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].

Continuo, Part 3

The core of this month’s column is a description of the approach that I suggest for drilling and learning the actual—improvised—creation of continuo parts at the keyboard. The fundamental reason that it is better to improvise continuo parts than to play from a part—a realization—written out in advance is that the most effective continuo accompaniment is one that is flexible. Even at the last minute, but certainly during any process of rehearsal and preparation, it is important to be able to make basic decisions about what notes to play in response to things that we hear from the other players: dynamics, accentuation, intonation, and so on. The earlier in the process the notes are fixed once and for all, the less flexible it is possible to be. So, in playing, unaltered, a continuo realization written by the editor of a published version of a piece, we are committing ourselves to having no flexibility whatsoever during the rehearsal and performance process. Most published realizations are very thick—four voices most of the time—and, in the judgment of many players and listeners, too busy, too noisy. (This is especially true when they are played on organ or harpsichord. At the piano the busy-ness can be made less of a problem by simply playing the part more quietly.) But any realization that is created beforehand, even a wonderfully musical and sensitive one, lacks this flexibility. A player who works out a continuo realization during preparation and rehearsal, and writes it down planning to play it as is, has the opportunity to make it a good realization. But in this approach, last-minute flexibility is still lacking. 

(Actual last-minute flexibility—the ability to change the notes of a continuo part in performance from what they were even a short time before in rehearsal—can be desirable for several reasons. Some of these are: a change in the room acoustics with the arrival of an audience; an unanticipated change in the way a colleague is playing his or her part; problems in performance that suggest that you must project the beat more forcefully; and—most happily!—the fact that a new and better idea occurs to you.)

It also turns out to be easier in the end to learn how to realize continuo parts at sight than either to write them out in advance or to edit existing, published realizations to make them suitable for a given occasion. (And “suitable” still doesn’t take the idea of flexibility into account.) My own reason for plunging into studying continuo realization in the first place—about twenty-five years ago—was not anything artistically significant, but rather extreme annoyance with the mechanics of writing out parts for myself: it was boring, and it took too long.

In the decades following the disappearance of continuo playing as a living art, the notation and technique of continuo realization—figured-bass realization—was borrowed to fill various roles in the teaching of theory, harmony, and counterpoint. It is routine, almost universal, nowadays that anyone who has studied music theory at the college level has spent time learning how to concoct and write out realizations of figured bass lines. Because this activity is done in order to further the learning of something other than actual continuo playing, the kind of realization that is being sought is very different from what is best in performance. Specifically, in theory class, or a similar setting, it is almost always considered necessary to realize in a certain number of contrapuntal voices—probably ideally four, or three to make it easier. The rules of voice leading of course must be followed, and perhaps it is expected that each voice will be kept mostly within a certain range. Often this kind of exercise is presented in two alternate versions: one with all of the added notes in what amounts to the right hand—say, middle C and above—and the other with the four voices more or less evenly distributed, creating a hymn-like texture. In any case, again, all of the rules must be followed. It is (mostly) the need to avoid parallel fifths and octaves that can make practitioners of this sort of exercise tear their hair out. 

It is often their experiences with figured-bass realization in such a context that leads students to believe that it is almost unimaginably hard to play continuo at sight. After all, if something is so difficult and awkward even when you have all day to puzzle over it, to try different things, and to write it out, study it, and think about it, then it must be effectively impossible to do it off the cuff while other musicians are actually playing and expecting you to keep up. This logic is good, but the facts are wrong. What you do when actually playing continuo bears very little relation to the “figured-bass as theory-learning tool” activity, and is in some ways directly opposed to it. The last thing that is desirable in a “real” continuo part is, of course, that the number of voices remain always the same. That immediately and utterly prevents us from using the realization process to influence rhythm, dynamics, texture, and so on. That is, it takes away the very reason for the existence of continuo accompaniment. 

