© copyright Laura E. Thompson, 2011

In order to examine interpretations of the Strauss sonata, it is important to look at the sonata itself as a starting point, to find out what information performers were given by the composer himself. Strauss wrote the violin sonata in 1887-1888 at the age of 23, just after he had completed his first tone poem, Aus Italien.ⁱ The sonata is considered a transitional work for Strauss, in that it follows an outwardly classical structure, though both its harmonic language and heroic melodic writing foreshadow much of his better known tone-poems and operas to come.ⁱⁱ It is also significant in that it is Strauss’s last work of chamber music, and arguably, his most mature take on the genre.ⁱⁱⁱ Following the model of many of his contemporaries and predecessors, the movement is framed into a fast-slow-fast structure of three movements.^iv I will focus my analysis upon the first and second movements of the sonata, the Allegro, ma non troppo, and the Improvisation: Andante cantabile. The first movement is in sonata form, with an exposition, a long development, and a recapitulation. The second movement is often compared to a song without words, and is composed in ternary form (ABA), with gentle beginning and ending sections surrounding a stormy middle. Despite this outward adherence to classical and romantic forms, Strauss uses chromaticism and modulations to distant keys which subtly signal this work’s avant-garde nature and point the way towards his later work.^v The outward appearance of the sonata is of something conforming to Romantic tastes, and indeed, it has been dismissed by many writers about Strauss as “a final fling”^vi which is “by no means uninteresting”^vii and “not considered to be at the pinnacle of violin literature.”^viii But others argue that it is actually a “significant artistic and maturational turning point,” which extended the technical capabilities of the violin.^ix

Prior to the invention of recording technology, scholars had to rely upon the writings of musicians and observers to try to reconstruct the gradual evolution in accepted practices among violinists both as technicians and as interpreters. The advent of recordings helped to create a clearer record of violin playing, but also accelerated the rate of change in which musicians found themselves moving away from the “old style” of violin playing.^x In the early years of the twentieth century, several significant changes were being made among violinists both in terms of their musicality and their technique. One was the introduction of new materials into playing, notably of steel E strings which were thought to project further and to remain in tune more reliably than gut strings.^xi In terms of technique, bow holds and methods of bowing were also evolving to give players more power.^xii As far as interpretation is concerned, the biggest changes came about in the pacing and placement of rubato, tempo choices, the use of portamento and slides, the use of vibrato, and in the demand for greater accuracy in performing notes and rhythms exactly as they are written upon the page.^xiii It is especially this last feature which was brought about by the advent of recording technology.^xiv Once audiences became accustomed to hearing technically perfect recorded performances of classical works (which were often many takes spliced together), the technical standard for live performance was transformed in that audiences expected technically perfect live performances as well.^xv This has benefitted audiences and musicians alike in that it inspires a higher level of technical mastery for performers, but takes away emphasis from the musical ideas and dramatic arcs of the works which previously artists had to strive to convey. As Robert Philip argues, in the current practice which emphasizes technical perfection, the musical content of works is almost an afterthought which is thought to come about through the technical mastery of playing the notes and rhythms accurately. This is of course an oversimplification of current performance practices, but is in line with the trend of trading the excitement and unpredictability of previous eras for interpretations that are, arguably, both technically and musically ‘safe.’^xvi

In this narrative of changes in violin playing and interpretation, Heifetz and Midori serve as key figures in the evolving performance practices throughout the 20^th century. Heifetz was a transitional figure from early twentieth century violin playing to late century practice, in that he was one of the first to adopt a continuous vibrato^xvii and to play with a higher bow arm and tilted wrist on the bow for greater projection, a style of bowing which became known as the Russian bow hold.^xviii But as evidenced by his widespread use of slides and portamento in the Strauss recording of 1934, as well as of his adventurous tempo choices, he adheres a bit more to the old style of playing in his conception of rhythmic flexibility and use of expressive slides. He even employs a form of rubato which has been lost from contemporary performance practice, in which one player takes rubato and the other continues with a steady beat, creating a momentary and purposeful dislocation in rhythm between the players.^xix

