INFORMATION TO USERS The most advanced technology has been used to photo­ graph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the au thor did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are re­ produced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. These are also available as one exposure on a standard 35mm slide or as a 17" x 23" black and white photographic print for an additional charge.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

UMÎ University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9014429

Song composers and their poetry choices: An analysis of the literary background and textual selections of twelve composers

Gregg, Thomas Andrew, D.M.A.

The Ohio State University, 1989

Copyright ©1989 by Gregg, Thomas Andrew. All rights reserved.

UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106

SONG COMPOSERS AND THEIR POETRY CHOICES;

AN ANALYSIS OF THE LITERARY BACKGROUND AND

TEXTUAL SELECTIONS OF TWELVE COMPOSERS

D.M.A. DOCUMENT

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Musical Arts in the Graduate

School of the Ohio State University

By

Thomas Andrew Gregg, B.M., M.M.

*****

The Ohio State University

1989

Document Committee: Approved by

H. Swank

E. Davis Adviser M. Barnes School of Music © 1989 Thomas A. Gregg To My Family, Mentors, Colleagues and Friends

11 ACKNOWLEDGEMENTS

I express sincere appreciation to both Professor Helen Swank and Professor Eileen

Davis for their guidance throughout my course of study, including work on this docu­ ment. Thanks also go to the other committee members. Dr. Marshall Barnes and Dr.

David Butler for their time and support. The assistance of Dr. Franz Keller of the French

Department at The Ohio State University, Betsy Baumeister and Lois Sims of The Ohio

State University Libraries, and Dr. James Porter of the Mathematics Department at the

University of Mississippi is also gratefully acknowledged. To the others who have provided help and emotional support, I offer many thanks: Carol Rausch, Cindy Cole­ man, Stephanie Carter, and especially Paul Gunzelmann.

Ill VITA

March 19,1958 ...... Bom — Marysville, Ohio

1980 ...... B.M., Capital University, Columbus, Ohio

1982 ...... M.M., University of Michigan, Ann Arbor

1983-1985 ...... Private studio voice teacher, free-lance singer, Washington, D C. area; Admin. Specialist, Arlington, Virginia Health Department, Public Library

1986-1988 ...... Graduate Teaching Associate in Studio Voice, The Ohio State University, Columbus, Ohio

1988-Present ...... Instructor (1988-89), Assistant Professor (1989-Present), Department of Music, the University of Mississippi, Oxford, Missis­ sippi

FIELDS OF STUDY

Major Field: Music

Studies in Voice, Vocal Literature, Vocal Pedagogy, Music Theory, Music Perception, and Composition.

IV TABLE OF CONTENTS

ACKNOWLEDGEMENTS...... iii

VITA...... iv

TABLE OF CONTENTS...... v

LIST OF TABLES...... vii

LIST OF FIGURES...... viii

RECITAL PROGRAMS...... xi

INTRODUCTION...... 1

CHAPTER PAGE I. TEXTUAL SOURCES FOR THE LIED COMPOSERS Introduction to Chapter I ...... 8 Franz Schubert (I797-I828) ...... 10 Robert Schumann (1810-1856) ...... 25 Johannes Brahms (1833-1897) ...... 39 Hugo Wolf (1860-1903) ...... 56 Summary for the Lied Composers ...... 71

n. TEXTUAL SOURCES FOR THE MELODIE COMPOSERS Introduction to Chapter I I ...... 84 Henri Duparc (1848-1933) ...... 86 Gabriel Fauré (1845-1924) ...... 95 Claude Debussy (1862-1918) ...... 107 Francis Poulenc (1899-1960) ...... 121 Summary for the Mélodie Composers ...... 133 m . TEXTUAL SOURCES FOR THE AMERICAN ART SONG COMPOSERS Introduction to Chapter III ...... 144 Châties Ives (1874-1954) ...... 147 Samuel Barber (1910-1981) ...... 167 Ned Rorem (b. 1923) ...... 184 Dominick Argento (b. 1927) ...... 197 Summary for the American Art Song Composers ...... 206

FINAL SUMMARY: THE TWELVE COMPOSERS...... 216

APPENDIX: TABLES 13-34...... 231

LIST OF REFERENCES...... 254

VI LIST OF TABLES

TABLE PAGE

1. The Songs of Franz Schubert ...... 17 2. The Songs of Robert Schumann ...... 30 3. The Songs of Johannes Brahms ...... 45 4. The Songs of Hugo W olf ...... 62 5. The Songs of Henri Duparc ...... 89 6. The Songs of Gabriel Fauré ...... 100 7. The Songs of Claude Debussy ...... 113 8. The Songs of Francis Poulenc ...... 126 9. The Songs of Charles Ives ...... 153 10. The Songs of Samuel Barber ...... 173 11. The Songs of Ned Rorem ...... 189 12. The Songs of Dominick Argento ...... 201 13. Textual Sources for the Songs of Franz Schubert (1797-1828) ...... 232 14. Textual Sources for the Songs of Robert Schumann (1810-1856) ...... 236 15. Textual Sources for the Songs of Johannes Brahms (1833-1897) ...... 238 16. Textual Sources for the Songs of Hugo Wolf (1860-1903) ...... 240 17. Textual Sources for the Songs of Henri Duparc (1848-1933) ...... 241 18. Textual Sources for the Songs of Gabriel Fauré (1845-1924) ...... 242 19. Textual Sources for the Songs of Claude Debussy (1862-1918) ...... 243 20. Textual Sources for the Songs of Francis Poulenc (1899-1963) ...... 244 21. Textual Sources for the Songs of Charles Ives (1874-1954) ...... 245 22. Textual Sources for the Songs of Samuel Barber (1910-1981) ...... 248 23. Textual Sources for the Songs of Ned Rorem (b. 1923) ...... 250 24. Textual Sources for the Songs of Dominick Argento (b. 1927) ...... 253

Vll LIST OF FIGURES

FIGURE PAGE

LIED COMPOSERS

1. Number of Songs ...... 73

2. Number of Sources ...... 74

3. Arithmetic Average Number of Songs Per Source ...... 75

4. Median Average and Modal Value of Songs Per Source ...... 76

5. Percentage of Texts from Authors Bom 16-100 Years Earlier than Composer ...... 77

6. Percentage of Texts from Authors Bom Within 15 Years (Plus or Minus) of Composer ...... 78

7. Percentage of Texts from Authors Bom More than 15 Years After Composer ...... 79

8. Percentage of Texts from Undated Sources ...... 80

9. Percentage of Texts from Authors Bom More than 100 Years Before Composer ...... 80

10. Percentage of Texts in Non-Native Languages ...... 81

11. Approximate Number of Authors with whom the Composer had Personal Contact ...... 83

MELODIE COMPOSERS

12. Number of Songs ...... 135

V lll 13. Number of Sources ...... 136

14. Arithmetic Average Number of Songs Per Source ...... 136

15. Median Average and Modal Value of Songs Per Source ...... 137

16. Percentage of Texts from Authors Bom 16-100 Years Earlier than Composer ...... 138

17. Percentage of Texts from Authors Bom Within 15 Years (Plus or Minus) of Composer ...... 138

18. Percentage of Texts from Authors Bom More than 15 Years After Composer ...... 139

19. Percentage of Texts from Undated Sources ...... 140

20. Percentage of Texts from Authors Bom More than 100 Years Before Composer ...... 141

21. Percentage of Texts in Non-Native I^guages ...... 142

22. Approximate Number of Authors with whom the Composer had Personal Contact ...... 143

AMERICAN ART SONG COMPOSERS

23. Number of Songs ...... 207

24. Number of Sources ...... 208

25. Arithmetic Average Number of Songs Per Source ...... 208

26. Median Average and Modal Value of Songs Per Source ...... 209

27. Percentage of Texts from Authors Bom 16-100 Years Earlier than Composer ...... 210

28. Percentage of Texts from Authors Bom Within 15 Years (Plus or Minus) of Composer ...... 211

29. Percentage of Texts from Authors Bom More than 15 Years After Composer ...... 211

ix 30. Percentage of Texts from Undated Sources ...... 212

31. Percentage of Texts from Authors Bom More than 100 Years Before Composer ...... 213

32. Percentage of Texts in Non-Native Languages ...... 214

33. Approximate Number of Authors with whom the Composer had Personal Contact ...... 215

THE TWELVE COMPOSERS

34. Number of Songs ...... 218

35. Number of Sources ...... 219

36. Arithmetic Average Number of Songs Per Source ...... 220

37. Median Average of Songs Per Source ...... 221

3 8, Percentage of T exts from Authors Bora 16-100 Y ears Earlier than Composer ...... 221

39. Percentage of Texts from Authors Bora Within 15 Years (Plus or Minus) of Composer ...... 224

40. Percentage of Texts from Authors Bora More than 15 Years After Composer ...... 225

41. Percentage of Texts from Undated Sources ...... 226

42. Percentage of Texts from Authors Bora More than 100 Years Before Composer ...... 227

43. Percentage of Texts in Non-Native Languages ...... 228

44. Approximate Number of Authors with whom the Composer had Personal Contact ...... 229 SfedlaLI

Thomas A. Grsgg, Tenor Nelson Harper, Piano

Tuesday, March 10,1987 at 8:00 p.m. Hughes Hall Auditorium, The Ohio State University Columbus, Ohio

Italian Songs (17th & 18th centuries) ed. by Luigi Dallapiccola Vittoria, mio cuore Giacomo Carissimi Se/ve amiche Antonio Caldara Già il sole dal Gange Alessandro Scarlatti

Rêves (1942) Darius Milhaud Marronniers Toi Confidence Le Mistral “Long Distance” Jeunesse

An Invisible Shore: Christopher Columbus Songs (1986) Thomas A. Gregg Provisions Where the Spices Grow Land is sighted Columbus: Bahia de Acul

Intermission

Letters from St. Paul (1970) Daniel Pinkham Hebrews 12: I&2, “Wherefore seeing we also are ..." Romans 8:35,37-39, “Who shall separate us...” Colossians 3:16, “Let the word of Christ .. I Thessalonians 5:1-6, “But of the times ...” Philippians 4:4-7, “Rejoice in the Lord alway” Romans 13:11 & 12, “Now it is high time to awake ..."

Songs of Joseph Marx Joseph Marx Waldseligkeit (1911) Nocturne (1911) SeligeNacht (1915) Hat dich die Liebe beriihrt (1908)

xi Recital II

January 29 & 31,1987 at 8:00 p.m. Mershon Auditorium, The Ohio State University Columbus, Ohio

LA RONDINE

by Giacomo Puccini

Producer/Director Roger L. Stephens Musical Director/Conductor Marshall Haddock Lighting Designer J. Patrick Merryman

The Cast

Magda de Civry, Rambaldo’s mistress Sarah E. Wilson Ruggero Lastouc, son of a friend of Rambaldo David Rives Rambaldo, Fernandez, a wealthy Parisian banker Maik Baker Lisette, Magda’s maid Susan Foster Prunier, a poet Thomas A. Gregg Yvette ) Suzy Hutchinson-Lane Bianca ) friends of Magda Tiffany Hartlage Suzy ) Mary Manofsky Perichaud ) Douglas Slusher Gobin ) friends of Rambaldo Craig Montgomery Crebillon ) Daniel May Georgette ) Nancy Davis Kiser Gabriella ) grisettes Mardie Millit Lolette ) Janet B. Johnson Butler Daniel May

XU Recital III

Thomas A. Gregg, Tenor Assisted by Ron Cook, Medieval Harp and Crumhom janice Cook, Vielle, Viola da Gamba and Crumhom James Yeager, Harpsichord Carol Rausch, Piano

Tuesday, February 16,1988 at 8:00 p.m. Hughes Hall Auditorium, The Ohio State University Columbus, Ohio

Early French Songs Ben pac d’ iver, et d’ esti Peire Vidal Lanquant li jorn sont long en mai Jaufré Rudel Phyton, le mervilleus serpent Guillaume de Machaut

Madrigals for Solo Voice Claudio Monteverdi Quel sguardo sdegnosetto Ohimè! ch’io cado Maledetto, sia I’aspetto from Die Schone Miillerin Franz Schubert Das Wandern Wohin? Halt! Danksagung an den Bach Am Feierabend Der Neugierige Ungeduld Intermission Cocardes Francis Poulenc Miel de Narbonne Bonne d’ enfant Enfant de troupe

Two Songs Montparnasse Hyde Park

Songs of Stephen Foster Stephen Foster Ah! May the Red Rose Live Alway! Beautiful Dreamer I f You’ve Only Got a Moustache

xiii Rgçital IV

Thomas A. Grcgg, Ténor assisted by James Y eager. Harpsichord Julianne Fish, Horn Timothy Huffman, Piano Stephanie Carter, Soprano James Shull, Violin/Viola Revecca Lively, Mezzo-Soprano Robert Bethel, Cello Robin Rice, Baritone

Saturday, June 4,1988 at 6:00 p.m. Hughes Hall Auditorium, The Ohio State University Columbus, Ohio

Cantata, “Ich weiss, daB mein Erloser lebt,” BWV 160 attributed to J. S. Bach with violin, harpsichord, and cello (1685-1750) Aria: “Ich weiss, dafi mein Erloser lebt” Recit: “Er lebt and ist von Toten auferstanden!” Aria: “Gott Lob, dafi mein Erloser lebt” Recit: “So biet ich alien Teufeln Trutz!” Aria: “Nun, ich halte mich bereit”

Auf dem Strom, Op. 119 Franz Schubert with horn and piano (1797-1828)

Quartette fur vier Solostimmer, Op. 92 Johannes Brahms vocal quartet and piano (1833-1897) O schoneNacht Spdtherbst Abendlied Warum?

Intermission

Four Hymns Ralph Vaughan Williams with viola and piano (1872-1958) Lord! come away! Who is this fair one? Come Love, come Lord Evening Hymn

Canticle II—Abraham and Isaac, Op. 51 Benjamin Britten with mezzo-soprano and piano (1913-1976)

XIV INTRODUCTION

A recent book devoted to the “art song” genre says that “[i]n most art songs the quality of the poem or text is of the utmost importance.... ” ' The conventional basis of vocal music is indeed the setting of a text. It follows, then, that the first step in the process of composing a song necessarily becomes the search for the words to be set. Ned Rorem confirmed this: “The composer’s initial job is to find an appropriate poem.”^

Every song text is selected for its suitability in expressing the feeling or emotion the composer wishes to portray in the song. “What do composers seek in a text? Words that have a melodious quality of their own, words that evoke definite moods, words that flow rhythmically, and above all words that release the musical inspiration within them.”^

The method the composer uses in finding each text is a personal matter, although external factors like taste in literature developed through formal education and suggestions of verses from friends may be influential. For each completed song a unique and permanent association is formed, linking the artistic expression in die composer’s music with that of the author. A secondary and perhaps equally important connection arises between the aesthetic values of the composer and his or her “authors” through the text selection process.

This document will examine these associations of composers and authors in a concrete manner, through a study of the textual selections made by a diverse group of song

'Barbara Meister, An Introduction to the Art Song (New York: TapHnger, 1980), p. 14.

^Ned Rorem, Music From Inside Out (New York: George Braziller, 1967), p. 39.

^Meister, op. cit., p. 17.

1 composers. Documentation of the poetry choices for each of the composers in this paper has been previously completed by other writers, either formally or in various biographical contexts, providing a starting point for the present study. By including composers from several time periods, song genres, and nationalities, however, this paper attempts to provide a broader, comparative view of the discussion of song composers and their textual choices.

DeOnitions

Several terms must be defined at the outset of this study. First, the use of the words songand art songwill be used inclusively in this document, i.e., these terms will apply to works for voice and piano in all national styles. The definition of “art song” given in The

New Grove Dictionary o f Music and Musicians, for example, allows for such diversity:

“A song of serious artistic purpose written by a professional composer, as opposed to a

folksong. The term is more often applied to solo than to polyphonic songs and embraces the 19th-century lied and mélodie.”^ Meistefs definition is confirmed here: “In this work we shall refer to serious songs of all nations, in all languages, as art songs.

The words lied and méiodie are more specialized terms which refer to the song genres of Germanic and French cultures, respectively. Whitton provides this initial explanation of

the term Lied:

Generalized definitions are always dangerous, but perhaps we might hazard the following: ‘A Lied—or perhaps more accurately a Kunstlied, literally an “art-song”—is the musical setting of an original poem in German.’ The important point is that the poem existed beforeihe music...

'^“Art Song,” The New Grove Dictionaiy o f Music and Musicians, 20 vols, ed. Stanley Sadie (London: Macmillan, 1980), I, p. 646.

^Meister, op. cit., p. 12.

^Kenneth Whitton, Lieder: An Introduction to German Song (London; Julia MacRae, 1984), p. 3. Benton and Noske give this definition for mélodie: “The accompanied French song of the 19th and 20th centuries, usually a setting for voice and piano of a serious lyric poem.”^

A prominent feature in the description of both Lied and mélodie is the use of pre-existing, lyric poetry, an item of particular importance for this document. Further background on these genres will be presented later, in the Introductions to Chapters I and II.

One final term in this document which needs an explanation is the word “generation.”

Webster's Dictionary defines “generation” as “The average period (about thirty years) between the birth of one generation and that of the next... .”8 The distinction of a thirty- year period for a generation will be upheld in the comparative dating of composers and authors. The composer’s own generation, for purposes herein, is defined as the period extending fifteen years before and after his year of birth.

Choice of Composers

The textual choices of twelve song composers are examined in this document. The composers were selected for their significant, widely recognized contributions to song literature. For comparative purposes it was decided to select composers from three song genres: German Lied, French mélodie, and American art song. The lives of the composers in each group cover a period of approximately 100 years, and each group is represented by four diverse individuals. The four Lied composers are Franz Schubert, Robert Schumann,

Joharmes Brahms, and Hugo Wolf. Henri Duparc, Gabriel Fauré, Claude Debussy, and

Francis Poulenc are the four mélodie composers. The American art song composers are

Charles Ives, Samuel Barber, Ned Rorem and Dominick Argento.

^Rita Benton and Frits Noske, “Mélodie,” The New Grove Dictionary of Music and Musicians, op. cit., XU, p. 112.

^Webster's New World Dictionary, Second College Edition, ed. David B. Guralnik (New York: The World Publishing Co., 1970), p. 581. Method of Study

A definitive list of the songs by each composer to be included in the study was first established, using recent and authoritative works-lists and catalogs. Published songs were counted in every case, while unpublished songs were included only when documentation was available. In an effort to maintain the “purity” of the study, only works for one voice and one piano were counted. The following were not a part of any composer’s final count:

1. incomplete or lost works 2. traditional folksong arrangements 3. works derived from choral music 4. works for more than one voice or more than one piano 5. works originally composed for voice and orchestra 6. vocal music originally part of a larger work (e.g., cantata or oratorio) 7. vocal music originally part of a staged work (e.g., play or opera) 8. works originally with obbligato instruments or instrumental ensemble

The source (author) of the text for each song was then determined. These sources were most often found with the song titles in the composers’ works-lists and catalogs. Some basic information for each source was then collected, generally including the author’s year of birth, year of death, and nationality, in addition to the number of texts the composer chose from each source. When this information was complete, the various averages and percentages (described in the next section) were finally calculated. In order to make both likenesses and divergences among differing nationalities and eras visible, calculated values in this document are stated in comparable relative terms. The results of this process are the essential numerical components of the study. Categories for Presentation

The infoimation and calculations gathered for each composer and his sources were then divided into seven general categories, which fonn the structure of the presentation for the individual composers, as well as for the three chapter summaries and the final summary.

1. Educational and Literary Background

The early life and formal education of the composer is briefly detailed, with an emphasis on his individual studies in languages and literature which may have been influential later in the textual choices. This is followed by a discussion of the ways in which literary concepts and people were significant in the life of the composer, the ways in which the composer manifests his literary interests (besides songwriting), and the influences any of these may have had on the textual selections.

2. Poetry Choices

This is a description of the actual texts which each composer selected. It is arranged in a slightly different marmer for each composer, e.g., by source or author, by the chronology of the compositions, or from the most frequently chosen sources to the least frequently chosen ones.

3. Number of Songs and Sources

Numerical information for the songs is presented in this category. The exact numbers of songs and sources are recorded first. The song catalogs and work-lists from which the number of songs and sources were derived are also cited here.

Three averages (arithmetic, median and modal) derived from this information follow.

The arithmethic mean is calculated from the total number of songs divided by the number of sources. The median is the number of texts selected from the source in the middle of each composei^s complete list of sources, arranged by highest to lowest number of choices. In cases where the total number of sources is an even number, the median is determined by averaging the amounts from the two middle sources. The modal value is the number of texts the composer most frequently selected from his sources; i.e., the most frequently occuring number in the complete list of sources. At the end of this category is a brief interpretation of the data.

4. Authors Categorized by Birth Date

This is the first of the categories which group the composers’ choices by date. Each text is dated by relating the birth year of its author to the birth year of the composer. The first group in this category includes the choices from authors bom 16-100 years before the composer (i.e., from “older generation” authors). The next group consists of selections from authors of the composer's own generation, bom no more than 15 years earlier or later than the composer (i.e., from “contemporary” authors). The final group contains choices

from authors bom more than 15 years later than the composer (i.e., from “younger generation” authors).

5. Choices of Undated and Old Sources

Two additional groups of sources are described in this section. The first includes those texts which the composer chose from unknown, anonymous, or undatable sources.

Undatable sources are identified authors about whom little or nothing is known, especially the year of birth. The second group in this category consists of texts by authors who were bom more than 100 years before the composer. These are referred to as “ancient” texts, or as being from “old” sources, distinguishing them from texts chosen from the “older generation” authors in the previous section. Note: the combined percentages in categories four and five are equal to 100% (i.e., all of the choices) for each composer.

6. Choices in Non-Native Languages

All texts the composer used which were originally in a language other than the composer’s native language are part of this category. These include texts from both identified and anonymous authors, set to music either in their original languages or in a translation. Information pertaining to each composer’s identification with foreign countries, languages, and the translators is mentioned when applicable. The percentage in this category is calculated independently from all previous groups, because it is not dependent upon the dating process used for the previous two categories.

7. Personal Contact with the Poets

As an appendage to the information gathered about each composer, this final section discusses those authors, from whom texts were chosen, who were in personal contact with the composer. “Personal contact” refers to the composer’s first-hand knowledge of the author through correspondence or in at least one actual, documented meeting. If the composer himself wrote one or more of the texts he used for his songs, he is also considered a “personal contact” in the count for the composer. Authors who were possibly in contact with the composer are mentioned at the end of this section if appropriate. CHAPTER I

TEXTUAL SOURCES FOR THE LIED COMPOSERS

Introduction to Chapter I

The Lied genre and its ties to literature are described in Eric Sams’ article on the subject in Tne New Grove Dictionary:

In the 19th century the German vernacular song developed into an art form in which musical ideas suggested by words were embodied in the setting of those words for voice and piano.... The genre presupposes a renaissance of German lyric verse, the popularity of that verse with composers and public, a consensus that music can derive from words, and a plentiful supply of techniques and devices to express that interrelation.... The Romantic lied directly mirrored [the 18th century] literary developments by combining the styles and themes of opera, cantata or oratorio with those of folk or traditional song, and reducing the result to terms of voice and keyboard. i

The Lied continued to evolve through the nineteenth century, and is now recognized as a significant component of western music. By the end of the nineteenth century, four composers had emerged as the dominant masters of the genre: Franz Schubert, Robert

Schumarm, Johatmes Brahms, and Hugo Wolf. The choice of texts for their vocal works represent the entire range of German lyric poetry before 1900, and includes works from the master poets, minor and obscure poets, anonymous and folk sources, as well as a variety of poems translated from other languages. A detailed analysis of the poetry chosen by each of these four composers is presented in Chapter I.

The educational and literary background of each Lied composer, also described in this chapter, shows that they had some basic formal education and a distinct interest in literature. Early experimentations in finding a type of poetry which met their individual

’Eric Sams, “Lied, §IV,” The New Grovs Dictionary o f Music and Musicians, op. cit., X, p. 838. 8 9 expressive needs resulted in the eventual creation of a mature body of vocal works set to poetry well suited to their individual musical styles. 10 FRANZ SCHUBERT

Schubert's Educational and Literary Background

Franz Peter Schubert (1797-1828) began his formal education, in addition to his earliest musical education, with instruction from his father Franz Theodor Schubert, a successful schoolmaster in Vienna. Through an open audition in 1808, at the age of 11, Franz won a scholarship to sing as a choirboy in Vienna’s court chapel and to study at the Imperial and

Royal City College, the Kaiserlich-konigliches Stadtkonvikt. He maintained good marks at school for four years, “...but in 1812 and 1813 his increasing devotion to composition began to affect his other work, so that he failed to maintain the highest standard, particularly in Latin and mathematics.’’^

In the autumn of 1813 he turned down an endowment to remain a student at the

K onvikt, and enrolled at the Normal High School of St. Anna, a teacher’s training school in Vienna. He passed a qualifying exam in August 1814 to become a primary teacher; this was the end of his formal education. This background enabled him to earn a modest living as a teacher for a few years and in later life to enjoy the company of many intelligent friends and acquaintances.

The informal education which Schubert received from these friends is perhaps a more important consideration when discussing his choices of song texts. Deutsch states that,

“As a composer of songs he showed a fair knowledge of German lyrical poetry since 1750, but better-read friends advised him.’’^ During his student years, Schubert’s friends introduced him to some fine lyric poetry (e.g., Goethe, Schiller, Matthison). Later he was led to the Romantic poets (e.g., the Schlegels, Novalis, Riickert) by his friend Franz von

^John Reed, Schubert, The Master Musicians Series (London: J.M. Dent, 1987), pp. 9-10.

^Otto Erich Deutsch, “Schubert the Man,” in The Music o f Schubert, ed. Gerald Abraham (Port Washington, New York: Kennikat Press, 1947, revised 1969), p. 11. 11 Schober, as well as Schober’s circle of literary enthusiasts. There is no doubt that their regular readings ofJiterary works, which Schubert attended in the early 1820’s, influenced his song text selections at that time. Even after his own knowledge of literature was fully developed, he continued to be influenced by his friends’ taste in poetry. In fact, Diirr takes

Deutsch’s statement one step further by saying that Schubert’s “. ..knowledge of the literature of his own time was certainly no less comprehensive than that of his friends in the

Schober circle.”'* Ernst Hilmar agrees: “Schubert’s level of literacy as a composer or performer at the time was remarkably high.... ”^

Schubert’s Poetry Choices

The quality of Schubert’s poetry choices has been an issue often discussed by scholars.

Many claim (based on impressions recorded by Schubert’s friends) that he set whatever poetry was at hand, good and bad. Smeed commented that “If Schubert has limitations... it would be in the field of.. .his choice of texts.”® Others such as Diirr have sought to prove this false: “Whilst Schubert selected poems by one author from a particular volume of poems, he did not choose them indiscriminately; only very seldom did he set adjacent poems in a volume.”^ Later in his analysis Diirr concludes that: “Schubert did not therefore indiscriminately choose the words of his songs from ‘any volume of poems which happened to be at hand’—he set to music whatever had a special attraction for him,

^Walther Dürr, “Schubert’s songs and their poetry: reflections on poetic aspects of song composition,” SchwbeTt Studies, ed. Eva Badura-Skoda and Peter Branscombe (Cambridge, England: Cambridge University Press, 1982), p. 11.

®Emst Hilmar, “The Literary Scene,” Franz Schubert In His Time, tr. Reinhard G. Pauly (Portland: Amadeus Press, 1988), p. 102.

®J.W. Sneed, German Song and Its Poetry, 1740-1900 (London: Croom Helm, 1987), p. 114.

^Walther Diirr, “Schubert’s songs and their poetry: reflections on poetic aspects of song composition,” op. cit., p. 5. 12 and rejected much that was put before him."* Hi^mer concurs with the latter: “It is not widely appreciated ihat Schubert typically evaluated and chose his texts with great care.... ”9

Schubert most frequently chose verses by Johann Wolfgang von Goethe. He continually returned to the works of this master writer, and songs using Goethe’s poetry were written throughout Schubert’s compositional life. By the end of 1814, he had set four of Goethe’s poems, and his famous setting of one of them, “Gretchen am Spinnrade”

(D118),1° is often considered the beginning of his mature songwriting period, and perhaps the beginning of significant lieder composition in the 19th century. The next year, 1815, saw further settings of Goethe’s works, including many short, simple, strophic songs, such as “Heidenroslein” (D257), as well as some works with a modified strophic cast, such as “Kennst du das Land?” (D321) from Goethe’s novel Wilhelm Meister. Schubert’s masterpiece song, the “Erlkonig" (D328), also dates from 1815.

During 1816 he set another declamatory song from Goethe, “An Schwager Kronos”

(D369), as well as the three “Gesange des Harfners” from W ilhelm Meister(D478).

Further significant Goethe settings include “Ganymed” (D544) in March 1817, and

“Prometheus” (D674) in October 1819. Two years later he set several of Goethe’s love poems, including at least one of the “Suleika” settings (D720). More Goethe songs were composed in December 1822, among them the brief yet masterful “Wandrers Nachtlied”

(D768). Schubert’s final Goethe songs from January 1826 were resettings of earlier choices from Wilhelm Meister, the last of which was a version of “Nur wer die Sehnsucht kennt” (D877), which was to become a text chosen by many later song composers.

^Ibid.,p. II.

®Emst Hilmar, “The Literary Scene,” Franz Schubert In His Time, tr. Reinhard G. Pauly (Portland: Amadeus Press, 1988), p. 101.

’°These numbers identify Schubert’s works in the catalog by Otto Erich Deutsch, Schubert: Thematic Catalogue o f ail his VPbritsCLondon, 1951). 13

Throughout his compositional life Schubert also selected poems from the works of

Friedrich Schiller, although many of the Schiller settings were written before Schubert was twenty. The length of Schiller's poems, coupled with Schubert’s attempts to pattern his early songs after the long ballads of earlier song composers such as Johann Zumsteeg, make the Schiller settings generally less successful, and most are infrequently performed today. Jack M. Stein is quite critical of the youthful Schiller settings: “In the early through- composed Schubert ballads, his tendency to pictorialize musically any and all dramatic elements emphasizes these individual features to a degree that becomes ludicrous. This is especially true of the Schiller ballads, because in many of them the melodrama and pathos approach the borderline of good taste.’’ * * Other later songs on texts by Schiller have been better received; considered the best among them are “Gruppe aus dem Tartarus” (D583) of

1817 and “Dithyrambe” (D801), usually dated 1824.

Like the earliest Schiller choices, Schubert’s settings of a number of poems from the works of the well-known writer Friedrich Klopstock in 1815-16 resulted in weaker songs.

Schubert also chose a number of poems by so-called “minor poets” in these years, especially the works of three writers: 29 by Friedrich Matthisson, 23 by Ludwig Holty, and 21 by Ludwig Kosegarten. Texts by Holty and Kosegarten were not chosen after

1816, and Matthisson was abandoned after 1817.

Later in 1816 Schubert set several poems by Matthaus Claudius; the best known of these is “Das Tod und das Madchen” (D531). A large number of works by various other writers were added to the list of Schubert’s choices in subsequent years. A partial list includes Johann Gaudens von Salis-Seewis, with choices ranging from solemn to lighter verses; Christian Daniel Schubart, author of the popular “Die Forelle” (D550); poetry in a variety of moods by Johann Georg Jacobi; Alois Schreiber, from whom Schubert selected

* * ^ Jack M. Stein, Poem and Music in the German Lied from Gluck to Hugo Vyb//'(Cambridge, Massachusetts: Harvard University Press, 1971), p. 59. 14 the long poem “An den Mond in einer Herbstnacht” (D614) for setting in April 1818; the obscure poet Georg/Phillip Schmidt von Liibeck, writer of “Der Wanderer” (D489 and

D493) on the “wandering” theme so important to Schubert and popular at the time; and one poem by the important writer and translator Ludwig Uhland, “Friihlingsglaube” (D686).

Schubert started setting the poetry of Wilhelm Müller in 1823, in the first of his two important song cycles, Die SchoneMiillenn of 1823-24. Maurice Browns tells us that, “In writing the story of the young miller and his doomed love for his master’s daughter, Müller used a conventional literary motive of the period.”'^ From these “conventional” poems,

Schubert fashioned a cycle of twenty songs, in a variety of formal schemes (strophic, modified strophic, episodic, and through-composed), and with a somewhat continuous flow of drama from begiiming to end.

He returned to MüMeris poetry in 1826-27 to create his other important song cycle

Winterreise. This set of twenty-four songs is not dramatically continuous as is the earlier cycle, but carries through the theme of “the wintry landscape through which the embittered lover wanders.”'^ While these two songs cycles are generally considered to be among the finest Lrederwritten, Jack Stein’s comments reflects a typical attitude about Schubert’s choice of the Müller poems:

[The poems] do not begin to measure up to the glorious music. The settings ennoble the verses to which they are written to an extraordinary degree. With these songs even the literary-minded listener would do better simply to let the words serve as a general verba! background for some of the most enjoyable works in the entire literature of song.*'*

While composing Die Schone Mui/erin Schubert was also attracted to the poetry of

Friedrich Rückert and chose five of his poems in 1822 and 1823. Maurice Brown

'^Maurice J.E. Brown, Schubert Songs(London: British Broadcasting Corp., 1967), p. 33.

13/h/d,p. 51.

^“^JackM. Stein, op. cit., pp. 95-96. 15

comments that, “All five are in his finest lyrical vein and are very p o p u l a r .”^ ’ o f these

five, the well-known “Du hist die Ruh” (D776) is usually considered the masterpiece.

Another poet whose works Schubert set during this time was Matthaus von Collin: of the

four songs set to his poetry, the most famous is certainly “Nacht und Traume” (D827).

Other poets’ works were selected for settings only at the end of Schubert’s life, such as the

two from the obscure Karl Lappe in 1825, nine from Ludwig Rellstab in 1827-28, and six

Heine settings in 1828. Seven of the Rellstab songs, along with the six Heine settings and

a version of Seidl’s “Die Taubenpost” were collected and published posthumously under

the title Schwanengesang by Haslinger in the spring of 1829.

Schubert’s friends were the authors of over a hundred of his song settings. Maurice

Brown wrote that, “Among Schubert’s friends were several tme poets, but most of the

members of his circle wrote verses.”’^ From his good friend Franz von Schober he took

twelve verses, including the famous “An die Musik” (D547). However, he used only five

poems by another friend, Franz von Bruchmann, in settings from 1822-23. Schubert

selected a substantial number of texts by his friend Johann Baptist Mayrhofer and set 46 of

his texts (plus one of his translations from the ancient Greek Aeschylus). Brown believed

that Mayrhoferis poetry would be obscure today without Schubert’s song settings: “Apart

from Schubert’s selection from his poems, his work is fbrgotton.”’^ Reed gave a positive

description of these songs:

The best Mayrhofer settings combine intellectual strength with lyrical power. They vary in style from the ‘simplicity’ of “Nachtviolen” [D752] and “Erlafsee” [D586] to the ragged intellectuality of “Freiwilliges Versinken” [D700] and “Der Sieg” [D805], and there are hardly any weak songs among them.’®

’ ^Maurice J.E. Brown, op. cit„ p. 38.

p. 45.

Ibid., p. 46.

’®John Reed, The Schubert Soug Companion (Manchester, England: Manchester University Press, 1985), p. 470. 16

Others in this last group of choices include two poems from Johann Ladislaus Pyrker, a friend of Schubert’s youth, and seven from Franz Xaver von Schlechta, a loyal friend and admirer of Schubert who wrote occasional verses.

Many other writers not mentioned here may be added to this group. The entire list is presented in Table 13, pp. 232-35, including all of the authors’ birth and death dates and the number of poems Schubert chose from each source. A few of the poems will be mentioned later, as part of the various categories of Schubert’s choices. 17

' Table 1. The Songs of Franz Schubert

Description ______Number_____% Total number of Songs 624

Total number of Sources 9 7

Arithmetic average number of songs per source: 6.43

Median number of songs per source: 2.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 383 61.38%

Texts from poets contemporary with composer (bom within 15 years of composer) 210 33.65%

Texts from poets younger than composer (bom 15 years after composer) 0 0.00%

Texts from undated sources (dates of poet or source unknown) 21 3.37%

Texts from old sources (dated more than 100 years before composer's birth) 10 1.60%

Texts from non-native sources (either in translations or in other languages) 45 7.21%

Approximate number of poets with whom the composer had personal contact: 19 18

Schubert’s Number of Songs and Sources

The exact number of Schubert’s songs has long been a point of debate, due to the high number of variant versions of songs, lost songs, unfinished songs, etc. The number has been set as low as 603 (Series XX of the older Gesamtausgabê) in the 1890’s and as high as 708 (by Maurice Brown) in 1966. John Reed more recently (1985) has carefully addressed this matter*®, and his total is 631. Reed’s total is the basis of the present study, although seven of the songs which he includes in his count have been dropped here (D183,

D352, D607, D645, D663, D943, and D965), based on the criteria set forth in the

Introduction. This makes Schubert’s total for this study 624 songs, the largest number among the twelve composers. This large number will be considered further later in this document, when compared to other lieder composers, as well as when compared to the entire group of composers at the end of the paper.

Reed further lists and totals Schubert’s songs by source \n Part II of his book^o, and also provides brief biographies of the authors of Schubert’s song te ;:ts. Using this as a primary guide, 97 sources are presently counted; the complete list of these sources is presented in Table 13, pp. 232-35. A number of other sources were also critically consulted for the compilation of this list and the data contained therein, particularly the

Work-List prepared by Eric Sams for The New Grove Dictionary o fMusic and

Musicians?-^

Schubert averaged 6.43 songs per source, based on 624 songs from 97 sources. This

figure does not entirely represent the overall pattern in his choices, for while he set 72 texts by Goethe alone, he chose only one or two texts from more than half of his authors.

*®Reed, “How Many Schubert Songs?,” Appendix I, ibid^ p. 483.

2®Reed, “Part II, The Authors,” ibid., pp.461-481.

2 ’ Brown, Maurice J.E., with Work-List by Eric Sams, The New Grove Schubert (London: MacmiUan, 1980,1982), pp. 125-158. 19 making the median average only 2 songs per source. In addition, the modal value (the most commonly occurring value) for his choices is even lower, at only one song per source. The first of these averages, 6.43 songs per source, is the second highest for the twelve composers in this study (see Figure 36, p. 220). The second (median) average, however, is relatively low among these composers (see Figure 37, p. 221). Furthermore, the modal value of one in this study is the lowest number possible for the given data.

These averages indicate that while Schubert chose a relatively high mimher of texts from a number of authors, most of the sources were used only once or twice (a relatively low number). This helps to explain the large discrepancy between the top of the list (e.g.,

Goethe, with 72 text settings) and the general pattern of his choices (40 sources used only one time).

Schubert’s Authors Categorized by Birth Date

A breakdown of Schubert’s choices of poets by birth date (see Table 1) reports that, of the 624 song texts, 383 texts were written by poets bom from 16 to 100 years before 1797

(between 1697 and 1781), 210 texts were by poets bom in the 15 years before and after

1797 (between 1782 and 1812), and no texts were by poets bom more than 15 years after

1797 (after 1812). These numbers may also be stated in percentages: 61.38% of

Schubert’s total selections were written by poets bom in the 16-100 years before 1797;

33.65% of the material was written by poets bom in the years 1782-1812; and 0.00% was written by poets bom after 1812.

There are several possible reasons for this distribution. First is Schubert’s short life span. Had he lived longer, he may well have set more poetry from his own generation and from the younger poets of his time whose works appeared later. This can only be speculated, but his pattem for choosing contemporary texts is established in the figures of the previous paragraph, and therefore probably would have continued. It will also be 20 shown later in this paper that the song composers with a life span greater than that of the short-lived Schubert did use poetry by younger writers.

Another factor in this discussion is that Schubert’s peak years for songwriting (largest

quantities composed) were between 1815 and 1817, ata time when he was only 18 to 20

years old. This relatively young peak age would likely indicate that much of the important

anchor published poetry of his own generation was not yet written, and that the works of

older writers were what he had most available for his use during those years. This can

only be a partial explanation, since he was already setting some contemporary poetry (much

of it unpublished) in his younger peak years, and he frequently returned to the writers of

the older generations later in his composing life. A further discussion of the cyclic patterns

in Schubert’s textual choices is detailed in Reed’s Appendix

Two final considerations related to Schubert’s own tastes and those of the general

public. The fact that such great poets as Goethe and Schiller were becoming highly

esteemed by the public mote than likely influenced his choices. Other poets outside the high public visibility of the time yet highly esteemed by Schubert, such as Wilhelm Müller and Johann Baptist Mayrhofer, were significant to his work in spite of their relative

obscurity, partly because they were of his own generation. It has often been stated that the

rise of nineteenth century Lieder was to a large extent due to the simultaneous rise of

German lyric poetry. If Schubert is to be considered the first of the truly great Lieder composers, then the numbers presented here show that a majority of the great poetry he set to music came from poets bom in the generations before him, and that the begiiming of the era of great poetry and of great songs did not exactly coincide.

^^Reed, “Schubert’s Poetic Progress: The Cyclical Method of Composition,” Appendix V, op. cit., pp. 499-501. 21 Schubert’s Choices of Undated and Old sources

Franz Schubert’s uses of anonymous sources were rare in comparison to his total song output: only 19 texts among the 624 are by unknown or anonymous authors, a small

3.04% of his total. Two other songs may be added to this undated category, one by

Bernhard Ambros Ehrlich and another by A. Pollack; both are authors about whom little or nothing is known. These two add another small percentage to the choice from undated sources, making the total for the group only 3.37% of his entire selection.

Schubert’s use of old textual materials (those from authors whose years of birth predate the composer’s by more than 100 years) forms an even smaller group. He set six English poems (all in German translations) from the 16th and 17th centuries, by Shakespeare,

Cowley, Cibber, and Pope, in addition to three texts from the 14th century Italian poet

Petrarca, also in German translations. One further ancient text which he set was by the 5th century B.C. Greek writer Aeschylus, in a translation by Mayrhofer (see Table 13).

These ten texts represent only 1.60% of Schubert’s total choices, making this group the smallest of any of his categories (aside from the 0.00% in the younger poets category). A number of his other choices may be indirectly linked to this category, including

Bruchman’s free translation of a verse by Anacreon from the 6th century B.C., and the ten settings attributed to an ancient Gaelic bard named Ossian. The latter are described in the next category, texts from Non-native languages.

Schubert’s use of ancient texts is infrequent compared to other composers in the present study; this will be further discussed later. Schubert’s choices in this category, however, do point out that he did know and set a number of poems from the distant past. It was during the years immediately after Schubert’s birth that a number of German philologists began searching out, translating, and publishing the important lyrics from both the past and from other countries, increasing the number of these poems generally available by the early years 22 of Schubert’s life. It is thus not surprising that he would select some, if not a great amount, of this older poetry for his song settings.

Schubert’s Choices in Non-native languages

Schubert’s textual choices in non-native languages were greatly influenced by the popularity and availability of poetry recently translated into German from other countries.

The 45 non-German poems which he set represent 7.21% of his total choices.

Schubert’s early studies with the Italian composer Salieri were another influence in this category. Salieri, who had learned much about writing music for the Italian stage from the great poet and librettist Pietro Metastasio, gave the young Schubert a number of

Metastasio’s verses as text-setting exercises. Schubert eventually set twelve Metastasio texts in the original Italian and one in a German translation by Heinrich Josef Collin.

Another group of ten settings in this category are the “Ossian songs.” These texts were from a volume by Scottish writer and schoolmaster James Macpherson, who falsely claimed they were written by the ancient Gaelic bard Oisin (Latin ‘Ossian’). Published in

1760, they quickly became a sensation in Germany; Schubert used the translation by

Edmund von Harold. Reed tells of the poems’ influence in Germany:

Between 1760 and 1770 at least 10 German translations of Ossian appeared, and for fifty years and more the Ossianic literature set a fashion of irresistible appeal to young and romantically minded artists. Its influence can be seen not only in Schubert’s settings of Ossian but in the ballads of his contemporaries, like Josef Kenner and Johann Mayrhofer (not to mention greater names like Goethe and Schiller), all of them full of ruined castles on storm-girt cliffs, mossy banks and cloudy seas.23

The remainder of his settings from texts outside the German language include the older

English and Petrarca settings mentioned above; eight by Sir Walter Scott from various translators; one anonymous Italian text (“II modo di prender moglie,” D902, no. 3); two by

23Reed, “Ossian Songs,” ibid., p. 351. 23 the Venetian author Jacopo Andrea Vittorelli (both originally attributed to Metastasio); and

one by Carlo Goldoni, a Venetian opera librettist. Reed describes this last song, “La

Pastorella al prato,” (D528) as a “charming exercise in the Italian style made shortly after the first visit of the Italian opera to Vienna.’’^-» Most of Schubert’s foreign-language text settings are not as stylistically related to the poem’s original language. The German translations enabled him to use a more Germanic (and thus more personal) treatment of these texts.

Schubert’s Personal Contact with the Poets

Franz Schubert had personal contact with at least 19 poets from whom he selected texts

(their names are marked with an asterisk in Table 13, pp. 232-35). Schubert never met or knew Goethe, but did send him some of his scores, which Goethe never acknowledged:

The story of Schubert’s attempts to win Goethe’s approval is one of disappointment and misunderstanding. The profound impression made on his friends by the first great masterpieces encouraged them to prepare an ambitious scheme to publish the songs in a series of volumes grouped under individual authors. The first two volumes were to consist of Goethe settings. Early in 1816 fair copies of selected songs were made and sent to Goethe with a suitably respectful introductory letter by Josef von Spaun.... The songbook was sent back without acknowledgement.... Later in 1816 a second volume was prepared, but it was never sent.25

Some of Schubert’s authors, however, were his closest friends, such as Josef von

Spaun, Edward von Bauernfeld, Franz von Schlechta, Franz von Schober, and Johann

Baptist Mayrhofer. Schubert’s contact with Mayrhofet’s personality and ideas played a particularly important part in his thinking and composing. Porter devoted an entire chapter to the collaborations with Mayrhofer in his book on Schubert’s songs.26 He commented

2‘*Reed, ibid., p. 465.

Ibid., pp. 463-464.

25 Ernest G. Porter, “Schubert and Mayrhofer,” Schubert’s Song Technique (London: Dennis Dobson, 1961), C huter IX, pp. 120-130. 24

on their like-mindedness: “Schubert could easily follow his friend’s flights not only when

he exclaimed passionately at his lack of recognition but also when he expressed the half­

hidden thought of all artists that they are of the immortals.”^^ Reed comments that

“Mayrhofef’s importance in Schubert’s work has still not been fully appreciated. ”28

Franz Ritter von Bruchman was another who influenced Schubert’s life and work,

through his link to the Romantic writers of Vienna and the “Schubertiads” held in his house

from 1820 to 1823. The rest of the authors marked in Table 13 had little to moderate

contact with the composer, or may have been in his circle for a year or two. Other authors

not marked in Table 13, such as Grillpaizerand Friedrich Schlegel, may have actually met

Schubert, but there is no documentation verifying it. Although Maurice Brown simply

states that “.. .his selection from their work was prompted by personal liking... ,”29 Walther

Diirr concludes that there is more to be learned from Schubert’s use of his friends’ verses:

There is now no doubt that Schubert’s relationship with these poets and their works was different from that with other writers whom he first came to know through reading and discussion. In the verses of his friends he could hear their own voices, their own delivery; he therefore knew what moved them to write. It is only natural that such texts had a direct and immediate appeal for him. These poems tell us little about his relationship to literature, but they tell us much about his relationship to their authors.^o

This Is also iiue of several of the other composers in this study, as will be seen later in this document.

Ibid., p. 125.

2®Reed, op. cit., p. 470.

29Maurice J.E. Brown, Schubert Songs {London-. British Broadcasting Corp., 1967), p. 43.