The process of actually learning to play continuo, therefore, does not go through the kind of theory-oriented figured-bass study that I describe above. That kind of study can serve a purpose similar to the reading exercise that I included in last month’s column, that is, to bring a student to the point of knowing the meaning of the figures with real immediacy and ease. (It is overkill for that purpose, in the amount of time and effort that it takes, but it does accomplish it.) For every aspect of learning continuo playing after the meaning of the figures is well established, work on “continuo as theory/harmony/counterpoint” is actually taking us in the wrong direction.

If a student develops a strong sense—simultaneously conscious and instinctive—of what constellation of keys on the keyboard any given note/figure combination is pointing towards, and this sense directs the fingers towards those notes without the need to think much about it, then that student can play continuo at sight. That is, when the student who can already pick up the exercise from the last month’s column and “look at the first note and say ‘F’, the second note and say ‘A, C, and E’, the third note and say ‘F and C’” can play those notes rather than say them, he or she can take on continuo parts from real pieces with other players also playing.

The most effective way to develop that sense goes like this:

1) Find a bass line with some figures. It doesn’t matter very much what the bass line is, although lines from harmonically dense choral or orchestral music can be harder to deal with than is ideal at this stage. Handel chamber music is one excellent source, among many. (A public domain edition can be found at this address: <http://216.129.110.22/files/imglnks/usimg/4/4d/IMSLP05632-Handel_19_Son…;. There are appropriate bass lines on more or less every page.) The bass line can come from a slow or a fast movement. For reasons explained below, this doesn’t matter at all. It need not be a complete movement of a piece or any coherent section, just some notes and figures.

2) Put this bass line up on the music desk of a keyboard instrument. For this purpose it doesn’t matter what instrument: harpsichord, organ, piano, electronic keyboard—anything with at least about four octaves of normal keys.

3) Prepare to play the line very slowly. Because the tempo at which you play this bass line and do this exercise bears no relation to anything about performing the piece from which you have extracted the line, it doesn’t matter what the tempo of that piece might normally be. Each note of the bass line must come along very slowly, regardless of whether it is printed as a whole note or a thirty-second note or anything else. For someone beginning this process, the notes of the bass line should come at a rate of no more than ten or twelve per minute. But that is just a guideline: slower is always fine; faster is also fine if it works.

4) As you play the bass line very slowly, try, for each note of the line, to play (in the right hand) some version—any version—of the notes suggested by the bass note and its figuring. Do not think about anything other than playing something that counts as the right notes: the playing equivalent of what you thought or said in doing the exercise from last month. Specifically, do not worry about the spacing of chords, the part of the compass of the instrument, or the nature of the transition from what you play with one bass note to what you play with the next. Do not worry in the least about parallel fifths or octaves or whether notes resolve correctly. 

5) If you cannot—more or less in tempo—think of any notes to add above a given bass note, simply move on. Do not worry about this. If, the first time through, you only add ordinary triads above the “8,5,3” notes—or even only above some of them—and nothing “fancier”, do not worry about this. 

6) After you have played the bass line and whatever notes you have added in this way once, do it again. Don’t increase the tempo. Try to add some notes where you didn’t the first time. Then, of course, do it a few more times. If it feels natural to let the tempo increase a little bit that is all right, but by no means necessary. However:

7) Do not play the same line more than several times. If after a while (four or five times through) you have not succeeded in providing right hand notes for all of the bass line, don’t worry about this either. The effectiveness of this drill does not depend on “solving” the entire bass line, but rather on developing a sense of spontaneity with those spots that you do solve. If you play over it too many times in a row, that sense of spontaneity will be lost and replaced by excessive concern for getting it all right. 

8) Choose another bass line, and do all of the above again. This can be another section from the same movement or piece, or something completely different. Practice this way with as many bass line passages as possible. Never stay with one of them so long that you feel like you know it and are simply repeating something that you have already learned: move on to another one. Try to use lines in different keys, but you need not seek out anything too unusual: two sharps or flats is far enough along the circle of fifths for now. If most of what you use is in keys with one or no sharps or flats that is OK. Just don’t stick to only one key. That can become a rut.

