On Teaching

Gavin Black is Director of the Princeton Early Keyboard Center in Princeton, New Jersey. He can be reached by e-mail at [email protected]. He writes a blog at www.amorningfordreams.com.

Substitution in thirds

The left-hand version of this exercise for practicing substitution in thirds starts as is shown in Example 1. Again, you should carry out the multiple substitutions in the most comfortable order. Try out various other fingering possibilities, and also try this, and other similar exercises, with added accidentals, as if it were in C minor, for example, or transposed to other keys. Substitution on black notes is physically different from substitution on white notes, since the black notes are thinner and spaced further apart. The principle is always the same: perform the substitutions in the right order, and plan out carefully the direction and angle from which each new finger arrives and in which each old finger departs. If the departing finger is released down and to the side, you have to be careful that it doesn’t inadvertently play an adjacent natural. 

You can also convert the Rameau passage shown in Example 2 (and discussed in the January 2014 column, as Example 9) into substitution exercises by tying the repeated notes and changing the repeated-note fingerings
into substitutions.

Substitution in fifths

Examples 3 and 4 are exercises for substitutions in fifths. Try these all over the keyboard: as written but with added accidentals, in other keys, an octave up or down. As you practice these exercises and a selection of transpositions, try carrying out the substitutions according to various different timings: 

1) As quickly and as smoothly as possible: instant substitution, but keeping the order correct. For example, in the above left-hand fifths it is better to execute the 4–5 substitution before the 1–2. This keeps the hand compact and avoids uncomfortable stretching. In doing the substitution instantly, as one gesture, this order can be preserved by carrying out something that feels like a rolling motion of the hand. (Technically the 4–5 is closer to “instant” than the 1–2, but the gesture is fast and smooth and should feel like one event.)

2) Very promptly and rapidly, but as a succession of separate quick gestures: in the case of these two-note-at-a-time substitutions, the timing of this approach is similar to that of a mordent.

3) Truly timed finger changes. This can be in a number of rhythms. For example, in the rhythm of the fifths above: triplet quarter notes (i.e., with the new chord, the first substitution, and the second substitution spaced out evenly); a quarter note and two half notes; a dotted quarter note and two sixteenth notes. This last rhythm shades over into the final timing concept:

4) Both substitutions as a quick one-piece gesture at the end of the held note, having almost the feeling of a before-the-beat ornament to the
next note.

I have listed these in a particular logical order: from closest to the beginning of the note to closest to the end of the note. It is important to avoid practicing anything at a faster pace than what you can carry out comfortably. Therefore, you should start with a slower (timed) execution of each substitution, and work towards the faster timings and the “instant” un-timed forms as you become fully comfortable with the fingerings and the shapes of the gestures.

Substitution in scalar passages

Examples 5 and 6 form an exercise that has the appealing feature of being simultaneously silly and efficient. It involves playing simple scale passages, and performing extravagant chains of substitutions on each individual note. This is beyond what you are likely ever to do in fingering a piece of music. There are, however, places where more than one substitution occurs on one note, as we will see below.

This can also be practiced in different timings. The “instant” version will of course take a discernable amount of time, since there are so many fingers involved one after another. Don’t try to practice this exercise at a tempo faster than it can be accomplished accurately and comfortably. It can still feel like one gesture: sort of sliding or slithering around on the note. As always, you should pay close attention to hand position and to keeping everything relaxed and comfortable. You may notice yourself occasionally inadvertently releasing a finger before the next finger has arrived to take its place: in effect converting the substitution to a repeated note. Do not try to correct this by holding the notes down harder. It is just a matter of timing. If this becomes a problem, slow down the exercise.

Substitution in counterpoint

Example 7 shows a Reger passage (from the chorale prelude Morgen-glanz der Ewigkeit—discussed in the November 2013 column as Example 10), demonstrating the practicing of separate voices, with a suggested fingering involving thoroughgoing substitution. (This is the left-hand part. I have written the fingerings for the two voices above and below the staff respectively, for clarity.)

The following is a detailed discussion of the logic behind these fingerings, but with an emphasis on the substitutions, and with comments on how best to carry out those substitutions. You should read it and correlate it in detail to what you see in the music before practicing the passage. If as you read this discussion you think of different fingerings that you want to try out, please do so. Make sure that you understand your own rationale behind those fingerings and that you are convinced that they will be comfortable and effective.

The choice of the first finger for the first note makes sense both because that enables you to reach down to the second note easily, and because it puts the hand in the best position to play the a# that is coming up. (This is a comfortable fingering in part because of where the passage lies on the keyboard. Try the same pattern two octaves higher. It will feel quite different and might need a different fingering solution, perhaps playing the opening note with 2, and substituting 1 at the last instant before playing the a#.) The choice of 5 for the d that is the second note of the piece is also obvious. The first substitution (4–5 on the note e) should be performed quickly, both to relax the hand and to enable the second finger to reach the c#. The substitution on the f# should be treated the same way for the same reasons. The substitution on the g# can be performed either instantly or on a measured basis (it is the first opportunity here to practice the latter).

Moving to the second measure, the substitution on the c# on the first beat has to be performed quickly so that the hand can move on to the next notes (d′, e′). In theory, the 4–5 substitution on the a need not be done until close to the time to play the b on the third beat. However, at the moment of the first beat itself, it will be more comfortable to carry out both substitutions quickly, with the 4–5 actually happening first. (This is to keep the hand compact and avoid uncomfortable stretching.) This should be carried out as an instant “rolling” double substitution. The b on the third beat of this measure is an interesting case. It is natural to play it with 4, and it also should have 4 holding it when it is ready to end (seven eighth-notes later, in the next measure). However, it is a good idea to hold it instead with 5 through most of the length of the note. This is to put the hand in the best position to reach the notes in the upper of the two voices, especially the f#′ that is the first note of the following measure. There is no particular reason not to do the substitution from 4 to 5 right away. (You could also postpone it until just before the end of the measure, in which case it would probably be more comfortable to play the d# with 2. The advantage to playing the d# with 3 is that it enables 2 to be poised to reach towards the upcoming f# as promptly as possible. This is a positive reason to do the substitution instantly.)

At the other end of this long-held b, the substitution back to 4 should be done only after the upper-voice substitution from 2 to 1 on e′. This is so as not to stretch the hand out uncomfortably. The purpose of that upper-voice substitution itself is partly to un-stretch the hand and partly to free the second finger to reach up to f#′. The substitution from 5 to 4 on the a should be done only after the thumb has played the f-natural. This is, of course, to keep the hand from being stretched out unnecessarily. On the fourth beat of this measure we come to the first substitution of non-adjacent fingers. The switch from 5 to 3 on the g# is motivated by the underlying rationale for substitution: it makes sense for one finger to play the note, but for another finger to be holding when it is time to move on to the next note. The reason for using 3 in the latter role is that the next two notes are in a downward direction. (It would also be possible to do a 5–4 substitution, and then another 4–5 substitution on the g-natural.) In any case, the 2–1 substitution on the e′ should be done first and very quickly.

The first substitution of the fourth measure, on b in the lower voice, is the first one we have seen that must be really instant: fully a part of the gesture that plays the note in the first place. This is, of course, because of the sixteenth-note motion in the upper voice. The hand must be in position to reach for and then play f#′ comfortably, essentially right away. The two remaining substitutions in this measure can be done at a somewhat more leisurely pace. On the third beat in the lower voice, my suggested 4–3 4 could be replaced by 4 5–4. Or indeed you could do a 3–2 substitution on e′ that is the lower-voice note on the second beat, and then play the rest of the lower voice in this measure with the fingers as they come. The 2–1 substitution on the f-natural in the upper voice on the fourth beat exists for the purpose of un-stretching the hand.

There are four substitutions that must be carried out in the half measure coming up, the last part of this excerpt. The switch from 2 to 1 on e′ in the upper voice and the switch from 3 to 2 on d′ in the lower voice (a note that was initiated in the previous measure) must be carried out in the order in which I just listed them, for the most basic possible reason. The “new” finger in the second substitution was just in use holding another note. It had to be freed from that note—by substitution—before it could take over its new note. These two substitutions are ideal to be played quite measured: the upper-voice substitution on the second half of the first beat, the lower-voice substitution on the second beat. On the second half of the second beat we encounter two substitutions, both of which must be carried out within the time span of an eighth note. For most hands it will be more comfortable to do the 2–1 substitution on d′ first, and then the 4–3 substitution on b. This means that the latter must be very fast indeed. The two should end up feeling like one gesture. 

The fundamental purpose of this fingering could be described as a way to play all of the notes legato without awkwardness or discomfort. This is achieved by a significant increase in the amount of fingering busy-ness: as fingered here, this passage involves sixty-one fingering events to play forty-three different notes. For comparison, Example 8 shows one possible fingering without substitution, based on a willingness to allow many of the notes to be played non-legato. (Remember, however, that Reger in his own hand marked this piece sempre ben legato.)

Try this fingering out, leaving aside for the moment its musical or historical appropriateness. Keep everything light, and make the non-legato gestures smooth and non-abrupt. Is one fingering easier than the other? What differences in feel do you notice?

(Note: Based in part on feedback from readers of The Diapason, I will possibly add further exercises and examples to the final version of this section on substitution. These will deal at greater length with substitution on black notes and with non-adjacent fingers. I will move on next month to exercises and approaches to learning to play with hands and feet together.) 

Organ Method XIII

This follows directly from last month’s column. For those with little or no prior keyboard experience, I have made this method’s exercises simple, direct, and systematic. At the same time, assuming that the student can remember and build on what has come before, the student should be able to take the right approach to practicing the exercises and be able to concoct his or her own exercises to some extent. A student who is already experienced on another keyboard instrument should be able to get something important out of this section, since the feel and sound of playing in chords and multiple voices is critically different on organ from what it is on other instruments. 

Before we look at exercises designed to get each hand used to playing patterns of more than one note at a time, there are a few general points to consider.

1) The clearest physical difference between playing one note at a time in one hand and playing more than one note in a hand, is that the latter places more limits on fingering choices. If you are called upon to play five notes at once in one hand—which is rare but not unheard of—then there is little or (usually) no actual choice about fingering. In the more common situation of two or three notes at once in one hand, there are often different fingering possibilities, but not as many choices as when playing only one line in a hand.

2) Hand position, already discussed earlier in this column, can be even more important when playing multiple notes at once than when the hand is playing one line, and can also be more difficult to manage well. In particular, the role of fingering choices for raised keys—sharps and flats—in determining comfortable hand position is crucial. In some passages, the position of the notes necessitates some compromise in hand position. Part of gaining experience and comfort with playing the most complex repertoire is learning how to manage these situations well. If a hand position is not ideally comfortable, then it is important to relax the hand away from that position promptly and smoothly.

3) In a piece of music, or a passage, in which each hand is playing only one note at a time, each hand’s part is a musical line or voice or melody. When either hand has more than one note at a time, that texture can be multiple voices or it can be chords, or it can be some combination of the two. 

4) Playing two or more separate melodies or voices with one hand in a way that sounds to an attentive listener like simultaneous melodies rather than chords is mostly a matter of attentive listening by the player. Exercises designed to address that aspect of playing are essentially listening exercises.

5) In most contrapuntal pieces with more than two voices in the hands, at least one voice migrates from one hand to the other. When this is the case, it almost always means that the piece, or that part of the piece, should be played on one manual: otherwise the sonority of that voice changes at essentially arbitrary times. It is also important that choices about which notes will be played by which hand be made carefully and sensibly. (In particular, it is important not to assume that every note printed in the upper staff should be played by the right hand, and that every note printed in the lower manual staff should be played by the left hand. These will be the tendencies, but the whole texture should be divided between the ten fingers in whatever way is most comfortable and gives the best musical results.)

6) Chordal passages often present articulation issues. If a series of chords in one hand is meant to be played legato, that presents fingering challenges, often involving finger substitution—a technique that will be dealt with later. If chords in one hand are meant to be played detached, then more fingering choices are available. Any detached fingering should be practiced first with enough space between the notes that the physical motion from one chord to the next feels easy. Once the gestures have been established, the spaces between the notes can be made smaller without abrupt gestures or tension.

7) Just as it makes sense to practice hands or feet separately, it makes sense to practice individual components of the note picture within one hand separately. Sometimes it also can make sense to focus on listening to one component of the texture of the part being played by one hand over the other parts.

Examples 1 and 2 are simple exercises with which to begin playing more than one note in each hand.

I have placed these exercises in regions of the keyboard that, for most players, will permit the note patterns to feel most comfortable, with the forearm and hand aligned well. However, as with earlier exercises in this chapter, you should move them around: up or down by octaves, or by other intervals, mixing versions with few or no sharps or flats with versions that have more. 

There is an obvious fingering for these exercises. For the right hand: 3/1–4/2–3/1–4/2–5/3–4/2, repeat; and for the left hand: 3/5–2/4–3/5–2/4–1/3–2/4, repeat). However, you should also try different fingerings, for example, playing all of the two-note chords with the same pair of fingers, and simply moving the hand smoothly and gently from one chord to the next, or a mixed fingering such as (for the first exercise) 3/1–4/2–3/1–4/2–4/2–3/1 (repeat). In any case, whenever you pick up the same pair of fingers to play the next chord, make that gesture as light, relaxed, and smooth as possible. In particular, do not try to make the space between the chords particularly short: use as much space as you need to allow the gesture to be completely without tension or any feeling that you are “snapping” from one chord to the next.

Keep the tempo slow for now, and do not worry if you hear the two notes of each chord not quite sounding at exactly the same time as each other. This is important: of course in the long run you need to be able to make multiple notes in one hand sound exactly together, and also indeed to make them sound not quite together in ways that you have decided on for musical effect. However, any attempt to ensure that each finger depresses its note at exactly the same time as each other finger does —before you have developed a fair level of comfort playing note patterns of this sort—will lead to a touch that is too crisp and too focused on driving each key to the bottom. This can sometimes lead to real physical tension and, in the long run, pain. The good news is this: any tendency of multiple fingers to play notes somewhat out of kilter with one another will go away naturally and of its own accord as you continue to practice. 

If you have studied other keyboard instruments, you might be impatient with the simplicity of these first exercises. However, the touch and sound of the organ are different enough from piano, harpsichord, and clavichord that both the physical act and the listening aspect of playing more complex textures is very different indeed.