2°WaIther Dürr, “Schubert’s songs and their poetry: reflections on poetic aspects of song composition,” Schubert Studies, ed. Eva Badura-Skoda and Peter Branscombe (Cambridge, England: Cambridge University Press, 1982), pp. 10-11. 25 ROBERT SCHUMANN

Schumann’s Educational and Literary Background

Robert Schumann ( 1810-1856) spent his early years comfortably in Zwickau, where his father was a very successful book dealer. Walker states; “O f all the great composers, with the exception of Mendelssohn, Schumann probably started life with the best material advantages."^ ^ He began his education with a private tutor, was placed in a private preparatory school at the age of six, and enrolled at the Zwickau Lyceum, where he studied until the age of eighteen. It was at the Lyceum that Schumann began his early music studies, but his early love of literature had also emerged:

Music apart, the other great passion of Schumann’s youth was literature. In pursuit of his classical studies he devoured Homer, Cicero, Horace and Sophocles; among German writers, Schiller and Goethe were his idols. At sixteen, he helped to run a German Literary Society among Lyceum students, asserting that ‘it is the duty of every cultivated man to know the literature of his country. ’ He also tried his hand at poetry... .Another great influence was Jean Paul [Richter, the German Romantic writer], whose high-flown literary style made such inroads into Schumaim’s impressionable mind that he described Jean Paul’s novel Die Flegeljahrezs in its own way like the Bible.’32

He successfully passed his school-end examinations in March 1828. His father, who had tried to encourage his musical studies, had died two years earlier in 1826. It was his

mother’s wish that he study law instead. He was forced to obey his mother’s wishes by

enrolling at the University of Leipzig, but he spent his time there pursuing both music and

literature, not his legal studies. A year later he convinced his mother to allow him to study

law with Justus Thibaut in Heidelberg (who happened to also be a musical philosopher),

3 ^Alan Walker, “Schumann and his background,” Robert Schumann: The Man and His Music, ed. Alan Walker. (London; Barrieand Jenkins, 1972), p. 1.

32Aian Walker, ibid., pp. 2-3. 2 6 but again Schumann found himself totally engrossed in music. His formal studies ended in

1828, and he began.his musical studies with Friedrich Wieck in 1830.

Schumaim’s early interest in literature was to stay with him into adulthood and deeply influence his music. Fischer-Dieskau says that, “Having grown up surrounded by books,

Schumann viewed the world through the eyes of a poet and spoke the poet’s language.... Experience and pleasure were defined and conditioned for Schumann by literary models, and these early artistic impressions profoundly influenced him throughout his life.’’33 Sams carries the concept one step further. “There is a sense then in which

Schumann is indeed the most lettered of composers. The verbal, the conceptual, the semantic, are of overwhelming importance in his music; and it follows that the words of his songs, though secondary, exert a profound influence of their own.’’^^ Schumann’s later important side-career as an editor and music critic (1835-1844) was one final, visible continuation of his literary endeavors.

Schumann’s Poetry Choices

Schumann set 38 texts by Heinrich Heine, the largest number chosen from one author.

Schumann produced many of his greatest songs with these texts, and all of them were composed in his “Liedeijahre” (Year of Song), 1840. The most famous are the sixteen of the Dichterliebecyclc, op. 48.^5 Another nine form the Liederkreis (Song Cycle), op. 24, the first of two groups of songs with that title. Three Heine settings were included in the collection titled M yrthen, op. 25, of which two, “Die Lotosblume” (no. 7) and “Du bist wie

^^Dietrich Fischer-Dieskau, Robert Scbumaim: Words and Music— The Vocal Compositions, tr. Reinhard G. Pauly (Portland: Amadeus Press, 1988), p. 11.

^‘'Eric Sams, The Songs o f Robert Schumann (London: Eulenburg, 1975), 2nd edition, pp. /.

^^The opus numbers given here for Schumann’s works were origmally assigned to them for their first publications. The authority for this document is the Works-List of Schumann’s songs by Eric Sams, in “Schumann, Robert,” The New Grove Dictionary of Music and Musicians, ed. Stanly Sadie, 20 vols. (London: Macmillan, 1980), XVI, pp. 857-863. 27 eine Blume” (no. 24) are . .certainly among the finest in all Schumann.”36 The other ten were written in 1840 but published later; they are found in op. 45 (no. 3), op. 49 (nos. 1 and 2), op. 53 (no. 3), op. 57, op. 64 (no. 3), op. 127 (nos. 2 and 3), and op. 142 (nos.

2 and 3). The last four of these were originally intended for the DJchterliebe cycle, op. 48, but they were rejected by the composer and published later. The most famous of these last ten songs is op. 49 no. 1, “Die Beiden Grenadiere” (The Two Grenadiers), with its rousing

“Marseillaise” quote.

Schumaim also produced well-known cycles using texts by two other poets,

Eichendorff and Chamisso. He set twelve Eichendorff poems in 1840 for the second and more famous Liederkreis cycle, op. 39. This set of songs (not actually a cycle in the sense of dramatic progression) contains more of Schumann’s most famous songs (e.g.,

“Waldesgesprach,” no. 3, and “Mondnacht,” no. 5). There are only four more Eichendorff settings: two from 1840 in op. 45 (nos. 1 and 2); one originally included in op. 39 but moved to op. 77 (no. 1) for publication in 1850; and one later setting composed in 1850, op. 83 no. 3. Eight of the Chamisso texts were set for Schumann’s famous cycle of 1840,

Frauenliebe und-Leben, op. 42. The other five Chamisso settings also date from 1840: op. 27 no. 3, three in op. 31, and op. 40 no. 5.

Schumaim returned to Goethe’s poetry several times, as Schubert had done, and set a total of twenty texts. There was one early Goethe setting from 1828, “Der Fischer,” and five more were included in the op. 25, M yrthen, of 1840. Two of these (nos. 5 and 6 of

26 songs) are contrasting verses from Goethe’s Schenkenbuch im Divan, and no. 9 is a

“Suleika” love poem setting. Nos. 2 and 8 in the collection, “Freisinn” and “Talismane” are the other two Goethe settings in op. 25.

All but one of the other Goethe choices were from 1849, the centenary year of Goethe’s birth. In that year Schumaim set three Goethe texts for the Lieder-Album fiir die Jugend

Stephen Walsh, Tbe Lieder o f Schumann {New York: Praeger, 1971), p. 23 2 8

(Songbook for Youth), op. 79. The third of these (no. 28, the final song in the collection)

is his setting of “Kennst du das Land,” his first choice from the Mignon songs in Goethe’s

Wilhelm Meister. The largest single group of Schumann’s Goethe songs are the nine of

op. 98A, more settings from the novel Wilhelm Meister. The first in this set was a reprint

of his earlier “Kennst du das Land” from op. 79, and the rest are three more Mignon verses, four Harper’s songs, and the “...tidy but uninspired ‘Singet nicht in Trauertonen,’ which serves i iainly for light relief.”^^ Desmond laments that these Goethe settings were so late in Schumann’s songwriting years. “One cannot help wishing he had composed

them earlier in his career, because the mental trouble from which he suffered periodically was not very helpful, or perhaps too helpful in setting such songs as the Harper’s in which he got too much involved.One final Goethe text was set in 1850, his highly regarded setting of Goethe’s “Über allen Gipfeln 1st Ruh,” op. 96 no. 1, which Schumann entitled

“Nachtlied.”

There are also twenty settings of poetry by Friedrich Riickert. The first and probably most famous is “Widmung,” no. 1 of the Myrthen, op. 25. There are four more Riickert settings in this collection (nos. Î 1, 12, 25, and 26) which are not as well-known. Eight

Riickert choices date from 1840; “Jasminenstrauch,” op. 27 no. 4, six of the twelve

“Liebesfriihling” songs of op. 37, and “Volksliedchen,” op. 51 no. 2. Six more were chosen later “Schneeglockchen,” op. 79 no. 26 and the four solos in the Minnespiel, op. 101 dating from 1849, and “Die Blume der Ergebung,” op. 83 no. 2 which dates from

1850. One final Riickert setting, “Glockentiirmers Tochterlein,” has no opus number and the date of composition is unknown.

^^Stephen Walsh, ibid., p. 85.

Astra Desmond, Schumann Songs, BBC Music Guide No. 22 (Seattle: University of Washinj^ton Press, 1972), p. 49. 29 Most of the nineteen verses by Justinus Kemer and nine by Emmanuel Geibel did not prove to be outstanding choices for Schumann. Five early Kemer songs date from 1828.

A larger group of twelve were published as op. 35. These were written at the end of his

“Liedeijahre,” ( 1840) and published a year later. These twelve songs are not considered on a par with the earlier, more famous works from 1840. Two more Kemer texts were set in

1840 emd intended for op. 35 but were published later, the “Sangers Trost,” op. 127 no. 1, emd “Trost im Gesang,” op. 142 no. 1.

Four ofGeibel’s verses date from August 1840: the three outgoing, masculine songs of op. 30 and the contrasting “Sehnsucht,” op. 51 no. 1. These are all considered fairly successful settings, unlike the other Geibel settings from 1849. Three are solos from the

Spanisches Liedcrspiel, op. 74: “Melancholie” (no. 6) for soprano, “Gestandnis” (no. 7) for tenor, and “Der Kontrabandiste” (no. 10) for baritone, the final song in the collection.

The other two were called “Zigeunerliedchen” (no. 7) in the Lieder-Album fiir die Jugend

(Songbook for Youth), op. 79, actually two songs grouped together under one title.

Schumann’s choices from other poets produced few important songs. His only setting from his friend Julius Mosen produced the famous song “Der Nussbaum,” op. 25 no. 3.

He set five Morike texts; op. 47, nos. 1 and 2; op. 79 no. 23; op. 107 no. 3; and op. 125 no. 3. Through the years these minor songs have become overshadowed by Wolfs settings of the same poems. Schumann also set some translations of poems by important

British writers, such as Bums, Byron, Shakespeare and Moore. These will be detailed later under the “Old Sources” and “Non-native Languages” categories. Many of the choices from the later years, such as from Lenau and Kulmann did not produce great songs. The remaining authors are relatively insignificant among Schumann’s choices and are listed in Table 14, pp. 236-37. 30

Table 2. The Songs of Robert Schumann

Description Number % Total number of Songs 257

Total number of Sources 50

Arithmetic average number of songs per source; 5.14

Median number of songs per source: 2.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 110 42.80%

Texts from poets contemporary with composer (bom within 15 years of composer) 116 45.14%

Texts from poets younger than composer (bom 15 years after composer) 10 3.89%

Texts from undated sources (dates of poet or source unknown) 15 5.84%

Texts from old sources (dated more than 100 years before composer's birth) 6 2.33%

Texts from non-native sources (either in translations or in other languages) 29 11.28%

Approximate number of poets with whom the composer had personal contact: 22 31 Schumann’s Number of Songs and Sources

The number of songs by Robert Schumann counted for this study is 257. This number was obtained through critical examination of the songs listed in Eric Sams’s Work-List of

Schumann’s compositions in The New Grove Dictionary o fMusic and Musicians?"^ A number of other lists of Schumann’s works were also consulted for comparison. Because

Sams’s Work-List is comprehensive (i.e., it includes all fragments, lost songs, vocal

ensembles, etc.), many individual works were not counted, as defined jn the Introduction

of this paper. Two special cases need to be mentioned here. The multiple sections of both

op. 53 no. 3 and op. 64 no. 3 form complete units, and are thus counted only once each.

The solo songs in the Spanishes Liederspiel (op. 74) and the Minnespiel(op. 101) are

included in this study (but only those for one voice), because they were conceived as

Liedcrand for concert (recital) performance, not for the stage. The group of Spanische

Liebeslieder(op. 138) was scored for piano duet; its solo movements were therefore not

counted.

The number and identity of the sources for Schumann’s songs were also derived using

Sams’s Work-List as a primary guide, followed by a similar comparison with other works

lists, especially Sams’s own earlier published list of Schumaim’s sources.^® The count

shows that Schumann used a total of 50 sources; they are individually listed in Table 14,

pp. 236-37, by name, dates of birth and death, and the number of their texts Schumann

used.

These two totals (257 songs from 50 sources) show that Schumann averaged 5.14

songs per source. Further examination of Table 2 indicates a median average of 2 songs

per source, however, and the modal value (most commonly occurring value) is only one

^^Work-List by Eric Sams, in “Schumann, Robert,” The New Grove Dictionary o f Music and Musicians, ed. Stanly Sadie, 20 vols. (London: Macmillan, 1980), XVI, pp. 857-863.

Eric Sams, “Appendix III: The Sources,” Ihe Songs o f Robert Schumann (London: Eulenburg, 1975), 2nd edition, pp. 278-281. 32

song per source. The first of these averages (5.24 songs per source) is among the highest

for the twelve composers studied, while the other two averages are comparatively low

among these composers. Schumann’s choices, like Schubert’s, show a disparity between

the large number of choices from a few sources (e.g., 38 from Heine) and the small

number of choices from many of the sources (e.g., 24 sources used once), and both of

these tendencies are relatively more extreme for Schumann than for Schubert. It appears

that while Schumann selected a fairly high number of texts from a few sources, he also selected a low number of texts from a relatively large numberof his sources.

Schumann’s Authors Categorized by Birth Date

Schumann’s choices of texts are categorized by their authors’ years of birth as compared to Schumann’s birth year (1810) in Table 2. A total of 110 texts are by poets who predate Schumann by 16-100 years (bom between 1710 and 1794), 116 texts are by poets considered his contemporaries (bora between 1795 and 1825), and 10 texts are by poets bora more than 15 years later than 1810 (after 1825). Stated in percentages, the three categories are as follows: 42.80% of the texts Schumann chose were from poets bora in the 16-100 years before 1810; 45.14% were from poets bora in the years 1795-1825; and

3.89% were from poets bora after 1825.

Schumann had a slight preference for contemporary poetry over texts from older writers. He selected texts from a small group of older poets and chose a high number of texts from many of them. On the other hand, he worked with a large group of contemporary poets but, with one exception (Heine), selected only a few texts from each of them. This suggests that he generally had an interest in fewer of the older authors but set many texts by those he particularly liked; conversely, he had an interest in more contemporary writers but chose a smaller number of texts from most of them. 33 For further explanation of this distribution, it must be noted that Schumann’s peak year

for songwriting was 1840, when he was 30 years old. This statement explains much about

his textual choices. Kenneth Whitton states that:

By the time that Schumann was writing songs, from 1840 on, the literary Romantic Movement in Germany had lost a good deal of its impetus, but Schumaim’s Lieder are, in the main, settings of poems of unimpeachable Romantic provenance: unrequited love, dreams, ‘Romantic’ mined, mediaevel castles, ladies in high towers, moonlight nights, all the literary topoi of the period are to be found there.'* •

By age 30, Schumann certainly had already learned of the poetrj' of the past, and he

was old enough to know the poetry of the present. He had also come to know and admire

the songs of Schubert (and consequently his choices of texts); Schubert’s song texts were

probably at least a partial influence on Schumann’s choices. He was also in love in 1840,

and was ready to express his love through music and poetry together in his songs. Sams

describes Schumann’s need to choose certain poems for their content: “From the volumes

he chose poems which either directly mirrored his own feelings or could be adjusted to

reflect them.”^^ His choices of texts, then, were generally more closely related to their

suitability of expression, rather than whether they were “old” or “contemporary.” It is not

surprising, given these influences and and circumstances, that a majority of Schumann’s

selections were almost equally divided between earlier authors and those of his own time.

The last group of Schumann’s choices in this section (10 verses by younger poets) is a

relatively high number among the composers studied (3rd highest). The first of the two

younger poets was a “Wilfried von der Neun,” whose real name was Friederich Wilhelm

Traugott Schopf, from whom Schumann set six verses for his op. 89 and two for op. 96

(nos. 4 and 5) in 1850. The other was Paul Heyse, from whom he took only two verses,

one in op. 125 (no. 5 in the thematic catalogue, no. 2 in the Complete Edition) in 1850 and

Kenneth Whitton, Lieder: An Introduction to German Song (London: Julia MacRae, 1984), p. 53.

'♦^Eric Sams, The Songs o f Robert Schumann, op. cit., p. 4. 34 the other in op. 107 (no. 4) in 1851. These selections are all among Schumann’s later works (he was 40-41 years old and mentally declining at the time), and both poets were in their early 20’s when the songs were composed. Schumann even asked the young poet

“von der Neun” to meet with him in late May and early June of 1850 about setting his texts, when they “.. .discussed music to be written for the setting of the poems—an unusual procedure.”43 In spite of the obscurity in which these songs now remain, they do show that Schumann knew and worked with some poets and poetry of the following generation, and, were he to have kept his sanity, lived longer, and composed more songs, might have continued to use the works of younger writers.

Schumann’s Choices of Undated and Old Sources

Robert Schumann’s choices of material by anonymous or undatable writers were relatively few: only 15 texts among 262 songs, or 5.84% of the total. These choices fall into two distinct groups:

1. Eight texts from Anonymous Sources:

Source ______Title. Gous Number. Date ______Old German poem Vom Reitersmaim, no opus no., before 1840 Des Knaben Wunderhom Die Ammenuhr, no opus no., 1848 Des Knaben Wunderhom Das Kauzlein, op. 79 no. 10, 1849 Des Knaben Wunderhom Marienwiirmchen, op. 79 no. 13, 1849 Anonymous poem Kinderwacht, op. 79 no. 21, 1849 Old Catholic poem Requiem, op. 90 no. 7, 1850 Anonymous poem Schneeglocken, op. 96 no. 2, 1850 Des Knaben Wunderhom Das Kauzlein II, no opus no., unknown date

43Dietrich Fischer-Dieskau, Robert Schumann: Words and Music— The Vocal Compositions, tr. Reinhard G. Fâüly (Fürîlaiid: Ainadeus Fress, 1988), p. 174. 35 2. Seven texts from Obscure Poets:

Source ______Title. Opus Number. Date ______Carl Christem Ich wandre nicht, op. 51 no. 3, 1840 Wilhelmine Lorenz Loreley, op. 53 no. 2, 1840 Lily Bernhard Madchen-Schwermut, op. 142 no. 3, 1840 Christian L’Egru Auftrage, op. 77 no. 5, 1850 Julius Buddeus Resignation, op. 83 no. 1,1850 Julius Buddeus Die Meerfee, op. 125 no. 1, 1850 Ferdinand Braun Friihlingslied, op. 125 no. 4, 1850

The first of these two groups shows that Schumann did occasionally set an anonymous verse. Three of these choices are in the Lieder-Album fiir die Jugend (Song Album for the

Young), op. 79, which includes many simple and folk-like poems by Hoffman von

Fallersleben and others, in addition to the anonymous texts. Schumann’s “Requiem,” op.

90 no. 7, was written as a tribute to the dying writer Lenau, and is a setting of an anonymous translation of an old Latin poem.

The second group of songs in the undatable category consists of settings of poems by amateur poets who were in contact with Robert and Clara Schumann; little is known about these authors except for their correspondence with the Schumanns, or in some cases their professional relationship with Robert (such as Carl Christem, who was a correspondent for the Neue Zeitschrift furMusik, and Wilhelmine Lorenz, who was the wife of the sub-editor under Schumann for the Neue Zeitschrift furMusik). These texts were indeed amateur, and did not often inspire Schumann’s best efforts at songwriting. Sams describes

Schumann’s two choices by Julius Buddeus as “sickly drivel’^M and “smooth inanities,”-»? and regarding his setting of Wilhelmine Lorenz’s verse “Loreley,” he says that, “Only

'»'»Sams,op. cit., p. 232.

’ 5/ôid, p. 247. 36 some domestic or family connection could ejq)lain how the greatest living song writer [of

1840] came to set such pitiful trash."^^ It is probably fortunate for Schumann that this type of poetry choice was rare!

The data collected on ancient poetry shows that Schumann used only six texts by authors bom over 100 years before him, the smallest group (2.33%) of his choices. This group includes one text from Shakespeare (in a German translation), “Schlusslied des

Narren,” op. 127 no. 5, one of his earliest songs of 1840. The other five texts in this group are attributed to Mary, Queen of Scots (Mary Stuart), in a translation by Gisbert von

Vincke, Schumann’s op. 135, set in 1852. These latter pieces were Schumann’s last efforts in song, and by 1852 he was no longer capable of choosing texts or producing good works: “That his taste and judgment deteriorated is evident in his choice of.. .unsuitable texts like those of the Mary Stuart songs.’’^^ Schumarm was clearly not dedicated to the use of old poetry for his songs, even with these few exceptions.

Schumann’s Choices in Non-native Languages

Schumann used 29 texts originally written in other languages, all set in German translations. This amounts to 11.28% of his total settings. As will be more clearly seen later in this document, Schumarm’s percentage in this category is in the mid-range among the twelve composers studied. All but 6 of the 29 choices are by writers from the British

Isles, including nine by Robert Bums, five by Byron, the five by Mary Stuart and one by

Shakespeare mentioned above, one attributed to Byron but probably by Catherine

Fanshawe, and two by Thomas Moore. These last two by Irishman Moore are actually on

Italian themes (op. 25 nos. 17 and 18, “Zwei Venetianische Lieder,” [Two Venetian

Songs]), making them a double departure from the Germanic norm. Five of the remaining

^^Ibid., p. 85.

Astra Desmond, Schumann Songs, BBC Music Guide No. 22 (Seattle: University of Washington Press, 1972), p. 63. 37 six in this category are by the famous Danish poet Hans Christian Andersen, four of which were translated by Chamisso and one anonymously. The last of Schumann’s choices in this category is the anonymous Old Catholic poem from the Latin (for the song “Requiem,” op. 90 no. 7) mentioned in the section on his undated choices.

A number of other poems which Schumann used are indirectly related to this group of non-native texts. Two of Chamisso’s poems among Schumann’s settings are German poems after the Frenchman Pierre Déranger (op. 31 nos. 2 and 3), and another (“Venatene

Liebe,” op. 40 no. 5) is based on a Greek poem. The works Schumann set by Geibel, too, are based on non-Germanic verses, such as the Spanisches Liederspiel, op. 74, which featured texts after Spanish poetry. These songs and their texts are considered influential in

Hugo W olfs later use of translated Spanish (and Italian) verses, which will be discussed later in the section on Wolfs poetry choices. The use of non-native language texts, while not common among Schumann’s songs, was nevertheless a viable, important, and influential part of his poetry selections.

Schumann’s Personal Contact with the Poets

Schumann was in contact with 22 of his poets, marked with an asterisk in Table 14, pp. 236-37. This number is confirmed by the names in Eric Sams’s list of Schumann’s song text authors labelled as “poets known to Schumann whether in person or in correspondence.”'** Julius Komer, counted in Sams’ list, was a translator of some of

Byron’s poems which Schumann used. These translations are counted among Byron’s texts in this study, and Komer is therefore not listed or counted here.

All but two (I. Kemer and Goethe) of the most extensively selected authors (use of more than 10 texts) were known to Schumann. He met Heine in Munich in 1828, but

“.. .the relationship made during Schumami’s short visit remained but a casual

'**Eric Sams, “Appendix III: The Sources,” Tie Songs o f Robert Schumann (London: Eulenburg, 1975), 2nd edition, pp. 278-281. 38 acquaintanceship. Many of the other authors listed here as known to Schumann were relatively unimportant in his life; he knew them primarily through correspondence. A number of the more obscure authors were discussed earlier in the “Undated sources” section, as was the poet “von der Neun” in the section on younger poets.

Schumann’s relationship with two of the others involves the trials over his opera

Genoveva. Friedrich Hebbel, first of the two, had helped write the original story, while the libretto was then developed by the poet Robert Reinick. ReinicUs version did not suit

Schumann, who then asked Hebbel to assist in rewriting the libretto. Hebbel declined, so

Schumann wrote the final version himself.

Schumann had a more pleasant relationship with Hans Christian Andersen (mentioned earlier under “Non-native sources”), who shared an evening in late 1840 or early 1841 with the recently married Robert and Clara in their new lodgings in Leipzig. Ultimately

Schumann’s poets of choice, unlike Schubert’s (and other composers’ as will be seen later), were not the closest people to him in his life. Schumann’s level of personal contact with his authors ranks the highest in this study. Other composers in this document may have had fewer contacts, but they were often more intimate and significant than

Schumann’s.

‘^^Hscher-Dieskau, op. cit., p. 22. 39

JOHANNES BRAHMS

Brahms’s Educational and Literary Background

Johannes Brahms (1833-1897) was bom into poverty. He developed an early love of reading, as an escape from the squalidness around him. His boyhood home was

Hamburg, an important city for bookselling at the time. This was an asset to the young

Brahms, who “.. .began to collect a small library of his own—largely bought from

Hamburg’s second-hand book stalls—which he treasured as much as his tin soldiers.”5° It is also reported that his music teacher Eduard Marxsen “.. .stimulated the boy’s passion for reading serious literature by lending him a steady flow ofbooks.”^' He immersed himself in world literature, as evidenced by a personal notebook which he compiled for himself.

Brahms called this collection of his favorite passages Des jungen Kreislers Schatzkastlein, deriving its name from the fantasy, KaterMurr, by E.T.A. Hoffmann. He even borrowed the name of the central character of the fantasy, Kreisler (an idealistic musician), signing some of his early works simply “Johannes Kreisler, Jr.” This collection is further described by Stohren

The 645 excerpts which Brahms copied into his Schatzkastlein represent the great names of world literature. From the ancient world he included Sophocles, Diogenes, Cicero, and Horace; among later authors are Dante, Tasso, Shakespeare, Swift, Pope, and Diderot. German writers naturally are the most prominent, and here too the scope of Brahms’ reading is impressive: Luther, Kleist, Klopstock, and Holderlin; Herder, Goethe, and Schiller, and finally the full heritage of the Romantics, above all, Jean Paul (Johann Paul Friedrich Richter). The Schatzkastlein contains 124 entries from this thoroughly romantic author, as compared with 68 by Goethe! Noteworthy also is the fact that while Brahms included in his collection practically every

^°Joan Chissell, Brahms, The Great Composer Series (London: Faber and Faber, 1977), p. 10.

5’-Kathleen Dale, Brahms: A biography with a survey o f books, editions & recordings (Hamden, CT: Archon Books, 1970), p. 9. 40 great writer available to him, he never used any of the poems in the Schatzkastlein for his solo songs.52

His formal education began at age six in a small private school, and after five years he moved on to a more progressive grammar school. His school education was completed during the 1847-48 academic year, at age fifteen. A few years later he began his travels as a musicizm, never returning to formal academic study. His only other notable encounters with academia were the several honorary degrees bestowed upon him, from Cambridge

University in England (which he twice refused) and from the University of Breslau.

One important feature of Brahms’s scholarly life continued past his school days: his collection of books eventually grew into an impressive personal library. This library was essential to his informal education, much like Schubert’s friends were for him, for it provided Brahms with a large resource from which to draw information (and poetry) for his songs. Musgrave describes its importance: “Brahms’s library is not merely a collection of books reflecting special enthusiasms, but a body of writings on many subjects which reflect the need to be thoroughly acquainted with the world of ideas, artistic and otherwise, present as well as past. ”53 His library contained many volumes of poetry, especially books of folk poems which were at the heart of his vocal writing. It was mainly from these books that he chose his song texts, along with the occasional verses he received from friends and acquaintances.

Brahms was not an educated man in the formal sense, yet he was certainly well-read, and his strong literary background evolved concurrently with his ever-growing book collection.

5^ Sister Mary Baptist Stohrer, The Selection and Setting o f Poetry in the Solo Songs o f Johannes Brahms (Ph.D. dissertation. The University of Wisconsin, 1974; Ann Arbor, Michigan: University Microfilms, 74-18,958), pp. 16-17.

53Michaei Musgrave, “The cultural world of Brahms,” Brahms: biographical documentary and analytical studies, ed. Robert Pascall (Cambridge: Cambridge University Press, 1983), p. 4. 41 Brahms’s Poetry Choices

It was not the great literary works that Brahms had read and copied down in his youth

which inspired his songs. These he revered only for their artistic worth, not for their use

as song texts. For his songs he turned to poetry from which he could create music more

easily: “.. .Brahms generally preferred to leave the very greatest poetry untouched, to

appreciate it solely as literature, and to choose lesser works out of which he could freely

fashion his own synthesis. In no instance would the ‘beautiful poem’ leave his mind: it

was the more modest one which needed music to make it memorable.

The largest single type of lext Brahrris used was folk poetry. It is definitely a modest

type of verse, well-suited to musical setting: “Folk song poetry fits itself to the music; it is

predictable, and extremely regular in meter, and it can be sung to many different tunes so

long as the rhythm of the words remains undisturbed.’’^^ This was not as crucial to

Brahms in his use of folk poetry as was his general liking for it, and his sense of preserving a cultural heritage by setting it to music.

The rest of his poetry choices were a mixture of both well-known and obscure authors.

Georg Friedrich Daumer was the single poet he set most often ( 19 times); today Daumer is

considered a second-rate poet. Eight of the Brahms songs on Daumefs texts were published together in op. 57, the largest group of his Daumer songs. ^6 Five more Daumer settings are poems “after Hafiz,” and these are detailed later in the paper under the “Non- native Languages” category. The other six are interspersed among several other opus numbers: op. 32 (no. 2), op. 46 (nos. 1 and 2), op. 59 (no. 6), op. 95 (no. 7), and op. 96

(no. 2). These poems are full of the moods and scenes Brahms felt would best associate

^‘'stohrer, op. cit.,p. 42.

^^Ibid., p. 23.

®®The opus numbers given for works by Brahms are those under which the songs were originally published. The main source consulted for this document was the works list by Heinz Becker, “Brahms, Johannes," The New Grove Dictionary o f MusicandMusicians, cà. Stanley Sadie, 20 vols. (London: Macmillan, 1980), III, pp. 174-185. 42 with the music he wanted to write at a given time. As Sams puts it, “Thus he chooses texts as a preacher might-r-less for their intrinsic merit than for relevance to a given topic or m o o d , ” 5 7 This is most applicable to his frequent use of Daumer’s verses.

Ludwig Tieck was the only first-rate author from whom Brahms selected a large number of texts: the 15 for the song cycle Romances from Tieck’s Magalone, op. 33. The lyrical intermezzi from Tieck’s novel Liebesgeschichte der schonen Magelone und des

Crafen Peter von A y venceappealed to Brahms for several reasons. First, the novel is a

German romance based on a French medieval tale, a combination which “suited Brahms’s taste. ”5* Second, Brahms saw much of himself in the character Count Peter, the hero of the story. Both left home against their parents’ wishes, both were articulate mainly through their art (lyrics for Count Peter, music for Brahms), and they were both known for their physical prowess. Brahms also identified with Count Peter’s wandering and homelessness, his unrecognized rank and merits, and the two very different women he loved, the virtuous Magalone and the exotic Sulima. Third, the incidental verses in the novel were the only way he could translate the novel into the L/edform. The song cycle is not, as a whole, considered his best work, but the choice of author, Brahms’s personal associations with the text, and the fact that it is his only song cycle give it and maintain its distinction among his vocal compositions.

Poems by Klaus Groth and Karl Lemke were also frequently chosen for his songs.

Groth was a close friend of the composer, and many of Groth’s books were in his library.

Storher states that “Groth is of particular importance as a poet because he was the first to raise Low German dialect to the status of a genuine literary vehicle.”^^ Brahms discovered

Groth’s poetry in 1873, when he responded to its themes of love and nostalgia. Several of

^^Eric Sams, Brahms Songs (Seattle: Univ. of Washington Press, 1972), p. 9.

^*Eric Sams, ibid., p. 25.

^^Stohrer, op. cit.. Appendix I, p. 207. 43 the eleven Groth settings are among his finest and most popular, such as “Heimweh II” (O wüsst ich doch den Weg zuriick), op. 63 no. 8 and “Wie Melodien zieht es mir,” op. 105 no. 1. Groth’s poem “Regenlied” was the only poem Brahms set twice; one was published as op. 59 no. 3 and the other without opus number was published posthumously in 1908.

O f the seven songs on Lemke’s poems, only “Im Waldeseinsamkeit,” op. 85 no. 6 is highly regarded, particularly for its “... forest evocations which communicate perhaps the most Germanic of all Romantic moods’’^® and for its harmonic content, “.. .where the emotional modulations of the middle part contrast with serene harmonies at the close.’’^*

Smaller groups of texts by at least eight important authors are found in his songs:

Heine (6 texts), Eichendorff (5 texts), Goethe (5 texts), Uhland (4 texts), Morike (2 texts),

Liliencron (2 texts), Riickert (I text), and Storm (1 text). The first three of these choices

(plus Riickert) reflect the influences of Schubert and Schumann; the Morike choices look toward Wolfs later monumental Morike song collection. Two of the most famous of his settings from these authors are “Der Schmied,” op. 19 no. 4, from Uhland and “Der Tod, das ist die kiihle Nacht,” op. 96 no. 1, from Heine. His singular setting of a poem by

Theodor Storm was “Über die Haide,” op. 86 no. 4. Friedlaender commented on this single choice: “It is strange that Brahms only set to music this one poem (but that a pearl among modem German lyrics) by Storm, although the latter came from the same part of the country as the musician and was in many respects a kindred soul.”62 Brahms knew and occasionally used the poetry of important writers for his songs, but it was not the type of verse which most suited his particular style of songwriting.

Brahms was sometimes led to completely obscure authors in his search for texts to meet his needs. The early song “Treue Liebe,” op. 7 no. 1, is a setting of a text by the short-

®°Max Harrison, The Lieder o f Brahms {New York: Praeger, 1972), p. 23.

Ibid., p. 35.

^-max Friedlaender, Brahms’s Lieder: An Introduction to the Songs for One or Two Voices, tr. C. Leonard Leese (London: Oxford University Press, 1928), p. 152. 44 lived and obscure poet Eduard Schulz, who used the pen-name “Eduard Ferrand.” The sensual love poem which Brahms dared to set, “O komme, holde Sommemacht,” op. 58 no. 4 is by Melchior Grohe, an author so unknown that his death date remains unrecorded.

A third and final example is “Lied aus Iwan,” op. 3 no. 4, another early song, on a text by

Friederich Bodenstedt. Bodenstedt’s verses were popular in their day but eventually waned.

Brahms selected texts from the Bible, as translated by Martin Luther, for his final vocal works, the Vieremste Gesange (FourSerious Songs), op. 121. The aging Brahms selected these three Old Testament verses and one New Testament verse on the subjects of death and love for setting as a tribute to ill and departed friends, especially his beloved

Clara Schumann. He chose these verses for their direct expressions of emotion. “Luther’s

Bible.. .spoke to him directly. It has a stark simplicity denied to rhythmically more complex forms. It rings even truer than folk song itsr if; it blends the direct speech of two peoples, German and J e w i s h . ’’^3 These texts fall into several of the classifications defined

in this study, and will be mentioned later within those contexts.

^^Sams, Brahms Songs, op. cit., p. 64. 45

Table 3. The Songs of Johannes Brahms

Description ______Number % Total number of Songs 201

Total number of Sources 47

Arithmetic average number of songs per source: 4.28

Median number of songs per source: 2.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 112 55.72%

Texts from poets contemporary with composer (bom within 15 years of composer) 29 14.43%

Texts from poets younger than composer (bom 15 years after composer) 9 4.48%

Texts from undated sources (dates of poet or source unknown) 39 19.40%

Texts from old sources (dated more than 100 years before Hîrffî^ 1 x 5 .9 7 %

Texts from non-native sources (either in translations or in other languages) 30 14.92%

Approximate number of poets with whom the composer had personal contact: 7 46 Brahms’s Number of Songs and Sources

The total number of songs by Brahms included in this study is 201. Two primary

sources were consulted and compared in determining this number: Becker’s list of works

by Brahms in The New Grove Dictionaiÿ>^ and A. Craig Bell’s book on the Lieder of

Brahms.^^ Several other lists of his songs were also consulted to confirm the survey.

Only his two songs of op. 91, with viola, are generally included in the lists but not

included here. Brahms’s songs are easier to document than Schubert’s and Schumann’s,

because there are fewer of them, more of them were published in his own lifetime, and

there are fewer questions about multiple versions, lost copies, etc., in spite of the difficulty

scholars have had dating the composition of individual songs.

The number of sources which Brahms used is totalled at 47 in this study. This figure

was also determined using the sources detailed above.^^ Stohrer’s list of songs by source

was of particular value in this area.^^ ^ complete list of these sources is found in Table 15,

pp. 238-39.

The average number of songs per source for Brahms is 4.28, based on the totals above

(201 songs, 47 sources). This average is in the upper mid-range among the twelve

composers studied, implying that it is neither extremely low nor extremely high. The

median average for Brahms is 2 songs per source, and the modal value is 1 song per

source, both relatively low numbers. Brahms chose more than ten texts from only four

sources (actually only three individual poets, plus the combined Anonymous sources), and

eighteen sources were used only once, a smaller percentage of single choices than for either

®^einz Becker, “Brahms, Johannes,” The New Cmve Dictionary o fMusic and Musicians, ed. Stanley Sadie, 20 vois. (London: Macmillan, 1980), III, pp. 174-185.

Craig Bell, The Lieder o f Brabms{Dailey, England: The Grian-Aig Press, 1979), Appendix 1, pp. vra-xn.

®%ecker, op. dr., and Bell, Appendix 11, pp. Xlll-XIV.

Stohrer, op. cit.. Appendix II, pp. 218-227. See also chart pp. 32-33. 47

Schubert or Schumann. In addition, Brahms used 2-7 poems from over half of his sources. All of these numbers indicate that Brahms was fairly uniform vnhXs number of texts used per source.

Brahms’s Authors Categorized by Birth Date

A breakdown of authors by birth years, shown in Figure 3, indicates that 112 texts were chosen from authors bom 16-100 years earlier than Brahms (between 1733 and

1817), 29 texts were chosen from authors bom 15 years before or after Branms (between

1818 and 1848), and 9 texts were chosen from authors bom more than 15 years later than

1833 (after 1848). These numbers stated in percentages are as follows: 55.72% were from authors bom 16-100 years before Brahms; 14.43% were from authors bom within 15 years (before or after Brahms); and 4.48% were from authors bom more than 15 years after him.

There are several explanations for this distribution of percentages. The first possible reason for his use of so many works from the older writers is that he began to read literature at an early age, and in the process leamed of the older, early Romantic poets long before he started writing the bulk of his songs. Stohrer says that, “This vast body of [early

Romantic] literature had a profound effect upon the young Brahms. He was thoroughly steeped in it, as well as a variety of earlier literature.”®* Another, related reason is that the lise of German lyric poetry peaked in the years before his birth. Stohrer comments on this:

“Brahms’s century was an unusually complex one from the literary point of view. By the time he was bom in 1833 German literature was second to none in the world.

In addition, there were the many people who influenced Brahms in his choices. His piano teacher, Marxsen, probably steered the young Brahms toward some of the older

®*Stohrer, ibid., p. 15.

®9/b/d.,p. 11. 48 poets. Later there were personal friends and professional colleagues who led him to poetry. An early example of this occurred in the summer of 1847, when Brahms spent time in the country with the Giesemann family. During this time Brahms and Lischen

Giesemaim, one year apart in age, became fast friends: “They shared a love of reading and he gave her piano lessons.. .with Lischen, Johannes made acquaintance with Tieck’s story of The beautiful Magelone and the Knight Peter, fifteen poems from which formed the basis of his song cycle Magelone Lieder, op. 33, composed in the 1860’s.”70 Another example, from the early 1870’s, is his use of Goethe’s “Dammerung senkte sich von oben,” op. .59 no. 1: “Brahms became acquainted with the poem through a composition by his friend Hermann Levi which stimulated him to write his.’’’’ Brahms even stole four bars of music from Levi’s song for his own setting!

The textual choices of song composers who preceded Brahms were also important to his own selections. Harrison writes that, “[mjany links can be pointed out between his songs, early and late, and those of other com posers.”’^ por example, he “.. .worshipped

Schubert and he knew the Austrian’s songs well....”’^ Important also was Schumarm,

“.. who influenced Brahms considerably in his early works.”’"* Harrison cites a number of links with Schumann, not just in the early songs.’^ Brahms had an intimate knowledge of

Schumann’s works, which included his choices for song texts; he also knew some of

Mendelssohn’s songs. The Romantic poetry these composers used certainly played a part in his choices.

’^Kathleen Dale, Brahms: A biography with a survey o f boohs, editions & recordings (Hamden, CT: Archon Books, 1970), p. 10.

’ ’Friedlaender, op. cit., p. 98.

’^Max Harrison, The Lieder o f Brahms (New York: Praeger, 1972), p. 126.

’ ^Kenneth Whitton, pp. ciY., p. 63.

’"^Friedlaender, op. cit., p. 3.

’ ^Harrison, op. cit., p. 127. 49 Beyond these considerations is a yet more intangible explanation for all of Brahms’s poetry selections: his need for personal expression. He sought out verses for their message and their close relation to his own inner feelings. The date or style of the poetry

(and many say the quality) were less important to him than were the words themselves as a vehicle for the emotion in the songs. This alone explmns many of his choices, from both older and contemporary poets.

The small group of nine verses Brahms used from younger poets are the works of four writers: one each from Adolf Frey and Max Kalbeck, three from Felix Schumann, and four from Hans Schmidt. Frey was a Swiss poet, and Brahms set “Meine Lieder,” op. 106 no. 4, in 1886. Kalbeck is best known today for his biography of Brahms; the setting of

Kalbeck’s “Nachtwandler” was published as op. 86 no. 3 in 1882. Felix Schumann was the last child born to Robert and Clara Schumann (Brahms was his godfather). The first of

Feli^^s three verses which he set in December 1873 is the famous “Meine Liebe ist Griin,” which Brahms titled “Junge Lieder I” and published as op. 63 no. 5. It was followed in op. 63 with another of Felix’s poems, “Weim un den Nollunder,” or “Junge Lieder II,” op.

63 no. 6, written in the summer of 1874. Friendlaender says that, “Brahms intended the composition as a Christmas present for the young poet and his mother.The third is

“Versunken,” op. 86 no. 5, written in December 1873. Brahms consciously derived his setting from Robert Schumann’s song style, for obvious reasons. Three of the four verses

Brahms set by Hans Schmidt are nos. 1-3 in his op. 84, published in 1882, and the fourth is the famous “Sapphische Ode,” op. 94 no. 4, published in 1884.

Several items must be mentioned about this group of texts from younger authors.

First, Brahms ranks very high in the twelve composers for this category (choosing works

from younger authors). Second, his long life allowed him the time to know these younger poets. He was in his 40’s at the time he used their verses. Third, he personally knew three

Friedlaender, op. cit., p. 110. 50

of the four writers, giving this group even further significance in his choices. Finally, two

of the nine settings in this group are rated with the best of his songs, which indicates that

the elder Brahms was quite capable of producing good settings from the younger poets’

texts, and that they were not just set through a sense of obligation to the authors.

Brahms’s Choices of Undated and Old Sources

Nearly a fifth of Brahms’s song settings use texts o f anonymous poetry. This is not unexpected, considering his status as one of the masters of the folk song setting. Not only did he write the 39 Lieder on folk texts counted in this study; he also compiled three large groups of folksong arrangements using the same type of texts, which are not counted in this study (the 14 “Folksongs for Children” of 1858, the 28 “German Folksongs” of 1858, and the 49 “German Folksongs” of 1894). There are two major differences between the 39

Lieder in this study and the latter volumes. Brahms composed new melodies for the Lieder in this study, while in the folksongs he merely wrote accompaniments for the original or traditional tunes. Additionally, the Lieder include many translated folk poems originating

from other countries, while the folksongs are all intentionally of German-language origin.

This last matter will be considered again under the “Non-native Languages” category below.

Brahms scattered his Lieder using anonymous texts among the other songs, a practice begun as early as the one in op. 6 of 1852 and ending with the one in his op. 105 in 1886.

Usually only one or two of them appear within a given Opus number, with three exceptions. They are the eight songs in op. 14, the seven of op. 48, and the nine of op.

69, each of which contains mostly settings of traditional and anonymous texts. His work in this area of setting and arranging folk materials is one of his greatest achievements, and it is certainly logical that his number of choices in this category is among the highest of the twelve composers studied. 51

It may also be logically expected that Brahms would use some old texts for his songs, given his extensive studies in ancient literature and music. The twelve settings in this group were taken from four diverse sources: the four Biblical texts of the VierEm ste

Gesange (Four Serious Songs), op. 121; one text translated by Herder from the thirteenth century poet Thibault, Count of Champagne, King of Navarre, “Ein Sonnett,” op. 14 no.

4; the “Five Songs of Ophelia” from Shakespeare’s Hamlet, translated by Tieck and

Schlegel, without opus number; and the two texts by the German Renaissance poet Paul

Flemming, for the well-known song “O liebliche Wangen,” op. 47 no. 4 and the lesser- known “An die Stolze,” op. 107 no. 1.

Additional texts are indirectly linked to this group. One is the set of “Romances from

Tieck’s M agelone," op. 33. In these poems the author Tieck, “...as a true Romantic, gazes longingly back to the distant spacious days of the old romantic Orient.”^^ Harrison calls them “.. .comparatively modem texts that looked back far into the past.”^® Some of the

Anonymous and Traditional texts Brahms used also have their roots in antiquity.

Friedlaender mentions this in his description of the famous “Wiegenlied” (Lullaby), op. 49 no. 4: “The poem is very old. Already at the end of the fifteenth century, a version of it was known as a love letter.”^9 Several other songs by Brahms have ties with the past, particularly his “Ich schell mein Horn in Jammertal,” op. 43 no. 3, “.. .which decidedly echoes the sixteenth century,”®° his “Entfiihrung,” op. 97 no. 3, with its tale of Lady

Judith’s kidnapping on the moors, and op. 71 no. 5, which has the ancient-sounding title o f“Mirmelied.”

^^Friedlaender, ibid., p. 37.

^®Harrison, op. cit., p. 5.

^^Friedlaender, op. cit., p. 80.

®°Harrison, op. cit., p. 3. 52 The percentage of poems Brahms selected from ancient authors ranks low among the composers studied. / His choices of texts which indirectly relate to this category, as well as his broad intellectual understanding of antiquity are not reflected in the low percentage.

The ancient world and its poetry were actually integral parts of his song writing endeavors.

Brahms’s Choices in Non-native Languages

There is a large carry-over from the last category (Undated and Old Sources) into this one for Brahms. Twenty of the 39 anonymous poems were translated from other languages. All but the two Flemming poems in the Old Sources group were translated into

German; the texts from Shakespeare, Thibaut, and the Bible. The total of 30 texts represents 14.92%, or a little over a seventh, of his total selections. This is a middle-range amount in the comparison of the twelve composers. Five additional poems are not included in the count but were derived from another language, the Daumer poems “after Hafiz,” an early Persian poet. Brahms set three as op. 32 nos. 7, 8, and 9. The last one (no. 9) is the famous “Wie bist du, meine Konigin." The other two of the five are in op. 47 (nos. i and 2). Op. 47 no. 1 is the famous “Botschaft.” Actually all of Daumer’s verses which

Brahms used are in some way influenced by non-Germanic poetry, especially those he took from Daumer’s books Hafîsanà Polydoia.

The folk poems in translation which Brahms set were from a wide range of original languages: Bohemian (Czech), English, Serbian (Serbo-Croation), Spanish, and Swabian

(an old dialect from southern Germany). His choices were translated by a number of scholars and writers. Herder, Heyse, Kapper, Kopisch, Schlegel, Tieck, Wenzig, and

Zuccalmaglio translated most of those he set. His use of these poems was tied less to his interest in the original languages and their cultures than to their basic appeal as texts in the translations. The fact that he set none of them in their original languages is evidence of this, as is the fact that the ones he chose were found in the widely read books of verse in 53 German translation available in his day. This is not to imply that he did not know about other languages: he travelled extensively, going to Italy many times, as well as to Holland and elsewhere on concert tours. Perhaps his early language studies were partly to blame: in school “.. .English always gave him trouble, French not so much, despite his execrable accent.”8i His love of things Germanic ultimately won out in the matter, however, and his

Lieder based on foreign-language texts (with the exception of the Four Serious Songs) are not the most important segment of his song compositions.

