poetic text describes the seemingly definitive convergence of the soul with the surrounding night, Schumann expresses the transcendental and incorporeal nature of the moment by opening up the ending of his musical form. In "Schone Fremde," however, Schumann evokes a sense of uncontrollable excitement by creating constant musical motion that has no clearly defined beginning or end point. We can describe the opening of this song as more extreme than that of "Mondnacht," because the dominant harmony that is prolonged in the Aufgedang is not even stated until halfway through the first strophe, and when it finally appears its harmonic function is left undefined until the Abge<>ang begins. But if the opening is thus weaker than that of "Mondnacht," then the ending is that much stronger, since Schumann again creates forward motion through the same harmonic progression but now brings it to its expected conclusion, resolving the climax of the song with a clearly articulated close. The difference in Schumann's setting of "Schone Fremde" is not only a reflection of its contrasting affect but also a response to its particular presentation of the theme of convergence. Where in "Mondnacht" Eichendorff expresses this theme through the ambiguous sense that nature has become conscious and the narrator's equivocal feeling that his soul is taking flight, in the final strophe of "Schone Fremde" there is no ambiguity or equivocation: the stars are looking at the narrator, and the distance speaks to him. And yet, paradoxically, as his awareness of the night's consciousness becomes more concrete and absolute his understanding of its message diminishes, and as the song ends we still have no idea what it is that the night has said. It is this gap between apparent certainty and actual obscurity that Schumann captures through his unusual manipulation of musical form in "Schone Fremde." The ending of the song is a closural gesture of great clarity and power, but it is a gesture that stands alone, with no analogous opening gesture to complete and enclose the form. It is much like the ruin that provides the setting for the poem, a wall that is strong and enduring but is not attached to an integral structure that would make its purpose clear. The cadence provides the potential for clarity more than it does clarity in and of itself, and in this sense it is a symbol for the message that the night imparts, a message that is absolutely clear in some time scale—perhaps in the past, -when the ancient gods really did patrol these walls, or in the future of which the distance speaks —but at the present moment is drunken and dreamlike and thus incoherent.
Conclusion I have presented my discussion of Schumann's Eichendorff Liederkreu as a series of five individual song analyses in order to suggest a new way of understanding the notion of coherence in the Romantic song cycle. It is not simply the use of closely related keys or similar melodic motives that associates the songs I have been discussing in the last three chapters but the exploration of the same set of compositional ideas and the use of the same formal strategies. My conception of
166
Schumann '
cyclic coherence is certainly not as definitive as the voice-leading structure that Arthur Komar describes in Dichterliebe or the motivic web that other scholars have observed in the Eichendorff Liederkreu. First, it is more relevant to some of the songs in the cycle than it is to others. "Mondnacht" and "Schone Fremde," for example, share virtually the same underlying form, a form that involves compositional ideas present in several other Eichendorff songs as well. However, while "In der Fremde" exhibits a number of the unusual formal elements that are found in the four other songs I have analyzed, it lacks some of the most significant, including the weak opening, -which I have identified as being especially characteristic of the cycle. Second, Schumann explores virtually all of the compositional ideas that I have described in the Eichendorff songs in other works as well, and so it is difficult to draw a clear line between the coherence of this particular cycle and and his overall compositional style. Although Schumann may be concerned with the same general questions about musical form in many of his song cycles, he often tends to focus on particular solutions in different ones, and this is one way in -which he imparts to each of them a particular stylistic and expressive character. We have seen, for example, that he uses several related techniques in the Eichendorff songs in order to create -what I have called an open ending, in which technically there is harmonic, melodic, and contrapuntal closure, but rhetorically the song is still open-ended. In the Dichterliebe songs, which Schumann began -within a couple of days after completing the Eichendorff Liederkreu, he continued to experiment -with song endings. But while there are some examples of open endings in Dichterliebe that are similar to those I have described in the Eichendorff songs, they are far more infrequent. What is extraordinary about the endings of many of the Dichterliebe songs is their extensive piano postludes, -which may be half again as long as the song itself and often introduce new musical material. These postludes fundamentally alter the structure and proportions of the songs and, in terms of expression, turn them into dialogues between voice and piano. In the Eichendorff Liederkrei), Schumann composes more conventional postludes, which never last for more than a phrase. And, as we have seen, these postludes appear to be motivated primarily by his desire to create an open ending, since they typically complement the vocal cadence in one of two antithetical ways, either providing resolution if it is lacking, as in "Friihlingsnacht" and "Mondnacht," or else undermining it if it is present, as in "In der Fremde" and "Intermezzo."30 We thus see an example of Schumann's interest in a single compositional problem in both cycles — how to find new, more expressive ways to end his songs — and his use of a different solution in each of them, the open ending in one case and the extended postlude in the other, which gives each cycle its individual character. Schumann's process of creating his song cycles, which is the subject of the final part of this book, provides further support for the idea that each cycle derives its coherence and distinctiveness in large part from his exploration of a small set of compositional ideas. The source evidence that survives suggests that he typically began to consider the arrangement of the songs only after he finished composing them, and so it seems unlikely that he had a unifying structure
Recompoditional Pairings
167
in mind from the start. However, he usually composed the songs of a cycle very rapidly, over a brief period of time and selected his texts from a single volume of poetry, written by the same poet. It seems almost inevitable that a group of songs composed in this •way would share the same stylistic and formal characteristics and that these characteristics would derive to some degree from the formal and thematic similarities of the poems being set. In his Dichterliebe songs, for example, Schumann uses the extensive piano postludes to respond to Heine's Romantic irony and to express the gap between the sense of the narrator's words and their underlying meaning. And I have argued throughout the last three chapters that the use of the open ending enabled Schumann to respond to the conditional and inconclusive way in which Eichendorff tends to close his poems and to convey the yearning for transcendence that is an important thematic element in so many of them. Because Schumann uses structural elements within each song to respond to the poetic texts that he sets and to create coherence within his cycles, we cannot explain these processes by focusing exclusively on the structure of the whole cycle. Rather, we need to engage in the detailed study and analysis of the individual song.
This page intentionally left blank
Part III From Songs into Cycles
This page intentionally left blank
SCHI/IVIANN'S PROCESS OF i COMPOSING A CYCLE
T!
Eichendorff Lifderkreui provides us with an especially good opportunity to reevaluate the way in which we conceive Schumann's song cycles, because there are two related circumstances concerning its genesis that have challenged our traditional assumptions about the genre. First, the poems that Schumann set •were selected by Clara Wieck from Eichendorff s collected edition and were never intended by the poet as a cycle. Second, the only complete set of manuscript sources that survives for the songs reveals that the order in which Schumann composed them bears virtually no relationship to the order in which they stand in the published opus. Our conventional understanding of the song cycle carries such powerful "implications of order and interdependence," as Patrick McCreless puts it, that "the omission of any of the songs, or the rearrangement of their order, constitutes a threat to or negation of its cyclic character."1 It is this premise that led to the formulation of Arthur Komar's theory of musical structure in Dichterliebe. It has also led to the expectation that the poetic texts of a song cycle should create a straightforward narrative with a clear beginning and end. Obviously, the genesis of opus 39 raises serious questions about this premise, and it is with these questions that I am concerned. In this chapter, I will consider what the surviving manuscript evidence reveals about Schumann's process of arranging songs into cycles, and then, in the following chapter, I will consider how Schumann created the texts of his cycles and, in particular, how he turned the poems of the opus 39 Liederkreid into a lyric cycle as he composed the work. The traditional division of Schumann's publications of songs into two categories— cycles and collections—presupposes a corresponding distinction between two compositional processes. In the case of Schumann's collections, he simply composed a number of songs and then arranged them into a particular order "when he was ready to publish them. Our belief that order is of crucial importance in Schumann's cycles, however, has led to the assumption that in these works he conceived of the arrangement as he composed the songs. The problem that scholars have faced in the case of the opus 39 Liederkreu is that in composing it Schumann apparently followed the procedure that is characteristic of his collections. The paradigmatic cycles -with which the Eichendorff LiederkreM is 171
172
From Song
usually compared are the Heine Liederkre'u, Frauenlie.be and Leben, and Dichterliebe. There is compelling evidence in the case of the first two of these cycles that Schumann did have a particular order in mind -when he began to compose them and wrote the songs in this order. But these are the only two sets of songs that Schumann composed in this way. If we consider all of Schumann's song publications that were composed in 1840—41, we can see that it is more typical for him to compose the songs first and then decide on an order afterward, even in the case of publications that he specifically designated as cycles. I will attempt to reconstruct the compositional history of the opus 39 Liederkreu in this chapter by comparing the manuscript sources that survive for the cycle with the more extensive source material that is extant for his other song publications. The complete body of Schumann's song manuscripts reveals that the process he employed in composing the Eichendorff Liederkreu is typical, both for his cycles and his collections. The source material thus provides further evidence that Schumann did not clearly distinguish between these two types of publications.
The Berlin Notebooks Schumann's self-proclaimed Liederjahr began with the composition of a Shakespeare song in German translation on February 1, 1840, and ended with the completion of a set of seven solo songs and two duets from Friednch Riickert s LLebe^fruhLing on January 16, 1841.2 During these months, he composed 139 solo songs, as -well as 12 duets and part-songs, all but 3 of •which were ultimately published in twenty-three opera between 1840 and 1858. Which of these publications are intended as song cycles, and -which are merely collections that were put together for commercial reasons? As I have observed in chapter 3, it is very difficult, if not impossible, to definitively answer such questions, because no clear distinction -was made between these two categories in the early nineteenth century. Certainly there are some cases among Schumann's output that are clear-cut. The last two publications that contain songs composed in 1840, for example, are the Fiinf Lieder and Geddnge, published in 1854 as opus 127, and the Vier Ge^iinge, which was published posthumously in 1858 as opus 142. Both of these collections consist of outtakes from two earlier cycles —Dichtertizbe and the Zwiilf Gedichte von JiMtiniM Kerner—rmx&A in with single settings of texts by various poets and were arranged for publication many years after the songs were composed. To argue that either of these is a song cycle would reduce the term to a meaningless concept. However, there are two collections of songs —Dicbterltebe and Frauenltebe und Leben—that have universally been regarded as cycles by scholars and performers and for which the evidence of Schumann's intention is unambiguous. But what of the Seclu Gedicbte von Reinlck, published as opus 36 in 1842, or the Zwoif Gedichte atu F. Rilckert<) Liebedfriibling, which appeared as opus 37 in 1841? Rufus Hallmark has recently included both of these among a list of Schumann's song cycles, but to my knowledge he is the first Schumann scholar to do so, and it is extremely rare for anyone to perform either work as a complete entity. In the same list, Hallmark excludes Schumann's second published
Schumann '<> Proceed of Composing a Cycle
173
song collection, Myrthen, because he believes that it lacks many of the defining characteristics of a song cycle, among -which he includes "a single poet, shared poetic theme or mood, narrative linkage, or prominent shared motives." It is true that this work is almost never performed as a complete cycle, either, and in fact, the composite range of the songs makes it difficult for a single singer to do so. And yet it is clear from both the public announcement for this work in the Neue Zeittichrift fiir Miuiik and the title page of the first edition that Schumann intended it as a cycle. In an issue of his journal that appeared on June 16, 1840, Schumann advertised the forthcoming publication of a "Liedercyclus von Gothe, Riickert, Heine, Burns and Byron, Op. 25." When he published the -work in October as Myrthen, he changed the subtitle to read: ''Liederkreu von Gothe, Ruck* ert, Byron, Th. Moore, Heine, Burns, und J. Mosen." We are not bound, of course, by Schumann's own designations in defining the song cycle. But even if we decide to exclude Myrthen from the genre ourselves, we should not ignore the fact that Schumann considered it a cycle.4 It is a question not merely of which works belong and which do not but, just as important, how we understand those works that do belong. Schumann's designation of Myrthen as a Liederkreu provides a significant piece of evidence that our understanding of the genre is quite different from his. We can find a great deal more evidence about how Schumann conceived of his song cycles in the three volumes of song manuscripts that now reside in the Berlin Deutsche Staatsbibliothek, which constitute the most complete source that we have from his Liederjahr. They contain drafts for all but 2 of the 139 solo songs that he composed in that year, as -well as 3 duets from the same year and A later songs. The drafts are numbered continuously throughout the three volumes, and most of them are dated. Each volume bears the title Liederbuch, folio-wed by the months during -which the drafts -were composed. The first volume, which contains forty-nine drafts, is dated "Februar, Marz und April 1840," the second, -which contains fifty-six, is dated "Mai, Juni, Juh 1840," and the third, with thirty-nine drafts, is simply dated "Von September 1840 bis."5 This final inscription reflects the fact that the feverish pace of Schumann's song composition slowed dramatically during the second half of 1840 and then ceased altogether in January 1841. It appears that Schumann did not really complete his collection of manuscripts but simply lost interest in it as his compositional activity trailed off. This impression is reinforced by the presence of several drafts at the end of the third volume that postdate the Luderja.hr — one from December 30, 1841, and three from as late as 1846 and 1847. We owe the existence of this remarkably comprehensive source both to the compulsive streak in Schumann's personality, which has left us such other valuable documents as the HaujhaLtbiicber and the BriefverzeichnLi, and also to the fact that he decided to compile the drafts of his songs as a gift for his bride. But despite the comprehensiveness of these notebooks and despite Schumann's apparent care in compiling them, we must bear in mind several caveats before we look to them for evidence about Schumann's composition of his song cycles. First, for most of Schumann's songs there were undoubtedly other manuscripts at one
174
From Song
time, even if many of them are lost today, •which means that the drafts only record one stage of each song's composition. Second, the drafts are a heterogeneous mixture — some served as Stichvoriagen, or engravers' copies, while others contain numerous corrections and differ substantially from the published version— so we must exercise care in comparing the drafts for different opera. Third, there are a number of inconsistencies in the dating and arrangement of the drafts. About two dozen songs have no dates at all, and another dozen have approximate dates, such as "Anfang April" or simply "Nov," which implies that Schumann dated some of the drafts long after he had composed them and did not always remember the exact date. There is also at least one specific date that is obviously incorrect, and there are several dated drafts that appear out of order, even though Schumann took the trouble to number all of the drafts consecutively as he placed them in the notebooks. Despite these inconsistencies, we can still discern a certain logic to the arrangement of the drafts that corresponds in some way to the ultimate disposition of the songs within the various published editions. Nineteen of the twentythree publications that include songs from the Liederjabr are represented in the Berlin notebooks by more than one piano draft.6 If we compare the placement of the drafts for each of these publications -with the published order of the songs, we can divide the publications into four categories, as follows: 1. The drafts appear in the exact order of publication, and they are clearly separated off as a unit, complete with a title page. This category includes three published sets of songs: the Heine Liederkreii, opus 24; Dichterliebe, opus 48; and Frauenliebe and Leben, opus -427 2. The drafts appear either in the exact order of publication or very close to it, but they do not have a title page and are not set off as a separate unit. This category includes five sets of songs: Fiinf Lieder on texts by Hans Christian Andersen, opus 40; and Drei zweuftimmige Lieder, opus 43, both of which appear in their published order; as •well as Drei Gedicbte von E. Geibel, opus 30; Drei Geddnge von A. von Cbamudo, opus 31; and Secbd Gedicbte von Reinick, opus 36, for each of which there is a single discrepancy between the order of the drafts and the published order of the songs. 3. There is no discernible relationship between the order of the drafts and the published order of the songs, but they still appear together in the same section of the notebook. This category includes four sets of songs: Myrtben, opus 25; the Eichendorff Liederkreu, opus 39; the Kerner Liederrei.be, opus 35; and Zwoif Gedichte aiu F. Riickertd Liebe
Schumann i Proc&M of Composing a Cycle
175
one for cycles and the other for collections — is an oversimplification. It is true that the first category consists of three song publications that are universally accepted as true song cycles. And the final category consists of those publications that are clearly miscellaneous collections and •were not conceived as coherent groups of songs. But the two categories that fall in between these extremes suggest a relationship that is the inverse of -what we would expect: those publications to which Schumann gave the title Liederkreu or Liederreihe are among the most scrambled, and those that were published simply as Lieder or Gedtchte and are rarely performed as cycles •were composed in what is essentially their definitive order. If -we -want to resolve this apparent paradox, we must consider the drafts more carefully and see how they fit into the larger body of Schumann's song manuscripts, both extant and lost.
Composing Dichterliebe Although the drafts in the Berlin notebooks are the only surviving autographs for many of Schumann's songs, additional manuscript sources exist for a number of them, and it is likely that numerous others have been lost. When he composed his early piano music, Schumann generally worked at the keyboard, improvising and sketching ideas that he gradually transformed into finished pieces. In a well-known letter to Clara Wieck of February 24, 1840, written shortly after he turned to song composition, he explains that this new genre requires a different procedure: "Mostly I compose them standing or walking, not at the piano. It is an entirely different kind of music, which does not come first through the fingers, but much more directly and melodiously.9 Schumann's new method of composition, which is apparently peculiar to his songs, is documented by the particular kinds of autograph material that he produced in the process of composing them, as Hallmark has observed. Hallmark designates the manuscripts that Schumann saved in the Berlin notebooks as "piano drafts" and suggests that in general these represent the second stage of the songs' composition. The first stage, which Schumann describes in his letter, is recorded in the form of what Hallmark calls "vocal sketches," -which consist of a vocal melody, usually complete, with no accompaniment.10 Vocal sketches survive for only a minority of Schumann's songs, but there is reason to believe that he wrote them for many others, perhaps for most of them. He made no effort to save these sketches, as he did the piano drafts, and those that survive are scattered among various libraries and private collections.11 In at least some cases, there was also a third stage of -writing down, a fair copy that was used as a Stichvorlage. There are again relatively few of these that survive, but the existence of others that have been lost can be postulated. It is clear, however, that Schumann sometimes sent a set of piano drafts directly to a publisher or at least intended to do so at the time he compiled them in the Berlin notebooks. This latter circumstance helps to explain the exceptional appearance of the drafts for the Heine Liederkreu, Dichterliebe, and Frauenliebe andLeben. In the case of the first of these cycles, the drafts -were indeed used as a Stichvorlage, as is clear
176
From Songj into Cycl&i
from the presence of engravers' markings. In the case of Frauenltebe arid Leben, there is a later fair copy, and Hallmark argues that there must have been one for Dichterliebe as well, even though no such manuscript survives. But these cycles were not published until three and four years after they were composed, and there is some evidence that Schumann originally intended their drafts as Sticbvorlagen as -well.13 In any case, the fact that these three cycles are set off as separate entities and given title pages suggests that their drafts represent a later stage in the compositional process than the other drafts in the notebooks. By the time he compiled them, Schumann had already decided which songs would be included in these cycles, as well as the order in which they would be published. In the case of the Heine Liederkreu and Fmuentiebe undLeben, the order of publication was determined even before Schumann composed a note of music, since they both set groups of poems that -were published as integral lyric cycles. The complete set of vocal sketches that survives for the latter -work documents the fact that Schumann adhered to this order when he Erst composed the songs. But -we must be careful not to generalize too much from this, since Schumann very rarely used such preexisting texts for his song cycles. The autograph sources for Dichterliebe, which presents a more complicated case, reveal that Schumann composed the songs in their definitive order as long as this was dictated by his literary source, Heine's Lyridched Intermezzo. But as soon as Schumann decided to stop following Heine's sequence, he adopted a more fluid approach, in which the order and contents of the cycle emerged as he went along. There are two autograph sources that survive for these songs: the set of piano drafts in the Berlin notebooks and a much smaller manuscript, now in the Robert-Schumann-Museum in Zwickau, which primarily consists of a partial set of vocal sketches. At the time Schumann composed these two manuscripts, between May 24 and June 1, 1840, he set a total of twenty of the sixty-six poems in Heine's cycle and gave them the title Gedichte von Heinnch Heine, It was only after the cycle was accepted for publication by Peters, in November 1843, that Schumann omitted four of the songs and changed the title to Dichterliebe.14 In his comprehensive study of the sources for Dichterliebe, Hallmark shows that Schumann went back and forth between the sketch manuscript and the drafts as he composed the first half of the cycle and then abandoned the sketches altogether, composing the later songs directly as piano drafts. He began by setting the first seven poems in Heine's cycle consecutively, in the form of vocal sketches, and then immediately elaborated the sketches as piano drafts. At this point Schumann stopped, and before he composed any more songs he made a list of twenty-one poems, in the order in which they appear in Heine's book. It is not entirely clear whether Schumann's intention was to set all of the poems on this list in order or he -was simply jotting down those texts that interested him, from which he would then make a further selection. It seems likely that it was the latter, since after setting the first poem in his list, "Im Rhein, im heihgen Strome," which he numbered 8, he began to skip around and became uncertain as to what the order of the remaining songs should be. The next three songs in the Berlin notebooks — "Ich grolle nicht," "Und wiifiten's die Blumen," and "Das ist ein
Schumann i Proceed of Composing a Cycle
177
Floten und Geigen" — demonstrate the danger of using the order in -which the piano drafts appear to determine the order in -which Schumann composed his songs. He -wrote the drafts for these three songs in the order in which he ultimately published them as part of Dichterliebe, but the sketch manuscript reveals that this is not the order in which he initially conceived or composed them. He first wrote just the title of "Ich grolle nicht," which is the second poem on his list, and numbered it 9. He then skipped to the fourth poem on his list, "Das ist em Floten und Geigen," and composed a piano-vocal sketch, which he numbered 10. But at this point he decided to place the sixth poem on his list — "Und wiifiten's die Blumen"—in between these two, and so he added its title, numbered 10, above the sketch of "Das ist ein Floten und Geigen" and renumbered the sketch 11. He then wrote the piano drafts for the three songs.15 There is no way to tell from the appearance of the drafts that the third song in the notebook is actually the first that Schumann composed. I do not wish to make too much of the discrepancies between the order in which these three songs were composed and the order in -which they appear in the Berlin notebooks. With one further exception, all of the remaining songs in the cycle were composed in their definitive order, and it is true, as McCreless claims ck Dichterliebe, that "the order of composition is essentially the order of the published version, with the occasional rearrangement of a few songs."16 But if we are going to rely on the evidence that the Dichterlie.be. sources provide in order to reconstruct the compositional history of the Eichendorff Liederkreit, for -which the surviving sources are apparently not complete, we need to consider this evidence carefully and decide which of it is applicable and which is not. Even though Schumann did not set the complete Lyrutcbej Intermezzo, he -was clearly guided by Heine's arrangement as he constructed his song cycle. Despite his frequent changes of mind about -which poems to include, Schumann ultimately did not alter Heine's arrangement very much at all, taking only three of the twenty poems that he set out of order. In this sense, the relationship between Dichterliebe and its literary source is much closer to that of the Heine Liederkreu and FrauenLie.be und Leben than it is to any of Schumann's other cycles. The fact that Schumann became uncertain about the arrangement of the songs at the moment -when he abandoned Heine's order is thus the most relevant point, since this is -where the process of composing Dichterliebe diverges most sharply from the process of composing the Heine Liederkreii and Frauenllebe und Leben and begins to resemble Schumann's more typical procedure. Furthermore, it is clear from the Zwickau sketch manuscript that Schumann did not begin to consider the shape of the cycle until after he had composed at least seven songs, and he did not have a very clear picture of that shape until he had composed several more. So, to say that Schumann "had an order in mind relatively early in the compositional process" is not entirely accurate.17 It would be closer to the mark to say that for one reason or another, he ended up publishing the songs of Dichterliebe in the order he composed them. As -we consider some of Schumann's other cycles, including the Eichendorff Liederkreui, we shall see that this is an important distinction, because it means that even if
178
From Songj into Cycled
Dicbterliebe more closely resembles Frauenliebe and Leben and the Heine Liederkreu in its final version, it was conceived in the same way as his other cycles: Schumann began by simply composing a group of songs and only subsequently decided how he would arrange them into a cycle.
Schumann's Two-Stage Process Because the ultimate order of the songs in Dichterliebe ended up being very close to the order of composition, this cycle does not present us with the most obvious example of Schumann's typical procedure. If we consider the manuscript sources for some of the cycles from my third category — those in which the order of the drafts bears no relationship to the published order of the songs—we find evidence that is less ambiguous. The most extreme example of a cycle that was composed first and arranged later is undoubtedly the ZwoLf' Lieder CUM E Riickerfo Lieb&ffriihling. This cycle is the only musical work that was published as a joint opus by Robert and Clara Schumann, appearing as his opus 37 and her opus 12 in September 1841. It includes the last songs that Schumann composed before turning his attention to orchestral music and thus brings us to the end of the Luderjakr. It also takes us to the end of the Berlin notebooks, since the drafts for seven of the Riickert songs appear, as a group, near the end of the third volume. Although only the first of these drafts is dated — January A, 1841 — Hallmark has used entries from Schumann's Haiuhaltbucb to determine that they probably appear in the order in which they were composed, with the last one dating from January 16. Schumann also composed two duets from Liebejfriihling during the same twelve-day period, but no drafts for these survive, presumably because he generally excluded his duets from the notebooks.18 After Schumann completed these drafts, he put them aside for several months and waited until Clara composed the three songs that she was to contribute. She presented these songs to Robert as a birthday gift on June 8, and two days later he recorded in his Haiuhaltbuch "the Riickert Lieder notated" (die Riickert'ttcben Lieder twtirt), which Hallmark takes as a reference to the fair copy of the songs. This manuscript, which is in Clara's hand, survives today in the Bibliotheque Nationale in Paris, although the three songs that she composed have been removed. Robert's seven solo songs appear in the order of the Berlin drafts, which suggests to Hallmark that Clara was copying from the notebooks. It also means that the Schumanns began to work on the order of the cycle only after the fair copy had been completed. Clara's manuscript served as the Stichvorlage, and the correct order of the cycle was indicated to the engraver by a series of red Roman numerals at the top of each song. Hallmark has discovered that these red numerals are written over partially erased numerals in black, and he has used the latter to reconstruct two earlier orderings for the cycle, which were both ultimately rejected in favor of the published version. The Paris manuscript thus documents the considerable effort that Schumann devoted to the arrangement of the Riickert cycle. He apparently continued to work on it for most of the summer, since he records the completion of the "Riickertiana" in his Haiwhalt-
Schumann '<> ProceM of Composing a Cycle
179
buck on August 19, more than two months after the fair copy was written and a month before the cycle was published.19 While the amount of time that elapsed between the composition of the Riickert songs and their arrangement into a cycle may be unusual for Schumann, the fact that these were two distinct stages of his compositional process is not. We can see that he employed a similar procedure, for example, in the case of his opus 25 Myrtken, which is one of the first song cycles that he composed and also one of the first to be published, appearing in September 1840. In addition to the piano drafts for these songs, which occupy more than half of the first volume of the Berlin notebooks, there are four vocal sketches and an incomplete fair copy that survive. But there are also two further documents that shed some interesting light on Myrtben'tt conception.20 The first is a letter that Schumann •wrote to the publisher Franz Kistner on March 7, 1840, proposing the publication of the cycle. In his letter, Schumann suggests that the title page read: "Myrthen: Liedercyclus in vier Heften von R. Sch.," and then lists the poets whose texts he has set, along with the contents of the first book. Schumann names five of the seven poets •whose texts were ultimately included in the published cycle, excluding both Julius Mosen and Thomas Moore, and his description of the first book differs considerably from the published version. Despite these discrepancies, Schumann evidently considered Myrthen to be a complete cycle at the time he "wrote to Kistner, and he expresses some impatience to get it into print. He tells Kistner that he "would like the first two [books] completed by the end of May, the latter two by the end of August." And he "writes that he expects to receive the fair copy of the first book by the following Monday, after which Kistner "could immediately begin with the engraving."21 All of this turned out to be premature, since Schumann •was not yet done with the arrangement of the cycle and, in fact, had not even composed all the songs that he would ultimately include. The letter to Kistner provides us "with a partial glimpse of Schumann's initial conception of Myrthen. In Table 7.1, I have compared the arrangement of the first book that Schumann proposed in his letter both with the numbering and dating of the drafts in the Berlin notebooks and also with the placement of the songs in the published edition. As Schumann altered this arrangement in the final version of the cycle, he retained Riickert's "Widmung" as the opening song and kept the last three songs consecutive in the same order, although he ultimately placed the second and third Riickert songs at the end of the second book and the folio-wing Burns song at the beginning of the third book. What is more significant, however, is the fact that the arrangement Schumann proposed to Kistner on March 7 bears no relationship at all either to the order in -which the drafts were composed or to their placement within the first Berlin notebook. This suggests that as Schumann composed and compiled the drafts he -was not thinking about the order in which he -would publish the songs. Schumann composed three more of the Myrthen songs after he wrote the letter to Kistner. The drafts of Thomas Moore's "Z-wei Venetianische Lieder" are dated March 13 and 14, and the Riickert song that Schumann titled "Aus den
180
From Songj into Cycled
TABLE 7.1 Myrthen, Book 1, As Proposed to Kistner on March 7 Kistner letter 1. 2. 3. 4. 5. 6.
"Widmung" by Riickert "Lotosblume" by Heine "Jemand" by Burns "Mutter, Mutter" by Riickert "Lass mich ihm am Busen hangen" by Riickert "Mein Herz ist im Hochland" by Burns
DStb. Mus. ms. 16/1 47
8 Feb. 12 16 Feb. 25 20 March 1 22 March A 10 Feb. 12
Op. 25 1 7 4 11 12 13
ostlichen Rosen," the last draft to be composed, is dated simply "Anfang April." It was at some point after Schumann composed the Moore settings, but apparently before he got to the final Riickert song, that he sketched out 'what is essentially the final arrangement of the cycle. His notation of this arrangement is preserved on an undated sheet that somehow made its way between two of the drafts in the Berlin notebooks. Schumann lays out the contents of all four books, listing the titles of the songs and the number of pages each 'would take up and, for some of the songs, the key and the name of the poet as well.22 The only discrepancy between this arrangement and the final version of the cycle is that the former includes Riickert's "Jasminenstrauch," but by the time Schumann published the cycle he had substituted the same poet's "Aus den ostlichen Rosen." These two documents—the letter to Kistner and the undated sheet—provide clear evidence that in the case of Myrtken Schumann did not decide on the order of the songs early in the compositional process, did not compose the drafts in the order he intended to publish them, and did not even bother to compile them in his notebook in that order. On the contrary, Schumann only began to arrange the songs into a cycle after he had essentially finished composing them. It may be objected that because Myrthen contains twenty-six songs and sets texts by seven different poets it is so different from Schumann's other cycles that it is not an appropriate point of comparison for the Eichendorff Liederkreu. But as I have noted earlier, it is we who have posited the uniqueness o^Myrtben and not Schumann. He published it with the same designation—Liederkreut—that he gave to both his first Heine cycle and his Eichendorff cycle. And if we consider the documentary evidence on its own terms, rather than imposing our distinctions and categories a priori, then it becomes clear that, in terms of the appearance of their drafts, Myrthen is the cycle that is most similar to the Eichendorff Liederkreu. In both cases, a majority of the drafts are consecutive in the order that they were composed, but several of the Myrthen drafts and one of the Eichendorff drafts have been placed much farther back in the notebooks than their dates would suggest. Furthermore, these two sets of drafts are the only ones among the first three categories that I outlined earlier — -which is to say, the only ones among those sets that Schumann conceived as coherent groups of songs—that have other songs from different publications mixed in.
