Printer Friendly

To be or not to be: notes on the muted E, Part 2.


THE CARRYING OUT OF THE FRENCH muted e in lyric diction is a complex issue that transcends linguistics. The causes and the motives that stirred and fed the passionate debate that unfolded around it at the close of the nineteenth century were outlined in the first part of this article. This section will take an experimental and more practical stance, and will survey the performances of singers that collaborated with the composers who broke ties with traditions.

Thanks to recordings made in the early 1900s, it is possible to evaluate if the innovative ways of notating muted es prompted a novel style in singing them--above all, if the performances on records are uniform. The performers selected are Reynaldo Hahn, singing or accompanying his songs; Debussy accompanying soprano Mary Garden, his first Melisande; Charles Panzera, Jane Bathori, and Ninon Vallin, who have together collaborated with, and premiered works by Faure, Debussy, Hahn, Ravel, and Satie, among others.

One short essay, published by Pierre de Breville and reviewed in Part 1, lays out several ways of carrying out the final muted e, but does not allow any leeway to the performer in its execution. According to Breville, each alternate notation corresponds to a unique interpretation. Such a radical method may be unrealistic with the elusive nature of the schwa.

Eleven melodies, Bizet's "Habanera," and a short excerpt from Pelleas were surveyed for this study. Evidently, thirteen excerpts represent a small fraction of the repertoire; they cannot be considered a definitive model, but the historical performances on recordings, shown in Examples 9 through 18, reveal patterns that clearly indicate a deliberate approach. Recognizing them will benefit current knowledge.

To save limited space in the figures, the conservatively written schwas were excluded when carried out according to the tradition (in 80% of the cases). The twenty-four exceptions were included in the examples.


VoceVista 3.2 was used in order to transcribe the excerpts. (1) The length of a syllable, consequently the note value corresponding to that note, was calculated uniformly starting from the onset of a vowel, and ended at the onset of the subsequent vowel (or at the end of phonation for ends of phrases). Although the readings with the spectrogram would not have been sufficiently defined to identify the formant values of the vowels, they were showing accurately their timing in microseconds. Note values were rounded up to the sixty-fourth note. VoceVista proved essential to recognize if a neutral vowel was phonated in words ending with a consonant. There was no ambiguity on the spectrograms.






The Notation Patterns: Pattern a through e

Altogether, 137 examples of final muted es notated in five alternative patterns were accounted for within the songs surveyed. Four of those patterns involved the use of a tie: eighty occurrences (58%) between two notes on similar pitches. Such occurrences will be referred as pattern a in this article (Example 1). Twenty-six occurrences (19%) on two different pitches will subsequently be referred as pattern b (Example 2). Six occurrences (4%) between a note and an appoggiatura will be referred as pattern c (Example 3), and three occurrences (2%) between a note and a rest correspond to pattern d (Example 4). The last alternative, pattern e, consists of a unique note for both penultimate and final unstressed syllables (Example 5). It occurred twenty-two times (16%).

Those five patterns represent the prevailing innovative notations of muted es, although additional notations (the use of parentheses, for example) occur randomly in other isolated works. (2)

Schwa Types: Type 0, Type 1, and Type 2

The performers on the recordings did not carry out the five patterns described above uniformly. What seemed so one dimensional in Breville's "Note on the muted e" (see Part 1) does not correspond to the practice. This article will introduce schwa types, a categorization that typifies schwas according to their sounding (how they are carried out), regardless of their notation. The purpose of such classification is to assess, organize, and analyze the selected performances of the next section in order to correlate more efficiently how composers notated schwas, and how performers carried them out.




Apocopated final muted es will be referred as type 0 schwas (Example 6). A type 1 schwa refers to when the penultimate syllable of a word is sung as written, without elongating its value, and the onset of the schwa is rhythmically accurate. However, the schwa is held for only a fraction of its actual value, but long enough to show on a spectrogram (Example 7). With the type 2 schwa, the vowel of the penultimate syllable is elongated and the schwa is delayed. Its value is reduced by half or by a quarter (Example 8).

The English speaker must bear in mind that in French, accentuation is carried out through duration of the vowel only (on average, an accented vowel is twice as long), and always falls on the last syllable of a phrase. (3) Therefore, type 2 schwas correspond to the intrinsic way of performing a feminine rhyme.