Brahms’s Personal Contact with the Poets

Only seven poets of songs texts by Brahms have been confirmed in this study as having been personally known to him. They are marked with an asterisk in Table 15, pp. 238-39. Brahms was known to have had many friends outside of music, but the music critic Eduard Hanslick and Joseph Widmann were his only close friends who were writers. Musgrave, however, mentions "...long relationships with Klaus Groth and Paul

H eyse.”*2 Groth was probably closer to him than any of his song text authors, for, as

Sams writes, “He was not only a personal friend of Brahms but came from the same part of

Germany; they both remembered the same boyhood.”®^ Sams also mentions Brahms’s friendship with Gottfried Keller, which “.. .may have predisposed him to “Abendregen”

[op. 70 no. 4, 1875].”®^ Brahms did not know the artist, historian, author, and composer

Franz Kugler, but he did know Kuglefs son and daughter-in-law in his later years. The

®*Joan Chissell, Brahms, The Great Composers Series (London: Faber and Faber, 1977), p. 10.

®^ichael Musgrave, “The cultural world of Brahms,” Brahms: biographical documentary and analytical studies, ed. Robert Pascall (Cambridge: Cambridge University Press, 1983), p. 3.

®®Sams, Brahms Songs, op. cit., p. 41.

Ibid., p. 43. 54 popular “Standchen,” op. 106 no. 1, uses a text from Kugler's Skizzenbuch of 1830, which was (according to Friedlaender) “one of Brahms’s favorite w orks.”®^

The younger poets (Felix Schumann, Hans Schmidt, and Max Kalbeck) described earlier were three more poets he did know personally. Sams concisely describes the

Brahms^Kalbeck relationship; “It was in 1874 that Brahms first met Max Kalbeck, whose lifelong devotion later expressed itself in the best-documented biography. Brahms in turn took a benevolent interest in the young man’s unpublished verses.”®®

He had one further meeting with a poet: “Brahms went to some length to meet the aging Georg Friedrich Daumer.”®^ Friedlaender describes the meeting: “On the way from

Nuremberg to Karlsruhe early in May 1872, he broke his journey at Wiiizburg. After some trouble he found the street and the house, and was considerably surprised when a little shrivelled old man was introduced to him as the German Hafis.”®® This relationship, like Schumann’s with Heine, was short-lived.

Brahms, it may be concluded, was not like those song composers who took an active part in the literary world of their time, taking from it and shaping it in the process, as is the case with several of the composers in this paper. The size of his personal library and his relatively small contact with the poets from whom he chose texts would indicate, rather, that Brahms relied on a less personal means for knowing and using poetiy. Schauffler adds to this argument by citing Brahms’s North German shyness:

This was so painfully intense that it doubtless cost him many a struggle to associate himself, as publicly as a song-writer must, with the poets and their passionate, intimate confessions of love, joy, or sadness. We have.. .seen how impossible it was for this extraordinarily self-contained man to confide his most private thoughts to even his best friends. These lyric struggles must

®®Friedlaender, op. cit., p. 182.

®®Sams, Brahms Songs, op. cit., p. 44.

®^Musgrave, op. cit., p. 3.

®®Friedlaender, op. cit., p. 32. 55 often have made his soul involuntarily shrink back into its shell, in a way which absolute music, with its greater soul-privacy, could never do.®®

Brahms, through his songs, was looking for a way to express publicly his private emotions. This, it seems, is the overriding factor in his choice of texts, in the style and age of the poetry, and in his general lack of closeness with the poets themselves.

®®Robert Haven Schauffler, The Unknown Brahms (New Yoik: Dodd, Mead, & Co., 1936), pp. 331-332. 56

HUGO WOLF

W olfs Educational and Literary Background

Hugo Wolf ( 1860-1903) had an incomplete formal education. He spent four years in elementary school in his Styrian home town of Windischgraz, one term at the Gymnasium in Graz, one year at the seminary school of St. Paul, Corinthias, and two years at the

Gymnasium at Marburg. He was advised to withdraw from the St. Paul school because of poor grades, and for the same reason "...was not a success at Marburg.”^® His restless personality and strong focus on music prevented him from completing any education except elementary school, although he did gain a wide appreciation of literature in these early years. In 1875, at the age of fifteen. Wolf went with his aunt to Vienna, to study at the

Vienna Conservatoire, which also eventually proved to be an unsuccessful venture. During this time he began composing his early works, discovered opera, began his following of

Wagner, and leamed of the works of earlier musical and literary masters in the library of the Conservatoire.

After a time of floundering in Vienna and leaving his Conservatory training. Wolf was forced by financial troubles to go to his home town, but he finally returned to Vienna in

1877, where he was taken under the wing of Dr. Adalbert von Goldschmidt. From the

Goldschmidt household Wolf was free to borrow books and scores to take to his own quarters for study, furthering his knowledge and love of literature. Further evidence of his strong love of reading dates from his idyllic stay in Maierling during the summer of 1880:

At Maierling he had.. .plenty of opportunity to indulge his passion for reading: Viktor Preyss [his host] kept him supplied with books from the city. Wolf is said to have busied himself with Wagner’s prose writings and the philosophy of Schopenhauer, although in the latter subject there is no evidence that he had got beyond the lighter reading of the Parerga und Paralipomena. Goethe’s Faust, Gottfried Keller, especially Dergriine Heinrich, Heine’s poems and

®°Frank Walker, Hugo Wolf: A Biography (New York: A. Knopf, 1952), p. 13. 57 Reisebilder, Lenau, Kleist, Immermaiin's Miinchhausen and Tulifantchen, E. T. A. Hoffmann, and Jean Paul—these names and titles give some idea of his tastes in German literature. Among foreign writers he discovered and liked Turgenev, Lermontov, Leopardi, Scott, Byron, Mark Twain, Dickens, Sterne, Molière, Rabelais, Mérimée, and Claude Tillieris M on Oncle Benjamin. We notice here, besides romantic tendencies, a pronounced taste for the masterpieces of humour.^'

His literary understanding was indeed exceptional.

After several indecisive years, Wolf became the music critic of the Vienna weekly

Salonblatt in 1884, a position he held for about three years. Frank Walker, in the biography quoted above, dismissed this period of Wolfs life as insignificant to his development as an artist. Eric Sams, however, believed that "...regular reviewing and concert-going gave him much-needed discipline and experience, as well as a new understanding of the nature of language and its relation to music, including his own.”^^

Wolf ended his career as a music critic in early 1887, and soon after began his brief years of monumental creative efforts in song composition, from 1888 to 1891. He had received little in the way of education, yet today he is considered one of the composers most concerned with the literary side of the Lied, one who was aware of and used only the best textual materials available, and one who used an “intellectual” approach to song composition.

W olfs Poetry Choices

Hugo Wolf experimented with song composition as early as 1875, and did not reach his maturity in vocal writing until the late 1880’s. Most of the songs he created in the early years were not published during his lifetime, and the poetry choices during this time were part of the experimentation. Some of the poets for the early songs were by virtually

91/6/d., p. 104.

9^Eric Sams, “Wolf, Hugo,” The New Grove Dictionary o f Music and Musicians, 20 vois., ed. Stanley Sadie (London: Macmillan, 1980), XX, p. 492. 58 unknown writers, such as Schokke, Herlossohn, and W olfs friend Paul Peitl. These writers were quickly abandoned.

Poems by more established writers, such as Matthison, Riickert, Chamisso, Hebbel, and Reinick were set along with those by obscure poets, but Wolf did not continue to choose texts from these authors in his mature period. Lenau ( 12 texts) and Heine ( 17 texts) were the two poets Wolf chose most often for his early songs, but they too were dismissed in later years (except for one published Heine setting written in 1888, “Wo wird einst”).

Only the six Goethe and the two Morike settings among the earliest songs anticipate the later, important collections from these authors. The complete list and tabulation of Wolf s selctions by sources is found in Table 16, page 240.

The first and largest collection of songs dedicated to a single poet was written in 1888, the Gedichte von Eduard Morike (the Morike songs). Regarding this collaboration, Camer writes that, “The conjunction Morike-Wolf seems to have been a case of elective afiinity felt on the part of the composer, whose fifty-three settings (published in 1889) represent one of the high points of his creative c a r e e r . ”^3 "Wolf with these songs began a process of investigating, through musical settings, the range of ideas expressed by a given poet.

Sams believes that “...the poems chosen are fully representative of each aspect of Morike’s art. The songbook as a whole is unique for its absorption of the essence of one great poefs work into music of a comparable quality.”^'* Wolf was generally against setting texts he felt had already been set successfully by other composers, yet Morike had been largely ignored (e.g., Schubert set none of Morike’s texts, Schumann set only five, and

Brahms only two), leaving Morike’s poetry to become a large part of Wolfs songwriting legacy.

^^Mosco Camer, Hugo W olf Songs (London: British Broadcasting Corp., 1982), p. 21.

^^Eric Sams, The Songs o f Hugo Wolf (New York: Oxford University Press, 1962), p. 35. 59

Wolf published a set of songs using twenty poems by Eichendorff in 1889. A majority

of these were composed in September of 1888, after producing most of the Morike songs.

Sams explains the choice of Eichendorff s atypical verses: “Wolf may well have thought

that no one could improve on Schumann’s sensitive settings o f ‘typical’ Eichendorff; or he

may have felt the need for some relaxation after the Morike outburst.”^^ Many of the

Eichendorff choices show Wolfs lighter compositional side, in poetry about soldiers,

minstrels, and students, while some of the settings show the more romantic side of both

poet and composer. Camer claims that: “The acknowledged masterpiece among these

romantic mood pictures is ‘Das Standchen,’ a song about the remembrance of things

past.”96 W olf later considered Eichendorff s poetry superficial, but his refreshing settings

of these verses proved to be a worthwhile diversion amid the grander volumes from the

poetry of Morike and Goethe.

W olfs Goethe settings date mainly from late 1888 and early 1889. They were

published as a collection of 51 songs in 1890. The first ten songs are on texts from

Goethe’s novel W ilhelm Meister. Wolf challenged the Goethe settings of previous

composers and intended to improve upon them with his own songs (whether he did or not

is a matter of opinion). Another group of texts in this songbook were taken from Goethe’s

late poetry collection Westostlicher Divan, poems in praise of wine, poetry, and love,

based on the works of the ancient Persian poet Hafiz (mentioned earlier in connection with

Brahms’s settings by Daumer also “after Hafiz”). The last three songs in the songbook

form a small sub-group of poems about man’s relationship to the gods (“Prometheus,”

“Ganymed,” and “Grenzen der Menschheit”), which were also settings in response to

earlier versions by other composers. The rest of his Goethe choices are a miscellany of

verses, often somewhat weak, for “...in his search for Goethe poems not previously set by

^^Sams, ibid., p. 97.

^^Camer, op. cit., p. 32. 6 0

Schubert and Schumann, Wolf was driven to look farther afield, selecting verses that did not naturally lend themselves to musical treatment. Wolfs settings of Goethe, while usually not as highly esteemed as his Morike settings, are nevertheless larger in number

(when the early Goethe settings are counted) and as important both historically and in terms of textual choices as any of his selections. W olfs settings of Goethe’s poetry marked the culmination of Lied composers in the use of this great author’s verses.

After Goethe, Wolf found inspiration in the translations of Spanish and Italian poetry, for his Spanisches Liederbuch (Spanish Songbook) and his Italienisches Liedeibuch

(Italian Songbook). Camer gives two reasons for these selections: “The first and general is that Wolf, alone among the great German songwriters, was most strongly attracted to the spirit and landscape of a Mediterranean of the mind.... Secondly, with his Goethe settings

Wolf had, as it were, used up German poetry of the quality he demanded of his poets and, probably spurred on by Schumann’s and Brahms’s examples, turned to these translations... .”98 The Spanish and Italian verses were translations by Heyse and Geibel, and the song collections were both divided into two parts. The 44 Spanish songs are split into 10 Geistliche (Sacred) and 34 Weltliche (Secular) songs, and the 46 Italian songs were composed and published in two volumes about five years apart, with 22 songs in volume 1 and 24 songs in volume 2.

Wolfs remaining poetry choices were less extensive, i.e., he chose only a few poems from the other writers, and most of them were relatively insignificant choices. His six songs from Gottfried Keller’s Alte Weisen (Old Tunes)were composed in 1890, between the Spanish and Italian songbooks, and were originally conceived as a tribute to the Swiss poefs seventieth birthday, but the poet died before they were finished. Keller was one of

Wolfs favorite novelists, but these settings gave him difficulties and are not considered

9^Camer, ibid., p. 42.

98Camer, ibid., p. 43. 61 some of his best work. Nor are the three Reinick settings considered great; they were composed in 1888,4896, and 1889 respectively. Wolfs last seven songs are more highly regarded, and the authors for these are Heine (1 song, 1888), and translations of

Shakespeare (1 song, 1889), Lord Byron (2 songs, 1896), and Michelangelo (3 songs,

1897). All but the Heine choice will be discussed later, in the “Old Sources” and “Non­ native Language” categories. 6 2

Table 4. The Songs of Hugo Wolf

Description Number % Total number of Songs 313

Total number of Sources 27

Arithmetic average number of songs per source: 11.59

Median number of songs per source: 2.50

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 155 49.52%

Texts from poets contemporary with composer (bom within 15 years of composer) 1 0.32%

Texts from poets younger than composer (bom 15 years after composer) 0 0 .00%

Texts from undated sources (dates of poet or source unknown) 84 26.84%

Texts from old sources (dated more than 100 years before composer's birth) 73 23.32%

Texts from non-native sources (either in translations or in other languages) 96 30.67%

Approximate number of poets with whom the composer had personal contact: 1 63

W olfs Number of Songs and Sources

Two groups of Wolfs songs must be considered in determining his total number for

this study: those published by the composer during his lifetime and those which are

unpublished or were published posthumously. The Works List by Eric Sams at the end of

the article on Wolf in The New Grove Dictionar^'^ was used as the primary source in

figuring the totals for both groups. This list was then compared to other (less

comprehensive) catalogues of his Lieder. The published songs number 742, and the

unpublished number 71, making a total of 313 songs counted for this document. The three

songs Wolf wrote as incidental music to Ibsen’s play Das Fest auf Solhaugaxe not included

from the published works, and Sams’ list of 103 unpublished songs contains 32 lost,

incomplete, or other works not countable as outlined in the Introduction.

The 27 sources counted for Wolf were also derived mainly from Sams’s list cited

above. The sources are listed by name in Table 16, page 240. There are several special

features in this list. First, eleven of the sources were used only for early and unpublished

songs. Second, the Spanish and Italian Songbooks are listed as individual sources, not

with the Anonymous sources. Third, the Spanish authors identified among Wolfs

Spanish Songbook settings are not listed separately but as part of the complete collection.

These last two items were decided upon after much thought, but the overriding factor in

both cases is the idea that Wolfs intentions were to choose and set these verses as complete

entities, rather than as random, ungrouped choices. More discussion will follow on these

two collections, in the sections on W olfs Undated, Old, and Non-native Languages

choices, and the Spanish poets will be listed by name in the “Old Sources” category.

Wolfs average number of songs per source is 11.59 (based on 313 songs from 27

sources). This is the highest average for any of the twelve composers (see Figure 36,

p. 220). His median average is also fairly high at 2.5 songs per source. His modal value

^®Eric Sams, “Wolf, Hugo,” The New Grove Dictionary o f Music and Musicians, ed. Stanley Sadie, 20 vols. (London: Macmillan, 1980), XX, pp. 492-498. 64 is one song per source, but this a somewhat weak average, since Wolf chose single texts from only ten ofhi&sources. These numbers clearly reflect Wolfs practice of choosing a large group of texts from a given poet, in order to make them into collections of songs showing the author in a variety of moods. This practice was employed primarily in his mature songwriting period. The evidence of this is that eight of the ten poets which he used only one time are from his early, unpublished song settings, and that all of the large selections of poetry are from the later period.

W olfs Authors Categorized by Birth Date

Wolfs textual choices are categorized by the birth year of their authors relative to

Wolfs birth year of 1860 in Table 4: 155 texts were chosen from authors bom 16-100 years earlier than W olf (between 1760 and 1844), one text was from an author bom within

15 years of Wolf (between 1845 and 1875), and no texts were from authors bom 15 years after Wolf (after 1875). Stated in percentages, 49.52% of his choices were by authors bom between 1760 and 1844, 0.32% were by authors bom between 1845 and 1875, and none were by authors bom after 1875.

This one-sided distribution is explained by Wolfs eventual and overpowering desire to set only quality poetry, virtually all of which (for him) came from the masters of lyric poetry bom in the years before him. For example, his three most frequently chosen authors (not counting the Spanish and Italian Songbooks), Goethe, Morike, and

Eichendorff were bom 111, 56, and 72 years before him respectively. This means that

Goethe is not even in the “past generations” group (he must be counted with the ancient sources for Wolf!). Further, two of the three (Goethe and Eichendorff) were dead before

Wolf was bom, and Morike died when Wolf was a teenager. Additionally, the author from the “past generations” group closest in age to Wolf was Joseph von Scheffel, who was bom in 1826, over 30 years before the composer. The only exception to his choices from 65 older poets was the one text he chose from his friend and contemporary Paul Peitl, for an early unpublished song composed in 1876, “EinGrab.”

Other reasons for the one-sided distribution of ages in Wolfs selections come from external influences upon the composer. Wolf knew at an early age the Lieder of earlier masters (and thus their choices of poetry), Newman states that, of the works Wolf studied immediately after leaving the Vienna Conservatory as a teenager, it was “Schumann’s songs in particular he examined most carefully and m i n u t e l y . ” **^ He also knew the music of other composers, including the song settings of Schubert, and as a music critic he wrote analyses of Mozart, Gluck, Schubert, and others. The poetry choices of these composers, many of which were from earlier poets, undoubtedly helped bring Wolf to his own decisions about authors. It is also important to realize, as Wolf must have, that the peak period for writing great German lyric poetry had preceded him by quite a few years, and that a major portion of his achievement in song composition was founded on the use of texts from these earlier authors.

WolTs Choices of Undated and Old Sources

There are 84 undated or anonymous texts which W olf used, representing 26.84% of his total choices. This category is divided into three parts. The first is small; three unknown texts and one anonymous text are found among his miscellaneous early songs:

Title______Author ______Date______1. Der Goldene Morgen Unknown 1876 2. Bescheidene Liebe Unknown 71876-7 3. Wanderlied Anonymous 1877 4. Morgentau Unknown 1877

*°®Emest Newman, Hugo W olf (New York: Dover, 1907, 1966 reprint), p. 15. 66

The second and much larger part of this category is the group of 34 texts from the

Spanisches Liederbuch which are by anonymous authors or authors of unknown dates.

Three of the ten Geistliche (Sacred) Lieder are by anonymous authors (nos. 5, 8, and 9)

and five more are by early Spanish poets about whom little or nothing is known:

Author Song No. Song Title 1. Juan Ruiz 1 Nun bin ich dein 2. Nicholas Nunez 2 Die du Gott gebarst, du Reine 3. Francisco de Ocana 3 Nun wandre, Maria 4. Lopez de Ubeda 6 Ach, des Knaben Augen 5. Don Manuel del Rio 7 Miihvoll komm ich und beladen

The other two Geistliche Lieder are by authors whose dates are known, and will be described below. A large majority (24 of 34 songs) of the Weitiiche (Secular) Lieder in this collection are by anonymous authors (nos. 2-8, 10-13, 15-21, 25-26, 28, 30-32, and 34). Two of the Weltliche Lieder are by obscure early Spanish poets:

Author Song No. Song Title 1. Alvaro Fernandez de Almeida 1 Klinge, klinge mein Pandero 2. Don Luis el Chico 31 Deine Mutter, susses Kind

The other eight Weltliche Lieder are by known authors and will also be discussed below (under Old Sources). The 46 anonymous Italian texts of the Italienisches

Liederbuch (Italian Songbook) complete this category of Undated texts.

Wolfs choices of old poetry are recorded here at 73 texts, or 23.32% of his total choices. This is an incomplete and misleading figure, for it actually represents only those texts which have been established by the date of the authors’ birth years (as outlined in the precepts for analysis in this document, in the Introduction). If the 46 texts from the 67

Italienisches Liederbuch and the 34 from the Spanisches Liedeibuchy/ere added to this category, a more cmnplete number of 153 texts in the Old Sources (48.88% of the total selections) would be the result.

The texts from the two songbooks, while technically undated, are actually known to have originated in earlier times (fifteenth through seventeenth centuries), and could therefore be included in this category, even though they are not datable specifically by authors' births. The types of Italian texts which Wolf used, for example, were “.. .regional variants of the courtly classical strambotto, dating back to the fifteenth century if not earlier.”

Eleven authors are represented in the 73 dated ancient texts; Michelangelo,

Shakespeare, Goethe, and eight early Spanish poets. Both the three Michelangelo settings and the one Shakespeare setting were published late (1897-98), although the Shakespeare setting, “Lied des transferierten Zettel” (a translation of Bottom’s song from A Midsummer

Night’s Dream) was written in 1889. The three Michelangelo Liederwere Wolfs last

songs: he “.. .completed the Michelangelo settings in March 1897. Before the year was

out, his mind sank slowly into growing night.” 102 Wolf himself noted their ancient

quality, in a letter to his friend Hugo Faisst, writing that “.. .they are truly antique, as far as

can be in modem music.”

Most of the 59 Goethe poems were set in 1888-89. Texts by Goethe form his largest

selection from any author, ancient or otherwise. Wolf was compelled to work with the best

of German poetry, and that certainly brought him to Goethe’s poems, no matter that by the

time he got to them they were century-old classics. He was further led to work with

Goethe’s verses by studying other previous settings (e.g., by Schubert and Schumarm), as

*°*Mosco Camer, op. cit., p. 56.

p. 70.

*°3Quoted in Walker, op. cit., p. 408. 68 mentioned above, either attempting to improve upon them or set those not previously used.

With his songs Wolf had brought the age of Goethe settings to a climax and close, much as he had done with other types of lyric poetry.

Ten remaining songs in the Old Sources category are by eight Renaissance and Baroque

Spanish poets whose dates are known, all bom over 100 years before Wolf, and all included in the Spanisches Liederbuch. They are:

Author Dates Song No. Incioit 1. Lope de Vega 1562-1635 4 Sacred Die ihr schwebet 2. José de Valdivivielso 1560-1638 10 Sacred Wunden tragst du... 3. Rodrigo Cota C.1405-C.1470 9 Secular Blindes Schauen... 4. Miguel de Cervantes 1547-1616 14 Secular Kopfchen, Kopfchen... 5. Ciistôbal de Castillejo 1490-1550 22 Secular Dereinst, dereinst... 6. Luis Vas de Camoens 1524-1580 23 Secular Teifim Herzen... 7. Joan Escriva C.1450-C.1520 24 Secular Komm, 0 Tod... 8. Gil Vicente C.1465-C.1536 27 Secular Und schlafst du... 9. Lope de Vega 1562-1635 29 Secular Weint iiicht... 10. Gil Vicente C.1465-C.1536 33 Secular Wehe der...

Camoens was actually Portuguese but wrote verses in Spanish. He, with Cervantes and de Vega, are the only great poets of the ten; the others wrote verses which were representative enough to be selected by Geibel and Heyse for their book of translations.

Wolf was committed to setting the poetry of the past, including ancient poetry, as is demonstrated in this category. Almost all of his choices were taken from ancient writers and from authors bom 16-100 years earlier. This makes it especially seem as if Wolf belongs more with the song composers bom a generation or two before him than with the post-Wagnerians, and it helps distinguish him as the last of the nineteenth century Lieder composers, not as the begitming of the twentieth century composers who followed. 69

W olfs Choices in Non-native Languages

Almost all of Wolfs choices in this categoiy have been previously discussed: they were all either Undated or Old Sources, or both. The only exceptions are the two Byron texts set by Wolf. The 96 choices from non-native sources were all set in German translations, not in their original languages, and were 30.67% of his total choices. This percentage is one of the highest of the twelve composers for this category.

Three basic factors account for Wolfs high number of foreign-language choices; all have been previously mentioned. First, Wolf was familiar with the world literature of the day, as shown in Frank Walker’s account of Wolfs reading habits, in the section on his literary background in this paper. Second, he was personally drawn to exotic lands, especially the Mediterranean. This is demonstrated by the foreign associations in many of his musical works, not just his Lieder. Third, the influences of other song composers on

Wolfs textual choices have been noted,.

W olfs Personal Contact with the Poets

One poet who was the author of a text which Wolf set to music has been identifled as a personal friend of the composer. Paul Peitl was the author of the text for "Bin Grab,” an early song of 1876. He was also the librettist of the opera KonigAlboin, for which Wolf sketched some music but did not flnish, also from 1876. Literary historians are not sure of

Peitfs exact year of death, either 1917 or 1922.*°^ For this study the optimistic 1922 was recorded in Table 16 for Peitl. He sometimes wrote under the pseudonym “Paul

Maimsberg.” Wolf was simply too young to have known most of the poets from whom he chose texts, and the rest were unknown or anonymous writers.

**^See “Mannsberg, Paul,” Deutsches Uteratm-Lexikon, 11 vols, to date, ed. Heinz Rupp and Carl Ludwig Lang (Bern: A Franke AG, 1968-86), X, p. 373. 70

Wolf did know many writers outside those of his song texts. For example, he spent time with the poet Detlov Liliencron and the writer Michael Georg Conrad in 1890.

Liliencron had sent Wolf two volumes of poems for possible musical settings, but in them

Wolf "...found nothing sufficiently lyrical to inspire him to song.”*°^ He also knew a good many writers through his constant search for a good opera libretto. His beloved

Morike he did not know, but his time spent in Tübingen in 1890 with Professor

Kaufmann, one of Morike’s friends, was the next best thing. Together the elder Kaufmann and Wolf talked of Morike, and Kaufmann "...read Wolf the letters that his father had received from the poet.”'°^ This was a special meeting for Wolf, and Walker describes it as "...wholly delightful, and undisturbed by the slightest disharmony.”'*)?

Hugo Wolf was not a composer who saw his poets face-to-face. That he was able to capture their essence so completely with such limited and indirect contact makes his achievements in song even greater.

'*)^alker, op. cit., p. 269.

' “^Walker, ibid., p. 272-3.

'0?7fcid., p. 273. 71

SUMMARY FOR THE LIED COMPOSERS

Lied Composers’ Educational and Literary Background

Biographers of Schubert, Schumann, Brahms, and Wolf have written that each had some basic formal education, but not a great amount. Schubert went as far as teacher’s college and had a school teacher for a father. Schumann’s formal education did not really reach the college level, but his early schooling was of the highest quality. Brahms’s family was poor, yet saw to it that young J ohannes was schooled. Wolf had a few years of schooling before his teenage failure at the Vienna Conservatory. In each case, it was the informal continuation of the study of arid interest in literature, combined with the need to compose music, which sooner or later led them to write vocal music.

The manifestation of each composer’s literary life was unique. Schubert surrounded himself with literary friends, and showed an interest in both the works of great writers and those of his intimates. Schumann became a serious writer himself in his younger years; this evolved into a later career as a music critic. Brahms turned to books to understand his world, and his library was a tangible product of his literary interests. Wolf took on aspects of the literary lives of the other three. Like Schubert, he knew a number of literary people; like Schumarm, he spent time writing as a music critic; and like Brahms, he kept his eyes open to the songs and texts of other composers and the great literature of the world.

Lied Composers’ Poetry Choices

The Lied composers showed a personal approach to working with their choices of authors. Schubert would usually set the texts of one poet at a time, producing either a few or a great number of songs in the process. He would abandon some authors after one session with their texts, while he returned to others several times (particularly Goethe). 72

Schumann and Wolf also generally concentrated on one poet at a time, tiying to set a variety of moods from a given author’s texts. Schumann’s results are found in his great song cycles, and W olf carried this process to its farthest limits in his settings of Morike and

Goethe, and in the two Liederbucbs. Brahms, on the other hand, usually chose a concept, such as love, separation, or death and then chose texts (by any author) which he could use to express the concept through music. While Brahms’s choice of authors seems unfocused, the mood of the poems themselves is actually the unifying factor in his choices.

Four authors were set at least once by all four Lied composers: Goethe, Heine,

Riickert, and Shakespeare. The first three of these are not surprising to see on all four composers’ lists, for they aie authors of great German lyric poetry, which lends itself well to musical setting, and their importance and fame were becoming established in or before

Schubert’s day and grew in later years. Shakespeare, the master writer of the English language, may seem out of place with this group of writers. However, German translations of Shakespeare (and other English-language authors) were highly favored in

the nineteenth century, and many of these verses (even in translations) were also well-

suited to musical settings. Discussion of the foreign influences in the works of the Lied

composers will be continued under the heading of “Non-native Languages’’ below.

Three of the four composers in this discussion chose at least one verse by Uhland,

Hofltnan von Fallersleben and Morike. Uhland’s folk-like Romantic poetry was not to

Wolfs taste. Schubert died too soon to know and use Morike’s poetry. Hoffman von

Fallersleben was a contemporary of Schubert, but his verses were written too late for

Schubert’s use.

A greater number of authors were favored by two o f the composers, and the rest were

unique to one composer. These singular choices were often friends of the composer,

obscure writers which held the interest of a given composer, or in the cases of Schubert

and Schumann, the composer himself. 73 Each of the Lied composers is now associated mainly with one poet: Schubert with

Goethe, Schumann-with Heine, Brahms with Daumer, and W olf with Morike. These associations are valid, given the high quality and high number of songs produced from these associations (with the possible exception of Brahms/Daumer). Yet all four composers used more than a few textual sources, from W olfs 27 to Schubert’s 97. Also,

there are many masterpieces in each composer's output using texts other than those by the

“main poet.” It is safe to say that while the generalizations about each composer’s primary association with a single poet have some truth in them, all choices of texts and songs for a given composer must be carefully evaluated before a complete understanding of the composer and his works can be gained.

Lied Composers’ Number of Songs and Sources

The number of songs written by the four composers in Chapter I which conform to the criteria outlined in the Introduction ranges from 201 by Brahms to 624 by Schubert

(Figure I).

______. I «94 Schubert warn Schumann 257 Brahms 201 Wolf 313

I I ! I I I I I 0 100 200 300 400 500 600 700

nCU RE 1: UED COMPOSERS Number of Songs

The Lied composers are the four most prolific songwriters of the twelve studied, and their latge outputs may be attributed to several factors. For Schubert and Brahms there was 74 a lifelong commitment to creating vocal works in a relatively steady stream. For Schumann and Wolf, on the other hand, there were more highly concentrated periods for songwriting, in which they produced the bulk of their Lieder.

Another more subtle reason for each composer’s large number of songs may be his relationship to opera. None of them were prolific or highly successful opera composers

(though Wolf tried hard to become a successful opera composer). An alternative way to bring words and music together was through song, and this was the medium into which these composers put their fullest energies (for Brahms it was choral and vocal ensemble music as well).

Additional factors, such as the availability of worthy singers, the appeal of the intimacy in the song performance, the public popularity of the song in the nineteenth century, and the likely possibility for publication may have increased the song outputs for these masters.

Finally, there were personal reasons for producing large numbers of songs, such as the love lives of Schumann and Brahms, which gave them much inspiration, and can be seen so clearly in Schumaim’s wedding gift to his wife Clara, the cycle Frauenliebe und -leben.

The number of sources these composers used ranged from 27 by Wolf to 97 for

Schubert, as seen in Figure 2.

Schubert (| Schumann V/////////\ 50 Brahms | 47 Wolf T'l r I I I I I I ri " 0 10 20 30 40 50 60 70 80 90 100

HGURE 2: LIED COMPOSERS Number of Sources 75

This number is not in itself important for comparison purposes, since the number of songs for all the composers is not constant. The purpose in knowing the number of sources is rather for obtaining a re/aftVe number of songs created from each source. This is determined through the three averages derived from the breakdown of the songs by source

(see Tables 13-16, pp. 232-240). The first of these, the arithmetic average, is shown in

Figure 3.

Schubert - . BSSinn ^ 6.43 Schumann 5.14 Brahms 4.28

Wolf 11.59 n ' T~l I I T 'l I I "T I I 0 1 23456789 10 11 12

HGURE 3: LIED COMPOSERS Arithmetic Average Number of Songs Per Source

This graph shows a range of 4.28 songs per source for Brahms to 11.59 songs per source for Wolf. Here is visible evidence of Brahms’s lack of interest in working with a given composer to any great extent, and Wolfs absorption into the works of a smaller group of authors. The mid-range averages for Schubert and Schumann indicated their large 76 use of some authors’ texts while using smaller numbers from many others. A look at the median averages and the modal values for these composers are combined in Figure 4.

Schubert i# 0 1 2 .0 0 Schumarm /'/ / / y \ 2.00 U Modal Value Brahms 1 2 00 ^ Median Average Wolf s/s/s/s/s/v 2.50

0 1

FIGURE 4: LIED COMPOSERS Median Average and Modal Value of Songs Per Source

The low range of these values presents the other side of the relative frequency in the choices; i.e., the high number of poets whose works were selected only once or twice by all four composers. It should be noted that the strength of the modal value for these four composera was the greatest for Schumann (who used 48% of his sources once) and the lowest for Wolf (who used only 37% of his sources once), with Schubert and Brahms somewhere in between. These averages show the composers’ high levels of experimentation with choosing verses and the considerable lack of commitment to using large numbers of works by many poets. 77

Lied Composers’ Authors Categorized by Birth Date

The first group in this category is the percentage of selections from authors bom 16-

100 years before the composer, compared in Figure 5.

Schubert 1 - 61.38% 42.80%

Brahms i1 ® 55.72% W o lf 49.52%

Percent 0 5 10 15 20 25 30 35 40 45 50 55 60 65

nCU RE 5: UED COMPOSERS Percentage of Texts from Poets Bom 16-100 Years Earlier than Composer

The range in this category is from 42.37% for Schumann to 61.38% for Schubert.

These relatively large percentages show that all four composers took roughly half their texts from poets bom in the preceding generations. Most of the possible reasons for this are explained in the sections on the individual composers. This comparison indicates that all of them often used the works of “established” older writers. 78 The second comparison is the relative use of texts from poets contemporary with the composer, i.e., bom 15 years earlier or later than the composer (Figure 6).

Brahms 14.43% W o lf \ 0.32% M i l l T T r I I "I Percent 0 5 10 15 20 25 30 35 40 45 50

nCURE 6: LIED COMPOSERS Percentage of Texts from Authors Bom Within 15 Years (Plus or Minus) of Composer

The range of use in this segment is much wider, from Schumann’s 45.80% to Wolfs

0.32%. Such factors as the age of the composer during his peak songwriting years and how many poets the composer knew personally have been given as factors in this discrepancy. Schubert and Schumann lived closer in time to the height of the Romantic age of lyric poetry, and were thus more apt to use texts by early- to mid-Romantic period authors for their poetry choices. The percentages indicated in Figure 6 confirm this. 79

Smaller percentages are recorded in Figure 7 for poetry choices from authors bom more than 15 years later than the composers.

Schubert 0% Schumann

Brahms 4.48%

Wolf 0% I I I I I I 0 1% 2% 3% 4% 5% FIGURE 7: UED COMPOSERS Percentage of Texts from Authors Bom More than 15 Years After Composer

Neither Schubert nor Wolf used any texts by younger authors, and the choices here for

Schumarm and Brahms are both under 5% of their totals. A variety of reasons have been given to explain these figures for each composer; e.g., short lives for Schubert and Wolf, longevity for Brahms, and the composers’ access to these poet’s works. The Lied composers clearly looked to the past for their song poetry, not toward the future. 80 Lied Composers’ Choices of Undated and Old Sources

The use of Undated (esp. Anonymous) sources became more imporant for the later

Lied composers, as the progression of increasing percentages in Figure 8 shows.

Schubert I 3.37% Schumann ^ 5.84%

Brahms 19.40%

Wolf 26.84% I I I I T ” T I Percent 0 5 10 15 20 25 30

FIGURE 8: LIED COMPOSERS Percentage of Texts from Undated Sources

Brahms’s strong interest in folk poetry and Wolfs large collections of anonymous

Spanish and Italian verse settings are the distinguishing features of this category.

Wolfs comparatively high percentage of texts taken from authors bom more than 100 years before the composer are in marked contrast to the other Lied composers, as seen in

Figure 9.

Schubert 1.60% Schumann ^ 2.33%

Brahms 5.97% Wolf 23.32%

Percent 0 10 15 20 25

nOURE 9: UED COMPOSERS Percentage of Texts from Authors Bom More than 100 Years Before Composer 81

Yet all four Lied composers had ties with old poetry, even if only small amounts.

Texts by such great poets from the past as Shakespeare, Petrarca, Michelangelo, and

Cervantes were familiar enough to consider them for musical settings. Both the universal greatness of the poetry and the public prominence of ancient literature in the nineteenth century are reflected in the Lied selections and settings of ancient works. These songs also acutely show the historical perspective and depth of understanding which each of these composers possessed. As will be seen at the end of this document, the American song composers studied here have continued and developed the trend of using old poetry as established by the composers of the Lied, while the French were generally more interested in the poetry of their own times.

Lied Composers’ Choices in Non-native Languages

Figure 10 indicates that the choice of texts from non-native languages (i.e., not originally in German) also increased progressively through the four Lied composers.

Schubert 7.21%

Schumann \/////////y / 11.28%

Brahms 14.92% T3 Wolf 30.67%

Percent 0 5 10 15 20 25 30 35

nCURE 10: UED COMPOSERS Percentage of Texts in Non-Native Languages

The prime alternative languages to German were Italian and English for Schubert,

English and Spanish for Schumann, a great variety of languages for Brahms, and Spanish 82 and Italian for Wolf. All but the Italian settings of Schubert were in German translations.

The growing prominence and availability of German translations of foreign poetry, mentioned several times earlier, are probably the most important influences on the choices in this category. The themes in the British verses, their outdoor and romantic moods, were particularly appealing to Schubert and Schumatm, as were the love themes of the Spanish and Italian verses for Wolf and the variety of moods in folk poetry for Brahms.

The universality of Shakespeare’s verses stand out in this category, because his verses appeared in settings by all four Lied composers. Translations of Shakespeare’s plays were an inîpüTtart element of Romantic drama; he (and the Spanish dramatist Calderon) “.. .seem almost omnipresent in Romantic statements on dramatic theory, and Romantic drama is greatly indebted to them as sources and m o d e l s . The appearance of settings of verses from his plays by Lied composers is thus an extension of Shakespeare’s influence upon the

German Romantic theatre, and one example of his influence on the German Romantic period in general.

*°*Roger Paulin, “The Drama,” The Romantic Period in Germany, ed. Siegbert Prawer (New York: Schocken Books, 1970), p. 180. 83

Lied Composers’ Personal Contact with the Poets

The approximate number of authors (from whom verses were chosen) in personal contact with the four Lied composers are listed in Figure 11.

Schubert i d " 19 Schumann 22 Brahms

Wolf w 1

T 10 15 20 25

HGURE 11: LIED COMPOSERS Approximate Number of Authors With Whom the Composer Had Personal Contact

A high degree of contrast in the personal contact with the authors selected is the most notable feature of Figure 11. Schubert and Schumann were both a part of the literary circles of their times, and as a result knew many of the writers from whom they selected poetry. Brahms and Wolf, however, were not as closely tied to the writers of their chosen song texts. They both knew many writers, but the associations were not generally in connection with their poetry selections for songs.

Thus for some composers this category (i.e., the personal link with the poets) has a strong bearing on their work, while for others this was simply not an important component of the text selection and songwriting processes. This category is of relative importance for different song composers, then, because the success or failure of a given song can rarely be linked to whether or not the composer knew the author of the text. CHAPTER II

TEXTUAL SOURCES FOR THE MELODIE COMPOSERS

Introduction to Chapter II

The style of song popular in France at the beginning of the nineteenth century was called the romance. Its simple style included a strophic form, often bad prosody, with a trite melody line and accompaniment. Many composers of the time wrote hundreds of these simple romances. External cultural influences, however, such as the importation of

Schubert’s songs into France, caused some composers to reconsider the songwriting procedure of the time. Two composers in particular, Berlioz and Liszt, tried to compose a higher quality of vocal work; the new type of song which they and other serious composers were writing in the middle of the century came to be called the mélodie.

What made the mélodie different from the romance was its varied formal structure and its more careful attention to prosody. In addition, the mélodie showed its greatest improvement over the romance in its poetry. Instead of insignificant verses, texts by the great French romantic writers, such as Lamartine and Hugo, were being used in the mélodies. The new style of song was quite developed by the 1860’s, as seen in the mélodies of Duparc and Fauré. These two composers expanded upon the style of French vocal music established by their teachers: Franck, Gounod, Niedermeyer, and Saint-

Saëns. The mélodies of Duparc and Fauré were especially distinguished as settings of the newer Parnassian verses. Parnassian poetry, like the mélodie, flourished in the 1860’s, and was conceived in reaction to the Romantic style of poetry, i.e., it followed a stricter formal structure and contained more precise imagery.

84 85 Debussy brought the art of mélodie composition into the twentieth century. He gave his songs a more complex harmonic language than Duparc and Fauré, and showed an extremely careful attention to prosody. His mélodies chiefly used Parnassian and post-

Pamassian, or Symbolist poetry.

Poulenc is today considered the best composer of mélodies in the twentieth century.

His vocal works contain both tonal and innovative harmonies, and his choices of texts were primarily from the Surrealist poets of his own time.

These four French composers, Duparc, Fauré, Debussy, and Poulenc, have been chosen for this study because of their importance as mélodie composers. Their educational and literary backgrounds, as well as their poetry choices will be carefully examined in this chapter. As will be seen, these four composers each had either some general or musical education, or both. Their artistic contacts, especially in the Parisian literary circles, were often quite influential to their choices of verses for their songs. These texts were most often from the French language, and most often from contemporary writers or writers from the generation before them. Finally, each of the four composers came to favor one author in particular, but successfully set poems from a number of authors. 86

HENRI DUPARC

Duparc’s Educational and Literary Background

The educational background of Henri Duparc (1848-1933) is not fully documented. He studied at the Jesuit College at Vaugirard, where he studied piano with César Franck. He eventually studied law and was a practicing lawyer. These are the only known facts about his formal eduation. He used both his musical and legal training to good effect. His life in music brought him close to most of the great French composers of the day, with whom he helped form the Société Nationale de Musique in 1871. Northcote described Duparc’s non­ musical life: “As to his legal duties, more especially as the genial mayor of Mames-la- coquette, it is evident he brought to them just as much care as he gave to his music.” •

Duparc was one of the best-educated of the twelve composers studied, at least in terms of non-musical training.

Duparc’s artistic sensibilities were fully developed as an adult. His tastes in art and

literature were of the highest order

He was among tl^ earliest French admirers of Tolstoy’s War and Peace, of Ibsen’s plays, of Jf rench primitive painting and of oriental art, not only Japanese prints but Japanese theatre and Cambodian dancing, both of which he saw at the Paris exhibition of 1900. He was a great Dante enthusiast, an early champion of Baudelaire and Verlaine, and deeply attached to Mistral’s poetry. 2

His musical tastes were shaped by his teacher Franck, who guided him to an admiration

of Bach and Beethoven. He also came to know important musicians of the day, such as

Liszt and Wagner. Duparc even offered advice to Wagner regarding his opera productions.

* *Sydney Northcote, The Songs o f Henri Duparc (London: Dennis Dobson, 1949), p. 55.

^Martin Cooper, “Duparc, Henri,” The New Grove Dictionary o f Music and Musicians, ed. Stanly Sadie, 20 vols. (London: Macmillan, 1980), V, p. 726. 87

Cooper commented that, “In this, as in other aesthetic matters, he was in advance of his time. This well-rounded, intelligent man withdrew from the world at the age of 37, suffering from a nervous disorder for nearly fifty more years. Today he is remembered exclusively for his small yet remarkable musical output.

Duparc’s Poetry Choices

Four of Duparc’s ten poets are considered some of the finest French poets writing in the middle of the nineteenth century. He set two verses from the great Charles Baudelaire’s important book Les Fleuis du M ai (1857), “L’Invitation au Voyage” in 1870 and “La Vie antérieure” in 1884. The latter was to be Duparc’s last song. He also chose two verses by

Théophile Gautier, for his early song “Au Pays où se fait la Guerre” in 1869 and for the later song “Lamento” in 1883. Duparc extracted four verses of Leconte de Lisle’s

“Phidylé” in 1882 for his song of the same title. Northcote praised both the poem and the song: “This is a truly magnificent setting of a beautiful poem.”^ The fourth of Daparc’s

“better” poets was Sully-Prudhomme, from whom he took verses for two early songs:

“Soupir” and “Le Galop,” both composed in 1868. Duparc later tried to destroy all published copies of “Le Galop;” it was considered lost until 1948, when it was re­ discovered and re-issued.

Duparc set one non-French poem, a translation of Moore’s lament of the death of

Robert Emmet. This song, “Elégie,” was composed in 1874, and will be mentioned again later in the “non-native language” category.

It was Northcote’s opinion that, “The remainder of Duparc’s poets would hardly win a place in a French anthology of poetry purely on their own merits.”^ Duparc set three

^Ibid

'^Northcote, op. cit.. p. 105.

^Ibid., p. 71. 88

verses by Jean Labor (pseudonym for Dr. Henri Cazalis), "Chanson Triste” in 1868,

“Extase” in 1878, and “Sérénade Florentine” in 1880. He selected two verses from

friends, “Testament” by Armand Silvestre in 1883 and “Le Manoir de Rosemonde” by

Robert Bonnières in 1879. Duparc’s “Romance de Mignon” of 1869 is a free adaptation of

Goethe’s “Kennst du das Land” by the obscure Belgian writer Victor Wilder. One final

song also dates from 1869, “Sérénade,” using a text by an obscure poet bom in the

Auvergne region of France, Gabriel Marc. These last two songs were also considered lost

for many years but were evenutally rediscovered. 89

/ Tables. The Songs of Henri Duparc

Description Number %

Total number of Songs 15

Total number of Sources 10

Arithmetic average number of songs per source: 1.50

Median number of songs per source: 1.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 6 40.00%

Texts from poets contemporary with composer (bom withiii. 15 years of composer) 9 60.00%

Texts from poets younger than composer (bom 15 years after composer) 0 0.00%

Texts from undated sources (dates of poet or source unknown) 0 0.00%

Texts from old sources (dated more than 100 years before ccmpcscr^s birth) v u.uuyo

Texts from non-native sources (either in translations or in other languages) 1 6.67%

Approximate number of poets with whom the composer had personal contact: 2 90 Duparc’s Number of Songs and Sources

Duparc’s total number of songs counted in this study is set at 15. The primary guide used to obtain this figure was Martin Cooper’s list of Duparc’s works in The New Grove

Drcrionary article.^ There are today sixteen surviving Duparc songs. Most sources on

Duparc’s works prior to Cooper list two songs as being originally for voice and orchestra.

Cooper, however states that, “Only one of Duparc’s song, ‘La vague et la cloche,’ was designed for the orchestra... Duparc’s last song, “La vie antérieure,” is the other song thought to be originally with orchestra. Cooper dates the original of this work (with piano accompaniment) at 1884 and the orchestrated version at 1911-13. It is with Cooper’s authority that “La vie antérieure” is included in the present count.

Duparc used ten sources for his songs. This number was also determined from

Coopers i:st of Duparc’s works, compared with other lists of his songs. The ten sources are listed in Table 17, p. 241.

The arithmetic average for Duparc is 1.50 songs per source, based on 15 songs and 10 sources. This is the lowest average for any of the twelve composers studied. Duparc’s median average and modal value are both set at one. This is a strong modal value, given that six of the ten sources were used only once. Duparc’s choices from his sources were consistently low, using most sources once and none of the rest more than three times.

Duparc’s Authors Categorized by Birth Date

Six of Duparc’s songs used texts from authors bom in the 16-100 years before him, nine were from authors bom 15 years earlier or later than his birth (1848), and none were from authors bom more than 15 years after his birth. The percentages of his total choices

^Martin Cooper, op. cit., p. 111.

Tibid. 91 from these three groups are as follows; 40.00% were by authors bom between 1748 and

1832, 60.00% were by authors bom between 1833 and 1863, and 0.00% were by authors bom after 1863.

