Introduction  |  Listening to Recordings as a Performer |  Recordings of Opus 15 and My Selection  | 

Selected Recordings of Opus 15 and How They Influenced My Perception of the Cycle: 

1. Margot Hinnenberg-Lefèbre - soprano, Helmut Roloff - piano  |  2. Carla Henius - soprano, Aribert Reimann - piano  |  3. Helen Vanni - mezzo-soprano, Glenn Gould - piano  | 

4. Brigitte Fassbaende - mezzo-soprano, Aribert Reimann - piano  |  5. Julie Kaufmann - soprano, Irwin Gage - piano  |  Conclusion 

Introduction

In my regular work as an accompanist prior to the project, listening to recordings to learn music was an extreme measure that I only employed when I was pressed for time and had to learn something new for the next day. Otherwise, I usually did not pay much attention to recordings of works I played. When I listen to other performers, especially early in the learning process, I take away part of the tension that is necessary for “making it work”. Considering the artistic process as a gravitational field that has to be fed continuously,1 I perceive listening to recordings from early on as diminishing the gravitational force, as I, consciously or subconsciously, decide this is how the music should go.

Approximately one and a half years into my project, when I decided to listen to a selection of recordings of Opus 15, I was well acquainted with the cycle, as I had practised it intensively both on my own and together with my first singer. I felt that listening to recordings at this stage could actually help me feed the gravitational field. Now, I had become accustomed to a specific way of playing. I had made certain decisions, perhaps even without being aware of them anymore. Also, I realised I might never have thought of certain other ways of playing this music. I expected the recordings to help me broaden my understanding of the cycle and inspire me to find new ways of playing. As a classical musician, and as an accompanist, in particular, I have to be familiar with different musical languages and ways of playing. Learning to perform a piece of classical music in an unfamiliar style might be compared to learning a foreign dialect of a known language. Getting into it, making it “mine”, requires me to hear others play it. It is necessary to hear them “speaking the dialect” to find my own voice. I assumed the recordings could also make me aware of aspects that other performers struggled with or reveal how others dealt with places that I perceived as challenging, for example, the complex eighth and fifteenth song.

My study of the recordings and a comparison to my own interpretation also helped me better to understand my artistic position in relation to other Lieder pianists. The selected recordings show, of course, only a fraction of my field. Consequently, this study cannot lead to generalisations about the performance of Opus 15 and define and describe its performance practice. As Fabian points out, “[i]t is impossible to know whether an observed element is typical until we have widespread evidence for it. Furthermore, personal traits cannot be distinguished from conventions until a representative sample has been thoroughly investigated.”2 Cook argues similarly that “patterns in performance practice may reflect local contingencies, biographical circumstances, or personal taste. They may reflect micro-narratives rather than the grand narratives so readily invoked in the history of performance.”3 A study of these few recordings can, however, shed some light on the diversity of the work’s performance tradition and contribute to my understanding of my personal artistic position.

Two of the recordings, in particular, had an impact on my playing as they influenced both my understanding of Opus 15 and my perception of how I had played it in the first concert. My way of listening to them also revealed to me some assumptions I had on collaboration in the Lied duo. My study of the recordings became thus a turning point in the artistic process and inspired my further work on the cycle.

In the following section, I discuss first briefly the issue of listening to recordings as a performer, both to those of my own rehearsals and performances and to those by others. Then, I give my reasons for choosing the five recordings I listened to before I describe those features of each recording that I found most interesting from my perspective as a pianist and that influenced me in my understanding of Opus 15.

Listening to Recordings as a Performer

Although the experience of a recording is not the same as that of a live performance as the sound has been altered through the recording process and the visual element and sense of occasion are usually missing,4 I often perceive recordings as performances. When I listen to a recording by another performer, I do so to a certain degree from an outside perspective. Like any other listener, I have not been through the process that led to the recording. However, unlike other listeners, I also listen from the perspective of a pianist. While I do not recognise another performer’s recording as “mine” like I do most of the time I listen to my own recordings, I seem to be able to recreate some of the missing aspects that are lost to other outside listeners. For example, I intuitively imagine some of the bodily sensations of playing what I hear. If it is my own recording, this sensation feels very natural, whereas other performers’ recordings can appear foreign, particularly when their ideas about timing differ considerably from my own. The degree of this kind of bodily immersion into a recording also depends on my familiarity with the music. If I am familiar with the music, I listen to a recording of Lieder to a certain degree in a similar way as I listen to a singer I play with and imagine, for example, how I would react to his or her collaborative cues.

As I wrote above, I often avoid listening to recordings by other performers, particularly at the early stages of the learning process. Although recordings can seemingly help a performer to learn music faster, the simple copying of effects heard in a recording creates inorganic, unnuanced and somewhat stiff performances in my experience. One of my singers, for example, learned her songs primarily aurally. As she only used one recording, she got fixed ideas about tempo and tempo changes and needed a lot of time to adapt to our collaboration. I tried to understand her ideas, but they did not make sense to me as they did not originate in her body and voice.

Apart from situations of dire emergency, I am therefore opposed to this way of listening to recordings as a quick fix. Nevertheless, they can be useful in the collaborative work. Comparing our understanding of different recordings helped my second singer and me to share our artistic ideas with each other. Listening to recordings can also broaden the range of a performer’s interpretive possibilities and guide them by giving a “cultural-aesthetic referential context”5 for canonic works. In the case of Schönberg’s Opus 15, where fewer recordings exist than, for example, of Schubert’s songs, however, the referential function of these recordings is less distinct, and the idiosyncrasies of individual performers might even inflect my understanding of the music in undesirable ways. While I can get new creative impulses from other performers’ recordings, I might also feel pressed to give up ideas that seem too different. Thus, listening to recordings might reduce the originality of my own interpretation. Therefore, I found it important to be conscious of my own way of playing before I listened to others.

While I did not listen to other performer’s recordings of Opus 15 until May 2016, my own recordings, both of rehearsals and performances, had been important from the beginning. I did not record much of my individual practice but almost all rehearsals with the singers. As I had little time to practise together with most of the singers, I depended on these recordings. Particularly at the early stages of a new collaboration or the work on an unfamiliar Lied, they helped me get a feel for the singer’s voice and way of phrasing or remember the approximate tempo we agreed on. However, the danger of practising “with the singer” through a recording lies in creating rather fixed interpretive ideas and rigid timing. Therefore, I avoided too extensive listening after the initial stages, though I still used recordings as an occasional memory aid. Taking an outside perspective rather than being “in the moment” of the rehearsal also helped me discover issues of ensemble coordination or intonation that I had not noticed during our practice when I was too involved in my own part.

Similarly, listening and watching recordings of the performances of the project helped me reflect on what I did. Performances involve complex decisions at the moment, and particularly during the first concert when the music was still new to me, I felt I did not have enough energy and mental capacity to process everything that happened. I found it important to have some distance to the performances before I listened to them, as I did not merely want them to confirm my views and reinforce my ideas the way prolonged practice tends to do. Rather, I saw these recordings as opportunities to scrutinise my playing from a more objective perspective, to compare it to my ideas about it and discover automatisms and idiosyncrasies that disturbed the communication of my ideas. Listening to my first performance was particularly important as it made me search for a more flexible approach to the music and thus determined to a certain degree how I listened to the recordings by other performers.

Recordings of Opus 15 and My Selection

Schönberg’s music has been recorded numerous times since 1922 when Walter Gieseking played the second of the Three Piano Pieces op. 11 on a piano roll.6 In 1994, R. Wayne Shoaf published The Schoenberg discography, which since then has been made accessible as a continuously extendable online catalogue at the website of the Arnold Schönberg Center.7 My overview of recordings of Opus 15 in Appendix B supplements this discography. I discovered more recordings than I expected to find when I started this project, and list, among others, twenty-seven commercial recordings of the complete work spanning from 1953 to 2017. Time constraints and difficulties of procuring some of the recordings made a comprehensive study of them impractical within this project. A whole research project could be devoted to the analysis of recordings of Opus 15, for example with the goal to reveal performance practices or to study the influence of the singer’s voice type or the performers’ native language on the interpretation.

