Hand Distribution I
In the last few months, several of my students have simultaneously wanted to zero in intensely on the question of how to decide which hand should play which notes of a passage, when there is any choice. I have been known to rely on a general comment, “if it’s unclear, try out different things, and go with what is most comfortable.” It is possible that I have over-relied on this level of generality; however, there is nothing wrong with this casual approach. It is the essence of what one should do in the end. Furthermore, just leaving it at that with a student could encourage that student to develop autonomy, to think for him- or herself.
But there are also many more specific and analytical things to say about how notes might best fall between the two hands. (In the series of columns from a few years ago that I headed “Working”—in which I examined the process of learning two particular pieces—I touched on this a little bit. Those brief discussions were tied to the specifics of certain passages, and not as theoretical or general as what I want to do here.) One of the purposes of this kind of analysis should be to widen the range of possibilities that students can see, to move the student farther from making limiting assumptions about what the choices are likely to be.
In this column and the next, I will analyze in detail my thinking about this. This analysis may seem too detailed: sort of fussy, or making too much out of something that could in fact be done efficiently through trial and error. Some students will happily take to going over these issues with this sort of fine-toothed comb—finding it interesting in the manner of a puzzle or detective story. Others will not, but will probably still learn something from doing some of it. Probably no one will go on analyzing these situations exactly like this permanently: it is a stage in learning to think about it and to develop intuition for it.
It is inadvisable to create a fingering for a passage without first deciding or knowing which hand will play which notes. However, perhaps because we talk so much about “fingering” as a crucial step—the crucial step?—in preparing a piece, students naturally want to plunge in to the actual fingering process as soon as possible. This may be one reason that students (and many players who are not students) often give more weight than is appropriate to the assumption that the upper staff is “the right-hand part” and the lower staff is “the left-hand part.” It seems to give a ready-made answer to the question of hand distribution. It would be willfully silly not to notice that there is a correlation: higher notes are more likely to be within the reach of the right hand, lower notes within the reach of the left hand. However, I think that it serves the player well to try to separate that simple logistic fact from the typography. The first principle of hand distribution, for me, should be:
Try not to think of the two manual staves as representing the two hands. Instead think of the combined staff—ten lines, eight normal spaces, one larger space in the middle, ledger lines, and so on—as representing music for your ten fingers to play in the best way possible.
Of course this—and all of what I am going to be discussing here—does not apply to music that is expressly written for two manuals.
This is just a change of attitude, but it makes a difference. Part of the difference it makes, if you really internalize it, is that there is no longer a “burden of proof” assigned to the notion of playing notes with the “other” hand. The sort of question that goes something like this—“Is it OK to play that note with the other hand?” or “Would it be all right to take that note over with the right hand?”—simply goes away. This saves time, for one thing. I have known performers to debate with themselves and take a long time to believe that it is (morally, ethically, or practically) acceptable to “switch” some note or notes to the “other” hand, when that is clearly the natural way to do it, and when the composer’s original way of writing out the music very likely didn’t have the notes divided the same way between the staves anyway!
Analysis for hand distribution
In order to determine how to distribute the notes between the hands, there is a sort of protocol that you can follow—a series of observations to make about the passage that will help you discover the most comfortable hand distribution for you. The way I am going to present it here is too cut-and-dried, really: in practice, all of these interact with one another and sometimes with other considerations. But these questions are a good starting place:
1) Is the manual part in (only) two voices? If so, then there should be a strong assumption that each hand will play one of those voices. This is the simplest case. It is almost not worth talking about, except that it brings us to the second great principle of hand distribution: when it is possible, ask your hands to share the work pretty much evenly. Students should be reminded not to play two voices in one hand while the other hand does nothing—at least not to do that reflexively, just because both voices are in one staff (as they might often be) or are both high or both low. Once in a while, two voices should be combined in one hand for the most practical of reasons: to free a hand up to turn a page or to manipulate stop knobs or something like that. This should only be done if the loss of ease in playing is very small.
2) If the passage is in three voices, then realistically the issue becomes “which hand will play the middle voice?” Again, this is a better way to ask the question than “where can one hand take over the middle-voice notes that seem to belong to the other hand?” This is probably the kind of texture in which these issues actually come up the most. The questions to ask in figuring out how to distribute the notes of a middle voice between the hands are:
i) To which outer voice are these notes closer?
ii) Is one outer voice more active, busier, or just plain harder than the other? If so, then can the inner voice be grouped with the less difficult outer voice?
iii) Is the middle voice more intricate than either of the outer voices? If so, then what can be done to make it as simple as it can be: can it be split between the hands, for example, or grouped with the simpler of the two outer voices?
iv) Are there ornaments—especially trills—in any voice, and does it make sense to isolate ornaments into their own hand, if possible?
v) Sometimes the need to prepare a good fingering for a repeated note suggests a way to proceed.
vi) Is there anything about hand position, independent of all of the above, that suggests that the middle voice notes can be played more comfortably with one hand than the other? This often comes about because of something to do with sharps or flats. Perhaps one hand would be forced to play a raised key with an awkward finger if also called upon to play the inner voice, while the other would not run into any such trouble.
