Latest news

Avior Byron will present a paper 'Huberman as Beethoven' in the 2010 Israel Musicology Conference.   

Subscribe to

Enter your email and press the 'Subscribe' button to receive blog posts via email:


Subscribe via RSS

What is RSS?

Follow @avior on Twitter

People reading Bymusic now

We have 1 guest online

Translate this page now

Recently at the Blog

Who is Behind

Avior Byron

My name is Avior Byron and I am a musicologist, blogger and composer. I write books, articles and a blog about music, performance, research, and theory. Read more at my about page

What next?

 Subscribe to the Blog

Click here to add site to your favorites

What I'm doing just now


What is THE best way to find scholarships?

Evaluating Sprechstimme: what early recordings tell us - the chapter

Evaluating Sprechstimme: what early recordings tell us

I am including here the first part of the chapter that I just wrote in the British Library as part of the Edison fellowship. You can read the whole chapter at my latest research page. This will be a chapter in a book that I plan to publish on Schoenberg and Performance:

After I wrote my article ‘Sprechstimme reconsidered’[1] I was sure that I finally solved what Boulez and Milhaud called ‘The Sprechstimme enigma’: namely, how should the vocalist in Pierrot lunaire perform the vocal part? As mentioned above, to many commentators the evidence seemed confusing: Schoenberg’s exact notation and demand to perform that notation without adding anything that is not notated, on the one hand, and his vague performance instructions at the preface of the score, and the recordings of him conducting the piece with Stiedry-Wagner not reproducing the notated pitch, on the other hand. I claimed quite confidently that ‘the test pressings of Pierrot lunaire confirm that a perfect reproduction was not Schoenberg’s intention.’[2] Stiedy-Wagner’s test pressings revealed a process of live improvisation in performance that was a very free (although not completely free) way of rendering the notated pitch.

            However, recently two books on Sprechstimme appeared. They were written by singers who perform Pierrot as part of their standard repertoire, and they argued the complete opposite of what I did. Aidan Soder suggested that Schoenberg did not have enough rehearsal time and that ‘the final product on Schoenberg’s recording is perhaps not how he heard it in his ear’.[3] If Soder is right, then perhaps my observations that are based on this recording should be seen as a compromise done by the composer.[4] In the second publication, Paul Mathews and the singer Phyllis Bryn-Julson gave preference to what they understand as Schoenberg’s ‘original conception of the sound’.[5] They argued that ‘the performer would likely prioritize a performance of the passage [in Pierrot] as notated, because … she will find correspondences of pitch and motivic shape in the surrounding texture’.[6] They claimed that ‘the correct interpretation of Sprechstimme is to emphasize the pitch and minimize the effects of “falling and rising”’ by doing glissando.[7] They seem to argue against (the aforementioned article by Stein that states) the idea that one can transpose the Sprechstimme part, since such transpositions, they believe, will cause ‘unintended consequences’.[8] They conclude that a performer that does not sing the notated pitch may feel ‘liberated’, yet many such performances ‘sound self-conscious and mannerist’.[9] At certain moments it seems as if they echo Eugene Narmour’s unfortunate claim that ‘many negative consequences’ will occur ‘if formal relations are not properly analyzed by the performer’.[10] Singers that have absolute pitch may feel it natural to perform the notated pitches accurately. One of them is Jane Manning who confessed: ‘From the outset I knew I wanted to try to adhere to the pitches the composer had written and to obey his every marking as far as I was able.’[11] The notion of the importance of being ‘faithful’ to the score is not shared only by some of the singers and musicologists that I have mentioned. As I wrote the current chapter, I stumbled upon a blog post by Maready of ‘The High Pony Tail’ that argued the following: ‘Would Schoenberg have taken such sweet care to imprison Pierrot inside a nightmare latticework of canons and free imitation, giving his instruments free rein to alternately mock and cradle and impersonate him, only to allow the singer to hit whatever notes she pleased? Pierrot’s predicament is that this dandified night music insists on being followed to the smallest workaday detail.’[12]     

