A new review based on history, musicology, interpretation and pedagogy of Johann Sebastian Bach’s Goldberg Variations at the piano.
Revised and fingered by Pierre tran
Copyright (c) 2020 Pierre Tran. All rights reserved.
Introduction
J.S. Bach’s Goldberg Variations have been one of the most important works of Western classical music since the 18th century. And were we to limit ourselves solely to keyboard works, then it is undeniable that only Beethoven’s Diabelli Variations enjoy a similar status.
Unfortunately, it is also undeniable that the work has a daunting reputation for being unplayable except by a musician qualified to the highest degree.
Moreover, the published editions are of little help in solving all the problems encountered when learning and performing the work. Bach left a whole layer of unwritten assumptions which are often difficult to grasp. A further interpretative challenge for those playing the work on the piano is that the Goldberg Variations were written for a double-manual harpsichord.
So our first concern in seeking to make the Goldberg Variations available to the widest number is to keep faith with the text, whilst conserving its authenticity within an approach which pairs Johann Sebastian Bach himself and the modern piano. The tools we used were simple, but complex to apply, and the effort required incalculable! While Czerny deserves praise for having made the first transcription of the piece for the piano, we cannot be confident that its original spirit was respected in the process; numerous dynamic and tempo markings distort its true liberalism, and deliberate transparency, which would be more in tune with its original essence.
Ferrucio Busoni undertook an exercise similar to Czerny’s in 1918. But the results are no longer tenable: Busoni did not have access to the original text, i.e. the Urtext, and as a composer he allowed his somewhat wilful temperament to tempt him into going beyond what ought to have been a rigorously philological exercise.
For our part, therefore, although we are reverting to a 19th century Italian tradition of re-reading Bach’s scores (Mugellini, Casella), we do so while seeking to strip away all historical inaccuracies – while at the same time allowing ourselves a certain individual creativity from an interpretative point of view: we have done so in the belief that an enlightened subjectivity cannot be harmful to a musical re-reading – in fact, the contrary applies. In Bach’s case, this approach allows us to get closer to his spirit and to experience more fully the essence of his music.
The tools uses were simple, but their application complex, and the effort required incalculable.
Our research has taken us down three different pathways, initially distinct, but then brought together in a single vision. These are: fingering, phrasing, and touch. Admittedly this last area is more intuitive than explicit.
Once the reader who has undertaken to follow us in this exercise has become accustomed to our methodology and absorbed the results of our work, we feel sure they will emerge from this meditative journey open to new ways of approaching music – and indeed of approaching life itself.
Myth and Historical Truth
The name ‘Goldberg’ associated with these variations belonged to a young harpsichordist known as Johann Gottlieb Goldberg, who served Count Herman Karl von Keyserling (1697-1764), a Russian diplomat based in Dresden.
The magical resonance of the name Goldberg echoes in the mind of many musicians for its close association with the famous and idolised recording of pianist Glenn Gould (see below), which was a catalyst for his rise to success. But we prefer to adhere to the historical truth, which does not correspond to the one the general public accepts.
Indeed, we know little to nothing directly concerning this piece composed by Bach between 1739 and 1740, which he engraved in 1741. It was only in 1802 that Bach’s first biographer Johan Nikolaus Forkel (1749 – 1818) told the story of the variations having been composed to the commission of Count Keyserling, to ease his mind during the nights when he suffered from insomnia. The 14 years old prodigy Goldberg was said to have played them in a room adjacent to the Count’s bedroom.
This story is little more than a persistent fairy tale and does not correspond at all with the truth. We believe that these pieces were destined for Wilhelm Friedemann, Bach’s elder son, who had already been the recipient of other dedications: the ‘Little Keyboard Book for Wilhelm Friedemann Bach’, the ‘Six Trio Sonatas for Organ’, and perhaps even the first book of the ‘Well-Tempered Clavier’ composed in 1722, not forgetting Bach’s ‘Two and Three Part Inventions’, an educational work composed for Wilhelm Friedemann when he was only 13 years old.