Ninon Vallin

Ninon Vallin (1886-1961) sang the first performance of Debussy's Le martyre de Saint Sebastien in 1912, and premiered with the composer at the piano his Trois poemes de Mallarme in 1914. She collaborated with several other composers, Albert Roussel, Reynaldo Hahn, and Joaquin Nin among the better known, and quickly established herself as an iconic figure of French opera. She recorded Bizet's "Habanera" in 1928 (Example 9). Her apocopation of all final muted es is equivalent to other notable recordings of Carmen made by Emma Calve (1907), Germaine Cernay (1942), or Solange Michel (1950). There is an unmistakable distinction between the carrying out of the schwas notated with tied notes, and the ones notated with individual notes, as Example 9 shows.


Mary Garden

Scottish-born soprano Mary Garden (1874-1967) recorded three excerpts from Ariettes oubliees (Debussy dedicated the second and revised edition of the whole set to her) with the composer at the piano in 1904, as well as the third act opening air from Pelleas. "Mes longs cheveux," "Il pleure dans mon coeur," and "Green" are shown in Example 10.4 She worked in close collaboration with Debussy, as well as Massenet, Charpentier, and Messager. In the excerpts studied, the majority of muted es (twenty) are notated without tie or slur, and Garden sings them with their full value. On three occasions (the words descendent and attendent from Pelleas and reve from "Green," included in Example 10), she minimized the value of the schwas. In all the instances where Debussy notated the schwas with tied notes, Garden altered their duration and carried them out, mostly as type 0 or type 2 schwas. There is one omission for the word tempete in "Green," but its musical context (the combination of ternary rhythm in the voice against the duplets in the piano, combined with the ongoing rallentando) makes it a marginal exception.


Jane Bathori

Jane Bathori (1877-1970) made her first public appearance singing Hahn's "Offrande" and "Prison," accompanied by the composer himself when she was twenty-one. Her exceptional musicianship earned her the esteem of the greatest living composers and conductors. Early in her career, she sang at La Scala opposite Caruso under the direction of Toscanini, but she quickly became one of the most sought after performers and advocates of new music. She premiered, among other pieces, Ravel's Histoire naturelles and Trois poemes de Mallarme, Debussy's Promenoir des deux amants, and Satie's Socrate. Trained as a pianist, she accompanied herself singing on the recordings she made in her early fifties. Hahn's "Prison" and "Offrande," as well as Ravel's "Le paon" were selected for this study. Since she played for herself, she was in complete control of the synchronization and the timing between the accompaniment and the articulation of the muted es, and her performances are most valuable.

In "Prison," she carried out all eight final schwas notated with a tie as type 2 schwas (Example 11). She also sang two of the four traditionally written muted es as type 2 schwas.

Ravel's Histoires naturelles brought Debussy's notion of "singing like a natural person" to such an extreme that Debussy himself rebuked Ravel's work. (5) Jules Renard's prose gave to Ravel all the freedom he needed to experiment with a new approach to lyric declamation. In "Le paon," Ravel makes use of the tie extensively in order to notate muted es. Bathori sang all of them as either type 0 or type 2 schwas. She made one omission on the word incapable, which she sang as if no tie were written. Example 12 displays in detail her performance as she recorded it. Nearly half of the schwas are apocopated (syncopated in the case of sirement). This includes all the ones which were not assigned an individual note (pattern e), but also many notated with a tie. When the schwas were sung, they were realized as type 2 schwas. The penultimate syllable is consistently elongated, the onset of the schwa delayed, and its duration shortened to a fraction of its value.


Charles Panzera

The youngest performer selected for this study, Charles Panzera (1896-1976), was nevertheless a protege of Gabriel Faure, and the aged composer wrote for and dedicated to him his last song cycle, L'horizon chimerique, in 1921. Panzera recorded extensively Faure and Duparc songs, and his performance reveals a slight tendency to carry out final schwas as type 0, although the majority of the traditionally notated schwas remain sung for their full value. Two songs by Reynaldo Hahn, "Prison" and "Cimetiere de campagne," and Debussy's second and third Villon ballads, were selected for this study because they contain muted es written out in both traditional and innovative manner. Example 13 displays the instances when the duration of schwas were significantly altered in "Ballade que Villon feit a la requeste de sa mere pour prier Nostre Dame." As the example shows, all of the schwas notated with a tie were mostly performed either as type 0 or type 2 schwas. Only one exception, the word chrestienne, was sung with its full value. On the other side, all but two (humble and pecheresse) of the eighteen schwas notated without ties and on separate notes were performed as written. Once more, only both exceptions were included in Example 13.

