The so-called Goldberg-Variationen by Johann Sebastian Bach constitute one of the absolute masterpieces of keyboard literature, one of the summits of keyboard technique, and a challenge for all performers tackling it as well as for all listeners venturing through this cycle for the first time.
Throughout the eighteenth and most of the nineteenth century, these Variations were considered as entirely unsuited for the concert hall, and potentially unplayable. In fact, their length and size were deemed to be disproportionate for the needs of the average concert-goer, and from the technical viewpoint they presented almost unsurmountable problems. True, piano technique had noticeably evolved from Bach’s times to those of Liszt or Busoni; however, Bach had conceived his work for a harpsichord with two manuals (i.e. two keyboards). While hand-crossings are comparatively common in piano technique, those resulting from the “translation” of Bach’s original to the piano cause very awkward situations: the two hands end up touching the same keys or keys very close to each other, and some passages are nearly impossible to play in a satisfactory fashion.
However, the movement known as Bach-Renaissance, which aimed at reviving Bach’s heritage and at bringing it on the concert stage, considered the Goldberg Variations as a masterpiece which could not be ignored or left silent. In the late nineteenth century, the idea of playing them on the instrument for which they had been designed, i.e. a two-manual harpsichord, was considered absurd; thus, some of the keenest promoters of Bach’s music started to imagine alternative possibilities for presenting them to the audience.
Ferruccio Busoni created an edition/arrangement for the piano which has been recorded in a Da Vinci Classics album, and in which the technically “unsolvable” passages were thoroughly rewritten (in same cases simply by dividing in a more rational fashion Bach’s notes between the pianist’s two hands, in other cases by re-scoring the passages, e.g. moving one or both hands an octave lower and higher). However, Busoni was also concerned by the imposing length of the cycle (even though his own solo recitals could last up to four hours!), and therefore suggested performing only twenty of Bach’s original thirty Variations, for the sake of brevity and concision. He also rewrote the conclusion, revealing the harmonic structure of the Aria da capo and presenting it in the form of a solemn Chorale.
Busoni’s ideas are questionable from the viewpoint of today’s concept of authenticity, but are motivated by his passionate love for Bach’s music; moreover, they offer a singular and personal perspective on the Variations, which may be (and surely is) very partial, but is also consistent and musically convincing.
Busoni’s version dated from as late as 1915: little more than a century ago, the Goldberg Variations were still considered as in need of some kind of “taming” before being performed in public. The version recorded here is one realized by Joseph Gabriel Rheinberger, which had been published already in 1883, and republished in the same year as Busoni’s version in a revised edition realized by Max Reger, who substantially modified many of Rheinberger’s dynamics and articulations.
In the short preface to the published version, Rheinberger stated the reasons for his choice, underpinning that the Variations are among the “most important” keyboard works by Bach. At his time, Rheinberg continues, “this impressive work is prized and played only theoretically” (i.e. it is studied by piano and composition professionals, but is not publicly performed). For him, the reason for this was mainly to be found in the technical issues posed by the transposition from two to one manual. Thus, Rheinberger had realized this “arrangement full of piety” (i.e. of respect) in order to let “musicians and music lovers” know “this treasure of true Hausmusik”. It is important to point out that Rheinberger was not yet envisaging the possibility of playing his version in concert, but rather he was proposing it for the enjoyment of pianists and of a limited audience within the framework of so-called Hausmusizieren, “home music-making”.
If the main reason behind Rheinberger’s operation was a very practical one – avoiding the knotted fingers which might result from performance of Bach’s original – of course it would have been both improper and disappointing to simply divide the Variations scored for two manuals into two parts to be played by two pianists (one playing right hand and one playing left hand), and to leave the remainder unchanged.
In the opening Aria, Rheinberger not only added nothing, but rather lightened Bach’s original by depriving it of many embellishments. The two pianos never play together, but are assigned each a half of the two parts in which the Aria is divided, thus pointing out the symmetrical and binary structure of the piece. What Rheinberger did add, throughout the cycle, is a wealth of performance marks, including articulation, dynamics (i.e. volume) and agogic (i.e. rubato), in conformity with the taste of the era.
Already in Variation I, however, the transcriber’s touch is more clearly observable. Here the part of Piano I is very similar to Bach’s original (again, with markings added), while Piano II demonstrates the dependence of this Variations’ harmonic scheme on that of the Aria (the roles are reversed in the second part). Different from many other famous Variation cycles (such as, to name one of the most celebrated, Mozart’s Ah vous dirai-je maman), Bach did not use the theme’s melody as the element to be varied, but rather its harmony, its bass line. Thus, the kinship between two different Variations or between some Variations and the Aria is not always apparent, unless one delves into the structural element of their harmonic principles.
If this process highlights the harmonic plan of the Variations, the polyphonic dimension is also enhanced by Rheinberger’s choices. As is well known, every third Variation of Bach’s original is a Canon (i.e. a musical genre in which the same melody is presented by two or more parts, starting with a temporal delay); moreover, the Canon’s numbering corresponds to the interval at which the Canon is realized. Thus, Variation III (i.e. the first Canon) is a Canon at the Unison, meaning that the canonic parts play exactly the same notes; however, in Variation VI, the intervals (i.e. the difference of pitch) of the canonic tune are strictly respected, but while the first part begins with a G, the second part begins with an A: G-A is an interval of a second, and in fact this is the second Canon.
Throughout the cycle, Bach composed Canons in two parts, which are in most cases accompanied by a bass line which is independent of the canonic rules; since normally pianists are two-handed, this implies that there are three parts to be divided between two hands. The challenge for the pianist is to bring into relief each individual line without letting the hearer be confused by the frequent intertwining of the parts; this is of course a complex technical issue, but one for which accomplished pianists are properly trained. Doubtlessly, however, if the melodic matter is assigned to two different pianists, the result will be much clearer, not only for technical reasons, but also due to the “stereophonic” impression one receives when listening to sounds coming from two different sources: this dramatically increases the clarity and intelligibility of the polyphonic structure.
The transcriber’s hand is more noticeable in those Variations in which he does not merely fill the harmonic structures, but adds new lines to polyphony, or intervenes on the texture by transforming its sound. Of course, the transition from one harpsichord to two pianos impacts dramatically on the aural result, but this effect is enhanced when the transcriber’s aesthetical perspective is more deeply determined by the taste of his time. This is especially observable towards the ending of the cycle, particularly in Variation 29, where the volume of sound and power displayed by the two pianos and created by the writing adopted by Rheinberger is utterly (and very convincingly) pianistic: the grandiosity of this Variation is clearly at odds with the thinner sound of the original, but the result is impressive and powerful.
Interestingly, Rheinberger’s published version does not include the indication to repeat the Aria after the concluding Quodlibet. The return of the Aria after the cycle’s ending is a very special and touching moment, which no listener of the Goldberg Variations is likely to forget; however, Rheinberger’s omission is significant, because it leads the musicians to conclude their performance on a jubilant and powerful note, instead of the more restrained and transfigured sublimity of the Aria in its simplicity. This resounds with Busoni’s idea of a majestic conclusion, and with his proposal to rewrite the Aria Da Capo so as to make it more imposing and solemn.
In this Da Vinci Classics album, however, the performing artists have decided to reproduce Bach’s original, in order to close the circle traced by the magnificent itinerary of this extraordinary cycle.
Album Notes by Chiara Bertoglio