In support, it is worth highlighting that nothing in the title given by Bach to the variations (‘Clavierübung IV’, which literally translates as ‘Exercise for harpsichord IV’) provides a link to Goldberg. To this title, Bach added the following enigmatic commentaries: ‘Consisting of an Aria followed by diverse Variations for harpsichord with double manual composed for connoisseurs, for the refreshment of their spirits’ – an enigma which we will investigate later in this text when we look at the structure of the piece.
Since their composition, the variations survived more than two centuries in semi-anonymity until the 1930s. Their existence went rather unnoticed during the 18th and 19th centuries and they first surfaced to fame in 1933, having been considered until then simple exercises for harpsichordists and pianists, which almost nobody except Arrau and Landowska thought of playing in public.
The public success of the variations can be attributed to harpsichordist Wanda Landowska, who was persuaded to champion one of the most beautiful pieces of music ever composed. She gave a complete performance in her property in Saint-Leu-La-Forêt on 14th May 1933, followed by a recording in November of the same year – a recording considered to be the first genuine one, although it is worth mentioning that an earlier soundtrack dates back to 1928, made by Rudolf Serkin on a player piano.
Two remarkable figures at the piano were responsible for the second re-birth of the piece – Glenn Gould and Rosalyn Tureck (though we should note that the first recording of the variations on the piano was made by Eunice Norton and Claudio Arrau in 1942). We find much to challenge in the approaches of Gould and Tureck, these two iconic pianists, but we must acknowledge the fact that each gave the public radically different but equally beautiful versions of the same piece. They enabled us for the first time to glimpse an unsuspected multi-dimensional aspect in the conceptualisation of a work which was previously thought to be an exercise of little interest for the fingers. Gould’s recording was made in 1955, while Tureck’s, first released in 1947 and later in 1957, traces back to her youth. At a later stage, Gould, the recording studios’ favourite, reworked the piece several times: in 1954 at English Radio Canada, in Salzburg in 1959, and especially in 1981, shortly before his death. This latter version is considered by the purists to be the best of all. Much later, Gould distanced himself in a very harsh way from his own 1955 version, in which he claimed no longer to be able to recognise himself. Tureck recorded the Goldberg seven times in total, including a version on the harpsichord.
It would be a pointless task to assemble all the versions ever recorded, for the Variations have become victims of their own popularity, with versions now played on instrumental combinations ranging from string trio, accordion, orchestra, to electronic synthesiser, as well in cinemas, jazz and video games. Among all these arrangements we find the one by Dmitry Sitkovetsky for string trio not lacking in interest. That said, András Schiff, an artist known for his remarkable interpretation of the Goldberg Variations, rejects all transcriptions, finding them each as bad as the next. Schiff states that the respect for the piece’s genuine origin is irrefutable: Johann Sebastian Bach’s Goldberg Variations are a masterpiece for the keyboard alone.
What instruments should one play the Goldberg Variations on?
At first sight it seems possible to assert without excessive risk that the Goldberg Variations are playable on the harpsichord, or the piano, or the organ. But it remains no less true that, on this question, history has offered no universal agreement, in fact it has supplied us with contradictory answers: and this within the larger debate as to whether one should play Bach on the harpsichord or the piano.
Wanda Landowska, who spent much of her musical life as an expatriate in the U.S.A., argued all her life in favour of confining Bach to the harpsichord – so much so that she seems to have traumatised a whole generation of younger musicians (Arrau for instance) who in all humility might have thought to begin playing genuine pieces of Bach on the piano. Arturo Benedetti Michelangeli also always refused to play Bach on the piano, in his case preferring the organ – apart, bizarrely, from the Italian Concerto and Busoni’s transcription of the Chaconne from the D minor Partita for violin. This transcription, existing as it does at some considerable remove from the original, which was written between 1717 and 1720, was nevertheless a cornerstone of Michelangeli’s repertoire. And then it is impossible to avoid thinking about the mimetic language of Glenn Gould’s playing. His gestural repertoire, particularly his idiosyncratic staccato articulation, at once fascinating or yet irritating in its insistence, evokes a piano that conforms to his image of a hypochondriac rather than one bearing any resemblance to a well-regulated modern piano.