Example 14 shows the results for "Ballade des dames de Paris" in their entirety. Panzera respected the articulation markings, notably the staccatos, but ignored all but three muted es notated with a tie, which he performs as type 2 schwas. He carried out a portamento on each of those ties, but did it too quickly to alter significantly the note values.



In "Prison" (see Example 11), Panzera delayed and shortened the value of six out of the eight schwas notated with a tie. The remaining two (palme and the final calme) are sung as they are written. On the other hand, he sings one of the four traditionally notated muted es, on the word cloche, as a type 2 schwa.

Reynaldo Hahn

Hahn, accompanying himself at the piano, recorded several of his own compositions as well as melodies by other fellow composers. Two of his earliest recordings from 1909, "Cimetiere de campagne" and "Offrande," are included in this study, since the former was also recorded by Charles Panzera and the latter by Jane Bathori, and their performances can be juxtaposed for comparison. A quick look at Example 15 reveals that Hahn's reading is more straightforward and conformed more to the score than Panzera's in "Cimetiere de campagne." The first system in Example 15 lays out all the instances where Panzera delayed and shortened the value of final muted es, or simply left them apocopated, regardless of Hahn's notation. In the same excerpts, Hahn observed the elisions as they were indicated, and placed the onsets of the remaining schwas accurately rhythmically. The second system of Example 15 displays all the instances where Hahn uses a tie to notate the muted es. It is interesting to see how both Panzera and Hahn carried out those schwas as type 0, type 1, or type 2 schwas, but how their performances did not match one another. Panzera sang them all but one (the word jeunesse) as type 2 schwas, which conformed to the instructions given by Breville in his notes. Hahn only sang the word herbe as a type 2 schwa, and in doing so he elongated the penultimate syllable as much as possible. He carried out most of the selected schwas as type 0, synchronizing the release of the consonants with the offset of the voice, but sang the word entiere without paying attention to the tie. However, since the word itself means "whole," Hahn's singing of the full value of the schwa adds to the connotation of the word.



In "Offrande," one observes the same tendency between Hahn's own reading, and the performance of Jane Bathori. The song contains only two schwas notated with a tie, and one with an acciaccatura. All the other schwas are notated in a traditional manner. The three distinctive schwas occur within the line "j'arrive, tout couvert encore de rosee" and are displayed in Example 16, first system. All three are carried out, delayed or simply shortened by Bathori. Hahn shortens two of them, while encore is apocopated. The second system in Example 16 displays five instances where Bathori performs traditionally written muted es as type 2 schwas. Hahn does the same thing on two of those words, reposee and reve, and matches perfectly Bathori's rhythmic alteration. However, his performance of branches, blanches, and tete, as well as all the other muted es of the song, is straightforward and conforms to his notation. Although it is highly subjective, one could argue that Hahn's freedom on both reposee and reve is an instinctive response to text painting.

Example 17 presents two songs that Hahn recorded accompanying Ninon Vallin, "Si mes vers avaient des ailes" and "Tyndaris" (from Etudes latines). Each song displays only one instance of muted e atypically notated. In "Si mes vers," Vallin respected Hahn's notation and carried out each of the final es accordingly to the rhythmic values assigned to them. They are not included in Example 17 since her performance replicates the printed music. However, the example displays how she differentiated both notations that Hahn used for the word ailes from the recurring line "si mes vers avaient des ailes." In "Tyndaris," Vallin's performance mirrored every detail of Hahn's notation as well. She sang feuillage, the only word whose schwa is notated with a tie, as a type 2 schwa.

"Tyndaris" shows an improvement in Hahn's notation. Words such as collines, plaintives, and sources fugitives are written out exactly as they are meant to be sung. Hahn notated the elongation of the penultimate syllable, instead of suggesting it in writing the schwa. An alternate notation that would suggest the same rhythmic declamation for those words is added beneath them in Example 17. The line "Et, sous l'epais feuillage il est doux de dormir" is a singular instance, and the use by Hahn of a tie on the word feuillage shows that this particular schwa is more elusive than the other schwas of the song. The line itself is an inversion (it would read as "il est doux de dormir sous l'epais feuillage," meaning "it feels good to sleep beneath the thick foliage"), and therefore carrying out the elision between feuillage and il would constitute a transgression, both phonetic groups being grammatically independent from each other. (6) With the tied notes, Hahn indicates that such an elision is not to be made, and that the muted e offeuillage must be sung, although as lightly as possible, which is what Vallin does.