Schumann 'j ProceoJ of Composing a Cycle
181
I would like to make one final observation about the evidence that the Berlin drafts contain concerning Myrthen. Although Schumann ended up placing the songs in an arrangement that is unique among his cycles, the drafts suggest that •when he began composing the songs he -went about it in a way that was quite typical. As with his settings of Eichendorff, Kerner, Chamisso, Reinick, Andersen, and Geibel and his later settings of Heine and Rtickert, Schumann composed the songs that eventually became Myrthen by taking one volume of poetry at a time and setting small groups of selected poems over a period of a few days. It is impossible to reconstruct the order in which the Myrthen. drafts -were composed with complete certainty, because eleven of them have either inexact dates or no dates at all and several of these appear in the notebook out of order, but I have presented a hypothetical reconstruction in Table 7.2 that is at least mostly accurate and that is entirely consistent with the dating of the drafts. Those songs that -were composed simultaneously with the Myrthen songs but published elsewhere appear in brackets. Schumann started with Goethe's W&it&itlicher Divan and set five poems from this collection. He then set Heine's "Belsatzar," a ballad that •was published in 1846 as his opus 57, and a single poem by Byron (in German translation). On February 11 and 12, he turned to two books by different poets — Heine's Buck 3er Lieder and Gerhard's Gedichte von Robert Burn*) — and set three poems from the former and four from the latter. Then, after setting Mosen's "Der Nufibaum" on February 16, he decided to set a complete cycle of nine poems from the Buch derLieder, which he completed on February 23, sending it off to Breitkopf und Hartel on the same day to be published as the opus 24 Liederkreu'. The next day, February 24, he returned to the volume of Burns poems and set four more of them, followed by single poems by Chamisso, Ferrand, and Byron, of which only the latter was published in Myrthen, and finally, beginning on March 1, he turned to Riickert's collected poems, setting at least four and probably five by the time he wrote to Kistner on March 7. As we have seen, the final three songs, two by Moore and one more by Riickert, were composed after he had begun arranging the cycle. If we consider the drafts in this way, they reveal two interesting points about the composition of Myrthen that have wider implications for our understanding of Schumann's cycles. First, the order in which the drafts were composed reinforces •what I have already been suggesting about Myrthen, which is that the composition of the songs and their arrangement into a cycle •were two distinct stages of the compositional process, and it appears that Schumann •was not even considering the latter stage at the time he was engaged in the first. We can tell that this was the case because there is a clear logic to the order in -which he composed the songs — he essentially took up four different poets in turn and set a group of poems by each of them—and this logic continued to operate up until the moment he wrote to Kistner with his proposal for publishing the cycle. There is, of course, a logic to the order in which he arranged the songs once they were composed, but a completely different logic, and it is hard to see how he could have been thinking in both ways at the same time. The assumption that Schumann either simply set a group of poems in the order he found them, or set them in the order he in-
182
From Song
TABLE 7.2 The Order of Composition of the Myrthen Songs DStb. Mus. ms. 16/1 Number Date 3
4 2/3 5, 62/4 37in Feb. [12/7 7 2/9 92/11 82/12 102/12 112/12 12 13
21Mid-Feb. 462/16 [27-36 142/24 15
162/25 19
[172/25 [182/28 49End Feb. 203/1 223/4 [233/4 47
26Mar. 24, 25 3/13-14 Beg. Apr.
Title
Poet
"Lied der Suleika" "Freisinn" "Aus dem Schenkenbuch" (2) "Taksmane" " Belsatzar" "Hebraischen" "Was will die einsame" "Lotosblume" "Mem Herz ist im Hochiand" "Hauptman's Weib" "Du bist wie eine Blume" "Hochlandisches Wiegen" "Im Westen" "Der Nufibaum" Liederkreis (9) "Hochlanderwittwe" "Niemand" "Jemand" "Weit, weit" "Was soil ich sagen" "Em Gedanke" "Rathsel" "Mutter, Mutter" "Lass mich ihm am Busen hangen" "Jasmmenstrauch" "Widmung" "Adagio" "Venetian Lieder" (2) "Aus den ostlichen Rosen"
Goethe Goethe Goethe Goethe Heine] Byron Heine Heine Burns Burns Heine Burns Burns Mosen Heine] Burns Burns Burns Burns Chamisso] Ferrand] Byron Riickert Riickert Riickert] Riickert Riickert Moore Riickert
Op. 25 Number 9 2 5,6 8 15 21 7 13
19 24 14 23 3 10 22
4 20 16 11 12 1 26
17, 18 25
tended to publish them thus sets up a false dichotomy. Second, we begin to see how difficult it is to generalize about Schumann's cycles and distinguish between those that follow a norm and those that deviate from it. When we consider Myrthen as a published work it may appear to be so different from Schumann's other cycles that it stands outside the genre altogether, but when we see how Schumann conceived the cycle we realize that it is not as unusual as it appears.
Coherent Publications and Miscellaneous Collections If we compare the manuscript sources for all five of the song cycles that I have considered so far—the Heine Li&derkreu', Myrthen, Dichterliebe, Frauenliebe and Leben, and the Riickert Lieder—it becomes clear that Schumann did not really have a single typical procedure for arranging songs into cycles. But as different
Schumann '<> Proce
183
as his compositional process may be for each particular cycle, there is at least a general principle that seems to apply to all of them. Schumann's first stage in composing all five of the cycles •was to take a book of poetry and set a group of poems over a period of a few days. The arrangement of the songs into a cycle was a second stage of the process, which usually began after the songs had been composed. The contrasting ways in which Schumann engaged in each of these stages exemplify the split between his impulsive and compulsive sides, a dichotomy that Schumann himself recognized in creating his two alter egos, Florestan and Eusebius. On the one hand, the incredible rate at which Schumann composed most of the songs during his Liederjahr and the fact that he usually revised only details of his initial settings suggests that the process -was almost improvisational, with the major compositional decisions coming intuitively. The arrangement of the songs into a cycle, on the other hand, was relatively labored and involved much reflection and many changes of mind. And so, in the most extreme case, we see Schumann compose nine Riickert songs in twelve days, in •what is essentially their final form, and then spend more than two months trying to decide -what order to publish them in. I believe that this principle is consistent with the evidence that -we have for all of Schumann's cycles, including the ones that were put into order at the beginning of the compositional process. With FrauenLie.be and Leben and the Heine LiederkreLi, it is not that Schumann decided on the order of the songs early on but that he never had to decide, since the order -was dictated by his poetic source. These two cycles are exceptional, then, in that Schumann was able to forgo the second stage of his process altogether. As for Dichterliebe, I have already observed that Schumann did not really arrange the songs into a cycle as he composed them, since the Zwickau manuscript documents a decision-making process that only began after he had already completed seven of the songs and continued as he composed the rest of them. This cycle is exceptional in that the two stages overlapped, but again this resulted from the unique nature of its poetic source. As -with his first Heine cycle, Schumann began to compose DichterLie.be by turning to a complete lyric cycle from the Euch der Lieder and setting the poems consecutively. But where the earlier cycle contained nine poems and •was thus an ideal length for a song cycle, the sixty-six poems of the Lyridched Intermezzo "were clearly too many, and so at some point Schumann had to consider the contents and arrangement of the cycle. However, the fact that he began to set Heine's cycle consecutively, despite its impossible length, demonstrates how intuitive the initial process of selecting and setting the poems was for Schumann. As with the Riickert Lieder, the arrangement of the songs into a cycle, once begun, required a great deal of conscious reflection and involved several changes of mind. In fact, Schumann did not make his final decision'—to eliminate four of the songs—until he was ready to publish the cycle in 1844, four years after he had composed it. The fact that the composition of the songs and their arrangement into a cycle were usually two separate stages can also help to explain the apparent paradox that I posed near the beginning of this chapter: why do the publications that Schumann designates as cycles tend to be more scrambled in the Berlin note-
184
From Songd into Cycle,)
books than those that are simply designated as Gedichte or Liederl In the case of Myrthen, the Eichendorff Liederkreu, the Kerner Liederreihe, and the Riickert Liefer, the order of the songs •was apparently an important consideration for Schumann, and so after he composed the songs he arranged them into a particular sequence. With the publications in the latter category—the Geibel and Reinick Gedichte, the Chamisso Gedcinge, the Andersen Lietter, and the opus 43 duets—he was not so concerned with the order of the songs, and so, with occasional rearranging, he simply published them in the sequence in which he had composed them. There may well be a relationship between the fact that Schumann did not designate these latter sets of songs as cycles, on the one hand, and did not bother to arrange them into a particular order, on the other. But there are coherent relationships among their texts that do not depend on order, and the fact that Schumann published them -with neutral designations does not necessarily mean that he considered them collections rather than cycles, as -we have seen in chapter 3. If we compare the drafts for all twelve sets of songs -within my first three categories — those that are set off as a unit within the Berlin notebooks, those that are not but are essentially in the order of publication, and those that are scrambled but are still in the same section of the notebooks—then they confirm what we can already see from the published works: Schumann arranged his song publications in a wide variety of ways. We are again confronted with the difficulty, and perhaps the irrelevance, of clearly distinguishing between a song cycle and a collection of songs. I believe that the more useful distinction is between these twelve sets of songs, on the one hand, and the seven publications in my fourth category—those •whose drafts are scattered throughout the notebooks — on the other. The former category consists of collections that were evidently intended as coherent publications at the time they were composed, and all -were published within a fewyears, half of them by the end of 1841. They sometimes have literary titles, and their designations invariably include the name of the poet, or poets, whose -work they set. The latter category consists of miscellaneous collections that were apparently put together at the time of publication. These generally appeared long after Schumann's Liederjakr, the earliest of them in 1843 and the latest, posthumously, in 1858. As might be expected, each of these latter collections includes songs that were composed at various times throughout 1840 and even beyond and that set texts by a variety of poets. In some cases it is clear how they were put together. I have already mentioned the last two of these that were published— opp. 127 and 142—-which essentially consist of outtakes from earlier cycles. In two other collections, opp. 45 and 49, Schumann put a pair of songs with texts by a single poet together with a third song that sets a different poet's text, presumably because he generally needed at least three songs in order to publish an opus. The generic titles of these collections provide further evidence that Schumann used them simply as a -way to publish songs that are otherwise unrelated. Most of them are part of two ongoing series of publications, titled Lieder unt) Gejange and Ramanzen un3 Balladen, that appeared throughout the late 1840s and early 1850s.
Schumann '<> ProceM of Composing a Cycle
185
The Eichendorff Songs Having surveyed the source evidence that survives for Schumann's 1840—41 song cycles generally, I will now turn to the opus 39 LiederkreL) and reconsider Patrick McCreless's provocative hypothesis about its compositional history. McCreless tries to reconcile the manuscript evidence that survives for the Eichendorff songs -with his assumption that the order of the songs in a cycle is immutable. Even though the evidence strongly suggests that Schumann did not have the definitive order of the songs in mind when he began to •work on the cycle, McCreless argues that Schumann still had some order in mind and this is the order in -which he composed the songs. In McCreless's view, -what makes the composition of the opus 39 Liederkrei) unique among Schumann's cycles is not that the arrangement of the songs emerged late in the process, but that Schumann changed his mind about his initial arrangement. McCreless derives the order of composition of the Eichendorff songs from a list in Herwig Knaus's book on the cycle, -which gives the dates of the songs as they appear on the Berlin piano drafts. I have reproduced the relevant portion of the list as Table 7.3.23 I have then represented the actual appearance of the drafts within the notebooks in Table 7.4, with the number and date that Schumann -wrote on each draft, as well as the page numbers. I have also included, in brackets, the drafts of six other songs that are mixed in with the Eichendorff drafts. The drafts are actually broken up between the first two volumes of the notebooks. Number 50, a setting of Chamisso's "Die Kartenlegerin" that was published in opus 31, is the final draft in the first volume, and number 51, Heine's "Die Grenadiere," -which appeared in opus 49, begins the second volume.24 If we consider the way in which the drafts actually appear, we can see that Schumann placed them in the notebooks quite casually, a significant point that is ignored in Knaus's list. Furthermore, there are two discrepancies between the order in which Knaus lists the drafts and the order in -which they appear. First, -while Knaus has "Die Stille" immediately after "In der Fremde (I)," Schumann placed the former draft near the end of the songs, after "Friihlingsnacht." Since "In der Fremde (I)" and "Die Stille" both have the same date, there is no way to determine -which of these two drafts was composed first. Second, although Schumann's dates suggest that "Wehmut" -was composed before "In der Fremde (II)," which is how they appear in Knaus's list, it is clear from the appearance of the actual drafts that this could not have been the case. We can tell that Schumann wrote the drafts of "In der Fremde (II)," "Wehmut," and "Friihlingsnacht" continuously, because for these three songs he began each successive draft on the same page that he ended the preceding draft. There is thus no question that they •were composed in the order in which they appear in the notebook. The likely reason for the discrepancy in Schumann's dating of these songs becomes clear when we examine the actual appearance of the dates on the drafts. Schumann wrote "17 u. 18 Mai 1840" above the top staff of "Wehmut" and to the right of the song's title. The placement of this date and the fact that it is in ink suggest that Schumann wrote it at the time he composed the draft. All of the other dates
186
From Songj into Cycles TABLE 7.3 Herwig Knaus's Compositional Order for the Eichendortt Songs Title
Date
"Waldesgesprach" "In der Fremde (I)" "Die Stille" "Mondnacht" "Intermezzo" "Schone Fremde" "Wehmut" "Friihlingsnacht" "In der Fremde (II)" "Zwielicbt" "Im Walde" "Auf einer Burg"
May 1 May 4 May A May 9 before May 16 May 16-17 May 17-18 May 18 May 18 May 19 May 20 May 20-22
on the Eichendorff drafts were scribbled in the left-hand margin, and some, including the one on "In der Fremde (II)," are in pencil. Schumann evidently added the latter date in a subsequent pass through the manuscript, perhaps when he was revising the songs, which he also did in pencil. It would have been very easy for him to miss the date on "Wehmut" and thus to misremember exactly when he composed "In der Fremde." The correct date is presumably May 17. So even if we accept McCreless's hypothesis, -we are still left-with at least one and possibly two discrepancies between his postulated compositional ordering and the actual order in which the drafts were composed. More significantly, the dating of the drafts provides evidence that they do not all represent the same stage of composition and thus may not be a reliable guide to the order in which the songs were initially conceived. May 16 appears to mark the beginning of a dramatic increase in Schumann's compositional activity. Before this date, he composed five drafts in the course of fifteen days, and then, beginning on May 16, he produced seven more drafts in the next four or five days.25 The appearance of the drafts themselves confirms that they •were produced at an increasingly feverish pace. Schumann began each of the first six drafts on a fresh page, but beginning with "In der Fremde (II)," as I have noted, he started writing the drafts continuously, and he proceeded in this fashion until he completed "Auf einer Burg," the last song that he composed.26 Given this evidence, it is surprising that Schumann reports to Clara Wieck in a letter of May 15 that "there are twelve Eichendorff [songs]. But I have already forgotten them and begun something new."27 Why does Schumann refer to twelve songs when he has composed only five drafts? And why does he write that he has already forgotten about the Eichendorff songs and started a new project and then begin the most concentrated period of work on the cycle ? The most likely explanation
Schumann '
187
TABLE 7.4 The Eichendorff Drafts in the Berlin Notebooks
Title
Number
Date
Page Numbers
DStb. Miu. nu. 16/1 "Waldesgesprach" "In der Fremde (I)" "Mondnacht" ["Der Nufibaum" ["Widmung" "Intermezzo" ["Rathsel" ["Die Kartenlegerin"
May 1 May 4 May 9 Feb. 16
110-13 114-15 116-18 119-22] 123-26]
48 49 50
May end of Feb. July 14
127-28 129-30] 130-34
51 52 53 54 55 56 57 58 59 60
May 12 May 16-17
3-7]
43
44 45 46 47
DStb. Mtu. nu. 16/2 ["Die Grenadiere" "Schone Fremde" ["Rothes Roslein" "In der Fremde (II)" "Wehmut" "Fruhlingsnacht" "Die Stille" "Zwielicht" "Im Walde" "Auf einer Burg"
May May May May
15 18 17-18 18
May 4 May 19 May 20
8-9 10-11] 12-13 13-15 15-17 18-19 20-22 22-25 25-26
is that Schumann had already composed all twelve songs in some form by May 15, probably as vocal sketches that have not survived.28 So perhaps the incredible speed -with -which he composed the remaining drafts did not result from a surge of creative inspiration but from his desire to -write down complete versions of songs that he had already created — in part on paper and in part in his head — so that he could be done -with them and move on to something else. What we have in the Berlin notebooks, then, is only a partial record of the composition of the cycle, -which documents two distinct phases of the process. At the time Schumann -wrote the first five drafts, he -was still conceiving the Eichendorff songs, and at least some of these drafts may be the first writing down of the music. By the time he began the sixth draft, for "Schone Fremde," Schumann was essentially making a -written record of songs that already existed in some form. Ultimately, there is not enough surviving manuscript material for us to come to any firm conclusions about the significance of the order in which the Eichendorff drafts -were composed. But if -we compare the appearance of these drafts to the others that Schumann collected in the Berlin notebooks, it seems probable that the order of composition does not reveal anything about the order in -which he intended to publish the songs. We have seen that when the drafts for a cycle appear in the notebooks in the same order in which the songs -were published,
188
From Songj into Cycl&f
they are set off from the surrounding songs and preceded by a title page. McCreless compares the Eichendorff Li£^erkreu to the cycles in this category and finds that Schumann's procedure is "atypical." But as I have demonstrated, there are three other cycles—Myrthen, the Kerner Lwderreibe,, and the Riickert Luder—whose drafts are much closer in appearance to the Eichendorff drafts and thus provide better points of comparison. If we take the similar appearance of these sets of drafts as evidence of a similar process of composition, then it is likely that Schumann began to arrange the Eichendorff songs into a cycle only after he finished composing them. The fact that he already refers to the songs as his "Eichendorffsche Zyklus" in a letter of May 22 to Clara Wieck suggests that, as with Myrthen, he may have come up -with at least a preliminary order right away. Although it was nearly two years before the cycle was published, he prepared a fair copy soon after he composed the songs and sent it to the publisher Haslinger on June 27. All that survives of the fair copy is the first song, "In der Fremde," the music of which is in Clara ^Vieck's hand. Schumann drew the braces, clefs, and signatures on the first page and wrote the titles of both the song and the complete cycle — Liederkreis von Joseph Freiherr von Eichendorff—at the top.29 It may be this manuscript that Schumann has in mind in the following passage, from a letter to Wieck of June A, in which he is welcoming her back to Leipzig: You will find much work to do, as •well, and I have already placed everything in readiness, and as a small reward, also, the Liederbuch, as a memento of me. Another one just as thick is already almost finished. It will open your eyes.30 The work Schumann has in mind is presumably the -writing of the fair copy, which he has placed in readiness by preparing the first page. This -would mean that he had decided on the order of the songs by the time he wrote this letter. The Liederbuch undoubtedly refers to the first volume of the Berlin notebooks, since this is its title, and -we know that he intended the notebooks as a gift for his fiancee. By the beginning of June, the first volume would have been complete and the second would already contain thirty of the fifty-six drafts that Schumann eventually compiled in it. To say that it was almost finished may have been an exaggeration, and as it turns out, Schumann did not complete it until the end of August. But at the time he was writing to \Vieck he had just composed seventy-eight songs in four months' time, including some of his most beautiful 31 music, and so his sense of excitement is understandable. By relying on the manuscript material that survives for the Eichendorff Liederkreu, as -well Schumann's apparent references to the work in his letters to Clara \Vieck, and comparing this evidence -with what -we know about the composition of Schumann's other song cycles, we can reconstruct a plausible compositional history for the cycle. Schumann conceived all twelve of the songs between May 1 and May 15, writing down five of them in the form of piano drafts and perhaps others m the form of vocal sketches that are now lost. By May 16, feeling that the songs -were essentially complete, Schumann became impatient to
Schumann '<} Proceed of Composing a Cycle
189
write the rest of them down, and so he quickly composed the seven remaining drafts, finishing the last one on May 20 or 21. It was at this point that Schumann began to consider the songs as a cycle and to decide on the order in which he would arrange them. He completed the arrangement and prepared for the copying of the Stichvorlage by June A and then turned them over to Clara Wieck, -who completed this last stage of -writing. If this history is indeed accurate and, in particular, if it is true that Schumann composed the songs first and arranged them into a cycle later, then we still must consider one final question: Why did Schumann compose the songs in the order that he did, if he did not intend to publish them in that order?
Schumann's Compositional Logic In contrast with some of his other cycles, such as Dichterliebe, Myrthen, the Heine LUderkreu, and the Riickert Lieder, Schumann did not begin the composition of the Eichendorff songs by leafing through Eichendorff's collected edition and selecting poems that he wanted to set. It was Clara Wieck -who performed this initial stage of work and then copied thirteen poems into a notebook, known as the Ab
190
From Song<> into Cycled
chapters, but I will briefly restate the points that are most relevant to the present comparison. The formal structure of "Mondnacht," a bar form, is completely straightforward, and the entire song is composed out of a remarkably small amount of musical material. In fact, most of the song consists of a single eight-measure phrase, which is repeated with little or no variation. Schumann establishes the tonality of the song entirely through the dominant harmony, which he prolongs throughout the first two strophes. The only tonic harmony that we hear, before we reach the end of the song, comes in the middle of the phrase, in a weak measure, and it functions as a ^ appoggiatura chord that resolves to V. When Schumann begins to move away from the dominant in the third strophe of the song, he substitutes V7/IV for the tonic harmony, and he uses the same substitution to undermine the final cadence, "which ultimately resolves through a plagal progression two measures after the vocal line descends to 1. Schumann creates the weak opening of "Intermezzo," which is in ternary form, through somewhat different means. He again begins with a static harmonic prolongation, as in "Mondnacht," but in this case he prolongs the tonic rather than the dominant. He uses a pedal in the bass and dissonant chromatic counterpoint in the inner voices to weaken the prolongation, so that the tonic feels transitional rather than stable. Where in "Mondnacht" the relationship between tonic and dominant is never in doubt, despite Schumann's emphasis on the latter harmony, in "Intermezzo" the weak initial tonic is followed by a tonicization of the dominant, as the first strophe reaches its cadence, and so the status of the tonic is not entirely clear as the second strophe begins. The nature of the opening strophe, •which is the A section of the form, prevents Schumann from simply repeating the section unvaried, as he does in "Mondnacht," because the opening would not make an effective recapitulation, and the cadence of the strophe cannot be used to conclude the song. Schumann varies the recapitulation by introducing the bass line that was implied at the opening of the song, thus creating a strong arrival, but as in the third strophe of "Mondnacht," he substitutes V7/IV for the tonic harmony and thus greatly exaggerates the weakness of the opening. Schumann finally does give us a definitive tonic cadence at the end of the recapitulation, but he undermines this cadence and creates an open ending, not by derailing the cadential harmonic progression, as in "Mondnacht," but by turning the cadential measure into an overlap with the beginning of the coda and using texture, register, and meter in the piano postlude to dissolve our sense of conclusiveness. In "Schone Fremde," Schumann returns to the bar form of "Mondnacht" and, as I explained in the preceding chapter, recomposes the structure of the latter song in a new guise. If-we consider "Schone Fremde" in terms of its order of composition, then we can see that, to some extent, Schumann alters the structure that he first created in "Mondnacht" by further developing compositional ideas that he introduced in "Intermezzo." As in "Mondnacht," for example, he once more prolongs the dominant throughout the first two strophes, and again there is a tonic harmony in a weak measure that functions as a * appoggiatura chord. But now Schumann concludes the first strophe by tonicizing the dominant, as in
Schumann J ProceM of Composing a Cycle
191
"Intermezzo," and so the relationship between tonic and dominant becomes ambiguous. And where in "Intermezzo" Schumann resolves the ambiguity by substituting a tonic cadence when the same music returns, in "Schone Frernde" Schumann tonicizes the dominant even more strongly at the end of the second strophe, so that we become convinced that it is the tonic of the song and are surprised when the actual tonic finally asserts itself in the third strophe. In the second strophe of "Mondnacht" Schumann repeats the music of the first strophe virtually unchanged, but in "Schone Fremde" the opening returns in the middle of the second strophe, and it is so heavily disguised that we are scarcely aware we are hearing the same music. But in one sense this return clarifies the tonally ambiguous opening, by revealing that it is, in fact, the middle of a phrase. In this sense, the unusual relationship between the opening and its return can be heard as a development of the varied recapitulation in "Intermezzo," which also clarifies the opening by exaggerating its lack of stability. Finally, there is the question of the opening itself. In "Mondnacht," there is never any ambiguity as to the identity of the tonic, despite its virtual absence. In "Intermezzo," as well, the tonic is clear as the song begins, but by the time we get to the end of the first strophe the dominant has been tonicized more strongly. "Schone Fremde" begins with a completely ambiguous harmony, and we have no sense of a key until the end of the first strophe, at which point, as I have noted, we are misled into hearing the dominant harmony as the tonic of the song. It is not until the final vocal phrase that the identity of the tonic is clarified. All of the analytic observations that I have been making about these three songs are related to a single compositional problem, which Schumann -was -working out in different -ways in each of them: how to compose a tonal piece without clearly establishing the tonic harmony at the beginning. In each successive song his solutions to this problem become more complex, and as he develops his idea he borrows and elaborates techniques from the preceding songs. If my description of these songs has taken the form of a teleological progression, it is because I am trying to point out the presence of a kind of logic — a compositional logic — that can explain why Schumann composed them in the particular sequence that he did. But I am not implying a value judgment about the relative merits of the songs as pieces of music. In fact, while "Mondnacht" may be the most straightforward of the three in many respects, it is also generally regarded not only as the greatest song in the cycle but also as one of the masterpieces of the nineteenth-century lied repertory, and it is performed as a separate piece more often, and more successfully, than any of the other Eichendorff songs. "Schone Fremde" may be the most complex, but it is also "something of an oddity," in the words of one commentator, and it -would be unimaginable to perform it outside the context of the cycle.3 In other words, there is, if anything, an inverse relationship between the development of the weak opening in each successive song and the ability of the song to stand alone as an independent piece. Schumann's exploration of the weak opening helps us understand the rationale behind the order in -which he composed "Mondnacht," "Intermezzo," and "Schone Fremde," but it does not suggest that he ever intended to publish them
192
From Song
in this order. His decisions about how to arrange the Eichendorff songs into a cycle, like his decisions about the order in which he composed the songs, were inevitably based on a myriad of complicated and varied considerations, and the best we can hope for is a tentative and partial understanding of the reasons for his choices. I have suggested that one reason Schumann placed "Mondnacht" and "Schone Fremde" together as a consecutive pair is that there are significant parallels between their texts and musically they are each a recomposition of the same underlying formal structure. The fact that Schumann did not compose these songs consecutively does not weaken the hypothesis that he intended them as a recompositional pairing. On the contrary, by considering the intervening draft of "Intermezzo" we can gain further insight into the way in which Schumann reworked particular aspects of their shared structure as he turned to the composition of "Schone Fremde," and we can illuminate some of the •ways in •which the pairing is related to the other songs in the cycle. If -we consider the first six drafts of the Eichendorff Lifderkreui as a group, then the openings of the songs reveal a different kind of compositional logic, as I have already observed. Schumann first composed three songs that have strong, somewhat generic openings and then another three with weak openings. These six songs then became the first half of the cycle, but Schumann did not simply leave them in the sequence of their composition. If he had, he would not have presented the contrast between the two kinds of openings in a particularly effective manner. By arranging them as he did, so that Songs 1, 3, and 4 have strong openings, while Songs 2, 5, and 6 have -weak ones, Schumann creates an alternation between strong and weak openings, and this formal idea becomes a way of organizing the first half of the cycle. As I have already acknowledged, my speculations about Schumann's process of composing and arranging the first six Eichendorff songs present a very incomplete picture. I have ignored the texts and have limited my account of the music to a single element, which mostly involves the accompaniments of the songs. Furthermore, as I pointed out in my discussion of the manuscript sources for the cycle, it is likely that Schumann composed at least some vocal sketches before he turned to the piano drafts, and so it is possible that the dates on these drafts only indicate the order in which Schumann wrote down the accompaniments and not the order in which he first decided to set the songs or wrote the melodies. But in a sense, the fact that my account is inadequate is precisely my point. I do not believe that there is a single straightforward relationship between the order in which Schumann composed the songs of his cycles and the order in •which he published them. The more elements that we consider as we study the songs, the richer and more variegated our explanation of the process will become. Schumann usually began composing his song cycles through an amorphous process of selecting poems that appealed to him (or, in the case of the Eichendorff cycle, to Clara Wieck) and turning them into songs. There are inevitably a variety of relationships between the poems that Schumann chose for each cycle and between the songs that he set them with, and when he arranged the songs into a particular sequence for publication he naturally found ways to
Schumann '<* ProceM of Composing a Cycle
193
emphasize and exploit these relationships. But just as there is no one reason for the various decisions that Schumann made in the process of composing and arranging a given cycle, there is no one explanation for the way in which that cycle coheres. In the words of his colleague Oswald Lorenz, the song cycle is a "loose coherence, . . . [an] exposition only lightly hinted at," and it is precisely in this flexibility and subtlety that the power and beauty of Schumann's cycles lie.33 As we shall see in the next chapter, the freedom with which Schumann arranged his songs into cycles reflects the literary tradition from which they emerged, and it is indicative of his sensitivity to the poetic texts that he set.
This page intentionally left blank
THE SONG CYCLE AS A LITERARY WORK
I
t is clear from the song reviews that -were published in the Neue Zeifochrift fur Mu
195
196
From Songd into Cycles The lied is the purest expression of our feelings, presented in the most natural form; it should truly express what one truly feels, but with both the appeal of novelty and the dignity of simplicity; it should speak to us with those little signs of authentication of the composer's own feeling, which irresistibly touch and ignite kindred souls. . .