But even then: what sort of harpsichord are we talking about? Certainly not the particular Pleyel model employed by Landowska designed according to her own plans, and more probably one of those owned by Bach himself, i.e. a five octave double manual instrument going down to the F below bottom C and which Peter Williams describes as ‘Bach’s intended harpsichord.’ These instruments were indeed rich in orchestral possibilities, allowing an extreme variety of effect.
András Schiff, a notably learned musician, offers a more plausible idea and one more in tune with contemporary thinking. This is the notion that the spirit of Bach is more important than the medium of its transmission, and that there is therefore no reason to forbid the use of the piano as a prime vehicle.
So, from now on, accepting parity between harpsichord and piano can perhaps restore harmony among competing musicologists, but at the same time sets a limit to exploration of the work in an approach that dares to be bolder.
The real question generated in the course of our research is therefore the following: is there a secret buried in the organic structure of each piece which has so far not yet been entirely discovered, one to which access can be gained only at the piano? And if so, what is its nature?
The present undertaking attempts to respond to these two burning questions by means which are both simple and yet powerful, such as innovative fingering, phrasing at once inspirational and revelatory, and a cantabile, never percussive, touch. One further point: the choice of piano for Bach’s Goldberg Variations is also important.
One should put a special care choosing the piano on which he (her) loves to play Bach’s Goldberg Variations. A French music critic recently asked why the work always seems to be recorded on a modern concert Steinway, on which the action is heavier and the hammers too hard, rather than on older instruments such as an Erard or a Pleyel from the early years of the twentieth century, where the action was lighter and the felt on the hammers was softer. A good question! In our research, we used a modern concert grand C. Bechstein model D 280. Buried deep within this particular instrument, many of the lost virtues of pianos from the golden age of the 1930s are still to be found: virtues of transparency and great luminosity.
Framework and symbolism
If we refer to the numerous scholarly publications on the Goldberg Variations (the first in the modern era dating back to 1935 by the American harpsichordist and musicologist Ralph Kirkpatrick), we find it hard to establish strong, undisputed guidelines. The viewpoints of many musicologists differ, and indeed conflict with one another: as was the case between Frederick Neumann and Robert Marshall, who brought their heated exchanges into the public domain. Few of them have had the courage or the openness to admit, publicly or privately, that their analytical efforts had gone too far: at times going well beyond what the composer in his genuine dedication to the music could have conceived.
We have opted to retain only information that seems to us essential, recognising at the same time that the essential is not necessarily where we think it is.
Any successful attempt to realise an architectural structure, no matter how ingenious from a technical or practical standpoint, or how ideal in its geometric proportions, remains meaningless if it fails to enlighten the soul from within. The emotional impact underlying the structure of these variations is evoked in the German terms ‘Gemüt’ and ‘Liebhaber’, which are key to understanding that they are all about soothing the music lovers’ soul: this was Bach’s desire (cf. Myth and Historical truth). In a word, this music is not intended to ‘entertain’ (in the etymological sense) an ‘audience’ but, on the contrary, to nourish the ‘soul’ (in the Pythagorean sense of the word) of ‘music lovers’. Here too we must understand the word ‘amateur’ in its etymological meaning: ‘the one who loves’.
This music is, therefore, of a profoundly mystical nature and reflects through its quasi-mathematical organisation the cosmic order dear to Bach. Professor James Sparks, a distinguished mathematician and an organist devoted to Bach’s music, showed us that these variations are a unique meeting place where the most liberal aesthetics come together with the most compelling logic in a climate of joy and love. Casella had already said something similar in his critical reading of the Well-Tempered Clavier (Book 2), where he reveals that Bach is capable of transforming barren mathematical games, through a kind of catharsis, into sonic jewels of stunning aesthetic beauty.