            All this made me think that perhaps my article conclusion concerning Schoenberg’s intentions was premature. As I describe in the first part of chapter …., the manuscripts, writings, letters and other evidence by Schoenberg and his circle, are highly contradictory. The picture is far from being clear. In spite of the fact that Stiedry-Wagner was the performer that Schoenberg often employed for many years, it could well have been that he would not be against a performance that would render the notated pitch without variance. Perhaps, due to various reasons, this was the best performance he could receive at the time the recording was made. The singer Martha Elliott, whom is an established Pierrot performer, wrote that today there are quite a few singers that are able to perform Pierrot as notated. She raised the question whether Schoenberg would have liked it performed that way. She concluded that since ‘what Schonberg said he wanted regarding the Sprechstimme in Pierrot and what he got in his lifetime were quite different, we can never determine what the “correct” style really is.’[13] Moreover, she ended her chapter on the Second Viennese School stating that ‘singers today can come closer than many of the original performers to what these composers actually asked for. But whether the composers would ultimately approve of this approach remains unanswerable.’[14] Indeed, all this seems to suggest that one can never really know how Schoenberg intended that the Sprechstimme will be performed. Bryn-Julson and Mathews acknowledge the contradictions, evolution and change in Schoenberg’s conception of Sprechstimme.[15] However, their preference on what they see as the original view of Schoenberg is a very subjective one. Why should one prefer his conception of 1912 rather than that of the 1930s? The contrary may be argued: during the 30s he had much more performance experience and could now really know how he wanted Sprechstimme to be done.   

Apart from voice limitations, taste, and performance traditions, one can try to build an interpretation based on convincing historical evidence. Part of such evidence was presented in the previous chapters. In this chapter I will explore further data that is revealed by an examination of early recordings from the early history of the interpretation of Pierrot lunaire.I will start by discussing reviews of these recordings. The chapter will end with a discussion on how the historical evidence of early recordings and the reviews of these recordings may help one define criterions for aesthetic judgments.

            My case study is the song ‘Parodie’ which contains canonical relationships between the voice and the instruments.[16] It is these relationships that most scholarly commentators have commented on when writing about the song. These relationships may have a connection to the text of the song. Jonathan Dunsby suggested that the texture of contrapuntally related lines are ‘apt in a melodrama entitled ‘Parodie’, featuring knitting-needles’.[17] The canons put before the vocalist the question of how and whether to present them in performance. Dunsby, for example, claimed that ‘the “voice”, for all the strict compositional relationships to be read from the score, is nevertheless still Sprechstimme, with no special instruction to convey the pitches.”[18] Aidan Soder suggested that rhythmic accuracy is enough in order to be aware of imitative relationships.[19] Yet Bryn-Julson and Mathews aforementioned argument, as well as some of the recent recordings[20] show that some performers do find it important to do a Sprechstimme that reproduces the notated pitch. From this perspective, this song is ideal as a springboard for discussing the way vocalists perform Sprechstimme in early recordings.

         In a lecture on Sprechstimme in Pierrot lunaire, the singer Jane Manning said the following: ‘I have a … preference for some of the early recordings, even though they are much less accurate than the recent ones, but they do seem to preserve the spirit of the age rather better than some of the modern ones’.[21] We will see in a moment that many critics heard these recording as quite distinct in character. Moreover, a close examination of early recordings shows a more detailed picture concerning the relation of these performances to notated pitch, as well as other aspects which are special to each performance.

        I will discuss four early recordings. The first one is the recordings by Stiedry-Wagner and Schoenberg in 1940. Three other recordings were recorded in the 1950s. The second was done by Ellen Adler, voice; and René Leibowitz, conductor, around 1951.[22] The next recording was done in 1954 by Leibowitz, once again, yet this time with Ethel Semser.[23] The forth recording that I will examine is from 1957 with Jeanne Héricard, voice and Hans Rosbaud as conductor.[24] One of the reasons behind choosing these early recordings is the strong connection between the conductors and Schoenberg. Leibowitz claimed that he studied with Schoenberg in the early 1930s (although there is no proof to substantiate this claim). He was in contact with Schonberg in 1945 and most of the correspondence between the two was done during the last decade of the composer’s life. Leibowitz promoted Schoenberg’s music after the Second World War by organizing concerts and writing Books.[25] Rosbaud corresponded with Schoenberg from the 1930s and up to the composer’s death. He performed Schoenberg’s music before and after the Nazi period in Germany. He premiered Beglietmuisk zu einer Lichtspielszene, Op. 34 (1930) and Vier Orchesterlieder, Op. 22 (1932). In 1948 he led the South-West German Radio (SWF) where he continued to promote modern music. This broadcast was recorded during two days with the SWF. In 1954 he gave the premier of Moses und Aron.

Where is the rest of the chapter?

You can read the rest of the chapter at my latest research page.

Related posts

How the British Library Sounds Archive helped my research

Evaluating Sprechstimme: what early recordings tell us

Sonic Visualizer - some comments as I go

Follow my research on Pierrot lunaire early recording live on Twitter

Early Performances of Pierrot Lunaire Op. 21 Research Proposal

10 Tips on how to find scholarships and funding

Review of The Glenn Gould Reader

An Interview with David Shemer: The Performance of Early Music - Part I

On being a critical author

How to choose a PhD, MA or DMA subject for a thesis

How to become a freelance musicologist

[1] Avior Byron, ‘The Test Pressings of Schoenberg Conducting Pierrot lunaire: Sprechstimme Reconsidered’, Music Theory Online (MTO), 12/1 (February 2006).

