At the end of the major-mode section of the Chaconne is the second passage where Bach provides only a series of chords and the direction “arpeggio” in the score:
We encountered this before; a lengthy section beginning in measure 89. In that case, though, Bach provided a specific way to play the arpeggio—at least, the first few measures:
But in this second passage there is only the harmony and the word; no suggestions about how to play it. So I have to be, in a limited sense, the composer.
This reliance on the performer to improvise or fill in is not unprecedented in music of the Baroque era. For example, a certain amount of ornamentation is expected in this music. Sometimes composers insert signs in the score to indicate an ornament, and while there are conventions and some treatises from the time to tell us how these ornaments are to be played it is still up to performers to decide how and where to play the ornaments. But in a sense this is like deciding where or if to put bumper stickers on your car: while they personalize the look of the car they don’t change its basic essence.
A more relevant example might be the use of figured bass in music of this time. Much Baroque music relies on the use of “basso continuo”—one or more instruments that provide harmonic support. A typical combination is cello (for the bass line) and harpsichord (for the bass line and the chords). The basso continuo part in a score is a bass line, with the actual notes written down. These bass notes are accompanied by figures that indicate how the rest of the harmony should be played.
While this combination of a bass note and a figure tells the performer what to play, it doesn’t tell how. It is up to the musician playing the harmony to realize the figures; that is, to decide how to play the notes of the indicated chord.
It’s a musical shorthand, like the “lead sheet” used today in jazz and popular music, where you are given the melody line with chord symbols.
All of this is to say that it isn’t all that unusual in the music of Bach’s time not to be told exactly what to play. So I have to “be” Bach and decide the final shape of this part of the composition. For inspiration and ideas I listened to some of my favorite violin versions. Interestingly, all of them—and then others that I listened to—all take an almost identical approach to this passage, breaking each chord into 2+2 double stops:
I found the uniformity of this approach striking and wondered why it was so. A little searching led me to a fascinating dissertation by a Brazilian violinist named Cármelo de los Santos on performance practice issues of the Chaconne. He addresses things like bowing, articulation, and the playing of chords by looking back over published editions and didactic works from the 19th and 20th centuries. It’s a fascinating document for anyone interested in this work and its performance history.
He devotes considerable time to the two arpeggio sections, and for context he looks at some contemporary 18th century treatises on playing the violin. It turns out that the sort of musical shorthand that Bach employs in these passages was pretty common in writing for strings. The treatises make clear that the performer can and should improvise “in good taste” and suggested arpeggio patterns are given.
The first “modern” edition of the Chaconne was made by Ferdinand David in 1843. His edition used the double stops for the passage beginning in m. 201 and almost all subsequent editions followed suit. That’s why all of the violin recordings treat these arpeggios in the same way.
De Los Santos demonstrates that the David version is consistent with accepted practices from Bach’s time, so it’s reasonable to think Bach would have approved and that it may well be what he intended.
Segovia’s version—and again, his was not the first guitar arrangement but the first one widely known—uses the same basic idea:
So Segovia channels David, and his guitar arrangement accurately reflects how a violinist would play the Chaconne. But as we have said all along, a guitar is not a violin and we needn’t be tied to its limitations. The guitar can play this in several ways and the trick is deciding which one works best.
It seems to me that there are two priorities. First, obviously, we want to keep Bach’s harmony intact. Second, and perhaps even more important, we want to bring out any moving voices:
My preliminary solution is to do something similar to the David/Segovia treatment, but using single pitches in the bass line instead of double stops. This approach keeps the motion in the bass line clear and makes it easier to control the shaping and dynamics. At the same time, I don’t think it gives up the mounting intensity of the passage.
But I am going to continue to experiment with other options, and the final performance version may well be different from this.