17

Here is an excerpt from Bach's BWV 796, measures 8–9. (Unfortunately, ABCjs renders the ties in exactly the wrong direction!)

X:1
T:Sinfonia No. 10 in G Major, BWV 796
M:3/4
C:J.S. Bach
%%staves [(1 2) 3]
L:1/16
K:G
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
V:1
g2e2c2A2 (f4 | f2)d2B2G2 e4 |
%
%
V:2
(c8 c)edc | (B8 B)dcB |
%
%
V:3 clef=bass
EDCB, A,G,F,E, D,CB,A, | C,CB,A, G,F,E,D, C,B,A,G, |

The right hand has two voices. There is an eighth note C that interrupts a half+sixteenth note C. The following measure has the same pattern, but a step lower.

At the technical level, a compromise has to be made to play these voices on one piano keyboard. Which one, and how does one ensure that each voice maintains its independence from the others?

  1. Play and hold down the short note:

    X:1
    M:3/4
    %%staves [(1 2) 3]
    L:1/16
    K:G
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    V:1
    g2e2c2A2 (f4 | f2)d2B2G2 e4 |
    %
    %
    V:2
    c4 (c4 c)edc | B4 (B4 B)dcB |
    %
    %
    V:3 clef=bass
    EDCB, A,G,F,E, D,CB,A, | C,CB,A, G,F,E,D, C,B,A,G, |
    
  2. Abandon the held note prematurely (and the eighth notes would probably be played somewhat staccato):

    X:1
    M:3/4
    %%staves [(1 2) 3]
    L:1/16
    K:G
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    V:1
    g2e2c2A2 (f4 | f2)d2B2G2 e4 |
    %
    %
    V:2
    c4 z4 zedc | B4 z4 zdcB |
    %
    %
    V:3 clef=bass
    EDCB, A,G,F,E, D,CB,A, | C,CB,A, G,F,E,D, C,B,A,G, |
    
  3. Strike the held note again as soon as the interrupting voice vacates it:

    X:1
    M:3/4
    %%staves [(1 2) 3]
    L:1/16
    K:G
    %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    V:1
    g2e2c2A2 (f4 | f2)d2B2G2 e4 |
    %
    %
    V:2
    c4 z2 (c2 c)edc | B4 z2 (B2 B)dcB |
    %
    %
    V:3 clef=bass
    EDCB, A,G,F,E, D,CB,A, | C,CB,A, G,F,E,D, C,B,A,G, |
    
3
  • 2
    I made a significant rewrite of your question to distinguish it from other similar questions. The distinction, IMO, is very worthwhile, and, I believe, in line with your intent. However, if you feel I've damaged your question, please roll back (or let me know, and I'll handle the rollback).
    – Aaron
    Dec 14 '20 at 3:52
  • 1
    I'll bet he wrote this for organ or dual-keyboard harpsichord! On a single keyboard, I'd definitely go for 1 Dec 15 '20 at 0:21
  • For completeness, a kind of dual to your (2) that is not unheard of is to abandon the short note (<< { g8 e r a } \\ c2 >>). I agree with the answers that option (1) is better, but I’ve heard this alternative, too.
    – wchargin
    Dec 15 '20 at 1:19
11

One of the things my first piano teacher taught me when looking at Bach was to isolate the voices and play the parts by themselves to get to know how the individual lines sounded like, not just their congregation. Because we are limited to only two hands of five fingers only, we have to make some sacrifices sometimes, as you correctly noted.

In your example, method 1 would be the correct way. It is the closest you can get to the original intended voicing, whereas method 2 completely disregards the latter two eighth notes and method 3 inserts a new note into the voicing

How to keep the voice independent? This is a bit trickier because we have already admitted to sacrificing the way it was written for playability. My best advice would be to practice the voices independently with the change and work on keeping the flow and dynamics in line with that voice

1
  • +1, but a small quibble: In this example, the hardest constraint is not our hands but that we only have one key per note. Feb 11 at 13:34
8

The short, general answer

All things (all voices) being equal, solution #1 is the correct technical approach. The shorter note should interrupt the longer one, but then continue to be held (allowing for additional interruptions) for the duration of the longest note.1,2 Solution #2 is appropriate in some situations (see below). Solution #3 is never appropriate.


The short, specific answer

For the measures given, I would slightly emphasize the long notes to help sustain them. Since they are being held (i.e., are static), the moving voice will naturally be perceived as superseding them in prominence. The allows the moving voice to be played dynamically just below the emphasis level of the long note. In that way, the re-articulation, slightly de-emphasized, will be heard as part of its own voice, but its sustain will be heard as the sustain/decay of the original, emphasized articulation.


Further discussion

The important distinction in this question from other similar ones is the focus on maintaining voice independence across the pitch repetition.3

It is customary when playing polyphonic music on the piano that voice independence is maintained by variations in dynamic level and articulation (touch) between voices.4 In the following discussion

*first voice* = the polyphonic voice first encountering the repeated pitch.
*second voice* = the polyphonic voice responsible for rearticulating the repeated pitch.

Voices distinguished by dynamics

Louder voice = first voice

It's a bit easier if the first voice is louder than the second, because the second voice's sustain more convincingly sounds like the natural decay of the first voice.

Louder voice = second voice

Should the second voice be louder, then a case can be made for releasing the note in the second voice rather than trying to preserve the sustain of the first voice: the interrupting, louder voice taking precedence.5 This is option 2 in the OP.

Voices distinguished by legato/staccato touch

Legato voice = second voice

When the second voice is the legato one, it allows for a blending of the natural decays of both voices.

Staccato voice = second voice

As with the dynamics case, if a staccato voice is interrupting a legato one, the a case can be made to curtail the first voice (but see note 4).


1 This is also true if a long note is interrupted by another long note. The interrupting note is held through all subsequent interruptions, even past the length of the original note, until the longest "current" value is exhausted.

2 The same principle is discussed in other similar questions, such as overlapping note on same staff.

3 For example, in the question mentioned in note 2, it is important to maintain the distinction between the held melody note and the accompanying chords that rearticulate it. If the chords are too loud, the effect will be as though the melody comprises repetitions of the same note.

4 In Bach's two-part Inventions, for example, students are often introduced to this concept by being taught to play sixteenth notes legato and/or louder while accompanying eighth notes are played staccato and/or softer.

5 In some cases, even when the second voice is louder, the decay of the first voice can be maintained. This involves making the second articulation and then briefly allowing the damper to just barely contact the string -- damping it somewhat, but not fully. This can be achieved either with the pedal, by a half-release, or by half-releasing the key itself, but does require technical circumstances that allow for such careful control.

1

Certainly not 3. We mustn't hear a note struck that isn't in the music. 1. seems about right. At a faster tempo it might tend towards 2.

1
  • You bring up a good point. At faster tempos, 2 would become indistinguishable from 1 Dec 15 '20 at 1:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.