Schoenberg, German speech, intonation, Pierrot Lunaire, Rapoport, intonation contours, intonation contour, progression, poetry readings, Otto Erich Hartleben, singing voice, Bottom Schoenberg, Frau Zehme, Albertine Zehme, Introduction Schoenberg, syllable structure, intonation patterns, intonation pattern, LSt, unstressed syllable, melodic contour
On the Origins of Schoenberg's Sprechgesang in Pierrot Lunaire ELIEZER RAPOPORT Introduction Schoenberg wrote Pierrot Lunaire in 1912, after being commissioned by Albertine Zehme, an actress who organized evenings of poetry reading
s and recitations accompanied by piano or small instrumental (music) ensembles. The texts were chosen from a collection of fifty poems written by the Belgian poet Albert Giraud, translated into German by Otto Erich Hartleben. Schoenberg's original style of sprechgesang (speech-singing) was an innovative use of the human voice in a musical work, as befitting the grotesque-macabre character of the poems, and the fact that Albertine Zehme was an actress rather than a singer. Indeed, she had well-defined personal views on the expressiveness of the speaking voice, which were manifested in her poetry readings, and formulated in the Program notes
of her performances. A very brief quotation from Simms (2000, 120-21): The singing voice
, that supernatural, chastely controlled instrument, ideally beautiful precisely in its ascetic lack of freedom, is not suited to strong eruptions of feeling.... Life cannot be exhausted by the beautiful sound alone. The deepest final happiness, the deepest final sorrow dies away unheard, as a silent scream within our breast, which threatens to fly apart or to erupt like a stream of lava from our lips. ...We need both the tones of song as well as those of speech. My unceasing striving in search of the ultimate expressive capabilities for the "artistic experience in tone" has taught me this fact. Schoenberg had discovered the musicality, and musical potential, of the speaking voice ten years earlier, when he had used a narrator in the final movement of Gurre-Lieder. He returned to this work in 1910, and orchestrated its second part. Hence the correspondence between Zehme and Schoenberg's views. The choice of 21 poems, divided into three groups, was made by Frau Zehme in her poetry reading evenings already in 1911, and was adopted by Schoenberg. These three parts form a narrative, considered by Herrewghe (1992) to be "a descent into the Inferno, and return." The second part, starting with the melodrama Nacht, and ending with Die Kreuze, constitutes the "Inferno," and is the climax of the work. This part is the subject of the present study. In a recent study, Rapoport (2004), pointed out the origins of Schoenberg's sprechgesang. The rhythm was determined by the syllable structure of the German language in Hartleben's texts, and the "melodic line" (sprechmelodie) of the vocal part was molded by intonation patterns in German speech. This was demonstrated by comparing the intonation contours of the texts of four poems being read aloud by two persons whose Mother Tongue
is German, with Schoenberg's melodic vocal line
s in the melodramas. It seems that Schoenberg's very sensitive musical ear perceived the musical melody in the speaking voice. He molded and crystallized it in musical frameworks of metrics, rhythms, and musical intervals, seemingly making musical "transcriptions" of the intonation contours. Atonality is particularly appropriate, as one cannot surmise that speech will conform to musical scales--a much later musical invention--that do
not appear in natural sounds. Thus, in Pierrot Lunaire, Schoenberg's atonality is not posttonality, but rather primordial pre-Tonal Music
. The artist uses his raw material
s and arranges them in stylistic frameworks, emphasizing those aspects that he wishes to emphasize, and, in twentieth-century music, freely, even with deliberate exaggeration and distortion. Nevertheless, in Pierrot Lunaire, it turns out that Schoenberg remained quite close to the origins given above. The purpose of the present work is a deeper investigation of THE RELATIONSHIP
between text, speech intonation, and Schoenberg's sprechmelodies, and the light they shed on his compositional processes. Some of the different musical transformations and variations that Schoenberg utilized are demonstrated. We concentrate on the seven melodramas of the second part of Pierrot Lunaire, of which three--No. 8 Nacht, No. 12 Galgenlied, and No. 14 Die Kreuze--were discussed in my previous study (Rapoport 2004). These poems, together with No. 1 Mondestrunken, were chosen previously as they represent four different facets and four different moods and poetic atmospheres. The addition of No. 9 Gebet an Pierrot, No. 10 Raub, No. 11 Rote Messe, and No.13 Enhauptung, in the present study, is aimed at completing the study of the set of poems comprising the second part--the climax of the work. Methods and Results of the Present Study The source material