Kerrigan analysed a selection of recordings of Opus 15 from the singer’s perspective.8 She chose them based on their value for the singer and the diversity of the performers’ artistic strengths and abilities.9 When I selected recordings to study, I was interested in encompassing a broad time frame and looked for something that seemed interesting from the viewpoint of a pianist. My selection was biased by expectations and assumptions. Apart from the recording by Brigitte Fassbaender and Aribert Reimann I had not listened to any of them previously.

The oldest recording I chose is from 1956. Margot Hinnenberg-Lefèbre (1901-1981), a German soprano, was well known for her interpretation of Schönberg’s music.10 She had her breakthrough in 1926 when she sang the soprano part of Schönberg’s String Quartet Op. 10 at the introduction of his masterclass at the Berlin Academy. Schönberg was enthusiastic about her interpretation and recommended her for further performances in Germany and abroad.11 She sang it repeatedly with the Kolisch Quartet, also on occasion of Schönberg’s 53^rd birthday in Vienna.12 In 1931, she sang the London premiere of Schönberg’s monodrama “Erwartung” Op. 17.13 In 1932, she married the musicologist, historian and critic of music Hans Heinz Stuckenschmidt, who took music analysis classes with Schönberg from 1931 to 1933 and later wrote extensively about Schönberg and his work.14

Her partner Helmut Roloff (1912-2001) studied piano at the Hochschule für Musik in Berlin (now Berlin University of the Arts) and privately with Wladimir Horbowski.15 He championed Schönberg’s and others’ modern music that was ostracised by the Nazi Regime and worked with the resistance group Rote Kapelle during the Second World War.16 In 1945, he began teaching at the Hochschule für Musik in Berlin, whose director he became in 1970.17

When the recording of Opus 15 was made in the Beethoven-Saal in Hannover on 8 March 1953,18 Hinnenberg-Lefèbre was fifty-one, and Roloff was forty years old. Although Roloff was not primarily known as an exceptional accompanist, the recording is of particular interest to this project because Hinnenberg-Lefèbre sang the cycle together with Schönberg’s preferred accompanist Eduard Steuermannin many European cities.19 She worked intensely with Schönberg, who later addressed her as “liebe Freundin” (dear friend),20 and with whom she studied his early songs, Tove from Gurrelieder and her soloist role in “Von heute auf morgen”.21 I assumed that she shared her vast, first-hand knowledge with Roloff.

The second recording I selected was made seven years later. German singer and music writer Carla Henius (1919 –2002) to whom some sources refer as a soprano22 whereas others call her a mezzo-soprano,23 discovered her love for contemporary music at the Darmstadt International Summer Course for New Music in 1949 and became the favourite interpreter of many of the young post-war composers, among others Pierre Boulez, John Cage and Aribert Reimann.24 Henius sang many world premieres, especially of works by Italian avant-garde composers, and worked very closely with Luigi Nono, who wrote La fabbrica illuminata for her.25

Aribert Reimann (*1936) began to work regularly with well-known singers, among others Dietrich Fischer-Dieskau, from around 195626 and became known to be a sensitive Lieder pianist.27 From 1974 to 1983, he taught Liedbegleitung at Staatliche Hochschule für Musik und darstellende Kunst Hamburg, and from 1983 to 1998 he served as a professor for contemporary Lied, a position that Hochschule der Künste in Berlin (now Berlin University of the Arts) established for him.28 Reimann is also a composer, whose works display clarity and a modern tonal language29 with influences by Wagner, Hindemith, Debussy,30 Berg, Webern and Indian music.31

When the recording was made in July 1963 in Copenhagen,32 Henius was forty-four, and Reimann was twenty-seven years old. I chose it because of Reimann’s reputation as a very sensitive accompanist. In combination with the fourth recording I selected, it also offered a comparison of a pianist’s work with two different singers. I assumed that the juxtaposition of these two recordings that were made twenty years apart might provide better insight into Reimann’s underlying interpretative approach. As Fabian points out, performers rarely change their playing radically: “Musicians tend to lay emphasis on their changing sensibilities and insights when asked to give reasons for re-recording a composition. Yet systematic comparative analyses more often result in identifying similarities than radical differences. […] What transpires, generally, is that subsequent recordings take similar ideas further; the underlying interpretative choices become more obvious with the passing of time, perhaps because the musician is more comfortable or confident about their view of the pieces […] It is plausible that frequent performing of works ingrains certain solutions that are hard to change radically.”33

The third recording I chose was made two years after the one by Henius and Reimann. American mezzo-soprano Helen Vanni (*1924), made her debut at the Metropolitan Opera in 1956.34 She had her recital debut in New York in 196035 and made a career as a concert singer beside her opera performances.36 She was known for her wide range and versatile voice. At the time of the recording, she sang mezzo-soprano roles, but later she also performed soprano parts.37

Glenn Gould (1932-1982) was a Canadian pianist, writer and composer. In 1955, he signed a contract with Columbia Records, for whom he recorded exclusively. He alleged musical and moral objections to the concert medium when he retired from his stage career and devoted himself exclusively to recording in 1964. Gould played a wide variety of classical piano music but mostly omitted early-Romantic and Impressionistic music from his repertoire. Bach and Schönberg’s keyboard music, most of which he recorded, was highly influential on his artistic ideas and his approach to playing the piano.38 Gould is known for very original and personal interpretations with extreme tempi, exaggerated phrasing and peculiar dynamics, as well as for his virtuosity and his intellectual way of playing the piano with an extraordinary ability to convey musical structures.39 Among his mannerisms were his way of sitting on a very low chair in front of the piano, bringing his head far down, and the habit to sing along with his performance even on recordings.40

When they recorded Opus 15 at 30th Street Studio, New York on 10 and 11 June 1965,41 Vanni was forty-one, whereas Gould was thirty-two years old and had just retired from public performance the previous year. I chose the recording because of Gould’s reputation as a Schönberg scholar and remarkable, though idiosyncratic pianist. It has been studied by Kerrigan,42 whose main conclusion was that, although it facilitates an “increased understanding of the intricacies of the piano writing”,43 the interpretation lacks depth as the pianist overshadows the singer. Murray, on the other hand, counts it among the three best American recordings. Especially Gould’s playing “demands the attention of anyone who seeks serious acquaintance with the cycle. Gould’s rhythm makes the difference: the often stark piano lines imply a high tensile strength – throughout, not just at moments of dramatic tension – which Gould uniquely maintains, as much through an exact grasp of where each phrase is going as through his own renowned metrical electricity. (He also understands the importance of a particular gait for each song.)”44

The fourth recording I selected was made twenty years after Reimann’s first recording of the cycle. Brigitte Fassbaender (*1939), a German mezzo-soprano, sang a broad range of operatic repertoire but was primarily known for her roles in operas by Mozart and Strauss.45 Beside her operatic career, she focused increasingly on Lieder and recorded, among others, Schubert’s Winterreise, Die schöne Müllerin and Schwanengesang that are usually sung by men. She is known for her dramatic expressivity and her colourful, idiosyncratic and psychologically imbued way of singing.46

When the recording was published in 1983,47 Fassbaender was around forty-four years old, and Reimann was around forty-seven. I chose this recording for similar reasons as Reimann’s first recording of Opus 15 with Carla Henius. Kerrigan also examined it in her comparative study. She admires Fassbaender’s passionate and sensual interpretation but laments the carelessness in pitching and argues “[o]ne can at best come to the conclusion that in this case the artists decided to interpret Opus 15 dramatically from the viewpoint of the texts, with careful consideration of the ensemble, but with little concern for exact pitching.”48 Murray, who also reviewed this recording, remarks that Reimann “is perceptive about the main drift but crude about detail. The nervy dotted rhythm of ‘Als Neuling trat ich’ declines shabbily into triplets, and the postlude is harsh and self-important.”49 He praises, on the other hand, the “formidable interpretation” by Fassbaender, against whom “everybody else sounds pallid.”50

The fifth recording I chose was made in 1992. American soprano Julie Kaufmann (*1950) studied piano and singing at Iowa University and continued her studies at the Opera Studio in Zürich and with Judith Beckmann in Hamburg, where she also took courses with Aribert Reimann.51 She sang lyric and coloratura roles and had an extensive Lieder repertoire with emphasis on songs from the 20th century.52