Banck admonishes us that the composition of lieder requires "deep selfcontemplation and feeling, the penetration of the language of nature and the human heart," since a composer -who has "never really considered the roots of his own feelings," and "who cannot use his powers of intellect to keep hold of life's onrushing emotions, will also be incapable of truly representing them." Interestingly, Banck makes no reference at all to the means through which this representation should take place. Although he begins by describing the lied as the child that is born from the love affair of music and poetry, these arts are conspicuously absent -when he begins to discuss what is required in order to create one. It is as if he wants to emphasize that it is not enough to excel at one's chosen art form in order to compose lieder. First and foremost, one must be a "complete person" who possesses a "sensitive spirit. 4 Later in the article, as Banck describes those lieder that represent the "highest attainment" of his ideals, he introduces the act of composing into his discussion in a way that reiterates the same point even more forcefully: One feels the contrivance, the experience, the craft in these compositions; they are the lieder of those practiced composers, who have the greatest e,(ue in handling the ordinary methods as well as their own characteristic methods for rendering the truth of feeling in music. The concern here is not the rebirth of poetry into the more beautiful and glowing life of song; rather, the composer has interpreted the words in his musical language, whose fluency provides his eloquence. He shrouds the radiant glances of his spirit with the fog of serenity, and expresses poetic language instead of literary ideas.5 In this passage, Banck is careful to bring out the relationship between the craft of composition and the true expression of feeling. When we hear a lied, it should sound like "the purest expression of our feelings, presented in the most natural form," but this is an effect that is created through the techniques of musical composition, and it requires practice and experience. Likewise, a poem is not miraculously reborn as music when it is set as a lied; rather, the composer must consciously and skillfully translate from the language of poetry into the language of music. A corollary of this is that the technical craft of composition is not an end in and of itself but the means through which the truth of feeling is expressed. A composer must always have this end in mind and make sure that he does not unnecessarily introduce extravagant compositional effects. Banck's article reads like a statement of purpose, and -we find echoes of his ideas in a number of subsequent Neue Zeitechrift song reviews. On the rare occasions when a critic refers to compositional technique, for example, it is typipally
The Song Cycle a>) a Literary Work
197
in terms of how the composer is responding to the text he is setting. Schumann, in his 1836 review of Carl Lowe's song cycle Either, describes how Lowe captures the "special tone" of each text and responds to the events of the narrative that are being related through his choice of key and mode, as we have already seen in chapter 3.6 Another Neue Zeitjchrift critic, writing a few months later, also makes a connection between tonality and text expression when he criticizes a setting of Heine's "Wenn ich in deme Augen seh'" by a composer named Kulenkamp. The reviewer objects to the large number of modulations within the course of a seventeen-measure song, which is primarily a compositional problem, but he expresses his criticism in terms of the demands of the text. He describes the song as "a little genre picture, in -which the great extravagance of strong lighting effects is not proportionate to the narrow frame."7 In other •words, the composer is indulging in showy compositional effects at the expense of an appropriate expression of the modest poem that he is setting. Schumann makes a similar criticism in an 1842 review of Henry Hugh Pearson's setting of six Burns poems: "To put it briefly, it appears to us that too much extravagance is squandered on this text; there are too many notes for such simple words."8 When the Neue Zeitdchrift critics discuss the creation of a song cycle out of a series of lieder, they again emphasize the literary aspect and evaluate it in terms of how sensitively the composer expresses the emotional and spiritual content of the text. In an 1844 review of Hieronymous Truhn's Liebedroman in 12 Liedern, •which sets poems by Heine and Emanuel Geibel, the critic begins by praising the "ingenious idea of putting together various lyric poems and attaining a narrative progress through the contrast of their distinctive emotional states" and later finds that the "chief merit of the composer" is "in the creative selection and combination of the beautiful poems."9 Similarly, in his review of Ferdinand Killer's setting of six poems from Riickert's Liebettfrilbling, which I have quoted in full in chapter 3, Oswald Lorenz observes approvingly that Hiller has selected and arranged the texts himself. In this case, however, the composer has not created a narrative progression but a "loose coherence," an "exposition only lightly hinted at," and it is precisely the suggestiveness of the arrangement that Lorenz finds appealing, because it enables Hiller to bring out the "sensitive and spiritual nature" of the poetry. As in Banck's 1835 article, Lorenz's review makes it clear that Killer's ability to express the truth of feeling is at least as important as his technical skill as a composer, if not more so. Near the beginning of the review, we read that "it is not the capability for musical treatment alone that determined the selection" of the poems but the creation of "a delicate spiritual thread [that] runs through the series." And toward the end Lorenz praises "the meaningful selection of the poems" and finds that it is not only "the ease and sureness of the work" but also "the tactful handling of seasoning and ornament" that reveals the composer's "skillful hand." Lorenz concludes, characteristically, by considering the lieder "from a purely human viewpoint" and finds them "so peaceful and cheerful in nature, so light and friendly in their expression of the phases of emotional life, that they can count on securing many warm friends."10 These reviews, which are among the very few in -which Schumann or his
198
From Songj into Cycled
colleagues discuss the creation of a song cycle, suggest that they conceived of the genre in much the same terms as they did the lied itself. The composer's primary purpose should be to faithfully express the emotional truth of the poetry that he is setting, and in the case of the cycle the selection and arrangement of the poems becomes another means to this end. Recent Schumann scholars, with the notable exception of Arthur Komar, have also defined the song cycle largely in terms of its text, but their approach has been quite different from that of the Neue Zeifacbri/t critics. Charles Rosen, for example, argues that Schumann reserved the term CycLiu exclusively for Dichterlie.be and FrauenLiebe und Leben because each of these works is "a monodrama, in which there is a single speaker and at least the skeletal suggestion of a narrative." This is characteristic of our modern-day conception, and it is in these terms that the Eichendorff Luderkreui is often judged to be anomalous. Jiirgen Thym, for example, compares the latter work to the same two cycles and observes that we cannot find a story in the series of poems that Schumann arranged as its text because there are no "characters which would be capable of action." And Patrick McCreless, too, distinguishes the Eichendorff Luderkreui from more typical cycles because there is no "story" that "holds the texts together."11 In the present chapter, I -will argue that the song cycle was conceived in the nineteenth century as a lyric genre and that by trying to transform it into a dramatic -work with a coherent narrative we have misconstrued it, much as we have by trying to turn it into a unified musical whole. Dichterliebe is again the paradigm and the Eichendorff LieSerkreid the exceptional case, so I will consider the reception of each in turn and conclude with my own interpretation of the latter work. First, however, I will clarify the distinctions that were typically made between lyric and narrative genres by nineteenth-century critics and composers.
Of Liedenpielen and Liedercyklen It is implicit in Bancks article that the lied is really a subgenre within the genre of the song, which is distinguished by its use of a lyric poem as its text. In the last part of his article, after surveying a variety of lieder by numerous composers, Banck turns to two publications by Lowe, titled Legenden and Der Bergmann: Ein Liederkreut in Balladenform, even though they do not entirely fit within the "Liedergenre." Banck tells us that one reason he includes these -works is that much of what he has to say about the lied should apply to song compositions (GuangcompOftitwnen) in general. But one suspects that the more decisive reason is that he considers Lowe to be too significant a song composer to ignore. Referring to the dramatic nature of Lowe's song texts, Banck explains that he feels compelled to expand upon his stated ideals, because "Herr Lowe has chosen to wrap his fertile imagination around us in these garments." WTnle Banck praises Lowe's ability to present "sharp characterization" and "visual depiction" in his musical settings, it is apparent that he considers Lowe's choice of texts to be a -weakness that the composer only partially overcomes. Banck writes, for example, that the "Bergmannsliecler but in BalLuler&orm waver too much between the two" and complains that
The Song Cycle cut a Literary Work
199
they tell their story and depict the scene "almost serenely," -which is to say that they are caught between the lyricism that is expected in a lied and the drama that is demanded of a ballad. Banck continues: "If we now acknowledge at the outset the natural influence of the text on the composition, then we are firmly convinced that Lowe's individuality is more suited than any other to resolve this problem in the happiest -way." Banck praises numerous aspects of Lowe's settings but tempers his praise -with two criticisms that reveal his bias against the use of the ballad as a song text. He tells us that "the form suffers from the length of the poems" and that the accompaniment is "certainly created skillfully, but too full, overloaded with tone painting, which often restricts the vocal line and reduces it to a subsidiary role."12 The ideal against which Lowe's songs are measured is clearly the lied, •which should be more artless and less theatrical. Schumann also criticizes Lowe, as -well as Schubert, two composers whom he regarded quite highly, for expressing too much of the "material" world in their songs. His criticism is mitigated, however, by the fact that he is contrasting them with their contemporary Bernhard Klein, who, "to the same degree," holds back too much when the music becomes "sensuous" or "picturesque."13 Schumann is more definitive about the fact that one should eschew compositional and dramatic effects in the lied in his review of Pearson's Burns heder, which I have cited earlier. He describes the typical Burns poem as "an effusion of poetic feeling, but always simple, brief, and concise" and -writes that "his words practically compose themselves into a lied, and most naturally into that form that is typical of a real folk song." He then turns to Pearson's settings: But this composer wants more than this; he mostly gives us large pieces that are fully developed, and which reveal the striving musician, but are contrary to the naive form of the poem; often his music even has a dramatic or theatrical touch, and here he appears to be further than ever from his goal.14 Another Neue 2^eit
200
From Song
Given the care with which these critics distinguish the lied from songs that set narrative or dramatic texts, it seems inconceivable that they would expect the LiedercycLu to tell a story or to present characters capable of dramatic action. And when we consider the nature of the Romantic lyric poems that were typically set as lieder in the early nineteenth century we can see how difficult it would be to create a coherent drama or a narrative discourse out of them. The Romantic lyric takes place within the realm of emotion and memory and not in the material •world. It is essentially an introspective utterance, in •which a firstperson narrator expresses his inner feelings •with little or no reference to concrete events or clearly defined characters. There -were attempts, nevertheless, to create narrative and dramatic genres out of the lied in the early years of the nineteenth century, and it is instructive to consider their fates. The first such attempt was the LiedenpieL, a genre that Johann Friedrich Reichardt invented and then defended in a lengthy and somewhat polemical article that appeared in the Leipzig Ailgeineine miuikaluche Ze.itu.ng in 1801.'6 Reichardt s Liederjpielen, which •were intended as a response to the empty virtuosity and shallow theatricality of German opera, essentially consisted of a play — complete with scenery, costumes, and spoken dialogue—that included lieder •with texts by Goethe and other contemporary German poets, set in a strophic, folklike style. What is peculiar about these lieder is that Reichardt provided them with orchestral accompaniment and intended them for the theater, rather than for private performance. But he tells us, in his 1801 article, that his first LiederjpUl, Lieb'and Treue, began as a domestic entertainment that was performed at a family celebration, and he subsequently published the lieder with piano accompaniment and -without the spoken dialogue or instrumental interludes. The Liedertipiel enjoyed a brief vogue in the early decades of the nineteenth century, both in the form of theatrical productions and as published volumes of lieder that were intended to serve as the basis for private dramatic entertainments.17 Schumann himself briefly revived the genre in his so-called second Liederja.br, 1849, composing two Liederjpielen on Spanish texts translated by Geibel and a third on texts from Riickert's Liebe
The Song Cycle cu a Literary Work
201
out of the lied that -would be comparable to Reichardt's Liederdpiel, we do occasionally find the use of the analogous term Liederroman, or simply Roman, to indicate that an early nineteenth-century song publication is intended as a coherent narrative. The most ambitious examples are two works on texts by the Berlin poet Christoph August Tiedge—Alexis und Ida: Ein Schaferroman in 46 Liedern, composed in 1814 by Friedrich Heinrich Himmel; and Gedange und Lieder zu Tiedgej Lifder-Roman Aennchen and Robert, composed in the following year by Sigismund Neukomm. Although these works were clearly intended for private performance and not for the theater, they are actually more dramatic than narrative in form.20 In each case, the two lovers are treated as separate characters who mostly sing alternating solo songs and occasionally get together for a duet, and there are some part-songs for additional characters as well as a concluding chorus. But there is no spoken dialogue, as in Reichardt's Liederdpielen, and in order to convey a complete story in a series of lieder Tiedge had to write a very large number of poems. Alexis and Ida consists of forty-six lieder and Aennchen and Robert an astounding eighty-eight, although in his musical setting Neukomm reduced the latter work to only thirty-three. The result is a long series of similar songs that, as one contemporary review of Alexis and Ida points out, nobody is likely to sing from beginning to end. However, the critic continues, if one selects individual songs out of the Roman, then one sacrifices the coherence of the whole.21 Barbara Turchin observes that the term Roman continues to be used "•well into the nineteenth century to describe song collections of a narrative nature,"22 but it is clear that as a genre, the Liederroman. -was even less successful than the Liederdpiet, and it did not enjoy even a brief spell of popularity. Recent historians of the song cycle have emphasized the connections between genres such as the Liederjpiel and the Liederroman, on the one hand, and the LitdercycLu or Luderkreu, on the other.23 But the contrasts are in many ways more interesting and can help us to understand how the latter genre is distinctive and why it is the only one of the three that is still of interest today. Let us reconsider, for example, the beginning of Arrey von Dommer's 1865 definition of the song cycle, which I have already quoted in full in chapter 1: Lieierkreu, Liedercydiu. A coherent complex of various lyric poems. Each is closed in itself, and can be outwardly distinguished from the others in terms of prosody, but all have an inner relationship to one another, because one and the same basic idea runs through all of them. The individual poems present different expressions of this idea, depicting it in manifold and often contrasting images and from various perspectives, so that the basic feeling 24 is presented comprehensively. Dommer reminds us that the nineteenth-century song cycle is essentially a lyric genre and that its text consists of a series of individual poems that are not intended to have the kind of narrative continuity or stylistic consistency that we expect from a play, a novel, or an extended poetic form such as a ballad. Although he begins his definition with the phrase "coherent complex" (ziuammen-
202
From Songj into Cycles
bcingender Complex), there is no implication that the coherence he has in mind is dramatic or narrative; rather, he simply refers to an underlying idea or feeling (Grundgedanke, Grundgefiikl) that is treated by all of the poems. This suggests a literary work that is different in kind from the Liet)er
The Song Cycle Oil a Literary Work
203
the physical -world and become reunited with the spiritual force of nature that are expressed in so many of the Eichendorff songs may sometimes appear in the form of a personal testimony, but this is still intended as a depiction of our common state. Similarly, the organization of the cycle does not derive from the story of an individual protagonist but from the tripartite schema of human history, so ubiquitous in Romantic literature, in -which we have fallen from a state of natural innocence and satisfaction into the more sophisticated but alienated state of modern existence and from here strive to attain a higher level of unity and selfknowledge.26 The natural world also plays a central role in the Wanderlieder cycle, both thematically and structurally. Poets and composers often use the rhythms ol nature as an organizing principle, either by depicting a particular landscape, such as the sea or the forest, or by alluding to seasonal and diurnal cycles. The idea of using the temporal cycles of nature as a device for arranging a series of songs or instrumental pieces into a musical cycle was already common in the eighteenth century.27 But where these earlier works usually present the stages of a single cycle in succession—the four seasons, the twelve months — and thus create a sense of completeness and closure, the Wanderlieder cycle typically emphasizes the unceasing cyclical movement of nature's time and turns it into a metaphor for the Romantic view of human existence as an ongoing quest for spiritual fulfillment and renewal. The progression from evening to nighttime or from -winter to spring may determine the order of two or more successive songs but usually not of an entire cycle. It is the cyclical aspect of the seasons' passage, and not their particular sequence, that is of significance to the Romantic poets, and so it is often merely through a reference to spring as a time of rebirth or to winter as a time -when life becomes quiet and hidden that the individual songs of a cycle suggest a larger, unifying context. Dichteriiebe, the Story If we consider the Eichendorff Lieclerkreu) as an example of the WanderLitder cycle, it becomes clear that there is nothing unusual about the -way in which Schumann arranged Eichendorff's poems into a song cycle text. The reason that scholars have considered this text to be unique is that we have come to regard the presence of a narrative discourse as one of the defining criteria for Schumann's cycles. This idea, which apparently originated within the twentieth century, directly results from the paradigmatic status of Dicbterliebe and Frauenlie.be uru) Leben. It is easy to see how this criterion applies to the latter work, in which the events of a woman's courtship and marriage are presented in the chronological order in which they occurred. But even in this case, it is not entirely accurate to speak of a narrative, as Rosen has observed, since the story that is told is a typical woman's life as idealized by a nineteenth-century man and the narrator is more an abstraction than an actual character.28 In any case, Frauenliebe and Leben is, if anything, more exceptional than the Eichendorff LiederkreLf, since it is one of Schumann's least Romantic cycles and sets poetry that is dramatic rather than
204
From Sought into Cycle*)
lyrical. The notion that Dichterliebe is a narrative discourse is far more problematic, and yet it is in the extensive scholarship on this cycle that the idea has been developed most fully. And so the reception of Dichterliebe once more provides the clearest example of how we have come to conceive the Romantic song cycle in the late twentieth century and demonstrates -why our conception is so much at odds with that of Schumann's day. A recurring idea in the literature on Dichterliebe is that in the process of choosing •which poems to set from Heine's LyrLtched Intermezzo Schumann extracted a clear narrative that Heine only vaguely implies. Rufus Hallmark argues that as Schumann reduced Heine's sixty-six poems first to twenty-nine, then to twenty, and finally to sixteen, he avoided or downplayed a number of themes that are central to the Lyruche
The Song Cycle cu a Literary Work
205
TABLE 8.1 The Plot oiDichterliebe, according to Turchin and Lewis Turchin
Lewis
I. Songs 1—6 exposition "the poet tenderly recalls a past love"
I. Songs 1-4 "the poet expresses his love . . . in the present" "direct mode of address"
II. Songs 7-11 development and crisis wedding in 9 is "dramatic center" 11 "objectifies . . . and provides a transition"
II. Songs 5-10 "the love is clearly lost" and exists in memory beloved is invoked "in images and symbols"
III. Songs 12-16 "the poet attempts to reconcile his feelings"
III. Songs 11-16 11 retells and generalizes the plot in a parable 12 — 16 do not add to story but comment on it "dreams and their fusion of past, present and future"
lected from the Lyruchej Intermezzo, and the story is the sequence of events that is the subject of this discourse. We can describe the latter as follows: 1. 2. 3. A.
The poet falls in love. He and his beloved carry on a romance. She ends the relationship. She marries someone else.
When we read the story in this straightforward form, it becomes clear that the order of the songs in Dichterliebe is not derived from the chronological order of the story upon which it is based, and, in fact, the events of this story are scarcely mentioned in any explicit way. The real subject of Dicbterlie.be is the poet's emotional and psychological response to the story, and the sequence of the cycle does not depend upon the logical conventions of storytelling but on the far more elusive laws that govern the unconscious realm of dream and memory. It is true that the poet starts off as if he is going to tell a story, which begins when he fell in love, "in the lovely month of May." But he interrupts himself almost immediately, and over the course of the next three songs he tenderly pleads with his beloved, expresses his love for her ecstatically and almost hysterically, and mocks her with bitter irony. He is no longer telling how the story began but responding in the present moment to its still untold ending. And he is clearly not speaking directly to his beloved or remembering an actual conversation that took place in the past but simply talking to himself. In order to hear these opening songs as a tender recounting of a happy love "in full bloom," with scarcely a hint of the tragedy to come, as both Lewis and Turchin suggest, one would have to be a willfully obtuse listener and ignore the continual changes in tone and nar-
206
From Songj into Cycled
rative voice, the subtle (and not so subtle) poetic inflections, and, of course, the sensitive musical setting, which communicates the poet's pain as clearly as the text does. I suspect that the reason these two scholars insist on such an implausible interpretation is that they are straining to turn Dichteriu.be into a narrative discourse, in -which the form is created by the dramatic unfolding of a. series of events and the gradual revelation of truth. So for Lewis it is only -with the fifth song that the poet suggests that he has lost his love and only in the seventh song that he reveals that she is "eternally lost." And for Turchin the crucial dramatic moment is the ninth song, "in which the beloved's marriage to another is revealed. " But while Songs 7 and 9 are certainly decisive moments in the cycle, it is more because they plumb the furthest to the depths of the poet's anger than because they reveal anything of significance about the story. They are, in other words, moments that mark a particular stage within the poet's psychological process and not a particular point in the sequence of the story. Kven though the poet describes the -wedding of his beloved in the present tense in "Das ist ein Floten und Geigen," it is clear from the context of the cycle that he is not literally "witnessing it; rather, his feeling of anger has brought the scene vividly to his memory. And while it is true that this song presents the first intimation that she has married someone else and is thus the kind of revelation that is typical of a narrative discourse, "we still do not perceive this scene as an event within an unfolding story, because, after the first song, the wedding is the only event that the poet describes in the entire cycle. We could perhaps consider Dickterliebe as a narrative in a metaphorical sense — it is not a recounting of a series of events but an inner psychological narrative. Rosen, for example, writes that Schumann's cycle "moves from the awakening of desire, through love, deception, rage, and despair, to a bittersweet ending of cynicism and regret." But even this description fails to capture the complexity of the cycle. Dichterliebe takes place entirely within the poet's memory, which does not operate in normal time. His feelings of love and desire, his rage and grief at his loss, his attempt to come to terms with the situation and attain an emotional distance—all of these conflicting feelings commingle in his memory throughout the cycle, creating a level of tension that coexists with and works against the logical sequence of emotions that Rosen describes. This is "why Rosen's claim that Dichterliebe is a cycle because its "order is chronological, each successive song representing a later moment than the preceding one," is problematic.32 In the fifth song, for example, the poet fantasizes about an erotic union "with a lily that leads the flower to "breathe a song of my beloved," which in turn reminds him of her kiss. Then, in the sixth song, a portrait of the Virgin ironically reminds him of the face of his beloved. In -what sense does the latter image depict "a later moment" than the former? Both are the product of the poet's imagination, and both remind him of a past time when he felt very differently about his beloved. But there is no temporal relationship between them, and our understanding of the two poems is not dependent upon the order in which we hear them.
The Song Cycle ad a Literary Work
207
There is another problem with interpreting Dichtertiebe as a narrative discourse, -which concerns the question of narrative closure. In "Die alten, bosen Lieder," which is the final text in both Heine's and Schumann's cycles, the poet describes how he will resolve the psychological tension of his unrequited love by burying "the ancient evil songs" and "the evil and angry dreams" in the sea. But he also makes clear the impossibility of the task, since his love has caused him so much grief that it would require a coffin larger than the great cask of Heidelberg, a bier longer than the Mainz bridge, and twelve giants to act as pallbearers, who must be stronger than Saint Christopher's statue in the Cologne cathedral. It is thus an ironic conclusion that denies the possibility of resolution, and one cannot help but read it as an implicit death wish. The Heine scholar S. S. Prawer argues that the poem is not really an integral part of the lyric cycle at all but a framing commentary in the poet's own voice. Prawer points out that the central metaphor of the poem, in which "the finished book, of the collection of lyrics, [is] a coffin at once of the songs it contains and of the emotion which inspired the songs," is one that Heine uses to conclude several other sections in the Buck der Lieder, as well. It is, in other words, a personal stylistic convention that enables Heine to provide a kind of rhetorical closure for the cycle, even as he prevents it trom attaining narrative closure. The piano postlude that Schumann uses to conclude his setting of the poem can also be heard as a commentary on the cycle, presented in the composer's own voice, that in some sense stands apart from the rest of the song. Schumann scholars have almost universally agreed that the effect of the postlude is to soften or even contradict the tone of bitterness and renunciation in Heine's poem and to substitute a sense of forgiveness and reconciliation. They have pointed to the fact that the postlude is in the major mode while the rest of the song is in minor; that it provides harmonic and melodic closure, which is lacking at the conclusion of the vocal line; and that it is a return of music that was first heard at the end of "Am leuchtenden Sommermorgen," a song in which the poet imagines that the flowers are encouraging him to forgive his beloved. In this way, it is argued, Schumann provides Dichteriiebe with the closure that Heine's original cycle lacks.34 But -while the ending of Schumann's cycle is more tender in affect than Heine's, it is no more conclusive. The poet sings that he has overcome his feeling of loss by sinking his love and his suffering into the sea, but Schumann emphasizes the doubtful nature of his claim by ending his melody tentatively, on 6. The postlude that follows is a wistful musical reminiscence that does not contradict the poet's failure to achieve psychological resolution but confirms it. The extraordinary effect of this postlude is largely due to the unusual nature of the return. It is not the vocal melody of the earlier song that comes back but its postlude, -which is unrelated to the vocal melody, and so, in both of its incarnations, this is music that is presented solely by the accompaniment, as an afterthought or commentary to each of the songs that it concludes. As Rosen astutely observes, it is for this reason that the postlude cannot provide the cycle with musical or narrative closure:
208
From Song
What returns is a postlude heard only a few songs back, three quarters of the way through the cycle. The return is consequently unmotivated by any convention of form or even by the demands of the text. It seems to be spontaneous, an involuntary memory, governed by a law of its own.35
The Reception of the Eichendorff Luderkrei) It is the narrator's obsessive focus on his ongoing inner quest to achieve psychological closure as well as the discontinuous and open-ended form in which this quest is presented, and not the presence of a unified narrative voice or the implication of an underlying story, that make DichterLiebe a Romantic song cycle. There are numerous examples of nineteenth-century song cycles, including Schumann's Eichendorff Liederkreid, in which there is no single narrator, nor even the hint of a plot. But even as scholars have acknowledged that the text of the Eichendorff cycle is not a narrative discourse in the traditional sense, they have relied on narrative as their point of departure, nevertheless, and have still looked for elements such as order, closure, and point of view to provide coherence for the cycle. In general, two strategies have been adopted in trying to understand the text of the Eichendorff Licderkrei). One is to subsume the songs of the cycle within the larger narrative of Schumann's life, so that the cycle becomes an explicit expression of his love for Clara Wieck. The other is to explain the cycle as a progression of moods that ultimately leads from the nostalgic melancholy of the opening song to the ecstatic jubilation of the final one. Although these strategies have enabled us to understand Schumann's cycle as a kind of quasi narrative, they have done so at the cost of ignoring the central themes and images of Eichendorff's poetry. One of the earliest modern discussions of Schumann's Eichendorff Liederkreu is Theodor Adorno's "coda" to his essay "Zum Gedachtnis Eichendorffs," which was presented as a radio broadcast on the centennial of the poet's death, in November 1957, and published the folio-wing year. Adorno already presents the central ideas that have dominated the reception of the cycle over the last forty years. He argues that there is a close relationship between the sequence of keys within the cycle and the arrangement of its poetic texts and traces a circular "modulatory path from the melancholy of the first [song], in Ft minor, to the ecstasy of the last, in the parallel major." He also uses tonality and text to divide the cycle into two parts, which have "the simplest symmetrical relationship, with the caesura after the sixth song." "Schone Fremde," in B major, "ends -with the Utopian dream of great future happiness, with presentiment \Ahnung~]," and the last song of the cycle, "Friihlingsnacht," up a fifth in Ft major, ends "with jubilation: 'Sie ist deine, sie ist dein,' with the present [Gegentvart]."36 Adorno suggests several different ways in which the final song provides the cycle -with narrative closure. First, its text describes the fulfillment of the future happiness that is predicted at the end of the first half. Second, the song returns to the tonal center -with which the cycle begins -while simultaneously moving to the farthest key on the sharp side of the circle of fifths that Schumann uses. Third, the climax of
The Sotig Cycle
209
the key progression corresponds to the strongest contrast in mood. As other scholars have considered the coherence of the Eichendorff Liederkre'w in recent years, they have essentially reiterated and elaborated on Adorno's basic conception. Turchin, for example, writes that "a progressive heightening of emotional states transpires over the course of the cycle," Thym describes "a. sequence of Stimmungen reaching from deepest melancholy to utmost ecstasy," and John Daverio finds "a great arch from melancholic alienation to ecstatic union •with the objects of the poet's longing." All three -writers make an explicit connection between the succession of moods and the sequence of keys, and they all divide the cycle into two halves, the first ending with the promise of future happiness and the second concluding -with the fulfillment of this promise.37 While Adorno's ideas have thus enabled scholars to impart a sense of order and closure to Schumann's Eichendorff cycle, some commentators have still been bothered by the lack of a consistent narrative point of view, a plot, and clearly identifiable characters. One solution has been to explain the cycle in biographical terms, so that Schumann himself becomes the implied hero. Thym, for example, has argued that the composition of the Eichendorff songs was "of considerable autobiographical significance for Schumann" and, in particular, that the choice of texts was influenced by his anticipated marriage, which took place just four months after he composed the cycle. It is for this reason, according to Thym, that Schumann selected poems "dealing with various aspects of love." And while "it is difficult to extract a well-defined story from the succession of poems," it is still the case "that the relationship between man and -woman is central" to the cycle. The clear implication is that if we focus on those aspects of the texts that are concerned with love and marriage, we can understand the Eichendorff cycle as a reflection of Schumann's own life story at the time that he composed the songs. Thus Thym points to the use of -wedding imagery in "Auf einer Burg," "Im Walde," and "Mondnacht," he describes "Intermezzo" and "Die Stille" as love songs, and he assumes that the future happiness referred to in "Schone Fremde" is the consummation of a love relationship, -which is finally fulfilled in "Friihlingsnacht."38 McCreless takes this idea one step further and argues that Schumann's impending marriage was responsible not only for the choice of poems but also for their order within the cycle. In fact, it is this circumstance that supplies him with his explanation for why Schumann decided to rearrange the order of the songs. McCreless divides the Eichendorff songs that "deal with love or with women" into two categories. The first consists of "more-or-less conventional songs of love, positive in tone, with fulfillment and/or marriage, explicit or implied," -while the second category includes songs "that take on a more ironic, questioning, and even bitter tone with regard to matters of love and marriage." McCreless finds that in the so-called "compositional ordering" it is the latter view that predominates, as the cycle "takes us from uncertainty and premonition in 'Waldesgesprach' to the sobbing bride of 'Auf einer Burg.'" He speculates that at some point Schumann decided that this order of the songs "projected a negative view of love, and indeed, painted a picture of marriage that was too grim for comfort"
210
From Sonfft) into Cycles
in the year of his wedding. So he "first moved 'Friihlmgsnacht,' the happiest song of all, to the end, thereby putting as bright a face on the matter as possible," and ultimately came up with the published order of the cycle, in which the negative songs "are placed in the larger context of the real Schumann love songs," as we progress "from isolation in the first 'In der Fremde' to marital union in 'Friihhngsnacht.' "39 Such interpretations of the Eichendorff Liederkreu are based on a curious assumption about the relationship between Schumann and his art. The extent to which a -work reveals something of the artist who created it has been a recurring theme in critical scholarship, at least since the advent of Romanticism in the early nineteenth century, but what is unusual in the case of the Eichendorff songs is that scholars have focused exclusively on the external circumstances of Schumann's life and not on his personality. These songs were composed in his "-wedding year," as Eric Sams keeps reminding us, and this is -why Schumann presumably read these texts as love poems, minimizing or even ignoring the far more significant theme of spiritual transcendence. It is Sams who states this idea most explicitly and then uses it to criticize Schumann's setting of "Mondnacht." Sams observes that Eichendorff's "poem is about spiritual communion with Nature" but insists that "Schumann is expressing its bridal imagery." He goes on to note that "Schumann's inspiration was always love, never religion," so it "would not be surprising if his music faltered, if only a very little, at a conclusion about the soul winging off to its home."40 Sams, like McCreless and Thym, supports his argument by citing Schumann's letter of May 22, 1840, to Clara Wieck, in •which he writes: "The Eichendorff cycle is certainly my most romantic and there is much of you in it."41 But -we can interpret this comment on a number of levels. Schumann may have been thinking of the fact that it was Wieck who selected the poems and wrote them out for him. Or he may have meant that in some general way she was his source of inspiration, and so there was much of her in all of the music he composed. Regardless of-what Schumann meant by this comment, which is, after all, a personal communication and not a statement of aesthetic intent, I think it is misguided to take it as evidence that in composing the Eichendorff songs his personal circumstances restricted his sensitivity to the texts. It is precisely because creative artists are able to universalize their own experiences and feelings that their work is of enduring value. It is not because he was expressing his love to his fiancee that we continue to perform and listen to Schumann's Eichendorff songs today but because his love inspired him to great artistic heights. The circumstances of Fraueniie.be andLeben'^s composition, and its subsequent reception, provide a useful point of contrast. On July 7, 1840, Friedrich Wieck filed a declaration with the Leipzig court that represented a major defeat in his legal attempt to prevent his daughter from marrying Schumann. Having been told six months earlier that he must offer evidence to prove his charge that Schumann was a drunkard, Wieck now conceded that he had none. He still withheld his consent to the marriage, however, so Clara and Schumann had to -wait until the court issued a formal decision in their favor. But, as Daverio tells it, the case
The Song Cycle aa a Literary Work
211
was essentially over, and the young couple had won. Schumann wrote: "Hurrah! Victory!" in his Haudhaltbuch entry for July 7, and the next day he and his fiancee 'Were apartment hunting." Daverio writes that, for Schumann, "the news of Wieck's concession unleashed the most powerful wave of creativity of this incredibly productive year," beginning with the composition of Fmuenliebe and Leben on July 11 and 12. Schumann's decision, four days after he had declared victory over \Vieck, to set these texts by Chamisso, which are explicitly and exclusively focused on marriage, certainly must have been a celebration of his impending wedding. But even in this case, there may have been deeper levels to Schumann's motivations, as Ruth Solie reminds us. She observes that Wieck still had the option of appealing the ruling, -which would inevitably mean spreading even more "scurrilous and slanderous reports of Schumann's character." For this reason, she suggests, Chamisso's texts "with their reassuringly noble and complacent portrait of a young husband must have fairly leapt off the page at Schumann. However we interpret Schumann's decision to set Chamisso's poems, it seems compelling that in this instance he was heavily influenced by the immediate circumstances of his life. And the poems that he chose — in which the woman is depicted as so devoted to her husband as to be quite literally selfless—have come to be regarded as among the weakest of all Schumann's song texts. The objectionable -way in which gender relations are portrayed in the cycle makes it "awkward" for us to listen to, as Solie observes, and the fact that it nevertheless continues to be a staple of the lied repertory only demonstrates Schumann's extraordinary ability to transcend an inferior text through his musical setting. However, in choosing the text in the first place Schumann doomed Fraiienliebe and Leben to, at best, an ambivalent reputation, because Chamisso's poetry is so limited in and of itself that it cannot possibly universalize his experience. Sams has it exactly backward -when he defends the cycle as follows: "In any event, the objection that a modern woman takes a quite different view of her love-life seems hopelessly irrelevant; -what these songs express is the quality of love that Robert and Clara Schumann in fact had for each other."44 To regard the cycle as a document of Schumann's relationship with Clara Wieck is to argue for its historical value. But as soon as we acknowledge that the view of women that is presented in the cycle is offensive to us, then we are admitting that its artistic value is in some way impaired. In the case of Frauz.nlLe.be. and Leben, we look to Schumann's biography to help us explain why he made what we consider to be a poor artistic decision. By contrast, the choice of Eichendorff's poems—which, we should remember, was Clara ^Vieck's and not Schumann's—does not require an external explanation, because we consider them to be among the finest poems that he set, and so we find adequate reasons for their selection in the texts themselves. This does not mean that the circumstances of Schumann's life are necessarily irrelevant to the decision to set these poems or to our understanding of the songs that he composed. But -we should only consider these circumstances to the extent that they illuminate and broaden our understanding, and we should certainly not allow them to restrict our interpretation a priori.