The last excerpt of this study displays the performance of dilettante tenor Guy Ferrant, partner of Reynaldo Hahn, who is accompanied by the composer in his melodie "Sur l'eau." In addition to the usual tied notes, Hahn used abundantly both appoggiaturas and ties over rests in this song. Without being systematic, Ferrant's performance shows recurrent patterns. As displayed in Example 18, the majority of the schwas notated with a tie over a rest were carried out as type 0 schwas; the majority of those notated with an appoggiatura were phonated very briefly as type 1 schwas; finally, the majority of those notated with tied notes were delayed and carried out as type 2 schwas.


The figures speak for themselves, and confirm that the performers selected for this study responded to the innovative notation used by composers in their performance of final muted es.



Overall, there were 137 occurrences of atypically notated schwas within the examples displayed. Fifty-one of them were apocopated by the performers; another twenty-four were significantly shortened to a fraction of their written value. Fifty others were sung delayed, and their penultimate syllables were elongated proportionally to the delay. These particular manners of carrying out schwas are respectively referred to as type 0, type 1, and type 2 in this article.


Figure 1 shows side by side those three types of schwas, and their distribution by notation patterns (as previously described). For three of those patterns, the outcome is unequivocal:

pattern e is performed apocopated (type 0),


The results are not as straightforward for both patterns a and b, which are carried out in all possible ways depending on the performers (as in Examples 15 and 17) and the contexts. Setting a word's final schwa on the similar pitch as its penultimate syllable, and then connecting both notes with a tie (pattern a), is the most recurrent pattern (eighty occurrences, 59%) among the innovative notations found in the selections surveyed. Half of those schwas were sung delayed (type 2), at the expiration of the voice, as Breville suggested.


Indeed, type 2 schwas correspond to the manner indicated by Breville to perform such notation. However, the performers on recording only comply with Breville half of the time. They perform twenty-six of those schwas apocopated,


They disregard the tie in only three occasions (less than 4%), which corroborates that they undeniably recognize the tied-note schwas as being specifically notated to sound different from traditionally notated schwas.

At first glance, the choices made by the performers when carrying out schwas notated with a tie seem arbitrary, but such is not the case. Type 0 schwas connote colloquial speech, where schwas are commonly apocopated. Vallin's use of them in Carmen reinforces the characterization of Bizet's heroine. Such treatment of the schwas is relevant to typify characters from the working and popular classes, as well as ingenues and younger antagonists.

Jane Bathori's apocopated schwas are omnipresent in Histoires naturelles, and emulate an unaffected pronunciation. However, French melodie is and will always remain an aristocratic art, the musical expression of literary texts. Ravel's setting of a prosaic text constitutes an isolated experiment in his melodies. Subsequently, the composer transposed the popular accent to his stage works. Several other schwas notated with a tie are apocopated by Mary Garden in Pelleas and Ariettes oubliees, Charles Panzera in the second Villon ballad and Hahn's songs, and Hahn himself in his songs. None of these examples can be justified as a characterization of a popular tone, and their number excludes the possibility of an oversight.

French composers notated final es on a tied note to signify their lessening. In the absence of formal guidelines, the performers have the freedom to delay their onsets, thus lengthening the penultimate syllable; they can voice them minimally at the release of the final consonants; or they can revoke them altogether. Their choice may depend on the tempo of the piece, on the context of the poetic line, but it can also vary according to the specific interpretation of the moment. There is one option the selected performers disregarded: to carry out the schwas as if no tie were written. There is no doubt that they recognized significance in the composer's specific notation of the muted e.

When schwas were tied to their penultimate syllables, but set on a different pitch (pattern b), the performers carried them out primarily as type 2 schwas. However, they disregarded the ties on nine out of twenty-six instances (35%), and sang those schwas for their full value. Six of those exceptions occurred in Debussy's "Ballade des dames de Paris," sung by Panzera.

Panzera's performance of "Ballade des dames de Paris" does not invalidate the general tendency observed. It is true that in this song, Panzera sang most of the schwas notated with tied notes without delaying their onset, albeit he clearly pays close attention to Debussy's detailed articulation and thoroughly observes portamentos, tenutos, and staccatos on the score. Nevertheless, the song's quick tempo and declamation inhibit the performer's rhythmic freedom in stretching and delaying note values. In other words, there is no time to be subtle with schwas in such a song. A closer look at the three occurrences of type 2 schwas in this third "Ballade" reveals that two of them, Calaisiennes and Italiennes, coincide with a significant slowing down of the pulse. A slower declamation facilitates the performance of type 2 schwas, and it is no surprise that this particular way of performing schwas is encountered mostly in slow-moving works.