Her partner, American pianist Irwin Gage (*1939), has played with renowned singers such as Elly Ameling, Arleen Augér, Dietrich Fischer-Dieskau, Gundula Janowitz, Christa Ludwig, Jessye Norman, Hermann Prey, Lucia Popp, and Peter Schreier since the 1970s.53 Gage is a renowned Lied pianist and got numerous awards for the many recordings he made. In 2001, he got appointed professor at the Musikhochschule in Saarbrücken.54 He is a specialist in the standard lieder repertoire and has done research in less known German and Austrian lieder.55 His interpretations are informed by an intense study of the text and the musical score and the consideration of historical context.56

When the recording was made at the Munich Residenz 27 to 31 July 1992,57 Kaufmann was forty-two and Gage was fifty-two years old. I chose it because of Gage’s high status as a Lied pianist. Like the recordings by Helen Vanni and Brigitte Fassbaender, it has been studied by Kerrigan, who praises the clarity in both artists’ performance but finds it unmoving and lacking in dramatic nuances.58

Selected Recordings of Opus 15 and How They Influenced My Perception of the Cycle

I listened to all five recordings individually in the order they were made, first without and then with the score, and wrote down my impressions in a diary. When I studied the recordings in May and June 2016, I was already very familiar with Opus 15, but I had not yet started practising it for the second concert in September. My impressions of the recordings were not only influenced by my musical experiences and my perception of Opus 15 at that moment in time, but also by what I saw on the cover and what I knew about the performers. In the following part, I point out the features I found most remarkable or inspiring about them when I listened to them in 2016. 

1. Margot Hinnenberg-Lefèbre - soprano, Helmut Roloff - piano59

Figure 1: Deutsche Grammophon 16129, front cover

As Hinnenberg-Lefèbre worked a lot with Schönberg, who seemed to have liked her interpretations, I expected her to be faithful to the score and to sing not very dramatically, guided more by the music than the text. The subordinate status of the pianist on the front cover suggested to me an unequal partnership between the two performers. The text on the LP sleeve, written by the singer’s husband, contains mostly a discussion of the work, its artistic background and importance while revealing little about the performers. Although this review acknowledges that voice and piano are equal partners in Opus 15, it is notable that while photographs of both musicians appear on the back cover, only Hinnenberg-Lefèbre’s achievements are mentioned.

The first song starts rather fast. The tempo appears, however, quite flexible, especially in the piano introduction. Roloff stretches the phrases around the martellato in bars 2 and 3 and almost doubles the tempo in bars 6 and 7 before the singer’s entrance. 

Audio example 1: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 1-7

The two performers appear to be together apart from a few places due to the singer’s habit of cutting long notes, unstressed syllables or rests. Nevertheless, the pianist seems to be a part of a well-rehearsed team rather than giving the impression of just following after the singer. In general, he articulates staccato notes less short than I do, for example, in the left hand in bar 12.

Audio example 2: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 11-12

In the second song, the piano is in the background and has a very soft and slightly dull and unclear sound, more likely due to sound engineering rather than Roloff’s playing. He starts the song with a slow arpeggio. I was fascinated by the two artists being independent yet together. The rhythmically intricate part around bar 7, in particular, appears convincingly synchronised and still very natural, apart from the singer’s pronunciation of the words “schillern” and “trillern”, which sound like [ʃɪln] and [trɪln]. Roloff’s smooth and rather long articulation of staccato notes, for example in bar 6, made me wonder if I over-articulated or tried too much to spell out the details of the score for the listener.

Audio example 3: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. II, bars 4-8

My first impression of the third song was that of intensity. Hinnenberg-Lefèbre conveys tension in the way she “draws” the lines and articulates the dotted rhythms. In this song, the two performers do not seem to play as well together as before. During my second listen, I also got the impression that the pianist does not feel the intensity of the dotted rhythms that permeate the entire song the same way the singer does. The different impressions I got from my first and second listen made me realise how important the singer’s part is for the overall effect of a Lied performance. Roloff rushes in bar 7 despite the indication “etwas breit” (somewhat broad). Although his dynamic range appears limited and he could listen more to the singer, I again admired his clear articulation at places. His dry non legato in the left hand at the beginning made me reconsider my articulation.

Audio example 4: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 1-8

The rather abrupt end of the song made me realise that I tended to think a fermata on the last note. The halting start of the fourth song with the first note almost twice as long as the second leads to the impression of a highly flexible tempo that seems to convey an atmosphere of unease. Again, the singer tends to cut notes and rests. The transition from the second to the third bar is unusually fast. 

Audio example 5: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IV, bars 1-6

Although the balance seems again in favour of the singer, the ensemble is well-coordinated with only a few exceptions, for example, in bars 13 and 14, where the pianist plays on the consonants instead of the vowels. In the fifth song, Roloff’s timing is not quite right for my taste. Sometimes, he plays on or even before the starting consonant of a word, for example on “Pfade” in bar 2. This, however, does not disturb the understanding of the text, as the balance is again much in favour of the singer. It seems as if the pianist tries to push the singer forward. 

Audio example 6: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. V, bars 1-4

The softness inspired me to try how soft it is possible to play the song without losing the images of the text. Roloff’s playing of this song also motivated me to experiment more with the balance in the harmonies, to pay more attention to my left hand, to search for a more varied articulation and to shape my phrases more independently from the singer. The sixth song has a very slow start and seems to lose intensity towards bar 2. Apart from a small rhythmic mistake by the singer in bar 5 (“neue”) and her last entrance in bar 16 (“wann”) the ensemble plays well together, but the performance seems almost too calculated in the slow tempo and lacks the intensity of earlier songs. In bars 6 and 7, Roloff plays each note very clearly and brings out the right hand despite Schönberg’s pianissimo indication. 

Audio example 7: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VI, bars 1-7

The interpretation of the seventh song did in my view not convey the intensity of emotions in the poem, possibly because of the balance, which again is in favour of the singer. The flexible tempo inspired me to work on playing freer, shaping phrases independently from the vocal part, where possible, to bring out the tension between the voices. Roloff never plays “ugly”. Although his articulation varies, his sound is always round and “under” the singer. He does not follow Schönberg’s staccato markings in bar 6 and seems to read bars 7 and 8 in two-four rather than six-eight time. 

Audio example 8: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VII, bars 1-8

The eighth song is extremely fast. It appears very light as the set-up of the recording seems to allow the singer to sing effortlessly without worrying about balance. Therefore, the interpretation does not manage to convey the intense longing I see in this piece despite the fast tempo. Although both performers seem to have the same “drive” and appear to be aware of certain coordination points, a lot of details are not synchronised. Rhythmic details cannot be realised properly in this impossible tempo. The dynamic contrasts of this song, reaching from pianissimo to forte fortissimo, are not fully exploited, but the end is very crisp, clear and loud. Roloff’s playing inspired me to experiment with using less pedal. Although the fast tempo appeared caricatural to me, it made me realise the heaviness of the song in my first performance. 

Audio example 9: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VIII

In the piano introduction of the ninth song, Roloff follows Schönberg’s indications, albeit without sharp dynamic contrasts, yet creates the impression of complete freedom. Throughout the song, I enjoyed both Roloff’s varied treatment of sound and the clarity in his playing and got again inspired to try out a freer phrasing that is independent and yet communicates with the vocal lines.

Audio example 10: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IX, bars 1-11

In the tenth song, the pianist conveys intense longing from the very beginning by stretching phrases and bringing out the different voices in a slow basic tempo. His playing motivated me to listen not only to the upper voice and the chromatically ascending left hand but to bring out more of the middle voices. 

Audio example 11: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. X, bars 1-10

I also liked his way of treating the right hand from bar 16, where he slightly stretches the second triplet.  

Audio example 12: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. X, bars 15-17

In the eleventh song, Roloff plays the beginnings of the two descending figures of sixteenth notes (bars 1, 20-21) very concrete before he gets softer and slower. The tempo is very flexible throughout, and each note seems to have a direction, also at the difficult place from bar 13, where the ensemble is very synchronised nevertheless. 

