But the passage of time has deepened our perception of the Goldbergs, and Sitkovetsky comes to his task with so much musical history and hindsight that the weight of responsibility must have been daunting in the extreme. It makes his achievement here all the more remarkable.
His desire to get right inside this music, to take possession of it, experience it as a string player, has made for some quite extraordinary transformations. Not once did I find myself thinking in keyboard terms. The variety in string texture, the subtle and sensitive use of solo instruments, the way in which he has kept faith with the spirit and character of these variations while reinventing them sonically - the level of invention here is truly worthy of the piece.
And always his musical licence bows to respect for the text. For the Aria itself, something personal, something private. To something shared: the first variation releases an explosive exuberance, effectively freeing the theme to journey forth. I love the stringy resilience of the all-dancing, upbeat variations, the great booming downbeats of Variations 10 and 14. And then we look inward again: Variation 13 offers such a wistful slant on the theme, Sitkovetsky's solo violin supported and caressed with the barest harmony and counterpoint from solo viola and cello.
So much of his transcription is gracefully understated (the cooling spring- rain pizzicatos of Variation 19 are obvious but charming), but then again he'll rise to the baroque splendour of the Ouverture that is Variation 16 with all concerted flourish, embellishment, opulence. And occasionally - as in Variation 25, the great Adagio of the set - I do believe he actually extends the expressive reach of the music, retaining the spareness and concentration of the harmonies while opening up intriguing new space and resonance. Quite an achievement. And a stunning recording.
A musical arrangement is a strange beast. When you change the instrumentation of a work, it becomes something new, even if the notes remain unchanged. Mussorgsky / Ravel (Pictures at an Exhibition) is no longer pure Mussorgsky; and as for Stravinsky / Pergolesi... So Bach / Sitkovetsky has to be treated the same way: no point in criticising it as another version of the Goldberg Variations: Sitkovetsky's view of it has to be judged on its own terms.
The trouble is, I don't really understand what those terms are. Sitkovetsky's use of solo and full strings sometimes recalls the concertato and ripieno alternations of a baroque concerto; at other times, it sounds either imaginatively free or completely arbitrary. It's the same with his sporadic use of harpsichord. A ghost of the old continuo haunts one or two of the variations, very effectively in Variation 6, where its sharp, tangy sound, punctuating the running string figures, throws a new light on Bach's original.
But that kind of fresh illumination comes all too rarely. The problem for me is that this arrangement doesn't convince either as fake baroque string music or as an original, modern re-interpretation. Most of it falls limply between the two stools. That impression of limpness is sometimes reinforced by the playing - often cautious or expressively neutral - and sometimes by the texture of the arrangements themselves (the frequent unimaginative pizzicato bass-doubling for instance).
Try as I might I can't help comparing this with the classic keyboard Goldbergs - Glenn Gould's inspired 1959 piano version (on Sony), or harpsichordist Kenneth Gilbert (on Harmonia Mundi). As a solo instrument, the harpsichord can be dreadfully dry and mechanical, but Gilbert's performance is a thing of soaring beauty and taut muscular energy beside this. If the experience of making this arrangement has enriched Sitkovetsky's understanding of Bach, all well and good - I look forward to his recording of the Violin Concertos. I don't think I'll be playing this again, though.
STEPHEN JOHNSONReuse content