212
From Song
Expressing the Ineffable Both of the interpretive traditions of the Eichendorff Lifderkreu that I have described in the foregoing section —the ideas that the songs create a sequence of moods and that they are intended as expressions of Schumann's love for Clara Wieck — reveal no awareness of the revolution in the reception of Eichendorff's writings that began with a surge of scholarly interest at the centennial of his death in 1957. Because of his simple, folklike prosody, his seemingly carefree evocation of the German countryside, and his use of a limited number of stereotypical images, most scholars had dismissed Eichendorff as a popular poet who was not worthy of serious consideration. By the early 1960s, however, a substantial body of scholarly work began to appear in which Eichendorff's landscape \vas reinterpreted as a far more complex and meaningful place than it had seemed, and scholars came to understand his cliche-ridden imagery as a symbolic web, •which constitutes the very means through which he expresses his poetic meaning.45 The key to understanding Eichendorff's poetry is realizing that one cannot separate the thematic content from the scenic background, because all that -will be left is a vague depiction of mood and atmosphere. It is more fruitful to keep the setting in the foreground and consider the way in -which Eichendorff's images take on increasing significance as they recur in various contexts and are subtly transformed through his sophisticated play of language and sound. Recent attempts by musicologists to understand the literary coherence of the Eichendorff Liederkrew have depended upon just such a separation of setting and meaning, and so it is not surprising that the cycle has been reduced to nothing more than a sequence of moods. Thym, for example, describes the "nocturnal forest landscape" as "the scenic background in most of the poems" of Schumann's cycle and subsequently observes that "Eichendorff's poetry in general is full of images and motifs which have the character of formulas," among which Thym includes murmuring brooks, rustling treetops, moonlit fields, and singing nightingales.46 But his insistence that love between man and woman is the central theme of the Eichendorff Liederkreu) leads Thym to ignore the way in which Schumann uses these apparently formulaic images to create meaning and coherence in the cycle. The poems that Clara Wieck selected for Schumann are primarily concerned with our quest to transcend the stifling and alienating reality of the material world and to regain a sense of balance by merging ourselves with the spiritual force of nature. Eichendorff presents this quest, metaphorically, as a fictive landscape that is in a continual state of dynamic motion, which symbolizes the inner growth of the lyric subject. As Schumann arranged his settings of these poems into his opus 39 Liederkreu, he showed an awareness of the crucial role that Eichendorff's imagery plays in depicting our ongoing process of spiritual growth. If we enter Eichendorff's landscape and follow his images through the cycle of poems, they begin to reveal the underlying meaning of the text. In "Die Stille," for example, -which is the fourth song of the cycle, there are two essential images — the night and birds—both of which have symbolic associations that develop and
The Song Cycle cu a Literary Work
213
intensify as they recur in other songs. "Die Stille" is typically described as a love song, undoubtedly because Eichendorff suggests, in the first strophe of the poem, that the source of the narrator's hidden joy is the fact that she is in love:47 Es weifi und rath es doch Keiner Wie mir so wohl ist, so wohl! Ach, wiifit' es nur Einer, nur Einer, Kein Mensch es sonst wissen soil!
Nobody knows or could guess How happy I am! Oh, if only one could know it, No other person should know!
And yet this implied reference to her beloved is the only one in the entire poem, so he has a rather tenuous identity. In fact, given Eichendorff's use of the subjunctive, we could easily read him as a completely hypothetical character: "I am overflowing with happiness that I cannot express to anyone; if only there were one person who could know it!" If we read the strophe in this way, then i becomes clear that the narrator is not happy because she is in love; rather, she •wishes she were in love because that -would provide an expressive outlet for her happiness. Such a reading makes even more sense when we consider that it is not only the beloved who is absent in the two further strophes that Schumann set from this poem but also the theme of love altogether. In the second strophe, Eichendorff uses the image of a -winter night, a time when the natural world has been silenced and concealed, to elaborate on the idea of hidden joy. So still ist's nicht draufien im Schnee, So stumm und verschwiegen sind Die Sterne nicht in der Hohe, Als meme Gedanken sind.
The snow outside is not so quiet, Nor so mute and hidden are The stars in the heavens, As my thoughts.
From this image -we turn to another, with no apparent connection:48 Ich wiinscht', ich ware em Voglem Und zoge iiber das Meer Wohl iiber das JMeer und welter, Bis dafi ich im Himmel war'!
I wish I were a little bird And could fly over the sea Far over the sea and farther still, Until I were in heaven!
In fact, this final image brings us back to the opening strophe of the poem (which literally does return in Schumann's setting) and further confirms that -what the narrator is really yearning for is transcendence. Her inability to express her inner feeling of happiness makes her feel isolated, and where in the first strophe this leads to her desire to be united with another person -who will understand her, at the end of the poem she yearns to be reunited with all of nature, beyond her physical existence. The implied depiction of death as a means of transcendence recurs in several poems of the cycle. We have seen, for example, that it is a central theme in the opening song, "In der Fremde. 49 The particular image that Eichendorff uses to represent this theme in "Die Stille"—a bird — is significant because of both of its special abilities: flight and song. It is only the former
214
From Songj into Cycled
that is presented explicitly within the poem, where it provides the narrator with a literal means of escaping from her earthly life as well as a physical manifestation of her joy. But there is also an implicit contrast between the image of the •winter night, -which is described as "still" and "stumm," and the image of a bird, which is one of nature's noisiest creatures. And, of course, Eichendorff draws our attention to this contrast through the title of the poem, "Die Stille," which emphasizes the fact that the narrator's desire to be a bird stems from her inability to express what she feels within. As a bird she can express through song what she cannot put into words. One might argue that in my latter interpretation of Eichendorff's bird image I am reading into the text something that is never actually stated. But as -with many of his poems, the connections in "Die Stille" are not spelled out very clearly and the reader is forced to provide -what the poet merely implies. Similarly, within Schumann's cycle recurring images become infused with symbolic associations and create a context for our understanding of each individual poem, even if the image is not entirely concrete in any one of them. There is no bird in "Intermezzo," for example, but the image, in the second strophe, of the narrator's song soaring into the air certainly implies one and at the same time makes explicit the connection between song and flight. A kind of symbiotic relationship thus develops between the texts of "Intermezzo" and "Die Stille" that provides both of them with a greater expressive power than either would have by itself. The tension between physical silence and inner expression, which creates the basic conflict in "Die Stille," is resolved in "Intermezzo" as a paradox. The narrator expresses his feelings in the form of a silent song, -which is miraculously transmitted through the agency of flight. In "Wehmuth," the ninth song in Schumann's cycle, Eichendorff again uses the image of a bird to metaphorically depict the theme of inner expression, but with a very different twist: Ich kann wohl manchmal singen, Als ob ich frohlich sei; Doch heimlich Thranen dringen, Da wird das Herz mir frei.
I can sometimes sing, As if I -were happy; But secretly tears come through, And so my heart is freed.
So lassen Nachtigallen, Spielt draufien Friihhngsluft, Der Sehnsucht Lied erschallen Aus ihres Kafigs Gruft.
Nightingales, When the spring breeze plays outside, Let their song of longing sound From the vault of their prison.
Da lauschen alle Herzen, Und Alles ist erfreut, Doch Keiner fiihlt die Schmerzen, Im Lied das tiefe Leid.
All hearts listen, And all are gladdened, None feel the grief, The deep sorrow in the song.50
In this poem, Eichendorff severs the image by placing the nightingale in a cage. The song now takes flight -without the bird and does not express the sor-
The Song Cycle (U a Literary Work
215
row and longing that the caged creature feels. Similarly, the narrator's song cannot express his grief, and it is instead his silent tears, which no one else is aware of, that free his heart and lead to transcendence. Where the narrator of "Die Stille" hides her joy through silence, in "Wehmuth" the narrator hides his grief by singing as if he were happy. Despite the fact that he can sing aloud, the narrator of "Wehmuth" is essentially as mute as the narrators of "Die Stille" and "Intermezzo," since, like them, he expresses his true feelings silently. W'e can understand the image of the bird in these three poems as a way of representing the theme of meffabihty. Our innermost feelings—whether happiness, love, or grief— cannot be expressed in •words and are thus hidden from the rest of humanity. We can only share them -with a kindred spirit, -who intuitively understands that which is left unspoken. This is one reason that, in our current state, we are so isolated and human communication is so fraught •with deception. Dusk and Darkness Eichendorff also associates the themes of ineffability and transcendence with the image of the night. In the second strophe of "Die Stille," the narrator compares the winter night to herself because it, too, is quiet, but her comparison suggests that just as her silence conceals her true feelings, the night's stillness is illusory as well. The snow gives the appearance of death, but hidden beneath it is a landscape that is merely dormant and will come back to life in the spring. Similarly, the natural world is covered by the silent darkness of night, but as it lies sleeping, the realm of the spirit secretly comes alive in dreams. The juxtaposition of the night image in the second strophe and the bird image in the third hints at the cyclical process of transformation that turns winter into spring and night into morning. And the fact that the latter image is expressed as the narrator's •wish places this transformation -within her imagination and not in the real world.51 The following song in the cycle, "Mondnacht," makes explicit both the seasonal cycle and its imaginative dimension, as the sleeping earth dreams that all of the seasons unfold in a single night. The image of the night is no longer a static tableau, as in "Die Stille," but a dynamic process that is set into motion by the heaven's kiss. And this process gradually merges with the narrator's feeling of transcendence, as his soul, much like the imaginary bird in "Die Stille," flies back to its heavenly home. The idea that the night itself expresses the ineffable is just barely suggested in "Die Stille" and is then presented more strongly in "Mondnacht," but it really becomes explicit only in the sixth song, "Schone Fremde." Now the night literally speaks to the narrator, but "confusedly, as if in dreams," and its actual words are never revealed. It is not so much what the night has to say that is of significance but the sense that it is communicating, because it is this sense that leads to the feeling of epiphany and oneness with nature that the poem depicts. The entire network of themes and images returns in "Friihlingsnacht," the final song of the cycle. The nighttime landscape is again in a state of dynamic motion, which is associated with the process of vernal rebirth. The spring night causes the narrator to experience a moment of transcendence, in which he loses
216
From Songj into Cycles
his sense of himself and of the boundary that separates him from the surrounding -world. He does not specifically say that he cannot express his inner feelings in -words, but his description of being so overcome that he -wants to laugh and cry at the same time suggests that this is the case. And it is again the night, and not the narrator, -who speaks at the end of "Fruhlingsnacht," in part through the nightingales' song. Finally, Eichendorff reveals what it is that the night says: It expresses the narrator's own inner desire. The text of the Eichendorff Liederkre'u is largely about our yearning to transcend our earthly state, in -which we have become alienated from one another, cut off from the natural -world, and divided within ourselves. The various narrators of the cycle express this yearning, alternately, as a desire to be united -with a beloved person or -with all of nature or to return home, but at its root it is essentially a spiritual yearning. The home to -which Eichendorff refers is our spiritual home, -where -we came from before our births and where -we shall return after our deaths. And although some of the narrators describe the ecstatic feeling that they have already become one with nature, as in "Mondnacht" and "Fruhlingsnacht," it is merely a glimpse of the afterlife to come. Even the theme of love is presented in spiritual terms. In "Intermezzo" it is described as the union of two souls, beyond the limits of physical existence, and in "Mondnacht" and "Fruhlingsnacht" it is simply an aspect of the narrator's epiphany. In the two "In der Fremde" poems, the objects of the narrators' love are dead, and so the desire for union is also a yearning for death. The imagery in all of these poems is based on the cyclical structure of human history that the Romantics adapted from the Christian narrative of fall and redemption. The past is an Edenic time of oneness and innocence, to which we shall return, having attained a higher level of spiritual growth and self-knowledge, in the future. In several of the poems I have been disussing — "Die Stille," the first "In der Fremde," and "Wehmuth" —Eichendorff contrasts this paradise of past and future with the present, •which is a time of alienation and loneliness. But it is in several of the other poems of the cycle — in particular "Zwielicht," "Waldesgesprach," and "Im Walde"—that the current state of human existence becomes the central theme, a theme that Eichendorff presents through a very different set of images. If the night is a time of spiritual transcendence, -when -we can feel our souls become one with nature, then the evening, the transition between day and night, is a time of deception and terror, -when nothing is as it appears and -we cannot read the landscape of the natural world. In "Zwielicht," -which Schumann set as the tenth song in his opus 39 Liederkreui, Eichendorff uses both thematic and formal elements to develop the symbolic significance of the twilight. In the first strophe, he depicts the narrator's sense of alienation from his surrounding landscape through his ambiguous and disorienting use of poetic language: Dammrung -will die Flilgel spreiten, Schaurig riihren sich die Baume, Wolken ziehn wie schwere Traume — Was will dieses Grau'n bedeuten?
Twilight is about to spread its -wings, Eerily the trees stir, Clouds drift like heavy dreams — What does this horror mean?52
The Song Cycle cut a Literary Work
217
The very first image, of twilight spreading its wings, suggests that, rather than descend and cover the earth, it is about to float upward. Eichendorff creates a similar reversal in the third line, when he compares the drifting clouds to heavy dreams, implying that they have substantiality and weight. And in the question with which the strophe ends, he plays on the ambiguous connotations of the -word Grauen, -which literally refers to our feeling of horror but also alludes to the grayness of the light at dusk, which obscures the landscape.53 It is, in other words, our very inability to recognize and make sense of the surrounding world that makes the moment so terrifying. As the poem continues, Eichendorff identifies his setting as the woods, a place of wildness and danger, and he fills his landscape with people, who are conspicuously absent in his poems of the night. In the second and third strophes, he suggests that our detachment from the world of nature stems from the mistrust and dishonesty that is inherent to the human -world: Hast em Reh du, lieb vor andern, Lafi es nicht alleine grasen, Jager ziehn im Wald' und blasen, Stimmen hin und wieder wandern.
If you love one doe above all others, Do not let it graze alone, Hunters roam the -woods and blow horns, Voices wander now and then.
Hast du einen Freund hienieden, Trau ihm nicht zu dieser Stunde, Freundlich wohl mit Aug' und Munde, Smnt er Krieg im tiick'schen Frieden.
If you have a friend here below, Do not trust him at this hour, Friendly, perhaps, with glance and word, He plots war in deceitful peace.54
The hunters -who roam the -woods and blow their horns constitute another image that has great symbolic significance for Eichendorff and that recurs in several other songs of Schumann's cycle.55 They are seductive and even violent figures, and the doe whom they threaten to shoot can be read, metaphorically, as a young woman -who is in danger of being raped.56 The spiritual inner bond of love that is depicted in "Intermezzo" and alluded to in some of the other Eichendorff songs is here replaced by its opposite, a crude and aggressive physical sexuality, which is driven by enmity rather than love. In terms of the poem as a whole, the depiction of human deception and hostility, in the two middle strophes, interrupts the cyclical progression from the uncertain gloom of dusk to the peaceful clarity of the dark night. It is as if -we have become frozen in this transitional state and it is our inability to trust and love one another that prevents us from transcending the uncertainty of the current moment. It is only in the last strophe that Eichendorff alludes to the continuation of the diurnal cycle, but at the same time he reiterates its ambiguity: Was heut mude gehet unter, Hebt sich morgen neugeboren. JVLanches bleibt in Nacht verloren — Hiite dich, bleib' wach und munter!
What today sinks -wearily down, Rises tomorrow newly born. Much gets lost in the night — Beware, be wakeful and alert!57
218
From Songj into Cycled
Eichendorff first reassures us that the apparent death of nightfall -will be folio-wed by the rebirth of dawn, but he then returns abruptly and inexplicably to the voice of warning that spoke throughout the middle strophes and suggests that the danger of death is real. And so the poem ends, as it begins, on a note of uncertainty and fear. It is largely through the use of formal discontinuities that Eichendorff evokes the transitional state of human existence that is symbolized by the evening in "Zwielicht." There is such a sharp break between the two inner strophes and the frame of the opening and closing ones that it almost seems as if he has inserted one poem -within another.58 And the sudden shifts in narrative point of view are equally unsettling, especially at the ends of the first and last strophes, when the objective third-person description is interrupted and the narrator addresses us directly. In "Im Walde," which immediately follows "Zwielicht" in the opus 39 Liederkre'u, the narrative voice is more coherent, but Eichendorff again uses disturbing juxtapositions and contrasts to highlight the themes of alienation and deception: Es zog eine Hochzeit den Berg entlang, Ich horte die Vogel schlagen, Da blitzten viel' Reiter, das Waldhorn klang, Das war ein lustiges Jagen!
A wedding passed along the mountain I heard birds singing, Many riders flashed by, the hunting horn sounded That was a merry hunt!
Und eh' ich's gedacht, war Alles verhallt, Die Nacht bedecket die Runde, Nur von den Bergen noch rauschet der Wald Und mich schauert im Herzensgrunde.
And before I realized it, all died away, The night covers it over, Only from the mountains the woods still rustle And I shudder deep within my heart.59
Now Eichendorff's narrator describes an apparently joyous and friendly scene of humanity, which takes place in the light of day. But by juxtaposing the wedding in the opening lines with the hunt in the second half of the strophe the poet draws on the symbolic associations of the latter image in order to question the meaning of the former. Is the wedding a celebration of loving union, or is it the culmination of a scene of seduction? It is also significant that the narrator is not a participant in the activity he describes but a passive onlooker. His implied isolation becomes more concrete in the second strophe, as he is left alone in the dark -woods. Again, as in "Zwielicht," Eichendorff depicts the frightening transition from day to night, in this case by presenting the two times of day in juxtaposition. As the silent darkness covers the surrounding landscape, the merry scene of humanity that has just transpired takes on a more ominous cast in the narrator's mind. He is reminded that while the fellowship we enjoy in our earthly life appears to bind us to one another as kindred souls, in reality it is transitory and, in the most fundamental sense, we are alone in the -world. The third song in the first half of the cycle, "Waldesgesprach," sets another text about human deception and seduction that corresponds to "Zwielicht" and "Im Walde" in the second half. The setting is again the woods at dusk, and the
The Song Cycle CM a Literary Work
219
protagonist begins by referring to the potential danger of both the place and the time of day: Es ist schon spat, es wird schon kalt, Was reit'st du einsam durch den Wald? Der Wald ist lang, du hist allein, Du schone Braut! Ich fiihr' dich heim!
It is growing late, it is getting cold, Why ride alone through the woods? The wood is wide, you are alone, You lovely bride! I will lead you home!60
As in the second strophe of "Zwielicht," it is the fact that the young woman is alone in the woods that makes her vulnerable, a fact that Eichendorff emphasizes through his use of language. In the second line the narrator asks why she is riding "einsam durch den Wald," and then, immediately repeating the word "Wald" early in the next line, he reiterates the question as a statement: "du bist allein." The syllable "ein" in the middle of the second line anticipates the rhyme that ends the third and fourth. And by pairing the words "allein" and "heim" Eichendorff calls attention to the contrast between the lonely danger of the present and the communal safety of past and future. As the young woman speaks in the second strophe, she reveals that she has, in fact, fallen victim to man's deceit, and her reference to the hunting horn suggests that she has been sexually violated: "Grofi ist der Manner Trug und List, Vor Schmerz mein Herz gebrochen ist, Wohl irrt das Waldhorn her und hin, O flieh! Du weifk nicht, wer ich bin."
"Great is man's deceit and cunning, Grief has broken my heart, The hunting horn sounds far and wide, Oh flee! You know not who I am."
However, her warning at the end of the strophe that she is not who she appears to be creates an ironic twist, and in the second half of the poem the two characters reverse roles: So reich geschmiickt ist RoG und Weib, So wunderschon der junge Leib, Jetzt kenn ich dich — Gott steh' mir bei! Du bist die Hexe Loreley.
So richly decked are steed and woman, So very beautiful her young form, Now I recognize you — God help me! You are the sorceress Loreley.
"Du kennst mich wohl—von hohem Stein
"You know me well —from the high rock Schaut still mein Schlofi tief in den Rhein. My castle silently gazes on the Rhine. Es ist schon spat, es wird schon kalt, It is growing late, it is getting cold, Kommst nimmermehr aus diesem Wald!" Nevermore will you leave these woods."61
It turns out that it is he who is alone and vulnerable and she who has the power to save or destroy him. In the final lines of the poem, she repeats his own words back to him and condemns him to remain forever in the -woods and thus
220
From Songj into Cycled
in a perpetual state of alienation and danger. In this seemingly simple poem, which retells the Loreley legend in the form of a dramatic dialogue, Eichendorff creates several layers of ambiguity. As the poem opens, the narrator's intentions are unclear and perhaps ambivalent. Is he offering to save the young woman from the danger that may befall her, or is his invitation really an attempt to seduce her himself? Likewise, is Loreley's response, in the following strophe, meant as an explanation of her current state, or a •warning that she recognizes the narrator's true, malicious intent? And as the narrator realizes who the young woman is, his attitude becomes conflicted in another sense. The legendary Loreley is not only a sorceress but also a seductress, -who tempts unwary young men to their deaths. While the narrator of "Waldesgesprach" expresses fear when he recognizes her, he primarily reacts by commenting on her beautiful appearance. The real danger comes from -within, since it is his own desire to cross into Loreley's forbidden world that keeps him trapped in the woods. His exclamation as he recognizes the sorceress—"Gott steh' mir bei!" — can be read more as a plea for inner spiritual strength than as a request for divine protection against an external threat. The twilit -woods of "Waldesgesprach" symbolize our temptation to remain engaged in the superficial activity of human society, just as the nocturnal fields and ruins in the songs that follow—"Die Stille," Mondnacht," and "Schone Fremde" — represent our potential for spiritual transcendence. In both cases, Eichendorff's landscape is not simply a setting but a symbol of an inner drama that takes place -within our souls and our imaginations. Schumann's inclusion of different poems that present stark contrasts between the imagery of dusk and of the night is one way in -which he lends a coherent meaning to the text of the Eichendorff Liederkreui. He turns the tension between our current state of confusion and alienation and our yearning to transcend this state and become reunited -with the natural world into a major theme and provides a context -within -which we can understand the symbolic significance of many of Eichendorff's other poetic images. As Schumann arranged the songs into a cycle, he took care to emphasize that both the diurnal progression that serves as the central symbol of the text as well as the spiritual quest that this symbol represents are cyclical processes, -which have no end. I believe that this is why he divided the cycle so clearly into two halves, each of which includes at least one song of the evening — "Waldesgesprach" in the first half and "Zwielicht" and "Im Walde" in the second — and each of-which ends -with a song of the night, first "Schone Fremde" and then "Friihlingsnacht." Schumann's arrangement implicitly hints at the progression from evening to nighttime and thus from the present to the future, but by presenting this progression twice he makes clear that its shape is cyclical and not linear. "Schone Fremde" and "Friihlingsnacht" are related not as a promise and its fulfillment but as two analogous moments, each depicting the ongoing growth of the human soul, and hinting at the potentiality of future transcendence. There is thus a close correspondence between thematic and formal elements in the text that Schumann created out of Eichendorff's poems. It depicts an inner process that does not have a clear beginning or ending point and that repeats itself again and again, both in the stages
The Song Cycle cu a Literary Work
221
of spiritual growth that take place within each individual and also in the various lives that unfold throughout the course of human history. Schumann does not cast this theme in the form of a closed and sequential narrative progression but as an open-ended cycle, whose two arches could potentially continue in either direction.
Motivic Relationships My explanation of why Schumann divided the Eichendorff Liederkrei) into two halves diverges quite sharply from the more traditional interpretation, which I discussed earlier in this chapter. A number of Schumann scholars have described a sequence of emotional states that takes us from the melancholy of "In der Fremde" to the joyous expectancy of "Schone Fremde" at the midpoint of the cycle and then to the jubilation and ecstasy of "Friihlingsnacht" at its end. My contrasting interpretation of Schumann's arrangement of the cycle's poetic text leads, in turn, to a different analysis of his musical setting. I will conclude by considering how my analytic approach diverges from that of other scholars and by explaining how it can help us understand Schumann's creation of an audible grouping within the Eichendorff Li£()erkreu>. Schumann's grouping of his Eichendorff songs is often analyzed in terms of a series of motivic relationships between "Schone Fremde" and both the first and last songs of the cycle. Daverio, for example, argues that the leaping fifth motive that appears in the accompaniment in measure 10 of "In der Fremde," a motive that I have discussed at some length in chapter A, recurs in the third strophe of "Schone Fremde," in measure 21, where "it expands to a sixth to highlight the closing proclamation of impending joy." It then appears in "Friihlingsnacht," in measure 24, again transformed, now with "its rhythmic pace quickened and its initial interval spanning a seventh," in order to convey the fulfillment of "Schone Fremde s" promise. As Daverio's motive recurs in these "two climactic songs," it marks the grouping of the cycle into two arches, and since its basic interval expands with each recurrence, it suggests a progressive development over the course of the cycle. At the same time, Daverio hears variants of this motive in several other songs. It expands from a fifth to an octave in the accompaniment of "Intermezzo" and becomes a series of falling fourths and fifths in the accompaniments of "Mondnacht" and "Auf einer Burg."62 Daverio thus argues that his motive serves two different purposes: not only does it articulate the formal division of the Eichendorff Liederkreu, but it also appears throughout the cycle "in a variety of guises," and in this way it unifies the songs. There is some question, however, as to whether the same motive can accomplish both of these purposes simultaneously. When Daverio claims that his motive appears in varied form through several successive songs, he is describing a process of thematic transformation, in which the constancy of the motivic repetition helps us to hear the continuity and coherence of the cycle. But if the motive is going to mark the sixth and twelfth songs as climactic moments and thus make audible the grouping of the cycle, it must return at these points, creating a
222
From Song
clearly recognizable connection to an earlier moment, when the same motive appeared. In order for a motive to return, however, it must first be absent, and so it cannot repeat throughout the intervening songs. If we hear Daverio's motive, at the end of "Schone Fremde," as one appearance within an ongoing series, then we cannot simultaneously hear it as a return that marks this moment as an arrival.63 McCreless alludes to this distinction between repetition and return and uses it as the basis for his argument that motivic relationships between the movements of a piece — or the songs of a cycle—can be either "unordered" or "ordered." His first category refers to thematic transformation, in which we cannot explain, analytically, why the various repetitions of a motive occur in the order that they do. McCreless acknowledges that his second category appears to be self-contradictory, since motives are unordered by definition, but he claims that when a motive returns at particular points within a cycle it may serve to articulate the grouping of the songs and that this provides a rationale for the order in which the appearances occur.64 McCreless's argument is based on a confusion between cause and effect, however. Schumann's placement of the same motive at the end of the first and sixth songs of the Eichendorff Liederkrei) may have the effect of articulating a grouping, but this placement does not explain the coherence of the motivic relationship. If Schumann reversed the order of the songs, the relationship -would still be audible and, for that matter, it would still articulate the grouping. If he placed the songs in entirely different positions — at the beginning and end, say, or as a consecutive pair — the relationship would no longer have the same grouping effect, but it would be audible, and musically coherent, nevertheless. Curiously, while McCreless goes to some lengths to argue that there is a meaningful difference between ordered and unordered motivic relationships, he ignores this distinction when he presents his examples from the Eichendorff Liederkreu and instead classifies them solely in terms of melodic interval. The motive that first appears in measure 10 of "In der Fremde" is an example of the "motive of the perfect fifth," and other examples of the same motive are found in "Intermezzo," "Waldesgesprach," "Die Stille," "Mondnacht," "Schone Fremde," "Auf einer Burg," and the second "In der Fremde." While Daverio hears the climactic leaps near the end of "Schone Fremde" and "Friihlmgsnacht" as transformed versions of this same motive, McCreless places them into a completely different category, as examples of the "motive of the sixth."66 What McCreless presents, in other words, is a taxonomy of the intervals that Schumann uses to construct his melodies, and this enables him to argue that almost all of the songs in the cycle are unified through the presence of motivic relationships. The crucial question, however, is not whether these relationships actually exist or even whether they are audible but what their significance is. If the use of perfect fifths relates Schumann's Eichendorff songs to one another, then it also relates them to many of his other works and, for that matter, to the works of numerous eighteenth- and nineteenth-century composers. The motive of the perfect fifth may reveal something about the style of tonal music in general, but it does not tell us anything significant about the Eichendorff Liederkreu.66 Another problem with
The Song Cycle M a Literary Work
223
McCreless's analysis is that he ends up arguing, as Daverio does, that Schumann uses the same motive both to unify the cycle and to mark its formal divisions. McCreless presents his perfect fifth motive as an example of an unordered relationship, variants of which appear in eight of the twelve songs. But when it appears in the accompaniments of both "In der Fremde" and "Schone Fremde," this perfect fifth becomes an example of an ordered relationship, because it now supports the grouping of the cycle into two halves. As with Daverio's expanding interval, however, it is impossible for us to hear McCreless's motive in both ways at once. These motivic analyses of Schumann's Eichendorff Liederkreu proceed from the same kinds of assumptions that have led scholars to interpret the text of the cycle as a sequence of moods. Daverio and McCreless are trying to impose a comprehensive unity and to explain the necessity for order and closure in a work that merely hints at a loose coherence and is inherently open-ended and cyclical. Just as our insistence on reading the text as a linear progression has forced us to ignore how Schumann's arrangement of the songs reflects the central theme of Eichendorff's poems, the claim that the cycle is motivically unified has prevented us from hearing how Schumann actually does use motivic relationships to express this theme in musical terms. Because they are primarily interested in motivic unity, McCreless and Daverio define motives in terms of a single musical parameter — melodic interval—and ignore the fact that Schumann relies on rhythm and meter, on musical texture, and on the harmonic and contrapuntal context •within which a given motive appears in order to make it audible and to give it musical meaning. If we take the latter parameters into account, then we cannot so easily argue that a single motive recurs throughout the songs, but we can explain how Schumann emphasizes a few crucial motives and uses them not only to articulate the formal division of the cycle but also to express the significance of that division. It is true, for example, that we can hear a relationship between the melodic gestures that appear in the vocal line in measure 21 of "Schone Fremde" and in the accompaniment in measure 24 of "Friihlingsnacht." And, at least to some extent, this relationship is based on the similarity of their melodic contours: both consist of an upward leap that is followed by a stepwise descent, which returns to the initial pitch. But in noting this I have done no more than identify these passages as typical examples of Schumann's stylistic language, since the melodic contour I have described is one that he uses often, in a variety of genres.68 If we want to understand how Schumann emphasizes the connection between these particular instances of this gesture and -why they have such expressive significance, we need to consider more carefully how he presents them. Another similarity between the two gestures, for example, which is just as important as their shared melodic contour, is their placement within the musical texture. As I have explained in my analyses of these two songs in part II, Schumann uses a variety of means in each of them to set the vocal line and the accompaniment against each other, ranging from contrasts of rhythm and register to more complex conflicts that involve phrase rhythm and counterpoint. The motive that appears in
224
From Songj into Cycled
the third strophe of each song is the first and only melodic gesture that is presented together in unison by the vocal line and the right hand of the accompaniment. This textural convergence is one way that Schumann emphasizes the motivic relationship between the two passages, and since both of the texts are concerned with the theme of the soul's convergence with nature, it is also an important way that Schumann lends expressive significance to the motive. I have explained, as well, that in both "Schone Fremde" and "Fruhlingsnacht" the upward leap that initiates the gesture comes as part of a climactic moment that leads to the final cadence of the song, and this analogous placement of the motive at a crucial point within the form is another way that Schumann highlights the relationship. If we compare these two cadences, we are led to reconsider Daverio's observations about Schumann's transformation of the motive in "Friihlingsnacht." It may be the case that the quickening of the rhythm and the expansion of the initial interval to a minor seventh intensify the climactic effect of the motive, but as I have argued in chapter 5, the ultimate result of the upward leap, in measure 24, is to weaken the resolution of the cadence that follows, since it disrupts the vocal line's stepwise descent to the cadential tonic, leading Schumann to substitute a downward leap, in parallel octaves with the bass. In "Schone Fremde," however, the motive helps to support the strength of the cadential resolution, since in this case, it is the second half of the motive that creates the stepwise descent to the tonic note. More generally, "Friihlingsnacht s" cadence is less definitive and conclusive than "Schone Fremde's" because Schumann turns the cadential tonic into a V 7 /IV and delays the harmonic resolution until after the vocal melody has concluded. It seems implausible, then, that he is using the motivic relationship between "Schone Fremde" and "Friihlingsnacht" to express promise and fulfillment, for if he were, he would certainly ensure that the motive strengthened the cadence in the latter song, and he would resolve it more conclusively than the former. Instead, he weakens the cadence of the final song, creating an open ending, and thus expresses the cyclical and ephemeral quality of the transcendent moment that each song depicts. There are several motivic relationships between "Schone Fremde" and the opening song of the cycle, "In der Fremde," that also serve to articulate the formal division of the Eichendorff Liederkreu into two halves. The most striking of these, as McCreless points out, is a leaping sixth figure that appears in the coda of each song. In my analysis of "In der Fremde" in chapter A, I explained that this figure is a varied reappearance of the leaping fifth that is heard in the accompaniment in measure 10, where it is associated with the narrator's yearning for transcendence. When the motive returns in the coda, it suggests that her yearning cannot be fulfilled, especially since the upper note of the leap, d2, is a dissonance that is left unresolved. When Schumann first composed the piano draft of "Schone Fremde," he concluded the song -with a brief postlude that goes directly from measure 24, the final measure of the vocal melody, to the last two measures of the song and simply arpeggiates the tonic B major harmony. In this initial version, Schumann already alludes to the coda of "In der Fremde," presenting the same sixth in a similar rhythmic form, but transforming it from minor
The Song Cycle cu a Literary Work
225
to major by turning the d2 into a dtt2. When Schumann expanded the postlude, at some point after he completed the draft, he made the connection to the earlier song even stronger by introducing a chromatic neighboring chord that prolongs the tonic harmony, as in the coda of "In der Fremde," and, even more strikingly, by including g^1 within this harmony, as an upper neighbor to ft1, and placing it prominently on the downbeat, in the uppermost voice of the texture. The motivic relationship between the codas of "In der Fremde" and "Schone Fremde," unlike that between the latter song and "Friihlingsnacht," is intended to convey the contrast between these two moments in the cycle. Schumann establishes an aural connection through his similar presentation of the sixth motive in each song. But when the motive returns in "Schone Fremde," he changes the tonal context and thus gives it a completely different expressive significance. Of particular importance is the fact that the melodic peak, which is now d)t2, is no longer an unresolved dissonance that is supported by the neighboring harmony but a member of the tonic triad, which arpeggiates down to the tonic note as the coda concludes. In "In der Fremde" Schumann uses the motive to question the stability of the tonic, but in "Schone Fremde" he is strengthening the tonic and reinforcing the closure of the final vocal cadence. The contrasting functions of this motive correspond to its different expressive purposes in each of the songs. In the former song, it expresses the protagonist's desire to become one with the natural world that surrounds her. The protagonist of "Schone Fremde" describes this experience as if it were actually happening, and the return of the motive at the end of this song, within a new tonal context that transforms it into a consonance, helps us to create an audible connection between these two passages. There is no narrative progression that unfolds between the texts of the two songs, and we cannot really hear them as parts of a single coherent discourse. But each records a moment within the development and growth of the human soul as it strives to move beyond the constraints of the material world and return to its spiritual home -within nature. Likewise, there is no continuous process of motivic transformation that leads from "In der Fremde" to "Schone Fremde." Rather, the appearance of the motive in the latter song is a thematic reminiscence that recalls to our minds a related but contrasting moment in the opening song.