All of the details above are important, but clearly step 4 is the essence of this exercise. Here are a few more specific thoughts about how to carry out that step.

a) It is perfectly all right for the tempo of the bass line not to be entirely steady. (This is certainly different from most types of practicing.) It doesn’t exactly need a tempo, but only be not too fast. If you need to draw one note out a little bit longer to think about what to play over that note, that is OK, as long as it is only a little bit. If you are really, in effect, stopping to figure something out, then that defeats the purpose.

b) You need not play all the notes that you add at the same time as the bass note or together with one another, though as you do more of this exercise you should discover that you can add the relevant notes with or close to the bass note more of the time. Initially it is perfectly acceptable to do something like this: set a metronome to 60; allow each bass note to last for eight metronome beats; expect to play the added notes on or near the fifth metronome beat; use the last beat or two to begin to look ahead at the next note. The numbers are arbitrary; the principle of keeping it slow and careful is crucial.

c) If you make certain kinds of mistakes about what the figuring means or what notes would be appropriate to add over a particular bass note, this doesn’t matter! One extraordinary thing about this exercise is that it usually leads a student to the right place even if it is done wrong. The most common way that this comes up has to do with un-figured notes. If you mistakenly assume that a passing tone is not a passing tone, and therefore add chords to bass notes that are not supposed to have anything added, this just constitutes more (fully useful) practice. If you interpret as a passing tone a note that really should have something added, and don’t add anything, that is a very minor wasted opportunity. It doesn’t mislead or do any harm. If you forget, for example, that “7” usually implies “7,5,3” and just play the pitch seven degrees above the bass note, that is still useful practice in developing the spontaneity that we are looking for. There is time to refine and fill in gaps in your awareness of what the figuring means and what the abbreviation conventions were later on. 

d) Likewise, leaving out things that are too complicated or unexpected—for example a figuring like “9, 7#, 4, 3b”—is not a problem. You have simply utilized one less practice note: no harm done. Reading really elaborate, complicated, counterintuitive figures can come later. In any case they are extremely rare. It is of course OK not to leave them out, but only if they are accurate and don’t slow the process up very much.

e) Of course, really fundamental mistakes—taking “6,3” to mean the notes one and four steps above the bass, for example, or anything else really egregious—will lead to trouble. Real misunderstanding at a fundamental level will be hard to eradicate later on. Therefore this exercise should come, as I said above, only after the student has comfortably learned the basic meaning of the figures.  

f) It is extremely important to resist the temptation to write anything down about a realization. The sole purpose of this drill is to develop the reading faculties as they apply to figured bass lines. Any time you write anything—a note or chord or a reminder perhaps expressed as a letter-name for a pitch—you have lost the opportunity to develop that reading, and in fact you are training yourself to be unable to do it. 

g) It is perfectly OK, though, to flesh out the figuring itself. The relative completeness of the figuring of the line that you happen to be using for practice is arbitrary. If you make it more complete before playing from it that is fine. (See, for example, the two versions of the Handel bass line that I included in last month’s column. Either of them is good material for this sort of practice.)

After doing a certain amount of this work, the student will be ready to begin thinking about how to shape an accompaniment for “real life” use, and to begin playing pieces with other musicians. This “certain amount” is often something like 25 or 30 bass lines, each eight to sixteen measures, each played five or six times. That is not a lot, but this method is extremely efficient. Some students will need or want to do more than that; some will be ready to move on to the next stage sooner.  

I will return to the subject of continuo playing and deal with approaches to entering that next stage in a future column. Not next month, however; I want to give readers a chance to digest what I have written about it so far and, if so inclined, to try out the drill suggested here or to have their students do so. I welcome both questions about that process as it unfolds, from anyone who is trying it, and any other feedback.

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

Rhythm and counting

The music that we as keyboard players work with on a daily basis involves time as well as pitch, timbre, and other qualities of sound. We are so used to this that it seems like something that does not need to be said. It is very close to the core of the way that we think about music. In fact, I have sometimes heard people define music as something like “sound with defined pitch organized according to regular time intervals” or words to that effect. This is, if not necessarily a valid definition of anything that could possibly be considered music, certainly a good start at defining one of the important components of the type of music that we typically hear and play. Of course, due to the nature of time and perception, whenever a sound succeeds some other sound, it has to do so on some kind of schedule, and indeed once in a while that schedule is irregular or subject to the whim (or judgment) of a player. But most of the time sounds in pieces of music follow one another at regular intervals. This phenomenon of pulse or rhythm or beat is as important as the notes (pitches) themselves, or close to it, in most musical situations.