Duparc was most concemed with setting the Parnassian poetry of his own generation and the one which immediately preceded it. Thomas Moore was the only author Duparc chose bom more than 50 years before the composer. Much of the poetry from the older generation of French poets was written in the later years of these poets’ lives, which were simultaneously the early years of Duparc’s life. This poetry wets essentially contemporary to Duparc, in spite of the actual age differences between composer and authors.

Northcote offered an explanation of Duparc’s interest in and use of contemporary poetry: “As one of the founders and first secretary of the Société Nationale he had shown his faith in contemporary musicians in a very practical manner. His choice of poets was probably due to a similar faith in contemporary poetry, neither more nor less.”* Duparc found that the post-Romantic poetry of his day was a suitable companion for his musical style. Northcote also wrote that the Pamassian poets gave their verses “...a chiselled refinement in rhythm and verbal sonority as well as a classical precision of form—all qualities which blended perfectly with a music which had developed the same native subtleties in clarity, logic and impressionism.”^ The older Romantic poetry weis to remain principally in the domain of the Romantic composers. Fauré was to parallel Duparc in his use of works by the contemporary Parnassians, as will be seen later in this document.

Duparc stopped writing music in 1885, at the age of 37. Had he continued on the same course he had established in his textual choices, he would have likely continued using contemporary poetry. Had he continued composing songs into his old age, he may have come to know and use the works of the Symbolist and Surrealist poets of the younger

*Northcote, op. cit., p. 62.

^Ibid., p. 63. 92 generations. While his long life would have allowed it, his nervous disorder prevented the world from knowing what might have been his latter-day poetic selections.

Duparc’s Choices of Undated and Old sources

Duparc set no anonymous or ancient verses. His strong ties with setting verses of contemporary writers had not given him the urge to go farther afield for his poetic sources.

There was no tradition yet established by the French post-romantic composers of his day in setting the great verses by French poets of the past (e.g., Charles d’Orléans, Villon, or

Ronsard), as would later be developed by Debussy and Poulenc. The speculation on his use of these writers’ texts is like that of his use of the younger writers: Duparc may have come to experiment with ancient poetry had he continued composing after 1885.

Duparc’s Choices in Non-native languages

One of Duparc’s texts is a translation of Thomas Moore’s verse “O breathe not his name,” for the 1874 song “Elégie.” This selection counts as 6.67% of his total choices.

Moore was an Irish poet and musician whose volumes of Irish Melodieswere published between 1808 and 1834. The French translations of these verses became quite popular in

France, especially for musical settings. Noske’s account of the origin of the term mélodie as a new type of vocal work implicated settings of Moore’s verses, which used the word mélodiesas translation o f the English m elodies!^

It is uncertain who wrote the prose translation of Moore’s poem Duparc actually set.

Northcote believed that Duparc wrote the version himself, citing “.. .Duparc’s expressive prose style, especially on elegiac and religious themes....” * * Other scholars think the translation might have been written by Duparc’s wife. Bom Ellie MacSwiney, she herself

*®See Frits Noske, French Song from Berlioz to Duparc, tr. Rita Benton (New York: Dover, 1970), pp. 23-24.

Northcote, op. cit., p. 73. 93 was Irish (her mother had fled to France as a widow during the Parnell riots). Whoever may have done the translation, it was to be Duparc’s only setting of a verse from a non­ native source.

One other of Duparc’s verees which has a close connection to this category is Victor

Wilder’s free adaptation of Goethe’s “Kennst du das Land.” Duparc set Wilder’s version for an early song “Romance de Mignon,” published in the Cinq mélodies, op. 2. Noske noted the German influences in this song: “In 1869 Duparc attended performances of

Tristan and Das Rheingoldxn Munich and had probably studied some of Wagner’s scores before the composition in 1868 of his Cinq mélodies. The airy chords of the [opening of

“Romance de Mignon”] certainly awaken echoes of Lohengrin.”^'^ Another Duparc song among the early Cinq mélodies, “Le Galop,” shows musical influences of Schubert, although the text is from a native French source.

Duparc’s ties to other lands, such as his Irish-born wife and his Wagnerian interests, may have had some influence on his textual choices, but for the most part the verses he used were French in origin. His percentage in this category is one of the lowest for the twelve composers studied.

Duparc’s Personal Contact with the Poets

Two poets from whom Duparc chose verses are identified as having been known personally by the composer: Robert de Boimières zmd Armand Silvestre. They are both marked with an asterisk in Table 17, p. 241. Bonnières was only two years older than

Duparc, and was known in his day more as a novelist and journalist than as a poet.

According to Northcote, “He was a close friend of the composer and at one time shared an apartment with d’Indy and Duparc.”’^ Silvestre’s relationship with Duparc was less

’^Noske, op. cit., p. 278.

^^Northcote, op. cit., p. 72. 94 intimate and not as well-documented. Northcote’s paragr^h on Silvestre mentions

“...Duparc’s disappointment about the critics’ indifference to his friend’s work.”'**

It is quite possible that Duparc may have known more of the authors of his song texts.

His keen interest in the cultural world of the time and his contemporaneity with the poets would suggest this. Duparc’s commitment to setting the poetry of his day was probably more related to his poetic instincts and to his involvement with the Parisian artistic community than to personal relationships with the poets. The small number of poets in this category implies, that with Duparc there was little or no direct correlation between the number of chosen poets he knew and his interest in using the poetry of his time.

!4 lid. 95

GABRIEL FAURF.

Fauré’s Educational and Literary Background

Gabriel Fauré (1845-1924) received all of his education at the Ecole Niedermeyer in

Paris. Fauré’s father (through the advice of interested musicians) decided that this school

"...could prepare his son for the profession of choirmaster while cultivating his natural

[musical] gifts. He took Gabriel to Paris (a three-day journey) in October 1 8 5 4 . ” *5 The main focus of the school’s educational program was musical study, but some time was given to instiuction in other subjects, especially languages and literature. “[School founder

Louis] Niedermeyef’s report for 1854 details the weekly timetable: ‘three lessons of

French language, two of Latin, one of arithmetic, one of geography, one of history and literature,’ to which was subsequently added a smattering of Italian.”*^ It was a disciplined

course of study.

Fauré stayed in school for eleven years: “His student career at the Ecole Niedermeyer

was completed on 28 July 1865..Six months later he began his professional career as

an organist; he returned to academic life many years later in the roles of teacher and

administrator. His early studies had given him enough exposure to poetry to begin writing

vocal music while still in school, as seen in his Hugo settings of 1864. The pattern for

songwriting was established in his school days, and continued through the rest of his adult

life.

*^Jean-Michel Nectoux, “Gabriel Fauré,” The New Grove Twentietb-Century French Masters (originally published in The New Grove Dictionary o fMusic and Musicians, ed. Stanley Sadie) (London and New York: Macmillan, 1980,1986), p. 1.

'^Jean-Michel Nectoux, Gabriel Fauré: His Life Through His Letters, tr. J. A. Underwood (London and New York: Marion Boyars, 1984), p. 15.

*^Jean-Michel Nectoux, “Gabriel Fauré,” The New Grove Twentieth-Century French Masters, op. cit., p. 3. 96

Fauré’s participation in the Paris salons brought him in contact with many cultural people of the day, and his poetry choices for songs were greatly influenced by some of these people. Count Robert de Montesquiou was one of Fauré’s chief literary advisors.

Nectoux suggested that when Fauré met Montesquiou in 1886 or 1887, the latter steered the composer away from such poets as Sully-Prudhomme and Silvestre and led him to others, such as Villiers de I’lsle-Adam. More importantly, “One of Montesquieu’s chief claims to fame will always be that it was he who introduced Fauré to the poetry of

Verlaine.”'*

Fauré was not always easily swayed by others, however. Some poems which friends proposed were not used, and "... while gratefully accepting all these suggestions he managed to preserve his creative independence intact; he never composed to oblige.”'^ His literary tastes could sometimes be influenced by others, while at other times they were completely autonomous, as seen in his choices (and rejections) of authors and texts.

Fauré’s weaker choices for song texts sometimes present him as being not of the highest literary mind, but he was an educated and intelligent man. His own skill at writing is evidenced in the articles he wrote as music critic for L e Figaro between 1903 and 1921.

He also personally knew several important writers of his day (especially Proust). Fauré chose some verses not for their high quality, but for “...their pliability, lack of reference to sounds and, particularly, lack of visual description that would restrict him....”^° Like

Brahms, his choices of poetry for his songs did not readily reflect his true intellect.

' *Jean-Michel Nectoux, Gabriel Fauré: His Life Tbrougb His Letters, tr. J. A. Underwood, op. cit., p. 200.

^^Ibid., p. 199.

^"Jean-Michel Nectoux. “Gabriel Fauré,” The New Grove Twentietk-Century French Masters, op. cit., p. 23. 97 Fauré’s Poetry Choices

The works of Gabriel Fauré are often grouped into three periods. One of the distinguishing features of these periods is the poetry he used for the songs in each of them.

The early songs of the 1860’s ate dominated by settings of Victor Hugo; the first seven are all by the great poet of the French Romantic period. By about 1870 Fauré began setting verses by other authors, such as Gautier (“Les matelots,” op. 2 no.2 and “La chanson du pêcheur,” op. 4 no. l),2i Leconte de Lisle (“Lydia,” op. 4 no. 2; “Nell,” op. 18 no. 1; and

“Les roses d’Ispahan,” op. 39 no. 4) and Baudelaire (“Hymne,” op. 7 no. 2). Some of the songs from the 1870’s were single texts chosen from obscure authors: Louis Pomrney

(“Aubade,” op. 6 no. 1), Marc Monnier (“Barcarolle,” op. 7 no. 3) and Paul de Choudens

(“Sylvie,” op. 6 no. 3) are three examples. In 1878 he chose three poems by

Grandmougin for his set of songs Poème d’un jour, op. 21.

Texts by Sully-Prudhomme, Armand Silvestre, and Villiers de I’lsle Adam complete the choices for his first period of song composition. Sully-Prudhomme’s verses were used for the well-known songs “Ici-bas!,” op. 8 no. 3; “Au bord de l’eau,” op. 8 no. I; and

“Les berceaux,” op. 23 no. 1. Vuillermoz called Fauré’s “Les berceaux” “...one of his most justly popular songs.”^^ Seven Silvestre poems were set in the early 1880’s, found in op. 23 (“Notre amour” no. 2 and “Le secret” no. 3), op. 27 (“Chanson d’amour,” no. 1 and “Le fée aux chansons,” no. 2), and op. 39 (“Aurore,” no. I ; “Fleur jetée,” no. 2; and

“Le pays de rêves,” no. 3). The two poems from Villiers de I’Isle Adam were chosen in

1886-87, for “Nocturne,” op. 43 no. 2 and “Les présents,” op. 46 no. 1.

Paul Verlaine was the predominant author for Fauré’s middle period of song composition. The first Verlaine setting was “Clair de lune,” op. 46 no. 2 in 1887. The

^^The opus numbers used were assigned to the songs at the time of publication. The authority for the opus numbers used in this document was the work-list by Jean-Michel Nectoux, “Gabriel Fauré,” The New Grove Twentietb-Century French Masters, op. cit., pp. 29-31.

^^Emile Vuillermoz, Gabriel Fauré, tr. Kenneth Schapin (Philadelphia: Chilton, 1960, 1969), p. 71. 98 song is important in Fauré’s work for its independent vocal line and accompaniment, as well as for its roccoeo theme, which create a work . .whose success does not happen to dull the power of its magical charm, [and] leaves us with the feeling of absolute perfection.”23 Other Verlaine settings followed: “Spleen,” op. 51 no. 3 in 1888, the Cinq mélodies, op. 58 in 1891, the nine songs in the great song cycle La bonne chanson op 61 of 1892-4, and for the song “Prison,” op. 83 no. 1 in 1894. Norman Suckling wrote of the significance of the VerlainevTauré collaboration:

[A] civilized sensibility was the distinguishing mark of most French art in the period of Fauré’s maturity—from about 1880 onwards—and his music thus came as the perfect counterpart to the poetry of the same age: that poetry which was labelled ‘decadent’ by critics who knew no better than to regard it as the mere disintegration of the Romantic poetry of a half a century earlier. In fact it was in great part a retum to artistic values which the Romantics had set aside; it fulfilled an essential condition of great art by being conscious of its feelings, not naively abandoned to them, and by treating emotional excitement as a thing not so much to be aroused in sympaùy as to be resolved into form. Verlaine, one of its leading exponents, though he sometimes fell into emotional extravagances which were a betrayal of his own best standards, had nevertheless a rich share of the central post-romantic virtue, an ability to use words with a fuller meaning than they carry in their ordinary application; which rendered him the apt partner for a composer whose musical ideas most fully tell their tale when not obscured by would-be ‘expressiveness’ of rendering: Fauré’s settings of Verlaine indeed reach a level not surpassed by the song-writing of any other age or nation.24

Fauré wrote songs for thirty more years after the Verlaine settings. The authors chosen at the end of his second period besides Verlaine were Samain, Mendès, and a retum to

Silvestre.

Fauré completed four major song cycles in his third and final compositional period

(1906-24). The first two. La chanson d’Eve, op. 95 and Le jardin clos, op. 106 used eighteen texts by Charles Van Leberghe, ten for the first and eight for the second. These were followed by Mirages, op. 113, four songs with texts by Renée (La Baronne) de

Brimont. The single song “C’est la paix!,” op. 114, on a text by Georgette Debladis was

23ib/rf., p. 72.

^■’Norman Suckling, “The Songs of Fauré,” The Listener, XXXIII (March 15, 1945), p. 305. 99 set by Fauré as part of a contest in 1919 (discussed further under Fauré’s “Undated

Sources”). His final song cycle was L ’Horizon chimérique, op, 118, four songs on texts by a poet killed during World War I, Jean de la Ville de Mirmont. The few other authors not mentioned here are included in the complete list of Fauré’s authors, Table 18, p. 242. 100

' Table 6. The Songs of Gabriel Fauré

Description ______Number______9 ^

Total number of Songs 102

Total number of Sources 26

Arithmetic average number of songs per source: 3.92

Median number of songs per source: 2.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 22 21.57%

Texts from poets contemporary with composer (bom within 15 years of composer) 47 46.08%

Texts from poets younger than composer (bom 15 years after composer) 25 24.51%

Texts from undated sources (dates of poet or source unknown) 8 7.84%

Texts from old sources (dated more than 100 years before composer's birth) 0 0=00%

Texts from non-native sources (either in translations or in other languages) 2 1.96%

Approximate number of poets with whom the composer had personal contact: 4 101 Fauré’s Number of Songs and Sources

Two recent and authoritative catalogues of Fauré’s works were compared to determine his number of songs and sources to be included in the study. The catalogues are Nectoux’s list of works for The New Grove Dictiorarÿ-^ and the one found in Appendix A of

Orledge’s biography o f Fauré.26 Nectoux catalogued 100 published songs from 24 sources. Orledge duplicated these and added two complete but unpublished songs from two other sources to his list. The two unpublished songs were included in the works counted for this study, making the totals for Fauré 102 songs from 26 sources. His three published songs written originally as incidental music for three plays were not included, nor was the “Vocalise étude,” which has no text.

Fauré averaged 3.92 songs per source, based on 102 songs from 26 sources. The median average in the breakdown of sources is two, and the modal value is 1. These averages are low-to-mid range for the twelve composers studied. The mid-range arithmetic average suggests Fauré’s somewhat frequent use of a few sources and more infrequent use of others. The low median and fairly weak modal value (10 sources of 26 used once) accurately tell of Fauré’s lack of commitment to setting texts by many of his authors. None of them were chosen for more than 20 texts and only three for more than 10 texts. The majority were chosen for less than five texts. He generally chose a fairly lowmxmhet of texts from each of his sources.

Fauré’s Authors Categorized by Birth Date

Fauré’s choices of authors, when separated into groups by birth dates (relative to the composer’s birth date), were a balanced mix of older, contemporary, and younger writers.

26jean-Michel Nectoux, "Gabriel Fauré,” The New Grove Twentieth-Century French Masters, op. cit., pp. 28-34.

26Robert Orledge, “Chronological List of Works by Genre,” Gabriel Fauré (London: Eulenburg, 1979), Appendix A, pp. 318-326. 102

Twenty-two of his texts were by poets bom 16-100 years before the composer (before

1845, between 1745 and 1829), forty-seven texts were by poets bom within 15 years of

1845 (between 1830 and 1860), and twenty-five texts were by poets bom more than 15 years later than 1845 (after 1860). The percentages for these three groups: 21.57% of

Fauré’s texts were from older poets bom 16-100 years before him, 46.08% were from contemporary poets bom between 1830 and 1860, and 24.51% were from younger poets bom after 1860.

Fauré’s desire to set the texts of writers from his own time is one significant feature of the percentages in this category. Also noteworthy is the relative balance of choices among the three groups; i.e., choices from one or two age groups do not heavily outweigh the other one or two. Other aspects of the percentages here are unusual as well. Fauré’s choices of texts from older generation poets are the lowest of the twelve composers (see

Figure 38, p. 223). At the same time, his choices of texts from younger generation poets are the highest of the twelve, by a wide margin (see Figure 40, p. 225). Fauré has often been described as a transition composer e.g., Nectoux wrote that “Fauré’s stylistic development links the last years of Romanticism with the second quarter of the 20th century, and covers a period in which the evolution of musical language was particularly rapid.”27 His choices for song texts, when dated in the manner prescribed for this document, provide further credence to the idea that Fauré looked backward to the nineteenth century and foward to the twentieth century.

Fauré’s three compositional periods are paralleled by his choices of poetry. His early period (1860-85) was influenced by Romantic French music, and his primary poet of the same period was the older generation Romantic poet, Hugo. The middle period ( 1885-

1906) showed his full maturity as a composer. This period was the time of his Verlaine

27 Jean-Michel Nectoux, “Gabriel Fauré,” TbeNewGrove Twentietb-Century French Masters, op. cit., p. 15. 103 settings, his largest choice of texts by a contemporary author. Finally, music from his third period (1906-24) reflects both the violence of the first World War and the serenity and wisdom of old age. His choice of poet during this time was Charles Van Leberghe. The selection of verses from a younger author brought the aged Fauré to his final, forward- reaching achievements as a song composer.

Fauré’s Choices of Undated and Old Sources

Fauré set six texts from three authors who remain obscure (to this author): four by

Renée de Brimont for his Mirages, op. 113; one by Louis Pommey for “Aubade,” op. 6 no. 1; and one by Georgette Debladis for “C’est la paix!” op. 114. Orledge says that

Baroness Renée de Brimont was “.. .the great-niece of the poet and liberal democrat

Alphonse de Lamartine.. .”2* as well as “...the confidante and literary champion of the mystic religious poet Oscar Milosz ( 1877-1939).”29 Neither Orledge nor other biographical sources about her include her date of birth or death. It is likely that she was younger than Fauré, based on her relationship with Milosz and the knowledge that some of her books were published between 1919 and the mid- 1930’s.

Mile. Georgette Debladis is more obscure: all that she is known for is her post-war verse, which won in a competition sponsored by L e Figaro. Fauré, a music critic for this periodical, had agreed to set the wirming poem to music. Fauré himself called it a ‘horrible little poem’ in a letter to his wife. 3° Georgette Debladis was also probably contemporary with or younger than Fauré, given the circumstances of the encounter and Fauré’s age at the time of the contest. Louis Pommey is the most obscure of the three in this group. His single poem which Fauré used is the only information about this author that is known.

2®Orledge, op. cit., p. 4.

^^Ibid., p. 192 note 3.

’^Quoted in Orledge, ibid., p. 175. 104

Two poems by anonymous Italian authors complete the Undated Sources category for

Fauré. The two poems were translated by his friend Romain Bussine. The song settings are both early: “Sérénade toscane,” op. 3 no. 2 and the famous “Après un rêve,” op. 7 no. 1. Most sources claim Bussine wrote or adapted the verses, but Nectoux’s list (the authority used for this document) labels them as “anonymous, translated by Bussine.

Tlie eight poems in the Undated category are 7.84% of his total choices. Fauré was not strongly attracted to anonymous verse, as can be seen in this small percentage. He was additionally unattracted to Old Sources (from authors bom more than lOO years earlier than the composer), as is shown by the complete lack of choices in this category (0.00%).

There are influences from ancient times in Fauré’s music, particularly the aesthetics of ancient Hellenism in his stage works and the Roccoco “fêtes galantes” style he used in some of his Verlaine settings. These did not translate into choices of actual ancient texts, however.

Neither Fauré nor Duparc used any ancient poetry, a strong indicator of their status as the two composers who brought new life to the French mélodie through settings of contemporary post-Romantic poetry. Once the use of modem texts had been firmly established before 1900, later French song composers began to explore the ancient French poetry, as will be seen later in the textual choices of Debussy and Poulenc. Fauré destroyed his last attempt at song composition, a sketch of a Ronsard verse (from the 16th century), “.. .when he discovered Ravel had already set the same text. ”^2 doing so

Fauré had bequeathed the setting of ancient verses to the younger song composers.

Jean-Michel Nectoux, “Gabriel Fauré,” TbeNewCmve Twentietb-Century French Masters, op. cit., p. 29.

--Orledge, op. cit., p. 225. 105

Fauré’s Choices in Non-native Languages

The two anonymous Italian texts translated by Remain Bussine discussed earlier are the only two texts Fauré set which are originally from a Non-native (i.e., non-French) language. These two songs were written when “.. .Fauré’s association with [his friends] the Viardots from 1872 to 1877 inclined him temporarily towards an Italian sty le ... .”^3

They represent 1.96% of his total choices. Fauré was capable of working in other

languages, as his many Latin choral pieces show. He also travelled widely, to Italy,

England, and Russia. For example, “Fauré often went to London for private festivals

organized by loyal friends...he returned almost every year between 1892 and 1900.

When it came to choosing texts for his mélodies, however, his taste was overwhelmingly

for native French verse.

Fauré’s Personal Contact with the Poets

Four authors from whom Fauré chose texts have been identified as having been in

personal contact with the composer Georgette Debladis, Albert Samain, Armand Sully-

Prudhomme, and Paul Verlaine. They are marked with an asterisk in Table 18, p. 242.

Fauré did not actually c&oosethe text by Georgette Debladis, as he had offered to set the

winning poem of Le Figaro’s cotAcsi. There is no documentation to substantiate their

meeting, but the contest situation made it extremely probable. Samain was the librettist of

the opera Bouddha, on the life of Buddha, which Fauré was interested in setting in 1891-2

^^Jean-Michel Nectoux, “Gabriel Fauré,” The New Grove Twentietb-Century French Masters, op. cit., p. 21.

3‘»76iV/., p. 9. 106 but never com pleted.^^ Three song settings were the result of this attempt at collaboration.

At least one letter from Fauré to Samain (dated 13 March 1900) survives.-’ '*

Contact with Sully-Prudhomme involves a letter from Fauré to the poet’s promoter

Gaston Pans, in which Fauré asked permission for his setting of Sully-Prudhomme’s poem “Le long du quai” to be called “Les Berceaux.” Fauré mentioned in a letter dating from July 1879 to his friend Marie Clerc that Sully-Pnidhomme had written him ‘a most flattering letter’ granting authorization for the title “Les Berceaux.”3^

Meetings with Verlaine were arranged in January 1891 to discuss collaborations on a project for the stage. Nothing ever came of the negotiations, however, and by September of that year the project was dropped. They did not become close friends, and their closest association was to be in Fauré’s 17 mélodies on Verlaine’s poems.

Victor Hugo knew of the young Fauré, through Paul Meurice’s correspondence with

Hugo asking for Fauré’s rights to use Hugo’s poetry, even though the two did not meet.^®

Fauré may have met author Charles Van Leberghe, who “.. .was present at the London première of Pelléas etMélisandein June 1898, though he never mentioned meeting

Fauré.”^® To complete this category, one of Fauré’s good friends was the singer Romain

Bussine, translator of the two anonymous Italian verses Fauré set discussed earlier. Fauré intimately knew relatively few of his authors, a surprising fact in light of his fascination with setting native, contemporary verses.

35see Jean-Michel Nectoux, Gabriel Fauré: His Life Tbrougb His Letters, tr. J. A. Underwood, op. cit., p. 114 and p. 159.

3®Mentioned in Orledge, op. cit., p. 16.

See Jean-Michel Nectoux, Gabriel Fauré: His Life Tbrougb His Letters, tr. J. A. Underwood, op. cit., pp. 86-87.

^*See Nectoux, ibid., pp. 17-19.

^^Orledge, op. cit., p. 191, Footnote 1. 107

CLAUDE DEBUSSY

Debussy’s Educational and Literary Background

Claude-Achille Debussy ( 1862-1912) was bom to poverty-stricken parents. They did not have money enough to educate their five children, and Debussy’s mother even gave four of them (all but Claude) over to her sister-in-law to raise. His mother taught him

"...to read, write and count, and with this his formal education came to an end.”'*®

Debussy’s inadequate schooling caused him to have trouble with grammerand spelling into adulthood. At the age of seven he was taken to Cannes to stay with his aunt, Mme.

Roustan, where he began his musical studies. Musical training continued when he retumed to his parents, and in 1872 was admitted to the Paris Conservatoire. He was not prepared for this new life: “When at the age of ten he was admitted to the Conservatoire, his awkward and unsociable manner, not to mention his poor education, hardly made him attractive to his fellow students.”'*^ He remained a student there for twelve years, however, and finally won the Premier Grand Prix de Rome in 1884.

By then Debussy had come to know the Vasniers, an upper middle-class family who took him in as almost a family member. Debussy’s early literary development began in their house: “M. Vasnier suggested books that he should read and became his guide in matters literary and artistic. Many of Debussy’s early songs were composed for Mme.

Vasnier, using some of the recommended texts.

'*°Victor I. Seroff, Debussy: Musician o f France (New York: G. P. Putnam’s Sons, 1956), p. 18.

Edward Lockspeiser, Debussy, The Master Musicians Series (London: J. M. Dent, 1936, 1980), 5th edition, rev. Richard Langham Smith, p. 10.

^ ^Ib id , p. 21. 108

He found his two years in Rome stifling and terrible, and upon his return to France he sought the company of the Parisian literary and artistic community. He had begun to associate with writers before he left for Rome: “As far back as Debussy’s early years at the

Conservatoire in the 1870’s he was discontented with music as it was then being taught, and preferred the company of literary men to that of musicians.’"*^ He renewed his sense of artistic liberty upon returning from Rome: “He found his freedom in the circles of the avant-garde, that is to say the Symbolists, and the number and calibre of the writers he knew is very striking....’’^ His literary sensibilities found their fullest development through the many acquaintances from the Parisian salons and cafés of the 1880’s and

1890’s.

Debussy showed himself to be a man of letters in four distinct ways. First he read a great deal. Lesure describes the "... impressive. ..list of the authors he read: Bourget,

Banville, Poe, Rossetti, Swinburne, Charles Cros, Jules Bois and Maurice Bouchor, not forgetting Baudelaire and Laforgue, both of whom provided many a quotation in his letters.

Later on he was to add to these Neitzsche and Schopenhauer, Carlyle and Conrad, Gerhart

Hauptmaim, Dickens, Alexandre Dumas, Jules Renard and the songs of Paul Delmet."^^

Second, he wrote both prose as a music critic and poetry for some of his own mélodies.

Third, his many letters to friends contain numerous literary references. Cobb pointed out that, “Although his biographers and others have mentioned his use of quotations from poems in his writings and identified a few here and there, no one has indicated the frequency with which such quotations are found or noted how naturally they are included

Arthur B. Wenk, Claude D ebussy and the Poets (Berkely: University of California Press, 1976), p. 1.

■François Lesure, Introduction to Debussy Letters, selected and ed. François Lesure and Roger Nichols, tr. R. Nichols (Cambridge, Massachusetts: Harvard University Press), p. xv.

^^Ibid. 109 in a given text.”'*® She presented a selection of his letters in her book which illustrates this idea. Finally, Debussy showed his ties with literature in his musical compositions. Many of his works are based on or inspired by literary texts, and projects for his incomplete stage works were taken from a wide range of literary sources. The choice of texts for his later mélodies especially show his sensitivity and taste for quality literature.

Poets and poetry were in fact at the very core of his life: “It has been said of Debussy, in broader terms and repeatedly, that it was poets and not musicians who influenced his art at ever>' stage of his career.”'’^ Jarocinski concurred with this statement when he wrote:

“To the end of his days, [Debussy’s] entire imagination, as a man and as a composer, was steeped in poetry.”'**

Debussy’s Poetry Choices

Debussy chose the texts of four poets for many of his early songs from the 1880’s:

Banville, Musset, Leconte de Lisle, and Bourget. He dropped all four poets as choices by

1892. Debussy’s ten settings on texts from the Parnassian poet Banville all date from 1880 to 1882. The most famous of his Banville songs is “Nuits d’étoiles.” Rohinsky commented that, “This was no great poetry, yet poetry harmonious and nostalgic enough to inspire young Debussy’s developing instinct.Only one setting on a text by the

Romantic poet Alfred de Musset is included in this study, the “Rondeau” from 1882.

Debussy set two earlier Musset poems around 1879, but they are now lost. One of the two

Leconte de Lisle choices, “Jane,” was not published until 1966; the other one, “La fille aux

^^Margaret G. Cobb, The Poetic Debussy: A Collections o f His Song Texts and Selected Letters (Boston: Northeastern University Press, 1982), Preface, p. xviii.

'*^Marie-Claire Rohinsky, The Singer’s Debussy (New York: Pelion Press, 1987), p. 33.

'**Stefan Jarocinski, Debussy and bis Poets (Introduction to the complete recordings of Claude Debussy’s mélodies) (Le Plessis Robinson: Blanchard, 1980, EMI 2C 165-16.371/4), p. 24.

'*9/bit/., pp. 35-36. 110 cheveux de lin,” remains unpublished. The latter song inspired Debussy’s piano prelude with the same title, composed in 1910. Seven of the nine songs on poems by Bourget were set in the early 1880’s.

The most famous of these is the first one Debussy composed, “Beau soir,” written in

1881-2. The other two Bourget texts were set in 1891, the Deux romances. Bourget was greatly admired in his day; Rohinsky suggested that Debussy’s “...choice of Bourget’s poems had been prompted in part by the Vasniers, and most likely by Madame Vasnier’s predilection for them.”^° Jarocinski added that Bourget’s texts “...were used by Debussy as subterfuges for, in love with Madame Vasnier yet anxious to conceal his feelings from a jealous husband, he sought out the most tender texts, even at the risk of their being really trite. ”5*

Many of Debussy’s greatest songs were settings of the poetry of Paul Verlaine.

Debussy began setting Verlaine texts in 1882. Rohinsky wrote that, “He had read

Verlaine’s Fêtes galantes in the Vasnier library: So began an association that would last more than twenty y e a r s .” 52 Debussy’s first five Verlaine songs date from 1882: “Clair de lune,” “En sourdine,” “Fantoches,” “Mandoline,” and “Pantomime.” In 1885 he completed a setting of Verlaine’s “Chevaux de bois” which remains unpublished. A second setting of this text was included in his six Verlaine songs published in 1888 as

Ariettes, paysages beiges et aquarelles, reissued in 1903 as the series Ariettes oubliéesin

1903. These songs are “C’est I’extase,” “H pleure dans mon coeur,” “L’ombre des arbres,” “Chevaux de bois,” “Green,” zuid “Spleen.”

Three more Verlaine settings were composed in 1891, the Trois mélodies. That same year Debussy composed the first set of Fêtes gantes, which included “En sourdine,” a

p. 36.

Stefan Jarocinski, Debussy and bis Poets, op. cit. p. 21.

^^Rohinsky, op. cit., p. 37. I ll revised setting of “Fantoches,” and “Clair de lune.” The second set of Fêtes galantes

appeared in 1904; these were to be his final songs using texts from Verlaine. The three

songs in this set are “Les ingénus,” “La faune,” and “Colloque sentimental.”

Debussy also chose texts by Baudelaire, Hyspa, Le Roy, and Louÿs in the years before

the turn of the century, in addition to setting his own poems. He developed a strong

interest in Baudelaire while in Rome (1875-77), and several years after his return to Paris

(1887-89) composed the Cinq poèmes de Baudelaire. Debussy set a single poem from each of two minor poets who were acquaintances of his: “La belle au bois dormant” by

Hyspa in 1890 and “Les angélus” by Le Roy in 1892. Rohinsky concluded that, “The circumstances behind Debussy’s choice of these two poets have not been ascertained.”^^

Pierre Louÿs was one of Debussy’s closer friends; the settings of three of the Chansons de

B ilitis by Louÿs were composed in 1897-8. “It was with these songs of Louÿs that

Debussy’s power of poetic insight reached its zenith; the intimacy of the aesthetic communion between the two artists brought forth what was recognized by music critics as the most accomplished of Debussy’s songs.’’^'* Debussy’s concern for the relationship between poetry and music was further highlighted in his settings of his own verses, the four Proses lyriques of 1892-93.

The use of ancient French poetry came with Debussy’s final, mature songwriting years, beginning in 1904. Two texts by the fifteenth century poet Charles d’Orléans (“Le temps a laissié son manteau” and “Pour ce que Plaisance est morte”) and one by seventeenth century writer Lhermite (“La grotte”) were chosen for the Trois chansons de France, composed in 1904. Two more songs on texts by Lhermite (“Crois mon conseil” and “Je tremble en voyant ton visage”) were written and published with “La grotte” in 1910 under

^^Ibid., p. 41.

^^Ibid., p. 42. 112 the title Xe promenoir des deux amants. In the same year Debussy composed the Trois

Ballades de Vi/Zon^hree songs on ballads by the fifteenth century poet François Villon.

Three of Debussy’s final songs were the Trois poèmes deMallarméoT 1913, written as a tribute to his friend, the great poet Stéphane Mallarmé. His last song was a setting of his own text, “Noël des enfants qui n’ont plus de maison.” Rohinsky presented the details of this last, sombre work:

This last song of Debussy’s was written text and music, on the eve of the first operation performed against the cancer that was to take his life two years later. Tlie song was prompted by Debussy’s patriotic feelings and inspired by the invasion of Northern France by the German armies during World War I. Such circumstances explain the style of the poem and the tone of the melody. 55

The few other poets from whom Debussy chose texts not mentioned here are listed with his entire textual selections in Table 19, p. 243.

55io/cf., p. 287. 113

/ Table 7. The Songs of Claude Debussy

Descript >Sîi "* Number % Total number of Songs 77

Total number of Sources 20

Arithmetic average number of songs per source: 3.85

Median number of songs per source: 2.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 47 61.04%

Texts from poets contemporary with composer (bom within 15 years of composer) 20 25.97%

Texts from poets younger than composer (bom 15 years after composer) 0 0.00%

Texts from undated sources (dates of poet or source unknown) 2 2.60%

Texts from old sources (dated more than 100 years before composer’s birth) 8 10,39%

Texts from non-native sources (either in translations or in other languages) 0 0.0%

Approximate number of poets with whom the composer had personal contact: 6 114 Debussy’s Number of Songs and Sources

Two lists of Debussy’s vocal works were compar ed to determine the number of songs counted for this study. The first was from Margaret G. Cobb’s recent, thoroughly researched book on Debussy’s song texts, the “Chronological List of Songs.’’^^ The other was the work-list prepared for The New Grove Dictionary by Roger Nichols and Robert

Orledge.^^ These two catalogues agreed on Debussy’s songs, with one exception: the completed but unpublished first setting of Verlaine’s “Chevaux de bois” from 1885 is part of Cobb’s list but not in the other. Both versions of “Chevaux de bois” were counted in this study, while several lost wc-vks and sketches from Debussy’s works were not. A total of seventy-seven songs by Debussy were thus included.

Debussy’s textual sources for his songs were determined using the same two catalogues. Debussy selected texts from twenty sources for his mélodies. They are presented in Table 19, p. 243.

Debussy’s average is 3.85 songs per source, given 77 songs from 20 sources. His median average is 2.00 and the modal value is 1.00 These averages are almost identical to

Fauré’s, and fall in the mid-to-low range for the twelve composers. All but three of

Debussy’s sources were used five times or less, and only one of the three more frequent choices (Verlaine) was used more than ten times. Eight of the twenty sources were used only once. His occasional single choices from some authors and his multiple settings from other authors are both reflected in the averages presented. Like Fauré, Debussy selected a

fairly lowtaxmhet of texts from each of his sources.

5®Cobb, op. cit., pp. 2-5.

^^Roger Nichols, work-list with Robert Orledge, “Claude Debussy,” The New Gmve Twentieth-Century French Masters (originally published in The New Grove Dictionary o f Music and Musicians, ed. Stanley Sadie) (London and New York: Macmillan, 1980,1986), pp. 114-115. 115 Debussy’s Authors Categorized by B irth Date

The texts Debussy used are divided into groups dated by their authors’ birth years relative to the composer’s birth year of 1862, shown in Table 7. Forty-seven texts were selected from authors bom 16-100 years earlier than the composer, twenty texts were from authors bom within 15 years (plus or minus) of the composer, and no texts were from authors bom more than 15 years after the composer. These numbers are restated in percentages as follows: 61.04% of Debussy’s choices were from writers bom 16-100 years before him (between 1762 and 1846), 25.97% were from writers bom within 15 years of him (between 1847 and 1877), and 0.00% were from writers bom more than 15 years after him (after 1877).

Three of Debussy’s most frequently chosen authors, Verlaine (21 texts), Banville (10 texts), and Baudelaire (5 texts) were bom 18, 39, and 41 years before him, respectively.

This places them in the “older generation” category, with six other authors (of 11 texts).

Paul Bourget (9 texts) and Debussy himself (5 texts) are the two most frequently chosen authors in his “contemporary” category, with four additional authors contributing 6 texts.

The largest percentage of Debussy’s choices were from older poets, when divided into the generational groups established for this document. In contrast, Cobb^^ and

Rohinsky 59 have highlighted the number of “contemporary” poets among Debussy’s choices, without establishing a definition of the term. Rohinsky even cited Banville,

Leconte de Lisle, Mallarmé, and Verlaine as “contemporaries of Debussy,”5° poets bom from 18 to 44 years before him. It is true that these poets were all alive within Debussy’s life span, but they were not his contemporaries in the stricter sense of being roughly the composer’s age (i.e., bom within 30 years of 1862).

5®Cobb, op. cit., pp. xiv-xv.

59Rohinsky, op. cit., p. 34.

^°Ibid. 116

Several factors help to account for Debussy’s percentages. His peak songwriting years

were from 1879 to early 1887, while in his late teens and early 20’s. Most of the poets he

chose in these years were Parnassians, whose poetry flourished in the 1860’s, around the

time of Debussy’s birth. These authors were bom a generation earlier than Debussy, and

Debussy had learned of them in his years at the Paris Conservatoire. The nine Bourget

settings also date from the peak years; Bourget was one of Debussy’s friends and contemporaries.

Debussy did not find inspiration for song texts in the poetry of younger writers. In his

later years of songwriting he remained loyal to the poetic style of Verlaine and Mallarmé.

Rohinsky gave the following explanation for his lack of interest in the “new” poetry, and the effect it had on Debussy’s textual choices:

No new poetic school was looming in the early part of the twentieth century. Mallarme’s promising young friend Valéry was waiting for a propitious time to speak out. Yet many poets of this period were turning to a more concrete view of the world and a more straightforward means of expression. In keeping with the development of his sensitivity and the simplification of his style of writing, Debussy turned for inspiration to older forms of poetry. French poets o f the late Middle Ages and the Baroque era influenced the production of new mélodies, characterized by subtlety, restraint, and suggestive qualities.® •

Debussy looked backward instead of forward for the poetry choices of his later years.

He did not live long enough to witness the rise of the post-war Surrealist poets, and he may not have set their poetry even if he had lived longer.

Debussy’s Choices of Undated and Old Sources

Two of Debussy’s texts were from undated sources. The first was a poem by André

Girod for the early song “Fleurs des Blés,” composed c. 1880. Nothing is known about

Girod, except for this poem and one other, “Histoire d’Automne,” which was “...adapted

p. 31-32. 117

to music by Alexandre Georges in 1 9 2 2 .”^2 Debussy also set one anonymous poem about

a year later, the “Rondel Chinois,” which remains unpublished. These two songs from

undated sources are 2.60% of his total choices. They are the exceptions to his normal

settings of texts by known (and datable) authors.

The eight texts wliitli Dtbussy set from old sources (authors bom more than 100 years

before the composer) form a more substantial group than the undated texts. These settings

were the prominent products of his third or mature pe rod, his final years of song

composition. The three authors for Debussy’s choices in this category were Tristan

Lhermite (3 texts), François Villon (3 texts), and Charles d’Orléans (2 texts). These

choices are 10.39% of his total choices. Debussy experimented with a new, more refined,

classical approach to composition in his later years, and the French poetry of the Middle

Ages and Baroque were chosen as complements to this approach. The choices of old

poetry also came at a time when Debussy was involved with a re defining of French

aesthetics; both the title and choices of poetry of the Trois chansons de France were signs

of Debussy’s work in this area.

His settings of poems by Louÿs, the Chansons de Bilitis, also have a connection with

this category. The author originally claimed that these poems were translated from an

ancient Greek writer who was a contemporary of Sappho. Louÿs carried out this hoax for

quite a while, and although the poems were written by Debussy’s friend and contemporary,

they were nevertheless inspired by the author’s enthusiasm for ancient Greek literature.

Debussy’s Choices in Non-native Languages

Virtually none of Debussy’s textual choices were originally in another language. His strong commitment to setting works by French authors is borne out in this statement, as is his great respect for and love of the French language itself. There are occasional hints of

^^Ibid., Note 12, p. 306. 118 oriental, Spanish, and other foreign influences in his songs, but they are rarely obvious.

Two examples of this are in the unpublished “Chinois Rondel” with its quasi-oriental text and “La Romance d’Ariel,” Bourget’s text adapted from Shakespeare’s The Tempest.

Debussy knew the writings of foreign authors, as mentioned earlier, and several of his unfinished orchestral works and operas were based on translated texts. He also travelled some, to Russia, Italy, Germany, and England. Yet he was a purist when it came to his mélodie texts: “What sets Debussy’s songs apart is that they are all settings of French poems; this fact alone emphasizes his unusual sensitivity to the word arrangement, the music, the rhythms and nuances of the French language.None of the other composers in this study used strictly native-language texts. This makes Debussy’s choices literally stand alone in this category.

Debussy’s Personal Contact with the Poets

There are five of Debussy’s song text authors marked with an asterisk in Table 19, p. 243, indicating that they were in personal contact with him. The composer himself is marked as well, for he was both author and composer for five songs.

Debussy probably met three of the authors in the Parisian artistic circles of the 1880’s and 1890’s: Grégoire Le Roy, Vincent Hyspa, and Paul Bourget. His friendship with

Bourget was brief but vital; Bourget at the time became one of his literary mentors.

Through him Debussy formed opinions about Shakespeare, and “[i]t was probably also

from Bourget, who had travelled widely in England, that Debussy derived his interest in

other English p o e t s . Rohinsky added that Bourget “.. .probably introduced Debussy to

^^Ibid., p. 46.

^“^Edward Lockspeiser, Z>e6ussy; H is L ife and Mind, 2 vols. (London; Cassell, 1962, 1965), I, p. 68. 119 the poetry of Jules L aforge.”®^ Laforge’s work was influential to Debussy, although he never set any of Lafbrge’s poetry for his mélodies.

Debussy’s relationship with the great Stéphane Mallarmé was developed through

Mallarmé’s weekly salon gatherings, of which Debussy took part. “It is not known how

Debussy met Mallarmé, but it has been assumed that he had known him since 1890. To be admitted to the inner circle was in itself a sure sign of intellectual superiority.”^®

Lockspeiser, however, suggested that Debussy was not always comfortable with his intelligent companions: “.. .one suspects that he was more conscious of the shortcomings of his early education than his friends have given us to believe; and a certain sensitiveness on this point must have prevented him from frequenting the company of Mallarmé and his admirers, to whose sensibility he was instinctively drawn. Debussy’s friendship with

Mallarmé gave rise to the famous Prélude à l’après-midi d’un faune in 1892-4, an orchestral piece after a Mallarmé poem. Nearly twenty years later Debussy paid a tribute to the poet (who had been dead 15 years) with his Trois Poèmes de Stéphane Mallarmé, three settings of Mallarmé poems.

Debussy’s most intimate and long-lasting relationship with any of his poets was with

Pierre Louÿs, which began in 1893 and continued for nearly a decade. Rohinsky summarized this partnership:

It was shortly before the composition of Proses Lyriques [ 1892-93] that the poet Pierre Louÿs is said to have discovered Claude Debussy, possibly at the Auberge du Clou (the Nail Inn) or at the Librairie de Tart indépendant (Independent Art Bookstore) or at one of Mallarmé’s gatherings. There has been disagreement on the part Louÿs played in Debussy’s career, precisely on his influence on Debussy’s aesthetic development. Certainly he provided Debussy with both moral and financial support, as their correspondence

®®Rohinsky, op. cit., p. 36.

p. 29.

Edward Lockspeiser, Debussy, The Master Musicians Series, op. cit., p. 44. 120 shows: Debussy was forever impecunious and Louÿs generous. Whatever the case, friendship struck with them and lasted a good many years.^®

The two eventually drifted apart, around the turn of the century: “The intimacy ot the association between poet and musician could not be sustained after their respective marriages.

There has been much speculation about whether or not Debussy personally knew

Verlaine. Lockspeiser speculated on this issue:

The fact that, depsite exhaustive research, no mention can be traced of a meeting in later years between Debussy and Verlaine is a mystery, but it may also be significant. One would expect there to have at least been some correspondence on the matter of the rights of setting Verlaine’s poems, similar to the published correspondence between Fauré and Verlaine. Had something occurred which caused them to avoid each other? Since they had many mutual friends they could hardly have done so successfully.’^

In the same paragraph, Lockspeiser presented a strong argument for the idea that as a child Debussy was witness to the difficult end of Verlaine’s marriage to Mathilde, the daughter of Madame Mauté, Debussy’s childhood piano teacher. The lack of concrete evidence that the two knew each other in later years, hov/ever, keeps Verlaine’s name from being included in the list of Debussy’s personal contacts for this document.

®®Rohinsky, op. cit., p. 41.

®^Wenk, op. cit., p. 173.

’^Edward Lockspeiser, Debussy: His Life and Mind, op. cit.. I, pp. 22-23. 121

FRANCIS POULENC

Poulenc’s Educational and Literary Background

Francis Poulenc (1899-1963) was bom into a cultured, well-to-do Parisian family, and was well-educated. He “...received a thorough academic training at the Lyceé

Condorcet.”^^ He did not attend one of the established French music schools, although his musical studies had begun at home at the age of six:

Both Poulenc and his mother wished him to enter the [Paris] Conservatoire, but his father’s preference for a general education prevailed, and he matriculated at a normal lycée, achieving a baccalauréat, without distinction, in 1917.72

His formal education was finished at the age of eighteen, and he never retumed to academic life.

Poulenc’s cultural education was fostered throughout his childhood by his family.

“From his mother the young Francis absorbed a love of poetry and painting, as well as of literature, drama, ballet, and cinema. ”73 In addition, Poulenc’s uncle “Papoum” took him to many artistic events as a child.

Poulenc’s childhood friend Raymonde Linassier brought him to his first-hand, deep understanding of contemporary literature:

Passionately fond of literature, she introduced Poulenc to the works of Claudel and Gide and later Proust and Joyce, and together they explored Verlaine, Mallarmé and especially Baudelaire. These early poetic experiences of Poulenc—he had known Mallarmé’s Apparitionhy heart at the age of ten—

71 Roger Nichols, “Francis Poulenc,” TbeNewCrove Twentietb-Century French Masters (originally published in The New Grove Dictionary o f Music and Musicians, ed. Stanley Sadie) (London and New York: Macmillan, 1980, 1986), p. 197.

72Keith W. Daniel, Francis Poulenc: His Artistic Deveiopment and Musical Style (Ann Arbor: UMI Research Press, 1982), p. 9.