Audio example 13: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XI, bars 13-16

The twelfth song is slightly faster than what I was used to at the time I listened to the recording for the first time. The forward phrasing gives the song intensity and seems to match Schönberg’s tempo and character indication “Mit bewegtem Ausdruck” (with animated expression). The transition to bar 16 appeared very convincing to me, although Roloff does not follow the molto ritardando indicated by Schönberg and instead accelerates towards the second half of the bar before he reduces the tempo only marginally. The ensemble is not together in bar 20, and the tempo is rather fast in bars 21 and 22 despite Schönberg’s indication “sehr ruhig” (very calm). 

Audio example 14: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bars 14-22

The singer conveys the threatening image of the white sand drinking the lovers’ blood by lengthening the dotted note on “weiße” in bar 24 while at the same time keeping in tempo as indicated by Schönberg. At the beginning of the slow thirteenth song, Roloff plays very clear and piercing accents on the repeated high f. He articulates one and one note in bars 3 to 5. 

Audio example 15: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIII, bars 1-6

I enjoyed Roloff’s little postlude, in which he manages to judge and direct the dynamic progression of all voices very well. The pianist starts the fourteenth song much faster than the singer who prefers an extremely slow tempo. Rather than concentrating on the fleetingness of the poem in general, they convey the different images. Especially the steps of the approaching guards (bar 5, left hand) are very convincing in this tempo. Roloff plays not only very dry as demanded by the score but also often non legato, for example in bars 4 and 5, where he manages to bring out the two different, crossing voices. 

Audio example 16: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIV, bars 1-7

The ensemble is together most of the time, although Roloff seems to prefer a faster tempo. In the last bar, he plays the third beat much earlier than the singer. Although I found the tempo too extreme, the recording inspired me to try playing this song at least slightly slower. Roloff starts the fifteenth song with a forward phrasing that continues throughout the whole song. Sometimes long notes are shortened when they stand alone, almost as if they are not interesting enough and disturb the forward drive, for example, the first note in bar 4. Listening for the first time, I found the singer’s entrance rather fast and free, but closer inspection revealed that Hinnenberg-Lefèbre continued in Roloff’s tempo and sang the exact rhythm, and I had gotten used to my singer shaping this place much slower and freer.

Audio example 17: Hinnenberg-Lefèbre & Roloff (1956). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XV, bars 1-15

The singer’s pronunciation and voice quality and the flexible way of shaping the music seemed at times rather strange yet oddly compelling to my ear, and both performers seem to feel the flexibility similarly. In general, I was surprised and fascinated by the interaction between the singer and the pianist and slightly appaled at the liberties the singer took with regard to the notated rhythms, which did not seem to match Schönberg’s ideal of following the score. Although the piano is often in the background, Roloff appears to be an (almost) equal partner to the singer. Both performers negotiate with each other and seem to work towards a mutual goal. Although they are not completely synchronised in all details, especially when the singer cuts long notes and rests, they agree on the main drift. Their constant, audible negotiation in the details brings out the tension between the voices rather than sounding like they did not practise enough together. Occasionally rather fast tempi, for example in the first and last, but especially in the eighth song were breath-taking but seemed slightly caricatural to me. Roloff occasionally changes articulations from what is written in the score. Their interpretation inspired me to focus more on my own voice and to work on being freer, although I still wanted a better synchronisation than this duo. A very soft fifth song made me consider the possibility to have even bigger dynamic contrasts in the cycle than what I did so far. 

2. Carla Henius - soprano, Aribert Reimann - piano60

Figure 2: BM 30 L 1523, front cover

Similar to the previous record, the sleeve text tells very little about the performers. Both artists’ names appear on the front cover but are far less prominent than the composers’ names. This suggests to me an approach that sees the performers as mediators of the composers’ intentions. On the back cover, Theodor Adorno discusses the work and its background with a focus on stylistic differences and similarities between George’s writing and the music of the Schönberg school. The discussion also has a distinctly educational purpose: Adorno elucidates what the listener should pay attention to.

The piano introduction of the first song had the biggest impact on me, whereas I found the recording little inspiring otherwise. Although it is difficult to judge details of tone colour in this old recording, I started to wonder if Reimann distributed the leaps with two hands. I had previously discarded this solution and did not even remember it until I heard the recording. Reimann starts the introduction in a much more stable tempo than Roloff, though he moves forward in bar 5. The singer enters too early. Her extremely clear enunciation of each syllable made the tempo appear so slow that I wondered if I played the record at the correct speed when I listened to it for the first time, although I realised later that the tempo is not extremely slow. The ritardando in bar 12 is inorganic due to the singer’s breath after “Stimmen.”

Audio example 18: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 1-12

The second song also starts slowly and sounds therefore very laboured. The pianist seems far away in the pianissimo from bar 5. The ensemble is most of the time very synchronised, apart from the difficult coordination at “Vögelreihen”.

Audio example 19: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. II, bars 1-8

The dotted rhythm of the third song seems very broad and uninteresting. Reimann follows Henius well and gives room for her breaths. The wrong note in the left hand in bar 23 is probably only audible to someone who is very familiar with the cycle.

Audio example 20: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 14-23

The fourth song starts very slowly and sounds contemplative and insecure but a little harmless as Henius does not sing the glissandi Schönberg indicated. The tempo increases slightly in the third bar, but remains so slow that the singer has to breathe before “prächtiges Gebiet”, which she stresses on every syllable as does Reimann underneath. The build-up to the climax of the song also seems very laboured.

Audio example 21: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IV, bars 1-11

The fifth song, in contrast, appears rather fast and uniform, as if the performers were mostly occupied with the connection in the one long sentence. Apart from bar 16 (“Schemel”), the ensemble is very synchronised. Reimann starts the sixth song with rather broad chords. He makes another mistake in the right hand of bar 6, where he plays g sharp instead of a sharp. The technically difficult fast jumps between chords in bars 8 and 9 sound laboured. He does not seem to distribute them between the hands.

Audio example 22: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VI, bars 1-11

The seventh song sounds again rather laboured with stresses on almost every note and a slow tempo. Reimann plays the staccato duro under “Sehnen” in bar 6 rather long. Bar 9 appears to be difficult for the singer due to the large leaps. Instead of pushing her, Reimann adjusts his playing to her slower tempo.

Audio example 23: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VII, bars 5-10

The right-hand accents of the eighth song seem to make it difficult for the performers to synchronise the song in all details. The singer’s breath before “seit ich dir” appears to help them in their coordination as Reimann plays the preceding chord quasi fermata.

Audio example 24: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VIII, bars 5-10

The ninth song is slow and intense, but after I listened to Hinnenberg-Lefèbre and Roloff, it appeared rather inflexible to me. Henius pitches well throughout the cycle but sings wrong in bars 14 and 15. Reimann opens the introduction to the tenth song with an almost brutal chord though he still manages to make the first beat of the second bar feel heavy. Otherwise, the slow alla breve creates a sensual atmosphere throughout the song, and the singer conveys the poetic images through the colours of her voice. Reimann seems to feel the very slow start of the eleventh song in eighth notes, although he still shortens some of the longer notes. The singer’s very calm phrasing conveys a feeling of blissful laziness. As the tempo would be too slow for her to sing the final phrase, they speed up and reach almost double tempo in the final three bars.

Audio example 25: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XI, bars 1-10

In the twelfth song, Reimann achieves the “animated expression” Schönberg called for by shortening long notes and rests between phrases in the piano introduction and thus moving forward. He does, however, play with rather uniform dynamics. Although I found his start too fast when I listened to the recording for the first time, it inspired me to consider the underlying anxiety more than the first poetic image of “sacred rest”. The singer’s breath before “unsrer Glieder” appears disturbing and unnecessary. Despite the indication “sehr ruhig”, Reimann plays very fast in bar 22 and the singer continues to move forward where Schönberg indicated “no ritardando” and articulates the dotted rhythms clearly to convey the horror of the poetic image.

