The string trio has in some ways a stronger claim than its more popular sibling, the string quartet, to be considered the summit of modern string chamber music. The Baroque trio sonata is a different beast, with or without continuo. However, during the early Classical period, two instrumental formations vied for favour, that of two violins and cello and that of violin, viola and cello. Haydn may have been the first to opt for the latter combination, which, a few exceptions notwithstanding, would eventually emerge victorious. Difficulties of balance and the restrictions of a three-part texture –mitigated from time to time by double-stopping – rendered the string trio a rarer beast and in many respects a greater combinational challenge. Mozart’s six-movement Divertimento in E flat major K. 563 remains a towering example; none should be fooled by the ‘light’ suggestion of ‘divertimento’. So does Schoenberg’s late String Trio, written in the context of a 20th-century revival of interest. Reger, greatly admired by Schoenberg, had written two, as had Hindemith. Webern had also pipped his teacher to the 12-note post by a couple of decades.
Developing Variation and a Reckoning with Memory
Responding to a commission from Harvard University’s music department for a 1947 symposium on musical criticism, Schoenberg’s contribution stood alongside works (for different forces) by Copland, Hindemith, Malipiero and Martinů – quite a symposium! One of Schoenberg’s fellow Californian exiles, Thomas Mann, recounts:
He told me about the new Trio he had just completed and about the experiences he had secretly woven into the composition – experiences of which the work was a kind of fruit. He had, he said, represented his illness and medical treatment in the music, including even the male nurses and all the other oddities [!] of American hospitals.
Schoenberg’s pupil and assistant, Leonard Stein, went further, telling Walter Bailey that Schoenberg ‘explained the many juxtapositions of unlike material within the Trio as reflections of the delirium which the composer suffered during parts of his illness’. ‘Thus, the seemingly fragmentary nature of the Trio’s material represents the experience of time and events as perceived from a semiconscious or highly sedated state. These unusual juxtapositions also represent […] the alternate phases of “pain and suffering” and “peace and repose” that Schoenberg experienced.’ We may or may not choose to follow quasi-programmatic cues and should remind ourselves that Schoenberg had conceived the basic outline of the work before his heart attack on 2 August 1946. Nevertheless, most of the actual composition took place in its aftermath, between 20 August and 23 September.
Such cues may at least offer a way in to what Michael Cherlin has described as music that ‘is full of abrupt and striking changes of texture and affect as musical ideas are broken off, interrupted by other ideas that are themselves interrupted’. Schoenberg’s method certainly extends beyond the programmatic; its concern with memory, alternating harsh dissonances and something reminiscent of, if not quite amounting to, tonality, may be understood in purely musical terms if we wish. The role of memory, as Cherlin has argued, relates not only to Schoenberg’s illness but also to one of his most esteemed precursors in this genre: Beethoven (albeit early, all four of his trios written and published prior to the turn of the 19th century). And yet, even here, biography intrudes, for if Schoenberg’s transformation of the fractures of late Beethovenian rhetoric are crucial both to formal outline and content – to adapt that slightly to Schoenberg’s own terms, to style and idea – then ‘for Schoenberg’, again quoting Cherlin, ‘Beethoven, and specifically late Beethoven, would have been nearly synonymous with the activity of deathbed composition’. Memory plays tricks, sometimes deliberately, mediating between youth and old age, then and now.
Waltzes from old Vienna, or rather fragments of such ‘remembered’ waltzes, likewise play a mediating role between the opposing forces of surface contrasts and unifying 12-note technique. That is a practice – or a way of understanding it – far from unique to this Schoenberg work; we might say much the same about Moses und Aron. The extremity of those contrasts remains, however, a crucial, almost neoexpressionist, aspect of the work. Indeed, one of the most intriguing characteristics of its form is its move away from the neoclassicism (broadly understood) that had characterized much of Schoenberg’s 12-note instrumental writing, towards the practice, familiar from Liszt’s Piano Sonata in B minor and Schoenberg’s own First String Quartet and First Chamber Symphony, of condensing four traditional sonata movements into a single sonata-form movement.
Here, that is perhaps more difficult to discern, at least on first hearing, partly because Schoenberg’s division of the work into three parts offers another equally important standpoint, related yet not identical. Its arch-like quality stands close to a favoured structure of Bartók, albeit without the symmetry, and, as was customary for Schoenberg, with the ‘developing variation’ he treasured in his ‘Brahms the Progressive’ at its heart (however much under attack). Three parts are interspersed with a first and second episode, those alternate phases of ‘pain and suffering’ and ‘peace and repose’ both characterizing and undermining the different sections’ identity. The complexity of the first bar alone, replete with tremolandos, harmonics, syncopations, sforzandos, and so on, sets the aural stage very well indeed, as well as any mere identification of the series. To the question as to whether the third part constitutes the recapitulation, further development, reconciliation or dissolution, the only possible answer can be: all of the above and more.