And of course our system of notation includes indications for rhythm—that is, lengths of notes in relation to one another. It is actually hard to call to mind any standard way of notating pitch in post-medieval music that does not also notate something about rhythm. When a composer wants to suggest that, in some way, notes are not being presented as having a pre-determined set of rhythmic relationships, the composer has to borrow rhythmic notation and then somehow indicate that the usual rhythmic meaning of the notes is suspended. That is the case, for example, with the unmeasured preludes of Louis Couperin and some other seventeenth-century composers, and with passages marked “free” or “unmeasured” or “molto rubato” or something to that effect.

However, organ teachers are rarely in the position of teaching students how to read rhythmic notation from scratch, or how to read notation at all. We usually take on students who have already learned “the basics” by studying piano—this includes notation along with at least some ideas about fingering, and of course a certain amount of actual keyboard dexterity. (Of course, many of us do teach these basics in settings other than organ instruction, and occasionally we are lucky enough to find a student who wants to study organ from the very beginning!)  On the other hand, there are many students whose sense of rhythm—even at the level of their comfort with reading the rhythmic dimension of the notation—is shaky. There are students who feel that they “can’t count” or who are at least not always confident about their reading of note values, and who are often right not to be confident. As with any other problems in studying or in playing, this usually stems from something’s having been taught inadequately or (and this might actually be more common) having been taught in a way that makes the matter seem more complicated than it has to be.

What follow here are ways to help students with some aspects of reading and producing correct rhythms. This is separate from, though in the end of course related to, the business of playing with a convincing sense of rhythm and pulse. Paradoxically or not, the convincing sense of pulse often comes from departing slightly from the scientifically (i.e., metronomically) correct reading of the rhythm. I am not talking about that here—rubato, timing, using accent to create pulse: all very important, but mostly for another time. One important aspect of the relationship is this: that all of the possible subtleties in the realm of rhythm are extremely difficult to execute for anyone whose playing is burdened by a fear that the basic rhythm is shaky—whether it actually is shaky or not. 

1) If it does fall to me to teach the basic meaning of the different “note values”—that is, the way that we indicate rhythm in written music—I try to do so in as simple and colloquial way as I possibly can. None of the jargon or the ways that we habitually put these things to ourselves in our heads once we know them really well is self-evident, and some of it is unnecessary. Take, for example, the phrase “note value” or the concept of the “value” of a note. It becomes second nature that this word refers to the aspect of the note that is the length of time that it occupies, but someone who has actually not studied the notation at all will not know this. I have actually seen someone more or less flummoxed by the statement that a dot “adds half the value of the note to the note.” And why not: unless you almost already know what that statement means, it is not particularly obvious what it means. I would rather say something like: “if you put a dot immediately to the right of a note, it makes that note one-and-a-half times as long as it would otherwise have been.” There are then two ways to concoct examples of this: first, if, say, a quarter-note takes as much time as two eighth-notes, then a quarter-note with a dot next to it takes as much time as three eighth-notes; and second, if a quarter-note is going to last a day, say, then a dotted quarter-note is going to last thirty-six hours.

2) The notation is all about ratios, and my last example above is meant as a reminder that a ratio is a ratio, regardless of the absolute times involved. A student learning the rhythmic notation from scratch needs to embrace the idea that the notation is all about ratios and not about absolute time—or for that matter about beats, pulses, measures, or time signatures. It is about the notion that some notes last twice as long—or three or four times as long—as others. That’s all. A whole-note lasts twice as long as a half-note; a half-note lasts twice as long as a quarter-note, and so on. Thus, if a quarter-note is going to last a second, then a half-note is going to last two seconds, and a whole-note, in turn, four seconds. If a sixteenth-note is going to last four years, then a thirty-second note is going to last two years, plus or minus an adjustment for the stray February 29. Once the student is comfortable with the idea of ratios, then learning which symbol is which note and what the ratios are is rather easy: there aren’t that many of them. Of course, the teacher must include in the discussion not just all of the “regular” notes, but also the concept of dotted notes, the meaning of ties, and the meaning of triplet signs. These are still all about ratios, and still do not amount to very many different things to learn.