After taking a first look at these exercises and moving them around on the keyboard a bit, you should practice them in a couple of different ways that involve breaking them up. For example, play each line (upper note and lower note of each chord) separately, as shown in Example 3. (And similarly for the other parts of the exercise.)

Or stagger the upper and lower notes, as in Example 4. (And similarly for the other parts of the exercise. Don’t make this too fast: for this purpose, the rhythm doesn’t much matter.)

Note that these deconstructed versions of this two-voice chord exercise are themselves simpler than the exercises and pieces that you have already been playing. 

Try playing three-note chord patterns,  such as those in Example 5. In this case, the fingering can well be 1–3–5 (or 5–3–1), and the same for each chord. It is important to move from one chord to the next smoothly, allowing the breaks between chords to be as long as necessary to keep the motion comfortable. Are there other fingerings that are possible or, perhaps, better? Play around with it. Move these patterns around to other notes and other regions of the keyboard as you have done with other exercises.

This set of exercises can also be broken into component parts—the lower two notes, the outer notes, the upper two notes—or played staggered. You can devise ways of moving from one of these components to the full three-voice texture yourself, as in Examples 6 and 7. Make sure that you use the same fingers for the components that you want to use for those notes when they are put back into the full texture.

An exercise such as that shown in Example 8 combines some of the above:

A traditional four-part chorale harmonization, such as that of Old Hundredth (shown in Example 9), provides material for continuing to practice moving each hand from one two-voice chord to another.

For the current exercise, you should break this hymn into short sections, and into separate hands. Then work out a comfortable, sensible fingering for each section, assuming that it is acceptable to allow a breath or break between each two successive chords: that is, to play detached. Do not worry at this point about how detached the chords are, but, no matter how much space you leave between chords, keep your hands light and flexible at all times. Release notes/chords smoothly and gently, and move to the next note or chord calmly. Do not necessarily expect to put the whole hymn together or to put the hands together: that is not the point at this stage, though you may very well return to it later and learn it as a piece, probably with pedal. You can find ample material for this sort of practice in any collection of chorales or hymns. ν

Next month’s column will continue this discussion, moving on to techniques for practicing the art of playing truly independent voices together in one hand.

Gavin Black is the Director of the Princeton Early Keyboard Center in Princeton, New Jersey. He can be reached at [email protected].

Some thoughts on ornaments I
The playing of ornaments is one of those areas that many—maybe most—students find intimidating. This is only partly because it can be genuinely difficult. It certainly can be difficult, although, like most physical tasks, it can be made much less so through the right kind of technical preparation and through an adequate amount of well-targeted practice. The intimidation factor with ornamentation comes, I believe, mostly from a fear of getting it wrong. There seems to be so much data about how this kind of trill was played in Italy in 1620 or how that kind of appoggiatura was played in Austria in the early nineteenth century that it can seem impossible to keep up with it all. One well-known book on ornamentation is nearly 600 pages long, and that is just one book of very many. Also, impeccably credentialed experts on the subject can disagree. It is easy, looking at a piece of music, to know what the “regular” notes are, though of course it may not be easy to play them. But it is not necessarily easy, or even possible, to know for sure what the notes of ornaments are or what the exact rhythmic shape of a given ornament should be. It is also a common experience for even very talented and “advanced” students to feel that they have learned to play certain ornaments, but that those ornaments just don’t sound very good. This is always frustrating, and extraordinarily so when it happens more or less all the time.
I believe that part of this frustration comes from, or is made worse by, a confusion among some of these issues. That is, students often assume that their ornaments sound bad because they don’t know what the notes of those ornaments should be, or they have gotten something else wrong in the realm of the historical or the musicological, when in fact they sound bad because the execution is awkward or the preparation before the ornament itself is wrong. It may be important to know whether a given trill should start on the main note or the upper note, or how long or how fast it should be, or whether a certain appoggiatura should be long or short. However, any of the above should be able to sound good—natural, fluid, graceful—whether or not it is the correct interpretation of the composer’s intent. It is important to sort these different aspects of playing ornaments out from one another in order to be able to work effectively on learning to play ornaments well.

So, let us consider several issues.
First of all, what is an ornament? On one level an ornament is a note pattern indicated by a sign, rather than by notes as such. If, for example, the three notes c–b–c are indicated by ordinary notes, they amount to an ordinary bit of music, a phrase or perhaps part of one. If those same notes are indicated—as they could be—by a mordent sign on the note c, then that entity is an ornament. Likewise, the three notes c–d–e could be indicated by three ordinary notes or by the note e with the sign for a slide, or the notes d–c–d–c–d–c by six notes or by a trill sign over the note c. This is basic and well known. So, what is the difference between notes indicated by an ornament sign and the same notes written out? Sometimes there might be little or no difference. In fact, there are plenty of pieces in the repertoire with parallel passages in which the same notes are one time written out and another time marked by ornament signs, with no reason to believe that they should be different one time from the other. (This may make it appear that our distinction between ornaments and other notes is at least sometimes arbitrary. This is true, and actually can be helpful in teaching students to play ornaments well and to be comfortable playing them. A significant part of the fear of ornaments comes specifically from identifying them as ornaments, as something other than just some notes to play.) However, when there is a difference, it is likely to be that notes indicated by ornament signs are meant to be quick and light or to deviate subtly from any rhythm that could be spelled out by notes in our rather simple system of rhythmic notation, or both of these.
In fact, from the point of view of execution or performance, ornaments are simply “quick, light notes” or perhaps the greatest exemplar of that kind of passage. This means that playing ornaments well can be achieved by applying the same kind of light, fluid touch that is in fact best for playing any note patterns on the organ. It also means that working on playing ornaments effectively can be one of the best ways of improving lightness of touch and freedom from tension in all playing.
There is an exercise that I use with students that I refer to as a trill exercise. It is extraordinarily effective at helping a player to develop the right kind of touch for playing trills. However, it is equally useful for teaching a light touch for any kind of fast playing, including both non-trill ornaments and any other kind of rapid passage. (It is in fact the only actual exercise that I normally suggest to students, given that in general I believe that it is best to practice pieces or note patterns drawn from pieces.)