Audio example 26: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bars 1-14

Audio example 27: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bars 20-28

The difficult polyrhythms in the thirteenth song are not synchronised. Reimann goes against Schönberg’s staccato marking as he holds the last chord of the phrase in bar 7 almost until the singer has finished singing “bin”, perhaps to give her time to adjust the pitching of her c against the c sharp or to convey the hopelessness expressed in the poem. He does the same in his later recording.

Figure 3: Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIII, bars 6-7

The fourteenth song is rather fast compared to the recording by Hinnenberg-Lefèbre and Roloff. Its shortness and the fast tempo, particular at the beginning, convey the fleeting images of the poem. The performers slow down slightly in bar 4 to bring out the word “zerschellen” (smashing). The singer leads throughout the song, but the ensemble is very synchronised, and the shaping of the entire song appears very thought through.

Audio example 28: Henius & Reimann (1963). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIV, bars 1-7

The piano introduction of the fifteenth song has again few dynamic contrasts. Although the performers are not together in all details, the rhythmically complex parts of the song appear well-rehearsed. Reimann seems to be familiar with Henius’ way of singing and manages, for example, to give room for the breath before “fahle” in bar 29 that appears to be motivated by the singer’s need for air rather than the poetic image.

When I listened to this recording for the first time, I was conflicted about it. My general impression, perhaps influenced by my previous listen to Hinnenberg-Lefèbre and Roloff, was that many songs were oddly stiff. As I tried to find a more communicative way of playing this music at that time, this was something from which I wanted to free myself, and therefore, I found the recording little inspiring.  Although Reimann seemed to be a better accompanist than Roloff and I could hear a lot of detailed work, I missed the freedom of Hinnenberg-Lefèbre and Roloff, which made me wonder how they were able to be together at all. Sound production, rhythms and ensemble coordination appeared very controlled to me, and I found many songs too slow and too declamatory causing the singer to breathe at strange places in the text. However, when I listened with the score afterwards, I found that she sang rather expressively and conveyed the intensity and discomfort that lies in the text. It seemed as if I understood their ideas better when I could follow them in the score but when I listened “from the outside”, some of it did not make sense.

Reimann accompanies well with occasional blunders in the synchronisation of the ensemble and very few wrong notes. The balance seemed better than that of the first recording although also here, I often got the impression of the singer leading. Similar to the record with Brigitte Fassbaender, the accompaniment seemed a little flat and not very exciting to me though it is competently done. Reimann appears to read the score rather precisely but does not exhaust the expressive possibilities.

3. Helen Vanni - mezzo-soprano, Glenn Gould - piano61

Figure 4: M2S 736, front cover

The recording of Opus 15 is part of a set of two LPs that belong to Columbia’s series “The Music of Arnold Schoenberg”. One LP contains Opp. 1, 2, and 15 with three different singers accompanied by Gould, the other Schönberg’s complete solo piano music performed by Gould. The cover art favours the pianist, possibly because of the second LP with solo piano music: Although the composer’s name is printed in a lighter colour and is thus more visible on the black background, both Gould’s and Schönberg’s names are written in capital letters of equal size, in contrast to the singers’ names that only appear in small print at the bottom. The only image is a photo of Gould.

Gould starts the introduction of the first song in a slow, rather stable tempo but moves forward towards the end of it (bars 4-6). He merges the martellato eighth notes with the long notes that follow them. In bars 9 and 10, he brings out the dissonances and follows the chromatic line from d sharp to d. 

Figure 5: Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 8-10

He plays the left-hand figures under “Fabeltiere” very articulated, particularly when they appear in the lower register, and the staccato tenuto under “Strahlen” in bar 14 very short. Rather than playing “flüchtig” in bar 16, he puts a little weight on each note of the sextuplet, at the beginning of which he takes time before he moves forward. When I listened to the recording, I had read Kerrigan’s study of it,62 in which she remarked that “no pianist should play quite as loudly as Glenn Gould and overshadow the singer.”63 I was therefore surprised that I did not experience the balance to be too much in favour of the piano with the only exception in bar 20 when the singer sings her final word and Gould brings out his left hand.

Audio example 29: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 1-17

The broken chord at the beginning of the second song stands very much on the first note in the left hand. The song is very synchronised, and each phrase has a direction. The part from bar 5 that is marked “a little slower” appears faster than the two preceding bars. Gould arpeggiates the chord in bar 7 under “Fischen” and plays it extremely short. Similarly, he plays the sforzati and the following octaves in bars 10 and 11 very short and dry. Although Gould seems to be more guided by the music than the text, it sounds as if the speaker reminds himself not to get lost in his surroundings. In contrast, Gould ends the song very dreamlike with much ritardando. The balance in this song is in favour of the piano, although I find it still acceptable.

Audio example 30: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. II

At the start of the third song, the singer and the pianist are not together. Gould seems to be the one to decide the tempo, and for the first time in the cycle, I get the clear impression that he plays for himself with a very flexible phrasing in the polyphonic voices while the singer has to follow him. It seems his left hand determines the phrasing more than the right hand which doubles the singer’s line. The performers do not move much forward in bar 12, although Schönberg marked the place “steigernd”, which to me suggests an increase in tempo. Also, Gould’s very soft mezzo forte pesante in bar 13 makes the speaker’s plea appear ingratiating rather than intense. His very free shaping of the postlude implies in my view that he actually limited his rubato at the beginning of the song to be somewhat together with the singer. He appears to feel the end in small gestures of half bars and is very free with regard to the notated rhythms, articulation and simultaneousness of notes.

Audio example 31: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 1-8

Audio example 32: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 20-26

Gould arpeggiates particularly many chords in the first five bars of the fourth song. He seems more interested in the phrasing of his left hand in bars 6 to 8 and plays diminuendo instead of the crescendo Schönberg indicated. He does, however, not only play for himself as he, for example, makes room for the singer’s slightly delayed entry in bar 13 after the interlude. His build-up to the climax in bar 18 is intriguing as he plays the right hand very softly and brings out the left hand. The short chord before the singer’s breath appears very dramatic.

Audio example 33: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IV, bars 11-20

In the fifth song, voice and piano seem independent of each other but are synchronised nevertheless. Gould marks the beginning of his own phrases clearly underneath the singer’s phrases. He often disregards Schönberg’s dynamics and plays, for example, mezzo forte instead of pianissimo in bars 9 and 10. I found it intriguing to listen to the very slow end and considered it amazing that the singer had the breath capacity for this slow tempo. Gould starts the sixth song with rather long and heavy chords. I found this song again very inspiring because of Gould’s clear polyphonic playing. He plays the difficult chord jumps very crisp and fast and also articulates the tenuto under “Lohn” as a staccato. From bar 14, when Vanni sings “die in schöner Finsternis”, he brings out the two voices that imitate each other by separating the sixteenth note upbeat instead of playing everything legato. On the other hand, he plays the following figures marked “etwas flüchtig” rather broad and holds the notes either with his fingers or the pedal.

Audio example 34: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VI, bars 8-18

Their interpretation of the seventh song is very slow and does not convey the speaker’s inner turmoil. Instead, Gould’s playing appears almost ridiculous to me. He starts very softly, and the staccato duro under “Sehnen” in bar 6 seems to get a scherzando character as he plays them very short in an accelerating gesture. He also plays the staccati under the singer’s final phrase (bars 14-15) extremely short.

Audio example 35: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VII, bars 1-6

The eighth song is remarkably slow with an average tempo of 65 bpm on the half notes in the first four bars, which is almost half the tempo Schönberg indicated. Again, Gould found his own articulation and plays the third and fourth note of the left hand rather short. The entire song, which appears to be felt in quarter notes rather than in half notes, seems very controlled and does not convey the almost panicky tension I read in the poem. The slow tempo allows the performers to coordinate this difficult song rather easily although they are not exactly together in the somewhat broader sixth bar. The tempo gets extremely slow from bar 9 before they reach an average of 71 bpm in the rather static following part that Gould ends with a ritardando.

Audio example 36: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VIII

Gould brings out the contrasting phrases of the piano introduction of the ninth song through differences in tempo rather than dynamics. The two performers almost ignore the caesura before “Kuss” as they accelerate towards it and take practically no time before they perform it subito piano. Again, Gould tends to break chords that are notated to sound simultaneously, most notably in the accelerando towards the song’s climax. He shapes the final figure of the postlude freely by prolonging the highest note before he accelerates and ends with a ritardando.

