This article is the first article of a four part series on the Goldberg Variations in which house writer Phoebus Kyriakoudis discusses different interpretations of the work
Legend has it that this work was written to help Count Kaiserling during his sleepless nights; not only is this, of course, the stuff of legend, but the Goldberg variations are far from sleep-inducing. Consisting of its famous Aria, reprised at the end, and 30 variations in between, this work explores virtually all aspects of emotional expression within 30 short variations on a given chord progression. Looking at the discography, one is dumbfounded at the sheer number of recordings of this piece both on the harpsichord and the piano. I aim to take a quick whistle-stop tour of some major recordings made over the years, beginning with the first ever harpsichord recording made by none other than Wanda Landowska in 1933 and even stopping by some of the newest additions by Lang Lang and Trevor Pinnock. This will be a 4-part article series, beginning this week with recordings by Landwoska and Richard Egarr as a quick overview of the realm of harpsichord recordings. I have also chosen to devote the entirety of my next article on the infamous Glenn Gould recording, talking about the legacy it has left on the recording history of this piece, so stick around for that.
Bach’s Goldberg Variations
First published in 1741, this work was dedicated to Johann Gottlieb Goldberg, a harpsichordist and organist who was also allegedly the first performer of the work. The work comprises an Aria, laying out the harmonic structure Bach will spend 30 variations exploring, until the Aria comes back at the end to bookend an Odyssey of a journey; the variations are a veritable kaleidoscope of emotions, ranging from bold and celebratory to pathetic and languishing. Bach specified that the work be performed on a double-manual harpsichord, which makes things much more complicated on a logistical level for pianists having to deal with awkward hand crossings and fingerings on only one keyboard. This could perhaps be why the Goldberg variations have been deemed by many as the hardest keyboard work by Bach. Every third variation is a canon written in an increasing interval every time. Thus, we start with a canon at the unison as variation 3, all the way until a canon at the ninth as variation 27. Normally, one would expect the final variation, 30, to be another canon, this time at the tenth; however, Bach replaces that with a quodlibet, a sort of 18th-century musical medley of different popular tunes played in counterpoint and with humorous undertones. As the variations progress, Bach’s music seems to become more and more unhinged, introducing dissonances and adventurous harmonies that sometimes makes the underlying progression hard to make out. The variation marking, perhaps, the emotional apex of this work, the so-called “black pearl” by Wanda Landowska, is variation 25. Here, it’s the third time we’ve switched from the original key of G major into that of G minor; a dark, mysterious variation, filled with passionate chromatic harmonies that don’t seem to make any sense, and yet work so masterfully together.
Wanda Landowska (1933)
In my previous article on Landowska’s recording of Scarlatti sonatas, I mentioned that she has gone down in history as largely responsible for resurrecting the harpsichord and bringing it once again on the foreground. Her monumental recording of the Goldberg variations is a definite must for an exploration like this, as it situates us perfectly on the initial issues that arose with the performance of Bach’s music, namely, the question of fidelity to the composer’s wishes. Like so many other 18th-century composers, Bach was not famous for supplying his performers with detailed instructions on how exactly to perform his music. One sees most of the time nothing but notes on a page, sometimes even missing a speed indication. Therefore, a lot is left up to the performer to decide. In the case of Landowska, nothing could be described as a better example of a deeply original and personal sound.
It all starts, of course, with the instrument Landowska was performing on. She is accredited with largely researching and exploring the harpsichord mechanism, supporting it as an instrument through her writings and her many concerts on her specially commissioned double-manual instrument by Pleyel. This instrument, with its sonorous 16-foot stop that still sends chills down my spine when I hear it in recordings, cannot however be deemed a faithful reproduction of a surviving instrument by any circumstance; one would say it is more a collection of sounds possible on different instruments at the time, constructing a sort of super-harpsichord.