of the present study is taken from recordings, made in 2000 by the author, of readings of texts from Pierrot Lunaire by Dr. Lutz Striem (LSt) already used in the author's previous study (Rapoport 2004). Dr. Striem was born in 1920 and educated in Berlin. In the author's view, he represents the closest possible German cultural environment
and speech intonation to Schoenberg's period. LSt was not previously acquainted with the texts. He read each text twice. The First Reading
represents the reader's first impression and spontaneous reaction to the text, and, in most cases, was used for the analysis in the previous study. However, significant changes were sometimes observed on the second (rehearsed) reading as the reader got acquainted with, and gained a deeper comprehension of the text. The FFT spectrograms of LSt readings were used to create transcriptions to the musical notes
, and were raised by an octave--to be compared to Schoenberg's sprechmelodies. As in the previous study, the transcriptions were made by measuring the sound frequency at the midpoint between the tone onset and its termination, assigning to it the note closest in frequency taken from the Tables. To increase precision, the frequencies of several harmonic partials were measured. One should be aware that the transcriptions do not represent perceived tones (Rapoport 2004). Recordings of readings by a young lady
, also of German origin (Avia Schoenleben, AS, from Heidelberg), representing the German intonation of the younger generation, were used in the previous study, leading to very similar results. Comparison between Speech Intonation Contours and Schoenberg's Sprechmelodies In Mahler's words (cf. Bauer-Lechner 1923): ...with me the melody always grows from the words. The words, so to speak, generate the melody.... this is the only way to achieve indissoluble unity of word and note....organic fusion of both elements.
More specifically, Schoenberg's sprechgesang was indeed shown to have developed from the text: the syllable structure of the German language generated the rhythm, and the German speech intonation generated the melodic contour (Rapoport 2004). In the comparison and analysis carried out here (and in Rapoport 2004), Schoenberg's sprechmelodie is divided into melodic segments, comprising a word or a group of words. The contours of these segments are compared with the corresponding segment contours in the transcriptions of LSt intonation. Only the shapes of the contours are considered, and not the specific musical notes or intervals. Various contour shapes were classified: increasing, (or decreasing) melodic progression, upward (or downward) jump, arch--convex melodic segment, or intonation contour (see Rapoport 2004, and the special notation developed therein). Considering Schoenberg's melodic segment contours as originating from the corresponding intonation patterns in German speech (such as LSt), one observes their interrelations as being of several types, either simple similarity, or derived from the corresponding intonation pattern (contour) by one of the following operations: 1. Inversion--meaning only inversion of the contour type: increasing shape inverted to decreasing shape, upward jump inverted downwards, etc. 2. Octave transposition (Stuckenschmidt 1978). 3. Emphasis on last syllable--jump (pitch rise) on last, unstressed, syllable. In German, the stress is usually on the first syllable of the word. This stress is characterized by its long duration and, usually, pitch increase. The word emphasis is used to denote pitch increase or pitch jump. It was observed in the previous study that, in special cases and special words
, in addition to the stress being on the first syllable, the second (usually unstressed syllable
) receives an extra emphasis with a pitch jump. This was encountered in speech intonation (LSt and AS), and reproduced by Schoenberg. Examples: "Die duerre Dir-ne mit langem Hal-se, In seinem Hir-ne, Steckt vie ein Na-gel" (Galgenlied), and "Heil'ge Kreu-ze" (Die Kreuze). Underlined syllables denote the emphasized last syllables. Schoenberg uses this device quite often, even without a counterpart in speech intonation. No. 9 Gebet an Pierrot This melodrama was the first of the cycle that was set to music by Schoenberg. A close look may yield insight into Schoenberg's Music composition
processes and artistic creativity
. Examples 1-3 present comparison between transcriptions of intonation contours in LSt readings and Schoenberg's sprechmelodies. The first phrase, opening the melodrama (lines 1, 2 in the text), is presented in Example 1, and the second phrase (text lines 3, 4) is presented in Example 2. Comparison of the corresponding parts in Examples 1-3 demonstrates the resemblance, down to the finest details, of Schoenberg's score to German speech intonation, and clearly indicates its origins in German intonation. Several particular melodic segments: 1. The opening word "Pierrot" (Example 1). A three-syllable word (Pie-er-rot), its basic intonation contour (LSt) is a jump from the first syllable to the stressed last syllable (F# to B), via an intermediate tone A#. This might also be viewed as a jump from F# to A#, resolved to B.
Example 1 Gebet an Pierrot, first stanza, first phrase. Top: LSt reading (first-sight reading); Bottom: Schoenberg's score Schoenberg kept this three-syllable (three-tone) structure, and modified it by lowering the intermediate tone, as if to enhance the jump. The word appears twice more in the poem: refrain, line 7, and line 13 (see below). 2. "Mein Lachen"--again a 3-syllable entity: the jump in LSt intonation contour (D# to B via intermediate A#) is similarly rendered still larger in Schoenberg's score, by lowering the intermediate tone. In both, the peak is on the second (unstressed) syllable "chen" (emphasized last syllable). 3. "hab ich verlernt": a descending minor second in the score (F# sharp to F) corresponds very closely to the four descending four 1/16 notes (F#-G-F#-E) in LSt reading. The large downward jump (D to low F sharp) in the word "verlernt" (LSt) indicates tension release and an end to the phrase. It is modified by Schoenberg in two steps, starting with a downward jump (F to A) followed by an upward jump to F sharp (octave transposition) indicating an exclamation mark. 4. "Das Bild des Glanzes" (Example 2): Jumps (LSt first reading), and jumps with glissando on "Bild" (F#-B) and on "Glanzes" (D-A#) in LSt second reading, are reduced by Schoenberg to a mere semitone, and a third, respectively. The stress is manifested by long duration.
Example 2 Gebet an Pierrot, first stanza, second phrase. Top: LSt first-sight reading; Middle: LSt second reading; Bottom: Schoenberg's score 5. "Zerfloss, zerfloss" (Example 2): In LSt first reading (Example 2, top), the downward jump in the first word signifies relaxation at the end of the phrase. Upon repeating the word, he seemed to have realized that actually this should have been an exclamation mark: Zerfloss! with an upward pitch jump-- which indeed he does on the second reading (Example 2, middle). The mild ascending pitch in the repeated word corresponds to a question mark. Schoenberg (Example 2, bottom) used an exclamation mark on the first word pair. The downward jump on repetition means tension release and termination. Thus the words are expressed: Zerfloss. Zerfloss? in LSt first-sight reading, Zerfloss! Zerfloss! in his second reading, and Zerfloss! Zerfloss. in Schoenberg's score. Schoenberg perceived the musicality of the German language and German speech intonation in his inner mind, and expressed it through the process of composing. Some further aspects of his artistic creativity, and the various musical transformations: inversion and octave transposition, are demonstrated in Example 3. Considering the refrain, lines 7, 8 (Example 3, bottom), Schoenberg set (the three-syllable) word: "Pierrot" in descending chromatic progression (Example 3, bottom; see also Example 4--fourth line, second column). This is to be compared to the ascending chromatic progression in LSt second reading (Example 3, top; Example 4--third line, first column). Example 3 (middle) demonstrates how, by inversion and octave transposition, Schoenberg's score (Example 3, bottom) can be obtained from LSt intonation (Example 3, top). We note that this is a repetition in both, whereas the first reading and the score's beginning are the first encounter. This is reflected in a smaller jump in the word "Pierrot," and a lower ambitus in the whole phrase. There also seems to be a quasi-tonality in LSt intonation pointing to F sharp major. Returning to the word "Pierrot," it appears three times in the melodrama (in lines 1, 7, and 13), and is rendered by Schoenberg in three different forms. The corresponding varieties in LSt, two readings, bring a total of 9 variations, presented in Example 4.
Example 3 Top: Opening of first stanza in LSt second reading. Middle: Top line transformed with octave transposition and inversion (first bar), and octave transposition (second bar). Bottom: Schoenberg's score of the refrain, lines 7, 8 (end of the poem's second stanza).
Example 4 The word "Pierrot": First column: The word "Pierrot" in the poem's line 1 (beginning). Second column: The word "Pierrot" in the refrain (line 7). Third column: The word "Pierrot" at the poem's end (line 13). Rows in Example 4: Top row: headings. Second row: LSt first-sight reading. Third row: LSt second reading; Fourth row: Schoenberg's score.
No. 10 Raub Examples 5 and 6 present the first stanza of melodrama No. 10, "Raub." Example 5 Raub, first stanza, first and second lines. Top: LSt reading; Bottom: Schoenberg's score Comparison of LSt reading and Schoenberg's setting in Example 5 reveals, in both, two descending progressions in the two phrases. Artistically, Schoenberg's setting has larger ambitus in the first phrase, and is more undulating in the second. It is also noteworthy that, in both, the opening tone is B in the first phrase (except for the 1/16 note-- C#--upbeat in LSt), and G# or (A flat) in the second phrase. The dotted rhythm on the words: Ro-te, Ru-bi-ne, Trop-fen al-ten Ruh-mes reflects the syllable structure: a long, stressed syllable followed by a short unstressed one, as in the melodramas: "Nacht," "Galgenlied," and "Die Kreuze" (Rapoport 2004). In Example 6, in both, the descending chromatic progression in the sub-phrase: "schlummern in den" is changed to an ascending progression on the following "Totenschreinen," terminating in a descending progression on the fourth text line ("drunten in dem Grabgewoelben"), with a local emphasis (small jump) on the syllable "Grab."
Example 6 Raub, first stanza, third and fourth lines. Top: LSt reading; Bottom: Schoenberg's score No. 11 Rote Messe Similarity of melodic segment contours and the corresponding speech intonation contours in the first stanza is demonstrated in Example 7. Example 8 presents the complete first stanza, with LSt intonation (top), and Schoenberg's sprechmelodie (bottom). We notice the emphasis on the 3-syllable word "grausem" (Example 7, column 1), with ascending glissando from B to A, resolved to G# on "sem" (LSt), simplified by Schoenberg by an upward leap (C-E flat), This emphasized last syllable originated in actual speech intonation. Example 7 Rote Messe, first stanza. Demonstration of similarity between melodic segments (groups of words) in LSt intonation (top), and Schoenberg's score (bottom). See text.
Example 8 Rote Messe, first stanza. Top LSt intonation; Bottom Schoenberg's sprechmelodie Noteworthy also is the similarity between the two sequences in "Abendmahle" (Example 7, column 2): G-F#-G# (LSt) and A flat-G-B flat (Schoenberg). An arch in "beim Blendeglanz" (Example 7, column 3) with its apex on "Blen-" (except for the D# upbeat in LSt) is common to both. For a description and the significance of the "arch" in speech intonation and in melodic segment contours (e.g. the phrase: "Ein geschlossnes Zauberbuch" in No. 8 "Nacht"), see Rapoport (2004). A large jump in "beim Flackkerschein" (Example 7, column 4) in LSt becomes a mere semitone rise in Schoenberg, whereas a gradual chromatic increase (E-F-F#) in "der Kerzen" (LSt) is transformed into an E flat-B flat jump by Schoenberg (emphasized last syllable). Example 9 similarly demonstrates the resemblance of corresponding melodic segment contours in LSt intonation and in Schoenberg's score in the second stanza of the melodrama. Noteworthy are: the jump in "Die Hand (Column 1), and three arched contours in columns 2-4, with apexes on "gott" (column 2), "grau-" (column 3), and "Blen" (column 4), respectively. The full second stanza is exhibited in Example 11.
Example 9 Comparison of corresponding melodic segment contours in LSt intonation (top), and in Schoenberg's score (bottom), in the second stanza of the melodrama Of special expressive significance is the sub-phrase "Zerreist die Priesterkleider," shown in Example 10. Table 10
Comparison of melodic segment contours "Zerreisst die Priestterkleider." Top: LSt intonation; Bottom: Schoenberg's score The emphatic upward jump in "zerreist" (LSt) is inverted by Schoenberg downwards, for dramatic effect. The rest of the phrase, characterized by very small ambitus (descending G# to E), is again inverted to a small ambitus ascending progression (C# to E flat), to create a jump on the last, unstressed syllable "der" of Priesterkleider (again, emphasized last syllable). A semitone pitch rise on this word's (stressed) first syllable ("Prie") is common to LSt and Schoenberg; the latter adds another small rise on the following non-stressed second syllable "ter" of the same word (emphasized last syllable). Example 12 compares corresponding melodic segment contours from the third stanza of the melodrama. Noteworthy are: 1. The low ambitus descending chromatic progression in LSt (line 1, column 1), compared to Schoenberg's more dramatic descending progression (line 2, column 1). 2. In column 2: the jump on "sein Herz" (LSt), is rendered smaller by Schoenberg. 3. In column 3, the arch contour in LSt (line 1) with its apex on the syllable "blut" is inverted by Schoenberg to a concave descending contour (line 2).