Audio example 37: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IX, bars 1-11

The introduction to the tenth song starts rather soft with little weight. Gould prepares the singer’s entry by playing in tempo from bar 10 instead of continuing the ritardando until she enters. He manages again to bring out the different polyphonic voices, for example by articulating the left-hand staccato rather than the notated legato in bar 18. His habit of humming along, although occasionally slightly disturbing, seems very understandable to me. Although I do not hum along during concerts, I often do it when I practise as it helps me to hear and feel connections between tones. Particularly at the end of this song, one can hear how Gould connects larger leaps with a kind of portamento feeling. The eleventh song captivates the listener with its extremely slow tempo and tender softness. It is also well worked out in the ensemble and appears, therefore, one of the most convincing songs of this recording. 

Audio example 38: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XI, bars 1-16

The twelfth song starts rather fast with few dynamic contrasts. The second part of the song that is marked “fließend” (smooth, fluently) seems to begin rather static. Gould plays, perhaps by accident, an A Minor chord in bar 16. In the following bars he brings out the expressive left hand by playing it louder than the right hand and following the tension of the descending chromatic scale. Gould plays c sharp instead of c on the sixth sixteenth note in bar 20, perhaps intentionally to show the relatedness to the end of the bar.

Figure 6: Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bar 20

At the end of the song, the performers do not follow Schönberg’s marking “without ritardando” in the singer’s last phrase, but due to Vanni’s clear articulation of the dotted rhythms and Gould’s staccato in the two first left-hand sixteenth-note figures, they manage to convey the horror of the poetic image nevertheless.

Audio example 39: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bars 16-28

Instead of bringing out the images of defensiveness in the thirteenth song, Vanni and Gould convey the speaker’s sorrow from the beginning. Gould plays, for example, the staccati in bars 3 and 4 legato. The ensemble is not together in the following polyrhythms. The fourteenth song is unusually fast. Gould ties the first two notes of the rising phrase together but articulates the rest of it staccato. The left-hand figures that symbolise the “steps of the destroyers” in bar 5 are played almost simultaneously. Gould clearly distinguishes between the two different voices in bar 7 and continues with a rather abrupt staccato in bar 8. He creates the sfppp effect in the following bar by breaking the chord together with the singer’s grace note and holding only the upper voice.

Figure 7: Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIV, bars 7-8

Audio example 40: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIV

In the fifteenth song, Gould’s extensive use of pedal in the rhythmically complex parts stands out. Due to their slow tempo – Gould seems to interpret the indication “nicht eilen” (do not rush) in bar 19 as molto ritardando – and his dynamic balancing between the different registers, he nevertheless manages to bring out the different voices. In bar 27, when the singer sings “jagen ruckweis”, I find the balance too much in favour of the left hand as the top voice is almost inaudible. Gould changes again the notated score in bar 29, where he plays b flat instead of b double flat in the descending line.

Figure 8: Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XV, bar 29

Audio example 41: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XV, bars 17-29

In the postlude, Gould repeatedly employs the effect of playing a chord strongly and releasing the upper notes immediately, keeping just the bass or the two lowest notes.

Audio example 42: Vanni & Gould (1966). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XV, bars 37-42

I was surprised at how much this recording inspired me. I had not heard many of Gould’s recordings before, as I had been “warned against him” by a former teacher. I agree that it might be confusing for a young piano student to listen to his playing due to his idiosyncrasies. Nevertheless, I found this recording the most enjoyable. Gould creates excitement through his very colourful way of playing that captivates the listener. Although I found the liberties he took with regard to notated articulation and rhythms exaggerated and did not agree with many of his tempi, I enjoyed the flexibility and direction of his phrases and the clarity and colourfulness of his voicing. He uses every aspect of piano technique to bring out polyphony, occasionally even playing chords not exactly together. Although the balance is sometimes too much in favour of the piano and his humming disturbs the perceivability of the singer at some points, only in the third song I got the impression he plays it for himself while the singer is just decoration. Vanni sings convincingly but appears slightly anonymous next to Gould. This recording was perhaps the one that inspired me the most to find a more flexible approach to Opus 15. I realised I should dare to take some liberties from the score and search for ways to bring out the different characters of polyphonic voices instead of stressing the top voice most of the time. I also got inspired to work on extending my dynamic range though I wanted to do so without disturbing the balance of the ensemble.

4. Brigitte Fassbaender - mezzo-soprano, Aribert Reimann - piano64

Figure 9: EMI 1C 067 1466851, front cover

The front cover displays a photograph of the two performers. The composers’ names precede the performers’ names, but both are written in the same size. The same applies to the back side, which lists all songs in order. The inside displays all song texts in German, and a short discussion of the works by music critic Karl Schumann gives background information in German, English and French. There is no further information on the performers or their ideas on the music apart from their voice type and instrument.

Reimann starts the piano introduction of the first song slightly slower than in his earlier recording and brings out the martellato by playing it almost in half tempo. He moves less forward in bar 5 than in his previous recording and separates the octaves in the left hand in bar 6. Kerrigan wrote about this recording that “there is empathy in tempo and interpretation between the pianist and singer”,65 but I find that Reimann mainly appears to serve the singer although he plays rather free introductions and postludes. For example, he seems to wait for her before “Fabeltiere” in bar 13 though he accelerated the fleeting figure before her entry. Having heard Gould’s recording, I found Reimann’s interpretation rather flat. He does not seem to distinguish much between the different voices and plays bar 13 rather unarticulated.

Audio example 43: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 1-15

Reimann plays the opening arpeggio of the second song rather fast but slows down towards the top. Instead of following the line under the word “Blütenwiesen”, Reimann accentuates the half note under “wiesen” before he waits for the singer’s breath. He plays the right-hand accents under “Vögelreihen” very broad and marked, and, possibly as a result of this, the ensemble is not exactly together. The dotted rhythms at the beginning of the third song appear not very articulated as Reimann seems to focus more on the middle voice of the right hand. He separates the last left-hand octave of each bar from the following bar but plays the first octave almost connected to the following fifth. The hairpin dynamics in the small polyphonic gestures from bar 3 seem rather flat although they appear to be played with a feeling of gravity in the opposing phrases. Despite Schönberg’s indication “etwas breit” (somewhat broad), Reimann moves forward in his little interlude in bar 7 which he does not play very loud despite Schönberg’s forte indication. Instead of bringing out the contrasting dynamics in bar 10, he gets softer already in bar 9 when the singer has entered. 

Audio example 44: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 1-10

Reimann supports Fassbaender in the long fortissimo lines at the song’s climax by moving slightly forward. The postlude is interesting as he brings out the middle voices under the melodic top voice which he plays quasi mezzo forte from the start while the singer sings her last word.

Audio example 45: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 21-26

At the beginning of the fourth song, Reimann starts slightly earlier than Fassbaender. The singer’s breath before “und” is necessary due to their extremely slow initial tempo but does not disturb the line. They move slightly forward in bar 3 and accelerate very much in bar 6, where Reimann marks every chord rather than playing an expressive line to bring out the urgent character (“etwas drängend”).

Audio example 46: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IV, bars 1-12

In the fifth song, the pianist is often slightly in front of the singer. While it makes sense to move forward to help the singer with the long phrase “zarte Seidenweben holen”, Fassbaender seems to want more time throughout the song. Perhaps Reimann’s way of playing in front of her was a conscious decision to keep the tempo so that the singer can sing the final phrase “Schemel unter ihrer Sohle” in one breath. Particularly in comparison to Gould, Reimann’s dynamics seem uniform.

Reimann begins the sixth song with very broad and heavy chords and creates the fp effect on the following octave by keeping only the bass note. He plays the difficult jumping chords unclear and rather broad and starts the thirty-second notes before early, so he stays together with the singer. Instead of following Schönberg’s diminuendo indication in bar 12, he plays the second right-hand chord short and accentuated to bring out the held left-hand e. In bar 16, his distribution of the not very soft thirty-two note figures between the hands is clearly audible.