Conclusion In contrasting my motivic analysis of Schumann's Eichendorff Liederkreu with the analyses of Daverio and McCreless, I have attempted to illustrate the way that my approach to Schumann's cycles differs from the interpretive tradition that they are -working -within. Where they both begin from the idea that there is a single motive that unifies the entire cycle and then discover particular instances of this motive in each of the songs, I begin from a more detailed and comprehensive analysis of the individual song and consider how Schumann uses various musical parameters to lend motivic significance to a given melodic gesture within that song. Motivic relationships between one song and another do not result simply from the presence of the same basic gesture in each of them but from
226
From Song<> into Cycled
a network of similarities and connections, -which are based as much on form and texture as they are on melody and rhythm. More generally, I do not find it all that fruitful to analyze the Eichendorff Llederkreid, or any other Romantic cycle, as a unified -whole. What makes the cycle such a compelling musical genre is the fact that it is made up of individual parts that are suggestive and imply a variety of potential interconnections. The open-ended quality that is so characteristic of Schumann's songs arouses in us a desire for unity and closure, and that is one reason that they are so well suited to publication and performance as cycles. Throughout the chapters of this book, I have explained this conception of the cycle in terms of several different perspectives. First, there is the group of poems that Schumann set. Each of them is a brief lyric utterance, in -which a feeling or an idea is expressed in large part through symbolic imagery. Eichendorff, like many of the poets -whom Schumann turned to in his Liederjahr, reveals very little about the situation and background in his poems or about the identity of his protagonist. The significance of Eichendorff's imagery is not always clear, and the reader is often left to make connections and supply meanings that the poet merely hints at. The cycle of twelve poems that Schumann created in composing his opus 39 Litderkreu leads us to understand this poetry on a new level, not because it places the individual poem -within a coherent discourse but because it creates a larger context of potential connections and meanings. This suggestive means of conveying thematic content derives in part from the nature of the content itself. Eichendorff expresses the quintessential Romantic yearning to transcend the material world, -which feels fragmentary and alienating, and to find the wholeness and unity that it lacks. This state of transcendence has not yet been achieved but only glimpsed, and the sense of the experience cannot be completely captured in language and imagery in any case. Second, -we might consider Schumann's process of composing cycles, to the extent that this can be reconstructed from surviving documents. He generally began -without a clear idea of the cycle's shape, and sometimes he did not even decide -which poems to include until after he had completed a number of songs. For Schumann, the initial stage of composing a cycle was improvisatory and intuitive and consisted of reading through a volume of poems and setting those that appealed to him. His songs tend to be quite brief, and he usually composed them in quick succession. He often responded to a poem by experimenting -with the typical song forms of his day—ternary, bar form, and strophic—and -with the underlying principles of tonal structure. I have compared Schumann to a visual artist -who is drawing a series of sketches and exploring the same ideas and techniques in each of them. As he arranges these musical sketches into a cycle, they imbue the whole with their improvisatory and open-ended quality. Third, there is the fact that Schumann was heavily influenced by the early Romantic movement that centered around Friedrich and August Wilhelm Schlegel, whose aesthetic ideals he shared. He viewed the construction of musical meaning as an interactive process, in -which the performer and the listener play as much of a role as the composer, since they must realize, in their awn minds, what he has merely implied. Schumann composed his songs with two different audi-
The Song Cycle 04 a Literary Work
227
ences in mind. He marketed them widely to amateur singers, who could not be expected to partake in this process, but he also considered the Luderkreut to be a private genre, intended for an exclusive circle of refined and sophisticated musicians. At the same time, Schumann knew that his audience was not necessarily restricted to the handful of Davidsbiindler and the mass of philistines -whom he saw around him. The Romantic conception of aesthetic understanding that reveals itself in his music is an outgrowth of the idealistic worldview represented in the Romantic literature that he drew on for his song cycles. The Romantics believed that the artistic and spiritual growth of humanity is an infinite quest. The significance and meaning of Schumann's songs continue to develop as sucessive generations perform them, analyze them, and listen to them.
This page intentionally left blank
NOTES
Chapter 1 1. Gerard Genette, Narrative Dutcourje: An E>
230
Note to Pages 8-9
7. Review of Clara Schumann's Sechj Lieder, opus 13, in Neue Zeitjchrift filr Miuik 20 (1844): 97. Die Lieder werden und wollen nicht einen gerauschvollen Triumphzug durch die Salons machen, aber in stiller Klause wird sich manch stilles empfangliches Gemiith an ihrer ungeschmiickten Anmuth, dem poetischen Dufte, der durch sie weht, erquicken. Review of Hieronymous Truhn's Bin Lieberjroman in 12 Liedern, opus 64, in Neue Zeitdckrift furMM& 20 (1844): 197. So eignet sich denn diese Sammlung wenn nicht zum Salon-Vortrage —wozu ja auch kein eigentliches Lied, und das sellenvollste am wenigsten gewahlt werden sollte—doch ganz besonders zur Verkiirzung einsamer Stunden oder zur anregenden Unterhaltung ines engeren gewahlten Kreises poetischer Seelen. 8. Ruth Bingham, The Song Cycle in German-Speaking Countries 1790-1840: Approaches to a Changing Genre (Ph.D. diss., Cornell University, 1993), 13—20. It was Bingham -who first discovered Dommer's definition and identified it as the earliest one. 9. In a review of Julius Stockhausen's 1856 performance of Schubert's Die jcbdne Mullerin, Eduard Hanslick describes the idea of performing a complete song cycle, "instead of the usual conglomeration of pieces that do not belong together," as something new. Edward F. Kravitt writes that Stockhausen gave repeat performances throughout the German-speaking world, to great acclaim. Nancy Reich, citing a documentary biography of Stockhausen, dates his first performance of Die
Note*! to Pages 11-17
231
end, was die Stimme hinsichts des Ausdruckes unerledigt lassen muss. Zur dramatisierenden Solocantate fehlt dem Liederkreise eigentlich nichts mehr als das Recitativ, und die arieartige Form der Gesange anstatt der liedartigen; im ubrigen wird man ihn der Cantate ziemlich nahestehend finden, oder als eine Mittelgattung zwischen durchcomponirtem Liede und Cantate ansehen. 11. In Arthur Komar, editor, Schumann's, Dichterliebe (New York: Norton, 1971), 63-94. 12. Rudolph Reti, The Thematic ProceM in Mu*tic (New York: Macmillan, 1951), chapter 2; Theodor Adorno, "Zum Gedachtnis Eichendorffs," uiNoten zur Literatur I (Frankfurt: Suhrkamp Verlag, 1958), 134-44. 13. David Neumeyer, "Organic Structure and the Song Cycle: Another Look at Schumann's Dichterliebe," Miuic Theory Spectrum 4 (1982): 92-105; Patrick McCreless, "Song Order in the Song Cycle: Schumann's Liederkreu, Op. 39," Mu*iic Analysis 5 (1986): 5-28; Peter Kaminsky, "Principles of Formal Structure in Schumann's Early Piano Cycles, "Music Theory Spectrum 11 (1989): 207-25. 14. Kaminsky, "Schumann's Early Piano Cycles," 207. 15. Turchm, Schumann'*! Song Cycled, 5—10. 16. Turchin, Schumann'*! Song Cycled, chapter 2. Turchin includes much of her discussion of Kreutzer's Wanderiieder in her article "The Nineteenth-Century Wanderlieder Cycle," Journal ofMu*>icology 5 (1987): 498—525. There is also a study of the cycle by Luise Eitel Peake, "Kreutzer's Wanderlieder. The Other Winterreiie,"MusicalQuarterly 45 (1979): 83-102. 17. Turchin, Schumann'*! Song Cycles, 220-22. 18. Bingham, The Song Cycle, 22, 26-28. 19. John Daverio, "The Song Cycle: Journeys through a Romantic Landscape," in German Lieder in the Nineteenth Century, edited by Rufus Hallmark (New York: Schirmer, 1996), 282. 20. Bingham, The Song Cycle, 237-42. 21. Turchin, Schumann'^ Song Cycle*!, 224—25. 22. Turchin, Schumann'*! Song Cycled, 9. 23. Turchin, Schumann'*! Song Cycled, 216. 24. Daverio, Robert Schumann, 213. 25. Carl Banck, who frequently reviewed lieder in the early years of the Neue ZeLtdchrift's publication, criticizes Der Bergmann, another cycle by Lowe that was published with the same designation, because the songs "waver too much between the two" genres. Neue Zeitdchri/tftir Mu*iik 2 (1835); 95. I will discuss this review in chapter 8. 26. John Daverio, "Schumann's Systems of Musical Fragments and Witz," in Nineteenth-Century Mudif am) the German Romantic Geology (New York: Schirmer, 1993). 27. Charles Rosen, "Fragments," in The Romantic Generation (Cambridge: Harvard University Press, 1995). Rosen relates the long history of his discussion of the fragment m the preface, x—xi. 28. Rosen's suggestion that there may be a connection between Schlegel's aesthetics and nineteenth-century music has also influenced scholars who are concerned with other composers. The most important book to explore this connection and the one that is most relevant to the present study is Richard Karmer, Distant Cycles: Schubert and the Conceivin of Song (Chicago: University of Chicago Press, 1994). In a more recent study, Kramer has considered the aesthetics of the fragment in relation to Schubert's large number of unfinished works. See "The Hedgehog: Of Fragments Finished and Unfinished," 19th-century
232
Noted to Paged 17-30
Miuic 21 (1997): 134-48. There is also a recent article that explores the influence of the fragment on Liszt's formal procedures by Ramon Satyanedra, "Liszt's Open Structures and the Romantic Fragment," Mudic Theory Spectrum 19 (1997): 184-205. 29. Daverio, Nineteenth-Century MIMIC, 53-54. 30. Leonard B. Meyer discusses the prevalence of this strand of organicist thought \nStyteandMudic (Philadelphia: University of Pennsylvania Press, 1989), 194-95. 31. Daverio, Nineteenth-Century Mudic, 75—86. 32. Rosen, The Romantic Generation, 79. 33. Rosen, The Romantic Generation, 53—55. I shall return to "Aus meinen Thranen spriessen" in the following chapter and discuss Schenker's analysis of the song. 34. Rosen, The Romantic Generation, 83. 35. Rosen, The Romantic Generation, 125. 36. Rosen, The Romantic Generation, 207—8. 37. August Wilhelm Schlegel, "Uber Zeichnungen zu Gedichten," in Athenaeum: Eine Zeitdchrift von Augudt Wdhelm Scblegel and Friedrich Schtegel, vol. 2, edited by Curt Grfltzmacher (Munch: Rowohlt, 1969). I shall discuss this essay in more detail in chapter 3.
Chapter 2 1. Arthur Komar, "The Music of Dichterliebe: The Whole and Its Parts," in Komar, ed., Schumann: Dichterliebe (New York: Norton, 1971), 63. All further references to Komar's essay will be in the form of parenthetical citations. The term cycle-hood is found in Patrick McCreless, "Song Order in the Song Cycle: Schumann's Liederkreid, Op. 39," MudicAnaLydid 5 (1986): 8. 2. For a transcription of Schumann's plan for Myrthen, see Barbara Turchin, Robert Schumann'd Song Cycled in the Context ofthe Early Nineteenth-Century Liederkreid (Ph.D. diss., Columbia University, 1981), 267. For a description of the draft for "Es leuchtet meine Liebe," see Rufus Hallmark, The Genedi) of Schumann 'd Dichterliebe: A Source Study (Ann Arbor: UMI, 1976), 91. Hallmark reports that the markings have faded too much for his reading to be definitive. 3. Of the song publications that include songs composed in 1840 and 1841, nine consist of miscellaneous songs and are clearly not intended to have any kind of cyclic unity. Among the thirty-eight songs within these collections, there are only two instances in which the succession of keys is not by third, by perfect fifth, or between parallel major and minor. David Neumeyer comes to a similar conclusion by comparing Schumann's song cycles with Schubert's waltz sets. See Neumeyer, "Organic Structure and the Song Cycle: Another Look at Schumann's Dichterliebe," Mudic Theory Spectrum 4 (1982): 95. 4. The exact pitch class of this second note is determined by calculating the sixth scale degree of the subdominant key of the first note. This is why B follows A when it is a major key at the beginning of the cycle, while Bt follows when A is minor in Song 8. 5. The example is my own and not Komar's. He does not provide an example that completely shows the workings of the harmonic and modal plans. 6. Turchin, Schumann',) Song Cycled, 332. 7. McCreless, "Song Order," 9; Neumeyer, "Organic Structure," 94-96; Peter Kaminsky, "Principles of Formal Structure in Schumann's Early Piano Cycles," Mudic Theory Spectrum 11 (1989): 208.
Notes to Pages 33-51
233
8. Heinrich Schenker, Derfreie Satz, 2d ed., Oswald Jonas, ed. (Vienna: Universal Editions 1956), §3, §10. 9. Schenkeer, Derfreie Satz, §§168-70. 10. This is not, of course, how Schenker explains the interruption in Derfreie Satz (see §§87ff.), but the relationship between the interruption and the parallel period is obvious and has been mentioned by a number of theorists. Schenker uses Beethoven's "Od to Joy" theme as an example of the transference of the interruption to the foreground level, and when he turns to sonata form in his chapter on form, he writes: "Only the pro longation of a division (interruption) gives rise to sonata form." Schenker, Derfreie Satz, §243; Figure 109e, 3; §312. See also Felix Salzer, Structural Hearing: Tonal Coherence in Music (New York: Charles Boni, 1952), 145-47; and William Rothstein, Phrase Rhythm in TonalMiutic (New York: Schirmer, 1989), 16-18. 11. Schenker, Derfreie Satz, Figures 22b; HOc, 2; 152, 1. 12. Allen Forte, "Schenker's Conception of Musical Structure," Journal of Mud ic Theory 3 (1959): 1—30. The section that is concerned •with Schenker's analysis of the song subsequently appeared in Komar, Schumann: Dichterhebe, 96—106; and the entire article was reprinted in Readings in Schenker Analysis and Other Approaches, edited by Maury Yeston (New Haven: Yale University Press, 1977). In all further citations, page numbers will refer to the 1977 edition. 13. Joseph Kerman, "How We Got into Analysis, and How to Get Out," CriticalIn «/ry6(1980):323. 14. Kerman, "How to Get Out," 326, 325. 15. Forte, "Schenker s Conception," 37, fn. 34; Komar, "The Music of Dichterliebe," 72. 16. Kerman, "How to Get Out," 325. 17. Heinrich Schenker, Harmony, translated by Elisabeth Mann Borgese and edited by Oswald Jonas (Chicago: University of Chicago Press, 1954), 221. 18. This is one of several instances in which Kerman misreads Schenker's graph. As I have noted, he describes Schenker's analyses of both measures as half-cadences. See Kerman, "How to Get Out," 325. 19. Allan Keiler, "On Some Properties of Schenker's Pitch Derivations," Music Perception 1 (1984): 214-15; and "The Syntax of Prolongation, Part I," In Theory Only 3 (1977): 19. 20. Schumann's recomposition of the first half of the phrase when it returns is irrelevant to this question, since it does not affect the structure of the second half or the function of the entire phrase. 21. Kerman, "How to Get Out," 325. 22. Neumeyer, "Organic Structure," 97. All further references to Neumeyer's article will be in the form of parenthetical citations. 23. Komar, Schumann'*! Dichterliebe, 8-11. See especially fn. 20 on page 11. 24. This is why instrumental transcriptions, such as Liszt's piano versions of the songs of Schubert and Schumann, are possible. 25. Schenker, Derfreie Satz, § 305. 26. In addition to the opening song of Dichterliebe, these include "Anfangs wollt' ich fast verzagen" from the Heine Liederkreis, which Neumeyer cites in his article (98), "Auf einer Burg" from the Eichendorff Liederkreui, and "Stirb Lieb' und Freud'!" and "Frage," both from the Kerner Liederreihe. Barbara Turchin discusses the endings of "Anfangs wollt' ich fast verzagen" and "Auf einer Burg" in Schumann's Song Cycled, 309—11, 322—24; as does Charles Rosen, in The Romantic Generation (Cambridge: Harvard University Press,
234
Noted to Pages 51-62
1995), 212—16. Jon W. Finson and Karen A. Hindenlang also discuss the archaism of "Auf einer Burg." See Finson, "The Intentional Tourist: Romantic Irony in the Eichendorff Liederkreis of Robert Schumann," in Schumann and Hid Word), edited by R. Larry Todd (Princeton: Princeton University Press, 1994), 166—67; and Hindenlang, "Eichendorff's Auf einer Burg and Schumann's Liederkreis, Opus 39," Journal of Musicology 8 (1990): 581-83. 27. Neue Zcibchriftfur Mtuik 23 (1845): 14. This review is quoted in full in chapter 4. 28. Schenker, Derfreie Satz, § 244.
Chapter 3 1. See, for example, Jerome J. McGann, The Romantic fdeolagy (Chicago: University of Chicago Press, 1983); Anne Janowitz, "Coleridge's 1816 Volume: Fragment as Rubric," Studies in Romanticism 24 (1985): 21—39; and Leonard B. Meyer, Style andMusic (Philadelphia: University of Pennsylvania Press, 1989), chapter 6. 2. Thomas McFarland, Romanticism and the Form** of Ruin (Princeton: Princeton University Press, 1981), 7. 3. McFarland, Forms of Ruin, 41-42. 4. Meyer, Style and Music, 176. 5. Morse Peckham, The Triumph of Romanticism (Columbia: University of South Carolina Press, 1970), 10. 6. Meyer, Style and Music, 196-99. 7. Meyer, Style and Music, 195. 8. Meyer, Style and Music, 200-201. 9. Peckham, The Triumph of Romanticism, 12. 10. See Charles Rosen, The Romantic Generation (Cambridge: Harvard University Press, 1995), 51; and John Daverio, Nineteenth-Century Music and the German Romantic Ideoiagy (New York: Schirmer, 1993), 54. 11. Meyer, Style and Music, 159-60. 12. See Edward A. Lippmann, "Theory and Practice in Schumann's Aesthetics," JAMS 17 (1964): 310. 13. Most of this collection was written by Friedrich Schlegel, with further contributions by August Wilhelm Schlegel, Novalis, and Friedrich Schleiermacher. For the attributions of specific fragments, see the text in Friedrich Schlegel, Kritische FriedrichSchlegel-Ausgahe, vol. 2, edited by Hans Eichner (Zurich: Thomas-Verlag, 1967). I will hereafter refer to this volume as KFSA 2. 14. See Philippe Lacoue-Labarthe and Jean-Luc Nancy, The Literary Absolute: The Theory of Literature in German Romanticism, translated by Philip Barnard and Cheryl Lester (Albany: State University of New York Press, 1988), 40; and Rosen, The Romantic Generation, 48-50. 15. A 116. "Andre Dichtarten sind fertig, und konnen nun vollstandig zergliedert werden. Die romantische Dichtart ist noch im Werden; ja das 1st ihr eigentliches Wesen, dafi sie ewig nur werden, nie vollendet sein kann ..." The text for this and all further quotations from Schlegel's -writings is from KFSA 2, unless otherwise noted. Fragments are designated by number, preceded by a capital letter to indicate which collection they come from: A for the Athenaeum Fragmente, I for the Ideen, and KF for the Kritische Fragmente, which were originally published in the journal Lyceum. All translations in this chapter are my own.
NotM to Page
235
16. See Hans Eichner, "Germany/Romantisch—Romantik—Romantiker," in "Romantic"and 1'fcf Cognates: The European History of a Word, edited by Hans Eichner (Toronto: Uni versity of Toronto Press, 1972), 105. 17. / 95. "Alle klassischen Gedichte der Alten hangen zusammen, unzertrennlich, bilden ein organisches Ganzes, sind richtig angesehen nur Ein Gedicht, das einzige in welchem die Dichtkunst selbst vollkommen erscheint." 18. KF 93. "In den Alten sieht man den vollendeten Buchstaben der ganzen Poesie: in den Neuern ahnet man den werdenden Geist." 19. A 116. "Die romantische Dichtart ist die einzige, die mehr als Art, und gleichsam die Dichtkunst selbst ist: denn in einem gewissen Sinn ist oder soil alle Poesie romantisch sem." 20. A 206. "Ein Fragment mufi gleich einem kleinen Kunstwerke von der umgebenden Welt ganz abgesondert und in sich selbst vollendet wie ein Igel." 21. Maurice Blanchot, "The Athenaeum," Studies in Romanticitm 22 (1983): 172. 22. Rosen, The Romantic Generation, 48. 23. For discussions of this distinction in Schlegel's writings, see Janowitz, "Fragment as Rubric," 26; and Rodolphe Gasche, foreword to Friednch Schlegel, Philosophical Fragments, translated by Peter Firchow (Minneapolis: University of Minnesota Press, 1991), viu. 24. A 24. "Viele Werke der Alten sind Fragmente geworden. Viele Werke der Neuern sind es gleich bei der Enstehung." 25. For a recent discussion o£BiH)ung and its relationship to Romantic literature, see Marc Redfield, "Romanticism, BiOung, and the Literary Absolute," in Le^onj of Romanticism: A Critical Companion, edited by Thomas Pfau and Robert F. Gleckner (Durham: Duke University Press, 1998), 41-55.
26. A 297. Gebildet ist ein Werk, wenn es iiberall scharf begrenzt, innerhalb der Grenzen aber grenzenlos und unerschopflich ist, wenn es sich selbst ganz treu, iiberall gleich, und doch iiber sich selbst erhaben ist. Das Hochste und Letzte ist, wie bei der Erziehung eines jungen Englanders, le grand tour. Es mufi durch alle drei oder vier Weltteile der Menschheit gewandert sein, nicht um die Ecken seiner Individualitat abzuschleifen, sondern um seinen Blick zu erweitern und seinem Geist mehr Freiheit und innre Vielseitigkeit und dadurch mehr Selbstandigkeit und Selbstgeniigsamkeit zu geben. 27. See Lacoue-Labarthe and Nancy, The Literary Absolute, 112; and KFSA 2:371. 28. KF 20. Eme klassische Schrift mufi nie ganz verstanden werden konnen. Aber die, welche gebildet sind und sich bilden miissen immer mehr draus lernen wollen. 29. A 22. ". . . die Fahigkeit, Gegenstande unmittelbar zugleich zu idealisieren, und zu realisieren, zu erganzen, und teilweise in sich auszufuhren." 30. Lacoue-Labarthe and Nancy, The Literary Absolute, 43—44. 31. Morse Peckham, The Romantic Virtuoso (Hanover: Wesleyan University Press, 1995), 11-12. 32. Peckham, The Romantic Virtuoso, 133—34. 33. NeueZeibchriftfOrMtuik 11 (1839): 161-63. Die Praludien bezeichnete ich als merkwiirdig. Gesteh' ich, dafi ich mir sie anders dachte und wie seine Etuden im grofiten Stil gefiihrt. Beinahe das Gegenteil; es sind Skizzen, Etudenanfange, oder will man, Ruinen, einzelne Adlersittige, alles bunt und wild durchaneinander.
236
Noted to Pages 65-68
34. KF A. "Es gibt so viel Poesie, und doch ist nichts seltner als ein Poem! Das macht die Menge von poetischen Skizzen, Studien, Fragmenten, Tendenzen, Ruinen, und Materialien." See John Daverio, "Schumann's 'Im Legendenton' and Friedrich bchlegel's Arabuke," 19th-century Music 11 (1987): 162.
35. NtufZeibchriftfiirMusik 14 (1841): 181. Andere haushalterischere Komponisten wiirden aus manchen Grundgedanken er Etuden ganze Konzerte und Sonaten aufgebaut haben; unser Komponist zieht es vor, nur anzudeuten und fluchtig anzuregen; sein uberwiegender Humor will es so, und auch der Schattenrift ist wilkommen.
36. NeueZcibchriftfiirMusik 14 (1841): 181. Es liest sich die Etudensammlung etwa wie ein Tagebuch. Mannigfaltige Meinungen smd hier nebenemander ausgesprochen, bittere Bemerkungen fehlen nicht, auch nicht liebe Erinnerungen. Der Kiinstler, der Philosoph, der Freund lafit sich darin gehen, als sane ihm kein Menschenauge zu, als gabe es keine Recensenten. 37. August Wilhelm Schlegel, "Uber Zeichnungen zu Gedichten," in Athenaeum: Eine Zeifochrift von August Wdheltn ScklegeL tint) Friedrich Schlegel, vol. 2, edited by Curt Griitzmacher (Munich: Rowohlt, 1969), 78. "Das selten aber entziickende Schauspiel des Zusammenwirkens zweier Kiinste, in Emtracht und ohne Dienstbarkeit." 38. Schlegel, "Uber Zeichnungen," 78. "Warum sollte es nicht eine pittoreske Begleitung der Poesie, nach Art musikalischen, geben konnen?" See Kristina Muxfeldt, Schubert Song Studies (Ph.D. diss., State University of New York at Stony Brook, 1991), 23. 39. Schlegel, "Uber Zeichnungen," 771 Ein Roman konnte vortrefflich sein und keinen einzigen tauglichen Moment fur die malerische Darstellung enthalten. Es wiirde hingegen keine sonderliche Tiefe verraten, wenn sich alles darin sichtbar machen liefie. Grade das Bedeutendste kann oft in der aufiern Erscheinung am wenigsten Evidenz hervortreten. Der Roman ist bestimmt, die zarteren Geheimnisse des Lebens, die nie vollstandig ausgesprochen werden konnen, in reizenden Sinnbildern erraten zu lassen. Die Poesie schmiegt sich hier vertraulich an die Wirklichkeit an und haucht ihr eine hohere Seele ein. Es ist nicht mehr die blofie Wirklichkeit, aber sie soil es noch scheinen. Es gibt keine Briicke, die den bildenden Kunstler aus seinem Gebiet in den Mittelpunkt einer solchen Dichtung hmuberfuhren konnte, und so sollte er sich auch fur zu gut halten, um an ihren aufiersten Grenzen herumzuschleichen While the term Raman is typically translated as navel or romance, Schlegel apparently intends it in a more general sense in this passage. Later in the chapter, I will return to the definition (A Roman when I discuss Friedrich Schlegel's use of the term. 40. Schlegel, "tiber Zeichnungen," 78. "Der bildende Kunstler gabe uns ein neues Organ, den Dichter zu fiihlen, und dieser dolmetschte wiederum in seiner hohen Mundart die reizende Chiffersprache der Linien und Formen." For further discussion of this passage, see Muxfeldt, Schubert Song Studies, 19-23. 41. Schlegel, "Uber Zeichnungen," 79-80.