"^^wid., p. I. 122 were calculated to develop his remarkable sense of musical prosody. In the meantime Raymonde Linassier and Francis Poulenc were frequently to be seen together at Adrienne Monnier's bookshop A u x Am is des Livies, in the Rue de rOdéan, the rendezvous of a literary élite that included Valéry, Gide, Claudel, Joyce, Valéry Larbaud, Apollinaire and Léon-Paul Fargue. Poulenc’s friends in this literary circle were André Breton, Louis Aragon and Paul Eluard, whose poems were later to inspire him.'^'^

Daniel reinforces the importance of this literary scene in Poulenc’s artistic development:

“His destiny as the leading composer of the French art song after Ravel was forged in

Moimiefs bookstore during the years of World War

Ricardo Vines, Poulenc’s piano teacher from 1914 to 1917, also introduced the young musician to many people in the Parisian artistic world. The most important of these introductions were to composer Erik Satie and writer Jean Cocteau, the two men responsible for Poulenc’s growing public visibility as a member of the group of young composers called Les Six, which flourished from 1917 to 1921. Cocteau in these years

“.. .became Poulenc’s cultural guide and mentor, introducing him to Count Etienne de

Beaumont and his wife, as well as to many other society figures and patrons; taking him to art exhibitions, concerts, poetry readings, and his newest passion, films; and making sure that his artistic contacts multiplied.”^^

Poulenc was fortunate to have this guidance from childhood until he was fully established as a major composer and performer. He continued to stay involved with the artistic and cultural world through the rest of his life. He wrote several books, including a biography of Emmanuel Chabrier and two autobiographical volumes, as well as numerous articles on musical subjects. More importantly, his understanding of 20th century French poetry was profound, and this knowledge was made manifest through his mélodies.

Claude Rostand once wrote that, “It is Poulenc who should be consulted by those who do

^'*Henri Hell, Francis Poulenc, tr. Edward Lockspeiser (New York: Grove Press, 1959), p. 6.

^ ^Daniel, op. cit., p. 11.

'^^Ibld., p. 16. 123 not entirely understand the meaning of the poetry of Max Jacob, of Jean Cocteau, of Louise de Vilmorin, and above all Guillaume Apollinaire and Paul Eluard; in his songs he uncovers all the mysteries."^^ Poulenc may not have had a long career as a student, but his cultural training made him one of the most cultured and literary composers of this study.

Poulenc’s Poetry Choices

Poulenc’s first choice of poet for his mélodies was the Surrealist writer Guillaume

Apollinaire, for the six songs of Le Bestiaire in 1919. Poulenc continued setting

Apollinaire’s verses for nearly 40 years, and many of these are considered some of his very best vocal works. Poulenc retumed to Apollinaire in 1931 for his Quatre poèmes and again in 1938 for the Deux poèmes and the single song “La grenouillère.” One of the Trois poèm es de Louise Lalanne, also composed in 1931 was actually written by Apollinaire (see the final paragraph of this section). One Apollinaire text was set in 1939, “Bleuet,” followed by five more in 1940, the Banalités. Two Apollinaire settings were completed in

1945, “Montparnasse” and “Hyde Park,” and two in 1946, “Le pont” and “Un poème.”

The seven Ca%/ammes followed in 1948. Poulenc thought these would be his last

Apollinaire songs, but two more were eventually written: “Rosemonde” in 1954 and “La souris” in 1956.

Poulenc did not begin setting texts by Paul Eluard until 1935, even though he “...had been attracted to Eluard’s poetry since adolescence.... ”7» The first five of Poulenc’s 34

Eluard settings were simply titled Cinq poèmes. These were followed two years later by the nine songs in the cycle Tel jour telle nuit. The two songs of Miroirs brûlants and the

^^Quoted in Pierre Bemac, Francis Poulenc: The Man and His Songs, tr. Winifred Radford (New York: Norton, 1977), p. 39.

Roger Nichols, “Francis Poulenc,” The New Grove Twentieth-Century French Masters, op. cit., p. 209. 124 single song “Ce doux petit visage” were composed in 1938. No further Eluard settings

were completed until 1947, when the two songs “...Mais mourir” and “Main dominée par

le cœur” appeared. Two major cycles on Eluard poetry were finished in 1950 and 1956,

the seven songs in La fraîcheur et le feu and the seven of Le travail du peintre. Poulenc

commissioned Eluard to write the poems for the latter work, which are verses on the lives

of seven important modern painters. Poulenc’s final Eluard setting was the song “Une chanson de porcelaine” of 1958.

Poulenc’s largest choices of texts from single authors after Eluard and Apollinaire were

from Louise de Vilmorin, Maurice Carême, and Max Jacob. His settings by Vilmorin

began in 1937, with the Trois poèmes. His popular set of six songs on Vilmorin texts

called Fiançailles pour rire were set in 1939. Other Vilmorin choices were for the three

Métamorphoses of 1943 and the “Mazurka” of 1949. The seven songs he set on child like

texts by Carême were chosen for his last song cycle. La courte paille, composed in 1960

“.. .for Denise Duval to sing to her little boy of six.”^9 Five of Poulenc’s seven choices

from Max Jacob were published as Cinq poèmes in 1931; the other two were written in 1954 and titled Parisiana.

Smaller numbers of texts were chosen from Poulenc’s poet friends: six from

Fombeure, three from Cocteau, two each from Aragon and Desnos, and one each from

Colette and Radiguet. The six texts from Maurice Fombeure were set in 1942 and called

Chansons villageoises; these songs were orchestrated in the same year. The three Cocteau poems were chosen for the early set Cocardes, written in 1919. Two poems by Aragon were set in 1943, which Poulenc called Deux poèmes d’Aragon. The first of these two is the emotional, popular song with a French nationalist theme, “C.” Poulenc’s two settings of poems by Robert Desnos are “Le disparu” of 1947 and “Dernier poème” of 1956. The

■^^Francis Poulenc, Diary o f my Songs (Journal de mes Mélodies), tr. Winifred Radford (London: Victor Gollancz, 1964, tr. 1985), p. 109. 125

only Colette poem Poulenc set was “Le portrait” of 1938, and the only Radiguet text he set

was “Paul et Virginie” of 1946.

Poulenc worked with a number of texts from earlier French poets, begirming with the

five Poèmes de Ronsard in 1924-25. A year later he set eight anonymous texts from the

seventeenth century, which he called Chansons gaillardes. Finally, four single choices

from earlier writers were set in subsequent years. In 1930 he set Malherbe’s “Epitaphe,”

in 1935 he returned to Ronsard with “A sa guitare,” in 1938 he set the “Priez pour paix”

by Charles d’Orléans, and Racine’s “Hymne” was set in 1947.

Several of Poulenc’s textual choices were out of the ordinary. For example, in 1927-

28 he set four poems by Jean Moréas, a poet which he detested, as an act of mutilation! To

his chagrin, the A irs cAantés became quite popular. The Trois poèmes de Louise Lalanne

of 1931 are unique in that there was no such poet. Apollinaire and his mistress Marie

Laurencin created the name as a hoax, and two of the three poems Poulenc set are by

Laurencin and the other is by Apollinaire. The Polish texts of the Huit chansons polonaises

are unlike his other textual choices, and will be discussed later under his texts from Non­

native sources. Poulenc’s final exceptional poetry choices were the three translated

Spanish poems by Garcia Lorca poems he set in 1947, simply called the Trois chansons de

F. Garda Lorca. These will also be mentioned again under the Non-native sources category.

The complete list of Poulenc’s sources for his textual choices, including the few not mentioned here, is presented in Table 20, p. 244. 126

y Table 8. The Songs of Francis Poulenc

Description ______Number______%

Total number of Songs 144

Total number of Sources 25

Arithmetic average number of songs per source: 5.76

Median number of songs per source: 2.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 49 34.03%

Texts from poets contemporary with composer (bom within 15 years of composer) 74 51.39%

Texts from poets younger than composer (bom 15 years after composer) 0 0.00%

Texts from undated sources (dates of poet or source unknown) 12 8.33%

Texts from old sources (dated more than 100 years before composer’s birth) 9 6.25%

Texts from non-native sources (either in translations or in other languages) 11 7.64%

Approximate number of poets with whom the composer had personal contact: 13 127 Poulenc’s Number of Songs and Sources

There are 144 Ppulenc mélodies counted for this study. Three lists of his vocal works were consulted to determine this number: Poulenc’s own list in his Diary o fMy Songs

(Journal de m es M é lo d ie s ) ,Bemac’s list in The Interpretation o fFrench Song,^^ and the list of Poulenc’s works prepared by Roger Nichols for The New Grove DictionaryP- Six songs Poulenc denoted as “chansons” are not included in this study: “Toréador,” the

Quatre chansons pour enfants, and the song “Les chemins de l’amour.” The 1927

“Vocalise,” which has no text, is also deleted from the present list. The early works Le

BestiareaaA Cocardes were composed simultaneously in versions with instrumental ensemble and piano accompaniment, and Poulenc included them in his list of mélodies.

They are therefore both counted in this study.

Poulenc selected verses from 25 sources, presented in Table 20, p. 244. Most of the sources were given in the three lists cited in the previous paragraph, except for the authors of the individual Polish texts of the Huit chansons polonaisesand the translator of the

Lorca poems Poulenc set, which were found in the appropriate published music scores.

The arithmetic average for Poulenc’s 144 songs from 25 sources is 5.76 songs per source. The median average is 2 and the modal value is 1. The first of these averages

ranks high (third) among the twelve composers, while the other two averages are on a par with most of the twelve composers. The modal value is fairly weak ( 10 of 25 sources used

one time each), Poulenc’s frequent settings of texts by Eluard and Appolinaire are reflected

in the high arithmetic average. These are in contrast to the fewer choices from the other

80 76/d, pp. 127-134.

8* Pierre Bemac, The Interpretation o f French Song (New York: Norton, 1970), pp. 270-273.

8^Roger Nichols, “Francis Poulenc,” The New Grove Twentietb-Century French Masters (originally published in The New Grove Dictionary of Music and Musicians, ed. Stanley Sadie) (London and New York: Macmillan, 1980, 1986), pp. 214-215. 128 twenty-three sources, seen in the low median and modal averages. Poulenc set a relatively high number of texts from two authors and a relatively lowmxmbtt of texts from the majority of his sources. His textual choices are quite similar to Schubert’s when analyzed in this maimer.

Poulenc’s Authors Categorized by Birth Date

Table 8 shows that most of Poulenc’s poetry choices were from contemporary authors.

He set 49 texts by authors bom 16-100 years earlier than 1899 (his birth year), 74 texts by authors bom within 15 years (plus or minus) of 1899, and no texts by authors bom more than 15 years later than 1899. These three groups are restated as percentages as follows:

34.03% of his choices are from authors bom 16-100 years before him (between 1799 and

1883), 51.39% are from authors bom within 15 years of 1899 (between 1884 and 1914), and none are from authors bom after 1914.

The percentage of Poulenc’s choices from older generation authors is raised considerably by the 33 Apollinaire texts. Apollinaire is placed in this category using the method for dating authors relative to the composers outlined for this paper. He was bom

19 years before Poulenc and died when Poulenc was 19 years old. The other authors in

Poulenc’s older generation group include Max Jacob, Jean Moréas, Colette, and the four

Polish authors from the Huit chansons polonaises.

The four Polish authors were all living at the time of the 1830-31 Polish Insurrection, and wrote verses which became popular with the Polish people during that period.

François Kowalski was a Polish officer at the time; Raynold Suchodolski was both a poet and a soldier who was killed in the uprising. Maurice Goslawski defected to Russia after the Insurrection and Stefan Witwicki was a popular poet of the time who wrote in the

Romantic style. 129

Except for Poulenc’s choices from Undated and Old sources (discussed later), the rest of his choices were from eleven contemporary authors. He personally knew all of them except Laurence de Beylié. Graham Johnson wrote that “Poulenc was proud of his.. .knowledge of twentieth-century T .ench literature.. .and his friendship with a number of his poets.’’*3 Poulenc’s large number of choices from contemporary authors for his mélodies demonstrated his pride in a concrete way. His percentage of choices from contemporary authors is exceeded only by Duparc’s for this study.

Poulenc’s Choices of Undated and Old Sources

Poulenc used twelve Undated texts, which are 8.33% of his total choices. They are divided into two groups, the first is the set of eight anonymous verses from the seventeenth century he set in 1926, the Chansons gaillardes. He found these somewhat obscene texts in tin anthology of seventeenth century verses. Poulenc thought that one of them,

“Invocation aux Parques’’ might be by the French poet Jean Racine (1639-99).^ The second group of anonymous verses Poulenc used were the four Polish texts among the

Huit chansonspolonaisesof 1934 (nos. 3,4, 7, and 8). These texts by unknown authors were popular in Poland during the time of the Polish Insurrection of 1831. This work will be mentioned again in the Non-native Sources category.

Poulenc was not as involved with setting folk poetry as Brahms or Wolf, but his percentage of choices from undated sources is the highest of the four French composers.

This is noteworthy, considering the high number of contemporary texts he set from authors he personally knew.

®^Graham Johnson, “Foreward to the English Edition,” Francis Poulenc, Diary o f my Songs (Journal de mes Mélodies), tr. Winifred Radford, op cit., p. 12.

^^Ibid., p. 25. 130

Poulenc composed nine songs using texts from four important early French authors.

These represent 6.25% of his total choices, a mid-range percentage for the twelve composers studied. Unlike Debussy, who worked with texts from ancient writers primarily in his later life, Poulenc set texts from ancient sources throughout his career. The six settings of Ronsard’s sixteenth century poems were Poulenc’s largest choice in this category. The single selections from the works of Charles d’Orléans, François de

Malherbe, and Jean Racine show that Poulenc was not overly concerned with multiple settings of the works of these ancient French writers. His choice of eight anonymous texts from the seventeenth century, the Chansons gaillardes, are also linked by date to this category. Poulenc’s choices of old texts continued the trend established by early twentieth century song composers (as represented here by Debussy’s choices) in setting ancient

French verses.

Poulenc’s Choices in Non-native Languages

Poulenc set two groups of texts originally in non-native languages. The first of these is the set of eight Polish poems he set in 1934, the Huit chansons polonaises. The published score of Poulenc’s settings show both the original Polish texts and French adaptations by

Jacques Lerolle. Four of them were by unknown authors, and the authors of the other four were mentioned earlier. Poulenc worked with these verses on a request from the Polish singer Marya Modtakowska (1896-1965), who wanted to sing them on a concert tour of

Morocco.

The second group of texts from a non-native source are the three Spanish poems by

Federico Garcia Lorca, which Poulenc set in 1947 using a French translation by Félix

Gattegno. These songs are not in the Spanish musical style, since the choice of these poems had more to do with Poulenc’s admiration for Lorca than with his interest in

Spanish literature or foreign musical styles. 131

Poulenc’s eleven selections from non-native sources are 7.64% of his total choices.

His percentage in this category is the largest for the French composers, and mid-range for the twelve composers studied. He essentially remained loyal to his use of French poetry, even though the numbers may suggest a slight deviance away from native sources. The two groups of non-native texts detailed above were not at the heart of Poulenc’s textual choices; they were more distractions from his more important settings of twentieth century

French poetry.

Poulenc’s Personal Contact with the Poets

Poulenc knew at least 13 of the poets from whom he took verses for his mélodies.

They are marked with an asterisk in Table 20, p. 244. These authors are identified as having been in personal contact with Poulenc based on a statement Pierre Bemac made in his book Poulenc: The Man and His Songs, regarding the first fourteen of Poulenc’s authors covered: “.. .with the exception of Laurence Beylié, [they] were all personally known to him .... Poulenc.. .confessed that he only felt musically at ease with poets whom he knew personally. ”85

He met all of these writers in his early years, during the time he frequented the company of the Parisian artistic and literary community. A few of them were especially important to him and his work. Cocteau, as mentioned earlier, was instrumental in presenting Poulenc to the general public. Another exzunple is the painter Marie Laurencin, who was Apollinaire’s mistress for a while and who wrote two of the verses attributed to

Louise Lalanne which were set by Poulenc. She was important to Poulenc as the designer for his ballet Les Biches, v/hich premiered in 1924. For this production, “It is possible that her name was suggested by Poulenc himself, as she was already known to him. She made

85pierre Bemac, iTanc/s/’ou/e/Jc; The Man and His Songs, tr. Winifred Radford (New York: Norton, 1977), p. 202. 132 contact with him in 1921, writing to compliment him on his musical settings for

Apollinaire’s poems from Le Bestiare ou cortège d’Oiphée.”^^

Poulenc knew and revered both Apollinaire and Eluard, whose poems he chose most frequently. He met Apollinaire in 1917 and knew him for only a brief time, for Apollinaire died a year later. Poulenc stayed in contact with Eluard for a much longer period of time, however. They met in the 1910’s, and their collaborations culminated in 1956 with the creation of Le travail du peintre, seven songs on the lives of painters for which Poulenc commissioned Eluard to write the verses. Poulenc was so dedicated to setting to music the poetry of these two authors that he wrote: “If on my tomb could be inscribed: Here lies

Francis Poulenc, the musician of Apollinaire and Eluard, I would consider this to be my finest title to fame.”®'^ With this type of statement, Poulenc readily admitted his wish that a part of his lasting legacy as a song composer could be from his collaboration with these poets and his insight into their texts.

®®Charlotte Gere, Marie Laurenc/jj (New York: Rizzoli, 1977), p. 22.

®^Ffaiicis Poulenc, Diaiy o f m y Songs (Journal de mes Mélodies), tr. Winifred Radford, op cit., p. 69. 133

SUMMARY FOR THE MELODIE COMPOSERS

Mélodie Composers' Educational and Literary Background

The amount of formal education that each of the four mélodie composers received varied greatly, from a full course of study for Duparc (including legal studies) and a good primary education for Fauré and Poulenc, to no formal education for Debussy. The four received varying amounts of formal musical training as well. Neither Duparc nor Poulenc studied at a music conservatory, while Fauré and Debussy trained for many years in this type of environment.

All four composers were acutely aware of the artistic climate of their time. They each read a great amount of literature by French and foreign writers. They also personally knew many important French writers, artists, and composers of their day. They each were a part of a cultural environment which helped shaped their opinions about artistic values and led them to the poetry choices for their mélodies.

Each of the four mélodie composers demonstrated his literary understanding in visible ways. Fauré and Debussy were both music critics, and Poulenc wrote many books and articles on music. Duparc was known to have been a multi-faceted aesthete, who could offer artistic advice to even the great Wagner. Debussy was probably the most “poetic” of the four composers, as evidenced in the poetic references in his letter and the poetry he wrote for his own mélodies.

Mélodie Composers' Poetry Choices

There are some generalizations which can be made about the texts chosen by the mélodie composers, based on the information gathered for this study. First of all, the 134

majority of their texts were taken from writers who were their contemporaries or from the generation of writers bom immediately before them. Second, they did not choose large amounts of undated or ancient texts. Third, their choices highly favored texts in French, their native language. Further discussion of these ideas, including the numbers which suggest them, will be presented later under the aporopriate categories.

Each of the four mélodie composers, like the four Lied composers, chose the texts for his songs in an individual manner. Duparc mixed choices from obscure authors with those

from more important writers, never really concentrating on the texts of any one author for multiple settings. Fauré went through three periods of text selection, beginning with

Romantic texts, then moving to the works of Verlaine and the Parnassians, and ending with texts by younger generation authors. Debussy’s compositions and textual choices were similarly divided into three groups, first with choices from authors such as Musset and

Banville, later to choices from the works of Verlaine, and ending with choices from French

Medieval and Baroque poets.

Poulenc’s choices were less predictable. He selected a group of texts or a single text from a given author for setting at any one time, depending on his inspiration. He returned to some authors several times throughout his life, unlike any of the other French composers studied. Finally, while most of his choices were from twentieth century authors, he also chose a small but noteworthy group of other texts, including those by ancient poets and some in foreign languages. 135

Mélodie Composers’ Number of Songs and Sources

The number of songs by each of the four mélodie composers counted for this study range from 15 by Duparc to 144 by Poulenc. The totals are shown in Figure 12.

Duparc 15 Fauré 102 Debussy 77 Poulenc 144 I I I I I I I I I 0 20 40 60 80 100 120 140 160

FIGURE 12: MELODIE COMPOSERS Number of Songs

The mélodie composers were not as prolifîc as the Lied composers. Several reasons account for the relatively small output of mélodies. Duparc’s systematic destruction of all but a few works stemmed from a compulsive perfectionism; the nervous disorder which prematurely ended his composing career further prevented a large output from him. Fauré was constantly limited in the amount of time he could devote to composition, because of the responsibility of professional duties throughout his adult life. Debussy’s time spent on works in other genres, especially on orchestral works, piano works and his opera Pelléas et

Melisandetook him away from mélodie composition. This was particularly troe in the later part of his life, after the 1890’s. Poulenc was the most prolific of the four composers in

Chapter II, but even he was distracted by writing works in other genres, especially works for the stage. 136 The number of sources the four mélodie composers used ranged from 10 sources for

Duparc to 26 for Fauré.

Duparc Faure Debussy Poulenc I I I I I I I 0 5 10 15 20 25 30

HGURE 13: MELODIE COMPOSERS Number of Sources

The numbers presented in Figure 13 were used to determine the arithmetic average number of songs per source for these composers, shown in Figure 14,

Duparc 1.50 Fauré 3.92 Debussy 3.85

I I I I I I I 0 1 2 3 4 5 6

FIGURE 14: MELODIE COMPOSERS Arithmetic Average Number of Songs Per Source

The arithmetic averages for the mélodie composers are somewhat low in the twelve- composer comparison; they are lower as a group than this average for the four Lied composers. Duparc’s is the lowest arithmetic average of any of the twelve composers; his consistent use of small numbers of texts from his authors are seen in this average. Fauré 137 and Debussy selected relatively more texts from fewer sources, as seen in their averages in

Figure 14, where the larger number of texts from some authors are counteracted by the small number of texts from others in their averages, Poulenc’s arithmetic average is the highest for the mélodie composers, and one of the highest of the twelve composers studied.

His large number of choices from a few sources significantly offset the small number of

choices from other sources.

An examination of Tables 17-20, pp. 241-244, provides the median averages and

modal values for the mélodie composers. These averages are both presented in Figure 15.

Duparc 1.00 both values Fauré 2.00 A A A * 1 im Modal Value Debussy 2.00 Poulenc 2.00

1

HGURE 15: MELODIE COMPOSERS Median Average and Modal Value of Songs Per Source

These two averages show the consistently high number of authors from whom the

French composers selected one or two verses. The French composers showed a high

amount of experimentation in their textual choices, as suggested in these averages, as well

as an overall lack of commitment to using large numbers of texts by many of their authors.

Mélodie Composers’ Authors Categorized by Birth Date

The French mélodie composers are less consistent as a group than the Lied composers

in their percentages from authors bom 16-100 years before them. Fauré’s percentage here 138 is a low 12.75% (the lowest of the twelve composers), while Debussy’s is a high 61.04%.

Figure 16 shows this inconsistency.

Duparc 40.00% Fauré ^ 12.75% Debussy 61.04% Poulenc 34.03% n " T I I I I I r I "f I I I Percent 0 5 10 15 20 25 30 35 40 45 50 55 60 65

FIGURE 16: MELODIE COMPOSERS Percentage of Texts from Authors Bom 16-100 Year? Earlier than Composer

Some explanations for the percentages in this category were given for each composer previously, in the individual discussion of this category.

The percentages of texts from authors bom within 15 years (plus or minus) of the

French composers are given in Figure 17.

Duparc 60.00% Fauré R 54.90% Debussy 25.90%

Poulenc 51.39%

— n 'T T I I I I I I I I I I Percent 0 5 10 15 20 25 30 35 40 45 50 55 60 65

FIGURE 17: MELODIE COMPOSERS Percentage of Texts from Authors Bom Within 15 Years (Plus or Minus) of Composer 139

These percentages, with the exception of Debussy’s, show the mélodie composers to be somewhat consistent in their strong intentions of using . 'temporary poetry.

The third percentage in this category shows the extent to which the French composers drew from poets bom more than 15 years after the composers, presented in Figure 18.

Duparc 0% ■m Fauré ?4.51%

Debussy 0% Poulenc 0% I I I I I I Percent 0 5 10 15 20 25

HGURE 18: MELODIE COMPOSERS Percentage of Texts from Authors Bom More than 15 Years After Composer

Fauré was alone among the French composers in using poetry from younger generation writers. Duparc stopped composing before this was viable for him, Debussy tumed to ancient poetry rather than the works of younger writers late in his composing years, and

Poulenc stayed with the authors who were his contemporaries (or older) to the end of his life. 140

Mélodie Composers’ Choices of Undated and Old Sources

The mélodie composers were not interested in setting large amounts of anonymous poetry (Figure 19).

Duparc Faure 7.84% Debussy 2.60%

Poulenc 8.33%

I I I I I I I I I I I Percent 0 123456789 10

FIGURE 19: MELODIE COMPOSERS Percentage of Texts from Undated Sources

None of them set more than 10% of their songs from undated texts. Much of the material in this group for the French composers is from identified authors about whom little or nothing is known, rather than from anonymous texts. Poulenc’s choices in this group are the only ones by anonymous authors. 141

The mélodie composers’ percentages of texts from authors bom more than 100 years before them are shown in Figure 20.

Duparc Faure Debussy 10.39%

Poulenc 6.25% I "j y r I 1 T' r I "I T Percent 0 123456789 10 11

HGURE 20: MELODIE COMPOSERS Percentage of Texts from Authors Bom More Than 100 Years Before Composer

Neither Duparc nor Fauré chose any texts in this group, as opposed to Debussy and

Poulenc, who both chose relatively small amounts of ancient poetry. As stated earlier, this

trend of setting ancient French poetry grew after the end of the nineteenth century. These

texts served as a substitute for new poetry in Debussy’s choices and as a diversion from

contemporary poetry for Poulenc. 142

Mélodie Composers’ Choices in Non-native Languages

Figure 21 shows that the French composers studied used a small amount of texts in languages other than French, or none at all.

uuparc 6.67% Faure 1.96% Debussy Poulenc 7.64% I I I I I I I I I Percent 0 12345678

FIGURE 21: MELODIE COMPOSERS Percentage of Texts in Non-Native Languages

Debussy was the only one of the twelve composers who set no texts in a language other than his native French. Duparc chose one poem from an English translation, Fauré used translated Italian verses, and Poulenc chose texts originally in Spanish and Polish, all of which were set in French translations (Poulenc’s Polish songs include both French and

Polish texts in the score).

The mélodie composers’ lack of significant foreign-language settings were not due to their disregard for non-native texts. They were all exposed to foreign materials, and they read works by foreign authors. French translations of foreign language texts became available in the late nineteenth and early twentieth centuries through the persistence of some

French writers, who “.. .encouraged literary exchange and saw to it that foreign works were published in literary reviews. Until the First World War, however, these exchanges 143 had remained sporadic and subject to the whims of personal tastes or connections, the hazards of chance readings or translations.”**

The four of them also knew the music of foreign composers, past and present. In addition, each of them personally knew many contemporary composers outside of France and travelled abroad.

Mélodie Composers’ Personal Contact with the Poets

All four French composers were in contact with some of the authors of the texts they chose for their mélodies.

Duparc Faure Debussy Poulenc 13

0 2 4 6 4 8 10 12 14

FIGURE 22: MELODIE COMPOSERS Approximate Number of Authors with whom the Composer had Personal Contact

As seen in Figure 22, Duparc and Fauré were in contact with fewer of the authors of their song texts than Debussy and Poulenc. Each of the composers knew more writers than are shown in Figure 22, but did not choose to set their poetry to music. Debussy and

Poulenc were quite close to several of their authors, and these associations were as significant to the French composers as this type of relationship was for Schubert.

**Germaine Brée, Twentietb-Centuiy French Literature, tr. Louise Guiney (Chicago and London: University of Chicago Press, 1983), p. 86-87. CHAPTER III

TEXTUAL SOURCES FOR THE AMERICAN ART SONG COMPOSERS

Introduction to Chapter III

A definition of “American art song” by Ned Rorem serves as an appropriate

introduction to this chapter

The setting is by a specific American composer as opposed to anonymous or collective authorship; is self-contained as opposed, say, to an aria, which is part of a whole; and is strictly as opposed to approximately notated like so- called popular songs..., which can be rendered by any voice in any arrangement at any speed. ‘Art song’ is our answer to the German L ied or to the French mélodie, which implies a through-written recital song as distinct from a chanson.^

The composition of “art songs” in the United States began to blossom in the decades before the beginning of the twentieth century. There was little in the way of a serious songwriting tradition in America before the middle of the nineteenth century. The main body of American vocal music before then consisted of folk songs and parlor songs, as well as popular tunes taken from ballad and pastische operas in the 18th century and minstrel shows in the 19 th century. Much of this popular music was derived or imported from European vocal music, as were the earliest art songs in America.

One composer who used his European-style training to create a personal, distinctly

American style of song in the years immediately before the start of the twentieth century was Charles Ives. He is often considered to be the first truly “American” song composer

“Although his output was not large in comparison to Schubert’s, Ives wrote songs that

*Ned Rorem, “The American Art Song,” Setting the Tone (New York: Coward-McCann, 1983), p. 225.

144 145 stand at the center of American song jest as those of Schubert are at the center of the

German lied. Ives's^songs are a national treasure, both on purely musical grounds and as a cultural effusion of exceptional richness.”^

Ives was only one of many song composers who went beyond European convention.

Donald îvey claimed that the influences of nineteenth-century European song were uniquely translated into the composers’ song styles of many countries, including the United States:

In America, the links with the past have been similar to those elsewhere in that there are some song composers who have assimilated much of the subject matter and harmonic-melodic approach of the previous century without entirely denying their own. Typical of this type of carry-over is the work of Charles Griffes and John Alden Carpenter, with their empathy for the impressionistic vocabulary... .Perhaps of greater significance is the work of Samuel Barber, whose more recent songs absorb a good many of the rhythmic, melodic, and harmonic techniques of the century without at the same time eliminating the tunefulness of the vocal line. 3

Samuel Barber was less concerned with creating a purely American song style than Ives was. When asked in 1978 if there is an American song tradition. Barber denied it with this casual reply:

Way back there [was] Stephen Foster, and there [was] all of that tum-of-the century salon stuff which tended to be Germany-oriented, but I don’t think that constitutes a real American tradition. But is there any real American symphonic tradition, for that matter? Composers like MacDowell, Parker, Chadwick, and Foote derived from either German or French music.

American art song was indeed slow in rise to prominence. Music publishers were reluctant to print this type of work, although a few were doing so by the 1930’s:

Cos Cob-Arrow, N ew Music, and later Music Press-Mercury brought out volumes of songs by... American composers whose music at the time was considered to be ‘advanced.’ Of course, what was published was a small fraction of the quantity of songs written in those years. But for the first time, there was a serious American art song repertoire that was quite distinct from

^Howard Boatwright, “The Songs,” The Music Educator’s Journal 61, no. 2 (October 1974), p. 47.

^Donald Ivey, Song: Anatomy, Imagery, and Styles (New York: The Free Press, 1970), p. 241.

‘‘Quoted by Phillip Ramey, “A Talk with Samuel Barber,” notes to the recording. Songs o f Samuel Barber and Ned Rorem (New World Records NW 229,1978), p. 2. 146 the earlier lied imitations and the ‘black and white’ songs (so-called because of the covers) that were the staple of every ladies’ music club meeting. 5

American composers had almost completely abandoned European Romanticism by the middle of the century. While some composers (such as Barber) continued in a more traditional vein, many other American song composers were developing new sounds. Ned

Rorem emerged in the 1950’s as one of the most important of them: “Possibly the most prolific song composer in America during the present era has been Ned Rorem. In any consideration of song he is important, not only because of the unusual attention he has given to it but also because of his several books in which he has shared some of his ideologies.’’®

Another less visible but highly regarded composer of Rorem’s generation is Dominick

Argento, who is known equally for his operas and song cycles. His “...delay in nationwide recognition.. .may be due to the fact that he is an original, revisionist, independent-minded musician whose traditional, eclectic, and literary approach to music is not shared by many of his confreres.’’^ His songs have gained wider popularity especially from American singers in recent years.

The choices of the four American art song composers for this study were less clear-cut than for the other groups. The decision to include Charles Ives, Samuel Barber, Ned

Rorem, and Dominick Argento was based on their prominence as American composers and the time span they collectively cover, in addition to their musical diversity and their strong commitment to song composition. All four of these men were well-educated and well-read; the eclecticism of their textual selections reflects this.

^Howard Boatwright, op. cit.

®Ivey, op. cit., p. 241.

^“Argento, Dominick,” Current Biography Yearbook 1977, ed. Charles Moritz (New York; H. W. Wilson, 1977), p. 32. 147

CHARLES IVES

Ives’s Educational and Literary Background

Charles Ives (1874-1954) was the son of George E. Ives ( 1845-1894), a prominent musician of Danbury, Connecticut. George Ives was a free-thinking individual, especially about music, a man who . .seemed to crystallize out of his environment all that was most inventive, ingenious, naive, and original. He was also the most important single influence on his son.”*

Charles attended public school in Danbury as a child. His early musical activities somewhat hindered his academic standing, but in 1888, “[h]e concentrated on his school work, and made the Roll of Honor.”® He then attended two private schools before entering college. “In 1891 he went from a public school to the newly organized Danbury

Academy....In 1893, he moved to Hopkins Grammar School, New Haven....”’® He attended the latter school “. ..to prepare for his entrance examinations to Yale.”” Wallach indicated that Ives struggled with his studies at this stage of his education:

The move away from the familiarity and security of Danbury was accomplished by several more tangible anxieties which Ives faced and energetically strove to conquer. Chief among these was his mediocre student record at Danbury Academy. During the Spring-Summer of 1893, he studied furiously to pass the preliminary examinations for Hopkins.

^Laurence David Wallach, The New England Education o f Charles Ives (Ph.D. dissertation, Q>lumbia University, 1973; Ann Arbor: University Microfilms, 74-17,913), pp. 71-72.

^Ibid., p. 135. Wallach cites the Roll of Honor report in the Danbury Evening News for July 7, 1888.

’®John Kirkpatrick, “Charles Ives,” TbeNewCrove Twentieth-Century American Masters (originally published in The New Grove Dictionary o fAmerican Music, ed. H. Wiley Hitchcock and Stanley Sadie) (New York: Norton, 1986, 1988), pp. 2-3.

’’Wallach, op. cit., p. 233.

’2jb/d 148

Ives continued to work towards his entrance into college; in the summer of 1894 he was

.receiving specif tutoring to prepare for his entrance examinations to Yale.”*3

He managed to enter Yale in the fall of 1894. In college he was not allowed to compete in athletics, in spite of his physical abilities: “...his father had forbidden it, knowing how precarious his grades would be."''* Ives continued to do poorly in school: "... the Yale

College gradebooks from 1894-98 show that his grades were barely passing except for the music courses for which he had credit as an upperclassman.”'^ Kirkpatrick suggested that

Ives was too busy with the large body of music he wrote at the time: “With all his composing—over 40 songs, various marches, overtures, anthems, and organ pieces, the quartet, and the symphony—his academic average of D+ at Yale is quite understandable.”"5 Ives graduated in June 1898, and thereby ended his formal education.

Kirkpatrick has concluded that Ives, “[w]ith his variety of interests.. .spread himself thin and was never a brilliant scholar."On the other hand. Perry concluded that “Ives was alert, intelligent, and well-read, the kind of man who was very much the product of intellectual trends and impulses in American civilization.”'^

He received two official awards in later life which recognized his achievements in music. The first was his election to the membership of the National Institute of Arts and letters in 1946. The second was the Pulitzer Prize for music, which he won in 1947 for his

Third Symphony.

'3 /6/d, p. 243.

'''Kirkpatrick, op. cit., p. 5.

'^Vivian Perils, C6ar/es/ves/?eme/n6ene

'^Kirkpatrick, op. cit., p. 6.

'7/b/d, p. 2.

' ^Rosalie Sandra Perry, Introduction to Charles Ives and the American Mind (Kent, Ohio: The Kent State University Press, 1974), p. xvi. 149

Charles Ives may have shown some weakness in his formal training, but he demonstrated his true capabilities by the creation of a highly successful insurance business.

He also revealed the depth of his literary mind through his essays, poems, and other written works. His prose texts encompass a variety of subjects, from important booklets on life insurance (written between 1912 and 1924)^9 to explanations of his aesthetic position.20 His poetry also covered a wide range of topics, such as personal childhood memories. World War I, political activism, and subconscious ideas which were rooted in the Transcendentalism of nineteenth century New England.

Several scholars have noted Ives’s strong connection between writing words and writing music. Yellin commented on this in a review of the 1961 edition of Ives’s collected essays:

For Ives verbalization was almost as important as composition. He was not satisfied to let his ideas, made tangible as music, speak for themselves. Some inner necessity forced Mm to articulate them in words as well. So it may not be too far-fetched to suggest that botii Ives’s music and Ives’s words are simply two aspects of one idea; tliat the Concord Sonata together with the Essays forms a kind of cantata.^*

Boatwright suggested that Ives’s vocal music stemmed from a strong literary propensity:

Songs occupy a special place in the music of Charles Ives. All through his creative life, he turned again and again to that medium, pemaps because along with being profoundly musical he was also highly verbal. The verbal side found its strongest expression in essay writing which began while he was in school, reached a climactic point in Essays Before a Sonata (a companion to the "Concord” [piano] Sonata), and continued on in a more fragmentary way in many drafts of articles, memos, and scribblings on the backs of envelopes and calendars and in the margins of manuscripts—long after ill health had forced him to abandon composition.^^

*^The full list is given by Perlis, op. cit., p. 36, footnote 1.

^°E.g., his Essay Before a Sonata and the “Postface” of his 114 Songs.

^ Wictor Yellin, review of Essays Before a Sonata, and Other Writings by Charles E. Ives, ed. Howard Boatwright, Journal o f the American Musicological Society 17 (Summer 1964), p. 229.

^^Howard Boatwright, op. cit., p. 42. 150

The poems Ives wrote for his songs are not as prominent or as highly regarded as his prose works are today, perhaps because of the conventionality of their subject matter and their style. Schoffman, for example, commented that “Ives was by no means a great poet, but he undoubtedly had literary talents, which manifest themselves both in his writings and in his song t e x t s . ” 23

The strength of Ives’s literary mind is emphasized in his musical compositions, which are full of references to important writers, especially the New England Transcendentalists.

Burkholder has even suggested that “The term literary music .. .describes the conception

and execution of the works of Ives’s maturity.”24 Wooldridge went as far as claiming that

“Ives had a comprehension of America’s seers and poets that would have been remarkable

for a composer or a businessman. For one who was both it was marvelous. ”25

Ives’s Poetry Choices

Charles Ives chose the texts for his songs from a large body of literature, which

included works from many styles, languages, and eras. “Viewed as a whole, Ives’s song

texts reflect a richness of outlook and experience consistent with his philosophy of taking

life in the full, not merely in specialized p a r t s .” 26 Schoffman equated Ives’s range of

textual choices with his musical ones: “The variety of styles in the song texts is quite wide,

somewhat analogously to the wide variety of musical styles in his w o r k s . ” 22

22Nachum Schoffman, The Songs o f Charles Ives (Ph.D. dissertation, Hebrew University of Jerusalem, 1977; Arm Arbor, Michigan: University Microfilms, 79-05,582), p. 236.

24j. Peter Burkholder, Charles Ives: The Ideas Behind the Music (New Haven and London: Yale University Press, 1985), p. 17.

250avid Wooldridge, Charles Ives: A Portrait (London: Faber and Faber, 1975), p. 5.

26Boatwright, op. cit., pp. 42 & 45.

22Schoffman, op. cit., p. 236. 151

Ives left much of the selection of his song texts to chance, rather than through careful consideration of individual poets and poetry. Boatwright explained this by saying that,

“Because he was a hardworking insurance executive, Ives had less time than he would have wished to read and contemplate possible texts for songs. He had to react to them when and where he found them, and that meant on more than one occasion that he took his texts out of newspapers or magazines."^^

Newman felt that the eclecticism of Ives’s poetry choices was both negative and contradictory;

[T]hough Ives was apparently widely read, it is doubtful he was a discriminating lover of great poetry, for in choosing song-texts his taste tended to be eclectic and inclusive. Often he accepted, along with the artistically perceptive, the sentimental and banal. In a sense this is curious considering his frequent vituperations against the maudlin and imitative in music. Nevertheless, there are reasonable explanations for these textual choices.29

The explsiiations Newman gave were that little quality poetry was available to Ives while he wzis composing songs (from about 1887 to about 1926), and what was available to him was typical (i.e., sentimental, nostalgic) 19th century verse. Newman also mentioned one exception to Ives’s “traditional” choices: “. ..Those texts which discuss political and social problems are a distinct departure from these nineteenth-century types.”3°

There was little chronological order or pattern to Ives’s textual selections. He was using the same types of verses when he began composing songs in the late 1880’s until he stopped nearly 40 years later. These consisted of a mixture of his own verses, anonymous and folk verses, poems from family and friends, foreign texts (many previously set by the

^*Boatwright, op. cit., p. 42.

^^Philip Edward Newman, The Songs o f Charles Edward Ives (J874-1954), 3 vols. (Ph.D. dissertation. The University of Iowa, 1967; Volumes I and II, Ann Arbor, Michigan: University Microfilms, 67- 16,823), I, p. 201.

30 Ibid., I, p. 202. 152 lied composers), a few ancient verses, and a large number of poems from both obscure and important British and American writers of the nineteenth and twentieth centuries. A large portion of the selections will be discussed later in this document, under the various categories outlined for this study. The complete list of Ives’s poetry sources is found in

Table 21, pp. 245-47. 153

Table 9. The Songs of Charles Ives

Description ______Number % ___

Total number of Songs 176

Total number of Sources 82

Arithmetic average number of songs per source: 2.15

Median number of songs per source: 1.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 62 35.23%

Texts from poets contemporary with composer (bom within 15 years of composer) 64 36.36%

Texts from poets younger than composer (bom 15 years after composer) 1 0.57%

Texts from undated sources (dates of poet or source unknown) 31 17.61%

Texts from old sources (dated more than 100 years before composer's birth) 18 10.23%

Texts from non-native sources (either in translations or in other languages) 37 21.02%

Approximate number of poets with whom the composer had personal contact: 6 154 Ives’s Number of Songs and Sources

There are 176 songs by Charles Ives counted in this study. The primary authority used to arrive at this number is the work-list which John Kirkpatrick prepared for The N ew

Grove Dictionary o fAmerican Music?^ Seven works in Kirkpatrick’s list were not included, based on the criteria established in th= Introduction of this document.

Kirkpatrick’s list was critically compared with Newman’s documentation in the comprehensive study he made of the Ives songs.^^ Newman presented a total of 205 songs, 30 of which were not countable for this study and one of which (“Two Slants”) was counted here as two separate songs. This also produced a total of 176 songs, which confirmed the number derived from Kirkpatrick’s list.

Ives composed many of his songs using his own pre-existing musical material. Most often these songs were new arrangements for voice and piano of instrumental or choral works or sketches. This was a pervasive and common practice in Ives’s compositional procedure, ai:d was not like re-scoring a song originally composed for voice and orchestra or extracting a song from a play or other stage work. For this reason, all of Ives’s songs in

Kirkpatrick’s list (with the few exceptions mentioned) are counted in the total for this document.

Ives chose his song texts from 82 sources. These sources are presented in Table 21, pp. 245-47. Kirkpatrick’s list (cited in the previous paragraph) was the primary guide used to determine the sources. The sources listed by Kirkpatrick were compared to the ones given in Newman’s detailed account of Ives’s s o n g s .

^*John Kirkpatrick, “Charles Ives,” TbeNewCrove Twentieth-Century American Masters, op. cit., pp. 35-43.

^^Newman, “Appendix: Songs in Numerical Order,” op. cit.. I, pp. 251-253.

” ib/d., II, pp. 23-419. 155 The occasional discrepancies between the two lists with regard to the sources were resolved individually. For example, Ives considerably altered some texts by known writers for his own use. Kirkpatrick attributed these verses to Ives, while Newman gave the original author as the source. In these instances the poems are counted here as texts by

Ives. In contrast, Kirkpatrick listed most songs in translations as being by the translator

“after” the original author, while Newman named the original author as the source in these cases. Following the pattern begun in Chapters I and II of this study (Tables 13-20), the original author is counted and presented, followed by the name of the translator if known.

Ives averaged 2.15 songs per source, given 176 songs from 82 sources. The median average for his songs is 1.00, as is the modal value. These averages are quite low among those for the twelve composers studied. With two exceptions (texts by anonymous or unknown authors and his own verses), Ives did not use more than 8 verses from any source. Only 17 of his 82 sources were chosen for more than one text, leaving 65 sources which he used one time each. While Ives selected a large number of both his own texts and those by anonymous or unknown writers, he most often chose a low numberoï texts from each of his sources.

Ives’s Authors Categorized by Birth Date

The breakdown of Ives’s selections by dates in Table 9 shows that 62 of his texts were taken from authors bom 16-100 years before 1874 (the year Ives was bom), 64 texts were by authors bom within 15 years of 1874, and one text was from an author bom more than

15 years later than 1874. The first group (from authors bom between 1774 and 1858) represents 35.23% of his total choices, the second group (from authors bom between 1859 and 1889) are 36.36% of his total choices, and the one choice from an author bom after

1889 is 0.57% of the total. 156

Ives’s textual selections from older generation and contemporary authors are almost identical in number and percentage. What distinguishes the two from each other is the number of authors in each: there are 42 authors in the older generation group compared to

14 authors in the contemporary group. The difference is largely explained by the high number of texts Ives wrote himself, which form a large part of the percentage by contemporary authors. Additionally, his relatively large choices of six texts by his wife

(Harmony Twichell Ives) and five by Kipling help to increase the percentage from contemporary authors. In contrast, Ives selected no more than five texts from any of the older generation authors except Heine, from whom he took eight verses.

The percentage of texts Ives used from older writers is primarily explained by the fact that he composed most of his songs between 1887 and 1921, in his teen and early adult years. As mentioned previously, the poetry of the nineteenth century was the most readily available material for his song texts at that time. Newman wrote that, “During the years

Ives was composing songs.. .there was little poetry of an advanced nature available to him for possible song-texts. Consequently,.. .most of these texts follow accepted formal designs and express familiar nineteenth-century themes... .”^4

Ives’s eclecticism is quite evident in his choices from older authors. For example, the nationalities of the 42 authors are as follows: 16 British (English, Irish, Scottish),

13 American, 9 Germanic (German, Austrian), 2 Danish, 1 Norwegian, and 1 French.

There are many important and famous poets on the list, such as Byron, Shelley, Heine,

Longfellow, Tennyson, and Holmes. Others are more obscure, such as H. Coleridge and

Ploug.

American writers are predominate on Ives’s list of contemporary authors. Ten of the

14 writers in this group are from the United States, 3 are British, and one (John David

McCrae) is Canadian. Important contemporary authors such as Kipling, Lindsay, and

34/b/J., I, p. 201. 157

Untermeyer provided texts for a few of his songs. Ives did not choose texts from several

other notable American poets who were his contemporaries:

The New Englander, Edwin Arlington Robinson [1869-1935] (the first to break away from the sentimentality of the fading nineteenth century and to give substance and firmness to his verse) and, particularly, Robert Frost [1874-1963] (with whom Ives should have had obvious kinship) arrived too late on the literary scene to have influenced him. Edgar Lee Masters [ 1869- 1950] and Carl Sandburg [1878-1967], however (both of whom brought a Midwestern liveliness to poetry), were seeing their early works published in the mid-teens, but these either went unnoticed by Ives or failed to suggest song settings. 3 5

Ives probably heard about some or all of these writers in his later years. In the early

1940’s, for example, Louis Untermeyer visited with Ives and mentioned the poetry of

Frost, but “...it evinced no response from Ives w hatsoever.”36 Perhaps the composer made no comment because he had ceased writing songs nearly twenty years before that meeting.

