THE GOLDBERG VARIATIONS
A PERFORMER’S HISTORY
(or, AN ESOTERIC HISTORY)
In 1955, the musical world was stunned by the release of Glenn Gould’s recording of J.S. Bach’s Goldberg Variations (GV). A recording full of humor, virtuosity and vitality, it became an instant best- seller that still has not gone out of print, and an iconic moment in recording history that went unmatched until his second recording in 1981, shortly before his death.
Gould’s performance came to listeners as a relevation. Familiar perhaps with the stately import of Wanda Landowska, the studied and sober Ralph Kirkpatrick, and the meticulous Rosalyn Tureck, Gould’s exuberance opened new doors. In the 30 years before his 1955 release, there were 10 recordings of the GV. In the 30 years after, there were 26; in the 25 years after that, from the 1980’s up until the present day, there have been more than 60.
We may be forgiven, then, for thinking that Gould “discovered” the GV, since what was niche before him became a rite of passage afterwards. In fact, almost since the time they were written, they have exerted a strong influence on musicians, extending even beyond the keyboard. As Bach’s art in later years became more and more collections of esoterica - obscure combinations of canon and fugal technique, dense elaborations on chorale themes, and so on - so have the GV had an esoteric history since their conception.
Though the GV are mentioned in some histories of music (Hawkins 1776), the earliest description of any detail comes from Forkel’s 1802 biography which set the legend of their performance history in stone: the Russian Count Kaiserling, ambassador to the Saxon Court, commissioned Bach to write a set of variations for his 14 year old house musician Johann Goldberg, to ease the Count’s insomnia. All modern scholars find the story disreputable, and perhaps it coincides with a contemporary feeling towards Bach - ETA Hoffmann described the GV in 1806 as a piece that would chase away an audience.
Still, throughout the 19th century there is ample evidence that the GV exerted influence on musicians of esoteric tendencies. Beethoven’s Diabelli Variations are frequently compared to GV, and though he found inspiration from many sources (including quoting Mozart) it is likely that his 33 Variations are intended to upstage Bach’s 30 (32 movements including the Arias), and in fact Diabelli advertised them that way. There are intimations of the GV in his variations in the Sonata op.109, including an apparent reference to the trills in Bach’s Variation 28. Beethoven’s student Czerny later published an edition for pianists, with restrained dynamics, that is notable mainly for its thorough fingering.
Johannes Brahms, in an 1869 letter, singled out the GV as being instructive of the true art of variation: “When I vary the melody, I can hardly do more than be clever or charming... on top of a given bass, I truly invent the new... I create. Look at Bach’s Variations in G.” In fact, he disowned earlier variation sets he had written on the grounds that they were only variations on a melody. His Variations and Fugue on a theme of Handel, op.24, showed how much he learned from studying the endlessly inventive changes in Bach’s bass line going through the GV.
According to preserved concert programs, Brahms’ rival Liszt played the GV in some fashion on tour between 1838 - 1848. No commentary exists of his performances. A student in 1885 recalls playing them for Liszt, and righteously scolds her colleagues for neglecting them.
It is perhaps around this time that the GV started to spread to the commercial musical sphere, gaining appreciation by lay audiences. Hans Bischoff, a Bach scholar from the late 19th century, made an edition for pianists in the 1880s also with dynamics and articulations, since Bach notated his completely without them - interesting in its date of appearance, but not particularly enlightening. Josef Rheinberger, a composer who wrote mainly for the organ, made an edition for two pianos, later edited by organist Max Reger, that added extra voices and harmony, steeped in an analysis of the way the theme related to the variations; perhaps that reflected mostly that era’s desire for greater volume and density.
In the 1910’s, the GV received two new treatments. The first was for one piano, four hands edition by Karel Eichler (a response to Rheinberger’s that eschewed adding extra voices but still edited dynamics and articulations). This publication implied there was a demand for at-home, amateur musicians. The second, and by far the most interesting and frustrating interpretation, was the 1915 Ferrucio Busoni edition.
Busoni wrote in his preface that he specifically wanted to find a way to adopt the GV to the concert hall (the commercial sphere). By so doing, he advised skipping all the repeats, omitting several variations, re-ordering others to provide a sense of climax, and introducing post-Lisztian piano technique in new arrangements of other variations. Among the variations he suggested omitting are ones that we today deem essential to the architecture and conception of the piece: the canon at the Unison; at the third; the overture (!); and still others. The “trill” variation, number 28, is an example of expanded piano technique. He adds a much wider range, turns single voices into double, moves the trills around the keyboard, and increases the size of the leaps. The da capo aria he rearranged as an organ chorale, entering the world of religious kitsch.
It was in this atmosphere of pianists struggling to make the piece palatable for contemporary audiences that Landowska decided to perform, and in 1933 record, the GV on her “harpsichord” (actually a hybrid harpsichord-piano with a steel frame from Pleyel). Her choice of instrument alone was esoteric, and her interpretation was as original as any edition preceding her, notwithstanding her stated purpose of playing Bach “his way.” For one, her instrument had varying dynamics according to touch, a luxury un-afforded by harpsichords of the 18th century. Also, she appended at the end of a few variations a repeat of the first eight bars of that respective variation, a choice seemingly designed to break up predictability, or poetically to add charming or nostalgic codas.
Landowska’s recordings and enthusiasm for the music and instruments of the past inspired a whole movement of historically aware performers, many of them scholars themselves. Ralph Kirkpatrick at Yale recorded a more “correct” version of the GV on harpsichord, but also made an interesting edition intended for those playing on one manual (ie pianists) showing how difficult hand-crossing passages could be simplified. Rosalyn Tureck recorded the GV on both harpsichord and piano, and her piano version from the 1940s is steeped in the awareness of the historical revival.
That takes us back to Gould’s 1955 version. As we can see, his approach was less a novelty than a revival of a tradition - a long tradition dating back to Forkel in 1802 of treating the GV in an idiosyncratic, personal way, adopting it to fit a private narrative, and musical, approach. He took the GV from the contemporary assumption of “correctness” and realized the full thrust of virtuosity and its adaptability to the modern concert stage – recalling and achieving the goal of the previous generations, Busoni, Rheinberger, Reger, Czerny, et al.
After Gould, the approach to Bach fractured. In ‘Piano Notes,’ Charles Rosen has a provocative insight: “Today, with no universally accepted standard, with neither authentic sound nor the sober academic cantabile able to command allegiance or conviction, performances of Bach on the modern piano range from the wildly eccentric, successful only if the pianist’s sensibility is exceptionally interesting, to a repressed unvarying drive through the piece...”
Today, the pianist has to consider all of the history of thought behind the GV, and find an approach that is both honest to history, yet personal. Hopefully you will leave the concert tonight with a renewed interest in the piece, and in Bach’s words on the original title page, a refreshed spirit.