Were Bach's Toccata and Fugue BWV565 and the Ciaccona from BWV1004 Lute Pieces?
Altschuler, Eric Lewin, Musical Times
FOR MANY LAY LISTENERS AND MUSICIANS Johann Sebastian Bach's Toccata and Fugue in D minor BWV565 is the prototypical organ piece. It was therefore most interesting when, in 1981, Peter Williams drew attention to some striking anomalies about the work: that it comes to us not in a JS Bach autograph but in a less than perfect copy; and that many of its features - for example, the frequent unison octave passages, rudimentary harmony and counterpoint, and extended solo manual writing - are too simple, if not simplistic, for an organ work by Bach.1 Williams then made the brilliant suggestion that if the piece were by JS Bach, the current version might be a transcription of a previous version for another instrument, such as the violin. The piece does not lie perfectly for violin, however, and requires transposition for maximal effect. Therefore, more recently, Mark Argent made the ingenious suggestion that BVW565 may have originated as a piece for five-string cello, with the top (E) string tuned down a step.2 But I should like to argue here that an even better idiomatic instrument for BVW565 - and for the final movement of Bach's Partita for solo violin in D minor BWV1004 - is the lute.
D minor Toccata and Fugue BWV565
Argent mentions that one of the most striking sections of the Toccata (bars 12-15, ex.1) and the fugue subject itself use the bariolage technique - the alternation of fast-moving notes with an equally fast but fixed note.3 Bariolage is idiomatic to the lute and Bach's use of the technique already hints at a possible lute origin for BWV565. Indeed BWV565 is so replete with lute techniques deployed in a strikingly effective manner as to greatly reinforce the notion that it may indeed have been conceived as a lute piece.
The octaves of the Toccata's opening bars are unnecessary: accordingly, Argent employs only single notes in his cello transcription. The chords in bars io to 12 (ex.i) are, by their duration, clearly to be contrasted to the rapid notes that have preceded them. Yet on the cello or violin such chords must be broken by multiple-stopping, thus attenuating the intended contrast. On a lute the chord can be played as one and held. Bars 27 and 28 present a similar situation.
The bariolage of the fugue subject has also already been mentioned. Bariolage is also employed in bars 39-46. The two-part counterpoint, for example in bars 32-39 (ex.3), is simple enough to play on the organ, but difficult to play at all on the cello and, significantly, is not included in Argent's transcription. It is, however, at a level of difficulty and complexity most appropriate to the lute. The extended arpeggio sections in bars 66-70 and 73-82 can certainly be played on a bowed instrument, but seem easier and as a more relaxed break in tension - on the lute. Bariolage returns in bar 86, bar 103 and, finally, in bars 120 and 124, to close the fugue. The chord which begins bar 130 and the chords in the piece's closing three bars again act as a contrast to the rapid preceding notes. As before, these chords are easily played as one on the lute, whereas they must be broken on the cello or violin. As a lute piece the Toccata and Fugue could be performed to a much larger audience than it would have had it required an expert performance on a fivestring cello. Manifestly, the piece suffers not at all from realisation on guitar.4 To my ears as a lute/guitar piece the Toccata and Fugue has a powerful, evocative, intimate feeling.
D minor Ciaccona from BWV 1004
The Ciaccona from the D minor Partita BWV1004 - at 257 bars, longer than the four preceding movements of the Partita combined (185 bars without repeats) - is much admired for its structural ingenuity and complexity. Yet the movement remains largely enigmatic: despite the fact that almost every great violin player for the last two centuries has performed the piece, there is still a sense that the most appropriate way to play it, especially the large burden of double, triple and even quadruple stops, has not been found. …