3) By the way, as far as rhythm and counting are concerned, it is probably true to say that nothing would change if all of the time signatures and all of the bar lines were magically erased from all of (at least) western classical music. The note lengths and the rhythms that arise from them are fully described by the note-heads, stems and flags. The bar lines do not change anything about that, and a time signature is—for basic rhythms—either redundant or incorrect. Sometimes a time signature suggests something about what pulses or groupings will arise out of the basic rhythm. But even with respect to that, if the groupings or pulses would not be there or would not be convincing without the time signature, then they are probably not really there with it either. I mention this partly to reinforce the idea that no one learning how to read rhythm in our common notation should be thinking about time signatures or bars or measures. This is just a distraction. Also, when a player who is past the stage of learning notation—who is presumably comfortable with the rhythm side of music reading—nonetheless has a problem reading the rhythm of a passage or with feeling secure about that reading, it is almost always a distraction to be thinking about the time signature or the phenomenon of “measures”. I have seen many students effectively prevent themselves from reading a fairly straightforward rhythm because they were not sure right off the bat how to relate that rhythm to something about the time signature or meter or the placement of bars. This leads me to the next point, an especially important one.

4) If the systematic counting of a passage is going to be useful in creating an accurate rhythm, then of course the counting must happen at a steady pace. It does not, however, need to use numbers that relate it to the measures. Often using the ordinary “one and two and three and . . . ” system for counting a passage is enough of a distraction that it hinders rather than helps. Also, sometimes a student puts so much stock in the fact that the numbers are present and in the correct order that he or she forgets to keep them absolutely steady. So we hear something like this: “one and two . . . and three . . .  andfourand.” The student believes that the passage is being practiced correctly because, again, those numbers are there, and they are the right numbers in the right order. So rather than counting a passage as in Example 1, I would suggest something like that shown in Example 2.

This starts with deciding to use the eighth-note as the steady beat. (The correct choice for the steady counting beat at the beginning of the process of drilling a rhythm is the smallest fairly common note value, unless the passage needs to be kept slow enough that that note value is too slow to follow easily: see 7) below.) Then it expresses the length of each note in the number of eighth-notes that that note value actually includes. Of course, it is still necessary to be sure that the numbers come at an even pace. But the second example zeros in on what the player needs to understand and to work on. Of course, as the passage becomes better learned and can go faster, the next step with this type of counting looks like Example 3.

5) The kind of counting described in 4) is also one of the best exercises for teaching the basics of rhythm notation, once the simple rhythmic meaning of the different note shapes has been learned. Taking a large number of fairly straightforward but not trivial rhythms—say of about the level of my example—and counting them out this way serves to drill the meaning of the different note values quite efficiently. For this purpose it is not necessary to play anything, just to count, dropping back to “1” at the beginning of each new note. At the same time, this type of counting really does work well to straighten out a tricky or recalcitrant rhythm. It is not just for beginners: I use it myself when I encounter a rhythm that I want to count out.

6) Sometimes the idea arises of practicing the rhythm of a passage completely separately from the actual notes. This usually takes the form of tapping the rhythm on a table or clapping it. This can be a good idea. It is based on the clearly sound notion that a tricky rhythm should be practiced some—initially or whenever it becomes a problem—without the distraction of worrying about fingering and hand position. It is a form of isolating and simplifying something difficult, philosophically similar to practicing hands or feet separately. I would suggest, as a modification of this, that a student can in effect practice a rhythm in isolation by playing it with random easy notes, perhaps just five adjacent notes played up and down by the five fingers: no choices to make about notes or fingering, but a sonority to hear. This seems to me to be more “true to life” and probably just more interesting for many players. It also gets around a problem that clapping and tapping both have: that they are usually carried out as “repeated note” gestures, and that fast “notes” are therefore harder than they need to be. If a student is in fact going to tap a rhythm on the table, then he or she should use two hands alternating, or two or more fingers alternating in what amounts to keyboard-playing gestures.