To enable the necessary structural understanding of these variations, we set out here some essential points, together with useful pedagogical advice. We come to this from the standpoint of a practising musician rather than a musicologist, aware as we are that musicology does not represent the whole of music.
Analytical summary (salient features of the work which we develop in the commentary on individual variations):
1) The theme consisting of eight notes in the bass which can be sung, recognised, and incorporated immediately. From the first note of the theme in the left hand, it can be seen as an Ariadne’s thread guiding us through all the variations (with the one exception of Variation 9): the unifying element that corresponds to the person we can always count on through the ups and downs of life. This underlying melodic line (used also by Handel in his Suite number 9 in G major) needs underlining, without heaviness, wherever it occurs. Some performers forget this elementary principle, to the extent that music critics have been obliged to draw attention to the lapse when reviewing their otherwise respectable recordings.
2) The number 3: Bach builds a triangular structure, the base of which contains 10 virtuoso pieces, the left side 10 genre pieces, and the right side 10 canons. Bach applies this to create cycles, each of three successive variations: a genre piece, a virtuoso piece and then a canon, continuing in the same pattern until the end (with the exception of Variation 30).
3) The division of the work into two equal parts. The first part ends with a canon at the fifth in G minor, whose metaphysical character plumbs the poignancy of life. The second part is introduced by a majestic Ouverture à la française; which naturally gives rise to the issue of whether a break should be made in a public performance: should there be a longer than usual pause after Variation 15, as for example at the end of a sonata movement, or should everything be played in one go?
4) The organising role of the number 2: the theme of the Aria is divided into two halves of sixteen bars each.
5) The catalogue of diverse musical genres which gives the work its colours, i.e. its inner life. Without offence to Bach, we have gathered support for a better understanding of the specific character of almost each piece from the titles given to individual variations by the Swiss harpsichordist and conductor Jörg Ewald Dähler. His list has been added to the explanatory page that accompanies each variation.
6) The need to make contact via a meditative thought and an inner listening process, in relation to the specifics of each variation. This must be carried out variation by variation, half by half, phrase by phrase, sometimes bar by bar. The architecture of the whole work as it imposes itself on us through the grace of a latent organisation (biological if not cosmic), is utterly unattainable by intellectual speculation.
7) Deep understanding of the subtleties of each variation through playing when time is suspended, so that it becomes natural and apparently effortless.
8) Differing types of touch that occur naturally during the stages of the assimilation work described above. These range from legatissimo to staccato, and must include a transcendent legato. Obviously at all times a non-percussive touch is required, even in forte playing – fortissimo being very rare in Bach. However, unlike Czerny and Busoni (cf. Introduction), we have avoided all notations relating to dynamics. A general contrast between piano and forte is undoubtedly necessary, but we have placed a few accents here and there with great care. Here an anecdote comes to mind: Bach advised his students to practice his works on a clavichord, the precursor of the pianoforte, whenever possible. That means he had foreknowledge of a pianistic approach, not an exclusive awareness of the harpsichord or of the organ.
9) Song and dance as the two main pillars, and there we know that Bach considered all his work to be ‘cantabile’. As regards the use of the pedal to enhance a singing tone, we leave the choice (as for tempi), to the individual performer, being reluctant to enter into a partisan dispute between those in favour (Perahia) and those fiercely opposed (Schiff).
In conclusion, we believe that the dialectical relationship between our musical phrasing and the fingering (both of which are unique and experimental, and which we have implemented with great care, self-criticism, love and musical intuition) is the keystone of all our work on this critical revision of the Goldberg Variations by Johann Sebastian Bach. If this work ‘speaks’ to our readers, our goal will be achieved; we will then share with them the unspeakable joy brought about by so many aesthetic and spiritual wonders, which will always remain beyond the reach of our intellectual understanding. For ‘music talks only to the heart’.
We want to let them know that our indications are not restrictive: one can and must put in best, so that this Music is fully accomplished under their fingers.