Audio example 47: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VI, bars 8-18

At the beginning of the seventh song, Reimann plays each chord rather marked. He gives room for the small caesuras, but his phrasing sounds rather uniform and does not seem to express the urgency and inner strife of the poem as much as the singer’s dramatic colours. He plays the staccato duro under “Sehnen” very broad. The following slower part sounds as if it is not together as Reimann accentuates the sixteenth notes that go against the singer’s rhythm. 

Audio example 48: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VII, bars 1-8

In the eighth song, Reimann clearly marks the sixteenth note before the first right-hand chord as Schönberg indicated, but continues immediately softer without emphasising the sixteenth notes when the singer enters. Their average tempo is 91 bpm on the half notes in the first four bars and 101 bpm at the end (bars 15-21). Although it is slightly slower than the recording by Hinnenberg-Lefèbre and Roloff, it appears still rather fast, particularly from bar 14, as Reimann does not articulate each note clearly. While the two performers seem to agree on the basic tempo changes, on closer inspection, they are not together at many places, most notably in bar 9, where Reimann, similarly to his earlier recording, waits on the chord before “seit” to coordinate the ensemble. 

Audio example 49: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VIII

The ninth song starts very slow and contemplative. Both performers lead towards “Kuss” almost without crescendo and perform it after a short breath almost pianissimo rather than forte as Schönberg indicated. The tempo remains slow throughout as Reimann plays little accelerando in bars 16 and 20 as indicated in the score. The introduction to the tenth song is very slow. Reimann’s way of phrasing by accelerating and slowing down rather than with dynamics creates a heavy monotony that fits well with the speaker’s waiting. Fassbaender sings very expressively, but after listening to Gould, I find Reimann’s playing colourless, particularly when he plays together with the singer. The piano introductions and postludes seem more interesting. Occasionally, he follows the singer’s phrasing rather than his own. In bar 23, for example, he seems to be more occupied with the idea of “und mild” than with the two voices in his right hand.

Audio example 50: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. X, bars 1-15

Audio example 51: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. X, bars 21-23

Reimann starts the eleventh song again with regular rather clearly articulated sixteenth notes in the left hand but plays slightly faster than in his first recording. In the rhythmically complex part from bar 13, he plays the thirty-two note figures much faster and has to wait for the singer before “schwache”. This way of playing ruins the connection in the left hand although the lively right-hand figures sound interesting.

Audio example 52: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XI, bars 13-16

In the twelfth song, Reimann starts less animated than in his earlier recording though with similarly few dynamic contrasts. His rubato has changed from speeding up and cutting rests to shaping more contemplative phrase endings. Reimann ignores Schönberg’s molto ritardando indications, both in bar 15, where he prepares the following flowing tempo (“fließend”) and in bar 20, where he seems to be occupied with the word “rasch” (quick) in the poem text. Consequently, he is not together with Fassbaender in the following bar that is marked “sehr ruhig”. He plays less wrong notes than in his earlier recording, but the two wrong chords in bar 9 are clearly audible. The performers ignore Schönberg’s indication to perform the end without ritardando.

Audio example 53: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bars 1-16

The most notable feature of the thirteenth song is the very slow tempo of the end, which results from  Fassbaender’s ritardando in bar 10 that Reimann continues in bar 11. He plays the grace note in the last bar very long. Reimann starts the fourteenth song even faster than in his earlier recording. He articulates the second voice in bars 4 and 5 in crisp staccato but holds the last of the “steps” in bar 6 longer than indicated in the score. The polyrhythms in bar 7 and 8 are very fast and seem a little sloppy. Reimann ends the song molto ritardando. 

Audio example 54: Fassbaender & Reimann (1983). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIV

He plays the introduction to the fifteenth song very articulated and dry with rather much rubato, often cutting long notes. The rhythmically complex parts are well worked out between the two performers, although it appears they have “coordination points” between which they move freely. Like in his first recording, Reimann plays the arpeggio in bar 19 before the beat.

When I listened to this recording for the first time, I reacted with mixed feelings like I did when I heard Henius and Reimann. Fassbaender sings very expressively and colours each word or sentence with great care. In the 15th song, she even gets close to Sprechstimme in her treatment of the voice. Although she often sings out of tune, I find her much more convincing than Reimann, but I heard this recording after I listened to the one by Gould to whose voicing and colourful playing it is difficult to be compared. Reimann seems to achieve expressivity more through timing than a big dynamic palette. He plays somewhat soloistic introductions and postludes but mainly seems to serve the singer otherwise. He also tends to move forward and then wait for the singer. The ensemble is together in their conception of overall shapes, but in small details, they are often not coordinated. Reimann seems to be more an accompanist than a truly equal partner in both of his recordings, though perhaps in this recording, he is even more overshadowed by Fassbaender’s very personal interpretation. Therefore, I did not find his playing as inspiring as Gould’s. Occasionally, I got the impression he and I perceive timing in relation to the voice slightly differently as he sometimes plays a little before Fassbaender and at other times just a little late.

5. Julie Kaufmann - soprano, Irwin Gage - piano66

Figure 10: Orfeo C 305931 A, front cover

A section of Gustav Klimt’s Damenbildnis from 1917/18 dominates the front cover of the CD. The composers’ names above it are most prominent, whereas the pianist’s name is written in the smallest font size, though all names appear in equal size on the back cover. The booklet contains an introductory text about the songs’ genesis with observations about musical, textual and psychological aspects, biographical notes for Kaufmann and Gage including portrait photos, and the song texts with English and French translations.

The first song starts with a very exact reading of the rhythms of the piano introduction, in which Gage seems to feel a stable pulse of eighth notes under each note, which results, in my opinion, in a lack of tension in the phrases. The introduction appears therefore rather fast, although it is slightly slower than Schönberg’s metronome marking. Gage almost seems to ignore Schönberg’s martellato and crescendo markings. His straightforward way of playing makes the beginning sound somewhat detached and clinical. The singer enters slightly slower and shapes her phrases in a more flexible manner. Gage seems to feel the left-hand thirty-second notes in bar 13 as grace notes as he plays them very close to the sixteenth notes, possibly because of the challenging jumps. He also uses a lot of pedal. The singer is slightly early on “aus den”, but otherwise the ensemble is together. Gage plays the sextuplet in bar 16 that is marked “flüchtig” rather slow and marked. He plays the final arpeggio almost together and clearly sforzato. 

Audio example 55: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. I, bars 1-17

Gage seems to feel the arpeggio at the start of the second song differently for both hands and plays the left hand faster than the right hand. Both performers start the first rallentando already in bar 3 with the word “Blütenwiesen”. The technically and rhythmically complex part in bars 5 to 7 seems very metronomic, and consequently, part of the polyphony gets lost. Due to the singer’s breath before “Vögelreihen”, the ensemble is not as exactly together as during the rest of the song. Gage plays crisp sforzato arpeggios in bars 10 and 11, followed by marked, long octaves. He is a nuance too late in bar 12. This soft left-hand octave is difficult to coordinate with the second syllable of “Eines”, and in my own experience, it is better to be slightly too late than to cut into the singer’s word.

Audio example 56: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. II, bars 1-8

In the third song, Gage plays the sixteenth note after the second beat rather early, most notably in the two first bars. Although he is not quite together with the singer at “Neuling”, his shaping of the phrases conveys the speaker’s uncertain, ingratiating approach to the beloved. He cuts the dotted note similarly in the postlude, which he plays even freer. Although the singer exhibits an amazing breath control in her long phrases, for example at “mit erbarmender Geduld, den”, her light voice, albeit conveying the speaker’s youth, does not quite capture the urgency of his plea.

Audio example 57: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 1-7

Audio example 58: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. III, bars 20-24

Kaufmann and Gage start the fourth song at a walking pace but slow down slightly towards the second bar, before they move forward already in bar 4 instead of bar 6 that Schönberg marked “etwas drängend”. The performers agree even in small details and are also careful to end their phrases together, for example in bar 20, which they perform with much ritardando. Although the overall shape is worked out well, and Gage is careful not to start the crescendo towards the climax too early, I find the interpretation a little boring and lifeless. The fifth song starts rather slow but moves forward in the long phrase “zarte Seidenweben hole” that can be difficult for a singer to sing on one breath. I get the impression that they share an idea about the shape of the song. Their tempo changes are planned well so that neither has to wait for the other. Gage opens the sixth song with sharp, seemingly rather short chords, though their quality is difficult to judge due to a lot of reverberation. He articulates each note clearly in the polyphonic part that begins in bar 5 but plays the different voices rather uniform. He manages to execute the difficult chord jumps from bar 8 clearly and plays the thirty-second note figures in the end very fast, before he ends with a fortepiano effect on the final octave by releasing the upper note immediately after the marked though not very strong attack.