Structuring of Variation and the Snares of Biography
Bach’s ‘Goldberg’ Variations have received a number of transcriptions and arrangements. Some have opted for supposedly Baroque forces and formations. Others have taken advantage of any number of modern possibilities, ranging from one of the earliest in 1883, by Josef Rheinberger (revised by Reger), for two pianos, to the 1938 orchestral version by Polish dodecaphonist Józef Koffler. More recent contributions have included two for string trio, from Dmitry Sitkovetsky (1984) and Federico Sarudiansky (2010). Sitkovetsky’s version came at a time of peak ‘authenticity’, when certain voices would frown upon any such reimagination. Following his 2009 revision, he recalled:
When I first wrote my transcription of Bach’s ‘Goldberg’ Variations for string trio, in 1984, it was both a labour of love and an obsession with the 1981 Glenn Gould recording. […] Generally, at that time, transcriptions were out of fashion and I recall that my own colleagues and managers were sceptical about such an audacious idea. […] Since then, my transcriptions have been played all over the world and moreover they have opened the floodgates of new interpretive possibilities for the piece, which have included solo harp, wind instruments of all kinds, saxophone quartets, Renaissance viols and even a fascinating concoction of Uri Caine, among many others.
The players of the Trio Zimmermann have decided to join that merry throng, offering an alternative to Sitkovetsky. Here they give a brief explanation:
As a string trio, we were originally exposed to Johann Sebastian’s Goldberg Variations, originally written for harpsichord with two manuals, through Dmitri Sitkovetsky’s transcription, which made it possible to approach this masterwork on other instruments. When the quite infinite world of the Goldberg Variations began to open up for us, we found ourselves captivated by the original text, with all its countless treasures and details. This intensive study moved us to develop a version for string trio, which, within the spectrum of possibilities available, represents neither an arrangement nor a transcription, but should serve the sole purpose of disclosing Bach’s score and the genius of his composition for harpsichord.
At the time of this ‘disclosure’, there remain sharp divisions amongst those who perform and listen to Bach’s music, although the climate is perhaps less polarized than once it was. Curiously, transcriptions now often fare better than ‘straighter’ performances on modern instruments – a ‘symphonic’ performance of the St Matthew Passion will swiftly be anathematized. What almost anyone will now agree on, however, is that, even if they does not like the way the music is being performed, the musical qualities of the work remain clear. The Trio Zimmermann offer yet another standpoint: something surely to be welcomed both in theory and in practice.
Ways to think about the Variations in question are at least as inexhaustible as ways to perform it; ideally, the two would enter into some sort of dialogue. The celebrated tale of the work’s origins, whether true, false or somewhere in between, will never leave the stage; why should it? Johann Nikolaus Forkel, Bach’s first biographer, tells the classic version:
We have to thank the instigation of the former Russian ambassador to the Electoral Court of Saxony, Count Kaiserling, who often stopped in Leipzig and brought there with him the aforementioned [Johann Gottlieb] Goldberg, in order to have him given musical instruction by Bach. The Count was often ill and had sleepless nights. […] Once the Count mentioned in Bach’s presence that he would like to have some clavier pieces for Goldberg, which should be of such a smooth and somewhat lively character that he might be a little cheered up by them in his sleepless night. Bach thought himself best able to fulfil this wish by means of variations, the writing of which he had until then considered an ungrateful task on account of the repeatedly similar harmonic foundation.
There is less to detain us interpretatively than in the case of Schoenberg’s delirium. This Aria and diverse variations – 30 of them, rounded off with a return to the Aria, the same sarabande, yet quite transformed by what has unfolded in the meantime – may certainly be understood without any recourse to imagining a sleepless count, unless we particularly wish to do so.
Bach’s structuring of the variations, each more concerned with the Aria’s bass line and harmony than with its melody, is as sophisticated as we might expect. Every third variation is canonical, the first at the unison, the second (Variation 6) at the second, and so on, until Variation 27 reaches the ninth. Variation 30 is not a canon but a quodlibet, confounding expectations with a humorous combination of favourite Bach family folksongs. While we may not find the intervention of ‘Cabbage and turnips have driven me away’ amusing, as was possibly intended, we can certainly admire Bach’s contrapuntal mastery.
There is much more, of course, to the internal structure, of which barely the surface can be scratched here. Ralph Kirkpatrick’s study of the work showed that genre pieces immediately follow the canons: Variation 4 is a passepied, Variation 7 a gigue, and so forth, with Variation 16 a full-blooded French overture (opening, as it were, the second half and offering further symmetry to the work as a whole). The variations following next but one after the canons (from No. 5 onwards) are characterized by Kirkpatrick as ‘arabesques’, swift flights of dazzling virtuosity of one variety or another. Crossing of hands, for instance, will necessitate a different strategy with a string trio, as we shall discover. Whatever the musical hardware, though, there is a special quality to the G minor variations (Nos. 15, 21 and 25), the final one unforgettably christened the ‘black pearl’ by Wanda Landowska – much to the chagrin of prosaic ‘authenticists’. In its chromaticism we might hear just a little of what led Schoenberg to declare Bach the first 12-note composer.
(This essay was first published as a programme note for the 2017 Salzburg Festival.)