Note^ to Page 68
237
Der wesentliche Vorteil 1st aber der, dafi die bildende Kunst, je mehr sie bei den ersten leichten Andeutungen stehenbleibt, auf eme der Poesie desto analogere Weise wirkt. Ihre Zeichen werden fast Hieroglyphen, wie die des Dichters; die Phantasie wird aufgefordert zu erganzen und nach der empfangenen Anregung selbstandig fortzubilden, statt dafi das ausgefuhrte Gemalde sie durch entgegenkommende Befriedigung gefangennimmt. . . . So wie die Worte des Dichters eigentlich Beschworungsformeln fur Leben und Schonheit sind, denen man nach ihren Bestandteilen ihre geheime Gewalt nicht anmerkt, so kommt es einem bei dem gelungenen Umrifi wie eine wahre Zauberei vor, dafi in so wenigen und zarten Strichen so viel Seele wohnen kann. 42. Schlegel, "Uber Zeichnungen," 80. "Zwar mufi man seine Phantasie schon malerisch geubt und vollstandige Kunstwerke viel gesehen haben, um diese Sprache gelaufig lesen zu konnen." 43. Schlegel, "Uber Zeichnungen," 78. Ein scharfsinniger Kenner hat vor kurzem auf die so oft vernachlassigte Foderung gedrungen, dafi jedes Kunstwerk sich selbst ganz aussprechen soil. . . . Aber die Freiheit, manchen Umstand als bekannt vorauszusetzen, auf den er nur anspielen kann, wird doch dem Kiinstler bleiben mussen, wenn er nicht gar zu enge eingeschrankt werden soil. Ein solcher Kreis von Mythen oder Legenden ist dann als das gemeinschaftliche Gedicht eines Volkes oder Zeitalters zu betrachten, womit man die Bekanntschaft jedem einzelnen zumutet. Eben jener Kunstrichter hat den Begriff einer Zyklus von Gemalden sehr belehrend ins Licht gesetzt, und gibt zu, dafi in der zykhschen Form Auftritte vorkommen diirfen, die erst durch vorhergehende oder folgende ihre voile Deutung erhalten. Da, wo nicht unabhangige und ausgefuhrte Werke aufgestellt werden sollen, sondern wo eine Kunst nur einen Teil ihrer Mittel gebraucht, um sich mit einer andern zu verbrudern, erstreckt sich die Befugnis natiirlich noch weiter. 44. Barbara Turchin, for example, reads Schlegel as suggesting that circles of myths and legends constitute "the subject matter which [he] finds most appropriate for such cyclic presentation." She does demonstrate convincingly that the terms KreLi and Cykltut entered the vocabulary of German literary criticism at the turn of the nineteenth century, largely as a result of teh renewed interest in the cycles of epic poems composed in ancient Greece and, ultimately, in medieval Germany. She is equally persuasive in her claim that the two terms •were used interchangeably by many critics, including Schlegel. But as Muxfeldt has pointed out, Turchin's interpretation of the present passage is implausible. Schlegel does not make any kind of connection between the "circle of myths or legends" ("Kreis von Mythen oder Legenden") in one sentence and the "cycle of paintings" ("Zyklus von Gemalden") in the following sentence, except perhaps in the most loosely analogous sense. He is primarily arguing that it is acceptable for a work to result from "the interaction of two arts" and that m such a case each genre's contribution may be incomplete and thus not fully comprehensible without reference to the contribution of the other genre involved. Works that are based on "circles" of legends, as well as those that are part of "cycles" of paintings, may be insufficient in the same way and thus provide precedents for Schlegel's hybrid artwork. See Turchin, Robert Schumann'^ Song Cycled in the Context of the Early Nineteenth-Century LieierkreLi (Ph.D. diss., Columbia University, 1981), chapter 3; and Muxfeldt, Schubert Song Studies, 37, fn. 25.
238
Notu to Pages 69- 71 45. Muxfeldt, Schubert Song Stadia, 24-25.
46. One such reference can be found in Goethe's critical writings, where he, too, recognizes "cyclic qualities" in painting but not in Itierature. See Glaus-Michael Ort, "Zyklische Dichtung," in Reallexikon der (leuttichen Literaturgejcbichte, vol. 5, edited by Werner Kohlschmidt and Wolfgang Mohr (Berlin: Walter de Gruyter, 1988), 1106. 47. 795. Als Bibel wird das neue ewige Evangelium erscheinen, von dem Lessing geweissagt hat: aber nicht als einzelnes Buch im gewohnlichen Sinne. Selbst was wir Bibel nennen ist ja ein System von Buchern. Uebrigens 1st das kein willktthrlicher Sprachgebrauch! Oder giebt es ein andres Wort, um die Idee eines unendlichen Buchs von der gemeinen zu unterscheiden als Bibel, Buch schlechthin, absolutes Buch? 48. A 116. " . . . alle getrennte Gattungen der Poesie wieder zu vereinigen, und die Poesie mit der Philosophic und Rhetonk in Beriihrung zu setzen." 49. KF 60. "Alle klassischen Dichtarten in ihrer strengen Reinheit sind jetzt lacherlich." 50. Friedrich Schlegel, Literary Notebooks: 1797-1801, edited by Hans Eichner (Toronto: University of Toronto Press, 1957), 116. " Jedes Gedicht eine Gattung fur sich." 51. A 139. Aus dem romantischen Gesichtspunkt haben auch die Abarten der Poesie, selbst die ekzentrischen und monstrosen, ihren Wert, als Materialien und Voriibungen der Umversalitat, wenn nur irgendetwas dnn ist, wenn sie nur original sind.
52. A 434. Die gewohnlichen Einteilungen der Poesie sind nur totes Fachwerk fur einen beschrankten Horizont. Was einer machen kann, oder was eben gilt, ist die ruhende Erde im Mittelpunkt. Im Universum der Poesie selbst aber ruht nichts, alles wird und verwandelt sich und bewegt sich harmonisch; und auch die Kometen haben unabanderliche Bewegungsgesetzte. Ehe sich aber der Lauf dieser Gestirne nicht berechnen, ihre Wiederkunft nicht vorherbestimmen lafit, ist das wahre Weltsystem der Poesie noch nicht entdeckt. 53. Hans Eichner, "Friedrich Schlegel's Theory of Romantic Poetry," PMLA 71(1956): 1018-41. 54. He explicitly excludes those of Fielding and Richardson, for example. See Eric A. Blackall, The NoveL) of the German Romantics (Ithaca: Cornell University Press, 1983), 22-24. 55. KFSA 2: 336. " . . . dafi ein Lied ebenso gut romantisch sein kann als eine Geschichte. Ja ich kann mir einen Roman kaum anders denken, als gemischt aus Erzahlung, Gesang und andern Formen."
56. KFSA 2: 336. Das Schauspiel soil auch romantisch sein, wie alle Dichtkunst; aber ein Roman ists nur unter gewissen Einschrankungen, ein angewandter Roman. Der dramatische Zusammenhang der Geschichte macht den Roman im Gegenteil noch keineswegs zum Ganzen, zum Werk, wenn er es nicht durch die Beziehung der ganzen Komposition auf eine hohere Einheit, als jene Einheit des Buchstabens,
Noted to Page* 71-73
239
iiber die er sich oft wegsetzt und wegsetzen darf, durch das Band der Ideen, durch einen geistigen Zentralpunkt wird. 57. Neue Zefochri/tfilr MiMik 10 (1839): 134. Einzzlne schone Erscheinungen dieser Gattung werden sicherlich hier und da zum Vorschein kommen und sind es schon; im iibngen aber, scheint es, hat die Form ihren Lebenskreis durchlaufen, und dies ist ja in der Ordnung der Dinge, und wir sollen nicht Jahrhunderte lang dasselbe wiederholen und auch auf neues bedacht sein. In addition to Linda Roesner and Charles Rosen (see notes 58 and 59, respectively), this passage has been cited by John Daverio, in "Schumann's 'Im Legendenton' and Friedrich ScUegel's Arabedke," 152; and by Nicholas Marston in Schumann: Fantasie, Op. 17 (Cambridge: Cambridge University Press, 1992), 26. The review from which this passage comes is quoted at length in Leon B. Plantinga, Schumann
240
Noted to Paged 74- 78
66. NeiieZeitdchriftfltrMudik 14 (1841): 181. "Die schonen Keime, die auch dieses sein letztes Werk in grofier Zahl enthalt, geben indes auf schonere Hoffnungen Anspruch" 67. NeueZeitdchriftftirMudik 15 (1841): 141. Chopin konnte jetzt alles ohne seinen Namen herausgeben, man wiirde ihn dock gleich erkennen. Darin liegt Lob und Tadel zugleich, jenes fur sein Talent, dieser fur sein Streben. Denn sicherlich wohnt ihm jene bedeutende Originalkraft inne, die, sobald sie sich zeigt, keinen Zweifel uber den Namen des Meisters ziilasst; dabei bringt er auch eine Fiille neuer Formen, die in ihrer Zartheit und Kilhnheit zugleich Bewunderung verdienen. Neu und erfmderisch immer im Aufierlichen, in der Gestaltung seiner Tonstiicke, in besonderen Instrumenteffecten, bleibt er sich aber Innerlichen gleich, dass wir furchten, er bringe es nicht hoher, als er es bis jetzt gebracht. For a similar interpretation of the same passage, see Meyer, Style andMiuic, 197. 68. Rosen, The Romantic Generation, 699. 69. See Roesner, "'Parallel' Forms," 266; and Joel Lester, "Robert Schumann and Sonata Forms," 19th-century Music 18 (1995): 190. 70. There are six quartets by Mozart in which the minuet is the second movement and the slow movement is the third movement—K. 170, 171, 387, 458, 464, and 499. B cause of the strength of the generic convention, we can easily identify these works as exceptional in this respect. 71. For this reason, Mozart's contributions to the Classical string quartet genre do not include every composition that he wrote for string quartet. The set of six quartets that he composed in Milan in 1772, K. 155—60, are in the Italian style, and each has only three movements. And although some of Mozart's Divertimenti were probably intended for string quartet, they are clearly not Classical string quartets, either. There are sixteen works in Mozart's oeuvre that qualify as string quartets in the Viennese Classical style: the set of six quartets that he composed in 1773 during a brief visit to Vienna, K. 168—173; the six quartets dedicated to Haydn, K. 387, K. 421, K. 428, K. 458, K. 464, and K. 465; the so-called Hoffmeuiter quartet, K. 499; and the three Pridian quartets, K. 575, K. 589, and K. 590. 72. Although all of the slow movements of Mozart's quartets are in diatonically related keys, there are a few exceptional pieces in the sonata-cycle works of Haydn and Mozart in which the slow movement stands in a chromatic relationship to the other movements. But again these are exceptions that prove the rule, as is demonstrated by the disproportionate attention that is paid to them in the scholarly literature. 73. Turchin, Schumann'd Song Cycled, 304-5. 74. The specific examples that he cites are a review of Schubert's Winterreiie in Allgemeiner mudikaluicher Anzeiger 1 (1829): 10—11 and a review of Schumann's songs in AUgemeine mudikalidche Zeitung 44 (1842): 30—33, 58—63. See John Daverio, "The Song Cycle: Journeys through a Romantic Landscape," in German Lieder in the. Nineteenth Century, edited by Rufus Hallmark (New York: Schirmer, 1996), 280-81. 75. Turchin, Scbumann'ti Song Cycled, 224ff., 238—40. She summarizes some of her supporting evidence in her subsequent article "Schumann's Song Cycles: The Cycle within the Song," 19th-century Madic 8 (1985): 232. 76. See Turchin, Schumann'd Song Cycled, 218; and Gerald Abraham, "Robert Schumann," in The New Grove Early Romantic Madterd I, edited by Stanley Sadie (New York: Norton, 1985), 124, 140.
Noted to Paged 79-80
1A1
77. NeutZcitocbriftfurMiuik 15 (1841): 14. Die Gedichte sind sammtlich aus dem Liebesfriihling genommen. Sie stehen hier nicht in dem urspriinglichen Zusammenhange, den ihnen die Anordnung des Dichters gegeben; der achtsamere Beobachter fiihlt indefi bald heraus, dafi ein feiner geistiger Faden die Reihe durchzieht und nicht die Rucksicht auf Fahigkeit fur musikalische Behandlung allein die Auswahl, und der Zufall die Ordnung bestimmte. So offenbart sich denn auch in der Auffassung der Gedichte, der verschiedenartigen technischen Behandlung ungeachtet eine Wahlverwandtschaft, welche indefi mehr zarter, geistiger Natur ist, wahrend im Formellen und durch wenige feine Ziige kaum angedeutet ist, dafi die in diesem oder jenem Liede ausgesprochene Situation nicht ein isolirtes, in sich abgerundetes Ganze, sondern das Glied einer Kette sei: So durch eine skizzenartige Gestaltung, wie die des dritten Liedes, oder durch eine nicht abschliefiende harmonische Schlufiwendung, wie im zweiten, die sich, das Lied als Selbststandiges genommen, wohl durch Frage des ersten, nicht aber fur die iibrigen rechtfertigen liefie. Eben aber in diesem losen Zusammenhange, in der nur leise andeutenden Ausfuhrung, offenbart sich, wie in der sinningen Auswahl der Gedichte, der gereifte Geschmack des gebildeten Kiinstlers, so wie in der ganzen technischen Zurichtung, Declamation, Harmonik u. dgl. die gewandte Hand des so geistreichen als schul- und stylfesten Tonsetzers sich leicht an der Leichtigkeit und Sicherheit der Arbeit, und an dem richtigen Tacte in Handhabung der Wiirz- und Ziermittel erkennen lafit. Von rein menschlichem Standpuncte aus betrachtet sind die Lieder so friedlicher, heitrer Natur, die in ihnen ausgesprochnen Stationen des Gemuthslebens sind so leicht und freundlich, dafi sie um so sichrer auf viele und warme Freunde rechnen konnen. 78. See Turchin, Schumann '
242
Notes to Pages 81-84
83. The delimtion, which I have quoted in full in chapter 1, is found in Musikalisches Lexicon auf Grundlage des Lexicon s von H. Ch. Koch, edited by Arrey von Dommer, 2d edition of-//. Ch. Koch's musikalisches Lexicon (Heidelberg: J. C. B. Mohr, 1865), 513—14. The review is in Allgemeine musikalische Zeitung, Wien 2 (1818): 478—79. It is cited by Barbara Turchin in "The Nineteenth-Century Wanderlieder Cycle," Journalo/Musicology 5 (1887): 501—2. 84. AUgemeine nuuikaiidche Zeitung 19 (1817): 73-75. Mit Freude, und mit Dank gegen den Meister, der sie ihm gewahrete, zeigt der Rec. diese, unter sich, in Dichtung und Musik, eng verbundenen Lieder an; gewiss, alle Leser, haben sie sich mit ihnen bekannt gemacht, werden mit ihm einstimmen, wenn er sie meist zu den schonsten zahlt, die wir seit mehren Jaahren erhalten haben; ja, ihm schwerlich widersprechen, •wenn er behauptet: besonders in Hinsicht auf Phantasie und Gefiihl gehoren sie unter die schonsten, die wir uberhaupt besitzen. Er kann sich nicht enthalten, Rechenschaft von dem Wesentlichsten zu geben, •was er hier gefunden hat; sollte dies auch, bey B.s Namen, und nach jener allgememen Erklarung, nicht nothig seyn. . . . . . . Nun erfolgt em kurzes, rasches Zwischenspiel, und tonet also aus, zugleich einen eben so zweckmassigen, als ungemem anmuthig ansprechenden Uebergang zum zweyten, etwas heiterern Liede bildende . . . Dies zweyte Lied gehet unrmttelbar in das dritte iiber, (G dur in As dur) . . . Desto herrlicher ist, nach kurzer Wendung durch C moll, die Riickkehr zur Tonart, so wie Gefuhlstimmung und Lieblichkeit des ersten Liedes, in dem letzten, wahrhaft vortrefflichen, das gegen das Ende, eben wie beym Dichter, jenes erste Lied selbst wieder anklingen lasst, nur zusammengedrangter, dann mit einem freyern, herzinnigen Schluss endiget, und so zugleich das Ganze, als einen wahrhaften Lieder-Jf/ri/, vollkommen befriedigend und nach Wunsch, abschliesst. 85. For discussion of the early reception of An 9t£ feme Geliebte, see Turchin, Schumann's Song Cycles, 62-69; and Ruth Bingham, The Song Cycle in German-Speaking Countries 1790-1840: Approaches to a Changing Genre (Ph.D. diss., Cornell University, 1993), 233-34. 86. Turchin, Schumann's Song Cycles, 216. 87. Ncue Zeitschrift fur Musik 5 (1836): 143. Fast alles A-Moll, wenig C-Dur. . . . Verlangen des Konigs nach Esther, drohender Ton, da sie jenen ausschlagt, drohendere Wiederholung. F-Dur geht nach Moll. . . . Im letzten Liede reiches A-Dur. Israel ist wohlhabend worden "Auch meine Zwillingstochter stehen wie Lilien Gottes aufgebliiht: doch muss ich still im Leide gehen" spricht Esther. Die Musik kehrt in das ursprungliche A-Moll ziiruck "das weisse Kreuz das ist das Zeichen, da find' ich meines Sohnes Grab." Leon Plantmga presents a reading of this review that is very similar to mine. See Schumann as Critic, 166—67. 88. Neue ZeitschriflfiirMusik 6 (1837): 141. "Em Cyklus von 5 Gesangen, von denen zwar jeder in sich musikahsch abgesclossen ist, die aber wie im Inhalt des Textes, so in der Wahl verwandter Tonarten zusammen ein Ganzes bilden." 89. The succession of keys is Bt major, Et major, C minor modulating to F minor, Bt minor, Bt major. See Turchin, Schumann's Song Cycles, 218, 238. 90. Allgemeincrmusikalischer Anzeiger 12 (1840): 114.
Noted to Paged 84-87
243
Wie von dem Inhalte der Worte bedingt, zieht sich auch durch das Ganze ein reizender Verbindungsfaden, einschliissig des gewahlten Tonarten-Cyclus, welche im analogen Wechsel D-dur, H-moll, H-dur, E-moll, E-dur, A-dur, D-moll und Dur den geheimen Nexus ihrer Wahlverwandtschaft beurkunden. Turchm identifies the author of this review as Ignaz Xaver Ritter von Seyfned. See "The Cycle within the Song," 243, fn. 9. 91. Felix Mendelssohn-Bartholdy, Briefe van Felix Mendelddohn-BarthoOy an Ignaz und Charlotte Modchekd, edited by Felix Moscheles (Leipzig, 1888), 207-8. Deine gutige Erlaubnis, ein Heft Lieder aus den zehn durch Dr. Becher erhaltenen zusammenzustellen, habe ich benutzt, und da Kistner sechs fiir ein Heft verlangte, folgende gewahlt: . . . Die Tonarten folgen sich zwar aufs Allertollste: F dur und H dur und alles durcheinander; aber ich habe immer gefunden, dafi Einem kein Mensch die schonste Tonartenfolge dankt, dagegen eine gewisse Abwechslung von langsam und schnell, ernst und heiter durchaus verlangt wird; daher verzeih das Tonartenfricassee. 92. Turchm, Schumann',) Song Cycled, 127—28. 93. NeueZeitdchrift/iirMiuik 23 (1845): 14. Diese Lieder tragen em ganz eigenthumliches Geprage; die Begleitung ergeht sich in graciosen Figuren und iiberraschenden Wendungen; die Melodien smd innig und frisch. Indem wir auf den Aufsatz in dies Blattern: "Robert Schumann nut Riicksicht auf Mendelssohn-Bartholdy" verweisen, in welchem Schumann's Eigenthiimlichkeit im Allgemeinem naher beleuchtet wird, gehen wir zu den emzelnen Liedern fiber, und heben diejenigen hervor, die uns am besten gefallen haben: Nr. 1 "Im wunderschonen Monat Mai." Die Begleitung leitet schembar nach Fis-Moll ein, die Singstimme wendet sich jedoch iiberraschend A-Dur zu und schliefit eigenthumlicherweise auf dem D-Dur Accord, wahrend die Begleitung mit Cis-Dur, als Dominante von Fis-Moll endet. —Nr. 5. "Ich will meme Seele tauschen," ganz besonders schon; ein eigner Zauber ist fiber das Ganze verbreitet. Aufiergewohnlich ist der Anfant—(das Lied ist in H-Moll) die Singstimme mit h im Auftact allein beginnend—die Begleitung sich mit den beiden Septimenaccorden cis-e-g-h, dann fis-ais-cis-e unter sie schmiegend in Harpegglen. —Nr. 7. "Ich grolle nicht und wenn das Herz auch bricht" besonders charakteristisch. —Nr. 8. "Und wiifiten's die Blumen"—die Begleitung, wie sauselnder Abendwind; die Singstimme sehr ausdrucksvoll. Von alien Compositionen, die wir iiber diese Worte kennen, scheint uns diese die beste zu sem. — Nr. 15. "Aus alten Mahrchen winkt es hervor" bewegt sich in dem Lieblingsreich des Compomsten, im Reich des Geisterhaften und Phantastischen, im Lande der Sehnsucht, in dem er besonders gluckhch ist. —Nr. 16. "Die alten bosen Lieder die lafit uns jetzt begraben" ist mit viel Humor componirt. — Schade nur, dafi viele Leder dieses Cyklus wie Schmetterkmge fluchtig an uns vorfibereilen und us gar keine Zeit lassen, an den schonen, schillernden Farben uns so recht con amore zu ergozen. 94. Rosen, The Romantic Generation, 41-48, 53-54, 204-20.
244
Note* to Paged 92-101
Chapter 4 1. David Neumeyer, "Organic Structure and the Song Cycle: Another Look at Schumann's Dichterliebe," MIMIC Theory Spectrum 4 (1982): 92-105; Patrick McCreless, "Song Order in the Song Cycle: Schumann's LiederkreL), Op. 39," MIMIC AnalydLt 5 (1986): 5-28; Arthur Komar, ed., Schumann: Dichterliebe (New York: Norton, 1971), 63, 12. 2. The latter song has been analyzed by Felix Salzer in Structural Hearing: Tonal Coherence inMudic (New York: Charles Boni, 1952), 1: 154, 2: 98. The former song has been analyzed by numerous theorists and musicologists. In the last two chapters, I have cited Charles Rosen, David Neumeyer, Arthur Komar, and Heinrich Schenker. 3. Several studies are concerned with Schumann's process of composing cycles. The most important is Rufus Hallmark, The Gene
Notes to Pages 104-18
245
to represent the actual process by which Schumann composed the phrase, and in fact, as I will discuss in a later section of this chapter, the piano draft of the song reveals that in his original conception of the phrase it was already expanded and had a completely different harmonic structure. 13. The analytic technique of rhythmic normalization also comes from Rothstein's work. See William Rothstein, "Rhythmic Displacement and Rhythmic Normalization," in Trends in Schenkerian Research, edited by Allen Cadwallader (New York: Schirmer, 1990), 87-113. 14. When Schumann first composed the song, he retained the neighboring note formula in measure 18 and compensated for the fact that there are fewer syllables of text in this phrase by slowing down the first measure, which sets the words "fiber mir." He subsequently revised the rhythm of the vocal line in this phrase twice, writing the final version directly on the engraver's copy, which was prepared by Clara Wieck. See Reinhold Brinkmann, Schumann and Eichendorff: Stiidien zum Liederkreis Opus 39 vol. 95 of MiufikKonzepte, edited by Heinz-Klaus Metzgere and Rainer Riehn (Munich: 1997), 74. I will return to these revisions in a later section of this chapter. 15. Rosen, one of the few scholars who discusses this type of ending, writes of the final cadence of "Ich wandelte unter den Baumen," from the Heine Liederkreis, that its "resolution is not questioned, but only left open to question." See Charles Rosen, The Romantic Generation (Cambridge: Harvard University Press, 1995), 655. 16. Schumann's first version of the coda is not exactly the same as the published version, but where the B section was recomposed in its entirety, the ending was revised only slightly. The most significant difference in the original version is that the vocal melody ends on ct2, rather than ft1, which means that the dissonant d2 is resolved, but the vocal melody as a whole is left open. In revising the passage, Schumann found a subtler way to achieve the same expressive effect. The piano drafts for the Eichendorff songs are found in Deutsche Staatsbibliothek, Berlin, Music manuscript autograph 16. For a photographic facsimile, see Herwig Knaus, Musiksprache und Werktitruktur in Robert Schumann^ Liederkreis, vol. 27 of Schriften ziLrMiisik, edited by Walter Kolneder (Munich: Musikverlag Emil Katzbichler, 1974). For a complete transcription of the "In der Fremde" draft, see Rufus Hallmark, "Schumann's Revisions of 'In der Fremde,' Op. 39, No. 1, presented at the Conference of NineteenthCentury Music, Nottingham, 1978. 17. Bayerische Staatsbibliothek, Munich, Mus. ms. 1589. 18. Reinhold Brinkmann regards Schumann's desire to change the rhythm and text setting of the fourth phrase to be the primary motivating factor m his revisions of "In der Fremde" and thus argues that there are three distinct versions of the song, the last of •which is documented in the Munich draft. See Schumann und Eichendorff, 73—74. 19. Viktor Ernst Wolff, Robert Schumann*) Lieder in ersten und spdteren Fassungen (Leipzig: H. S. Hermann, 1914), 54—58; Knaus, Musiksprache und Werkstruktur, Barbara Turchin, Robert Schumann's Song Cycled in the Context of the Early Nineteenth-Century Liederkreis (Ph.;D. diss., Columbia University, 1981), 314-18, and "The Cycle within the Song," 238—40; Hallmark, "Schumann's Revisions." 20. Hallmark, "Schumann's Revisions." 21. Leonard B. Meyer, Explaining Music (Berkeley: University of California Press, 1973), 23. 22. Turchin, "The Cycle within the Song," 238. 23. Turchin, "The Cycle within the Song," 233, 238-40. 24. Brinkmann, Schumann und Eichendorff, 73—74. 25. Brinkmann, Schumann und Eichendorff, 75.
246
Noted to Paged 121-36
Chapter 5 1. Barbara Hernstein Smith, Poetic Closure: A Study of How Poenw End (Chicago: University of Chicago Press, 1968), 2. 2. Edward T. Cone, Miutical Form arid Miuical Performance (New York: Norton, 1968), 16-20. For further discussions of the evasion of closure in Romantic music, see Leonard B. Meyer, Style and Miuic (Philadelphia: University of Pennsylvania Press, 1989), chapters 7 and 8; and Charles Rosen, The Romantic Generation (Cambridge: Harvard University Press, 1995), chapters 1 and 2. 3. Smith, Poetic Closure, 2. 4. Meyer, Style and Miuic, 303-25, 203-4. 5. Meyer, Style and Miuic, 246—52. There is also a discussion of complementary melodies in his Explaining Miuiic (Berkeley: University of California Press, 1973), 175—83. 6. Because the complement of the melody is not an exact inversion of the model presented m measures 2—3 and 4—5, it does not imply the specific pitch, fi1, that occurs in the vocal melody on the downbeat of measure 7. But the pattern does imply a leap, rather than the stepwise motion that is dictated by the parallel tenths in the accompaniment. Compare measure 24, the analogous spot in the recapitulation. 7. In fact, this motive is not really new but is derived from the sequential descending line in the right hand of the piano in measures 1—4. 8. Joseph Freiherr von Eichendorff, Samtliche Werke de<> Freiherrn Joseph con Eichendorff: HLftorLich-Krituiche Aiugabe, vol. 1, bk. 1, edited by Harry Frohlich and Ursula Regener (Stuttgart: Kohlhammer, 1993), 239. The translations in this chapter from this work are my own. When he set the poem, Schumann made two minor alterations, changing "Ueber'n" to "Uber'm" in the first line and changing "Und in Traumen" to "Und im Traumen" in line 10. 9. See, for example, Theodor Adorno, "Zum Gedachtnis Eichendorffs," in Noten zur Literatur I (Frankfurt: Suhrkamp Verlag, 1958), 137; Jiirgen Thym, The Solo Song Settings of Eichendorff j Poenu by Schumann and Wolf (Ph.D. diss., Case Western Reserve University, 1974), 212; Barbara Turchin, Robert Schumann'*! Song Cycled in the Context of the Early Nineteenth-Century Liederkreut (Ph.D. diss., Columbia University, 1981), 279-80; Charles Burkhart, "Departures from the Norm in Two Songs from Schumann's Liederkreu," in Schenker Studies, edited by Hedi Siegel (Cambridge: Cambridge University Press, 1990), 164; and John Daverio, Robert Schumann: Herald of a "New Poetic Age" (New York: Oxford University Press, 1997), 216. I will return to this interpretation and to the question of how the text relates to the rest of the cycle in chapter 8. 10. In Eichendorffs collected poems, in fact, "Fruhhngsnacht" appears m the section titled "Friihling und Liebe." 11. Eichendorff, SamtLiche Werke, 79. 12. See Lawrence Radner, Eichendorff: The Spiritual Geometer (Lafayette, IN: Purdue University Studies, 1970), 305. For discussions of "Intermezzo" as a love poem, see Eric Sams, The Songs of Robert Schumann (London: Methuen, 1969), 94; Thym, Solo Song Settings, 220; Barbara Turchin, "Schumann's Song Cycles: The Cycle within the Song," 19thcentury Miutic 8 (1985); 241: and Daverio, Robert Schumann, 216. 13. See August Wilhelm Schlegel, "Uber Zeichnungen zu Gedichten," in Athenaeum: Eine Zeitjchrift von August Wilhelm Schlegel und Friedrich Schlegel, vol. 2, edited by Curt Griitzmacher, (Munich: Rowohlt, 1969). There is a discussion of this essay in chapter 3. 14. This recomposition of the opening, at the moment of its return, is very similar to Schumann's procedure in "Aus meinen Thranen spriessen," from Dichterliebe. See my discussion in chapter 2.