[2] Ibid., 4.8.

[3] Soder, Aidan, Sprehstimme in Arnold Schoenberg’s Pierrot lunaire: A Study of Vocal Performance Practice (Lewiston, Queenston, Lampeter: The Edwin Mellen Press, 2008), p. 18.

[4] Although she underestimated the authority of the recording as reflecting Schoenberg’s ideal intentions, she does not advocate a performance that renders the notated pitch in a perfect manner.

[5] Bryn-Julson, Phyllis and Paul Matthews, Inside Pierrot lunaire: Performing the Sprechstimme in Schoenberg’s Masterpiece (Lanham, Maryland; Toronto; Plymouth, UK: The Scarecrow Press, 2009), p. 49

[6] Ibid., p. 57.
[7] Ibid., p. 62.
[8] Ibid. p. 58.
[9] Ibid.

[10] Eugene Narmour, ‘On the relationship of analytical theory to performance and interpretation’, in Eugene Narmour and Ruth A. Solie (eds.), Exploration in Music, the Arts, and Ideas (Styvesant: Pendragon, 1988), 340. Quoted in Rink02, p. 36. 

[11] Jane Manning, ‘A Sixties “Pierrot”: A Personal Memoir’, Tempo, Vol. 59, July 2005: 17-25.

[12] Erika Sziklay: ‘Pierrot lunaire’, 23 August 2009, Retrieved on 24 August 2009

[13] Martha Elliott, Singing in Style: A Guide to Vocal Performance Practices (New Haven and London: Yale University Press, 2006), p. 238.

[14] Ibid., 250.

[15] It is strange that Bryn-Julson and Matthews claimed that ‘Byron argues that Schoenberg’s view of Sprechmelodie remained fairly consistent’ (Inside Pierrot lunaire, p. 76) when I actually wrote that ‘The history of Schoenberg’s conception of Sprechstimme proves that he understood it differently in different periods.’ Byron, ‘The Test Pressings of Schoenberg Conducting Pierrot lunaire: Sprechstimme Reconsidered’, Music Theory Online (MTO), 12/1 (February 2006), [4.12].

[16] For analyses of such relationships in this song see Dunsby, Pierrot lunaire, pp. 64-65, and Bryn-Julson and Mathews, Inside Pierrot, pp. 187-191.

[17] Dunsby, Pierrot lunaire, p. 65.

[18] Ibid.

[19] Soder, Sprechstimme in Arnold Schoenberg’s Pierrot lunaire, p. 91.

[20] For example, Christine Schäfer, voice; Pierre Boulez, conductor (recorded: IRCAM/Espro, Paris, France, September 1997) *Deutsche Grammophon 457 630-2 GH stereo DDD (1998) CD.

[21] Jane Manning, ‘Schoenberg’s Pierrot lunaire in performance’, Saul Seminar, 7th June 2005, British Library (1CDR 0022875).

[22] Ellen Adler, voice; Paris Chamber Ensemble (Jean-Pierre Rampal, flute & piccolo; Ernest Briand, clarinet; André Dupont, bass clarinet; Francine Villers, violin; Colette Lequien, viola; Sean Barati, violoncello; Claude Helffer, piano); René Leibowitz, conductor; Dial DLP 16 mono (1951?) LP  

[23] Ethel Semser, soprano; Virtuoso Chamber Ensemble (Edward Walker, flute & piccolo; Sidney Fell, clarinet; Walter Lear, bass clarinet; Lionel Bentley, violin; Gwynne Edwards, viola; Willem De Mont, violoncello; Wilfrid Parry, piano); René Leibowitz, conductor (recorded: 1954?)  *Argo RG 54 mono (1955?) LP

[24] Jeanne Héricard, voice; members, Sinfonie-Orchester des Südwestfunks, Baden-Baden (Kraft-Thorwald Diloo, flute; Otto Voigt, piccolo; Sepp Fackler, clarinet; Hans Lemser, bass clarinet; Günther Weigmann, violin; Ulrich Koch, viola; Anton Käsmeier, violoncello; Maria Bergmann, piano); Hans Rosbaud, conductor (recorded: Musikstudio, Südwestfunk, Baden-Baden, West Germany, 4-5 April 1957) *Wergo WER 6403-2 (286 403-2) mono AAD (1993) CD

[25] For example, René Leibowitz, Schoenberg and his School (philosophical library, 1949).


One Response to “Evaluating Sprechstimme: what early recordings tell us - the chapter”

Add your own comment...

Copyright Avior Byron 2017 .