A trill exercise
This exercise is not written in music notation, and, although it involves playing notes at the keyboard, it is really a kind of relaxation/breathing/meditation exercise. It can be carried out at the organ or at the harpsichord. It can also work on the piano, as long as the player remembers not to care about producing a robust or loud sound. It goes like this:
1) Sit at the keyboard, and identify the place on the keyboard where each hand can meet the keys with the arm, wrist, hand, and fingers more or less in a straight line. This is usually at the notes written near the top of the treble clef for the right hand and at the notes near the bottom of the bass clef for the left hand, though it varies a bit from one person to another. It is fine to let the elbows float away from the sides. If you are sitting at an instrument with more than one keyboard, choose the keyboard that it is most natural and comfortable to reach.
2) Pick two adjacent (natural) notes and two fingers. Initially it is a good idea to use adjacent and “good” fingers, perhaps 2–3 or 3–4. Later it is fine to do the exercise with any pair of fingers with which you might ever want to play a trill. The two notes should feel as similar to each other as possible. On many organs this is not an issue, though it is on some. It certainly is an issue on many harpsichords. Choose a quiet registration: a Gedeckt or Dulciana, perhaps, or, on a harpsichord, one 8-foot stop by itself.
3) Play one of the notes lightly and comfortably with the finger assigned to it, and hold the note. While holding this note, let your hand relax as much—as thoroughly—as you possibly can. This can be aided by moving the arm around a bit in the air—still holding the note—or by flexing the wrist a little bit, up and down, or by taking calm deep breaths. When you feel that your hand is fully relaxed:
4) Play the other note and then the first note again, as quickly and as lightly as you can. As you do this, you should have as little as possible of a feeling that your hand is bearing down into the keys. Instead, it should feel as if the hand is almost floating up and away—just failing to do so enough to allow the fingers to play the notes that they are trying to play.
5) After you have played these two rapid notes, you will notice that your hand has lost at least a little bit of its relaxation, that it has picked up a bit of tension or at least a bit more muscle tone than it had just before playing those two notes. So, the next step is, while continuing to hold the note that you are holding, again wait for your hand to become fully relaxed. You should then repeat the process described above, that is, the rapid playing of two notes. It can be repeated several times—four or five, maybe up to a dozen. It should never happen according to a beat or a schedule. Each time, while holding the note chosen as the first note, you must wait until your hand is perfectly relaxed before executing the rapid two-note gesture for the next time.
6) After doing this several times in a row, do the same thing but start with the other note and the other finger.
This exercise should be done with each hand, with various combinations of fingers. It is not a good idea to segue directly from doing this exercise to practicing or playing a trill or any other note pattern. Rather, it should simply be done by itself, perhaps for ten minutes or so at some point—or at two different points—during each practice session. Then, when actually practicing or trying to play a given trill (or other rapid ornament or other rapid passage), the idea is to remember and recapture the feeling in the hand, wrist, arm, shoulders, and body that you experienced during this exercise.
I have never known this exercise to fail to help a student, or any player, beginner or advanced, who spent some time with it. It can be used not just to develop a better feeling for the touch of trills, but also to train recalcitrant fingers to play trills and to play rapidly with control. In particular, it is very fruitful to do this exercise with 4–5, after having first done it with more “normal” trill fingers. Almost everyone I know believes that he or she “can’t” play trills with those fingers. In fact, almost everyone can after having applied this exercise to the task.
(I should mention that the original idea behind this exercise was suggested to me by my friend the late David Margeson in the early 1980s when he was a graduate student in organ at Yale. I have refined the idea and adapted it somewhat to the specifics of organ and harpsichord.)

Fingerings
A real necessity in playing ornaments well is planning good fingerings. This has several elements to it. First, of course, is choosing fingers for the notes of the ornament itself. In spite of the claim I made just above, it is a good idea to use the “best” fingers whenever possible. For most people, these are the middle three fingers, or indeed specifically 2 and 3. It is a good idea to use whatever fingers the player is most comfortable with—why compound difficulty by not doing so?—but it is also important not to be so tied to those fingers that passages before and after an ornament end up suffering from convoluted and unnecessarily difficult fingerings. For example, a player who can only play trills or rapid mordents with 2–3 will frequently get into trouble of this sort. A player who is also comfortable using 3–4 will get into much less trouble. Fingerings such as 4–5, 1–2, 1–3 are also useful, though the actual need for them arises less often. A consideration in choosing fingering for an ornament should always be the effect of that fingering on hand position and, in particular, the ability of the player to keep the fingers from migrating too deeply into the keyboard. So, for example, if one note of an ornament is a raised key and the other a natural, then it is wonderful to end up playing the raised note with 3 and the natural with 2 or 4 as appropriate. Reversing this leads to some kind of awkward hand position, and thus makes it harder to maintain a light, comfortable touch. The logistics of this vary at different points along the compass of the keyboard and also from one player to another depending on the relative lengths of the different fingers. The important thing is to remember to pay attention to the hand position that results from a fingering choice with an ornament.
If the note immediately before an ornament is the same note that actually begins the ornament, it is very important indeed to play the two successive iterations of that note with different fingers. This is an approach that I always prefer with repeated notes (see The Diapason, January 2009), but for preparing ornaments it is especially crucial. This is because, again, a light touch and a relaxed hand are absolutely essential to playing ornaments in a way that feels and sounds good. It is very common for a trill that should begin with the upper note to be preceded by that same note. The best way to work out this fingering is to decide first on the best fingering for the trill, based on the player/student’s preferences and on the logistics of the particular notes, then select a finger to play the preceding (same) note from among the fingers not designated to play the first note of the trill. This choice should be made based on the shape of the passage leading into the trill. If it is impossible to make that passage work without using the same finger for the final note before the trill and for the note that starts the trill itself, then the trill fingering should be changed if at all possible. I have very rarely indeed been unable to devise a good solution in this very common situation—perhaps never. The point for the teacher to make to the student is that it is both fairly easy to work this out and abundantly worth doing so. Awkward starts to trills are usually the result of simply not having thought out the fingering both of the trill itself and (especially) of the notes leading into the trill.
It is also very common for the note before an appoggiatura to be the same as the note of the appoggiatura itself. In this situation, using different fingers for the two iterations of that note will not only make the whole pattern of notes sound more natural and give greater control over timing and articulation, but it will specifically create the right accent relationship amongst the three notes: the note before the appoggiatura, the appoggiatura itself, and the note following it (the “main” note). Trying this example with each of the indicated fingerings (both for the right hand) can make this difference seem clear.
One of my frequent chamber music colleagues recently made the following comment to me about a (non-keyboard) musician with whom we both play a lot: “I’ve figured out why so-and-so’s ornaments always sound so good. He plays them quietly.” The above discussion about fingering and the suggested exercise are essentially aimed at helping students to develop an organ and harpsichord equivalent of playing ornaments quietly. Next month I will deal with the sometimes vexing questions about what the notes and rhythms of ornaments should be—on the beat or before, starting on main or auxiliary notes, and so on. I will also address how to use that information to help students feel freer in their playing of ornaments rather than more constrained, and how to help them approach the subject creatively. 

It has always struck me as interesting that changing fingers on repeated notes and substitution are so similar to one another in what they actually involve physically. Thus it makes sense to me to use one of them to introduce the other. It is also important to keep them straight: it is extremely common for students to fall into the habit of doing a substitution when they think that they are changing fingers from one note to the repetition of that note.   

This continues without a break from last month’s column.

Gavin Black is Director of the Princeton Early Keyboard Center in Princeton, New Jersey. He can be reached by e-mail at [email protected]. He writes a blog at www.amorningfordreams.com.

A further practice step is to try patterns in which the hand plays more than one note at a time (Examples 1–4). The fingerings given above and below each line are alternates. There are other possibilities, for example, involving pairs such as 1/4 and 2/5. You can adapt these exercises in ways that occur to you, such as using black notes. Once again you should experiment with articulation. You can make non-repeated chords legato, and repeated chords any degree of non-legato; or try to match, as nearly identically as possible, the articulation of each of the motions from one chord to the next; or use a variety of non-legato articulations. Keep the hands light and relaxed, especially while releasing notes. Pay attention to the direction in which you release each finger when another finger is preparing to play that same note: up, down, slightly (or fully) to one side or the other. These logistic possibilities all have their place. They work out differently for players with varying relative finger lengths, and also for varying note patterns. It is your job to pay attention as you work on these exercises and figure out the most comfortable ways.