Note to Pages 137-45
247
15. Interestingly, while the parallel tenths clarify the counterpoint, they simultaneously obscure the pattern of appoggiaturas, by emphasizing the b1 on the second beat of measure 18. Chapter 6 1. For recent discussions of the opus 5 Impromptus, see Charles Rosen, The Romantic Generation (Cambridge: Harvard University Press, 1995), 658—69; and John Daverio, Robert Schumann: HeraQ of a "New Poetic Age" (New York: Oxford University Press, 1997), 106-9. 2. NeueZeitsckriftfiirMusik 5 (1836): 63. Der die ersten Variationen ersonnen (doch am Ende wieder Bach) war gewifi kein iibler Mann. . . . Denn gewifi ist in keinem Genre unsrer Kunst mehr Stiimperhaftes zu Tag gefordert worden—und wird es auch noch. Von der Armseligkeit, wie sie hier aus dem Grunde bliiht, von dieser Gemeinheit, die sich gar nicht mehr schamt, hat man kaum einen Begriff. Sonst gab's doch wenigstens gute langweilige deutsche Themas, jetzt mufi man aber die abgedroschensten italienischen in funf bis sechs wasserigen Zersetzungen nacheinander hinterschlucken. For an extended discussion of Schumann's views on virtuoso variations, see Leon B. Plantinga, Schumann a
248
Notu to Paged 145-57
10. See Reinhard Thum, "Cliche and Stereotype: An Examination of the Lyric Landscape in Eichendorff's Poetry," Philological Quarterly 62 (1983): 450—51; and Egon Schwarz, Joseph fan Eichendorff (New York: Twayne, 1972), 100. 11. For a discussion of this aspect of the poem, see Brinkmann, Schumann and Eichendorff, 20-21. 12. Eichendorff, Samtliche Werke, 39. Schumann changed the punctuation in lines 7 and 8 so that they read as follows: "Was sprichst du wirr, wie in Traumen, zu mir, phantastische Nacht!" He also changed "grofiem" to "grofien" in line 12. 13. "Schone Fremde," which was first published in the Roman Dichter und ibre GejeLLen, appears in Eichendorff's collected edition of his poems in the section titled "Wanderheder," and "Mondnacht" appears in the section "Geistliche Gedichte." 14. See Rosen, The Romantic Generation, 698. 15. Brinkmann argues, however, that in the A section of the song it is unclear whether the key is E major or B major. And Jon W. Finson writes that the ending of each eight-measure phrase "creates a plagal progression that constantly reinterprets a B-major chord as tonic every two lines." See Brinkmann, Schumann und Eichendorff, 25—26; and Finson, "Schumann's Mature Style and the 'Album of Songs for the Young,'" Journal of Miuicology 8 (1990): 237-38. 16. The disagreement results from the fact that the vocal melody and the accompaniment do not correspond in terms of their phrase rhythm. Brinkmann and Finson both follow the accompaniment and hear measures 6—13 as the first eight-measure phrase. But Rosen and Charles Burkhart, following the vocal melody, hear measures 7—14. The only commentator who has mentioned the disjunction between melody and accompaniment is August Gerstmeier. He analyzes the vocal melody of the first strophe as a series of threemeasure phrases, each of which is followed by a silent fourth measure, beginning in measure 7. And he analyzes the accompaniment in the same strophe as follows: a fivemeasure phrase, 5—9, two four-measure phrases, 10—13 and 14—17, and a five-measure phrase, 18—22. See Brinkmann, Schumann and Eichendorff, 22; Finson, "Schumann's Mature Style," 237; Rosen, The Romantic Generation, 698; Burkhart, "Departures from the Norm in Two Songs from Schumann's LicderkreLi," in Schenker Studies, edited by Hedi Siegel (Cambridge: Cambridge University Press, 1990), 148; and Gerstmeier, Die Litder Schumann* (Tutzing: Hans Schneider, 1982), 22-25. 17. For an explanation of the harmonic structure of the introduction, see Edward Aldwell and Carl Schachter, Harmony am) Voice Leading, vol. 2 (New York: Harcourt Brace Jovanovich, 1978), 130-31. Burkhart follows their analysis, but Turchin, Gerstmeier, and Rosen all describe the opening harmony as a V9 chord. See Burkhart, "Departures from the Norm," 154; Turchin, Schumann 4 Song Cycles, 320; Gerstmeier, Die Lieder Schumann^, 24; and Rosen, The Romantic Generation, 698. 18. See Burkhart, "Departures from the Norm," 153; and Brinkmann, Schumann und Eichendorff, 28-29. 19. See Gerstmeier, Die Lieder Schumann*), 22. Finson, however, argues that "the medieval cast of Langze'denvert) would be inappropriate" for "Mondnacht" and instead reads the poem as "iambic trimeter, with a feminine ending at the end of each odd line." See "Schumann's Mature Style," 237. 20. Rosen, The Romantic Generation, 698. 21. It is interesting that when el' first appears in this measure of the phrase, in measure 14, and again in measure 36 it is an anticipation. Now Schumann reveals that this is indeed its proper position. 22. Brinkmann interprets the ending of "Mondnacht" quite differently from me. He
Note* to Paged 157- 72
249
observes that Schumann had a great deal of trouble with the ending, revising both vocal line and accompaniment extensively in several stages on the Berlin autograph and then going back and forth between two possible endings for the vocal line as the song was published and then twice reprinted. Brinkmann explains Schumann's difficulties as the result of his "obvious inclination towards a definite ending, in the indicative tense, because of his desire to establish and emphasize 'nach Haus,' while at the same time it is certainly clear (and was presumably clear to Schumann as well) that Eichendorff's poem does not bear out this inclination." Schumann and Eichendorff, 36—43. 23. Turchin, Schumann'^ Song Cycled, 368. 24. Turchin, Burkhart, and Rosen all call attention to this similarity. See Turchin, "The Cycle within the Song," 241; Burkhart, "Departures from the Norm," 158; and Rosen, The Romantic Generation, 213. 25. Burkhart, "Departures from the Norm," 162. 26. Gerstmeier observes that the indication in Schumann's manuscript draft of "Schone Fremde" is "sprechend, flusternd" and suggests that Schumann's intent at this point in the song is a sudden change into declamatory prose, to be sung in a free parlando style, an effect that is facilitated by the ritardando. See Die Lieder Schumann^, 63. 27. Burkhart analyzes the phrase rhythm of the song quite differently from me and hears the second phrase as an eight-measure expansion of an underlying four-measure phrase. See "Departures from the Norm," 162. 28. The upper two voices in my example are inverted in the song, with the middle voice displaced above the soprano. 29. Both Burkhart and Gerstmeier discuss the unusual and mysterious nature of this return. But at least two commentators apparently do not hear measures 11—15 as a return at all. Theodor Adorno writes that the third strophe "is clearly of the nature of an Abgesang; but on the whole, the song dispenses with the symmetry of repetition." Similarly, Jiirgen Thym writes of "Schone Fremde" that it is "through-composed and has no large elements of repetition that define the form." See Burkhart, "Departures from the Norm," 162; Gerstmeier, Die LUder Schumann^, 72; Adorno, "Zum Gedachtnis Eichendorffs," in Noten zurLiteraturI (Frankfurt: Suhrkamp Verlag, 1958), 141; and Thym, Sola Song Settings, 207. 30. The postlude of "Schone Fremde" is, exceptionally, more like the postludes of Dichterliebe, in that it does not seem to be motivated by the vocal cadence but essentially introduces a new musical idea.
Chapter 7 1. Patrick McCreless, "Song Order in the Song Cycle: Schumann's Liederkn'u, Op. 78," MiuicAnalyoii 5 (1986): 5. 2. Although the February 1 draft of "Schlusslied des Narren" is the earliest surviving source for any of Schumann's mature songs, several scholars have suggested that he may have begun to compose songs several months earlier. Fritz Feldman refers to a letter of October 10, 1839, in which Schumann writes that he has begun fifty new compositions, and argues that at least some of them may have been songs. More recently, Barbara Turchin has called attention to an even earlier letter, of August 30, in which he writes to August Pott that he is sending a "Gesangstiick" for a Mozart album that Pott •was compiling, although she notes that a few days later he referred to the piece in a different letter as a small "fughetta." See Feldman, "Zur Frage des 'Liederjahres' bei Robert Schu-
250
Notes to Paged 173-76
mann," Arckiv filr Mudikwiddendchaft 9 (1952): 248—50; Turchin, "Schumann's Conversion to Vocal Music: A Reconsideration," Mudical Quarterly 67 (1981): 402-3; and also Rufus Hallmark, "Robert Schumann: The Poet Sings," in German Lieder in the N ineteenth Century, edited by Hallmark (New York: Schirmer, 1996), 77. If Schumann did compose any songs before writing the February 1 draft, however, they are presumably lost, since he dated almost all of the song drafts that he composed in 1840—41, as I shall explain later. John Daverio, for one, believes that the compositions referred to in Schumann's letter of October 10, 1839, are for piano and that "Schlusslied des Narren" is his earliest mature song. See Daverio, Robert Schumann: Herald of a "New Poetic Age" (New York: Oxford University Press, 1997), 533, fn. 28. 3. Hallmark, "The Poet Sings," 83, Table 8.3; 114, fn. 5. 4. As I have observed in part I, Schumann and his contemporaries are inconsistent in how they designate their publications of songs. But while it is common for works intended as song cycles to bear a neutral designation, it would be very unusual for a miscellaneous collection of songs to be designated as a Liederkreid or a Liedercyklud. If this is what Schumann intended in the case ofMyrthen, it would be unique among his song publications. 5. For a complete description of the notebooks, including a listing of all of the drafts, see Viktor Ernst Wolff, Robert Schumanru Lieder in erdten und dpateren Faddungen (Leipzig: H. S. Hermann, 1914), 9—24. Barbara Turchin lists the numbers and dates from the drafts for six of Schumann's cycles. See Robert Schumann d Song Cycled in the Context of the Early Nineteenth-Century Liederkreid (Ph.D. diss., Columbia University, 1981), appendix 2. 6. The four other publications are: opus 57, which was published in 1846 and consists of a single song, Heine's "Belsatzar"; opus 77, published in 1851, which includes Eichendorff's "Der frohe Wandersmann," a second 1840 song for which no draft survives, and three later songs; as well as opus 29 and opus 34, both published in 1841, whose drafts were not included in the Berlin notebooks because they consist of partsongs and duets. 7. The drafts for Dichterliebe include twenty songs and bear the title Zwanzig Lieder and Heined Liederbuch. 8. Schumann placed drafts of his seven solo songs from this set into the notebooks. He presumably left out the drafts of the remaining five songs because two are duets and the other three are by Clara Schumann. See Rufus Hallmark, "The Riickert Lieder of Robert and Clara Schumann," 19th-centuryMudic 14 (1990): 6. 9. Robert Schumann and Clara Schumann, Briefe einer Liebe, edited by Hanns-Josef Ortheil (Konigstein: Athenaum Verlag, 1982), 263. Translation is mine. "Meistens mach' ich sie stehend oder gehend, nicht am Klavier. Es ist doch eine ganz andere Musik, die nicht erst durch die Finger getragen wird—viel unmittelbarer und melodioser." 10. Rufus Hallmark, The Genedid of Schumann'd Dichterliebe: A Source Study (Ann Arbor: UMI, 1976), 19. See also his essay "Die handschriftlichen Quellen der Lieder Robert Schumanns," in Robert Schumann—ein romantidched Erbe in neuer Fordchung, edited by the Robert-Schumann-Gesellschaft, Diisseldorf (Mainz: Schott, 1984), 99-117. 11. There is a list of the extant vocal sketches and fair copies for opp. 24, 25, 35, 39, 42, and 48 in Turchin, Schumann'd Song Cycled, appendix 2. For a more complete but much older account of the manuscript sources for Schumann's songs, see Georg Eismann, "Nachweis der internationalen Standorte von Notenautographen Robert Schumanns," in Sammelbande der Robert-Schumann-Gedeltdchaft, vol. 2 (Leipzig: VEB deutsche Verlag fur Musik, 1966), 7-37. 12. See Turchin, Schumann'd Song Cycled, 410.
Noted to Pag^ 176-86
251
13. Hallmark, Genedu of Dichterliebe, 19. 14. Hallmark describes the extant sources for Dichterliebe and tells their history in Gcnu'uaf Dichterliebe, 15-23 and 110-28. 15. Hallmark, Gen&fif of Dichterliebe, 112-14. 16. McCreless, "Song Order," 18. 17. The phrase is from McCreless, "Song Order," 18. 18. Hallmark, "Riickert Lieder, 5-6. 19. Hallmark, "Ruckert Lieder, 8-11. 20. The source material for Myrthen has not been studied as comprehensively as the manuscripts for some of Schumann's other cycles. Turchin presents a brief narrative account of its compositional history and a list of the surviving manuscript sources. See Schumann'd Song Cycled, 264—68 and 410—12. 21. Robert Schumann, Robert Schumann'*! Leben and deinen Brie/en, vol. 1, edited by Hermann Erler (Berlin: Verlag von Ries and Erler, 1887), 236-37. Der Titel 1st: Myrthen Liedercyklus in vier Heften von R. Sch. Die Texte sind von Ruckert, Gothe, Heine, Burns und Byron. . . . Die zwei ersten [Hefte] wfinschte ich bis Ende Mai, die letzten zwei bis Ende August beendigt. . . . Das erste Heft enthalt: "Widmung" v. Riickert, "Lotosblume" von Heine, " Jemand" von Burtu, zwei Brautlieder v. Riickert, und "Mein Herz ist im Hochland" v. Burnd. Montag erhalte ich die Remschrift des Isten Heftes und Sie konnten gleich mit dem Stich beginnen lassen. Schumann's impatience to finish Myrthen was largely due to his intention to give his bride the published work as a wedding gift. He may have been encouraged by the quick pace of opus 24, which was already being engraved on February 28, five days after he finished composing the songs. 22. For a transcription of this sheet, see Turchin, Schumann'd Song Cycled, 267. 23. Herwig Knaus, Mudikdprache utu) Werkdtruktur in Robert Schumann^ Liederkreu, vol. 27 of Schri/ten ziir Mudik, edited by Walter Kolneder (Munich: Musikverlag Emil Katzbichler, 1974), 14. Two of Knaus's dates are conjectural. As shown in Table 7.4, the draft of "Intermezzo" is dated simply "Mai 1840" and "Auf einer Burg" is undated. 24. The draft numbers are continuous throughout the three volumes, but Schumann paginated each volume separately. Wolff presents a table of all three volumes in his dissertation. As Turchin points out, however, his date for "Der Nufibaum," May 16, is a misreading. Wolff, Robert Schumann,! Lieder, 12—19; Turchm, Schumann d Song Cycles, 266, fn. 25. Although "Intermezzo" has no exact date, Knaus's supposition that this song •was composed before May 16 is probably correct, since it appears in the first volume of the notebooks. Although Schumann added a draft that he composed on July 14 to the end of this volume, the latest date on any of the other drafts is May 9, and the second volume begins with a draft dated May 12, so it is likely that "Intermezzo" was composed between these two dates.
252
Noted to Paged 186-96
26. "Friihlingsnacht" ends at the bottom of one page, and "Zwielicht" begins at the top of another. Schumann placed the draft for "Die Stille" in between these two drafts in the notebook. 27. Robert Schumann, Jiujendbriefe von Robert Schumann, edited by Clara Schumann (Leipzig: Breitkopf und Hartel, 1885), 314. "Eichendorffsche sind es zwolf. Die hab' ich aber schon vergessen und etwas Neues angefangen." 28. This is Rufus Hallmark's supposition as well. See "Schumann's Revisions of 'In der Fremde,' Op. 39, No. 1," presented at the Conference of Nineteenth-Century Music, Nottingham, 1978. 29. For a description of this source, see Hallmark, "Schumann's Revisions." 30. Robert Schumann, Robert Schumann in deinen Schriften uni Brief en, edited by Wolfgang Boetticher (Berlin: Hahnefeld, 1942), 341. "Auch findest Du vieles zu arbeiten vor und ich habe schon alles bereit gelegt, und zur kleinen Belohnung auch das Liederbuch zum Andenken an niich. Es ist fast ein ebenso dickes schon wieder fertig. Da wirst Du Augen machen." 31. Both Hallmark and Turchin interpret the Liederbuch as a manuscript copy of the Eichendorff Liederkreid and the other, "just as thick," as a copy of the Dichterliebe songs. But this reading of the passage seems implausible for two reasons. First, Schumann completed his manuscript of Dichterliebe on June 1, three days before he wrote the letter, so why would he describe it as "almost finished"? Second, as Hallmark himself has argued, the fair copy of the Eichendorff Liederkreid was apparently written by Wieck herself, so how could Schumann be offering it to her as "a memento"? See Hallmark, Genedid of Dichterliebe, 23; and Turchin, Schumann'd Song Cycled, 284. 32. See Charles Burkhart, "Departures from the Norm in Two Songs from Schumann's Liederkreid," in Schenker Studied, edited by Hedi Siegel (Cambridge: Cambridge University Press, 1990), 157. 33. Neue Zeitdchrift fur Mudik 15 (1841): 14. The review from which this quotation comes is quoted in full in chapter 3.
Chapter 8 1. Neue ZeitdchriftfilrMudik 13 (1840): 119. "Auch Veit wendet auf die Wahrheit des musikalischen Ausdrucks in Wiedergabe der Worte die treuste Sorgfalt. Dies Lob steht fiber jedes andere." 2. Neue Zeitdchrift filrMudik 13 (1840): 118. Die Komposition ist in schmerzlicher Zeit entstanden, tiefmelancholisch, aber zur innigsten Teilnahme anregend, und wahr. Wahr—zittert euch nicht euer kleines Herz, Kompomsten, wenn ihr dieses ^Vort hort! Bettet euch immer weicher in eure schonen Gesangesliigen, ihr bringt's doch nicht hoher, als von einigen andern Judaslippen gesungen zu werden, vielleicht verfuhrerisch genug. Aber, tritt dann wieder einmal ein wahrhaftiger Sanger unter euch, so fliichtet nut eurer erheuchelten Kunst, oder lernt Wahrheit, wenn es noch moglich ist. Wahr ist denn auch Burgmuller durch und durch; noch mehr, er giebt die Wahrheit auch meistens in schonem Gewand. 3. NeueZeitdchriftfiirMudik
2 (1835): 83.
Das Lied ist der reinste Ausdruck unsers Gefuhls, in der nattirlichsten Form gegeben; es soil wahr ausdriicken, was man wahr empfunden, aber zugleich mit
Note to Pagu 196-97
253
dem Reiz der Neuheit und der Wiirde der Kinfachheit; es soil zu uns reden mit jenen kleinen Beglaubigungszeichen der Selbstempfindung, die unwiderstehlich in verwandten Gemiithern anklingen und ziinden . . . As is typical for the Neue 2Zeitjchrift, the article is signed not with the author's name but with a number, 6. This number was typically used by Banck. See Leon B. Plantinga, Schumann cw Critic (New Haven: Yale University Press, 1967), 13, fn. 45. 4. NeueZeitocbriftfurMtuik 1 (1835): 83. Aber wie zuletzt alle Darstellung in der Kunst auf tiefer Selbstanschauung und Empfindung, auf dem Eindringen in die Sprache der Natur und des menschlichen Herzens beruht, so auch hier. Wer . . . nie die Grundtone seines eigenen Geftihls . . . recht beobachtet hat, und nicht mit Geisteskraft die im Leben voruberrauschenden Gemiithsbewegungen festhalten kann, der vermag auch nicht, sie uns treu wiederzugeben. Das alles aber ist die Aufgabe eines feinen Geistes, der seiner Mittel vollkommen Herr ist. 5. NcueZeitMhriftfiirMiuik
1 (1835): 92.
Diesen Compositionen fiihlt sich das Gemachte, die Routine, das HanStverk durch; es sind die Lieder jener geiibten Tonsetzer, die grofle Leichtigkeit in Handhabung der gewohnhchen oder auch ihnen eigenthumhchen Mittel fur Wabrheit der Empfindung und ihrer musikalischen Uebertragung haben. Hier handelt es sich nicht um die Wiedergeburt der Poesie zum schoneren, bliihenderen Leben im Gesange, sondern der Tonsetzer hat die Worte mehr zu Dolmetschern seiner musikalischen Sprache gemacht, deren Gelaufigkeit er fur Beredsamkeit halt. Er umhiillt die strahlenden Blicke seines Geistes mit dem Nebel des Gleichmuths, gibt statt dichterischer Gedanken poetische Floskeln. 6. NcueZci&chriftfiirMiuik. 7. NcueZeikicbriftfUrMMik
5 (1836): 143. 6 (1837): 135.
Wenn ich in deme Augen seh" em klemes Genrebildchen ist, in welchem der grofie Aufwand von starken Lichteffecten in keinem Verhaltmfi zu dem engen Rahmen steht. Es beginnt in As-Dur, weicht nach 4 Tacten nach Es, dann nach Des-Dur aus, fallt ohne Weiteres nach A-Dur und gleich darauf nach As-Dur und schliefit in F-Moll; und das Alles in 17 Tacten. 8. Neiie Zeitjchrift furMiwik 17 (1842): 151. "Um es kurz zu sagen, es scheint uns zu viel Aufwand gerade an diese Texte verschwendet; es sind zu viel Noten zu den einfachen Worten." 9. NeueZeihtchr^tfurMiuik 20 (1844): 197. Der geistreiche und bereits schon, wenn auch nicht in dieser Ausdehnung, benutzte Gedanke, verschiedene lyrische Dichtungen zusammenzustellen und durch den Gegensatz der darin ausgepragten Gefuhlszustande einen erzahlenden Fortschritt zu erlangen, ist in dem vorliegenden Werke auf eine sehr anzuerkennende Weise verwirklicht worden. . . . Das Hauptverdienst des Compomsten wurde wohl in dem ansprechenden Grundgedanken des Ganzen und in der schopferischen Auswahl und Zusammenstellung der schonen Gedichte zu suchen sein . . .
254
Noted to Pages 197-99
10. Neii£ ZeitjcbriftfiirMiuikik 15 (1841): 14. For the complete German text, see chapter 3, note 77. 11. Charles Rosen, The Romantic Generation (Cambridge: Harvard University Press, 1995), 208; Jiirgen Thym, The Solo Song Settings of Eicheni)orff'
5 (1836): 144.
Wo die Dichtung aber sinnlicher, malerischer, indischer wird, bleibt die Musik meistens zu ungefallig zuriick; man will dann mehr, weichen fleischigen Ton; im gleichen Grad, wie hier Franz Schubert, Lowe und viele der neueren oft zu materiell auftragen, thut es Klein zu wenig . . . 14. Neue ZeiUchnft farMiuik 17 (1842): 151. Die Burnsschen Gedichte . . . sind wohl Ergiisse einer wahrhaften Dichterstimrnung, aber immer schlicht, kurz und biindig; darum lieben ihn die Komponisten auch so sehr, darum fiigen sich seine Worte wie von selbst zum Liede, und am naturlichsten in jene Form, wie sie dem wirklichen Volksliede eigen ist. Der Komponist wollte aber mehr als diese; er giebt meistens grofie ausgefuhrte Stiicke, die wohl einen strebsamen Musiker verraten, mit der naiven Form der Gedichte aber im Widerspruche stehen; oft hat seine Musik sogar einen dramatischen oder theatralischen Anstrich, und hier scheint er am weitesten vom Ziele zu sein. 15. Neue ZeitMhrift filr Muoik 6 (1837): 136. Damit haben wir iiberhaupt einen Hauptvorzug dieser Lieder ausgesprochen: mit wenigen Mitteln ist Viel gewirkt. Ein zweiter ist die echt liedermafiige
Noted to Paged 200-204
255
Selbststandigkeit der Melodic, die zwar nicht darin besteht, dafi sie wie beim eigentlichen Volkslied der Harmonisirung ganz entbehren konnte, die aber ihrer nicht bedar, um erst eine individuelle Farbung zu erhalten, oder um iiberhaupt erst Musik, Gesang zu werden aus bloser Declamation. Man verlangt dramatische Wahrheit auch vom Liede. Wir stimmen bei, nur unterscheide man dramatisch und theatrlisch, was nicht immer geschieht. 16. The most exhaustive history of the Liederdpiel is found in Luise Eitel Peake, The Song Cycle: A Preliminary Inquiry into the Beginnings of the Romantic Song Cycle and the Nature of an Art Form (Ph.D. diss., Columbia University, 1968), chapter 4. There are also discussions of the genre in Barbara Turchin, Robert Schumann d Song Cycled in the Context of the Early Nineteenth-Century Liecterkreid (Ph.D. diss., Columbia University, 1981), 12-18; and in Ruth Bingham, The Song Cycle in German-Speaking Countries 1790-1840: Approaches to a Changing Genre (Ph.D. diss., Cornell University, 1993), 69-85. Bingham also presents a translation of Reichardt's 1801 article as her appendix 1, 251—64. 17. See Turchin, Schumann's Song Cycles, 17—18. 18. For a discussion of Schumann's Liederdpielen, see John Daverio, Robert Schumann: HeraOofa "New Poetic Age" (New York: Oxford University Press, 1997), 402-4. 19. Allgemeine mudikalidche Zeitung 9 (1807): 219. This review is cited in Turchin, Schumann'*! Song Cycled, 16. 20. Peake, in fact, includes them as examples of Liederjpielen and suggests that they would have been performed as private theatrical productions. See Preliminary Inquiry, 68— 77. 21. Allgemeine mudikalidche Zeitung 17 (1815): 162. This review is cited both in Turchin, Schumann's Song Cycled, 22—24; and in Luise Eitel Peake, "The Antecedents of Beethoven's Liederkreis," Miuic and Letter 65 (1982): 248. 22. Turchin, Schumann d Song Cycled, 21. 23. See, for example, all three of the dissertations listed in note 16. 24. Mudikalitched Lexicon auf Grundlage i)ed Lexicon '<) von H. Ch. Koch, edited by Arrey von Dommer, 2d edition of//. Ch. Koch'd musikalidched Lexicon (Heidelberg: J. C. B. Mohr, 1865), 513. For the complete German text, see chapter 1, note 10. 25. Barbara Turchin, "The Nineteenth-Century Wanderlieder Cycle," Journal of Mudicology 5 (1987): 498-525. 26. There are numerous discussions of this Romantic theme. See, for example, Herbert Lmdenberger, "Theories of Romanticism: From a Theory of Genre to the Genre of Theory," in The Hid ton/ in Literature: On Value, Genre, Indtitutiond (New York: Columbia University Press, 1990), 61—84; M. H. Abrams, Natural SupematiiralLfm: Tradition and Revolution in Romantic Literature (New York: Norton, 1971), especially chapter 4; and Harold Bloom, "The Internalization of Quest-Romance," in Romanticism and Condcioudnedd: Edsayd in Criticism, edited by Bloom (New York: Norton, 1970), 3—23. Turchin discusses the relationship between this theme and the Wanderlieder cycle in "WancterLieHer Cycle," 499. 27. See Bingham, The Song Cycle, 42—49. This is, of course, one of the oldest means of formal organization, not only in music but in other arts as well. For a survey of its history in literature, see Claus-Michael Ort, "Zyklische Dichtung," in Reallexikon derdeutdcken Literaturgedchichte, vol. 5, edited by Werner Kohlschmidt and Wolfgang Mohr (Berlin: Walter de Gruyter, 1988), 1116. 28. Rosen, The Romantic Generation, 175. 29. Rufus Hallmark argues that the dramatic nature of Chamisso's poetry is one of the principal ways in which Frauenliebe und Leben differs from Dichtertiebe, since we "per-
256
Noted to Paged 204-11
form or hear the Heine songs as direct lyrical utterances of the singer, the Chamisso songs as role-playing." See "Robert Schumann: The Poet Sings," in German Lieder in the Nineteenth Century, edited by Hallmark (New York: Schirmer, 1996), 81. 30. Rufus Hallmark, The Genedid of Schumann d Dichterliebe: A Source Study (Ann Arbor: UMI, 1976), 115, 119, 117. 31. See Turchin, Schumann',* Song Cycled, 288-90; and Christopher Lewis, "Text, Time, and Tonic: Aspects of Patterning in the Romantic Cycle," Integral: The Journal of Applied MiuicaL Thought 2 (1988): 47-50. Other scholars who argue that Schumann created a narrative out of Heine's poems include David Neumeyer, in "Organic Structure and the Song Cycle: Another Look at Schumann's Dichterliebe," Mudic Theory Spectrum 4 (1982): 96; and Daverio, in Robert Schumann, 214. 32. Rosen, The Romantic Generation, 207-8. 33. S. S. Prawer, Heine: Bitch 3er Lieoer (London: Edward Arnold, 1960), 48-49. In fact, we find the same metaphor in the final poem of Schumann's Heine Liederkre'u, opus 24. 34. See Hallmark, Genedid of Dichterliebe, 120; Eric Sams, The Songd of Robert Schumann (London: Methuen, 1969), 124; Turchin, Schumann'd Song Cycles, 291, 338; Neumeyer, "Organic Structure," 97, 102; and Daverio, Robert Schumann, 216. 35. Rosen, The Romantic Generation, 210. 36. Theodor Adorno, "Zum Gedachtnis Eichendorffs," in Noted zur Literatur I (Frankfurt: Suhrkamp Verlag, 1958), 134-39. The reason Adorno uses such apparently inappropriate terms to describe the texts of these two songs —Ahnung arid Gegenwart—is because he is alluding to the title of Eichendorffs Roman, which was the original source for four of the poems that Schumann uses in his cycle (although neither "Schone Fremde" nor "Friihlingsnacht" is among them). Adorno is perhaps implying that the Roman provides a narrative framework for the cycle. Herwig Knaus makes a similar point more explicitly in a monograph published nearly thirty years later. He argues that "Ahnung und Gegenwart could scarcely have escaped Schumann's wide reading" and insists that "the assumption that Schumann would at least have been prompted to compose the cycle through this reading is corroborated by the placement of the poems within it—-with the exception of 'Waldesgesprach' they correspond to the unfolding plot of Ahnung und Gegenwart." But Knaus never explains how the cycle as a whole relates to the prose narrative and, as Turchin points out, provides no real evidence that Schumann was familiar with the book. See Knaus, Mudikdprache und Werkdtruktur in Robert Schumann 'd Liederkreid, vol. 27 of Schriften zur Miidik, edited by Walter Kolneder (Munich: Musikverlag Emil Katzbichler, 1974), 13—14; and Turchin, Schumann'd Song Cycled, 278. 37. Turchin, Schumann'd Song Cycled, 279, 313; Thym, Solo Song Settingd, 219—20; Daverio, Robert Schumann, 214—16. 38. Thym, Solo Song Settingd, 212—13. 39. McCreless, "Song Order," 23-25. 40. Sams, The Songd, 98. 41. Robert Schumann, Robert Schumann indeinen Schriften und Briefen, edited by Wolfgang Boetticher (Berlin: Hahnefeld, 1942), 340. "Der Eichendorffsche Zyklus ist wohl mein aller Romantischstes und es steht viel von Dir darin." See Sams, The Songd, 92; McCreless, "Song Order," 23; and Thym, Solo Song Settingd, 213. 42. Daverio, Robert Schumann, 194. 43. Ruth Solie, "Whose Life? The Gendered Self in Schumann's Frauenliebe Songs," in Mudic ant) Text: Critical Inquiried, edited by Steven Paul Scher (Cambridge: Cambridge University Press, 1992), 222, fn. 17.