7) Another way to practice the rhythm of a passage with difficult notes is to slow the passage down enough that the notes become easy. This is, of course, always the pillar of good practicing, as far as I am concerned, whatever the particular circumstances. Sometimes, however, in order to make the notes of a passage easy enough that the student can afford to think about the rhythm, the tempo has to become so slow that the rhythm begins to seem even more non-intuitive or not really there: almost as if the sixteenth-notes were indeed lasting years. The way to deal with this is not to be afraid of subdivision. If the line that I used for my examples above were hard enough that it had to be practiced at eight quarter-notes per minute, for example, then the correct choice for the counting beat would probably be the thirty-second note, at sixty-four per minute. This would indeed mean that the opening whole-note would be counted (steadily!) from one to thirty-two. This might seem—or indeed be—annoying. However, there is no shortcut to practicing slowly enough, and attempting to time a note that lasts half a minute by counting only “one and two and three and four and” is not going to succeed. The numbers and the “ands” are too far apart to be meaningfully related to one another in counting them out. 

8) There is sometimes a fear of subdivision or of building rhythms up by counting out the smallest components. The fear is, I believe, that too much of this will make a performance seem choppy, make it not convincingly reflect the underlying pulse. My own experience is that this is just not a problem. Having the player’s conception of what the rhythm actually is—what the note lengths and their relationships are—be both correct and really solid is the absolute requirement for achieving a convincing pulse. Counting small beats—subdivided beats—accurately is the most sure-fire way to be certain that the rhythm being drilled is accurate. As a passage becomes solid and as the tempo is able to approach performance tempo, the player’s focus on the smaller beats will naturally melt away. The ease with which both the correct notes and the correct rhythm can be executed will free the player up to listen for the beat groupings and the underlying pulse along with any and all other artistic or rhetorical aspects of the music. Also, any rhythm that needs to be slowed down a lot and treated to a really extreme subdivision will be the exception: an especially hard passage. Every player will have the experience of playing many passages that have been learned from the beginning counting only the time-signature beats. If it is indeed easier to get a convincing overall pulse and shape in these passages—which I rather doubt—then the player can consciously transfer the feeling of playing and hearing those passages to others that have had to be taken apart more finely.

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

Leaving notes out

Last month I wrote about using altered rhythm as a practice technique. This month I am writing about another way of practicing that also involves changing the note picture away from what is actually on the page for the purpose of playing what is on the page better. Purposely leaving notes out of the texture of a piece at certain stages of practicing is very different from using altered rhythms. However, both techniques involve changing the notes—and both, therefore, are departures from “regular” practicing. The philosophy behind normal practicing at a keyboard instrument is, as I see it, simply that accurate repetition of a physical gesture will, in due course, lead to that gesture’s becoming second nature. Most of what I have conveyed about practicing is more or less a gloss on that idea. Therefore, I think it is worth pointing out when the physical gestures being suggested are actually not those that will directly lead to playing the passage or piece accurately. 

The positive effect on the learning process is less direct. Any student who is interested in altered rhythms as discussed last month, or in the technique discussed this month, must always remember that these techniques only supplement basic, regular practicing and cannot replace it. It is important to employ these unusual practice techniques in small doses, alternating with regular practicing of the same passages, so that the connection between the specialized practicing and the feeling of actually playing the passage being practiced is clear and vivid.

Expression through timing 

and articulation

At the organ—and also at the harpsichord—we cannot create accent, shape, phrasing, or rhythm by playing certain individual notes louder or softer than other notes. That is basic: in fact, for many pianists first coming to study organ or harpsichord it is almost the definition of those instruments. However, we can create accent, shape and so on at the organ or harpsichord—we do so by using, generally speaking, timing and articulation. (An interesting way to think of those two concepts, by the way, is this: that timing concerns initiating notes, and articulation concerns their release. This is also basic and perhaps obvious, but a good way of organizing thinking about execution of notes.) 