By contrast, Midori is a violinist who plays squarely in the modern style. She uses a wide and constant vibrato, and was first discovered as a child prodigy of unusual technical gifts. Though she has made highly regarded recordings, “… her tone as heard in the concert hall is small and it is difficult to avoid the conclusion that she is a creation of the compact disc age. Her interpretations are sound but rather bland.”^xx I find her performance of the Strauss sonata to be very musical, but extremely measured, in that she, in accordance with contemporary performance practices, limits the amount of rubato and portamento she takes, and generally creates long melodic phrases, rather than moving quickly forward and behind impetuously as might have been common earlier in the 20^th century.^xxi

One of the first aspects of each player’s technique that comes to the fore when listening to their recordings of the Strauss sonata is their choices regarding the use of vibrato. Both of them employ a continuous vibrato, which has become standard practice today, but was still a conscious choice in 1934, when Heifetz was recording. Robert Philip writes that in the 19^th century and early twentieth century, scholarly evidence points towards the conclusion that vibrato was not used as a connecting and constant undercurrent in violin playing, but as a tool to demarcate special moments melodically or harmonically. ^xxii In contrast to its current use as a force which rounds out and unifies a performer’s sound to ‘make it sing,’ vibrato was then something that would highlight unique moments. As Leopold Auer (Heifetz’s professor in St. Petersburg who also taught Heifetz his aforementioned Russian bow hold) writes in his book Violin Playing as I Teach It, “The purpose of vibrato… is to lend a more expressive quality to a musical phrase, and even to a single note of a phrase.”^xxiii He goes on to write, “Like the portamento, the vibrato is primarily a means to heighten effect, to embellish, and to beautiful a singing passage or tone. Unfortunately, both singers and players of string instruments frequently abuse this effect just as they do the portamento, and by doing so they have called into being a plague of the most inartistic nature.”^xxiv And so by implementing a constantly undulating vibrato, Heifetz went against the advice of his teacher, but also greatly influenced the direction of modern use of vibrato. Philip writes in the Grove Dictionary of Music that, “Ysaÿe encouraged a trend towards a more liberal (though delicate) use of vibrato, but it was Kreisler, closely followed by Heifetz, who initiated the continuous use of vibrato on the violin, which was echoed in viola and cello playing.”^xxv Midori and Heifetz are relatively alike in their constant use of vibrato throughout the Strauss, though Heifetz tends to use a narrower form of vibrato, especially in the opening passages of the second movement.

One of the most interesting aspects of the interpretations of the Strauss sonata is the way in which the performers interpret rhythm, including choosing tempos and implementing rubato. At the end of the 19^th century and the beginning of the twentieth century, performers were encouraged to vary their tempos to express different moods. They were advised to linger on especially expressive notes and to use faster tempos to express excitement and intensification of the musical content.^xxvi Robert Philip writes, “In any movement containing contrasts of mood and tension, it was general practice in early 20^th century recordings to underline contrasts through big changes of tempo.”^xxvii In general, much less emphasis was placed on accuracy in conveying the musical text in early twentieth century performances and recordings than it is now, and more stress was placed on bringing out the character and dramatic arc of musical works.^xxviii Because performers were not expected to play every note clearly and perfectly in tune, faster maximum tempos could be chosen in the early twentieth century.^xxix Heifetz was a phenomenon for his time because he played remarkably accurately and virtuosically, but he still chose slightly faster tempos in general than Midori. Heifetz chose a brisk tempo for the opening and close of the first movement, of =126. He is remarkably consistent by returning exactly to this tempo, but in the interim, plays expressive and calmer passages at =108-112 (see page two annotated Heifetz score) and returns to =126 for exciting and soaring passages such as that in the last line of page two in the score. In passages marked espressivo, Heifetz interprets this mark as a signal to slow down, or to use a contrast in tempo to bring out the cantabile nature of these phrases, such as at the appassionato ed espressivo on page 14, when he suddenly pulls back to the tempo to =86. But generally speaking, the main tempo is =126 for the main theme and energetic passages, and =108-112 for gentler passages. Midori uses slightly slower tempos throughout the movement than Heifetz and a smaller range of changes in tempo. Her general tempo for the beginning is =116-120 and she keeps most of the piece between = 112-122. There are a few exceptions to this rule, especially at the end, when she is building to a dramatic ending by playing the tempo tranquillo on page 17 at =96 and the tempo primo on page 18 suddenly very fast at =138.