Note,* to Pages 211-18
257
44. Sams, The Songj, 129. 45. Some of the most important essays written about Eichendorff during this period have been published in two collections: Paul Stocklein, ed., EichendorffHeute: Stimmen der Fonchung (Darmstadt: Wissenschaftliche Buchgesellschaft, 1966); and Alfred Riemen, ed., Atuichten zu Eichendorff: Beitrage der Forck-KritutcheAiMgabe, vol. 1, bk. 1, editred by Harry Frohlich and Ursula Regener (Stuttgart: Kohlhammer, 1993), 206. Translations from this work in this chapter are my own. Schumann changed "soil" in line 4 to "sollt'." 48. In the original poem, there is an additional strophe that provides a transition between these two: "Ich wunscht', es ware schon Morgen. / Da fliegen zwei Lerchen auf, / Die iiberfliegen einander, / Mein Herze folgt ihrem Lauf." Schumann omitted these lines when he set the poem. 49. This demonstrates one way in which the description of the cycle as a sequence of moods inevitably oversimplifies our understanding. If the significance of "In der Fremde" within the cycle is merely its mood of melancholy, while that of "Die Stille" is simply happiness, then we miss the fact that the two songs are related by their presentation of the same theme. 50. Eichendorff, Samtliche Werke, 76. In line 5, Schumann changed ''So" to "Es," and in line 8, "Kafigs" to "Kerkers." 51. In the third strophe of Eichendorff's poem, which Schumann omits in his setting, the narrator's desire for night to give way to morning becomes explicit. (See note 47.) 52. Eichendorff, Samtliche Werke, 11-12. 53. See Seidlm, Venuche liber Eichendorff , 244; and S. S. Prawer, German Lyric Poetry: A Critical Analyjui of Selected Poenu from Klop^tock to Rilke (London: Routledge and Kegan Paul, 1952), 132-33. 54. Eichendorff, Samtliche Werke, 11. 55. For a discussion of the symbolic significance of hunters in German Romantic poetry, see Susan Youens, Schubert, Milller, and Die schone Miillerin (Cambridge: Cambridge University Press, 1997), chapter 2. She discusses Eichendorff's poetry in particular on pages 64—65 and 72—76. 56. While the word Reh literally means "doe," it commonly refers to a young woman in German poetic usage. See Keith Spalding, An Historical Dictionary of German Figurative Wage (Oxford: Blackwell, 1959-), 1969. 57. Eichendorff, Samtliche Werke, 12. In Schumann's setting, the first line of this strophe reads: "Was heut' gehet mude unter." He also changed "bleibt" in the third line to "geht" and "bleib"' in the fourth line to "sei." 58. Seidlin, Venuchc ilber Eichendorff, 242-43.
258
Note* to Pages 218-23
59. Eichendorff, Sdmtiiche Werke, 11. Schumann changed "schauert" in line 8 to "schauert's." 60. Eichendorff, Sdmtiiche Werke, 366. Eichendorff's title is "Waldgesprach." Schumann also changed "es wird schon kalt" to "es ist schon kalt" in line 1 and line 15. 61. Eichendorff, Sdmtiiche Werke, 367. 62. Daverio, Robert Schumann, 215—16. See, also, his "The Song Cycle: Journeys through a Romantic Landscape," in German Lieder in the Nineteenth Century, edited by Rufus Hallmark (New York: Schirmer, 1996), 290-91. 63. For a general discussion of the distinction between these different uses of motive, see Leonard B. Meyer, Explaining MLMIC (Berkeley: University of California Press, 1973), 44ff. 64. McCreless, "Song Order," 14. 65. McCreless, "Song Order," 16-17. 66. Meyer makes a similar point in his criticism of Rudolph Reti's analysis of Brahms s Second Symphony in Explaining Mimic, 70-72. 67. McCreless, "Song Order," 15-16. 68. Rudolph Reti, in fact, identifies a very similar motive in "Von fremden Lander und Menschen," the first piece of Kinderjzenen, which he claims is the "thematic source" for the entire cycle. See The Thematic ProceM in Miuic (New York: Macmillan, 1951), 32-35.
WORKS CITED
Abraham, Gerald. "Robert Schumann." In The New Grove Early Romantic Masters 1. Edited by Stanley Sadie. New York: Norton, 1985. Abrams, M. H. Natural Supernaturalism: Tradition arid Revolution in Romantic Literature. New York: Norton, 1971. Adorno, Theodor. "Zum Gedachtnis Eichendorffs." \nNotenzurLiteraturI. Frankfurt: Suhrkamp Verlag, 1958. Aldwell, Edward, and Carl Schachter. Harmony and Voice Leading. Vol. 2. New York: Harcourt Brace Jovanovich, 1978. Allgenuine musikalische Zeitung. Leipzig, 1798-1882. Allgemeiner musikalischer Anzeiger. Vienna, 1829-41. Bingham, Ruth. The Song Cycle in German-Speaking Countries 1790-1840: Approaches to a Changing Genre. Ph.D. diss., Cornell University, 1993. Blackall, Eric A. The Novels of the German Romantics. Ithaca: Cornell University Press, 1983. Blanchot, Maurice. "The Athenaeum." Studies in Romanticism. 22 (1983): 163-72. Bloom, Harold. "The Internalization of Quest-Romance." In Romanticism and Consciousness: Essays in Criticism. Edited by Harold Bloom. New York: Norton, 1970. Brinkmann, Reinhold. Schumann und Eichendorff: Studien zum Liederkreis Opus 39. Vol. 95 of Miuik-Konzepte. Edited by Heinz-Klaus Metzger and Rainer Riehn. Munich, 1997. Burkhart, Charles. "Departures from the Norm in Two Songs from Schumann's Liederkreis." In Schenker Studies. Edited by Hedi Siegel. Cambridge: Cambridge University Press, 1990. Cone, Edward T. Musical Form and Musical Performance. New York: Norton, 1968. Dahlhaus, Carl. Esthetics of Music. Translated by William W. Austin. Cambridge: Cambridge University Press, 1982. Daverio, John. "Schumann's 'Im Legendenton' and Friedrich Schlegel's Arabeske." 19thcentury Musk 11 (1987): 150-163. . Nineteenth-Century Music and the German Romantic Ideolagy. New York: Schirmer, 1993. . "The Song Cycle: Journeys through a Romantic Landscape." In German Lieder in the Nineteenth Century. Edited by Rufus Hallmark. New York: Schirmer, 1996. . Robert Schumann: Herald of a "New Poetic Age." New York: Oxford University Press, 1997. Eichendorff, Joseph Freiherr von. Sdmtliche Werke des Freiherrn Joseph von Eichendorff: Historisch-KritischeAusgabe. Vol. 1, bk. 1. Edited by Harry Frohlich and Ursula Regener. Stuttgart: Kohlhammer, 1993. 259
260
Work* Cited
Eichner, Hans. "Friedrich Schlegel's Theory of Romantic Poetry." PMLA 71 (1956): 1018-41. . "Germany/Romantisch—Romantik—Romantiker." In "Romantic" and ltd Cognates: The European HLftory of a Won). Edited by Hans Eichner. Toronto: University of Toronto Press, 1972. Eismann, Georg. "Nachweis der internationalen Standorte von Notenautographen Robert Schumanns." In Sammelbande der Robert-Schumann-GeseUschaft. Vol. 2. Leipzig: VEB deutsche Verlag fur Musik, 1966. Feldman, Fritz. "Zur Frage des 'Liederjahres' bei Robert Schumann." Archiv fiir Miuikwiueruchaft 9 (1952): 246-69. Finson, Jon W. "Schumann's Mature Style and the 'Album of Songs for the Young.' "JournaLerf'Musicoiogy 8 (1990): 227-50. . "The Intentional Tourist: Romantic Irony in the Eichendorff LuderkreLi of Robert Schumann." In Schumann and Hi) World. Edited by R. Larry Todd. Princeton: Princeton University Press, 1994. Forte, Allen. "Schenker's Conception of Musical Structure." Journal of Music Theory 3 (1959): 1—30. Reprinted in Readings in Schenker Analysis and Other Approaches, Edited by Maury Yeston. New Haven: Yale University Press, 1977. Genette, Gerard. Narratice Discourse: An Essay in Method. Translated by Jane E. Lewin. Ithaca: Cornell University Press, 1980. German Life and Letter 42/3 (1989). Gerstmeier, August. Die Lieder Schumann,). Tutzing: Hans Schneider, 1982. Goebel, Robert O. Eickendorff's Scholarly Reception: A Survey. Columbia, SC: Camden House, 1993. Goethe, Johann Wolfgang von. Goethes Werke Hamburger Ausgabe. Vol. 6. Edited by Erich Trunz. Hamburg: Christian Wegner Verlag, 1958. Hallmark, Rufus. The Genesis of Schumann's Dichterliebe:/! Source Study. Ann Arbor: UMI, 1976. . "Die handschriftlichen Quellen der Lieder Robert Schumanns." In Robert Schumann—ein romantisches Erbe in neuer Forschung, edited by the Robert Schumann Gesellschaft, Dusseldorf (Mainz: Schott, 1984). . "The Riickert Lieder of Robert and Clara Schumann." 19th-century Music 14
(1990): 3-30. . "Robert Schumann: The Poet Sings." In German Lieder in the Nineteenth Century. Edited by Rufus Hallmark. New York: Schirmer, 1996. . "Schumann's Revisions of'In der Fremde,' Op. 39, No. 1."Presented at the Conference of Nineteenth-Century Music, Nottingham, 1978. Hanslick, Eduard. Auo dem Concertsaat. Vienna: Wilhelm Braumiiller, 1870. Hindenlang, Karen A. "Eichendorffs AufeinerBurg and Schumann's Liederkreis, Opus 39." Journal of Musicoiogy 8 (1990): 569-87. Janowitz, Anne. "Coleridge's 1816 Volume: Fragment as Rubric." Studied in Romanticism 24 (1985): 21-39. Kaminsky, Peter. "Principles of Formal Structure in Schumann's Early Piano Cycles." MIMIC Theory Spectrum 11 (1989): 207-25. Keiler, Allan. "The Syntax of Prolongation, Part I." In Theory Only 3 (1977): 3-27. . "On Some Properties of Schenker's Pitch Derivations." Music Perception 1 (1984): 200-228. Kerman, Joseph. "How We Got into Analysis, and How to Get Out." Critical Inquiry 6 (1980): 311-31.
Work* Cited
261
Knaus, Herwig. MiutiLtprache and Werkdtruktur in Robert Schumann '<> Liederkreif. Vol. 27 of Schriften zur Musik. Edited by Walter Kolneder. Munich: Musikverlag Emil Katzbichler, 1974. Komar, Arthur, editor. Schumann: Dichterliebe. New York: Norton, 1971. Kramer, Richard. Distant Cycle,): Schubert and the Conceiving of Song. Chicago: University of Chicago Press, 1994. . "The Hedgehog: Of Fragments Finished and Unfinished." 19th-century MIMIC (21) 1997: 134-48. Kravitt, Edward F. "The Lied in 19th-century Concert Life." JAMS 18 (1965): 207-18. Lacoue-Labarthe, Philippe, and Jean-Luc Nancy. The Literary Absolute: The Theory of Literature in German Romanticism. Translated by Philip Barnard and Cheryl Lester. Albany: State University of New York Press, 1988. Lester, Joel. "Robert Schumann and Sonata Forms." 19th-century MIMIC 18 (1995): 189— 210. Lewis, Christopher. "Text, Time, and Tonic: Aspects of Patterning in the Romantic Cycle." Integral; The Journal of Applied Musical Thought 2 (1988): 37-73. Lindenberger, Herbert. The History in Literature: On Value, Genre, Institutions. New York: Columbia University Press, 1990. Lippman, Edward A. "Theory and Practice in Schumann's Aesthetics." JAMS 17 (1964): 310-45. Marston, Nicholas. Schumann: Fantasie, Op. 17. Cambridge: Cambridge University Press, 1992. McCreless, Patrick. "Song Order in the Song Cycle: Schumann's Liederkreui, Op. 39." Music Analysis 5 (1986): 5-28. McFarland, Thomas. Romanticism and the Fornu of Ruin. Princeton: Princeton University Press, 1981. McGann, Jerome J. The Romantic Geology. Chicago: University of Chicago Press, 1983. Mendelssohn-Bartholdy, Felix. Briefe von Felix MendeLisohn-Bartholdy an Ignaz und Charlotte Moscheles. Edited by Felix Moscheles. Leipzig, 1888. Meyer, Leonard B. Explaining Music. Berkeley: University of California Press, 1973. . Style and Music. Philadelphia: University of Pennsylvania Press, 1989. Musikalisches Lexicon auf Grundlage 'd&a Lexicon d von H. Ch. Koch. Edited by Arrey von Dommer. Second edition ofH. Ch. Koch's musikalischeJ Lexicon. Heidelberg: J. C. B. Mohr, 1865. Muxfeldt, Kristina. Schubert Song Studied. Ph.D. diss., State University of New York at Stony Brook, 1991. Neue Zeitscbrift fur Musik. Leipzig, 1834—. Neumeyer, David. "Organic Structure and the Song Cycle: Another Look at Schumann's Dichterliebe" MIMIC Theory Spectrum A (1982): 92-105. Newcomb, Anthony. "Schumann and the Marketplace: From Butterflies to Hausmusik." In Nineteenth-Century Piano Music. Edited by R. Larry Todd. New York: Schirmer, 1990. Ort, Claus-Michael. "Zyklische Dichtung." In Reallexikon der deutscben Literaturgeschichte. Vol. 5. Edited by Werner Kohlschmidt and Wolfgang Mohr. Berlin: Walter de Gruyter, 1988. Peake, Luise Eitel. The Song Cycle: A Preliminary Inquiry into the Beginnings of the Romantic Song Cycle and the Nature of an Art Form. Ph.D. diss., Columbia University, 1968. . "Kreutzer's Wanderlieder. The Other Winterreise." Musical Quarterly 45 (1979): 83102.
262
Works Cited
. "The Antecedents of Beethoven's Liederkreis." Music and Letters 63 (1982): 24260. Peckham, Morse. The Triumph of Romanticism. Columbia: University of South Carolina Press, 1970. . The Romantic Virtuoso. Hanover: Wesleyan University Press, 1995. Plantinga, Leon B. Schumann as Critic. New Haven: Yale University Press, 1967. Prawer, S. S. German Lyric Poetry: A Critical Analysis of Selected Poems from Klopstock to Rilke. London: Routledge and Kegan Paul, 1952. . Heine: Buch der Lieder. London: Edward Arnold, 1960. Radner, Lawrence. Eichendorff: The Spiritual Geometer. Lafayette, IN: Purdue University Studies, 1970. Redfield, Marc. "Romanticism, Bildung, and the Literary Absolute." In Lessons of'Romanticism: A Critical Companion. Edited by Thomas Pfau and Robert F. Gleckner. Durham: Duke University Press, 1998. Reich, Nancy B. Clara Schumann: The Artist ant) the Woman. Ithaca: Cornell University Press, 1985. Reti, Rudolph. The Thematic Process in Music. New York: Macmillan, 1951. Riemen, Alfred, editor. Ansichten zu Eichendorff: Beitrdge der Forschung 1958 tis 1988. Sigmaringen: Jan Thorbecke, 1988. Roesner, Linda Correll. "The Sources for Schumann's Davidsbiindlertanze, Op. 6: Composition, Textual Problems, and the Role of the Composer as Editor." \n Mendelssohn and Schumann: Essays on their Music and Its Context. Edited by Jon Finson and R. Larry Todd. Durham: Duke University Press, 1984. ."Schumann's 'Parallel' Forms." 19th-century Music 14 (1991): 265-78. Rosen, Charles. Sonata Forms, rev. ed. New York: Norton, 1988. . The Romantic Generation. Cambridge: Harvard University Press, 1995. Rothstein, William. Phrase Rhythm in Tonal Music. New York: Schirmer, 1989. . "Rhythmic Displacement and Rhythmic Normalization." In Trends in Schenkerian Research. Edited by Allen Cadwallader. New York: Schirmer, 1990. Salzer, Felix. Structural Hearing: Tonal Coherence in Music. New York: Charles Boni, 1952. Sams, Eric. The Songs of Robert Schumann. London: Methuen, 1969. Satyanedra, Ramon. "Liszt's Open Structures and the Romantic Fragment."Music Theory Spectrum 19 (1997): 184-205. Schenker, Hemrich. Harmony. Translated by Elisabeth Mann Borgese. Edited by Oswald Jonas. Chicago: University of Chicago Press, 1954. Translation oiHarmonie. Vienna: Universal Edition, 1906. . Derfreie Satz, 2d ed. Edited by Oswald Jonas. Vienna: Universal Edition, 1956. Schlegel, August Wilhelm. "Uber Zeichnungen zu Gedichten." In Athenaeum: Eine Zeitschrift von August Wilhelm Schlegel un3 Friedrich Schlegel. Vol. 2. Edited by Curt Griltzmacher. Munich: Rowohlt, 1969. Schlegel, Friedrich. Literary Notebooks: 1797-1801. Edited by Hans Eichner. Toronto: University of Toronto Press, 1957. . Kritische Friedrich-Schlegel-Ausgabe. Vol. 2. Edited by Hans Eichner. Zurich: Thomas-Verlag, 1967. . Philosophical Fragments. Translated by Peter Firchow, foreword by Rodolphe Gasche. Minneapolis: University of Minnesota Press, 1991. Schumann, Robert. Jugendbriefe fan Robert Schumann. Edited by Clara Schumann. Leipzig: Breitkopf und Hartel, 1885.
Work* Cited
263
. Robert Schumann '<> Leben CUM
525. Wolff, Viktor Ernst. Robert Schumann^ Licder in er<>ten und tipateren Fajjungen. Leipzig: H. S. Hermann, 1914. Youens, Susan. Schubert, Mailer, and Die schone Miillerin. Cambridge: Cambridge University Press, 1997.
This page intentionally left blank
INDEX
Adorno, Theodor, 11, 208-209 Allgemeine mujikatLfche Zeitung (Leipzig),
Eichendorff, Joseph Freiherr von collected edition, 4, 146, 189
"Friihlingsnacht," 129-131, 215-216
80, 81-82, 200
"ImWalde,"218 "In der Fremde," 95-96, 213 "Intermezzo," 131-132, 214-215 "Mondnacht," 144-147, 156, 215
Allgemeiner miuikalucher Anzeiger, 84 analysis motivic, 12, 18, 222 of text and music, 55—56 voice-leading, 12, 21-22, 31-35, 37, 38-39, 42, 44
poetic style, 93, 212, 226 "Schone Fremde," 144, 146-147, 215
"Die Stille," 212-215 Banck, Carl, 195-196, 198-199 Beethoven, Ludwig van, 60, 75 AndieferneGelicbte, 10-11, 13, 14, 19-20, 75, 81-82 Bingham, Ruth, 13, 14 Blanchot, Maurice, 63 character piece, 71, 73-74, 75, 77 classical style, 34, 39, 66, 73-74, 75-76 closure, 58, 91-93, 106-107, 121-122 Cone, Edward T., 121 cycle, Romantic, 17-19, 65, 68-69, 7577 (fiee aLo song cycle) Daveno, John definition of cycle, 13-14, 15-18 on the Eichendorff Liederkreii, 209, 221-222, 223, 224 on nineteenth-century song criticism,
77-78 on Schumann's life, 210-211 on Schumann's review of Lowe's Ed ther, 15-16 Dommer, Arrey von, 8-10, 201-202
"Waldgesprach," 218-220 "Wehmuth," 214-215 "Zwielicht," 216-218 Eichner, Hans, 70 Finson, Jon W., 94 Forte, Allen, 38-39, 40 fragment, 17-19, 62-66, 69, 106-107 Genette, Gerard, 3 genre, 3, 5, 7-9, 19, 69-77, 227 Hallmark, Rulus onDichterliebe, 176,204 "In der Fremde" piano draft, transcription of, 110, 111, 112, 114-115 on the Riickert Lieder, 178 on Schumann's song cycles, 172 — 173 on Schumann's song manuscripts, 175176 Heine, Hemrich Buck der Lieder, 207 Lyri)che<> Intermezzo, 56—57, 65, 176 — 177, 204
265
266
Index
Kaminsky, Peter, 12, 30 Keiler, Allan, 44 Kerman, Joseph, 39, 40, 41, 42,
key
47,48
relationships in the sonata cycle, 75-76 relationships in the song cycle, 28-30, 76, 78, 82-85, 87 unity, 27, 30-32, 49 Knaus, Herwig, 110, 185-186 Komar, Arthur analysis of Dichterliebe, 12, 26-38, 40, 52-53, 57 definition of cycle, 11, 15, 26-28, 91-92 and Schenker's theory, 27, 31-35, 37, 38, 52, 57 on song texts, 49-50 Kreutzer, Conradin, 13 Lacoue-Labarthe, Philippe, and Jean-Luc Nancy, 64 Lester, Joel, 75 Lewis, Christopher, 204-206 lied. See song Liedercycliu, Liederkreif. See song cycle Liederroman, 201 Ludenpiel, 200-201 Lorenz, Oswald, 78-79, 80, 83, 197 manuscript sources Berlin notebooks, 21, 173-175, 178, 188 for Dichtertube, 176-177, 183 for the Eichendorff Liederkreui, 4, 109, 171-172, 180, 185189, 192 for "Es leuchtet meine Liebe," 29 for Frauent'Lebe ant)Leben, 5, 176 for "In der Fremde," 95, 109110, 111-112, 114-115, 116-118 (orMyrtt>en,29, 179-182 piano drafts, 175
for the Riickert Lieder, 178-179 for "Schone Fremde," 224 Stichvorlagen, 174, 175 vocal sketches, 175 McCreless, Patrick definition of cycle, 12, 171 on Dichterliebe, 177 on the Eichendorff Liederkreu, 91, 185-186, 188, 198, 209-210, 222-223 on Komar, 30 McFarland, Thomas, 59 Mendelssohn-Bartholdy, Felix, 84 85 Meyer, Leonard B., 59-60, 61, 116, 122, 125 Mozart, Wolfgang Amadeus, 75 — 76 Muxfeldt, Kristina, 67, 68-69 Neue Zeittickrift fiir Miufik reviews (ttee aLo Schumann, Robert: reviews) of Ferdinand Miller's Riickert Lieder (by Oswald Lorenz), 78-79, 80, 197 of Kulenkamp's Sechj deuttche Grange, 197 "Liedercompositionen" (by Carl Banck), 195-196, 198-199 of Carl Lowe's Gregor auf dem Stem (by Oswald Lorenz), 80 of Leon de St. Lubin's Secbj Lieder, 199 of Clara Schumann's Sechti Lieder, 8 of Robert Schumann's Dichterliebe, 51, 86-87, 92 of Hieronymous Truhn's Bin Liebedroman, 8, 197 of Georg Vierling's Cyclic arakifcher Dichtung, 80 song criticism in, 14-15, 78, 85, 195, 196-198 Neumeyer, David, 12, 30, 48-51,
53-56, 58, 91, 143-144
Index open ending, 93, 106-107, 108109, 122, 128-129, 140, 156 orgamcism in the cycle, 3-4, 12-15, 18, 21-22, 25-27, 38, 49, 50, 55, 57-58, 59, 60-61, 86, 88,91 contradictions of, 59—61, 63 dynamic, 61 formal, 61, 64-65, 88, 91 as poetic metaphor, 57 in Schenker's theory, 27, 32—33, 34-35,38,47-48,49,50 in tonal music, 27, 31, 49, 60 pairing Baroque origin of, 50—51 Lorenz on, 80 recompositional, 141-144, 192 transitional, 54-55, 143-144 Peckham, Morse, 60, 61, 65 Phrygian cadence, 51, 107 poetic themes deception, 216 homelessness, 94, 202 ineffability, 93, 132, 214-215 love, 132, 202, 209-210, 212213, 216 nature, union with, 94, 144, 202-203, 212, 216, 224-
225 quest, spiritual, 202, 220, 225 transcendence, 122, 129, 139, 144, 212, 216, 220, 226 Prawer, S. S., 207 Reichardt, Johann Friedrich, 200 Reti, Rudolph, 11, 142 Roesner, Linda Correll, 72, 75 Roman, 67, 69, 70-71 Romanticism versus Classicism, 6, 62-63, 66, 69-70, 73-74, 85-86 and genre, 5, 69-71, 73-74 influence on Schumann, 6—8, 11-12, 16-19,58,61-62, 226-227
267
musical, 60-61, 66, 73-74, 77, 121-122 and orgamcism, 59—61 Rosen, Charles on Chopin's Preludes, 18 — 19 definition of cycle, 19-20, 198 on DicblerLiebe, 18, 20, 87, 206, 207-208 on the fragment, 17, 18-19, 63 on Frauenlicbe and Leben, 203 on genre in Romantic music, 74 on "Mondnacht," 153 on Schumann's view of the sonata, 72—73 Sams, Eric, 210, 211 Schenker, Heinrich analysis of "Aus meinen Thranen spriessen," 38—48, 57-58, 87 analysis of "Im wunderschonen Monat Mai," 51-52, 87 analytic notation, 32—33, 40, 42-43, 45-47, 52, 54 critiques of, 22, 39, 47-48 on Handel's Suite #2 in F major, 50-51 interruption technique, 33—34, 39,43 and organicism, 27, 32—33, 34 — 35, 38, 47-48, 49, 50 passing tone, 33—34 structural levels, 35, 37, 46-47 theory, 22, 27, 31, 39, 42, 4344,58 Unatz, 32-33, 39, 47, 49 Schlegel, August Wilhelm, 6, 61 on the cycle, 68—69 on the sketch, 22-23, 66-68 Schlegel, Friedrich, 6, 61, 85 Athenaeum Fragmente, 62-63, 64, 70 "Brief iiber den Roman," 71 on the fragment, 16-17, 62-64 on genre, 69—71 Keen, 63, 69 Kritutche Fragmente, 63, 65, 70 literary notebooks, 70
268
Index
Schubert, Franz, 8, 20 Schumann, Clara Wieck, 6, 7, 210-211 and the Eichendorff LiederkreLi, 4, 94, 110, 171, 188-189, 210 and the Riickert Lieder, 178 Schumann, Robert "Die alten bosen Lieder" (opus 48, #16), 207-208 "Aus meinen Thranen spriessen" (opus 48, #2), 18, 39, 4042,43,44-45,54-57, 143 Baroque techniques, use of, 50— 51, 107, 142 as critic, 6, 14-15, 16, 59, 60, 71-74, 85, 195 cycles coherence in, 5—6, 17—20, 23-24, 166-167, 192193, 225-227 composition of, 5, 21, 23—24, 166-167, 171-175, 181184, 192-193, 226 formal organization of, 58, 65, 75-77, 88, 225-226 reception of, 4, 6-8, 11-12, 19, 21, 58, 75-77, 88, 227 text of, 108-109, 167 DichterLiebe (opus 48) analysis of, 12, 21-22, 2638, 48, 57-58, 91-92 coherence in, 23, 166—167 composition of, 176—178, 183 formal organization of, 65, 76 as paradigmatic song cycle, 4, 11,20,21, 172, 198,203 reception of, 25-26, 86-88, 91-92, 203-208 review of, 51, 86-87, 92 text of, 204-208 Eichendorff LiederkreLi (opus 39) coherence in, 23, 92-93, 121122, 139-140, 141, 165167, 202-203, 208-209, 214, 220-221, 223-226 composition of, 4, 92-93, 171-172, 177, 180, 185192, 209-210, 220
order of songs in, 92, 185— 188, 191-192, 209-210 reception of, 3-4, 21, 198, 208-210 text of, 21, 93, 94, 119, 147, 171, 202-203, 208-221, 225, 226 Etudes jymphoniquzti (opus 13), 142 Fun/Lieder und Grange (opus 127), 172 Frauentiebe and Leben (opus 42) composition of, 174, 210-211 as paradigmatic song cycle, 4, 20, 172, 203 reception of, 211 "Der frohe Wandersmann" (opus 77, #1), 93-94 "Friihlingsnacht" (opus 39, #12) analysis, 123-129, 223-224 form of, 123 text of, 119, 122-123, 129131, 139-140,215-216 Heine LiederkreLi (opus 24), 172, 174 review of, 84 "Ich will meine Seele tauchen" (opus 48, #5), 92 Impromptus (opus 5), 142 "Im Walde" (opus 39, #11), 218 "Im wunderschonen Monat Mai" (opus 48, #1), 18, 51, 55-57, 92, 107, 143 "In der Fremde" (opus 39, #1) analysis, 97-108, 111-118, 224-225 composition of, 95, 109-119 form of, 98, 106, 111, 118-119 as opening of the Eichendorff Liederkreu, 93-94 text of, 95, 96-98, 101, 104, 105-106, 111, 114, 116, 117-119,213 "Intermezzo" (opus 39, #2) analysis, 133—139 composition of, 189—192 form of, 123, 133, 190-191 text of, 119, 122-123, 131133, 139-140, 214-215
Index Liederjahr, 4, 21, 172, 173 LieJerdpielen, 200 "Mondnacht" (opus 39, #5) analysis, 148-158 composition of, 189-192 form of, 141-142, 153, 164165, 190-191 text of, 119, 144-147, 149, 156-157, 164-165, 215 Myrtken. (opus 25), 173, 179-182 NoveUetten (opus 21), 7, 18, 76 piano music, 5—7, 142 Reinick Gedicbte (opus 36), 172 reviews of Norbert Burgmtiller, Heinrich Esser, and W. H. Veit's lieder, 195 of Chopin's Nocturnes, Ballade, and Waltzes, 74 of Chopin's Preluded, 65—66 of Stephen Heller's EtuSed, 66, 73, 74 of Bernhard Klein's lieder, 199 of Carl Lowe's Either, 15 — 16, 83, 197 of Mendelssohn and Schubert's piano sonatas, 72-73 of Henry Hugh Pearson's Burns Gedichte, 197, 199 of piano sonatas, 71—72 of piano variations, 142 "Die Rose, die Lilie" (opus 48,
#3), 37 and Romanticism, 6—8, 11 — 12, 16-19, 58, 61-62, 226227 Riickert Lieder (opus 37), 172, 178-179, 183 "Schone Fremde" (opus 39, #6) analysis, 157-164, 223-225 composition of, 189-192, 224-225 form of, 141-142, 164-165,
190-191 text of, 144, 146-147, 162,
164-165, 215 sonata form movements, 74 —75
269
"Die Stille" (opus 39, #4), 212215 Vier GedHnge (opus 142), 172 "Waldesgesprach" (opus 39,
#3), 218-220 "Wehmuth" (opus 39, #9), 214215 "Zwielicht" (opus 39, #10), 216-218 sketch {dee a/dff manuscript sources, vocal sketches) as metaphor for poem, 132 as metaphor for Romantic art, 22-23, 66-67 as metaphor for song, 36, 79, 80, 92-93, 132, 189, 226 Smith, Barbara Herrnstein, 121, 12^ Solie, Ruth, 211 sonata form, 5, 34, 71-74 song, 7, 38, 67, 75, 77 song criticism, nineteenth-century, 7,8, 11,77-78,87 song cycle coherence in, 3-7, 13-20, 23-
30, 78-80, 85-88, 91-92, 165-167, 193, 197-198, 200, 201-202 versus collection, 5—6, 12—13,
19, 22, 27-28, 29, 77-78, 80-81, 82, 85, 86, 171-
173, 174-175, 184 definition of, 3-24, 25-26, 75-78, 201-202 designations of, 13, 16, 80-81 history of, 11, 12-16 order of songs in, 23-24, 2730, 171-172, 185 key relationships in, 27—30, 76,
78, 82-85, 87 text of, 21, 197-198, 201-202 text and music, relationship between, 49-50, 55-56 Thym, Jurgen, 198, 209, 212 Turchin, Barbara, 30, 76, 201 definition of cycle, 12—13, 15 on Dichterliebe, 204-206 on the Eichendorff Liederkreif, 94, 142-144, 157, 209
270
Index
Turchin, Barbara (continued) on "In der Fremde" piano draft, 110, 116-117 on nineteenth-century song criticism, 13, 14-15, 7778, 83, 84 on Schumann's review of Lowe's Either, 15-16, 83
Wan^erltedef cycle, 21, 94, 202-203 weak opening, 23, 36, 37, 93, 121123, 129, 139-140, 141, 152-153, 189-192 Wieck, Clara. See Schumann, Clara Wieck Wieck, Friedrich, 210-211 Wolff, Viktor Ernst, 110