The composers who disapproved of the innovative approach to notating and singing schwas were primarily terrified of seeing the sophistication of their art jeopardized. Such an agenda was never in the minds of their challengers. Indeed, by creating new options of realizing the muted e, they updated and refined lyric declamation to new heights. The complexity of their notation system, which remains to this day elusive, contributed in keeping the art of singing from the uninitiated. They sought a natural declamation, but never a simplistic method.

This is above all true in Debussy's work, where innovative and traditional notations are used alongside. Very little (if anything) is left ambiguous in Debussy's creative process, and to assume that his treatment of the muted e was no better than "hit or miss" is to make a critical mistake. Indeed, the key to the understanding of his text settings and their poetic correlations often lies in grasping his reasons for setting the muted e in one particular way or another. The muted e bore a multitude of connotations in the nineteenth century composer's psyche. To overlook them is to remove a part of the significance of the lyric works written within those years.

Finally, observing the notation that a composer adopted for his final es is not so much a question of bon gout, but rather a question of being faithful to the score, in the same line as being accurate with notes, rhythms, articulation, tempos, and dynamics. As long as a composer is specific in notating his intentions, it is part of the performer's responsibilities to make an effort in respecting them.


Banville, Theodore de. Petit traite de poesie francaise. Paris: Le Clerc, 1872.

Bremont, Leon. L'art de dire les vers. Paris: Charpentier et Fasquelle, 1903.

Calvocoressi, Michel. "Le vers, la prose et l' <<e>> muet." Le Guide Musical 50, no. 44 (30 October 1904): 795.

Debussy, Claude. Monsieur Croche et autres ecrits. Paris: Gallimard, 1987.

Koschwitz, Eduard. Lesparlersparisiens. Paris: Welter, 1893.

Saint-Saens, Camille. "La Question de l'E muet au double point de vue litteraire et musical." Le Figaro no. 232 (19 August 1904).

Souza, Robert de. "Le role de l'e muet dans la poesie francaise." Mercure de France 13, no. 61 (January 1895).

Strauss, Richard, and Romain Rolland. Correspondance. Fragments de journal. Paris: Albin Michel, 1953.

Vives, Vincent. La beaute et sa part maudite. Aix-en-Provence: Publications de l'Universite de Provence, 2005.

Vives, Vincent. Vox humana, Poesie, musique, individuation. Aix-en-Provence: Publications de l'Universite de Provence, 2006.


(1.) The author thanks Sarah Khatcherian for her assistance in operating VoceVista, and for sharing her competence in reading the results.

(2.) A random occurrence in Debussy is shown in Part 1 of this article (Example 2). The use of parentheses is observed in Dukas' Barbe-bleue, or Charpentier's Louise and Julien.

(3.) See treatises and essays by Pierre Delattre, Pierre Fouche, Maurice Grammont, etc.

(4.) Extensive rubato in their performance made "L'ombre des arbres" impossible to transcribe objectively.

(5.) Claude Debussy, Monsieur Croche et autres ecrits (Paris: Gallimard, 1987), 63.

(6.) A phonological [phonetic] word is what constitutes at once a flow unit, a grammatical unit, significance unit. Jean-Claude Milner and Francois Regnault, Dire le vers (Paris: Seuil, 1987), 28-9.

Quebec-born pianist Martin Neron completed his doctorate on scholarship at the Manhattan School of Music. Praised as "an attentive partner" (OperaNews), Mr. Neron is an active recitalist, vocal coach, and educator. He has recorded different collections of French, Greek, and English art songs on the One Soul label, and a program of songs by Mikis Theodorakis on the Romanos label. In addition, Dr. Neron has recorded several concerts as a soloist and as a collaborative artist for the Canadian Broadcasting Corporation (CBC). His first book, a study guide on the songs of Francis Poulenc, was recently published by Leyerle Publications.

Leslie De'Ath, Associate Editor
COPYRIGHT 2012 National Association of Teachers of Singing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Neron, Martin
Publication:Journal of Singing
Article Type:Essay
Geographic Code:1USA
Date:May 1, 2012
Previous Article:An appreciation of the Bozeman and Miller descriptions of Formant--harmonic relations in singing.
Next Article:Voice pedagogy for aging singers (including the author).

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters