Lost and Found in Translation

There are inevitable challenges involved in translating a piece of music from one instrument to another. In general, it’s probably fair to say that any good composer will write music in a way that takes advantage of an instrument’s particular construction and abilities. Moving it to a different instrument means adapting to a different set of capabilities. Let’s look at a specific example that occurs in moving the Chaconne from the violin to the guitar.

One of my favorite passages in the piece starts in measure 229 and lasts for 12 measures; I’ve reproduced the first part here (measure 228 is grayed out):

The movement of one voice around the static A in the other creates a palpable sense of tension. Harmonically, since A is the dominant of D it also creates a sense of anticipation, since we know that the A is going to have to give way to D at some point.

Bach uses this technique a lot in his organ music (as did many Baroque composers), and it became known as a pedal point since the held pitch was usually played on one of the organ pedals. The pedal point often appears near the end of a piece, where it is used to build anticipation before a final climactic passage. It has the same function here in the Chaconne.

On the violin, Bach takes advantage of the open string tuned to this pitch, so every other note in the passage (as marked in my example) is played on this open string; the violinist uses a rocking motion to alternate between this string (the second highest on the violin) and the other pitches fingered on the two lower strings. This is what it sounds like on the violin:

Hilary Hahn, violin

Because it’s played on an open string, the pitch continues to ring out between strokes of the bow, creating the illusion that A is sounding continuously like a drone.

The guitar can play this passage exactly as Bach wrote it (allowing of course for the fact that the guitar sounds one octave lower). Here is the first portion of the passage in Segovia’s arrangement:

Most guitar arrangements that I have seen follow this same example. While it works well enough, it’s very difficult to create the sort of droning effect that the violin can achieve with its open A string. Segovia makes it almost seamless in his performance, but it requires a lot of shifting of the left hand, making it very difficult (if not impossible) for mere mortals to create a uniform tone on the repeated A:

Andrés Segovia, guitar

If you’ve been paying attention you might recall that the guitar does have an open A string, but it is a bass string. Using that as the drone string doesn’t work; it drowns out the moving lines and just sounds wrong.

However…moving up and octave does work. I can place my 4th finger on the fifth fret of the top E string and leave it there for the duration of the passage. Furthermore, I can use the same right hand finger to play that high A every time, lending it a nice consistency if I pay attention and play it correctly.

Here is an score excerpt of the version I’ve settled on along with an excerpt to show how it sounds:

Having the pitch played at the same place on the same string by the same finger gives me the closest thing possible to playing the A on an open string and creates the same kind of drone effect as the original. As with any good translation the particulars are changed in order to retain the spirit of the original.

I’ll be Bach

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:

Bach: Chaconne, mm. 201-208

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. 

Bach Sonata for Violin and Continuo, BWV 1021

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.

BWV 1021 continuo part, “realized”

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.

Ferdinand David’s 1843 edition of Bach’s Chaconne. Compare with the original above.

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:

Segovia’s guitar version of the Chaconne, mm. 201-2

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.

A Major advance…

As you might guess from the relative lack of journal activity in July, I took a bit of a break from the Chaconne. I wasn’t away from the guitar, though. I put together a program of pieces to play for an outdoor wine and cheese party at a friend’s condo in Harlem, and also played at my friend Jeff’s annual gathering of guitar players and fans: GuitarBQ. Both events were fun despite the summer swelter. But now it is back to Bach.

At the beginning of August I began to work on the section in D major that begins in measure 133. After 132 measures of D minor the change in color is striking. Even more striking, to me, is the way that Bach manages the transition. After the flurry of activity that ends the first the minor section —rapid scales and big chords— he introduces the new key very gently, with only two voices and a slowing of the rhythmic activity from 32nd and 16th notes to eighths and quarters. The effect is not like a great parting of storm clouds to reveal brilliant sunshine, but more like the first glimmers of daylight after a long, restless night.

Actually, it’s not quite correct to think about this section as being in a new key. We are still in the world of D—in musical terms that remains the tonic or key note—but the mode has shifted from minor to major. This highlights one of the more remarkable thing about this piece: that it contains no modulations. This is Bach playing the compositional game with one hand tied behind his back. Modulating, or changing key, is one of the best tricks in the composer’s repertoire. Changing key creates a sense of departure, and it can heighten tension or give a feeling of instability. The eventual return to the starting key feels like a kind of homecoming and gives us a kind of emotional satisfaction. By relying on the same basic harmonic progression for every 4 measure unit in the Chaconne, Bach sets himself a real challenge to keep things interesting in other ways. Fortunately, he is more than up to the task.

This major-mode section of the Chaconne is half the length of the section that precedes it. Unlike the first half of the piece, which is for the most part structured in couplets—pairs of four measure units with related material—this section is structured in larger groups.

The first group of 16 measures establishes the new mode with straightforward harmonies and stately eighth-note rhythm.

At the start of the second group in m. 149, we go dancing off in 16th notes into a dialogue between upper and lower voices.

This transition leads us to the next new idea. Arpeggios spell out the chords, ascending to a high point in m.158.

 Bach introduces a threefold repetition of the note A that permeates the arpeggios for the next 8 measures.

In the third group, starting in m. 169, three repeated notes become four and the texture thickens. The insistent rhythmic figure (like Beethoven’s “Fate” motif from the Fifth Symphone—short, short, short, long) creates a sense of tension or expectation.

The rhythm slows dramatically in the fourth group, starting in m, 177, and we have a sense of resolution from the preceding tension. 

A few measures later we come back to the original rhythmic figure that opened the piece. Here Bach adds another wrinkle as he scatters in the pitch C natural—not part of the D major mode—which creates a little instability in what until this point has been rock-solid tonality. Are we at long last modulating to a new key?

No. But something is happening. Even though the rhythmic pace remains slow, the tension increases as the pitches start to rise. Then, in a final twist, the last 8 measures are marked arpeggio (but the exact nature of the arpeggio is left to the performers imagination).

What happens next is my favorite moment in the piece. Indeed, it is one of my favorite moments in all music. But you’ll have to wait to read about that.

“This suspense is terrible. I hope it will last!” — Oscar Wilde, The Importance of Being Earnest

All of the recorded examples in this post come from Hilary Hahn’s excellent recording.

Smoothing out some rough edges

The topics of my two previous posts about my master class with Petra Poláčková—the interpretation of dotted notes and the use of arpeggio—are both about performance choices and interpretation. She made two other observations that were more about choices I made in my edition of the piece, but ultimately they are about performance as well. Both relate to fingerings, a topic I discussed back in February (See The Fickle Fate of Flying Fingers, Part 1 and Part 2), and both concern potential obstacles to smooth legato playing.

Her first observation was that I am not taking enough advantage of open strings in my fingerings. As an example, here are my fingerings for measures 4-5:

Petra suggested using the open 3rd string instead:

And, indeed, after playing it a couple of times it is clear that using her suggestion makes it easier to play smoothly. That single open string gives my hand a little extra time to move, and the g rings into the next pitch.

Sometimes open strings and stopped strings can be combined in clever ways to allow several pitches in a scale passage to ring out, as they would on a piano if one held down the sostenuto pedal while playing several keys in succession. The guitarist Douglas Niedt gives an excellent explanation of it (with examples) on his website.

Petra pointed out some small examples in the part of the piece that I played for her, but the idea provides  the fingering solution to a passage that has been vexing me for some time. It happens in measure 88, just before the start of the long arpeggio section. My original version was clunky, with the shifting first finger a particular problem. Try as I might, this passage always stuttered and ruined the transition to the next measure.

But using the campanella idea makes it much easier to play. Note in the revised version that no two consecutive pitches are played on the same string (shown with the red underline), and I use three open strings (marked by arrows). Playing this little scale like an arpeggio helps me get to the actual arpeggio much more fluidly!

Petra’s final observation was that I am not using enough slurs; that is, using a right hand finger to play the first note of a pair (or trio) and sounding the subsequent pitch(es) with the left hand alone, either by pulling a finger off the string to pluck a lower pitch or hammering a fingertip down to sound a higher pitch. I have been very sparing with the use of slurs in my edition of the piece in a desire for consistency in articulation, but Petra helped me to hear that this consistency can easily become monotony. There is not one particular place I can point to as an example, but I’ll be looking for opportunities to employ slurs more liberally.

I’ll conclude this series of posts on my master class with Petra by saying that a week ago she emailed me a scan of her playing edition of the Chaconne with all of her fingerings and performance notes. It was a very generous gesture that says a great deal about her, and about the wider community of guitarists. I feel fortunate to be in such company and to have such support for my project.

Chaconne à son goût

The first question Petra Poláčková asked after I played through the first part of the Chaconne in her master class at the Cleveland International Classical Guitar Festival was “Do you think it should be a little more French?” It might seem an odd question, Bach being German and all. And the real title of the piece is the Italian ciaconna, not the French chaconne. But I knew exactly what she was asking, and why.

The history of written music is a topic that has occupied many musicologists. Thomas Forrest Kelly lays out the fascinating early roots in Capturing Music, tracing the origins of our system of notes, rests, and staves through the earliest manuscript sources. The development of movable type and the printing press fueled the same explosion of music publishing as it did for the written word. Thanks to music notation we can hear music from hundreds of years ago.

However, the performance of older music is not without its problems. Technical difficulties aside, we don’t know—and probably can’t know—exactly how older music was played at the time. Starting in the 19th century composers began including very specific performance directions in their scores: indications for dynamics, tempo, and even emotional character. We also have some recorded legacy for understanding how such music is to be played, since musicians recorded in the very early days of the new technology learned from teachers who were passing on performing traditions and styles they had witnessed and absorbed. But printed and handwritten music of earlier times is almost entirely free of such performance cues, and the living memory of those who taught that first generation of recorded musicians doesn’t extend back so far as Mozart, to say nothing of Bach and those who came before him.

This uncertainty about how earlier music should sound has nagged at musicians for a long time, and it gave rise to the historically-informed performance movement. The idea was (is) that by studying the available evidence outside of the scores themselves, like treatises, method books, descriptive accounts, and even historical instruments from the time, we can arrive at an understanding of how early music was performed at the time it was written. Some 50 years on, the idea remains surprisingly controversial, along the lines of a great religious schism. Adherents claim that “authentic performance practice” is the only way to really understand early music, while detractors claim that authenticity is a meaningless and unattainable standard.

Petra’s question about the “Frenchness” of my reading of the Chaconne arises out of the HIP movement, and it has to do with the interpretation of dotted rhythms. In much French music of time leading up to Bach the use of “double dotting” is common. In notation, placing a dot after a note indicates that the duration of the dotted note is half again as long as the original note—a dotted quarter note has the duration of 3 eighth notes rather than 2.

If you place a second dot after a dotted note, the duration is extended by half the value of the first dot. So a double-dotted quarter note has the duration of 2 eighth notes (for the quarter note) + 1 eighth note (for the first dot) + 1 sixteenth note (for the second dot).

To apply this idea to the Chaconne, the score as written looks like this:

And, as written, sounds like this:

But in the “French-ified interpretation” it is played as if it looks like this:

And it sounds like this:

There is evidence to suggest that the practice of double-dotting was used in performance even when the music was not so notated; that such a stylistic practice was taken for granted by the composers and performers. The analogy might be to the notation of jazz, where a melody would be written in regular eighth notes but played with swing style.

We have for, example, this advice from composer Johann Joachim Quantz (1697-1773):

“The quavers [eighth notes] that follow the dotted crochets [quarter notes] in the loure, sarabande, courante, and chaconne must not be played with their literal value, but must be executed in a very short and sharp manner.”  He also wrote that stringed instruments must “detach the bow during the dot” of a dotted quarter note. This would leave a little space and, coupled with the shortening of the eighth, greatly intensify the rhythmic pattern. 

Quoted in Dance in the Music of J. S. Bach by Meredith Little and Natalie Jenne (Indiana University Press)

There are contrary arguments, including the fact that Bach in some cases wrote double dots; it’s not as if he didn’t understand notation. 

In the end it is not a question that can be resolved on evidence alone; we simply cannot know how Bach would have played it, or wanted us to play it. It may just come down to a matter of personal taste.

As I said to Petra during the class, the decision to play the dotted values as written was a deliberate but not necessarily final one on my part. I felt it best to follow the literal score in the beginning, leaving open the possibility of changing my mind and my playing as I get farther along with the piece. I do like the way the double-dotted approach lends a dance-like feel to the piece even at a slower tempo, and I will experiment with this “French” idea going forward.

Chaconne à son goût!

On the other other hand

In “On the other hand” we looked at some of the general principles of choosing fingerings for the right hand. In this post we’ll consider a specific example: a single measure from the Chaconne that requires thoughtful planning for the right hand.

Let’s begin by looking at the measure in question—60— from the violin score.

m60_urtext
Measure 60 from the violin version of the Chaconne. The slurs are also found in the autograph manuscript.

Here, for the first time in the piece, Bach’s writing for the solo violin implies a four-voice, polyphonic texture. If this were a passage from one of Bach’s chorales for soprano, alto, tenor, and bass, it might look like this:

m60_satb
An SATB choral setting of measure 60 from the Chaconne

It’s the independence of the moving parts that makes this measure tricky for the right hand (and the left, for that matter).

Another thing to notice about the violin original is the placement of the slurs—the little curved lines that join pairs of notes. In music for violin, such a notation indicates that the slurred notes are taken in the same upward or downward stroke of the bow. The effect of this is to create a softer articulation of the second note: not dahdah but dahah.

Here is how this particular measure is handled by two of the eminent editions of the Chaconne for guitar: the versions by Andrès Segovia and Abel Carlevaro.

m60_segovia
Segovia’s version of measure 60

One of the first things to notice about Segovia’s version is that he adds an extra note; the low A on the first beat. Next, notice that Segovia also uses slurs in the same places that the violin version does (except for the final pair). And finally, notice how few RH fingerings Segovia provides for this complicated measure! However, what he does provide gives hints to what he likely intended the player to do:

m60_segovia_implied fingering
Segovia’s version with the implied fingerings written in.

This appears to violate one of the principles established in the previous post: note the repetition of fingers, for example. But the intervening slurs give enough time for each finger to get ready for the next pitch.

m60_carlevaro
Abel Carlevaro’s version of measure 60 from “Guitar Master Class – Chaconne by J. S. Bach

Carlevaro’s edition provides a lot more information to the guitarist than Segovia’s; in fact, his edition is called a “master class” and includes extensive commentary on performance and the technical aspects. Carlevaro does not add the extra bass note on the first beat; nor does he carry over the slurs from the violin version.. But the clever p-i-p-i alternation can make the sixteenth-notes sound smooth, and articulating all of the notes can help to bring out the counterpoint.

m60_cfj
My version of measure 60. Note the similarities and differences with the Carlevaro version above.

The version I arrived at owes a lot to Carlevaro, as you can see. The primary difference is how I treat the final 4 notes in the top voice, preferring a-m-a-m to his a-i-m-i. However, as with all of these editorial decisions about finger, this could change as I spend more and more time working on this section of the piece.

This is the last post that I will write about technical matters for some time. I’ve been doing more reading on the Chaconne itself and have found a few really interesting articles that have opened up some new insights into the music itself. I look forward to sharing those with you.

Deciding which notes to play

Which notes?

Before I can start playing the Chaconne I have to figure out what notes to play. That’s not usually a problem when starting to learn a new piece of music—I simply learn the notes that the composer wrote down in the score! But it’s more complicated than that in the Chaconne.  First, as we already know, the Chaconne is written for the violin and not the guitar, so some kind of adaptation is necessary. Second, even in the original violin version there are two sections where Bach doesn’t tell the performer exactly what notes to play.Read More »