Midori and Heifetz both employ significant rubato in this sonata. Both of them have pianists who play the first measure virtually out of time, with a large breath between the first and second notes of the sonata. Robert Philip writes about the “lost language” of early twentieth century rubato, especially that which relies upon the dislocation of melody from accompaniment.^xxx Heifetz uses this form of rubato at one moment in the second movement of the sonata, on page three, on the third line from the top, in the third measure. Mathematically speaking, the violin part is supposed to enter on the fourth beat of that measure, coinciding with the piano’s last note. However, Heifetz enters one sixteenth note too soon according to the pacing of the underlying piano part. It is possible that this was a mistake or an oversight by Heifetz that has been preserved for posterity by recording technology, but at least equally likely is that it is an appearance of rubato which pulls the violin and piano parts apart, and by coming in early, the piano part can remain steady, while the violin part lingers upon the first note of its entry. To a less obvious extent, this form of rubato continues as the piano plays triplet sixteenth notes at a constant pace while the violin plays with rubato above it. Midori plays this part straight in time.

Another major aspect of both Midori and Heifetz’s interpretations of the Strauss sonata is their respective uses of portamento or expressive slides. In the very beginning of the 20^th century, portamento was widely used at every position change, but was becoming outmoded with the advent of recording technology, and players generally used it more and more sparingly as the century progressed.^xxxi In the first movement of the Strauss, I have examined the portamento tendencies of both Midori and Heifetz. Midori slides at almost every instance of a big jump in register on a slur, such as the last two notes of measure 13 and the first two notes of measure 14. This slide between notes that are separated by the interval of a third or larger is by far Midori’s most common form of slide, and in this movement she uses it 29 times. Heifetz also uses the large interval slide, and employs it 33 times in the first movement, a very slight difference. What is more significant in terms of portamento is his use of other kinds of slides. In terms of slides between notes separated by the interval of a second, Midori uses them only four times during the whole movement but Heifetz employs that kind of slide 14 times. Both Heifetz and Midori also sparingly use a type of slide between a note played on one string to the same note on another string, Midori once and Heifetz twice. Heifetz uses a fourth kind of slide four times in the movement that Midori does not include, which is a slide between two notes that are separated by a large interval, but in which he slides into the second note from the opposite direction as the previous note. This is feasible during string changes and creates an unexpected (and to modern ears, somewhat schmaltzy) effect for listeners. One example of this type of slide can be found on page 2 of the second movement, in the third measure of the second line, between the F (played on the A string) and the B-flat (played on the D string). Though the F is higher is register than the B-flat, Heifetz approaches the B-flat from below for this effect.

Recordings both coincided with major changes in violin playing and influenced them profoundly. The ability to record a performance, and to splice together separate takes into a seamless recording that was perfectly in tune and accurate in terms of reproduction of notes on the page created an expectation among audiences that performances would be of the same standard. While this has benefited performers and audiences alike in that it raised the general level of violin playing in terms of technical accuracy and ability among many players in the wake of recordings, it has taken away from performers’ freedom and ability to improvise rubato, and to take chances in performances. I think that the earlier emphasis on bringing out characters and drama in performance would be a healthy one to add to our current culture of performing things at a high technical standard. It is because of this link to the previous tradition of nearly improvisational rubato and tempo choices, the emphasis on characters and tone colors, and the room for creativity in terms of portamento and expression that I prefer the Heifetz recording to the Midori one, because it straddles ideals of both early twentieth centuries performance practices and those of the latter part of the century (which he helped to establish). His constant use of vibrato sounds beautiful to my modern sensibility, but his use of rubato and portamento have an other-worldly and quaint charm of a past era. But I also appreciate the Midori recording as a great performance, very well played (completely in tune in a live performance), and musically dramatic, despite claims they she is a bit boring as an interpreter (see Grove entry). I find her use of rubato and dynamics well-conceived, and more academically analyzed than those of Heifetz, which are seemingly thrown into the music at a moment’s fancy. But it is that spontaneous quality of the Heifetz recording that gives it a spark of surprise and charm. Both are very strong recordings, and excellent representations of these artists at their best both as individuals and as performers working within the contexts of their own eras.