Repeated notes often occur in the context of ornaments, especially trills. The exercise in Example 5 allows you to practice that, assuming that you start each trill with the upper (auxiliary) note.You can play the opening note with 3 and each of the trills in succession with 4-3, or play the opening note with 2 and each of the trills with 3-2, or perhaps other patterns. You should adapt this exercise to other specific note patterns, including some involving black notes and the left hand. Do not worry about making the trills especially long or fast: the focus of practicing is the repeated note that initiates each trill.

Another ornament-based repeated note exercise, involving mordents, is shown in Example 6. You can play each quarter note with 3 and each mordent with 2-3-2, or other fingering patterns. For the purpose of this exercise it is only necessary that the final note of each mordent be played with a different finger from that which you want to use to play the following quarter note. Again, adapt this exercise to different specific note patterns and to the right hand.

Playing repeated notes with different fingers, in addition to giving the player more control over the timing, articulation, and sound of the repeated note patterns, also gives the player a free chance to re-position the hand. It can actually clarify and simplify fingering patterns for the passage around the repeated notes themselves. The excerpt from Rameau shown in Example 7 (part of the fifth of six variations on a Gavotte in A Minor) is an example of this, so extreme that if Rameau hadn’t written it, anyone discussing the fingering of repeated notes would have had to do so.

For all players except those with the very largest hands, changing fingers on the repeated notes in each group of four sixteenth notes is actually necessary to permit the playing of the other sixteenth notes. The same is true in the left hand sixteenth note pattern in the sixth variation from the same piece (Example 8).

But in being necessary it also guides the shaping of all of the rest of the fingering in such a way that the passages are actually quite natural and straightforward to play. Each decision about what fingers to use on the first and second notes of each pair of repeated notes should be based on where your hand is coming from and where it is going. Example 9 shows one possible fingering for the left hand part of the preceding example.

In Example 10, from the Brahms chorale Mein Jesu, der du mich, there is a moment, at the beginning of the second full measure, where the use of a different finger on a repeated note makes it possible to set up a simple and effective fingering for the succeeding passage. (My suggested fingering is not the only way to do it.)

The musical advantages of using different fingers to play repeated notes can only be heard and felt if the hand is very relaxed and the touch smooth and fluid. Any repeated-note moment (such as the one in this Brahms example) is a good place to remember, recapture, and apply the feeling of lightness gained from the trill exercise described above. 

Substitution

As opposed to changing fingers on repeated notes, the technique known as “substitution” is changing fingers on held notes. While these two techniques serve very different musical and technical purposes, and indeed are most typically associated with different historical periods and repertoire, they have so much in common technically as to be essentially versions of one another. 

There are several things to bear in mind when beginning to work on substitution:

1) A substitution can be either measured—the new finger placed silently on the note at a predetermined time, probably defined in relation to the beat of the passage, or instant—that is, the new finger slides in to replace the original finger as part of the gesture whereby the original finger played the note in the first place. (Whereas the timing of finger change in a repeated-note passage is determined by the timing of that passage’s notes.) A substitution can also be somewhat in-between: that is, not instant, not a one-gesture slide, but not specifically timed to be on a beat or subdivision of a beat. This latter is probably the most common in practice, though all are quite useful.

2) In any substitution there is likely to be something to observe about the specific direction in which the original finger departs and the direction from which the new finger arrives. It may make sense to get the original finger out of the way by lifting it up, moving it sideways, allowing it to curve downward, or something else, or some combination. The new finger can slide in under the old, or from above it, or from one side or the other. All of this affects or is affected by hand position and by the relative lengths of the fingers. It is not—since the substitution is silent in any case—something that affects the musical results. It is about comfort and reliability. 

3) Substitution is generally associated with legato. The usual reason for introducing an extra gesture into the act of playing is to permit the hand to be in a position to play the next note or notes without having to release the existing note(s) in a way that creates an unwanted break. Sometimes, however, substitution simply seems to make a passage easier. Different players develop different degrees of comfort with substitution and use it to differing extents. 

4) Substitution is more likely to be necessary or to provide an appropriate solution for creating true legato in situations in which a hand is playing more than one note: counterpoint or chords. In single line textures, substitution is rarely necessary to effect a particular musical result. (When it is necessary, that is usually a result of something having to do with very wide intervals.) That is not to say that it is not often comfortable or convenient. Sometimes it can serve the same function as changing fingers on repeated notes in that it can allow the hand to reposition itself efficiently.

5) Substitution—unlike most of what most performing musicians do while playing—creates physical gestures that do not correspond to anything that the player or the listener actually hears. This can break or weaken or generally interfere with the player’s ability to experience the rhythmic vitality of the music through the kinesthetic experience of playing. For some players this sense—almost of dancing to the music while playing, but doing so with the playing gestures themselves rather than by literal dancing—is a real and valuable aid to vivid and convincing performance. If the feeling that the hands (and perhaps feet) are doing things that aren’t part of the rhythmic flow of the music seems, to a particular player, like a problem, then that player might well be inclined to use substitution less than other players. There are also ways of counteracting or compensating for that effect. At an early stage of learning organ, and of becoming comfortable with substitution, this is something to file away at the back of the mind, in case it seems like an issue to be dealt with later.

6) Sometimes a tendency to rely on substitution as an all-purpose way of finding notes (scrambling for notes, in effect) can lead a player—whether a student or otherwise—to cut short the process of working out good, efficient fingerings and then practicing those fingerings with enough focus and dedication to learn them. In this way, a heavy reliance on substitution—especially by a beginning or “intermediate” student—can actually damage the learning process, sometimes seriously. This is far from being a reason not to learn and work on substitution, since it is a valuable tool, and for some purposes a necessary one. It is simply something to watch out for.

The second exercise given above for changing fingers on repeated notes can be adapted as a good beginning point for practicing substitution, simply by tying the repeated notes, and keeping the fingering the same (Example 11).

And this same note pattern can be used with an extremely wide variety of fingerings, since in principle any substitution is possible and is worth practicing. For example, the right hand fingering could be 3-4-5(1)-2-3-2-1(5)-4-3(5) (The parenthesis indicates substitution. In this fingering, the tied g’ going from the second to the third measure does not have a substitution.) Another possibility would be 1-2-3(1)-2-3(1)-4-3(1)-3-2(1). These fingerings are musically random: their purpose is to help you get the feeling of different substitution patterns.

The two-note chord exercises above can also be adapted as substitution exercises (Example 12). With the same-note chords tied, the fingerings would be carried out as substitutions rather than as changes of fingering on newly played notes. This can be tried with other specific fingerings, and other similar note patterns, and of course also in the left hand.

In carrying out substitutions with multiple notes, it is important to do the individual substitutions in the correct order. The correct order is the one that is the most comfortable and natural physically. (Again, since the substitutions are silent, this is about physical comfort and reliability rather than any audible result.) Usually that means the order that keeps the hand small: that doesn’t stretch the hand out any more than necessary. So, in the example above, the substitutions on the lower notes of the two note chords should be done first. It is always possible to figure out by trial and error which way is best. Sometimes it is also possible to figure it out in advance by analysis of hand position. Performing multiple substitutions in the correct order also has the effect of allowing the whole hand to move in one gesture towards the next note or notes or towards its next position. It is extremely important to get this right. That can make the difference between a substitution’s being easy and natural and its being both difficult and a potential source of strain or even of real injury.