This is, however, is where a lot of the magic of Landowska’s work lies. It is very clear from this recording that she is attempting to defend the instrument’s capabilities by more than making up for its apparent lack of dynamic variation with a large arsenal of different effects and textures, which she doesn’t miss a chance to test out in the variations. Take, for example, variation 7: a joyful gigue throughout, until the very end, when Landowska decides to reprise the final couple of phrases of the variation to try out a new combination of what sounds like a 4-foot stop and a lute stop. Bach didn’t specify for any such repeats, and in any case, they would have been absolutely out of style for 18th-century standards. Landowska, however, isn’t concerned with that. It is an attempt to reconcile modern audiences with the instrument and deal with the fidelity of archaic recording techniques – thus presenting it at its full power and capabilities. To me, that is a brave and respectable action and one of the many reasons Landowska’s work is praised as ground-breaking nowadays.
Apart from that, her style of playing is very much a straightforward one, with a focus on rhythmic organization and avoiding any musical ornamentations of repeated sections. This can tend to sound quite mechanical, especially in the more virtuosic variations which can almost sound like etudes. However, Landowska’s smooth and connected playing always adds a touch of truthfulness to every bar, making even the most plain-sounding variation something of importance. Overall, Landowska poses a unique interpretation of the work that certainly has left its mark in history.
Richard Egarr (2006)
Jumping almost a century later, and attempting thus to encompass the entire historical evolution of harpsichord recordings of the Goldberg variations, would be a big mistake. Since this, however, is but a whistle-stop tour at various milestones in the recording history of this piece, I shall linger now to the recording made by Richard Egarr in 2006, for he encompasses, I feel, two core beliefs that developed during these 70 or so years, namely, the ever-present question of fidelity to the composer’s vision and that of emotional truthfulness.
The first was characteristically embodied by the historically informed performance movement that gained popularity in many recordings of the 80s. Musicians turned to historical sources and authentic manuscripts in search of clues on how best to recreate the soundscape for which a piece of music was conceived. One big result of this we can immediately hear from the opening notes of this recording: Egarr’s harpsichord is tuned to baroque pitch, a tuning standard that was lower than our modern A=440 Hz, in which Landowska’s instrument was tuned. This creates a more mellow and softer sound which, coupled with Egarr’s overtone-rich harpsichord and smooth playing, create an ethereal, otherworldly sound at times that really adds to the music’s lyrical quality. Gone now are the imposing bourdon sounds and playful lute stops of Landowska’s powerhouse of an instrument. Egarr’s is a faithful reproduction of an authentic harpsichord, and thus maintains an inherent historical authority as a resurrected voice from the past. Finally, we notice that Egarr’s additions to Bach’s music take the form of ornamentations of repeated sections, which always bring an element of anticipation and excitement to find out what musical stylings he’ll think of adding to this variation.
Apart from sounding like a cold remodeling of 18th-century performance, however, Egarr’s account of the Goldberg variations brings a level of humanity and emotional balance that many of these historically informed recordings seem to lack. A good example would be perhaps variations 24 and 25. Egarr always seems to want to take his time with the music, letting the expressive voice it carries unfold in every bar he plays, while his manipulation of rhythm for expressive purposes is simply masterful. Even the fast variations, when they are played fast, are never virtuosic or flashy; the focus is always on a singing line, a concept that frankly a few performers could do well with remembering from time to time about Bach’s music. What remains is again a personal account of Bach’s work, one which is on the one hand rooted in a historically informed approach to music, but which doesn’t forget what it’s trying to interpret: music.
In conclusion, I feel it only mandatory to once again point out that this is by no means an exhaustive account of recordings on the harpsichord of the Goldberg variations. There are countless other recordings out there, each with its merits and its weaknesses, that I encourage each and every one of you to explore for yourselves, for you will always find something new to be gained from a fresh take on this magnificent work. As I previously mentioned, my next article will be about the legacy of Gould’s famous recording, moving from now on to the world of piano recordings.
Phoebus Kyriakoudis, January 2021
Phoebus has also written: