Goldberg Variations BWV988 (c1741)

arranged for string trio by Dmitry Sitkovetsky (b1954)

Aria Variation 16 (Ouverture)

Variation 1 (free) Variation17 (duet)

Variation 2 (free) Variation 18 (canon at the sixth)

Variation 3 (canon at the unison) Variation 19 (free)

Variation 4 (free) Variation 20 (duet)

Variation 5 (duet) Variation 21 (canon at the seventh)

Variation 6 (canon at the second) Variation 22 (free: Alla breve)

Variation 7 (free) Variation 23 (duet)

Variation 8 (duet) Variation 24 (canon at the octave)

Variation 9 (canon at the third) Variation 25 (free)

Variation 10 (Fughetta)                    Variation 26 (duet)

Variation 11 (duet) Variation 27 (canon at the ninth)

Variation 12 (canon at the fourth) Variation 28 (free)

Variation 13 (free) Variation 29 (free)

Variation 14 (duet) Variation 30 (Quodlibet)

Variation 15 (canon at the fifth) Aria

The story goes that the Goldberg Variations were written for Bach’s pupil Johann Gottlieb Goldberg whose night job it was to play the harpsichord to Count Hermann von Keyserling, insomniac Russian ambassador to the court of Saxony. Even if we believe the story – which was apparently passed to Bach’s biographer Forkel by the composer’s son Wilhelm Friedemann – we obviously don’t know whether Keyserling ever fell asleep before he had heard all thirty variations. After a few nights listening to Goldberg’s harpsichord, however, he would be thoroughly familiar with the opening Aria at least. If, given that familiarity, he could have heard the Aria played by a string trio, he would surely have sat up in astonishment. Certainly, in the string-trio arrangment made by the violinist Dmitry Sitkovetsky in 1984 it sounds delightfully fresh and absolutely natural.

Like the Aria itself, many of the variations, including all but one of the nine canons, are in three parts and perfectly suited to the string trio. At least for keyboard purposes eleven of the others are in two parts, which is usually causes no difficulty. Where problems are most likely to arise is, of course, in the five variations in four parts, where double-stopping has to be employed. Except in those few cases, the great virtue of Sitkovetsky’s trio version is that the textures, and therefore the contrapuntal voices, are so much clearer than in the keyboard version. The complex canon at the third in No.9 is just one instance. Another virtue, which is already evident in the Aria, is that long notes can be held by the strings for their full value – which, for example, makes the harmonies of variations like No.6 (the canon at the second) more explicit and clearly defines a sustained inner voice like that of No.13.

It could even be argued that the basic structure of the work is enhanced in the trio version. Since the variations are not on the melodic line of the Aria but on its harmonic structure, which is reflected in all thirty of them, the ability of the cello to sustain the bass line of the Aria in all clarity, without having to accommodate a middle voice as the left hand must do in the keyboard version, is a distinct advantage. The symmetrical aspect of the structure – the occurrence of a canon (at ever wider intervals) with every three variations from No.3 onwards, the developing virtuosity of every third variation from No.5 onwards, the intervention of an imposing French overture to mark the beginning of the second half of the work in No.16 –

is, of course, equally clear in the two versions.

Some of the virtuoso variations (devised for two manuals in the orignal) suit the trio better than others. No.26, which simultaneously sets running semiquavers in 18/16 against a dotted motif in 3/4, is brilliantly effective, largely because of the use of double stops in the 3/4 material. The chordal textures at the end of No.23 are perhaps not as successfully transcribed as those that occur throughout the splendidly sonorous No.29. The string trio does have the advantage of a winder range of colour, which is put to such delightful use in the pizzicato accompaniment to the melodic line in No.19 and to such expressive use, each voice with its own personality, in the lovely slow movement (marked Adagio in Bach’s own copy of the work) that is No.25 G minor. If there is the occasional textural confusion in some of the the four-part variations, the quodlibet of No.30 (which is not ideally clear in the harpsichord version) loses nothing of its joy in applying learned contrapuntal treatment to popular tunes of the day.

Gerald Larner ©2006