Next month’s column will continue with more exercises for substitution and examples drawn from the repertoire.

Gavin Black is director of the Princeton Early Keyboard Center in Princeton, New Jersey. He can be reached by e-mail at .

Repeated notes
The playing of repeated notes on organ and harpsichord has always been an issue unto itself. If two notes in a row are the same, they cannot be treated like two notes in a row that are not the same. The reason for this is simple: in order to repeat a note that you are holding, you must first release it. This seems so obvious to those of us who play only these instruments that it is worth noting that this is not true in all kinds of musical performance. It is not true at the piano, except in situations that rule out the use of the damper pedal. It is not true with plucked string instruments. In singing, the repeated note phenomenon is only rarely an issue in itself. With bowed string instruments and most wind instruments, the relationships among articulation, technique, and pitch are complicated, with repeated notes as such only sometimes being a special concern.
One way to describe the situation with repeated notes at the organ or harpsichord is this: in general, any pattern of notes that doesn’t involve repeated notes can be played legato (though of course it doesn’t have to be), but repeated notes actually cannot be played legato. Therefore, patterns of non-repeated notes have, in theory, the full range of articulation available to them, from “as short as physically possible” to a full overlapping legato. Repeated notes have most but not all of that range of articulations available.
Since repeated notes cannot be (fully) legato, the more legato the overall style of a given performance is—whether because of the performer’s preference, or because of something that is known about the composer’s own style—the more any repeated notes are in danger of standing out, of sounding different at the very least and maybe stylistically wrong, and in any case amounting to a problem to be solved.
This, in turn, may be one reason that repeated notes have often been considered a problem—or again at least a particular issue that needs to be addressed—in hymn playing, since there is a strong tradition of playing hymns legato. Repeated notes are sometimes seen as a source of a disruptive choppiness in hymns, and thus, for some players in some circumstances, are considered worthy of being eliminated through tying.
In addition to obvious repeated notes—instances of the same note occurring two or more times in a row in one melody or one voice—there are various kinds of hidden repeated notes. These arise from voices crossing or from one voice playing a note that was just played by another voice or that is being held by another voice. They can also arise because of ornaments—when there is no repeated note printed on the page, but one arises from the notes implied by the ornament sign.
Of course, repeated notes occur in all sorts of rhythmic contexts. Sometimes the first note is an upbeat and the second a downbeat, sometimes the other way around; sometimes they are two successive weak or light beats, sometimes two successive downbeats. (Of course there are chains of more than two repeated notes in which more than one of the above may occur in succession.) Repeated notes can be fast or slow.
In all of these circumstances the same underlying fact applies: it is necessary to release the first note before playing the next one. It is certainly possible, and often necessary or a good idea, for a student or other player to think analytically about how long or short to make any note that is about to be repeated and to think about how the articulation and timing allows it to fit in to the rest of the music. This has been the subject of extensive discussion, analysis, and debate by teachers and players over many years. For example, David N. Johnson has a detailed and interesting discussion in his Instruction Book for Beginning Organists. Marcel Dupré is famous for having described a very clear-cut system for counting out the amount by which notes should be reduced prior to being repeated. (Perhaps I should say “infamous” since his system is widely considered to be too cut-and-dried to be artistically valid. However, it is worth remembering that he almost certainly intended his guidelines to be a stage in learning, not an end result.)
Rather than suggesting specific musical answers to repeated note issues, I would prefer to begin by helping students to do two things: first, to develop the greatest, most comfortable, and most reliable technical control over the physical act of playing repeated notes; and second, to develop the habit of listening closely to every part of any repeated note transaction—the articulation prior to the first note, the beginning, middle, and end of the first note, the space between the notes, the beginning, middle, and end of the second note, and so on. Once a student has made good progress on these things, then he or she will be able to make choices about how to play repeated notes in various different contexts, and these choices will be able to reflect the whole range of possibilities.
There is, I believe, a simple key to developing the greatest possible technical command of the playing of repeated notes: play them with different fingers, one from the other. That is, if you have played the first note with finger x and are holding it with finger x, then it is appropriate to play the second note (that is, the repetition) with any finger other than x. It is not OK to play it with x. This means that a note repeated more than once can be played with fingers x-y-x-y etc., or with fingers x-y-z-a-b-c etc., until the fingers run out, but not, again, x-x-x-x etc.
When a player repeats a note with the same finger that is holding it, that finger must travel both up, off the key, and back down, to play the note again, in the time that makes up the space between the two notes. This sets up a conflict between making that space short—playing the notes close to legato, at least—and executing the gesture comfortably. If the physical gesture involved is not comfortable, then the musical gesture will almost certainly sound awkward; playing a repeated note with the same finger greatly reduces the extent to which the gesture can come across as musically continuous. That is, either the repetition will have a large enough space between the notes to sound significantly disconnected, or it will have an awkward “hiccup” quality caused by an effort to push the two notes as close together as possible. The part of the “staccato to legato” spectrum that is unavailable to repeated notes intrinsically—because of the nature of the instrument, as discussed above—is made artificially greater by playing the notes with the same finger, and the range of possible, successful, articulations is narrowed.
It is also true that the act of moving one finger up and then back down is, among all of the gestures we make at the keyboard, one of the ones that is most likely to create tension in the hand. The “u-turn” that the finger makes at the top of that arc is a motion that is prone to tension. If it is not dealt with in some way, this tension can build up and, since essentially every passage of music has some repeated notes in it, this can lead to tense playing overall, even for a player who is consciously trying to play in a relaxed, light way.
In repeating a note with a different finger, the player can prepare the new finger in advance, and then release the initial finger smoothly while bringing the new finger into position to play the note and then playing the note. This is an intrinsically smooth, relaxed gesture, and it can actually serve to reduce tension that might have begun to accumulate in the hand.
François Couperin wrote in his L’Art de Toucher le Claveçin that he could tell by ear alone the difference between a note repeated with the same finger and one repeated with different fingers. (This was in the context of the playing of ornaments, which I will discuss briefly below.) When I first read that claim, years ago, I thought it was more or less impossible: that it was probably an exaggerated boast by someone whose eminence was great enough to permit him to get away with it. I would now make that same claim: I believe that, except in rare circumstances, I can detect that difference just by listening.
Once any teacher, student, or other player begins to be able to hear that difference, the motivation to work on playing repeated notes with different fingers follows automatically. Fortunately, it is an extremely easy thing to do. It is no harder, by and large, than playing those notes with the same finger. In fact, once it becomes second nature, then the fact that it is easier—that is, smoother, more natural—physiologically, makes it seem easier as a practical (and psychological) matter.

For a student to get accustomed to the feel and sound of repeated notes played this way, the best exercises are simple enough that they scarcely need to be written out (see Example 1). In this example, the student can play the notes at a variety of different tempos and with a variety of different fingerings: all the notes with any one finger (for comparison); pairs such as 2-3, 3-2, 3-4, 4-3, 3-1, or any others; or chains of fingers such as 2-3-4-5, 1-3-4-3, etc. The student should also experiment with repeating the same note but changing the rhythmic grouping. This can be done such that rhythmic groupings correspond to fingering patterns (that is, a duple grouping with a paired fingering such as 3-2, or a four-finger pattern such as 2-3-4-5; and a triple grouping with a three-finger pattern such as 4-3-2 or a six-finger pattern such as 2-3-4-5-4-3). Or it can be done with rhythmic groupings that are different from the fingering groups, such as a triple grouping with a paired fingering. In this case, the downbeat of each group shifts a finger from one time to the next.
It is very important to remember that repeating a note with a new finger does not mean slipping the new finger onto the note silently while still holding it and then repeating it with that (new) finger, which is now holding the note. This is a temptation—probably subconscious—that many students experience. Of course this is identical to repeating the note with the same finger: the supposedly new finger has become the incumbent finger.