The largest number of texts by one author in the contemporary group were written by the composer himself. Newman presented several viable explanations of why Ives used so much of his own textual material:

That Ives’s own texts mounted in number as time went on, may signify a maturing taste (though he never lost touch with certain homespun subjects) and, possibly, a concurrent inability to find material akin to his musically unconventional direction of thought. Also, and perhaps more likely, his vivid and often literal imagination, with its strong ethical overtones, may simply have demanded a verbal as well as musical expression.37

One of the texts Ives selected, “Afterglow,” was written by a poet from the younger generation. James Fenimore Cooper, Jr., bom 18 years after Ives, was a great-grandson of the famous novelist (1789-1851). He graduated from Yale in 1913, emd afterwards

35/6/£/., I, p. 2 0 2 .

3®Inverview of Louis Unteimeyer about Charles Ives, October 13, 1971, in Perlis, op. cit., p. 211.

37 AM , I, p. 203. 158

travelled in Europe and the western United States for several years. Friedberg reported

that he then "...made an unsuccessful attempt to study law at Harvard. He had about

decided to settle at Cooperstown, New York (founded by the novelisfs father), and pursue

a writing career, when World War I drew him into the start of what seemed to be a

promising career in the military.”^» Cooper eventually was named Captain of the 308th

Field Artillery, but died of pneumonia at the age of 26 at Camp Dix, Wrightsiown, New

Jersey, in February of 1918.^^

A collection of Cooper’s poems was published posthumously by Yale University Press

in 1918. A second, revised edition was published in 1919, the same year the 45-year-old

Ives composed his setting of “Afterglow.” Friedburg suggested that Ives may have known

about the young Cooper’s poetry before publication of the posthumous volume, since

several of the poems had previously appeared in the Yale literary magazine.'*® Ives was probably drawn to this text by its references to nature and memory more than through a personal knowledge of the author, although their mutual connection with Yale may have

influenced the choice.

As with most of the composers in this study, Ives used little poetry from younger writers. His single text from a younger generation writer only further implicated him as a traditionalist in his textual choices. Ives lived a long life, which would have allowed him the opportunity of working with the poetry from the middle of the twentieth century.

Before this was possible, though, he had ceased his creative efforts in song composition.

It could only be speculated what Ives’s textual choices might have been in the years after

1926.

^®Ruth C. Friedberg, Aimerican Art Song and American Poetry, 3 vols. (Metuchen, New Jersey: Scarecrow Press, 1981-87), I, p. 82.

^®Newman, op. cit., II, p. 316.

■*®Friedberg, op. cit. 159

Ives’s Choices of Undated and Old sources

Ives set 31 texts from anonymous or unknown sources. This number is 17.61 % of his total choices, a mid-range percentage for the twelve composers. Nine of Ives’s undated texts were from nine identified authors about whom little or nothing is known:

Name Comments Case, Lizzie York probably American, died 1911. Collins, Anne Timoney American, friend of Ives, poems and plays during 1920’s, an article written in 1956. Hill, Susan Benedict probably American, no information available Ives, Miss A. L. American, poem published in her father’s M usical Spelling Book of 1846, no relation to Charles Ives Morris, Maurice American, Pseudonym for Morris W. Pool, poems published 1921. O’Malley, D. J. American cowboy poet, late nineteenth century Chézy, Wilhelmine von German Countess, early nineteenth century, poet, librettist, and playwright Porteous, Baronness Unknown, but probably German, probably mid­ nineteenth century Chauvenet, W. M. No information available, the poem which Ives used was in English

The other segment in this undatable category consists of 22 texts by anonymous or unknown authors. Two of them are folk texts: the traditional Irish “Son of a Gambolier” and the German lullaby folk text made popular by Brahms, “Wiegenlled.” Two others are foreign texts from unknown authors: “Qu’il M’irait Bien,” a French text which Ives used in the original language, and “The White Gulls,” a Russian text translated into English by

Maurice Morris (Morris W. Pool). Newman’s opinion was that “[t]he remaining texts, whose authors, in some cases, Ives may have known but failed to record, range in 160 q uality.... "41 Ives may have written some of them himself but wished them remain anonymous, such as “I Knew and Loved a Maid.”42

The explanations for Ives’s textual choices from unknown and folk sources are fairly straightforward. He had a haphazard method of finding and selecting song texts, and frequently did not provide adequate documentation for the texts he used. Secondly, and more importantly, he had a strong need to include elements of common, everyday life in his works. His use of anonymous or folk texts was one way of fulfilling that need.

Ives’s choices from old sources were also eclectic. A total of 18 texts from 15 poets were chosen from authors bom more than 100 years before Ives. These are 10.23% of his total choices, an upper mid-range percentage for the twelve composers in this study. Ives chose these texts exclusively from European authors, and five of the verses date from antiquity.

The oldest of these was by the ancient Roman poet Manilius (or Manlius, who lived in the 1st century BC), from whom Ives took a single line of Latin text for his brief song

“Vita.” Another was by the 12th century Frenchman, Bernard of Cluny, in an English version by J. M. Neale (“Country Celestial”). The remaining three were by the 13th-14th century Italian poet Folgore da San Geminiano. Ives used the English translations of these verses by D. G. Rossetti (“August,” “September,” and “December”).

The other thirteen texts from old sources are grouped by time period and nationality as follows:

4* Newman, op. cit.. I, p. 222.

42ib/d, II, p. 146. 161 Time Period Number Nationality Names of (Century) ______of Texts of Authorfs) ______Author!s) ______15th-16th 1 Italian Ariosto late 16th 1 English Shakespeare 17th 2 English Milton, Tate 18th 3 English Percy, Cowper, Robinson 18th 2 German Tersteegen, Holty, Goethe 18 th 1 French Florian 18th-19th 2 English Wordsworth

These authors date from each of the last five centuries, and many of them are important

European authors. Ives showed himself to be far-reaching in his choices from old sources,

at least in terms of time periods. The significance of the European background of these

authors is important. Perry wrote that Ives was "...devoted to old customs and ways of the

past—but of his own past, not a Europeanized past.”^^ These choices of distinctly

European song texts might suggest otherwise.

Ives’s Choices in Non-native languages

Ives composed 37 songs with texts from non-native languages. These foreign-

language texts comprise 21.02% of his total choices, one of the highest percentages in this

category for the twelve composers studied. Ives set fourteen of the 37 non-native texts in

hnglish translations. These translated texts came from seven original languages:

'*^Perry, op. cit., p. 1. 162 Original Language Number of Texts German 6 Italian 3 Norwegian 2 Danish 1 Latin 1 Russian 1

Ives set 19 of his non-native texts in their original languages. The majority of these were in

German, but three other languages were also represented:

Language Number of Texts German 14 French 3 Italian 1 Latin 1

Finally, there are five special cases with regard to languages in this category. Ives set two songs from non-English translations of foreign texts and three in more than one language:

Song Title uriginai Language Language! s) for Ives's Setting Sehnsucht Danish German Du alte Mutter Norwegian German Rosamunde German German and French translation Feldeinsamkeit German German and English translation Marie German German and English translation

A majority of Ives’s songs in non-native languages were composed in the decade 1892-

1902, while Ives was a student. Newman gave two reasons why Ives worked especially with German poetry at that time: “Obviously the German Lied, along with other European songs, was a required part of his musical studies at Yale. Also, he was no doubt 163 influenced by his father’s serious study of both the German language and German music. Wallach added to this explanation: “Most of the German texts had been previously set by Schubert, Schumann, Mendelssohn, Brahms, Cornelius, or Franz. Ives was striving to fit himself into the European tradition in the traditional genre he knew best.”‘’5 Ives studied several foreign languages at Yale: German, Latin, and French, in addition to Greek literature.^^

One of Ives’s more notable choices from a non-native text was the portion of Bernard of Cluny’s poem. De Contemptu M undi, which he used for the song “Country Celestial” in

1891. Ives used a hymn-verse translation of the Latin text by John Mason Neale ( 1818-66) for this song. A year later Horatio Parker set 35 stanzas of De Contemptu Mundi in the original Latin for one of his most important works, the oratorio Ham Novissima. Parker and Ives had not yet become teacher and student at Yale when these works were composed.

Ives returned to the poem in 1898, perhaps influenced by Parker’s use of it. In that year he composed one of his more important works, the choral cantata The Celestial Country, this time using excerpts from the free translation of De Contemptu Mundi by Rev. Heniy

Alford.

The relatively large percentage of Ives’s texts from non-native sources emphasize the strong foreign (esp. European) influences in his life, particularly from his student days while studying with Parker at Yale. His use of texts from more than one non-native language are one further example of the widely varied nature of his work.

'^'’Newman, op. c lt, I, pp. 150-151.

^^Wallach, op. cit., p. 245.

^^Appendix 6, “Ives’s scholastic record (1894-98),” Charles E. Ives, Memos, ed. John Kirkpatrick (New York: Norton, 1972), pp. 180-183. 164

Ives’s Personal Contact with the Poets

Ives personally/knew five of the authors of his song texts. Six names, including Ives’s own, are marked with an asterisk in Table 21, pp. 245-47.

Two people who contributed verses for Ives’s songs were his relatives. One was his wife, Mrs. Harmony Twichell Ives, who wrote six of the poems Ives used. Harmony first met Ives in the summer of 1896 through her brother Dave Twichell. She enrolled in the

Hartford Hospital Training School for Nurses in 1898 and graduated as a Registered Nurse in 1900. She then spent several years afterwards as a practicing nurse. She again met Ives in 1905, and after a long engagement they were married in June of 1908.^7

Ives set four of her nature poems in the years of their engagement (“The World’s

Highway,” “Spring Song,” “Autumn,” and “The South Wind”). Her poem and his setting of “Mists” both date from 1910. Mrs. Ives wrote this poem as a lament on the loss of her mother in that year. Finally, her poem “To Edith” and its setting date from a trip to North

Carolina the Ives family made in early 1919. The name Edith refers to their adopted daughter Edith Osborne Ives (1914-56). One other poem, “Two Little Flowers,” was penned by the composer and his wife together (it is counted in this study as written by

Charles), with the musical setting composed in 1921. The “two flowers” were daughter

Edith and her playmate Susanna Mintem.^*

Ives used one other text from a relative. His first song (ca. 1887) was called “A Slow

March,” a dirge on the death of the family dog. The text for this piece was written collectively by the Ives family (i.e., Charles’s mother, father, and paternal grandmother),

“...but mostly by [his uncle] Judge Lyman D. Brewster.... ”'*9 Ives took a memorable trip

^^Much infonnation on the life of Mrs. Ives can be found in Appendix 19 of Charles E. Ives, Memos, Ibid., pp. 274-280.

48/bid., p. 279.

4'Newman, op. cit., II, p. 23-24. 165 with his uncle to the Chicago area in late August/early September of 1893, which included two visits to the Chicago World’s Fair.K irkpatrick noted that “All his life Lyman

Brewster wrote verse, of which a booklet was pnnted in 1900 by the Danbury News, entitled Youth and Yale.”^^ In the summer of 1903 Ives considered using a play by his uncle as the text for an opera which was never written.^-

Ives used two verses written by his friend and champion Henry Bellamann, for two late songs, “Peaks” and “Yellow Leaves” (both 1923). Bellaman was a musician, novelist, poet and lecturer. His article “Charles Ives, the Man and His Music” in the Musical

Çuarter/y( 19/1, January 1933, pp. 45-48) was one of the first important articles on Ives and his songs. Bellaman learned of Ives about 1918 or 1919, and wrote to him requesting a copy of the Concord Sonata. They exchanged letters about Ives and his music during

1920-21. Bellamann “...met Ives personal’y the first time in New York City about July

19, 1922.. ..”53 Bellamann and his wife visited the Ives family several times during 1923, the year Ives set the Bellamann poems. During one visit Mrs. Bellamann (a trained singer) sang some of Ives’s songs for him.54 The two couples remained in contact into the

1940’s, although Henry did not live to witness the prominence Ives and his music gained later. Mrs. Bellamann wrote a congratulatory letter to Ives after he won the Pulitzer Prize, in which she commented that her husband “would have rejoiced so in this recognition of the greatness of the music he believed in.”55

5°Charles E. Ives, Memos, op. cit.. Appendix 20, p. 281.

^^Ibid., §31, p. 83.

53paul Parthun, “Concord, Charles Ives, and Henry Bellamann,” Student Musicologists at Minnesota 6 (1975-6), p. 66.

^'^Ibid., pp. 72-73.

55/b/t/., p. 75. 166

Ives was in contact with Anne Timoney Collins through correspondence. Mrs. Collins was the author of the text for one of Ives’s best-known songs, “The Greatest Man” (1921).

Newman reported that “Ives discovered this poem printed in the New York Evening Sun for 7 June 1921 .”56 Ives wrote the author asking her permission to use this text, to which she reponded affirmatively. The two letters survive, along with one other from 1928, a note from Ives to Collins thanking her for the enjoyable time he and Mrs. Ives had had at a production of her play Bottled in Bond. 57 Collins was previously mentioned in the undated category, since little other information was available about her life.

The final author of a song text known personally to Ives was Louis Untermeyer. The poet “...visited with Ives only once in 1943 or 1944...to discuss the wedding of text and music in songwriting.”5* Ives had set Untermeyef’s poem “The Swimmers” almost thirty years earlier, in 1915. Untermeyer was interviewed in 1971 about his visit with Ives, and reported that, “Ives and I exchanged a couple of letters and he asked me to stop by the house in West Redding. He was close to seventy and I was close to sixty.”59 The two mainly discussed their mutual admiration for Emerson. Ives’s unresponsiveness to

Untermeyeris mention of Robert Frost was noted earlier, in the discussion of Ives’s use of contemporary poetry.

No other relationships between Ives and his poets were discovered. Ives was perhaps too independent to be greatly involved with individual poets or their works. While the themes of the poems he used were quite dear to him, it is clear that most often the poets themselves were of little importance.

5® N ew m an, op. cit., II, p. 398.

57john Kirkpatrick, A Temporary Mimeographed Catalogue o f the Music Manuscripts and Related Materials o f Charles Edward Ives (New Haven: Library o f the Yale Schcwl o f Music, 1960, reprinted 1973), p. 208.

5®Perlis, op. cit., p. 211.

59lnterview of Louis Untermeyer, in Perlis, Ibid., p. 211. 167

SAMUEL BARBER

Barber’s Educational and Literary Background

Samuel Barber (1910-1981) was a doctor’s son, and spent his early childhood with his family in the comfortable surroundings of West Chester, Pennsylvania. He began writing small compositions while only seven years old, and in these earliest compositions he showed an inclination towards writing music for the voice. “He was supported and encouraged in this respect by his mother’s sister, the contralto Louise Homer, and her husband Sidney Homer, a composer who concentrated almost exclusively on song.”^° His formal academic study was completed at West Chester High School, from which he graduated in 1926. In 1924 he began his formal musical training by enrolling at the newly formed Curtis Institute of Music in Philadelphia, where he was enrolled as a major in piano, voice, and composition. Friedberg reported that Barber studied French, German, and Italian as part of his curriculum at Curtis.^ i He graduated after eight years of study, ending his formal education in 1938.

Several years later he briefly entered the academic world again: “Barber returned to the

Curtis Institute in 1939 as a teacher of composition and remained there until 1942, though he had no great liking for this kind of work and never again accepted a teaching position.”^^ His achievements as a composer were publicly recognized with his election to the National Institute of Arts and Letters in 1941 and in 1958 to the American Academy of

^%chard Jackson, with work-list by Barbara Heyman, “Samuel Barber,” The New Grove Twentieth- Century American Masters (originally published in The New Grove Dictionary o f American Music, ed. H. Wiley Hitchcock and Stanley Sadie) (New York: Norton, 1986, 1988), p. 243

^^RathFnedbeig, American Art Song and American Poetry, 3 vols. (Metuchen, New Jersey: Scarecrow Press, 1981-87), III, p. 8.

“^Richard Jackson, op cit., p. 246. 168

A rts and Letters.63 He won two Pulitzer Prizes formusic, for his opera Vanessa in 1958 and for his Piano Concerto in 1962. In 1959 he received an honorary doctorate from

Harvard University.

Barber began his interest in literature at an early age. Broder wrote that Barber had

“.. .early developed a passion for reading. Barber himself admitted that, “I have always had a sense of the written word, and have sometimes thought that I’d rather write words than music.”65 He did write the text for one of his earliest songs, “October Weather,” although Kreiling commented that “...this rare evidence of Barbefs interest in writing words does not reveal a significant talent....

He won the first of many composition prizes in 1928, which provided a chance to study in Italy. He travelled abroad many times in subsequent years; this offered him a wide appreciation of other cultures and a broad perspective of the artistic world of his time.

Significant in this perspective was his exceptional understanding of world literature.

Barber’s widening literary development may be seen in his early considerations for song texts. In one of his sketchbooks dated “c. 1933,” now housed at the Library of

Congress, he compiled a list of literary selections to be considered for song settings.

Kreiling recently examined the sketchbook and described its contents: a translation of a

Chinese poem, a German folk text, and a list of foreign authors, “.. .apparently intended to be purchased and/or read, [which] includes the names of Thomas Mann, Turgenev,

Chekhov, Montaigne, and Suetonius. This reading list is dominated, however, by British

®^Nathan Broder, Samuel Barber (New York: G. Schirmer, 1954), p. 11.

^^Quoted by Phillip Ramey, op. cit.

Jean Louise Kreiling, The Songs o f Samuel Barber: A Study in Literary Taste and Text Setting (Ph.D. dissertation, The University of North Carolina at Chapel Hill, 1986; Ann Arbor, Michigan: Unversity Microfilms, 86-18564), p. 46. 169 names, including Dickens, Marlowe, Pope, Thackeray, and Woolf.Other segments found in the sketchbook were a portion of a sonnet by Gerald Manley Hopkins and one of

Barber’s unpublished Joyce settings.

Barber was quite involved with reading and studying world literature as an adult. In the 1954 biography of the composer, written at the height of his career. Broder indicated that Barber “.. .continues to read a great deal. One of his favorite authors is Stendhal, whom he admires for his combination of classic clarity and passion. Others whom he has studied through the years are Dante, Goethe, Joyce, Proust, and Melville.”^® Friedberg noted that only one of these authors was American, and that “.. .this one American preference takes on cloudy origins with the realization that Barber’s deep attraction to

Melville and his never fulfilled desire to base a musical work on his writing began on the occasion of his reading M o b yD ickin an Italian translation!

The composer himself later confirmed some of these literary investigations: “I do enjoy

reading contemporary poetry, not only in English but in German and French, and I’ve made a real study of Dante and Goethe in their original languages.”'^° He also mentioned a

study of Joyce: “I’m not unlearned in Joyce; I’ve read quite a few books on him.”^i

Finally, Barber showed himself to be deeply associated with literary themes by using

them repeatedly in his instrumental music. Three of his large orchestral works were each

given the title “Essay for Orchestra” (op. 12, 1937; op. 17,1942, and op. 47, 1978). In

each of them he attempted to suggest the logical development of a musical theme, much as a

written essay would do with an idea. His Overture to “The School for Scandal, ”op. 5,

(•TIbid., pp. 51-52.

®®Broder, op. cit., p. 45.

^^Friedberg, op. cit.. Ill, p. 10.

^°Quoted by Phillip Ramey, op. cit., p. 2.

'libM 170 refers to an eighteenth century comedy by Richard Sheridan, and his M usic fora Scene from Shelley, op. 7^ is based on a segment of Shelley’s Prometheus Unbound. Finally,

Barber’s op. 44, Fadograph o fa Yestem Scene, takes its title from Joyce’s Finnegans

Wake.

Barber’s Poetry Choices

Barber wrote over forty songs which he did not publish. He experimented with choices from a variety of authors for these works. Three songs were completed while he was still a teenager: “Why not’’ (text by K. Parsons, 1917), “In the Firelight” (author unknown, 1918), and “Isabel” (text by John Greenleaf Whittier, 1919). One song, “Thy

Will be Done” (text from “The Wanderer”) was written in about 1922, and the Seven

Nurseiy Songs (texts from Mother Goose) were composed in 1920-22. The latter were first performed, with most of the other songs completed before mid-1927, on a program given April 25, 1926, at the Barber home in West Chester, Pennsylvania.^^

Texts for some of the other songs on this recital included one by Barber himself

(“October Weather”); others were by black poets. Dr. Robert Thomas Kerlin (“My

Fairyland”) and Langston Hughes (“Fantasy in Purple”); tum-of-the-century authors Jessie

Belle Rittenhouse and Laurence Binyon (Two Songs ofYouth)and important British writers, Percy Bysshe Shelley (“Music When Soft Voices Die”) and Elizabeth Barrett

Browning (“Thy Love”). Four unpublished songs not on the 1926 recital but composed before 1930 are settings of texts by Sir Thomas Wyatt (“An Earnest Suit to his Unkind

Mistress,” 1926), Walter Savage Landor (“Mother I cannot Mind my Wheel,” 1927),

Algernon Swinburne (“There’s Nae Lark,” 1927), and an zmonymous verse (“The

Shepherd to his Love,” 1928).

reprint of the program for this recital shown in Ramey, ibid. 171

Barber composed ten more unpublished songs in the 1930’s. Three early settings of texts by James Joyce during this decade foreshadowed Barber’s later important Joyce settings: “O f that so Sweet Imprisonment,” 1935; “Strings in the Earth and Air,” 1935; and

“In the Dark Pinewood,” 1937. Three songs were composed on verses by the English poet

William Henry Davies (“Love’s Caution,” 1935; ' Night Wanderers,” 1935; and “The

Beggar’s Song,” 1936). Two were settings of ancient verses in translations: “Love at the

Door” (from 1934), by the ancient Greek philosopher and poet Meleager, and “Peace”

(from 1935), by the ancient legendary Hindu poet Bhartrihari. The remaining two unpublished songs from the 1930’s were “Serenades” (1934, text by George Dillon) and a

1935 setting of Robert Frost’s “Stopping by Woods on a Snowy Evening.”

Barber composed thirty-six songs which were published during his lifetime. The eclecticism and sophistication of his textual choices continued with these songs. Ten of them were written by the end of 1940. The three songs in op. 2 appeared first,’^ with texts by Irish poet James Stephens (no. 1, “The Daisies,” 1927, and no. 3, “Bessie Bobtail,”

1934) and British poet A. E. Housman (no. 2, “With Rue my Heart is Laden,” 1928). The three songs in op. 10 were completed in 1935-36, using poems from the book Chamber

M usic by the Irish writer James Joyce (no. 1, “Rain has Fallen,” no. 2, “Sleep Now,” and no. 3, “I Hear an Army”). The four songs in op. 13 were composed between 1937 and

1940. Barber used one text from each of four authors for these songs: English poet

Gerald Manley Hopkins (no. 1, “A Nun Takes the Veil,” 1937), Irish poet William Butler

Yeats (no. 2, “The Secrets of the Old,” 1938), American writer James Agee (no. 3, “Sure on This Shining Night,” 1938) and another American, Frederic Prokosch (no. 4,

“Nocturne,” 1940).

^^The opus numbers given for Barber’s songs were assigned to them for publication. See the work-list by Barbara Heyman in “Samuel Barber,” The New Grove Twentieth-Centuiy American Masters, op. cit., pp. 255-258. 172

Three published songs were composed in the 1940’s. Two of them are in op. 18:

“The Queen’s Face on a Summery Coin” (1942, text by Robert Horan) and “Monks and

Raisins” (1943, text by José Garcia Villa). The third was a setting of a passage from

Joyce’s Finnegans Wakeîor the song “Nuvoletta,” op. 25, written in 1947. Two

important sets of songs were completed in the 1950’s. Tire first of these was op. 25,

Mélodies passagères, five songs on French texts by the German poet Rainer Maria Rilke,

composed in 1950-51 for the French baritone, Pierre Bemac, who premiered the set in

January of 1952. The other set was op. 29, the H erm it Songs, ten settings of ancient Irish

clerical verses in English translations. These were commissioned by the Coolidge

Foundation of the Library of Congress, and first performed by soprano Leontyne Price,

with the composer accompanying, in the fall of 1953.

Two final groups of songs were written and published in the 1960’s and 197ü’s. The

five songs in op. 41, Despite and Still, were composed in 1968, and were premiered in

April 1969. Three writers were represented in these texts: English poet Robert Graves

(no. 1, “A Last Song;” no. 3, “In the Wilderness;” and no. 5, “Despite and Still”),

American poet Theodore Roethke (no. 2, “My Lizard”), and one further Joyce setting

(no. 4, “Solitary Hotel”). Barber’s last three songs were finished in 1972, his op. 45.

Three translated poems were chosen for these songs, one by Swiss writer Gottfried Keller

(no. 1, “Now Have I Fed and Eaten Up the Rose”), one by Polish author Jerzy

Harasymowicz (no. 2, “A Green Lowland of Pianos”), and one by German writer George

Heym (no. 3, “A Boundless, Boundless Evening”).

The complete list of the sources for Barber’s songs, including the few not mentioned here, is presented in Table 22, pp. 248-49. 173

" Table 10. The Songs of Samuel Barber

Description Number %

Total number of Songs 80

Total number of Sources 40

Arithmetic average number of songs per source: 2.00

Median number of songs per source: 1.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 30 37.50%

Texts from poets contemporary with composer (bom within 15 years of composer) 11 13.75%

Texts from poets younger than composer (bom 15 years after composer) 1 1.25%

Texts from undated sources (dates of poet or source unknown) 31 38.75%

Texts from old sources (dated more than 100 years before composers birth) 7 8.75%

Texts from non-native sources (either in translations or in other languages) 22 27.50%

Approximate number of poets with whom the composer had personal contact: 3 174 Barber’s Number of Songs and Sources

The Work-list for Samuel Barber which Barbara Heyman prepared for The New Grove

Dictionary o f Amencan Musiâ^ was consulted as the primary guide for this study. It was one of the most recently compiled lists of Barbel’s works, and presents a comprehensive catalogue of both his published and unpublished songs. Her list was compared with the one by Don A. Hennessee.^^ Heyman’s inventory contains sixteen unpublished songs not mentioned by Hennessee, all of which are included in this study. The count of 80 songs include 44 unpublished songs and 36 published songs. Several of Barbel’s important vocal works (e.g., his op. 3, DoverBeacH) could not be included in the count, based on the criteria established for this study.

The 40 sources Barber used were also determined from Heyman’s Work-list.

Unfortunately, no comparisons of the sources given by Heyman for the unpublished songs could be made, since hers is the only works-list that identifies these sources. The complete roster of Barber’s textual sources is presented in Table 22, pp. 248-49. There is one special case in this area for Barber the group of ancient anonymous Irish texts wh>ch he used for the Hermit Songsate counted and listed separately from other anonymous texts in

Table 22, because they were chosen and set as a distinct group of texts (as was done by

Hugo Wolf in his Spanish and Italian songbooks).

With 80 songs from 40 sources. Barber averaged 2.00 songs per source. The median average and the modal value are both 1.00. These averages are quite low in the twelve- composer comparison. The large majority of his sources (32) were used once each. He selected two to eight poems from six of his authors, and his largest sources were two sets of anonymous texts, the ten Irish poems and the fifteen miscellaneous anonymous verses.

^^Don A. Hennessee, Samuel Barber: A Bio-Bibliography (Westport, Connecticut and Lcndc Greenwood Press, 1985). 175

Like Charles Ives, Barber was eclectic in his poetry choices, and was not committed to using large amounts of material from his authors. Like Henri Duparc, Barber used a

consistently lownxaxibtT of texts from his sources.

Barber’s Authors Categorized by Birth Date

Barber’s preference for older, romantic poems for his songs is to be seen in the breakdown of his choices by date. Thirty of the texts were by authors bom 16-100 years before him, eleven were by authors bom within 15 years of him, and one was by an author bom more than 15 years after him. Restated in percentages, 37.50% of his total selections were by authors bom between 1810 and 1894; 13.75% were by authors bom between

1895 and 1925; and 1.25% from an author bom after 1925.

Sixteen authors were represented in Barber’s choices from older generation writers.

They originated from four countries: seven were English, while three each were Irish,

American, and German writers. The largest choices in this group were the eight texts from

Joyce and the five from Rilke. Other texts here were by such important writers as

Housman, Yeats, and Frost. The lives of all but two of the authors in this group (Keller and Kingsley) spanned both the nineteenth and twentieth centuries.

The contemporary poets from whom Barber selected texts formed a much different type of group. All but one of the nine contemporaries were Americans and Barber used only one text from each of them. Two of these poets are marginally American: José Garcia Villa is a Filipino who has lived as an adult in the U. S., and American-bom Frederic Prokosch has spent much of his life in Europe. The true exception in both nationality and frequency of choice among Barber’s contemporaries was Robert Graves, an English writer from whom he selected three texts.

Barber was aware of contemporary poems and used a relatively small amount of them for his songs, but he did not find them as satisfactory as the verses from older writers. He 176 confessed: “I did once use Theodore Roethke poems, but they were really not very near to me.”^® Two of his contemporary authors were also his friends, James Agee and Robert

Horan; this certainly influenced his choices from their works. They will be further discussed as personal contacts of the composer. Friedberg suggested that the contemporary yet conservative Agee was a logical choice for the maturing composer:

By 1938, Barber had been successfully setting poetry for over a decade, and in all these early songs had exhibited the free-flowing Romanticism of the late nineteenth-century tradition to a degree that at once appalled and delighted the avant-garde saturated audiences of the 1930s. Not surprising, then, is the strong artistic affinity which Barber developed for Agee, whose literary style also carried Romantic overtones, and whose life, as well, embodied the passion, conflict, and self-destructiveness of the archetypical Romantic hero.^^

A text by a younger poet was used for one of Barber's last songs. The surrealist verse was by the Polish author Jerzy Harasymowicz, translated by Czeslaw Milosz, used in the song “A Green Lowland of Pianos,”op. 45 no. 2. Both the composer and the author were middle-aged at the time the song was composed: in 1974, Barber was 64 years old and the poet was 41.

Barber was drawn to Harasymowicz’s text more for its sarcasm and humor than for a desire to work with poetry from younger authors. In fact, he confessed that he did not much care for new poetry: “I wouldn’t think of setting Allen Ginsberg.... I admit that I often get bored looking through modem poetry for texts, so little of it seems suitable—the wrong poetry for me.”^®

The single choice from a younger writer was a late exception to Barber's choices from older and contemporary sources. While it demonstrated how far afield Barber could go for

^®Quoted in Ramey, op. cit., p. 2.

^^Friedberg, op. cit.. Ill, p. 10.

78Quoted in Ramey, op. cit., p. 2. 177 a song text, Barber’s lack of real interest in “new” poetry may be s-^en in his lack of further choices by younger writers.

Barber’s Choices of Undated and Old Sources

Barber’s use of undated texts was the highest for the twelve composers studied. He used thirty-one undated or anonymous Texts, 38.75% of his total choices.

Seven of the anonymous texts were used for the early, unpublished “Mother Goose

Songs” of 1920-23. Heyman calls them “Seven Nursery Songs.”^^ There were eight additional unpublished songs which had either anonymous or unknown authors.

Song Title Date Source

Sometime 1 9 1 7 Unknown

In the Firelight 1 9 1 8 Unknown

Thy Will be Done C 1 9 2 2 “The Wanderer,” Exeter Book

Dere Two Fella Joe 1 9 2 4 Unknown

Lady When I Behold the Roses 1 9 2 5 Anonymous Elizabethan

Au clair de la lune 1 9 2 6 Unknown

Hey Noimy No 1 9 2 6 Christ Church Manuscript

The Shepherd to his Love... 1 9 2 8 Unknown

Three of these texts are from ancient anonymous British sources; they will be mentioned again under Barber’s choices from Old Sources.

79-H-,Heyman, op. cit., p. 255. 178 Heyman also named five authors (for six of Barber’s song texts) which have not been identified or dated by this author:

Sonc Title Date Author Why Not 1917 K. Parsons Little Children of the Wind 1924 F. MacCleod Longing 1924 F. MacCleod La Nuit 1925 A. Meurath Ask Me to Rest 1926 Edward Hicks Streeter Terry Watchers 1926 D. Cornwell

Edward Hicks Streeter Terry was found to be an American writer who had three books of poems published in Philiadelphia between 1910 and 1916, but his birth and death dates were not discovered.*^ Heyman’s Work-list is the only place many of these songs are mentioned. Hennessee’s bibliography offers no assistance because, as mentioned earlier, it does not include many of Barber’s unpublished songs, nor are there indications of the authors or sources for those which are given.

The most prominent of Barber’s choices of anonymous texts were the ten ancient Irish texts he used for the H erm it Songs. These verses were the private, personal expressions of monks and scholars, written in the margins of the manuscripts they were copying or illuminating. Barber studied these Irish writings extensively. The texts for the ten songs came from five different sources: Philip Allen’s The Romanesque Lyric, Kenneth

Jackson’s A Celtic Miscellany, Sean O'^aoXkwLsJhe Silver Branch,anà two sets of translations prepared for Barber by Chester Kallman and W. H. Auden.**

The National Union CatalogPre-1956 Imprints, 685 vols. (London and Chicago: Mansell, 1978), DLXXXVII, p. 358.

** Kreiling, op. cit., p. 246. 179 The anonymity of the authors was surely a factor in Barber’s choice of these verses.

. .[T]he texts of the H ermit alternately portray human sharing and human isolation; indeed, the cycle as a whole celebrates both community and solitude.” The isolation portrayed in the texts is made more poignant by the very lack of identification of the authors themselves.

Seven of Barber’s texts were from Old Sources, i.e., from authors bom more than 100 years earlier than the composer (for Barber, before 1810). These comprise 8.75% of his total choices. There are seven authors in this group, from a variety of eras and nationalities.

Author Nationalitv Centurv Song Title Song Date Bhartrihari Indian 5th-6th Peace 1935 E. Browning English 19th Thy Love 1926 W. S. Landor English 18th-19 th Mother I cannot... 1927 Meleager Greek 1st BC Love at the Door 1934 P. B. Shelley English 18th-19 th Music When... C1926 J. G. Whittier American 19th Isabel 1919 T. Wyatt English 16th An Earnest Suit... 1926

As this chart shows, all of Barber’s songs from Old Sources were early compositions, and none of them was published. The wide scope of the authors in this group also emphasize Barber’s literary depth.

Many of Barber’s choices from anonymous sources also date from antiquity, such as the three among his early, unpublished songs. Barber combined three verses from a poem called “The Wanderer” for the song “Thy Will be Done.” “The Wanderer” was recorded in the Exeter Book, which dates from the 10th century.He also composed a setting of

82“Xhe Exeter Book" and “The Wanderer,” Tbs Oxford Companion to English Literattire, ed. Margaret Drabble, 5th ed. (Oxford: Oxford University Press, 1985), pp. 333 and 1042. 180

“Lady When I Behold the Roses,” an anonymous Elizabethan text first printed in 1598.*3

The song “Hey Nonny No,” is based on a verse from the Christ Church Manuscript, and dates from the early seventeenth century (Ault dated the poem “not iater than 1620”).*^ A more important group related to antiquity are the Irish texts in the Hermit Songs, which date from the 8th to the 13th centuries.

The choices from authors dated as Old Sources seem relatively insignificant in Barber’s overall selection. The variety of these authors, however, shows the broad scope of

Barber’s textual choices. Further, when combined with his choices of anonymous texts which date from antiquity, the full importance and magnitude of this aspect of his selections may be fully comprehended.

Barber’s Choices in Non-native Languages

Twenty-two of Barber’s texts were originally in languages other than English. This number represents 27.50% of total choices.

The largest group of his non-English texts were the ten Celtic verses he used in the

Hermit Songs. As previously noted, these texts were taken from several sources, and

Barber commissioned Kallman and Auden to prepare some of the translations especially for this project. The breakdown of the songs by translator is as follows:

Translator Song Numbers Sean O’Faolain 1, 4, 6, and 10 Howard Mumford Jones 2 and 5 Chester Kallman 3 Kenneth Jackson 7 W. H. Auden 8 and 9

J. Kennedy, An Introduction to Poetry, 3rd ed. (Boston: Little, Brown, and Co., 1974), pp. 103-4.

^Elizabethan Lyrics, 3rd ed., ed. Norman Ault (New York: William Sloane, 1949), p. 516. 181

Kreiling’s review of the composer’s sketches for the Hermit Songs revealed important background details regarding his study and use of these Irish texts. ^5 For example, Barber consulted at least two additional sources on Celtic verse besides those from which the final selections were taken. One of them contained an alternate translation of “The Hermit’s

Song” (which Barber set as no. 10, “The Desire for Hermitage”); Barber probably replaced two lines of O’Faolain’s translation of this text with ones derived from the secondary version.*^ Finally, at least one poem originally considered for musical setting was later dropped from the final group.^?

The second major group of songs Barber composed on non-English texts are the five songs in his Mélodies Passagères, op. 27. The author of these poems, Rainer Maria Rilke, was even more of an international traveler than Barber. He was German by birth, lived in

Paris for several years, and “...spent substantial amounts of time in Russia, Italy, Austria, and Switzerland, eventually applying for citizenship in Switzerland.”^8 He wrote poetry in three languages (German, Russian, and French), and made German translations of poems originally in French, Italian, Russian, Danish, and English.*®

The composer himself provided some comments on his use of Rilke’s French verses:

Almost everything I’ve done has been in English, but I happen to speak fluent French, so there seemed no reason not to set those Rilke poems. I was living in France at the time [1950-51 ] and was, I suppose, in a French mood. Those particular poems were written while Rilke was living in Paris (he was secretary to Rodin), and he actually wrote them in French.®®

*^Kreiling, op. cit., pp. 245-266.

86/6i

Ibid., p. 254.

^^Ibid., p. 274-5.

*®7b/d., p. 275.

®®Quoted by Ramey, op. cit., p. 2. 182

Barber’s seven remaining songs using texts from non-native sources are a mixture of published and unpublished works, from five languages:

Language Author Title Date Publication French Meurath La nuit 1925 Unpublished French Unknown Au clair de la lune 1926 Unpublished Greek Meleager Love at the Door 1934 Unpublished Hindi Bhartrihari Peace 1935 Unpublished German Keller Now I Have Fed... 1972 Op. 45 no. 1 Polish Harasymowicz A Green Lowland... 1972 Op. 45 no. 2 German Heym O Boundless... 1972 Op. 45 no. 3

Kreiling noted the diversity of the final three songs: “With op. 45, Barber displayed an understanding of three poems whose national origins, styles, and themes could hardly have varied more.... Like Barber’s settings of American texts, these represent less a national consciousness than a literary one....”®' The use of the Keller poem relates more about

Barber’s interest in James Joyce, the translator of the version Barber used, than the original poet.

Choices of texts from non-native languages figured prominently in the songs of Samuel

Barber; his percentage in this category is one of the highest of the twelve composers.

Barber’s life of international travel, his ability to speak and read in foreign languages, and his interest in world literature are all mirrored in these textual selections.

Barber’s Personal Contact with the Poets

Samuel Barber had limited contact with the poets of his song texts. Three names are marked with an asterisk in Table 22, p. 248-49. One of the three is the composer himself.

Ibid., pp. 315-316. 183

He wrote one of his own texts for an early, unpublished song (“October Weather”). James

Agee and Robert Horan are the only two writers Barber personally knew.

James Agee and Barber, according to Kreiling, “.. .maintained a personal friendship:

Agee read his poems aloud at Barber’s Capricorn estate, and Barber encouraged Agee to write opera lib retto s.”®^ Robert Horan also visited Barber at the Capricorn house:

“.. .Horan, a friend of Barber and Menotti, wrote his collection of poems entitled A

Beginningat their Capricorn estate, and dedicated the volume to them b o t h . ” ^ 3

Two of the translators of some of the Irish poems Barber used for the Hermit Songs were in close contact with the composer. Barber asked Chester Kallman and W. H. Auden to translate verses for this project. This connection, although not an official part of the study, was certainly important to the creation of one of Barber’s most important groups of songs. In addition. Broder reported Barber’s contact with several authors while searching for a libretto: “He has discussed opera projects with Thornton Wilder, Stephen Spender, and Dylan Thomas, among o t h e r s . ” 94

Dr. Robert Thomas Kerlin may have been known to the young Barber. Kerlin

completed his study of Negro Poets and their Poems (Washington, D. C.: Associated

Publishers) in West Chester, Pennsylvania, in 1923, while Barber still lived at home. A

year later Barber used one of Kerlin’s poems for the unpublished song “My Fairyland.”

Personal contact with the authors of his songs texts was not a prominent part of

Barber’s life and work. His concern with the themes of the poems was more important in

his selection process than with knowledge of the poets themselves. As demonstrated by

his background study of James Joyce, he compensated for this lack of direct contact with

his authors by research and reading about them.

92ibid., p. 162.

p . 2 0 9 .

^'*Broder, op. cit., pp. 45-46. 184

NED ROREM

Rorem’s Educational and Literary Background

Ned Rorem (b. 1927) wrjs bom in Richmond, Indiana; his father was teaching accounting at Earlham College. His family moved to Chicago when he was eight months old.95 He graduated from high school in June, 1940, when at “...age sixteen, despite low grades, [he] emerged from U-High..., and the following autumn...was accepted at

Northwestern University’s School of Music, despite the same low grades, on [his] creative potential. "96

He remained at Northwestern for two years, attended Philadelphia’s Curtis Institute in

1943, and eventually earned his two college degrees at the Juilliard School in New York city, a BS degree in 1946 and a MS degree in 1948. He completed his formal education with these degrees.

Holmes succinctly described Rorem’s hectic life in the years after 1948:

The following year he went to France; after studying on a Fulbright scholarship with Arthur Honnegger and spending extended periods in Morocco, he settled in Paris (1952), where his intelligent charm and sometimes outrageous behavior won him the patronage of the Vicomtesse Marie Laure de Noailles and acceptance in the cultural milieu of Jean Cocteau, Francis Poulenc, and Georges Auric. During eight years spent in Morocco and France, Rorem produced many works in a wide variety of forms but concentrated mainly on songs.97

Rorem permanently returned to the U. S. in 1958. Since then he has earned his living mainly as a composer, although he has held three teaching posts, at the University of

96Ned Rorem, “The Piano in My Life,” Setting the Tone (New York: Coward-McCann, 1983), p. 18.

9676id., p. 22.

97james Holmes, “Rorem, Ned,” The New Crovc Dictionsiy o f American Music, 4 vols., ed. K. Wiley Hitchcock and Stanley Sadie (New York: Macmillan, 1986), IV, p. 87. 185 Buffalo (1959-60), the University of Utah (1965-66) and the Curtis Institute (1980-86).

Ke was also composer-in-residence at the University of Iowa in 1965-66.

He has earned many scholarships, awards, and commissions, including two

Guggenheim fellowships (in 1956-57 and 1978-9) and election to the Institute of the

American Academy of Arts and Letters in 1979. He won the Pulitzer Prize for music in

1976, for his orchestral suite A ir Music.

His literary tastes began developing early. He claimed that “At fourteen [he] knew by heart the poems of Cummings.’’^» A list of authors he was reading in Hyères, France, in

1954, includes Petronius, Isherwood, Maun?.c, Musset, Maupassant, Balzac, Ray

Bradbury, and Christopher Fry, in addition to the biography of Apollinaire and proofs of a novel by James Lord.^c He admitted to another side of his reading habits a year later:

Yes I read and I read and there is never a time when I haven’t two or three books under way. But I read junk too. I read more than I compose. I never read what bores me, and cannot read to leam. For we find in books we like only a rewording of what we already knew.'oo

Many years later he commented further on his literary pursuits: “My need for poetry is utilitarian. I read it not for pleasure but to keep abreast of friends’ work or to seek texts for music. I don’t dislike poetry, though it is hard for me. Because I cannot understand it I make songs of it.”i°i

Ned Rorem has demonstrated his literary side most dramatically through his own writings. He has authored twelve books. The earliest ones, from the 1950’s were personal reflections in the form of diaries. Later volumes, beginning with books from the

1960’s, feature formal discussions and casual commentaries on a variety of musical topics,

®*Ned Rorem, The Paris and N ew York Diaries (San Francisco: North Point Press, 1983), p. 247.

^^Ibid., p. 154.

p. 190.

*°*Ned Rorem, PUns Contraption (New York: Holt, Rinehart and Winston, 1974), p. 142. 186 especially song composition. Many of his most recent books . .combine some personal memoirs with criticism, reviews, and the author’s unique observations on the contemporary artistic s c e n e .

He considers these literary endeavors to be secondary to his work as a composer. He stressed that he is “.. .a composer who happens to write, not an author who happens to compose....[his] prose and [his] music are responses to unrelated d r i v e s , ” '®^ in spite of these comments, Rorem’s skills as a writer and composer show the fullness of his intellect, and place him in company with such writer-composers as Schumann, Wolf, Debussy, and

Poulenc.

Rorem’s Poetry Choices

Ned Rorem was recently asked about his method of choosing the poetry for his songs.

He gave the following response:

For song texts I choose whatever, as the Quakers say, speaks to my condition. I might love a great poem but feel that music would be superfluous, while composer X might find the same poem ideal. Another poem, less great, might be more appropriate. I don’t need to understand a poem (whatever that means) so much as to react kinetically to it; sometimes the very act of setting it reveals a meaning—a meaning perhaps far from the poet’s own. Different poems satisfy different requirements at different times of life.'"'^

Rorem has most frequently chosen the poetry of Paul Goodman for his song texts.

Many of Goodman’s poems were used for early settings, such as the Seven L ittle Prayers of 1946, the Three B lu eso f 1947, and a number of individual works. Perhaps the most famous Rorem song with a Goodman text is the “Lordly Hudson,’’ also composed in 1947. lO^Friedberg, op. cit.. Ill, p. 210.

103“jiie NATS Bulletin Interviews Ned Rorem,” The NATS Bulletin XXIX:2 (November/December 1982), p. 5.

^0‘*lbid. 187

Three settings of Goodman texts which date from the 1950’s were published as Thee

Poems o f Paul Goodman, “For Susan” (1953), “Clouds” (1953), and “What Sparks and

Wiry Cries” ( 1956). Rorem stopped setting Goodman’s verses by the mid-1950’s.

Texts from other contemporary American authors are prominent among his choices.

Songs on a group of six poems by Harold Norse were written in 1949, with the title Penny

Arcade. Rorem has composed thirteen songs on verses by Theodore Roethke, most of which date from the 1950’s. Ten verses by Howard Moss were set as a group in 1960-61, called King Midas. Six Kenneth Koch poems were chosen for the cycle Hearing, composed in 1965-6.

He has also selected small numbers of texts from other friends and contemporaries. He composed songs using two texts by Elizabeth Bishop (“Conversation” and “Visits to St.

Elizabeths,” both 1957) two by Jack Larson (nos. 6 and 12 in Poems o fLove and the

Rain, 1962-3), and two by John Ashbery (“Fear of Death” and “Thoughts of a Young

Girl” in The Nantucket Songs, 1979). There are also single choices from the poetry of

Donald Glaze (“A Journey,” 1974) and Adrienne Rich (“The Stranger,” in Women's

Voices, 1975-6).

His textual choices also include works from important nineteenth century authors, such as Whitman and Tennyson. Miller commented that, “Like so many other composers, but in his own individual way, [Rorem] has reacted to the rugged lines of Walt Whitman.”^°^ An early group of Whitman songs was published as Five poems o f Walt Whitman. Rorem’s important cycle War Scenes, composed in 1969, “.. .shows a grimmer side of the poet and a new angularity in the m u s i c . ” He returned to Whitman for the Three Calamus Poems, composed in 1982. One of the Tennyson songs was written in 1949, “The Sleeping

Palace;” the other three date from 1963 (“Ask me no more,” “Far-far-away,” and “Now

*°^hilip Lieson Miller, “The Songs of Ned Rorem,” Tempo CXXVII (December 1978), p. 26. 188

Sleeps the Crimson Petal”). These were published together in 1963 as Four Poems o f

Tennyson.

Rorem has also chosen texts from a number of ancient writers. One early song

(“Catullus: On the Burial of His Brother,” 1947) uses a translation of a verse by an ancient

Roman poet. For another early setting he used a verse by the medieval English poet

Chaucer. In 1964 he composed the Two Poems o f Plato. Nine of his songs are on biblical texts, including the popular Cycle o fHoly Songs aï 1951.

Works by foreign authors are also part of his selection. Rorem translated and set three prose-poems by Julien Green in 1951 and titled the set Another Sleep. Green is an

American novelist who has lived in Paris much of his life and writes in French. A group of six songs, Poèmes pour la paix(l9S3), are settings of verses by French poets on the theme of peace. His Anacreontiche 0954) uses four poems by the Italian Jacopo Andrea

Vitorelli. Over a century before Rorem, Schubert had set two Vitorelli verses. Another group of three songs from 1954 (“Abel,” “Guilt,” and “The Lord of Fear”) are settings of

English texts by the Greek poet Demetrios Capetanakis.

The authors mentioned here are most of Rorem’s important zind unusual textual selections. Some of his other sources for song texts will be detailed later; the complete list of Rorem’s sources is found in Table 23, pp. 250-52. 189

Table 11. The Songs o f Ned Rorem

Description Number %

Total number of Songs 186

Total number of Sources 64

Arithmetic average number of songs per source: 2.91

Median number of songs per source: 1.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 31 16.67%

Texts from poets contemporary with composer (bom within 15 years of composer) 77 41.40%

Texts from poets younger than composer (bom 15 years after composer) 0 0.00%

Texts from undated sources (dates of poet or source unknown) 14 7.52%

Texts from old sources (dated more than 100 years before composer’s birth) 64 34.41%

Texts from non-native sources (either in translations or in other languages) 17 9.14%

Approximate number of poets with whom the composer had personal contact: 14 190 Rorem’s Number of Songs and Sources

A total of 186 of Rorem’s songs were counted for this study. This number was gathered from three lists: a compilation at the end of one of his books,one from a magazine article, and the works-list prepared by James Holmes for The New Grove

Dictionary o fAmerican Music,^^ which Mr. Rorem personally recommended to the author. * Each of these three lists contained a fair number of items not on the other two, making the final count more a compilation than a comparison of the lists.

The majority of the works counted here are published songs; only a few known unpublished works surfaced in the investigation. Holmes mentions “ 125 unpublished songs” which were not catalogued individually and were thus not available for study. ’ ‘ ■

Several important song cycles were not included in the count because they were originally written for voice with orchestra or with obbligato instruments.

Rorem used a total of 64 sources for the songs in this study. Most of these were confirmed from the lists mentioned earlier, although Mr. Rorem himself revealed the sources for two obscure, unpublished songs mentioned in the 1968 list, “Poem for F” and

“Boy With a Baseball Glove.”' *2 The complete list of his sources is found in Table 23, pp. 250-52.

The arithmetic average for Rorem is 2.91 songs per source, using 186 songs from 64 sources. The median average and modal value are both 1.00. Rorem’s averages are

107«Published Musical Compositions by Ned Rorem,” Music and People (New York: George Braziller, 1968), pp. 235-237.

108«The Songs of Ned Rorem,” postscript to “The NATS Bulletin Interviews Ned Rorem,” The N A T S Bulletin XXIX:2 (November/December 1982), pp. 48-49.

'°®Holmes, op. cit., pp. 88-89.

' '^Letter to the author, July 22, 1988.

"'Holmes, pp. cit., p. 89.

' '^Letter to the author, August 1, 1989. 191 among the lowest for the twelve composers. About half of his sources were used for one text, indicating a weak modal value. Only four sources were used more than ten times; he took two to nine texts from the rest. With a few exceptions, he selected a fairly low number of texts from each of his sources.

Rorem’s Authors Categorized by Birth Date

Strong ties with contemporary poetry are seen in Rorem’s choices in this category.

Thirty-one of his song texts were written by authors bom 16-100 years before him, while seventy-seven were by authors bom within 15 years of his birth. No texts were chosen from authors bom more than 15 years after him. The percentage for these groups: 16.67% of his choices were by authors bom between 1823 and 1907; 41.40% were by authors bom between 1908 and 1938; 0.00% were by authors bom after 1938.

Rorem has presented arguments for using both older and contemporary poetry. For example, he has commented that, “Serious composers have always set words from the past far more than words by their contem poraries.” * He also suggested that the reason composers favor older poetry stems from a lack of usable, quality lyric verse by contemporary writers:

The modem composer understandably prefers to set poetry by his national contemporaries, but the pickings are slim. Lyric poetry is out of fashion today.... Not infrequently the composer will revert to the Bible, Shakespeare, or romantic schools of the eighteenth and nineteenth centuries to search out texts for his lyrical songs. *

He continued this discussion by advocating the use of verses by contemporary authors:

“Although there may be a scarcity of suitable poetry, the practice of making songs from the

H3Ned Rorem, “The American Art Song,” Setting the Tone, op. cit., p. 228. i *^Ned Rorem, Music From Inside Out (New York: George Braziller, 1967), pp. 42-43. 192 work of living writers is desirable. Artists of today, whether they know it or not, have basically more in common with each other than they have with artists of the past.”* *5

Rorem’s own choices reinforce the duality of these statements. While the majority of his texts are by contemporary authors, there are actually more authors represented in the

“older generation” group. He generally used higher numbers of texts from his contemporary authors, but no more than four texts from any of the older writers.

Rorem’s Choices of Undated and Old Sources

There are 14 Rorem songs from undated sources. These are 7.52% of his total choices. The largest part of this group are the nine extracted Biblical texts he used for song settings. Six of these are from the Psalms: four in the 1951 CycleofHolySongs^Xmiwo others, “A Psalm of Praise” and “A Song of David” (both 1945). The other three are from the New Testament: “An Angel Speaks to the Shepherds” (Luke 2: 9-15), “The

Ressurection” (Matthew 27: 62-6,28), both composed in 1952, and a 1957 setting of “The

Lord’s Prayer. ” His 1946 song using the single word “Alleluia” is also counted as an undated text and should be mentioned here.

The other four texts in this group are from early anonymous sources. These include one from the 15th century, “The Mild Mother” (1952), two from the 16th century, “The

Nightingale” (1951) and “Christmas Carol” (1952), and one from about 1650, “Epitaph on

Eleanor Freeman” (1953).

Rorem was not committed to using anonymous texts, as demonstrated by the low percentage of these selections. With the exception of the Cycle o fHoly Songs, these

settings of undated texts are not frequently performed, suggesting that they may not be

some of his best efforts.

*‘^/ibid'., p. 44. 193

Rorem’s 64 choices of ancient poetry form a substantial group of authors and texts.

They comprise 34.41% of his total selections, the highest percentage of ancient texts for the

twelve composers. Texts by 27 authors from many eras and nationalities are part of this group. The poetry dates from ancient Greek and Roman civilizations, as well as from the

14th to the 19th centuries. The breakdown by nationality for these writers is as follows:

16 English writers, 5 French, 3 American, and 1 each of Italian, ancient Greek, and ancient

Roman. The early anonymous texts from the undated category are also related to this group.

It should be noted that Whitman is considered an “ancient” poet in relation to Rorem using the dating procedure prepared for this study, since he was bom 102 years before the

composer. This is a similar relationship to the one between Goethe and Hugo Wolf

discussed in Chapter I.

The high percentage in this category shows Rorem’s strong inclination for poetry from past times. It also reveals the full extent of the eclecticism in his textual selections.

Rorem’s Choices in Non-native Languages

Rorem used 17 texts which were from non-native languages. These represent 9.14%

of his total choices. Four of these were set in English translations: the three by Julien

Green which Rorem translated himself, and the Catullus poems translated by Beardsley.

The story behind the Julien Green settings began in 1951, when Green presented Rorem

with the gift of a small book soon after they first met. Rorem completes the story:

The little book was a new edition of his 1930 memoir. L ’autre sommeil. I read it on April 15 in the waiting room at Marignane before boarding a plane for Casablanca. En route, I translated three extracts, and during the following week in Marrakech composed a baritone cycle of this English prose, calling it Another Sleep ... .My translations are not good, nor are they really Green, nor yet me.>

llôNed Rorem, Pure Contraption, op. cit., pp. 108, 110. 194

The other thirteen of his non-English texts were set in their original languages. Seven are in French, including the six by various poets in the set Poèmes Pour la Paixanà the single song “Jack L’Eventreur” on a text by his Parisian friend, Marie Laure de Noailles.

Rorem vented his feelings about Americans who use French verse for song settings:

Frustration awaits the American composer impelled to write songs in French, for those songs will seldom be heard. The rare French recitalist who programs an American song will make an effort to leam one in English. Meanwhile American singers find it more ‘legitimate’ for their French group to be by Frenchmen.

It is logical that Rorem, during the years he was living in France, would be compelled to set French verse. Less obvious are his choices and settings of six poems in Italian and

Greek, the four Vitorelli verses set in 1954 and the two Plato verses ten years later.

It is somewhat surprising that Rorem did not make further use of non-English texts

(especially French verses), given his extensive travels and language abilities. The influences of European culture did not have a strong impact on the textual choices for

Rorem’s songs, as they did for the other American composers in this study.

Rorem’s Personal Contact with the Poets

Fourteen of Rorem’s authors have been identified as having personal contact with the composer. They are marked with an asterisk in Table 23, pp. 250-52.

The most significant of these relationships was with Paul Goodman. The two were friends from childhood (Eduoard Roditi was another childhood friend). Rorem felt that his

Goodman settings are some of his best, and has called the poet his Goethe, his Blake, and his Appolinaire.*’® He eloquently eulogized Goodman’s death in 1972.

• l"^Ned Rorem, Introduction to the published songsPoèmes Pour la Paix (New York; Boosey & Hawkes, 1969).

^**Rorem, Pure Contraption, op. cit., p. 98.

^^^"Remembering a Poet," Ibid., pp. 97-101. 195

Many of Rorem’s other contacts are New York poets: John Ashbery, Kenneth Koch,

Frank O’Hara, Kenneth Pitchford, and Donald Windham. Mr. Rorem described Ashbery

as “a valued acquaintance .” '20 He also admitted that choices from the New York poets can

cause problems: “Each setting provokes a grave contretemps with the poet, while the few

remaining [New York] poets I’ve not set implore me.”'2i

Several of the authors have been close friends. He has written much about his close

relationship with Marie Laure de Noailles while living in France. He still calls Los Angeles

writer Jack Larson “a dear friend .” '22 Rorem has known some of the authors on a more

limited basis. His contact with Julien Green was intense, but limited to France and the

1950’s. He counted Harold Norse as “a good friend of the 1940’s, but [has] scarcely seen

him since.”'23 He was in contact with Elizabeth Bishop through correspondence for many

years.'24 He also mentioned Jean Garrigue as “a passing acquaintance”'2‘ and indicated

that he has met Adrienne Rich twice.'26

Personal contact with the authors of the song texts is clearly important to Rorem.

Evelyn Pool reinforced this aspect of his work: “Rorem believes in consulting the poet, if

living.”'27 The composer himself once wrote: “Poets’ names are seldom seen on song

'2"Letter to the author, August 1, 1989.

'21/&yj., p. 143.

'22%jetter from Ned Rorem to the author, August 1, 1989.

'23jrfeid

'24jhis correspondence is mentioned in Rorem’s article “Poetry and Music,” most recently reprinted in Settling the Score (New York: Harcourt, Brace, Jovanovich, 1988), pp. 294-295.

'2^Letter from Ned Rorem to the author, August 1, 1989.

'26/6id.

'27Evelyn Ivora Pool, “A Study of Ned Rorem and his song cyclePoems of Love and the Rain, ” Master’s Thesis, The Ohio State University, 1966, p. 8. 196 covers and usually omitted from printed song programs. Yet where would the song be without them?” *2® -

' “®Rorem, “Poetry and Music," Settling the Score, op. cit., p. 296. 197

DOMINICK ARGENTO

Argento’s Educational and Literary Background

Dominick Argento (b. 1927) is a highly educated man. He grew up in York,

Pennsylvania and attended public school there. He graduated from York High School in

1945 and afterwards enlisted in the U. S. Army. From 1945 to 1947 he served as a cryptographer in North Africa. His college education began in 1947 at the Peabody

Conservatory in Baltimore, Maryland; he graduated in 1951 with his BM degree. A

Fulbright grant took him to Florence, Italy, in 1951-52, where he studied piano and composition.

Argento returned to the U. S. and the Peabody Conservatory to work on a Masters degree, which he completed in 1954. He was then “eager to pursue his studies f u r t h e r , " so he entered the Eastman School of Music in Rochester, New York. He earned his doctorate in music there in 1957.

In that same year he returned to Florence on a Guggenheim fellowship, but returned in

1958 to begin his long teaching career on the faculty at the University of Minnesota. He has continued in this capacity to the present (1989), and was named Regents Professor in

1979.

His literary development began early. He was known to be an avid reader as a teenager, *30 a practice which has continued into adulthood. The high level of his taste in literature is best seen in his choices of texts for his operas and songs: “Much of the music

Baker’s BiographicalDictionary o f Musicians, 7th ed., ed. Nicolas Slonimsky (New York: Schirmer Books, 1984), p. 73.

*30«^gento, Dominick,”Current Biography Yearbook 1977, ed. Charles Moritz (New York: H. W. Wilson, 1977), p. 32. 198 that he has composed for the voice reflects his taste in literature and, by enhancing the texts, shows the appreciative quality of his reading.”‘3i

Argeoto’s Poetry Choices

Argento has composed four song cycles for voice and piano, each of which concentrates on the textual material of either one author or a group of related authors.

His first cycle. Songs about Spring, was written in 1954, while he was still an undergraduate student at the Peabody Conservatory. His composition teacher had suggested that he try writing some songs, since he had submitted only instmmental music as his composition projects. Argento recalled the process of finding appropriate texts for his assignment:

I obediently went to the library, checked out an armful of books of poetry, and within half-an-hour I selected a group of poems by e.e. cummings which seemed the right length. I finally picked three (later I added two more) all dealing with spring and made a little cycle out of them for soprano and piano. The poems themselves are very pretty, but I did not choose them because spring had any more significance for me than any other season, or because the texts were particularly inspiring. I selected them quite coolly and dispassionately because I had to have some songs ready for my next lesson. *32

The second cycle he composed was the now-popular set of Six Elizabethan Songs of

1958. The process for choosing these texts was as random as the one used for the student pieces:

My wife and I were living in Florence the year after we finished our schooling at Eastman. A friend of ours, the tenor Nicholas DiVirgilio, still back in Rochester, asked me to write a group of songs quickly for him to perform on his graduation recital. The bookstores of Florence offer a very limited hunting ground for English poetry, but I did find a collection of Elizabethan lyrics and said to myself: ‘this will have to do.’ Again, hardly a question of inspiration or sudden revelation. 1^3

p . 34.

^^^Domlnick Argento, “The Matter of Text,” The N A T S Journal 44 /4 (March/April 1988), p. 6. p. 7. 199

Argento admitted . .that in both of these cycles.. .[he] more or less set to music the first convenient text that came to hand; they had no special ^p eal for [him] apart from their readiness to become songs....He also believed that “After these, however, [his] seicciion of texts became more and more discrim inating. ’’^^s

As indicated by the title, eight entries from Virginia Woolfs diaries were chosen for his third song cycle. From the Diary o f Virginia Woolf. This cycle was composed in 1974.

The diary entries he used dated from April 1919 to March 1941; the songs follow the chronological order of the entries. In this cycle, Argento musically “.. .elaborated on a single tone-row, accommodates plainsong, café tunes, and a virtual operatic scena in the course of eight songs that trace a woman’s life to the brink of suicide.”'^^ The cycle won the Pulitzer Prize for music in 1975.

Argento’s most recent song cycle was written in 1982, The Andrée Expedition. The texts for this set of thirteen songs are from the letters and journals of three Arctic explorers tragically killed on an expedition in 1897. The trip was an attempt to reach the North Pole by hot-air balloon, headed by Swedish explorer Salomon August Andrée. Two young men accompanied Andrée as assistants, Knut Fraenkel and Nils Strindberg. They left from

Danes Island, Spitzbergen, on July 11th, followed their course for three days, but bad weather forced them o ff course. They travelled by land for many miles, but died together in early October.

The three explores were presumed lost after many attempts to find them failed. Thirty years later their remains were found by accident, when a group of fisherman found their camp, completely intact, on the remote island where they had ended their joumey. The

*^®Maiy Ann Feldman, “Argento, Dominick,” The New Grove Dictionary o f American Music, 4 vois., ed. H. Wiley Hitchcock and Stanley Sadie (London and New York: Macmillan, 1986), I, pp. 64-65. 200 cause of their deaths was never conHrmed. Stefansson speculated on the matter in

1938,‘37 but the 1952 report by a Danish doctor, E. A. Tryde, which implicated the eating of poisoned polar bear meat, produced the most viable cause of death.‘38 Each of the explorers recorded his story in separate journals, which were found and recovered with the other remmns of their camp. These journals were soon published together in a number of translations, including one in English.‘3^

The themes of self-realization and personal tragedy begun with the Woolf cycle (and other works, esp. Argento’s operas) were continued in The Andrée Expedition. One reviewer of the February 1983 premiere wrote that, “Again, as in the Woolf cycle, Argento manages to find in the material a kind of universal significance and to imbue the text with a range of emotion that makes these explorers’ apparently futile quest seem pathetic but ultimately heroic. A great deal of life is lived in these songs.”‘‘^

‘ 37 Vilhjalmur Stefansson, “How did Andrée die?,”Unsolved Mysteries o f the Arctic (New York: Macmillan, 1939), pp. 192-321.

‘38gee the Introduction to the historical novel on the subject by Per Olof Sundman,The Flight o f the Eagle, tr. Mary Sandbach (New York: Pantehon Books, 1970), p. 7.

‘39 AM., p. 5.

‘^‘‘Michael Anthony, review of the premiere performance ofThe Andrée Expedition, Hi Fi/Musical America 33/6 (June 1983), p. 30. 201

Table 12. The Songs of Domioick Argento

Description ______Number % Total number of Songs 32

Total number of Sources 10

Arithmetic average number of songs per source: 3.20

Median number of songs per source: 3.00

Modal value (most occurring) number of songs per source: 1.00

Texts from poets older than composer (bom 16-100 years earlier than composer) 26 81.25%

Texts from poets contemporary with composer (bom within 15 years of composer) 0 0.00%

Texts from poets younger than composer (bom 15 years after composer) 0 0.00%

Texts from undated sources (dates of poet or source unknown) 0 0.00%

Texts from old sources (dated more than 100 years before composer’s birth) 5 18.75%

Texts from non-native sources (either in translations or in other languages) 13 40.63%

Approximate number of poets with whom the composer had personal contact: 0 202 Argento’s Number of Songs and Sources

The four song cycles by Argento counted in this study are comprised of 5, 6,8, and 13 songs (in chronological order), making a total of 32 songs. The most comprehensive list of Argento’s works, used as a guide for this document, is Feldman’s in The New Grove

Dictionary o f American Music.^^^ Several less-comprehensive lists of his works were consulted for comparison. It should be noted that Argento’s important song cycles which use instruments in addition to or besides piano were not included as a part of this study, in adherence with the established procedure.

The ten sources for Argento’s songs were not all given in Feldman’s list. The names of the authors of individual texts of the Elizabethan Songs and The Andrée Expedition were gleaned from the published scores of these two works. The list of sources for

Argento is given in Table 24, p. 253.

The arithmetic average of 3.2 songs per source for Argento is based on 32 songs from

10 sources. The median average is three and the modal value is one. These averages show that Argento often used more than one text from each of his sources, while not using great amounts from any of them. Four of the ten authors were chosen once each, making the modal value somewhat weak. None of the authors was chosen more than eight times.

Argento chose a fairly lowiaimber of texts from his sources.

Argento’s Authors Categorized by Birth Date

Argento’s textual choices are chronologically one-sided. Twenty-six of his texts were from authors bom 16-100 years before him, while none were from his contemporaries

(bom between 1912 and 1942) or younger poets (bom after 1942). The first group (from authors bom between 1827 and 1911) are 81.25% of his choices; the other two groups are both 0.00% of the total.

‘'**Feldman, op cit., p. 65. 203

There are no readily discernable reasons for these percentages. For example, the fact that Cummings was%a writer from an older generation is essentially irrelevant, in light of

Aigento’s randomness in the selection of these texts. A more important consideration in this instance might be the prominence and availability of the books of Cummings’s poetry in the Peabody Conservatory library (the place where Argento found these books) in the

1950’s. The texts from Woolf and the Arctic explorers are more significant to Argento for their private and self-revealing qualities than for the actual choice of authors.

Consequently, the dating of these v/riters relative to the composer in these cases is also a somewhat unimportant consideration.

The lack of selection from contemporary and younger-generation authors is worth noting. Argento is the only composer of the twelve who has not chosen any texts from contemporary writers (even Wolf tried it once!). Argento has lived long enough to use works by contemporary and younger authors. Whatever the reason, his “one long exercise in self-discovery” •'*2 as a composer has so far been at the exclusion of more recent writers.

Argento’s Choices of Undated and Old Sources

Not one of Argento’s texts is from an undated or anonymous source (0.00% of the total). For his first two song cycles he chose from established (and therefore datable) authors. The later two cycles focus on the self-revelations of individuals as expressed in their personal writings. Anonymous texts would not be suitable for Argento’s “self- discovery” theme, because the loss of identity is part of the essence of anonymous poetry.

The six Elizabethan texts Argento used are his only choices from authors bom more than 100 years earlier (in his case, bom before 1827). They are 18.75% of his total choices. Two of these texts are by Shakespeare, with one each by Constable, Daniel,

Jonson, and Nash. As w«th the other categories, these choices came less from a propensity

*'’2Argento, op cit., p. 10. 204 for the poetry of a given era than from their sheer availability at the time and the circumstances surrounding the need for them (as described earlier). The fact that no further texts from old sources were chosen (under less pressured conditions) reinforces this idea.

The dates for the authors and texts in this category, as with those discussed earlier, are relatively insignificant to Argento’s motives for using them.

Argento’s Choices in Non-native Languages

The thirteen prose texts which Argento used for his cycle The Andrée Expedition were his only selections of texts originally in a non-native language, in this case Swedish. They are 40.63% of his total choices.

The songs were commissioned by the Schubert Club of Saint Paul, Minnesota, fora program on the Club’s 100th birthday recital season to be given by the Swedish baritone

Hakan Hagegard. Argento was probably influenced by the nationality of the performer in using the Swedish explorers’ texts. Anthony commented in his review of the premiere that Hagegard’s "... Swedish accent added a note of authenticity to the text.” Argento used a 1930 English translation by Edward Adams Roy for these songs.

The theme of the texts was more important to the composer than their original language or the nationality of the authors. Argento described what he found to be the central message of the collection; “In The Andrée Expedition, three very different kind of men in a difficult and eventually tragic situation come face to face with the true reasons for their behavior.”^'*^ The original language for these texts was a secondary consideration in their selection.

’^^Anthony, op. cit., p. 29.

^^Ibid, p. 30.

*^^Argento, op. cit., p. 9. 205

It would seem likely for Argento to use texts in other non-native languages, especially

Italian. Argento’s parents were U.S. emigrants from Sicily, and the composer has spent much time with his wife in Florence. These have had little influence on his choices for song texts, which have been exclusively in English.

Argento’s Personal Contact with the Poets

Dominick Argento was not in personal contact with any of the authors of his song texts. W oolf and Cummings were the only two even alive when Argento was bom, and

Woolf died when Argento was a teenager. Cummings was the only author he could have had known, but there is no documentation of contact between them. Argento’s search for song texts which suit his needs have not brought him into personal contact with his authors. 206

SUMMARY FOR THE AMERICAN ART SONG COMPOSERS

American Art Song Composers’ Educational and Literary Background

The American composers in this study were all well-educated. Each completed at least one college degree, and two (Rorem and Argento) complete graduate degrees. Barber,

Rorem and Argento have all spent time teaching at the college level. Ives concentrated his post-college energies on building a successful insurance business.

The literary backgrounds of these composers are equally impressive. Ives wrote both poetry and prose, made many references to writers and literature in his instrumental music, and was familiar with a wide range of literature. Barber was extremely well-read, and studied the works of several European authors. He also used literary references in his

orchestral music. Rorem, too, is well-versed in world literature, and is himself the author

of many books. Argento has showed himself to be an avid lover of literature through the

variety and quality of the sources for his vocal works.

American Art Song Composers’ Poetry Choices

The best word to describe the textual choices of the American art song composers is

eclectic. All of them found inspiration in works by a number of American and European

(especially British) authors, and poetry from all eras are found among their choices.

There are also personal aspects to their use of texts. Ives worked extensively with his

own poetry, unlike any of the other three. Barber particularly liked the writings of James

Joyce, while Rorem favors contemporary American writers, such as Paul Goodman.

Argento has not concentrated on the works of a particular author, but sets more than one

text from each selected author to create a cycle or set of songs. 207 American Art Song Composers’ Number of Songs and Sources

The number of songs produced by the American composers which were part of this study are shown in Figure 23.

Ives fmi g 176 Barber ►♦4 ♦ ♦ ♦ 4 80

Rorem 186 Argento 32

I I I I I I I I I I I I 0 20 40 60 80 100 120 140 160 170180190

FIGURE 23: AMERICAN ART SONG COMPOSERS Number of Songs

Rorem and Ives are the most prolific song composers here. Rorem has also composed a large group of songs which remain unpublished, in addition to a number of songs with instrumental forces which were not included in this study. Barber’s song output was small yet significant. His published works are less than half of those counted in this study, and he also composed songs for voice with instruments or orchestra which were not included.

Argento has divided his time between opera and song composition, and a large percentage of his songs are with obbligato instruments and were not a part of his final count for this document. 208 The number of sources from which the American composers chose their song texts are found in Figure 24.

Ives 82 m m m Barber 3 Rorem ^ « Argento

I I I I I I I I I I I I I I I I I I 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

HGURE 24: AMERICAN ART SONG COMPOSERS Number of Sources

Ives, Barber and Rorem selected texts from a large number of sources, Argento’s groups of songs have featured the works of ten writers. The totals in Figures 23 are divided by the number of sources in Figure 24 to produce the average number of songs per source (Figure 25).

Ives 2.15

Barber t t t t i 2.00 Rorem 2.91 Argento 3.20

I I I I I I I 0 1 2 3 4 5 6

FIGURE 25: AMERICAN ART SONG COMPOSERS Arithmetic Average Number of Songs Per Source 209

These averages are all among the lowest for the twelve composers. A high level of experimentation is suggested in these low averages, Argento chose more than one text from many of his sources, but none were selected for more than eight texts.

The median average and modal values for the American composers are presented together in Figure 26.

Ives 11.00 both values

Barber 100 both values Q Modal Value

Rorem 1.00 both values 0 Median Average Argento 3.00

0 1 2 3

FIGURE 26: AMERICAN ART SONG COMPOSERS Median Average and Modal Value of Songs Per Source

The eclecticism and experimentation in the choices from Ives, Barber, and Rorem are further emphasized in this chart. Conversely, Argento’s choices of multiple verses from many of his sources are shown by his high median average.

The strength of the modal value varied among the composers. Ives and Barber selected single texts from a large number of their sources, and therefore had very strong modal values, Rorem chose relatively fewer single texts from his sources, making his modal value only somewhat strong. Aigento chose even fewer single texts from his authors, and his modal value was the weakest of the four Americans, but on a par with many of the other composers in this study. 210 American Art Song Composers’ Authors Categorized by Birth Date

The percentages of texts chosen from authors bom 16-100 years before the composer are shown in Figure 27.

Ives 35.23%

Barber 37.50% Rorem 16.67%

Argento 81.25% I I I ! T I I "T I I "T I i n I I ' I I Percent 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

HGURE 27: AMERICAN ART SONG COMPOSERS Percentage of Texts from Authors Bom 16-100 Years Earlier than Composer

As this Figure indicates, Argento used a very high percentage of texts from “older generation” authors, the highest for the twelve composers. Ives and Barber were about equal in their choices from this group, both low-to-mid-range percentages. Rorem’s interests in contemporary and ancient poetry minimize his choices in this group. 211

Figure 28 gives the percentage of texts each American selected from contemporary authors.

Ives 36.36% B aA er 13.^5%

Rorem 41.40% Argento o.OO%

Percent 0 5 10 15 20 25 30 35 40 45

HGURE 28: AMERICAN ART SONG COMPOSERS Percentage of Texts from Authors Bom Within 15 Years (Plus or Minus) of Composer

This percentages in this group vary greatly among the Americans, from no choices of contemporary poetry for Argento, to the much higher percentage for Rorem and Ives.

BaAer was more interested in the works of older generation and anonymous writers than in contemporary poetry, as seen in his low percentage in this group.

The small percentages of texts Ae Americans selected from authors bom more than 15

years after the composer are seen in Figure 29.

Ives g 0.57% BaAer 1.25% Rorem | o.OO% Argento I 0.00% I I I I I I Percent 0 5 10 15 20 25

FIGURE 29: AMERICAN ART SONG COMPOSERS Percentage of Texts from AuAots Bom More than 15 Years After Comooser 212

Ives and Barber each used only one text from a younger generation writer. Rorem and

Argento have used none so far, but they are both still composing. Perhaps their latter-year efforts will include newer texts from the young writers of the present.

American Art Song Composers’ Choices of Undated and Old Sources

The percentages of texts the American composers chose from anonymous or undated sources are displayed in Figure 30.

Ives 17.61%

Barter 38.75% Rorem 7.52% Argento | o.oo%

I I I I I I I I I Percent 0 5 10 15 20 25 30 35 40

HGURE 30: AMERICAN ART SONG COMPOSERS Percentage of Texts from Undated Sources

Barber’s large percentage of anonymous or undated texts is the prominent feature of this Figure. Ives and Rorem have used relatively smaller amounts of anonymous and undated texts. In sharp contrast to Barber’s percentage is the complete lack of anonymous texts among Argento’s choices. Argento’s concern for revealiiig the inner personality of the author has led him to the more personal writings of identified authors. 213

The percentages of texts from authors bom more than 100 years before the composer are given for the Americans in Figure 31.

Ives 10.23% iîîttîtta8.75% Rorem 34.41%

Argento D- 18.75%

Percent 0 10 15 20 25 30 35

FIGURE 31: AMERICAN ART SONG COMPOSERS Percentage of Texts from Authors Bom More Than 100 Years Before Composer

Each of the Americans used at least a small amount of ancient poetry. Rorem’s percentage of choices from these authors is the outstanding feature of Figure 31. Argento also used a relatively high percentage of ancient poetry for his songs. The percentages for

Ives and Barber, while comparatively smaller, were also significant in the variety of nationalities and eras they represent. 214

American Art Song Composers’ Choices in Non-native Languages

Figure 32 graphs the Amercan composers’ percentages of texts in non-native

languages.

Ives 21.02% B arber 27.50% Rorem 9.14%

Argento 40.63% 1' I ” 1 I I I I I I I Percent 0 5 10 15 20 25 30 35 40 45

HGURE 32: AMERICAN ART SONG COMPOSERS Percentage of Texts in Non-Native Languages

Argento’s extensive selection of texts by the Swedish explorers place his percentage ahead of the other Americans. Each of the other three American composers set a number of

texts in their original languages, while Argento’s were exclusively in an English

translation. Ives worked primarily with non-native language texts in German and French,

Barber set Celtic verses in translation and French poems in their original language, and

Rorem set verses in French, Italian, and Greek.

Both the European influence and international outlook of modem American composers are reflected in the lives and works of the four composers in this document. Ives was especially influenced by nineteenth century European culture, as can be seen in his settings of German Lied texts. Barber, Rorem, and Argento have all spent significant amounts of time in Europe and learned to speak foreign languages. Their understanding of and interest in the literature of other cultures are evidenced by their choices of texts for their songs. 215

American Art Song Composers’ Personal Contact with the Authors

The ^proximate number of authors with whom the American composers had personal contact are presented in Figure 33.

Ives Baiber g g j Rorem 14 Argento I o I T I I I I I I I 0 2 4 6 4 8 10 12 14

nCURE 33: AMERICAN ART SONG COMPOSERS Approximate Number of Authors with whom the Composer had Personal Contact

As with many of the previous categories, liiere is a diversity in these numbers for the

American composers. Rorem has made contact with many of his authors; a high number of them are personal friends. Ives and Barber knew only a few of their authors, and Argento has been in personal conctact with none of his writers. FINAL SUMMARY: THE TWELVE COMPOSERS

The Twelve Composers’ Educational and Literary Background

The twelve composers in this survey came from diverse family situations. Brahms and

Debussy were from poor homes, while Schumann and Poulenc were from well-to-do homes. The others were essentially from middle-class families.

All but one of the twelve composers had at least a small amount of formal education, and several of them completed college degrees. Debussy received no schooling as a child except from his mother and his later musical training at the Paris Conservatory. Duparc earned a law degree, all of the Americans finished college, Rorem earned a Masters degree, and Argento earned a doctorate. Brahms, Barber and Rorem received honorary doctorates later in life.

Many of these composers remained in academia by becoming teachers. Schubert worked for a brief period as a school teacher. Fauré was a teacher and administrator, and eventually became the head of the Paris Conservatory. Barber taught at the college level for a few years, Rorem has been on several college faculties, and Argento has held the same teaching position at the University of Minnesota for over thirty years.

The literary backgrounds of these twelve composers show a common love of literature through a diversity of manifestations. Schubert, Schumann, Debussy, and Poulenc were directly involved with the literary circles of their times, and knew many writers as friends.

Rorem continues in a similar association with American poets. Brahms loved book collecting, and acquired a large library through the years. Duparc was quite involved with the cultural people and issues of his time. Barber read a great number of books in several

216 217 languages, and studied the lives of important literary figures. Argento’s love of reading, developed in his youth, continues to the present day.

Many of the composers studied were themselves writers. This is perhaps their most important association with literature, in that those skills would increase their appreciation and understanding of other writers’ works. Schumann, Wolf, Fauré, and Debussy had side careers as music critics. Poulenc and Ives both wrote articles and books, and Rorem has particularly distinguished himself as an author of diaries and commentaries on musical subjects. Many of them also wrote poetry, including Schubert, Schumaim, Debussy, Ives, and Barber. The verses Debussy and Ives wrote for their own songs are especially important in their oeuvres.

The Twelve Composers’ Poetry Choices

The process of selecting and using texts for songs was found to be an individual matter for each of the composers. Often they would abandon the works of some authors chosen in the earliest stages of their development, in favor of other writers who better suited the more mature expressive needs and musical styles of the adult composer. Sometimes a composer would return to the works of a particularly favored author after a period of choosing from other writers, producing a large body of important songs from that author’s texts. Other composers constantly searched for novel sources of song texts.

The actual lists of sources for the twelve composers (Tables 13-24) reveal the great richness and diversity in the textual choices. Very few of the authors are found among the sources of more than one composer. The sources include most of the great authors of lyric verse from several nationalities and eras, in addition to a large group of more obscure writers. The latter group often included friends or acquaintances of the composers. All of the composers have revealed much about their personalities through these textual selections. 218

The Twelve Composers’ Number of Songs and Sources

The total number of songs counted for each of the composers is given in Figure 34.

Schubert 624 Schumann 257 Brahms 701 Wolf 313 Duparc | 15

Fauré 102

Debussy 77

Poulenc 144 Ives 176

Barber 30

Rorem 186 Argento | 32 I I I I I I I I 0 100 200 300 400 500 600 700

FIGURE 34; THE TWELVE COMPOSERS Number of Songs

Schubert was the most prolific composer of songs in this study by a large margin. In general, the Lied composers produced the highest numbers of works, followed by the

American and mélodie composers. Perhaps the simple structure and use of the strophic form in the Lied enabled the German composers to create higher numbers of works. The mélodie and contemporary art song genres showed more complex formal designs and more careful attention to prosody, which may have hindered the completion of higher numbers of

French and American works. 219

The number of sources from which the twelve composers selected their song texts is found in Figure 35.

Schubert ...... Schumann V////////A^^ Brahms Wolf Duparc Faure Debussy Poulenc

Barber 40

64 Argento g jq

0 10 20 30 40 50 60 70 80 90 100

HGURE 35: THE TWELVE COMPOSERS Number of Sources

Tlie numbers in Figure 35 range from 10 for Duparc and Argento to 97 for Schubert.

This group of numbers tells little about the comparative amounts of texts chosen by the composers. For this purpose, the numbers of songs in Figure 34 are divided by the 220 numbers of sources in Figures 35 to calculate an arithmetic average number of songs per source for the composers. The results are found in Figure 36.

Schubert Schumann Bratuns

11.59 Duparc Faure Debussy Poulenc

Barber Rorem Argento I" I I "T‘ I 1 I I T T I I 0 1 2 3 4 5 6 7 9 10 11 12

FIGURE 36; THE TWELVE COMPOSERS Arithmetic Average Number of Songs Per Source

This is a more accurate picture of how the choices of the twelve composers compare.

Wolfs average reflects the high numbers of texts he set by a few authors. The multiple choices from smaller numbers of authors in the song cycles and collections of Schubert,

Poulenc, and Schumann are part of their higher averages. Brahms, Fauré, Debussy,

Argento, and Rorem used some sources for high numbers of texts, while choosing fewer

texts from other sources. Ives, Barber, and Duparc more often used only a few texts from

each of their sources. 221

The second average computed was the median average, which is the number of texts taken from the middle source in the complete list, arranged by most-frequent choices to least-frequent choices (i.e., from the tables in the Appendix). The median averages for the twelve composers are presented in Figure 37.

Schubert Schumann Brahms

2.50 Duparc Faure Debussy Poulenc

Ives K fisa 1 ( Barber }$$$${1.00

Rorem 1.00 Argento | 3.00

FIGURE 37: THE TWELVE COMPOSERS Median Average of Songs Per Source

There is a smaller range in these averages, from 1.00 to 3.00. Argento’s small list of sources, combined with his multiple choices from most of his authors have given him the highest median average. His average is followed closely by Wolf, who also took higher numbers of texts from many of his sources. The great amount of experimentation, with 222 one or two choices from at least half of their sources, are seen in the median averages for the rest of the composers.

The modal value, or most commonly occurring number, in the list of sources for every composer is set at one, and is therefore not presented in any Figure in this summary. All of the composers selected more single texts from their sources than any other amount. For some composers this was a relatively strong value, since they took single texts from the majority of their sources. Relatively smaller groups of single choices by the other composers resulted in weaker modal values, but the fact remains: all twelve song composers were most often compelled to select one text from a source.

The Twelve Composers’ Authors Categorized by Birth Date

Most of the other comparisons in this study refer to the relative dating of the texts, through the use the birth years of the authors and composers. The first such group is the 223 percentage of texts the composer chose from authors bom 16-100 years earlier than the composer, seen in Figure 38.

Schubert 68.38% Schumann V////////////////A 42.80% Brahms 55.72%

49.52%

Duparc 40.00%

12.75% Debussy 61.04% IXXXXXXX X XX Poulenc 34.03%

35.23% Barber ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ♦ ' 37.50% Rorem 16.67%

Argento 81.25% T T~n M i l l I I I Perceni 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85

FIGURE 38: THE TWELVE COMPOSERS Percentage ofTexts from Authors Bom 16-100 Years Earlier than Composer

As may be seen, there is a wide variance in the percentages for this group. Some composers made extensive use of texts from the “older generation” authors. Argento,

Schubert, Debussy, and Brahms selected over half of their texts from this group of writers.

Most of the others showed at least mild interest in texts from these authors, although Fauré and Rorem showed significant departures from the norm for this group.

It is expected that the poetry of “older generation” writers would be a strong part of the choices. This poetry, for any given time frame, is traditional and more established, while at the same time enjoys its highest level of popularity. It is also more readily available to a 224 composer who wishes to use it for song composition: other texts are somewhat more obscure, including the most ancient and the most modem poetry, and the composer must actively seek out these types of verses.

The percentages of texts taken from authors bom within 15 years of the composer (plus or minus), shown in Figure 39, are as variable as the percentages in the previous category.

Schubert « 33 65%

Schumann 45.14%

Brahms ; 14.43% Wolf 0.32% Duparc 60.00%

Fauré L 54.90%

Debussy 25.97%

Poulenc 51.39%

36.36%

Bather ]$#$$${13.75%

Rorem 41.40% Argento o.OO% I I I I I I I I I I I I I Percent 0 5 10 15 20 25 30 35 40 45 50 55 60 65

HGURE 39: THE TWELVE COMPOSERS Percentage ofTexts from Authors Bom Within 15 Years (Plus or Minus) of Composer

Five composers were particularly interested in using poetry from contemporary authors:

Duparc, Fauré, Poulenc, Schumaim, and Rorem. It is important to note that four of the five (all but Duparc) were in personal contact with a high number of their authors, suggesting a conn between contact with poets and the use of contemporary poetry. 225

Ives and Debussy also used moderate amounts of contemporary verse, Brahms and

Barber chose small amounts of texts in this group, Wolf chose one poem from a contemporary author, and Argento used none. The individual character of the textual selections of the twelve composers is most vividly demonstrated by the variation of percentages in this Figure and the one before it.

Figure 40 shows that five of the twelve composers selected one or more texts from authors bom more than 15 years after the composer.

Schubert 0.00%

Schumann ^ 3.89%

Brahms 4.48%

Wolf 0.00%

Duparc 0.00%

Fauré 24.51%

Debussy 0.00%

Poulenc 0.00% Ives 1 0.57% T> 1 g3 1.ZJ70

Rorem 0.00%

Argento 0.00%

1 1 1 1 1 1 Percent 0 5 10 15 20 25

HGURE 40: THE TWELVE COMPOSERS Percentage ofTexts from Authors Bom More than 15 Years After Composer

Fauré was the only composer who chose a significant number of texts from younger

generation authors. Brahms and Schumann used small amounts of texts in this group, 226

while Ives and Barber each set only one verse by a younger author. Factors such as the life

span and time period for composition of each composer, as well as their individual tastes

for new poetry, have been suggested as explanations for the distribution of percentages

here.

The Twelve Composers’ Choices of Undated and Old Sources

The amount of anonymous and undated poetry used by the twelve composers is shown

in the percentages in Figure 41.

Schubert I I 3.37%

Schumann ^ 5.84% Brahms 19.40%

Wolf 26.84%

Duparc 0 .0 0 %

Fauré * » : 7.84% Debussy 2.60%

Poulenc 8.33 % Ives sssssssa 17.61% Barber 38.75% Rorem 7.52%

Argento 0 .0 0 % n" nr r i i i i Percent 0 5 10 15 20 25 30 35 40

FIGURE 41 : THE TWELVE COMPOSERS Percentage ofTexts from Undated Sources

Barber’s early Lse of texts by anonymous and obscure writers, combined with his later use of anonymous Irish texts, gave him the highest percentage in this category. Wolf and 227

Brahms also used large amounts of anonymous and folk verses for their Lieder. The others chose smaller amounts of undated texts, or in the case of Duparc and Argento, none at all.

Percentages for texts from authors bom more than 100 years before the composer are displayed in Figure 42.

Schubert Schumaim Brahms 23.32%

0.00 %

Faure 0.00 %

Debussy 10.39% Poulenc 6.25%

10.23% Barber Rorem 34.41%

Argento 18.75%

Percent

FIGURE 42: THE TWELVE COMPOSERS Percentage ofTexts from Authors Bom More Than 100 Years Before Composer

Extensive use of ancient poetry in the songs of Rorem and Wolf is the prominent feature of this category. It should be noted that, for both of these composers, one significant author was included in this category who was bom slightly over a century before the composer. This was the case of Goethe for Wolf and WTatmari for Rorem. 228

A chronological pattern in the choices from ancient poetry is found for each of the three

song genres. The Lied composers chose texts from these sources in increasing amounts,

from Schubert’s small percentage to Wolfs later high percentage. The percentages for

both the French and American composers are divided in half. The first two composers in

each group chose significantly fewer amounts of ancient texts than the other two.

The Twelve Composers’ Choices in Non-native Languages

The use of texts in non-native languages figured in the choices of all but one of the

twelve composers, as seen in Figure 43.

Schubert 7.21% Schumann V///À 11-28% Brahms 14.92%

Wolf 30.67% Duparc 6.67%

Fauré 0 1.96% Debussy | o.OO% rs/vvy Poulenc 1 ^ 7 . 6 4 %

Ives 21.02%

t t t t t t t t t t t t i 27.50% Rorem 9.14%

Argento ^ 40.63% II I r ~ \" \ I "'I I Percent 0 5 10 15 20 25 30 35 40 45

FIGURE 43: THE TWELVE COMPOSERS Percentage ofTexts in Non-Native Language

Debussy is the exception in this category, having used exclusively French verses.

Argento and Wolf both used significant amounts of non-native texts; in each instance the 229 texts were translated into the composer’s native language. Several of the composers worked with texts from three or more non-native languages, often in translations to the language of the composer. Many set some non-native texts in their original languages, most notably Schubert, Poulenc, Ives, Barber, and Rorem. Ives and Rorem set texts in more than one foreign language; Rorem translated some his choices of non-native texts.

The Twelve Composers’ Personal Contact with the Poets

Figure 44 shows the composers’ numbers of personal contacts with their authors.

lj!| Schubert 19

Brahms 7

Duparc Faure Debussy Poulenc 13 BSSSH 6 Barber r tt

Rorem 14 Argento I q

0 10 15 20 25

FIGURE 44; THE TWELVE COMPOSERS Approximate Number of Authors With Whom the Composer Had Personal Contact 230

The literary circles which Schubert and Schumann frequented proved to be rich resources in finding the literary substance for their songs. In more recent times, Poulenc and Rorem have known many of the authors from whom they selected verses for songs.

Poulenc was proud of his associations with the French Surrealist authors, and Rorem has written much about his dealings with poets.

For the rest of the composers in this study, personal contact with the authors is less signficant to the selection and use of the poetry. In many cases the composer did not know those poets with whom they produced their finest works, such as Schubert with Goethe and Wolf with Morike. For these composers, the authors’ poems spoke for themselves. APPENDIX: TABLES 13-24

231 232

Table 13. Textual Sources for the Songs of Franz Schubert (1797-1828)

Year of Year of Texts No. Name of Author Birth Death Used

1 . Goethe, Johann Wolfgang von 1749 1832 72 * 2 . Mayrhofer, Johann 1787 1836 46 3. Müller, Wilhelm 1794 1827 44 4. Schiller, Friedrich von 1759 1805 44 5. Matthisson, Friedrich 1761 1831 29 6 . Holty, Ludwig 1748 1776 23 7. Kosegarten, Ludwig 1758 1818 21 8 . Anonymous and Unknown Authors ND ND 19 9. Schlegel, Friedrich von 1772 1829 16 * 1 0 . Komer, Theodor 1792 1813 14 1 1 . Claudius, Mattias 1740 1815 13 1 2 . Klopstock, Friedrich Gottlieb 1724 1803 13 13. Metastasio, Pietro ( 1 tr. H. J. Collin) 1698 1782 13 14. Salis-Seewis, Johann Gaudenz von 1762 1834 13 *15. Schober, Ritter Franz von 1796 1882 12 16. Leitner, Karl Gottfried von 1800 1890 11 17. Seidl, Johann Gabriel 1804 1874 11 18. Ossian (James Macpherson, tr. Harold) 1736 1796 10 19. Schulze, Ernst 1789 1817 10 2 0 . Rellstab, Ludwig 1799 1860 9 2 1 . Stolberg-Stolberg, Friedrich Leopold zu 1750 1819 9 2 2 . Scott, Walter (various translators) 1771 1832 8 23. Jacobi, Johann Georg 1740 1814 7 *24. Schlechta, Franz Xaver Freiherr von 1796 1875 7 25. Schlegel, August Wilhelm von 1767 1845 7 26. Baumberg, Gabriele von 1796 1839 6 27. Heine, Heinrich 1797 1856 6 tr. = translated by ND = No Date Available * = Personal Contact with Composer 233

Table 13 (continued) Year of Year of Texts No. Name of Author ______Birth Death Used

28. Novalis, Friedrich L. von Hardendenberg 1772 1801 6 29. Riickert, Friedrich 1788 1866 6 *30. Biuchmann, Franz Ritter von 1798 1867 5 31. Motte-Fouqué, Friedrich de la 1777 1843 5 32. Uz, Johann Peter 1720 1796 5 *33. Collin, Mattaus von 1779 1824 4 34. Schreiber, Alois 1763 1841 4 35. Schubart, Christian Daniel 1739 1791 4 36. Fellinger, Johann Georg 1781 1816 3 *37. Kenner, Josef 1794 1868 3 38. Petrarca, Francesco (tr. Cries) 1304 1374 3 *39. Pichler, Karoline 1769 1843 3 *40. Rochlitz, Friedrich 1769 1842 3 41. Shakespeare, William (various translators) 1564 1616 3 42. Stoll, Josef Ludwig 1778 1815 3 43. Bertrand, Friedrich Anton Franz 1757 1828 2 44. Castelli, Ignaz Franz von 1781 1862 2 45. Craigher, Jakob Nicolaus 1797 1855 2 46. Ermin (Johann Gottfried Kumpf) 1781 1862 2 47. Lappe, Carl 1773 1843 2 *48. Platen-Hallermiinde, August von 1796 1835 2 *49. Pyrker, Ladislaus von Oberwart 1772 1847 2 50. Schütz, Christian Wilhelm von 1776 1847 2 *51. Seim, Johann Chrysostomus 1795 1857 2 52. Silbert, Johann Peter 1772 1844 2 *53. Stadler, Albert 1794 1888 2 54. Széchényi, Louis, Count von 1781 1855 2 55. Vitorelli, Jacopo Andrea 1749 1835 2 56. Werner, Friedrich Ludwig Zacharias 1768 1823 2 tr, = translated by ND = No Date Available * = Personal Contact with Composer 234

Table 13 (continued) Year of Year of No. Name of Aiithor Birth Death

57. Willemer, Marrianne von 1784 1860 58. Aeschylus (tr. J. Mayrhofer) 5th c. BC 5th c. BC *59. Bauemfeld, Eduard von 1802 1890 60. Bernard; Josef Karl 1780 1850 61. Cibber, Colley (tr. J. N. Craigher) 1671 1757 62. Collin, Heinrich Josef Elder von 1722 1811 63. Cowley, Abraham (tr. J. F. Ratschky) 1618 1667 64. Deinhardstein, Johann Ludwig 1794 1859 65. Ehrlich, Bernhard Ambros ND ND 6 6 . Gerstenberg, Friedrich von 1780 1838 67. Goldoni, Carlo 1707 1793 6 8 . Getter, Friedrich Wilhelm 1746 1797 69. Grillparzer, Franz 1791 1872 70. Hell, Theodor (Karl Winkler) 1775 1856 *71. Hiittenbrenner, Heinrich 1778 1830 72. Kalchberg, Johann von 1765 1827 73. Kind, Friedrich 1768 1843 74. Klenke, Karoline Luise von 1754 1802 75. Kopken, Friedrich von 1737 1811 76. Kuffner, Christoph 1780 1846 77. Leon, Gottlieb von 1757 1832 78. Lubi, Michael 1757 1808 79. Majlath, Count Johann 1768 1855 80. Mikan, Johann Christian 1769 1844 *81. Ottenwalt, Anton 1789 1845 82. Pfeffel, Gottfried Konrad 1736 1809 83. Plainer, Eduard (?Anton) 1787 1855 84. Poliak, A. ND ND 85. Pope, Alexander (tr. Herder) 1688 1744 tr. = translated by ND = No Date Available * = Personal Contact with Composer 235

Table 13 (continued) Year of Year of Texts No. Name of Author ______Birth Death Used

8 6 . Prandstetter, Martin Joseph 1760 1798 87. Pratobevera, Adolf von 1806 1875 * 8 8 . Reil, Friedrich 1773 1843 89. Reissig, Christian Ludwig 1783 1822 90. Roos, Richard (Karl August Engelhardt) 1769 1834 91. Sauter, Samuel Friedrich 1766 1846 92. Schmidt von Liibeck, Georg Philipp 1766 1849 *93. Schubert, Franz Peter 1797 1828 94. Schiicking, Clemens August 1759 1790 *95. Spaun, Josef von 1788 1865 96. Tiedge, Christoph August 1752 1841 97. Uhland, Ludwig 1787 1862

tr. = translated by ND = No Date Available * = Personal Contact with Composer 236

Table 14. Textual Sources for the Songs of Robert Schumann (1810-1856)

Year of Year of Texts No. Name of Author Birth Death Used

* 1. Heine, Heinrich 1797 1856 38 * 2 . Riickert, Friedrich 1788 1866 20 3. Goethe, Johann Wolfgang von 1749 1832 19 4. Kemer, Justinus 1786 1862 19 *5. Eichendorff, Joseph von 1788 1857 16 * 6 . Chamisso, Adelbert von 1781 1838 13 *1. ITofTmann von Fallersleben, August 1798 1874 11 * 8 . Lenau, Nikolaus 1802 1850 11 9. Bums, Robert (tr. Gerhard) 1759 1796 9 * 10. Geibel, Emmanuel 1815 1884 9 11. Anonymous (Volkslied, Traditional) ND ND 8 * 12. Neun, Wilfried von der 1826 1916 8 13. Kulmann, Elisabeth 1808 1825 7 *14. Reinick, Robert 1805 1852 6 *15. Andersen, Hans C. (tr. Chamisso et. al.) 1805 1875 5 16. Byron, George Gordon (various translators) 1788 1824 5 17. Morike, Eduard 1804 1875 5 18. Stuart, Maria (tr. Vincke) 1542 1568 5 19. Pfarrius, Gustav 1800 1884 3 2 0 . Schiller, Friedrich von 1759 1805 3 * 2 1 . Schumann, Robert 1810 1856 3 * 2 2 . Buddeus, Julius ND ND 2 23. Heyse, Paul 1830 1914 2 24. Moore, Thomas (tr. Freiligrath) 1797 1852 2 25. Uhland, Ludwig 1787 1862 2 *26. Ulrich, Titus 1813 1891 2 *27. Bernhard, Lily ND ND 1 tr. = translated by ND = No Date Available * = Personal Contact with Composer 237

Table 14 (continued) Year of Year of Texts No. Name of Aiithor ______Birth____ Death Used

*28. Braun, Ferdinand ND ND *29. Candidas, Carl 1817 1872 *30. Christem, Carl ND ND 31. Fanshawe, Catherine (tr. Kannegiesser) 1765 1834 32. Fenand, Eduard (Eduard Schulz) 1813 1842 33. Frohlich, Abraham 1796 1865 34. Halm, Friedrich (Eligius Freiherr) 1806 1871 *35. Hebbel, Friedrich 1813 1863 36. Immermann, Karl 1796 1840 37. Jacobi, Johann Georg 1740 1814 38. Kinkel, Johann Gottfried 1815 1882 39. Kletke, Hemiann 1813 1886 *40. L’Egru, Christian ND ND *41. Lorenz, Wilhelmine ND ND *42. Mosen, Julius 1803 1867 *43. Müller von Konigswinter, Wolfgang 1816 1873 44. Platen, August von 1796 1835 45. Schulze, Ernst 1789 1817 46. Seidl, Johann Gabriel 1804 1875 47. Shakespeare, William (tr. Tieck & Schlegel) 1564 1616 48. Strachwitz, Graf Moritz von 1822 1847 *49. Wolff, Oskar Ludwig 1799 1851 50. Zimmermann, Georg Wilhelm 1794 1835

tr. = translated by ND = No uate Available * = Personal Contact with Composer 238

Table 15. Textual Sources for the Songs of Johannes Brahms (1833-1897)

Year of Year of Texts No. Name o f Author Birth Death Used

1. Anonymous (Volkslied, Traditional) ND ND 39 * 2 . Daumer Georg Friedrich 1800 1875 19 3. Tieck, Ludwig 1773 1853 15 *4. Groth, Klaus 1818 1899 11 5. Lemke, Karl 1831 1913 7 6 . Candidas, Carl 1817 1872 6 7. Heine, Heinrich 1797 1856 6 8 . Holty, Ludwig 1748 1776 6 9. Halm, Friedrich (E. F. J. von Miinch) 1806 1871 5 1 0 . Eichendorff, Joseph von 1788 1857 5 11. Goethe, Johann Wolfgang von 1749 1832 5 12. Hoffmann von Fallersleben, August 1798 1874 5 13. Platen, August von 1796 1835 5 14. Schenkendorf, Max von 1783 1817 5 15. Shakespeare, William (tr. Tieck & Schlegel) 1564 1616 5 16. Reinhold, Christian (C. R. Kostlin) 1813 1856 4 17. Biblical (tr. Martin Luther) ND ND 4 *18. Schmidt, Hans 1854 1923 4 19. Uhland, Ludwig 1787 1862 4 * 2 0 . Keller, Gottfried 1819 1890 3 2 1 . Schack, Adolf Friedrich von 1815 1894 3 * 2 2 . Schumann, Felix 1854 1879 3 23. Flemming, Paul 1609 1640 2 24. Geibel, Emmanuel 1815 1884 2 25. Hebbel, Friedrich 1813 1863 2 26. Lilliencron, Detlev von 1844 1909 2 27. Morike, Eduard 1804 1875 2 tr. — tionslatcu by ND = No Date Available * = Personal Contact with Composer 239

Table 15 (continued) Year of Year of Texts No. Name of Author Birth Death Used

28. Reinick, Robert 1805 1852 2 29. Simrock, Carl 1802 1876 2 30. Alexis, Willibald (G. W. Haring) 1798 1871 31. Ferrand, Eduard (Edward Schulz) 1813 1842 32. Allmers, Hermann 1821 1902 33. Bodenstedt, Friedrich 1819 1892 34. Bretano, Clemens von 1778 1842 35. Frey, Adolf 1855 1920 36. Grohe, Melchior 1829 ND 37. Gruppe, Otto F. 1804 1876 *38. Heyse, Paul 1830 1914 *39. Kalbeck, Max 1850 1921 40. Kopisch, August 1799 1853 41. Kugler, Franz 1808 1858 42. Lingg, Hermann 1820 1905 43. Meissner, Alfred 1822 1885 44. Rousseau, Johann B. 1802 1867 45. Riickert, Friedrich 1788 1866 46. Storm, Theodor 1817 1888 47. Thibault (tr. Herder) 13 th c. 13th c,

tr. = translated by ND = No Date Available * = Personal Contact with Composer 240

Table 16. Textual Sources for the Songs of Hugo Wolf (1860-1903)

Year of Year of Texts No. Name of Author Birth Death U sed

1. Goethe, Johann Wolfgang von 1749 1832 59 2 . Morike, Eduard 1804 1875 57 3. Italienisches Liederbuch (tr. P. Heyse) ND ND 46 4. Spanisches Liederbuch (tr. Heyse & Geibel) ND ND 44 5. Eichendorff, Joseph von 1788 1857 26 6 . Heine, Heinrich 1797 1856 18 7. Reinick, Robert 1805 1852 13 8 . Lenau, Nikolaus 1802 1850 12 9. Keller, Gottfried 1819 1890 6 10. Anonymous or Unknown ND ND 4 11. Hoffmann von Fallersleben, August 1798 1874 4 12. Hebbel, Friedrich 1813 1863 3 13. Michelangelo (tr. Robert-Tomow) 1475 1564 3 14. Byron, George Gordon (tr. Gildemeister) 1788 1824 2 15. Chamisso, Adelbert von 1781 1838 2 16. Riickert, Friedrich 1788 1866 2 17. Scheffel, Joseph Viktor von 1826 1886 2 18. Herlossohn, Georg Karl Reginald 1804 1849 1 19. Kemer, Justinus 1786 1862 1 2 0 . Komer, Theodor 1792 1813 1 2 1 . Matthisson, Friedrich von 1761 1831 1 * 2 2 . Peitl, Paul 1853 1922 1 23. Roquette, Otto 1824 1896 1

24. Shakespeare, William (tr. Schlegel) 1564 1616 1 25. Sturm, Julius 1816 1896 1 26. Zshokke, Heinrich Daniel 1771 1848 1 27. Zusner, Vincenz 1803 1874 1 tr. = translated by ND = No Date Available * = Personal Contact with Composer 241

Table 17. Textual Sources for the Songs of Henri Duparc (1848-1933)

Year of Year of Texts No. Name of Author ______Birth Death Used

1. Labor, Jean (Dr. Henri Cazalis) 1840 1909 3 2 Gautier, Théophile 1811 1872 2 3. Sully-Prudhomme, Armand 1839 1907 2 4. Baudelaire, Charles 1821 1867 2 *5. de Bonnières, Robert 1850 1905 1 6 . Leconte de Lisle, Charles Marie René 1818 1894 1 7. Marc, Gabriel 1840 ND 1 8 . Moore, Thomas (tr. unknown) 1797 1852 1 *9. Silvestre, Armand 1837 1870 1

10. Wilder, Victor (after Goethe) 1835 1892 1

tr. = translated by ND = No Date Available * = Personal Contact with Composer 242

Table 18. Textual Sources for the Songs of Gabriel Fauré (1845-1924)

Year of Year of Texts No. Name of Author ______Birth Death Used

1. Van Leberghe, Charles 1861 1901 18 * 2 . Verlaine, Paul 1844 1896 17 3. Silvestre, Armand 1838 1901 11 4. Hugo, Victor 1802 1885 9 5. Leconte de Lisle, Charles Marie René 1818 1894 5 6 . Brimont, Renée (La Baronne) ND ND 4 7. Gautier, Théophile 1811 1872 4 8 . de Mirmont, Jean de la Ville 1886 1914 4 9. Baudelaire, Charles 1821 1867 3 10. Grandmougin, Charles 1850 1930 3 * 11. Samain, Albert 1858 1900 3 * 12. Sully-Prudhomme, Armand 1839 1907 3 13. Anonymous Italian (tr. R. Bussine) ND ND 2 14. de LTsle-Adam, Villiers 1838 1889 2 15. Mendès, Catulle 1842 1909 2 16. Richepin, Jean 1849 1926 2 17. Bordèse, Stéphan 1847 ND 1 18. de Choudens, Paul 1850 1929 1 *19. Debladis, Georgette ND ND 1 2 0 . Dominique, Jean (Marie Closset) 1873 1952 1 2 1 . Fort, Paul 1872 1960 1

2 2 . Monnier, Marc 1827 1855 1 23. Moréas, Jean 1856 1910 1 24. Pommey, Louis ND ND 1 25. de Régnier. Henri 1864 1936 1 26. Wilder, Victor 1835 1892 1

tr. = translated by ND = No Date Available * = Personal Contact with Composer 243

Table 19. Textual Sources for the Songs of Claude Debussy (1862-1918)

Year of Year of Texts No. Name of Author ______Birth Death Used

1. Verlaine, Paul 1844 1896 21 2 . Banville, Théodore, de 1823 1891 10 *3. Bourget, Paul 1852 1935 9 4. Baudelaire, Charles 1821 1867 5 *5. Debussy, Claude 1862 1918 5 6 . Lhermite, Tristan 1601 1655 3 *7. Mallarmé, Stéphane 1842 1898 4 * 8 . Louÿs, Pierre 1870 1935 3 9. Villon, François 1431 ND 3 10. d’Orléans, Charles 1394 1465 2 11. Gautier, Théophile 1811 1872 2 12. Leconte de Lisle, Charles Marie René 1818 1894 2 13. Girod, André ND ND 1 14. Gravollet, Paul (Jeulin) 1863 ND 1 *15. Hyspa, E. Vincent 1865 1938 1 *16. Le Roy, Grégoire 1862 1941 1 17. de Musset, Alfred 1810 1857 1 18. Renaud, Armand 1836 1895 1 19. Unknown Author ND ND 1 2 0 . Valade, Léon 1841 1894 1

li. — iionslatcu by ND = No uate Available * = Personal Contact with Composer 244

Table 20. Textual Sources for the Songs of Francis Poulenc (1899-1963)

Year of Year of Texts No. Name of Author ______Birth Death Used

* 1. Eluard, Paul 1895 1952 34 * 2 . Apollinaire, Guillaume 1880 1918 33 *3. Vilmorin, Louise de 1902 1969 13 4. Anonymous, seventeenth century ND ND 8 *5. Carême, Maurice 1899 1978 7 * 6 . Jacob, Max 1876 1944 7 *7. Fombeure, Maurice 1906 1981 6 8 . Ronsard, Pierre de 1524 1585 6 9. Anonymous Polish, ca. 1831 ND ND 4 10. Moréas, Jean 1856 1910 4 * 11. Cocteau, Jean 1889 1963 3 * 12. Lorca, Federico Garcia (tr. F. Gattegno) 1897 1982 2 *13. Aragon, Louis 1897 1982 2 *14. Desnos, Robert 1900 1945 2 *15. Laurencin, Marie 1885 1956 2 16. Beylié, Laurence de 1893 1968 1 17. d’Orléans, Charles 1394 1465 1 *18. Colette, Sidonie-Gabrielle 1873 1954 1 19. Gaslawski, Maurice 1802 1834 I 2 0 . Kowalski, François 1799 1862 I 2 1 . de Malherbe, François 1555 1628 1

2 2 . Racine, Jean 1639 1699 1 *23. Radiguet, Raymond 1903 1923 1 24. Suchodolski, Raynold 1803 1831 1

25. v/itwicki, Stefan 1802 1847 1

tr. = translated by ND = No Date Available * = Personal Contact with Composer 245

Table 21. Textual Sources for the Songs of Charles Ives (1874-1954)

Year of Year of Texts No. Name of Author ______Birth Death Used

* 1. Ives, Charles Edward 1874 1954 40 2 . Anonymous or Unknown Authors ND ND 22 3. Heine, Heinrich (3 translated) 1797 1856 8 *4. Ives, Mrs. Harmony Twichell 1875 1969 6 5. Kipling, Rudyard 1863 1935 5 6 . Johnson, Robert Underwood 1853 1937 4 7. Moore, Thomas 1779 1852 4 8 . Folgore da San Geminiano (tr. D. G. Rossetti) ND ND 3 9. Vinje, Aasmund (tr. Lobedanz, Corder) 1818 1870 3 * 10. Bcllamann, Henry 1882 1945 2 11. Byron, George Gordon 1788 1824 2 12. Holmes, Oliver Wendell 1809 1894 2 13. Lyte, Henry Francis 1793 1847 2 14. Shelley, Percy Bysshe 1792 1822 2 15. Turnbull, Monica Peveril 1879 1901 2 16. Whittier, John Greenleaf 1807 1892 2 17. Wordsworth, William 1770 1850 2 18. Aldrich, Thomas Bailey 1836 1907 1 19. Allmers, Hermann (tr. H. G. Chapman) 1821 1902 1 2 0 . Ariosto, Ludovico 1474 1533 1 2 1 . Arnold, Matthew 1822 1888 1 2 2 . Bernard of Cluny (tr. J. M. Neale) 1100 1156 1 23. Bowring, John 1792 1872 1 *24. Brewster, Judge Lyman D. 1832 1904 1 25. Brooke, Rupert Chawner 1887 1915 1 26. Browning, Robert 1812 1889 1 27. Bulwer-Lytton, Edward G. E. 1803 1873 1 tr. = translated by ND = No Date Available * = Personal Contact with Composer 246

Table 21 (continued) Year of Year of No. Name of Aiithor Birth Death

28. Case, Lizzie York ND 1911 29. Chauvenet, W. M. ND ND 30. Chézy, Wilhelmine von ND ND 31. Coleridge, Hartley 1796 1849 *32. Collins, Aime Timoney ND ND 33. Cooper, James Fenimore, Jr. 1892 1918 34. Cornelius, Peter 1824 1874 35. Cowper, William 1731 1800 36. Durand, Henry Strong 1861 1929 37. Emerson, Ralph Waldo 1803 1883 38. Florian Jean P. C. de 1755 1794 39. Gallet, Louis 1835 1898 40. Goethe, Johaim Wolfgang von 1749 1832 41. Gottschall, Rudolf (tr. Rucker and Ives) 1828 ND 42. Groth, Klaus 1819 1899 43. Heyduk, Adolf (tr. N. Macfarren) 1835 ND 44. Hill, Susan Benedict ND ND 45. Holty, Ludwig 1748 1776 46. Hunt, Leigh 1784 1859 47. Ives, Miss A. L. ND ND 48. Keats, John 1795 1821 49. Kingsley, George Henry 1811 1884 50. Landor, Walter Savage 1775 1864 51. Lenau, Nikolaus 1802 1850 52. Lindsay, Nicolas Vachel 1879 1931 53. Longfellow, Henry Wadsworth 1807 1882 54. Lowry, Robert 1826 1899 55. Manilius (or Manlius) 1st c. BC 1st c. BC 56. Markham, Edwin 1852 1940 tr. = iransiatea oy ND = No Date Available * = Personal Contact with Composer 247

Table 21 (continued) Year of Year of Texts No. Name of Author Birth Death Used

57. McCrae, John David 1872 1918 58. Meredith, Geoige 1828 1909 59. Merrill, Charles Edmund, Jr. 1877 ND 60. Milton, John 1608 1674 61. Morris, Maurice (Morris W. Pool) ND ND 62. Müller, Wolfgang 1816 1873 63. O’Malley, D. J. ND ND 64. Percy, Thomas 1729 1811 65. Peterson, Frederic 1859 1938 6 6 . Ploug, Parmo Karl (tr. Kappey) 1813 1894 67. Porteous, Baroness (tr. unknown) ND ND 6 8 . Robinson, Rev. Robert 1735 1790 69. Rossetti, Christina Georgina 1830 1894 70. Scollard, Clinton 1860 1932 71. Shakespeare, William 1564 1616 72. Sprague, Charles 1791 1875 73. Stevenson, Robert Louis 1850 1894 74. Stieler, Karl 1842 1885 75. Symons, Arthur William 1865 1945 76. Tate, Nahum (with Nicholas Brady) 1652 1715 77. Tennyson, Alfred Lord 1809 1892 78. Tersteegen, Gerhard (tr. J. Wesley) 1697 1769 *79. Untermeyer, Louis 1885 1978 80. Whitman, Walt 1819 1892 81. Winter, William 1836 1917 82. Winther, Christian (tr. E. Lobedanz) 1796 1876

tr. = translated by ND = No Date Available * = Personal Contact with Composer 248

Table 22. Textual Sources for the Songs of Samuel Barber (1910-1981)

Year of Year of Text No. Name of Author Birth Death Usee

1. Anonymous or Unknown Authors ND ND 15 2 . Irish texts, 8th-13th c. (various translators) ND ND 10 3. Joyce, James 1882 1941 8 4. Rilke, Rainer Maria 1875 1926 5 5. Davies, William Henry 1871 1940 3 6 . Graves, Robert 1895 1985 3 7. MacCleod, F. ND ND 2 8 . Stephens, James 1882 1950 2 *9. Agee, James 1909 1955 1

* 10. Barber, Samuel 1910 1981 1

11. Bhartrihari (tr. P. E. More) 5th c. 6th c. 1 12. Binyon, Laurence 1869 1943 1 13. Browning, Elizabeth Barret 1806 1861 1 14. Cornwell, D. ND ND 1 15. Dillon, George 1906 1968

16. Frost; Robert 1874 1963 1

17. Harasymowicz, Jerzy (tr. C. Milosz) 1933 -- 1 18. Heym, George (tr. C. Middleton) 1887 1912 1 19. Hopkins, Gerald Manely 1844 1889 1 * 2 0 . Horan, Robert 1922 1981 1 2 1 . Housman, Alfred Edward 1859 1936 1 2 2 . Hughes, Langston 1902 1967 1 23. Keller, Gottfried (tr. J. Joyce) 1819 1890 1 24. Kerlin, Robert Thomas 1866 1950 1

25. Kingsley, Charles 1819 1875 1

26. Landor, Walter Savage 1775 1864 1 27. Meleager of Gadara (tr. J. A. Symonds) 140 BC C.70 BC 1 tr. = translated by Nu = No Date Available * = Personal Contact with Composer 249

Table 22 (continued) Year of Year of Texts No. Name of Author Birth Death Used

28. Meurath, A. ND ND 29. Parsons, K. ND ND 30. Prokosch, Frederic 1908 — 31. Rittenhouse, Jessie Belle 1869 1948 32. Roethke, Theodore 1908 1963 33. Shelley, Percy Bysshe 1792 1822 34. Swinburne, Algernon Charles 1837 1909 35. Terry, Edward Hicks Streeter ND ND 36. Villa José Garcia 1908 37. Whittier, John Greenleaf 1807 1892 38. Wolfe, Humbert 1885 1940 39. Wyatt, Sir Thomas 1503 1542 40. Yeats, William Butler 1865 1939

tr. = translated by ND = No Date Available * = Personal Contact with Composer 250

Table 23. Textual Sources for the Songs of Ned Rorem (b. 1923)

Year of Year of Texts No. Name of Author ______Birth Death Used

*1. Goodman, Paul 1911 1972 21 2. Whitman, Walt 1819 1892 16

*3. Moss, Howard 1922 — 13 4. Roethke, Theodore 1908 1963 13 5. Biblical ND ND 9 6. Herrick, Robert 1591 1674 9 7. Fletcher, John 1579 1625 6

*8. Koch, Kenneth 1925 -- 6

*9. Norse, Harold 1916 -- 6 10. Anonymous ND ND 4 11. Tennyson, Alfred Lord 1809 1892 4 12. Vitorelli, Jacopo Andrea 1749 1835 4 13. Capetanakis, Demetrios 1912 1944 3 14. Dickinson, Emily 1830 1866 3

15. Green, Julien (tr. N. Rorem) 1900 — 3 16. Rossetti, Christina Georgina 1830 1894 3 17. Wylie, Elinor 1885 1928 3 *18. Ashbery, John 1927 — 2 19. Auden, W. H. 1907 1973 2 *20. Bishop, Elizabeth 1911 1979 2 21. Cummings, E. E. (Edward Estlin) 1894 1962 2 22. Hopkins, Gerald Manley 1844 1889 2

*23. Larson, Jack C.1926 — 2 24. de Magny, Olivier 1529 1560 2

*25. Pitchford, Kenneth 1931 — 2 26. Plato C.428 BC C.348 BC 2 27. Sitwell, Dame Edith 1887 1964 2 tr. = translated by ND = No Date Available * = Personal Contact with Composer 251

Table 23 (continued) Year of Year of Tex No. Name of Author Birth Death Use

28. Waller, Edmund 1607 1687 2 29. Williams, William Carlos 1883 1963 2 *30. Windham, Donald 1920 — 2 31. "Alleluia" ND ND 1 32. Bamefîeld, Richard 1574 1627 1 33. Boleyn, Queen Anne 1507 1536 1 34. Bradstreet, Anne 1612 1672 1 35. Browning, Robert 1812 1889 1 36. Cattullus (tr. A. Beardsley) 84 BC 54 AD 1 37. Chaucer 1343 1400 1 38. Coleridge, Mary Elizabeth 1861 1907 1 39. Daurat, Jean 1505 1588 1 40. de Baïf, Jean-Antoine 1532 1589 1 41. Eddy, Mary Baker 1821 1910 1 *42. Garrigue, Jean 1914 1972 1 43. Gibbons, Orlando 1583 1625 1 44. Glaze, Andrew 1920 — 1 45. Herbert, Mary, Countess of Pembroke 1561 1621 1 46. Hillyer, Robert 1895 1961 1 47.Jonson, Ben 1573 1637 1 48. Landor, Walter Savage 1775 1864 1 49. Leigh, Mary, Lady Chudleigh 1656 1710 1 50. Lodge, Thomas 1558 1625 1 51. Mew, Charlotte 1870 1928 1 *52.' de Noailles, Marie Laure 1902 1970 1 *53. O'Hara, Frank 1926 1966 1 54. Pearse, Padhiaic 1879 1916 1 55. Regnier, Jehaim 1392 1470 1

*56. Rich, Adrieime 1929 — — 1

If. = ifansiated by ND = No Date Available * = Personal Contact with Composer 252

Table 23 (continued) Year of Year of Texts No. Name of Author ______Birth Death Used

57. Ridge, Lola 1883 1941

58. Roditi, Edouard 1910 -- 59. de Ronsard, Pierre 1524 1585 60. Shelley, Percy Bysshe 1792 1822 61. Spenser, Edmund 1552 1599 62. Stein, Gertrude 1874 1946 63. Stevenson, Robert Louis 1850 1894 64. Yeats, William Butler 1865 1939

tr. = translated by ND = No Date Available * = Personal Contact with Composer 253

Table 24. Textual Sources for the Songs of Dominick Argento (b. 1927)

Year of Year of Texts No. Name of Author ______Birth Death Used

1. Woolf, Virginia 1882 1941 8 2 . Andrée, Salomon August (tr. E. A. Ray) 1854 1897 5 3. Cummings, E. E. (Edward Estlin) 1894 1962 5 4, Fraenkel, Knut (tr. E. A. Ray) 1870 1897 4 5. Strindberg, Nils (tr. E. A. Ray) 1872 1897 4 6 . Shakespeare, William 1564 1616 2 7. Constable, Henry 1562 1613 1 8 . Daniel, Samuel 1562 1619 1 9. Jonson, Ben 1573 1637 1 10, Nash, Thomas 1567 1601 1

tr. = translated by ND = No Date Available * = Personal Contact with Composer LIST OF REFERENCES

General References

Benton, Rita, with Frits Noske and David Cox. “Mélodie,” The New Grove Dictionary o f Music and Musicians, 20 vois., ed. Stanley Sadie. London: Macmillan, 1980, XII, 112-115.

Chew, Geoffrey. “Song §9 and §10,” The New Grove Dictionary o fMusic and Musicians, 20 vols., ed. Stanley Sadie. London: Macmillan, 1980, XVII, 518-521.

Hall, James Husst. The Art Song. Norman: University of Oklahoma Press, 1953.

Ivey, Donald. Song: Anatomy, Imagery, and Styles. New York: The Free Press, 1970.

Meister, Barbara. An Introduction to the Art Song. New York: Taplinger, 1980.

Sams, Eric. “Lied §4 (The Romantic Lied),” The New Grove Dictionary o f Music and M usicians, 20 vols., ed. Stanley Sadie. London: Macmillan, 1980, X, 838-844.

Seaton, Douglass. The A rt Song: A Research and Information Guide. New York and London: Garland, 1987.

Stevens, Denis, ed. A History o fSong. London: W.W. Norton, 1960.

Webster’s New World Dictionary, Second College Edition, ed. David B. Guralnik. New York: The World Publishing Co., 1970.

Chapter I

Abraham, Gerald, ed. The Music o fSchubert. Port Washington: Kennikat Press, 1947, reissued 1969.

Abraham, Gerald, with Work-List by Eric Sams. “Schumemn, Robert,” The New Grove Dictionary ofM usic and Musicians, 20 vols., ed. Stanley Sadie. London: Macmillan 1980, XVI, 857-863.

Becker, Heinz. “Brahms, Johannes,” The New Grove Dictionary o fMusic and Musicians, 20 vols., ed. Stanley Sadie. London: Macmillan, 1980, III, 174-185.

Bell, A. Craig. The Lieder o fBrahms. Darley, England: The Grian-Aig Press, 1979.

254 255

Boylan, Paul C. “The Lieder of Hugo Wolf: Zenith of the German Art Song,” Ph.D. dissertation, University of Michigan, 1968.

Brown, Maurice J. E., with Work-list by Eric Sams. The New Grove Schubert, originally published in The New Dictionary o fMusic and Musicians, 20 vols., ed. Stanley Sadie. London: Macmillan, 1980, 1982.

Schubert Songs, BBC Music Guides. London: British Broadcasting ^corporation. 1967.

Capell, Richard. Schubert’s Songs, 3rd edition. London: Duckworth, 1928, revised 1957, 1973.

Gamer, Mosco. Hugo W olf Songs, BBC Music Guides. London: British Broadcasting Corp., 1982.

Chissell, Joan. Brahms, The Great Composers Series. London: Faber and Faber, 1977.

Cooper, Martin. “The Songs,” Schumaim: A Symposium, ed. Gerald Abraham. London: Oxford University Press, 1952, 91-137.

Dale, Kathleen. Brahms: A biography with a survey o f books, editions & recording':. Hamden: .\rchon Books, 1970.

Desmond, Astra. Schumaim Songs, BBC Music Guides. Seattle: University of Washington Press, 1972.

Deutsch, Otto Erich. Schubert: Thematic Catalougue o fall his Works. London, 1951.

Deutsches Literatur-Lexikon, 11 vols, to date, ed. Heinz R.i:,)p and Carl Ludwig Lang. Bern: A. Francke AG, 1968-86.

Diaz-Plaja, Guillermo. A History o fSpanish Literature, tr. Hugh A. Harter. New York: New York University Press, 1971.

Diirr, Walther. “Schubert’s songs and their poetry: reflections on poetic aspects of song composition, §1 and §2,” Schubert Studies, ed. Eva Badura-Skoda and Peter Branscombe. Cambridge, England: Cambridge University Press, 1982,1-11.

Fiedler, H. G. “Schubert’s Poets,” Music & Letters WVl, 68-77.

Fischer-Dieskau, Dietrich. Robert Schumann: Words and Music— The Vocal Compositions, tr. Reinhard G. Pauly. Portland: Amadeus Press, 1988.

Schubert: A Biographical Study o f his Songs, irans. Kenneth S. Whitton. London: Cassell, 1976.

Forster, Leonard, ed. The Penguin Book o f German Verse. Harmondsworth, Middlesex, England: Penguin Books Ltd., 1957.

Friedlaender, Max. Brahms’ Lieder. An IntroducUon o f the Songs for Gne or Two Voices, tr. C. Leonard Leese. London: Oxford University Press, 1928. 256

Friedrichs, Elisabeth. Literarische Lokalgrossen 1700-1900. Stuttgart: Metzlersche, 1967.

Geiringer, Karl. Brahms: His Life and Work, tr. H. B. Weiner and Bernhard Miall. Boston and New York: Houghton Mifflin, 1936.

Hallmark, Rufus E. The Genesis o fSchumann's Dichterliebe: A Source Study. Ann Arbor: University Microfilms International Research Press, 1979.

Harrison, Max. The Lieder o fBrahms. New York: Praeger, 1972.

Hilmar, Ernst. Franz Schubert In His Time. Portland: Amadeus Press, 1988.

Holmes, Paul. Brahms: His Life and Times. Kent, England: The Baton Press, 1984.

Kalbeck, Max. Johannes Brahms, 4 vols. Berlin: Deutsche Brahms Gesellschaft, 1908- 1914.

Kosch, Wilhelm. Deutsches Literatur-Lexikon, 4 vols. Bern: A. Francke AG, 1956.

Latham, Peter. Brahms, The Master Musicians Series. New York: Pellegrini and Cudahy, 1949.

Newman, Ernest. H ugo W olf. [1907] New York: Dover, 1966 reprint.

Niecks, Frederick. Robert Schumaim, ed. Christina Niecks, London: J. M. Dent, 1925.

The Oxford Guide to Spanish Literature, ed. Philip Wood. Oxford: Clarendon Press, 1978.

Robert Pascall, ed. Brahms: biographical, documentary and analytical studies. Cambridge: Cambridge University Press, 1983.

Prawer, Siegbert, ed. The Romantic Period in Germany. New York: Schccken Books, 1970.

Pulver, Jeffrey. Johannes Brahms. London: Kegan Paul, Trench, Trubner; J. Curwen & Sons, 1926.

Reed, John. Schubert, The Master Musicians Series. London: J. M. Dent, 1987.

The Schubert Song Companion. Manchester, England: Manchester University Press, 1985.

Sams, Eric. Brahms Songs, BBC Music Guides. Seattle: University of Washington Press, 1972.

. The Songs o fHugo Wolf. New York: Oxford University Press, 1962.

. The Songs o f Robert Schumann, 2nded. London: Eulenburg, 1975. 257

“The Songs,” Robert Schumann: The Man and His Music, ed. Alan Walker. London: Barrie and Jenkins, 1972, 120-161,

.. “Wolf, Hugo,” The New Grove Dictionary o fMusic and Musicians, 20 vols., ed. Stanley Sadie. London: Macmillan, 1980, XX, 475-502.

Schauffler, Robert Haven. The Unknown Brahms. New York: Dodd, Mead, 1936.

Schochow, Maximilian, and Lilly Schochow. Franz Schubert: Die Texte seiner einstimmingkomponierten Liederund ihre Dichter, vols. 2 Hildesheim: Georg 01ms, 1974.

Smeed, John William. German Song and Its Poetry, 1740-1900. London: Croom Helm, 1987.

Stein, Jack M. Poem and Music in the German Lied from Gluck to Hugo Wolf. Cambridge, Massachusetts: Harvard University Press, 1971.

Stohrer, Sister Mary Baptist. “The Selection and Setting of Poetry in the Solo Songs of Johannes Brahms.” Ph.D. dissertation. The University of Wisconsin Madison, 1974.

Walker, Alan. “Schumann and his background,” Robert Schumarm: The Man and His M usic, eà. Alan Walker. London: Barrie and Jenkins, 1972, 1-40.

Walker, Frank. Hugo Wolf: A Biography. New York: Knopf, 1952.

Walsh, Stephen. TheLiederofSchumarm.NewYoTkiPrdeget, 1971.

Whitton, Kenneth. Lieder: An Introduction to German Song. London: Julia MacRae, 1984.

Wilpert, Gero von. Deutsches Dichterlexikon. Stuttgart: Kroner, 1976.

Chapter II

Bemac, Pierre. Francis Poulenc: The Man and His Songs, tr. Winifred Radford. New York: Norton, 1977.

. The Interpretation o f French Song. Nc«' York: Norton, 1970.

Brée, Germaine. Twentieth-Century French Literature, tr. Louise Guiney. Chicago and London: University of Chicago Press, 1983.

Cobb, Margaret G. The Poetic Debussy: A Collection o f His Song Texts and Selected Letters. Boston: Northeastern University Press, 1982.

Cooper, Martin. “Duparc, Henri,” The New Grove Dictionary of Music and Musicians, 20 vols., ed. Stanley Sadie. London: Macmillan, 1980, V, 726-7. 258

Daniel, Keith W. Francis Poulenc: His Artisdc Development and Musical Style. Ann Arbor: UMI Research Press, 1982.

Debussy Letters, selected and ed. François Lesure and Roger Nichols, tr. R. Nichols. Cambridge, Massachusetts: Harvard University Press, 1987.

Gere, Charlotte. M arie Laurencin. New York: Rizzoli, 1977.

Hell, Henri. Francis Poulenc, ir. Edward Lockspeiser. New York: Grove Press, 1959.

Jarocinski, Stefan. “Debussy and His Poets,” Introduction to the complete recordings of Claude Debussy’s mélodies. Le Plessis Robinson: Blanchard, 1980, EMI 2C 165- 16.371/4.

Koechlin, Charles. Gabriel Fauré, tr. Leslie Orrey. London: Dennis Dobson, 1946.

Lockspeiser, Edward. Debussy: His Life and Mind, 2 vois. London: Cassell, 1952, 1965.

.. Debussy, The Master Musicians Series. London: J. M. Dent, 1936, 5th edition. rev. Richard Langham Smith, 1980.

Nectoux, Jean-Michel. Gabriel Faure: His Life Through His Letters, tr. J. A. Underwood. London and New York: Marion Boyars, 1984.

.. “Gabriel Fauré,” The New Grove Twentieth-Century French Masters, originally published in The New Grove Dictionary o fMusic and Musicians,ed. Stanley Sadie. London and New York: Macmillan, 1980, 1986, 1-37.

Nichols, Roger, work-list with Robert Oriedge. “Claude Debussy,” TheNev/ Grove Twentieth-Century French Masters, originally published in The New Grove Dictionary o fMusic and Musicians, ed. Stanley S ^le. London and New York: Macmillan, 1980, 1986,41-125.

“Francis Poulenc,” The New Grove Twentieth-Cenmry French Masters, originally published in The New Grove Dictionary o f Music and Musicians, ed. Stanley Sadie. London and New York: Macmillan, 1980, 1986, 197-218.

Northcote, Sydney. The Songs o fHenri Duparc.Lonàorv. Detmis Dobson, 1949.

Noske, Frits. French Song from Berlioz to Duparc,tr. Rita Benton. New York: Dover, 1970.

Oriedge, Robert. Gabriel Fauré. London: Eulenburg, 1979.

Poulenc, Francis. Diary o fmy Songs [Journal de mes Mélodies], trans. Winifred Radford. London: Victor Gollancz, 1964, trans. 1985.

Rohinsky, Marie-Claire. The Singer’s Debussy. New York: Pelion Press, 1987.

Seroff, Victor I. Debussy: Musician of France. New York: G. P. Putnam’s Sons, 1956. 259

Suckling, Norman. Gabriel Fauré, The Master Musicians Series. New York: Pellegrini & Cudahy, 1951.

. “The Songs of Fauré,” The Listener, WXIW (March 15, 1945), 305.

Vuillermoz, Emile. Gabriel Fauré, tr. Kenneth Schapin. Philadelphia: Chilton, 1960, 1969.

Wandycz, Piotr S. The Lands o fPartitioned Poland 1795-1918. Seattle: University of Washington Press, 1974.

Wenk, Arthur B. Claude Debussy and the Poets. Berkeley: University of California Press, 1976.

Wood, Vivian. Pouhmc’s Songs: An Analysis o f Style. Jackson: University Press of Mississippi, 1979.

Chapter III

Anthony, Michael. “Review of the premiere performance of Argento's The Andrée Expedition. ” IJiFi/MusiœlAjneriœXXXJIl/6 (June 1983), MA 29-30.

“Argento, Dominick,” Baker’s Biographical Dictionary o fMusicians, 7 th ed., ed. Nicolas Slonimsky. New York: Schirmer Books, 1984.

“Argento, Dominick,” Contemporary American Composers: A Biographical Dictionary, 2nd ed., comp. E. Ruth Anderson. Boston: G. K. Hall, 1982.

“Argento, Dominick,” Current Biography Yearbook 1977, ed. Charles Moritz. New York: H. W. Wilson, 1977.

Argento, Dominick, “The Matter of Text,” The NATS Journal 44/4 (March/April 1988), pp. 6- 10.

Block, Geoffrey. Charles Lves: A Bio-Bibliography. Westport, Connecticut: Greenwood Press, 1988.

Bloomquist, Marvin Robert. “Songs of Ned Rorem: Aspects of the Musical Settings of Songs in English for Solo Voice and Piano.” Ph.D. dissertation. University of Missouri, Kansas City, 1970.

Boatwright, Howard “The Songs,” The Music Educator’s Journal 61, no. 2 (October 1974), 42-47.

Burkholder, J. Peter. Charles Lves: The Ideas Behind the Music. New Haven and London: Yale University Press, 1985.

Contemporary Authors, Nc'm Revision Series, 127 vols, to date, ed. Christine Nasso and Ann Evory. Detroit: Gale Research, 1962-89.

Cowell, Henry, and Sidney Cowell. Charles Lves and His Music. London: Oxford University Press, 1955, reprinted 1969. 260

De Lerma, Dominique-René. Charles Ives, 1874-1954: A Bibliography o fHis Music. Kent, Ohio: Kent State University Press, 1970.

Elizabethan Lyrics, 3rd ed., ed. Norman Ault. New York: William Sloane, 1949.

Feldman, Mary Ann. “Argento, Dominick,” The New Grove Dictionary o fAmerican M usic, 4 vols., ed. H. Wiley Hitchcock and Stanley Sadie. London and New York: Macmillan, 1986,1, 64-65.

Friedberg, Ruth C. /imerican Art Song and American Poetry, 3 vols. Metuchen, New Jersey: Scarecrow Press, 1981.

Hart, James D. The Oxford Companion to American Literature. New York: Oxford University Press, 1983.

Holmes, James. “Rorem, Ned, ” The New Grove Dictionary o fAmerican Music, 4 vols., ed. H. Wiley Hitchcock and Stanley Sadie. London and New York: Macmillan, 1986, IV, 87-90.

Ives, Charles E. M em os, ed. and with appendices by John Kirkpatrick. New York: Norton, 1972.

Jackson, Richard, with Work-list by Barbara Heyman. “Samuel Barber " The New Grove Twentieth-Century American Mastersiori^naWy published in The New Grove Dictionary o fAmerican Music, ed. H. Wiley Hitchcock and Stanley Sadie). New York: Norton, 1986, 1988, 241-259.

Kennedy, X. J. An Introduction to Poetry, 3rd ed. Boston: Little, Brown, and Co., 1974.

Kirkpatrick, John. A Temporary Mimeogmphed Catalogue oftiie Music Manuscripts and Related Materials o f Charles Edward Ives 1874-1954. New Haven: Library of the Yale School of Music, 1960, reprinted 1973.

_. “Charles Ives,” The New Grove Twentieth-Century American Masters{ongm.a\\y published in The New Grove Dictionary o f American Music, ed. H. Wiley Hitchcock and Stanley Sadie). New York: Norton, 1986, 1988.

Kreiling, Jean Louise. “The Songs of Samuel Barber: A Study in Literary Taste and Text- setting.” Ph.D. dissertation. University of North Carolina at Chapel Hill, 1986.

Meleager of Gadaia. The Poems o f Meleager, tr. Peter Whigham and Peter Jay. London: Anvil Press Poetry, 1975.

Miller, Philip Lieson. “The Songs of Ned Rorem,” Tempo no. 127 (December 1978), 25- 31.

The National Union CatalogPre-1956Imprints, 685 vols. London and Chicago: Mansell, 1978.

“The NATS Bulletin Interviews Ned Rorem,” The NATSBuIletin29/2 (November/December 1982), 5ff. 261

Newman, Philip. “The Songs of Charles Ives,” 3 vols. Ph.D. dissertation. The University of Iowa, 1967.

North, William Sills Wright. “Ned Rorem as a Twentieth Century Song Composer.” Ph.D. dissertation. University of Illinois, 1965.

The Oxford Companion to English Literature, 5th ed., ed. Margaret Drabble. Oxford: Oxford University Press, 1985.

Parthun, Paul. “Concord, Charles Ives and Henry Bellamann,” Student Musicologists at M innesotaS (1975-1976): 66-86 .

Perils, Vivian. Charles Ives Remembered: An Oral History. New Haven and London: Yale University Press, 1974.

Perry, Rosalie Sandra. Charles Ives and the American Mind. Kent, Ohio: Kent State University Press, 1974.

Pool, Evelyn Ivora. “A Study of Ned Rorem and his Song-Cycle Poems o f Love and the Rain.” M.A. thesis. The Ohio State University, 1966.

Ramey, Phillip. “Songs of Ned Rorem” and “A Talk with Ned Rorem.” Notes to recording, “Songs of Samuel Barber and Ned Rorem,” New World Records, NW 229, 1978.

_. “Songs of Samuel Barber” and “A Talk with Samuel Barber.” Notes to recording. “Songs of Samuel Barber and Ned Rorem,” New World Records, NW 229, 1978.

Rorem, Ned. An Absolute Gift: A New Diary. New York: Simon and Schuster, 1978.

. Critical Affairs: A Composer's Journal. New York: George Biaziller, 1970.

. Letter to the author, July 22, 1988.

. Letter to the author, August 1, 1989.

. Music andPeople.'New York: George Braziller, 1968.

. Music from Inside Out. New York: George Braziller, 1967.

. The Paris and New York Diaries. San Francisco: North Point Press, 1983.

. Setting the Tone. New York: Coward-McCann, 1983.

. Settling the Score. New York: Harcourt Brace Jovanovich, 1988.

Rossiter, Frank R. Charles Ives and His America. New York: Liveright, 1975.

Schoffrnan, Nachum. “The Songs of Charles Ives.” Ph.D. dissertation, Hebrew University of Jerusalem, 1977. 262 Wallach, Laurence. “The New England Education of Charles Ives.” Ph.D. dissertation, Columbia University, 1973.

Woolridge, David. Charles Ives: A Portrait. London: Faber and Faber, 1975.

Yellin, Victor Fellin. Review of Essays Before a Sonata, by Charles E. Ives. Journal o f the American Musicological Society \1 (Summer 1964): 229-231.