Example 11 Rote Messe, second stanza. LSt intonation (top); Schoenberg's score (bottom) Example 12 Rote Messe, third stanza. Comparison of LSt intonation (top) with Schoenberg's score (bottom)
Example 13 Rote Messe, third stanza. Top: LSt reading; Bottom: Schoenberg's score No. 12 Galgenlied The close resemblance of Schoenberg's sprechmelodie and German speech intonation was demonstrated in Rapoport (2004), where male (LSt), and female (AS) intonation patterns were presented and discussed. In addition, the concave contour in the opening of the second phrase: "In seinem Hirne," presented in Example 14, is interesting because in both LSt intonation and
Schoenberg's sprechmelodie it spans a whole octave: B flat to B flat (LSt), C# to C# (Schoenberg). This fact, pointed out by Dr. Avi Gilboa, and gratefully acknowledged, might indicate a resemblance in their respective personal voices, as borne out also in other melodramas. In the transcriptions, LSt voice pitch is raised by an octave. It could be suggested here that Schoenberg possibly transcribed his own voice, and raised it by an octave for the female singing-speaking vocal role. Example 14 Galgenlied, beginning of the phrase: "In seinem Hirne." Left LSt intonation; Right Schoenberg's sprechmelodie No. 13 Enthauptung Comparison of corresponding melodic segment contours of the third stanza of the melodrama are presented in Examples 15-18. Example 15 Enthauptung, third stanza. Comparison of melodic segment contours in in LSt intonation (top), and in Schoenberg's score (bottom) Noteworthy in Example 15 are the emphasized last syllable in column 1 (Schoenberg), and column 2 (LSt and Schoenberg), and the descending contour terminating in a jump on "Knie" (Schoenberg), as compared to a mere semitone rise on the same word in LSt (column 4). Example 16 Enthauptung, third stanza. Comparison of melodic segment contours in in LSt intonation (top), and in Schoenberg's score (bottom)
Noteworthy in Example 16 is the relationship (between LSt and Schoenberg) in a downward octave transposition, putting emphasis on the word "zusammen" (column 1), and the apexes on the words "sau-se," and "stra-fend," teminating in a descending contour (column 2). Example 17 Enthauptung, third stanza. Comparison of melodic segment contours in LSt intonation (top), and in Schoenberg's score (bottom) In Example 17, column 1, the jump on the word "Suen-den-hals" followed by a descending progression (LSt and Schoenberg) is worth noting. A jump by means of an upward octave transposition on the word "nie-der" (Schoenberg) puts an extra emphasis on this word, rendering it more interesting. While the descending contour continues from the apex on "Suen-den-hals" to the end of the stanza (LSt), Schoenberg's octave transposition on the word "nieder" continues also to "der Mond," and to "Tuer-ken-schwert" (column 2). Conclusions A detailed examination of melodic segment contours, and comparison between transcribed German speech intonation (LSt) and the corresponding Schoenberg's sprechmelodies elucidates the origin of the latter from the former. Some of the different types of transformations employed in Schoenberg's creative composition have been demonstrated here.
Example 18 Enthauptung, third stanza. Top: LSt intonation; Bottom: Schoenberg's sprechmelodieReferences Bauer
-Lechner, N. Recollections of Gustav Mahler
. English translation
, Dika Newlin, 1923, p. 50. Faber Music. London: Faber & Faber, 1980. Herreweghe, P. Booklet companion to CD Harmonia Mundi HMC 901390, 1992. Rapoport, E. Schoenberg Hartleben's Pierrot Lunaire: Speech Poem Melody Vocal Performance. J. New Music Research 33: 71-111 2004. Simms, B. R. The Atonal Music of Schoenberg 1908-1923. Oxford University
Press, 2000. Stuckenschmidt, H. H. Schoenberg, His Life, World, and Work. New York: Schirmer Books, 1978, p. 354. Acknowledgment The author wishes to thank Prof. Eitan Avitsur for his encouragement, and Mr Morel Koren for his help with various computer facilities. Abstract This work investigates in detail the relationship between speech and sprechgesang in Schoenberg's Pierrot Lunaire, and shows the relations and transformations between the intonational contours of the spoken text
and the corresponding melodic segment contours in Schoenberg's score. It gives additional support to a previous study, in which it was demonstrated that Schoenberg's sprechgesang originated from the syllable structure of Hartleben's German text (rhythm), and from German speech intonation (melodic segment contours). With this present study and the previous one, completion of the analysis of the seven melodramas of Part II of Pierrot Lunaire is achieved.