Example 1 shows the Alla Breve section of Bach’s Prelude and Fugue in D Major, BWV 532, beginning on the last beat of m. 16 of the piece. It provides all sorts of material for thinking about hand distribution in three-voice manual textures. The rest of this column will be taken up with a detailed analysis of parts of this section. (For simplicity, from here on I will provide examples with only the manual staves, since we are only looking at the manual writing, though there is pedal almost throughout.) In this entire 80-measure section, there is no spot at which the right hand actually cannot play both the upper voice and the middle voice—though in many spots doing it that way would create awkward fingerings and articulations that would probably not make either the player or the listeners happy; the left hand can reach both the lower and the middle voices for all but an occasional measure or half measure. Therefore this is a good passage to think about what is best, not what is necessary or inevitable.
The middle-voice f#′ on the last beat of measure 16 can be taken very easily with either hand. It is almost equidistant from the upper and lower notes, and there is nothing in particular going on: on that beat, no voice is, or is about to become, busy or intricate. So, is there a way to make the choice, other than at random? (“At random” is fine when there is no reason to prefer one over the other: this is often the case. The point is to recognize when it is not the case.) The thing that I notice about this beginning is that there are several repeated notes coming up. I would prefer to finger them with different fingers for each repetition. Of course, different hands implies different fingers, so right away it seems as though there might be some good to be derived from moving the middle voice back and forth between the hands a bit. So, if we take that f#′ with the right hand, then the left hand is free to finger the repeated d′ in the lowest voice in whatever way seems best, and then the left hand can play the repetition of the f#′. Example 2 shows one specific version fingering modeled on that idea.
And there could be others. A player who does not want to change fingers on the repeated notes, might use fingerings shown in Examples 3 and 4, or something else.
At each point over the next few measures, we see one voice being more active than the others, as shown in Example 5.
In the first half of measure 18, it is clear that the middle voice should be taken in the left hand, both because that way the more active voice is left alone in one hand (point ii above), and because the notes of the middle voice are closer to those of the lower voice (point i above). In the second half of measure 18, the situation is more complicated, since the voice that is more active than the others is the middle voice itself. The three last eighth-notes in the middle voice, c#′′–b′–a′, can be reached by either hand. Does it matter which hand plays them? The c#′′ is closer to the upper voice, and the other two notes are closer to the lower. That might suggest splitting that line that way. However, the upper voice is in itself simpler than the lower voice through that part of the measure, which would suggest just playing the eighth-notes of the middle voice in the right hand. Here personal habit or something about the shape of the player’s own hand might determine the choice: therefore it is a place to try it a number of ways and see what works.
However, that might be influenced by what is coming up. The first few notes in the middle voice of measure 19 can also be reached by either hand. However, here the choices might make more of a difference to the musical effect than in the previous measure. The second eighth-note of the measure must first sound like it slips in and takes over the position of the e′ that is being held: that is, in such a way that it doesn’t sound like that note is being repeated in the same voice, even though as a key on the keyboard it is in fact being repeated. (This notion comes up again very soon.) Then it has to be held over into the a′ half-note, just long enough to suggest that the lower voice is still there. Furthermore, the a′ half-note itself begins a new pattern (that is, through the lens that we are using here). It begins a long passage in which the upper voice is clearly more active than the lower two voices, and should therefore be left alone in the right hand if at all possible. It is possible: beginning on that second beat of measure 19, the two lower voices fit comfortably in the left hand for the rest of this excerpt.
So if that a′ should be in the left hand, then so should the e′ that precedes it in the middle voice. (Otherwise there would be an unnecessary and awkward hand crossing.) If the three e′s in a row will be played in the left hand, then that hand needs all the flexibility it can get, in order to come up with a fingering that creates the right flow: in particular, this is a case where changing fingers on that note is a tremendous aid in creating the non-repeated note repetition that I described above. The left hand is the most free to finger those e′s if the right hand plays the g#′. That should then also be taken into account in planning the approach from the previous measure to the g# on the first beat of this measure.
To be continued . . . ν