Further exercises can put the experience into a musical context. These can begin with something simple, such as Example 2. This can be fingered in a number of ways, such as 2-3-4-5-4-3-4-3-2-3, or 3-4-5-4-3-2-3-2-3-4, or (again, for comparison) 2-3-4-4-3-2-2-1-1-2. The student should remember to keep everything as light, relaxed, and supple as possible. (It is possible to lose the advantages of using different fingers on repeated notes by playing with stiffness or tension.) The student should try different articulations: for example, making all of the non-repeated notes legato, and the repeated notes as smooth as possible; or making everything lightly detached so that the repeated notes are not articulated any differently from the rest of the line.
A chord pattern such as that in Example 3 can be tried with various fingerings, such as RH: 1,3,5/2,3,5, or LH: 5,3,1/4,2,1, and, for comparison, RH: 1,2,3/2,3,5, and LH: 5,3,2/3,2,1.
In Example 4 there is a hidden repeated note. If the two middle-Ds are played with the same finger, it will be difficult or impossible to make the two voices clear. The final quarter-note of the first measure will sound like a released and repeated note in the lower voice. A fingering such as 5,2/3/1//4,2 or 5,1/3/2//5,3 will make it possible for the middle-D to sound like it arises from the upper voice. This comes about because the necessary early release of the whole-note D can be smooth and unobtrusive. In this example, it would also work well for the left hand—any finger—to play the whole note, and for the right hand to play all of the other notes.
In many ornament situations such as this common one in Example 5, there are hidden repeated notes (assuming that the trill starts on C). A prudent way to work out a fingering here is to decide first of all which fingers should play the trill—say 3/2—and then to make sure that the note immediately before the trill is played with a different finger, say 4 or 2. Many problems that students (and others!) have playing ornaments are in fact problems with setting the ornaments up correctly. If, in this example, the student plays the C with the third finger and then repeats the C with that same finger as the first note of the trill, the attempt to play the trill will be undermined by tension before it has even begun. If the eighth-note C is played with 2, and the C that begins the trill is played with 3, then the trill will get off to a lighter, more fluent start.
Students and teachers can invent exercises to try different repeated-note fingerings, and can extract repeated-note situations from repertoire to use as exercises, before going on to finger and practice such passages in their original contexts. It is important to try different fingering, including those same-note fingerings that I would not recommend, in order to learn what the differences are between them. After a while, if a student finds the approach described here convincing it, becomes second nature, and, if anything, extra thought is required to play a repeated note with the same finger. (I sometimes need to do this as a demonstration, and I often fail to do so, out of habit!)
Sometimes a note pattern is such that it is actually impossible to change fingers on repeated notes. This is because the relevant fingers are doing something else. When this happens, then a student can draw on what he or she has learned through practicing the technique described here to be aware of what the goal is—in both feeling and sound—for those repeated notes. That awareness gives the student the best chance of coming close to achieving that sound or feeling even when the best technique for achieving it is not available. This can involve first isolating the repeated notes from the rest of the texture and practicing them separately with a good different-fingered fingering. After this, with all of the notes back in place, the memory of what the repeated notes would ideally sound like—and a generally very relaxed, smooth touch—will enable the student to get the best results under the circumstances.

Hand Distribution III

Continuing our trek through the Alla Breve section of Bach’s D Major Prelude and Fugue, BWV 532—looking closely at issues involving hand distribution—we come to a brief section that is influenced by something other than the music itself:

Example 1 shows that if there were nothing else to think about, clearly there is reason not to distribute the two voices between the two hands. That is the first principle of hand distribution, after all. However, in most editions of this piece, there is a page turn right about here. Therefore, the player can gain a bit of ease with that page turn by taking all of these notes in one hand (most likely the right hand). It is entirely possible that the various editors have chosen to position these measures at a page turn in order to help out in this way. Of course, for a player who memorizes the piece this won’t matter in the long run, but it might still help during the learning process. 

This is a special case—sort of a diversion. In fact, analyzing it like this is a useful way to help a student to relax: talking about something practical and not artistically intense, but relevant. However, it is not an unreal concern, and there are other reasons for taking clusters of notes in one hand in order to deal with something else while playing. The main one is probably the act of changing stops. Even something as simple as grabbing all of the notes of the final chord of each verse of a hymn in one hand to change stops with the other is a branch of decision-making about hand distribution.

The rest of this Alla Breve section mainly presents the same issues that we have already seen, with perhaps a few twists. I will go through it all, but concisely, since it is more or less “review”. 

The next short passage (Example 2, measures 48–49) has an outer voice that is more active than the other voices. Therefore it will make sense to keep that voice by itself in one hand, for the most part. Some players may want to break up the middle voice by taking the d′ at the end of measure 48 in the right hand. There may be other modifications that could make sense, but tracking the entire middle voice in the right hand would significantly increase the difficulty of the passage.

Example 3 (measures 50–51) shows the next measure, which has an intricate middle voice. All of the notes of that voice can be reached by the right hand; however, it might make sense to take some or all of those notes that can also be reached by the left hand, to break up the physical act of combining that line with another. The candidate notes are probably the opening c#′, and the b and the a in measure 50.

The next, which is fairly lengthy, has the fast-moving figures in the upper voice. However, the two slower voices are not close enough to one another to be taken in the left hand, clearing the right hand just to track the intricate line (see Example 4, measures 52–59).

For most players, the easiest and most natural way to finger the passage will involve taking in the left hand all of the middle-voice notes that the left hand can actually reach, and taking the notes that the left hand cannot reach in the right hand. On the second quarter note of each of the odd-numbered measures, where the two higher voices come together, there is a special issue to think about. Which hand can best project to the listener the illusion that this is two notes, one of which moves away as part of the upper eighth-note line and one of which is part of the middle-voice quarter-note line? It is actually a trap in a spot like this to try to play the note with two fingers at once, one from each hand. (No one would suggest this on purpose, but students will indeed fall into doing it, probably through indecision.) The choice of hand (and finger) should be made clearly, even though it can be made either way.

For most of the next nine measures, there are no real questions to think about, either because the (manual) writing is in only two voices or because the balance of more intricate and simpler writing makes it clear. 

At measure 65 (Example 5) there is an interesting subtlety to examine. The middle voice takes over the note being held—presumably in the left hand—by the lower voice. Which hand should play the note? The left hand is right there, but with the “wrong” finger—since whatever finger is holding the note, the hidden repeated note will sound better if it is played with a different finger. This is not hard to manage. The articulation and timing of the move from the c#′ to the a in the middle voice might seem to be under more natural control if both those notes are played in the same hand. However, it is entirely possible to practice towards making that gesture effective across the two hands, as I will discuss below. It might seem better to take that eighth- note a with the left hand to give the right hand more time to get up to the c#′′ on the second quarter note of the measure. However, to me that “leap”—the tenth from the a to the c#′′ over the time-span of an eighth note—is the main reason to take the a in the right hand. The physical gesture of moving the right hand up the distance of that tenth will—like a bowing gesture in string writing—give the player the best chance of shaping the articulation and timing of the musical gesture in an effective and natural way.