Audio example 59: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VI, bars 8-18

The seventh song appears rather uninteresting and harmless. Instead of bringing out the dynamic contrasts and differences in articulation, Gage plays rather uniformly and continues for example just as before in the staccato duro under “Sehnen”. First after the following comparatively long caesura, the sound changes completely. Gage accommodates the singer by waiting for her after her breath in bar 11. He plays the eighth song seemingly with a lot of pedal. Particularly bars 6 and 7 sound rather unclear. Gage appears to follow Schönberg’s indication to bring out the sixteenth notes before the chords although it is difficult to follow the details due to the mishmash of sounds. While Kaufmann is slightly in front at the beginning, Gage pushes forward at the end so much that they are clearly not together from bar 18. He does not wait for her like Reimann in bar 9 and is therefore early on the left-hand g octave. 

Audio example 60: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. VIII

The ninth song starts slow and contemplative with little dynamic contrasts. Although he takes time, Gage does not lift the pedal for the caesura before “Kuss”, which he plays subito pianissimo. His diminuendo into nothingness in bar 14 (“die ihn ungenossen”) is impressive. He keeps the pesante feeling from bar 21 almost to the end of the postlude and takes away his last right-hand chord very precisely as he leaves the left hand to sound on.

Audio example 61: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. IX, bars 9-23

Gage plays the piano introduction of the tenth song very fast, broadening it only slightly towards bar 7. Both performers take much time for “weiß und mild” and “wie süße”. The eleventh song starts rather fast and not very soft with even sixteenth notes. Gage’s broadening of the sixteenth notes in bar 2 make the phrase appear very free. His separation of the left-hand notes in bars 5 and 6 creates an interesting effect. He articulates clearly separated, crisp chords in his interjection in bar 10 (“spürten”). In the polyrhythms in bar 13, he plays the thirty-second notes in the right hand clear, even and slightly louder than the left hand. The coordination with the singer works very well. 

Audio example 62: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XI, bars 1-16

The twelfth song starts rather calmly, but moves slightly forward around bar 11, possibly to convey the nestling hands of the poem’s characters. Gage does not observe the molto ritardando in the interlude, although he creates an impression of broadness through dynamics and articulation. He moves very much forward under “Schatten” and plays bar 20, which the ensemble coordinates remarkably well, without pedal. He shapes the following fast figure freely and turns the sixteenth notes on the third beat in bar 22 almost into thirty-second notes. While Kaufmann gets slower in her final phrase despite Schönberg’s indication “ohne ritardando”, Gage conveys the threatening image of the poem through the rubato in his left-hand motives.

Audio example 63: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XII, bars 14-28

Kaufmann and Gage manage to bring out the speaker’s resignation through their very quiet tempo throughout the thirteenth song. Gage articulates the staccato in bars 3 and 4 very shortly and clearly. The following polyrhythms are coordinated extremely well. Gage does not only play no crescendo as indicated by Schönberg but gets even softer towards “rollst”. The very slow tempo and incredible softness of the end convey the speaker’s hopelessness.

Audio example 64: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XIII, bars 1-6

In the rather slow fourteenth song, Gage plays a long and accentuated grace note together with the singer’s consonants in bar 2. Possibly inspired by Gage’s dry and clear playing, the singer enunciates each word very clearly. Gage plays the first half of bar 7 much slower than the second. He plays the beginning of the piano introduction to the fifteenth song with few dynamic contrasts but makes a rather intense crescendo in bar 10. He gives the impulse for Kaufmann’s tempo with the last figure of the introduction. The rhythmically complex part is not well coordinated. Gage rushes at the end of bar 20 and again in bar 23, where he seems to notice his mistake as he corrects himself while Kaufmann sings “morschen”. Both performers seem to get nervous afterwards, as they appear to struggle to stay together. Gage also uses a lot of pedal in bar 24. They are better coordinated in bar 27, where Gage does not distinguish between the staccato and legato articulation Schönberg indicated. He shapes the interlude before the singer’s final phrase freely and moves forward in the first half of bar 32 despite the molto ritardando indication.

Audio example 65: Kaufmann & Gage (1993). Arnold Schönberg: Das Buch der hängenden Gärten, Op. 15, No. XV, bars 19-34

My impression of this recording has perhaps changed the most during the last one and a half years of the project. At the time I listened to it for the first time, I perceived Reimann’s interpretation, for example, to be rather “sloppy” and admired Gage for his technically clean way of playing, the ease with which he seemed to manage challenging rhythms or difficult jumps and the coordination of the ensemble that I perceived to be almost perfect apart from the eighth and fifteenth song. Their timing rarely sounds unnatural or inorganic as everything appears to be well planned. Although I lamented already in 2016 that most songs seemed slightly too smooth, perfect and little intense, I find the recording even less interesting now. It seems that my understanding of the music and my role in the duo has indeed changed. Nevertheless, I still find Gage’s almost perfect nestling to Kaufmann’s voice inspiring and perceive their collaboration as the most balanced of all five recordings.

Conclusion

Listening to these five recordings opened my eyes and ears for interpretive possibilities that otherwise might have been lost in the repetitiveness of the practice process. My understanding of these recordings, among other things, contributed to my more flexible approach to the music. Particularly Gould’s polyphonic playing and Hinnenberg-Lefèbre’s and Roloff’s ensemble inspired me to reconsider some of my (partly unconscious) interpretive decisions. The recordings also made me aware of my views on the music, the text and collaboration at that time. I was very occupied with ensemble coordination and balance and listened with the idea that the pianist is solely responsible for both though I claimed to be an equal partner. Although I disliked Kerrigan’s emphasis of the importance of finding the “correct” tempi and expressions as notated in the score,67 I realised that I was similarly preoccupied with a “correct” reading of the score and clean way of playing with an ideal of technical ease and smoothness similar to what I perceived in Gage’s recording.

While the change in my perception, both of the music and my role as a pianist felt huge to me at the time, I realise that it only gradually and in small nuances translated into my practice and is perhaps not that audible from an external perspective. Nevertheless, I believe that my experiences with these recordings contributed to the process of finding my own voice as a collaborative pianist of Opus 15, of reaching both independence from and accord with the score and the singer as evident in my final performance of the cycle.

Rather than copying specific effects as one of my singers did when she learned her songs by listening to one recording, I integrated the creative impulses I got into my understanding of the work and was thus able to create a final performance together with my second singer that I perceive to be both consistent and original. Our approach differed from the performance practice of Schönberg’s School as we did not seek to distance ourselves from the musical work,68 something which Adorno saw as a requirement of any Schoenbergian artist, and which Kerrigan emphasised in her approach to the cycle.69 Our performance seems to have been guided to a larger degree by the poem texts than that of other artists. I can hear gestural qualities in my playing that I cannot sense in the other recordings I listened to, starting already in the rising left-hand figures in bars 10 and 11 of the first song. Our understanding of the text resulted in a more dramatic, extroverted interpretation of the fourth song and a rendering of the seventh song that brings out the intensity of the speaker’s inner turmoil. I find our interpretation of the eighth song the most convincing as it balances chaos and control and we manage to convey the character changes clearly despite the fast tempo. While Gould and Reimann, in particular, create tension through their very soft and slow interpretation of the eleventh song, my second singer and I tried to create an atmosphere of relaxed sensuality. She and I shaped the cycle out of a shared understanding. We inspired each other as is audible, for example, in her “piano-like” articulation in bar 3 of the thirteenth song. Unlike in my earlier collaborations, in which I often followed the singer although I thought of myself as an equal partner, I seemed to have reached the right balance between giving and taking in the final concert.