NEW YORK (NYT NEWS SERVICE).-
I suppose I have some explaining to do to my perplexed fellow musicians, as well as to Glenn Gould devotees. Why? I decided to transcribe Bachs Goldberg Variations for harp.
Im the first to admit that my project a recording comes out Sept. 18 on the record label of Kings College, Cambridge can sound outré or precious. But I come by it honestly: My musical path has been a tad unorthodox. The child of a harpist and a trombonist, I was home-schooled in rural Tennessee to allow for a weekly rotation of lessons on harpsichord, organ and piano, intermingled with youth orchestra and choir practice (and my mother yelling at me from the kitchen about my harp technique).
At 16, I headed to a small British boarding school before studying history at Kings College while serving as organ scholar there. I then returned to the United States to spend two years learning historical performance at Oberlin and then two with the modern harp at Juilliard.
Thats a lot of different repertories, but the Goldberg Variations were one strand of continuity. That continuity also brought some persistent dissatisfaction. When it came to Bach, I was unhappy about the pianos awkwardness with hand crossings, the harpsichords lack of dynamic vitality and the tootiness of organ pipes.
I kept struggling with what my ideal Goldbergs might sound like. I wanted the raw pluckiness of the harpsichord, but with the expressive qualities of the piano. About five years ago, I came to realize that the way to hear this work and most of Bach, for that matter as I wanted would be to use my first instrument, the modern pedal harp.
Thinking that a piece known almost exclusively on keyboard could be transmuted to harp isnt so fanciful: Bach himself appears to have often been agnostic on matters of instrumentation. Like many composers of his time, he was constantly borrowing and rearranging his own compositions. The Double Violin Concerto, composed around 1719, turns up some 20 years later as a concerto for two harpsichords. The Preludio from the Third Violin Partita, written in 1720, reappears in 1731 as a Sinfonia to Cantata 29, rescored for organ obbligato and orchestra. And the Siciliano from the second sonata for viola da gamba is better known as Erbarme dich, from the St. Matthew Passion.
In the 18th century, transcription and arrangement were a means of preservation and dissemination. Bach himself produced solo organ and harpsichord transcriptions of violin and oboe concertos by Vivaldi, Alessandro Marcello and Telemann. His cantata Tilge, Höchster, meine Sünden is a re-orchestration in full of Pergolesis Stabat Mater, using a Lutheran translation of Psalm 51 in place of the original Latin text. Fast forward to the end of the century, and we find transcriptions of fugues from Bachs Well-Tempered Clavier in Mozarts and Beethovens hands, rescored for string ensembles.
Outright ambiguity exists in some of Bachs best-known works. The Art of Fugue and the ricercars from The Musical Offering have no indication as to what forces ought to perform them. There is debate about whether The Well-Tempered Clavier was intended for harpsichord or clavichord. And in the curious case of the Fantasia in G, Bach included a single pedal note outside the playable range of any instrument he had access to, but one that would have been commonplace on larger organs in France.
The musicologist Donald Tovey wrote that Bach wrote on the principle, not that music was written for instruments, but that instruments are made for music. Since World War I, many musicians have showed us other sides of his work by switching up what the pieces are played on. Wanda Landowska was one of the original iconoclasts, making history with the first harpsichord recording of the Goldbergs, after they had been performed solely on the piano for over a century. Stokowskis and Weberns atmospheric re-orchestrations of Bach fugues; Wendy Carlos mind-blowing Brandenburg Concertos on the Moog synthesizer; Chris Thiles fast-as-lighting mandolin treatments of the Violin Partitas: With many of Bachs works, theres now a general recognition that transcription is not only fair game, but even an expectation.
And yet things have been different when it comes to the Goldberg Variations, for which the boundaries of performance remain largely defined by the recordings made by the harpsichordist Gustav Leonhardt and by Gould, who recorded them twice on piano. Perhaps because of the works purity or austerity, depending how you look at it transcriptions of the Goldbergs are usually seen as novelty projects, somehow stepping on the keyboardists turf. While orchestral transcriptions of organ fugues and ricercars have become mainstream, and Busonis piano rendition of the great D Minor Chaconne, originally for solo violin, is considered standard rep, no transcription or adaptation of the Goldbergs has yet stuck.
There are those who prefer to hear the work as they imagine Bach might have done on the harpsichord while others would rather take the variations in on the modern piano, our cultures go-to instrument (like the harpsichord presumably was in Bachs). The results are very different: The harpsichord allows for very distinct articulation and encourages rhythmic flexibility, while the pianos natural suaveness suggests a more straightforward approach.
My solution? Take the piece to the harp. In the opening of Variation 1, keyboardists spend hours trying to amplify instances of implied counterpoint, whereby the left hand jumps around so much that it conceivably represents two voices rather than one. The pianist can differentiate with volume, making the lower notes a little heavier than the smaller notes on top.
The harpsichordist, on the other hand, lengthens the lowest notes as long as possible, to feign some dynamic contrast, while making the upper notes shorter by contrast.
On the harp, one neednt choose. As the instrument has no damping mechanism, the two voices keep sustaining, while creating some harmonic ambiguity.
In Variation 20, a pianist has to figure out how to make multiple voices ring while the hands are fumbling around one another.
While this is perhaps easier on the harpsichord, the lack of dynamics accentuates the natural dryness of the instrument.
But the harp allows for the bass line to sustain while playing other voices, and is less complicated to perform, as the hands approach the strings from opposite directions.
Everywhere in the Goldbergs, the harps lengthy overring allows harmonies to take over, rather than melodies. This is perhaps apt, as the melody heard at the beginning and end of the work never resurfaces. Indeed, one thing that makes the Goldbergs so interesting is that the theme is not in the right hands melody, but in the left hands harmonic pattern. If one looks to the score, the left hand of the opening Aria is composed in style brisé (broken style), indicating the continuity and sustenance of multiple voices, akin to the arpeggiated style typical of performances by plucked instruments like the lute and harp.
The harp isnt perfect. It struggles with intense chromaticism, since the harpist must use his feet in an elaborate pedal mechanism to achieve sharps and flats. And we only play with eight fingers, as the pinkies are too short. As a result, some tempos have to be slower and the aura of the work becomes quieter and more intimate. (This can actually be an advantage.)
The elephant in the room is that Bach never wrote for the harp, and its likely he could not have conceived of an instrument that looked and sounded the way it does. But I never felt I had gotten into Bachs brain until I took the plunge into transcription. To my mind, his music is written for all instruments and none, and the harp is just another instrument as invisible to Bach as his mind is to us. Times distance prevents us from asking the master any questions, so why should we place any restrictions on how and what we inquire of his music?
© 2020 The New York Times Company