At measure 69 (in Example 6) there is a brief passage in which any and all of the notes of the middle voice could be taken by either hand.

This is a good spot at which to remember once again that it doesn’t matter on which staff the notes are printed. The decision about which middle-voice notes to take in which hand should be based on comfort and logistics. This is not a bad time to mention that this will vary with the particular hand shapes of different players. For example, it is quite likely that a player with relatively short thumbs will gain more comfort from taking the d′ in measure 70 with the left hand than a player with relatively long thumbs will. 

Measures 71 through 78 display a texture in which the upper voice is mostly holding long notes, while the other two voices are fairly active. A sample of that passage is shown in Example 7.

It makes sense to take the eighth-note middle voice in the right hand, just accepting that one finger (the fifth finger) of that hand is unavailable since it has to hold long sustained notes. 

At measure 79 there is another opportunity to use hand distribution to make the playing of repeated notes sound natural, and to avoid letting those repeated notes disrupt the flow of the voices. My suggestions are indicated by letters, and are shown in Example 8.

The next complicated or involved spot begins at measure 89 (Example 9). This is a longer example of the sort of writing found at measure 36 and discussed in last month’s column. In this case, however, the eighth notes in the middle voice can all be reached by either hand. The player has a free choice as to which hand should play any of these notes and therefore what pattern to follow through the passage. The teacher’s role is mostly to point this out to the student, and to help the student notice the implications of different choices for hand position and articulation (and of course the implications of articulation preferences for hand distribution choices: the more interested a student is in playing the upper half-note line legato, for example, the more middle-voice notes the student will want to take in the left hand). I myself would probably take the third eighth note of each beat in the left hand—those that are a third higher than the lower voice left-hand notes, closest to them—and the others in the right hand. There are other ways to do it.

The next few measures (Example 10 measures 94–96) end the section of the piece that we are analyzing. Again, either hand can reach the middle-voice notes. Choices can be made based on the usual factors: closeness of notes to one another, hand position, and so on. However, this passage also has a special feature. A player might find that the shaping of the timing and articulation of the syncopations/suspensions in the upper voice feels more natural either 1) with those notes isolated on their own in the right hand, or 2) played with the middle-voice notes in the right hand, using a kind of rocking motion to reinforce the feeling of the pacing and articulation of those notes. This is an individual thing: I can easily imagine doing it either way.

When a student (or any player) has made all of the decisions about which hand should take which notes of a (middle) contrapuntal voice, and worked out the actual fingering, then the next step is to practice the passage in such a way that that voice sounds the way that the player wants it to sound. If the hand distribution and fingering are right (comfortable) then this should not be categorically different from practicing any other sort of passage. 

However, there is one concern. It is undeniably a little bit more difficult—or perhaps it would be more accurate to say “less intuitive”—to shape the timing and articulation of the transitions from one note to the next in a contrapuntal voice when those notes are in different hands than when they are in the same hand. It is very important not to let this fact lead a player into making awkward hand distribution choices. (Sometimes it can and should influence those choices when other factors are fairly evenly split). But it should be kept in mind and addressed in practicing. 

The main way to address it is to practice that voice by itself, but split between the hands with the correct, worked-out fingering. This is partly physical practice, but even more it is listening practice. It is easiest to attune the ears to the flow of the line when the line is not covered by other notes, and this will make it easier to hear and follow the line in the context of the full texture. It is always straightforward to extract the line once the fingering process has been accomplished. It can be a good exercise for a student to write out—or type out—the line by itself, add the chosen fingerings, and practice it from that. However this is probably not necessary. 

For the bulk of this practicing it is important not to change the chosen fingering—and it is crucial not to do so accidentally or at random. (It is always OK to rethink fingering consciously, if there is a reason to do so.) It is also important to listen carefully during this practice to the transition moments, where the voice crosses from one hand to the other. It is possible, especially with a line that is physically not hard to play, to play short sections of the line in one hand at this stage to listen for the continuity, and then put it back into the correct (two-hand) fingering, trying to match the one-handed effect. It is probably a good idea not to do very much of this: just once or twice through a given short section of the line being practiced. If a student finds this to be disruptive (that is, if it is hard to go back to the fingering that is really being practiced after visiting the one-hand fingering) then he or she should not do it. 

When a student has practiced a line this way and is ready to put the whole texture back together, he or she should try at first to listen only to the line that passes between the hands and to pay no attention to the voices around it. (Unfortunately, it is impossible by definition to solo out this line, since in all of the passages of the kind that we have been studying both hands and all the voices are—and have to be—on one keyboard.) This is an exercise in focusing, and of course it can’t be achieved literally. You will always hear the other notes, but you should try to focus on the line that passes between the hands, to be conscious of that line and the sonority of all of its notes.

It can be a good exercise to take any line of music—say the top line of a hymn, or one voice of a two-part Invention, or a cantabile melody from the slow movement of a Mendelssohn sonata—assign it an arbitrary fingering that shifts back and forth between the hands, and practice that fingering. (The fingering can be worked out arbitrarily, but should be written in and not changed at random.) This is not to end up playing that line that way regularly, but as training in listening to and executing the transitions from one hand to the other.

Often the issue is not that of passing a line between the hands. In non-contrapuntal music, the question of how to divide the notes between the hands (assuming, as always in this context, that the whole texture is meant to be played on one keyboard) should usually be determined as simply as possible by trying out the physical comfort, simplicity, and convenience of any of the various possibilities. In fact, very often, just remembering that it is perfectly all right to distribute the notes between the hands however they fall most easily is the most important as well as the first step. The rest follows from that. 

It is interesting that the impulse to play upper staff notes in the right hand and lower staff notes in the left hand can be pervasive. I recently took part in a conversation about the wide left-hand chord on the fourth beat of measure 8 of the Widor Toccata (Example 11). For many players, it is impossible (or nearly so) to play all four notes of this chord in the left hand, and for even more players it is at least awkward. The player who initiated the discussion absolutely could not reach those four notes. Nonetheless, the conversation revolved around such issues as which note or notes it was best to leave out, or whether there was a solution based on arpeggiation, or whether Widor’s left hand was really big enough for him to be able to play this chord easily and nonchalantly. 

It took a while for someone to notice the obvious solution, namely that the highest note of the so-called “left hand” chord is within easy reach of the notes of the upper voice, and can perfectly well be played in the right hand. Doing it this way opens up some performance issues similar to some of those discussed above. The timing and articulation of that note must be just right, as a match to the other notes of the chord. That is intuitive if all of the notes of the chord are in the same hand—and less intuitive, more challenging, if the notes are split between the hands. This is analogous to the issues involved in passing a voice back and forth between the hands. It is also important to keep the articulation of the top line going the way you want it while adding an extra note for the right-hand thumb. A player who absolutely cannot reach the entire chord can take on the task of practicing to get these things right. A player for whom the chord is possible but awkward can decide where the balance lies as to what is easiest and what will give the best results.