It is possible to spell out certain principles about the relationship between timing and articulation on the one hand, and accent, rhythm, phrasing, etc., on the other. For example, in general, notes that are held a bit longer than the metronome would suggest come out subjectively louder, notes that are preceded by a space—an articulation—likewise seem louder or accented, and so do notes that are delayed a little bit. Notes that are reached through strong legato often seem softer or unaccented, as do notes that are a bit shorter than the metronome would suggest, assuming that they are not so staccato as to draw attention to themselves. And so on: all of this is a bit over-simplified, though generally valid. None of it, however, is unfailingly true: a lot depends on the subtleties of the exact situation.

There is a lot that can be done, in any situation, to plan out the use of articulation and timing to create accent and shape, and doing so is important. However, there is also a danger. Schemes of articulation that are carefully thought out and mapped out can become stiff and lifeless. The act of thinking consciously about those schemes while actually playing can lead to stiff playing or can be a kind of distraction that decreases security and accuracy. I am deliberately somewhat overstating this problem: it happens sometimes to most of us; it is probably more of a problem for students. They often are vividly aware of their teachers paying attention, and they may be implementing articulation schemes and other performance ideas that the teacher helped to create or, in some case, just created. However, it is something that can happen to any player.

The point is this: if a player learning a passage can train the ears to hear more prominently those notes that are supposed to be in some way more prominent—accented notes, points of arrival, “louder” notes, notes that should seem to bloom or grow, as if a string player or singer were leaning into them after the initial attack—then the player’s subconscious mind will find ways to make those notes more prominent. This process can both supplement and to some extent bypass or replace the process of logically working out what notes should be longer or shorter and how exactly the timing should be adjusted to give the desired effect. This phenomenon—the direct link between hearing prominent notes and projecting them as prominent—has an intuitive, improvisatory feel to it, which is intrinsically non-stiff and which serves as an antidote to stiffness.

It was Professor Eugene Roan who first mentioned this idea to me—somewhat cryptically. He said that “if you can hear it, it will move through your elbows to your hands” (or something like that) and invoked the idea of “magic.” Whether it is magic or, as I suggested above, something to do with the subconscious—or both—I have found it very effective. From Prof. Roan’s remark I have developed some specific techniques. I routinely use these myself with most of the pieces that I learn. 

Practice procedures

Some musical lines lend themselves to the technique of leaving notes out in a way that seems almost too obvious, too easy—usually lines in which the rhythmic hierarchy of the notes is the clearest. The fugue subject from the Bach D-minor Toccata BWV 565 is such a line (Example 1).

The repeated A’s—all off the beat—are notes that almost everyone analyzing this line would agree should be lighter, less accented, than the (changing) on-the-beat notes. If the theme were being played by a violin, the player would almost certainly make those notes quiet. Leaving those lighter notes out gives a line that looks like Example 2 or perhaps a line that should be thought of as that in Example 3, with variably detached eighth notes.

Notice that I have made the judgment that the first three sixteenth notes of the theme form a three-note upbeat and should not be stripped down. Someone else might see that differently and render the beginning of the theme as Example 4 and so on. This would be fine. The exercise is not about right or wrong: it is about thinking about what you want to hear, and then moving your ears closer to being able to hear it.

Once you have made a decision about which notes to leave out, the procedure for practicing is something like this: 

1) Play the rewritten line several times. There is probably no reason not to use the fingering or pedaling that will be used for those notes when the whole thing is put back together. It is also not terribly important to do so. It is extremely important to keep the touch light—the success of this technique depends on that.

2) Put the missing notes back, and play the passage several times. At this stage (and always) it is still important to keep the touch light and fluid. Let your ears follow, as much as possible, the notes that you played in step 1. Don’t pay too much attention to the notes that you have added back.

3) Repeat 1) and 2) a couple of times if you wish.

4) Now do something else: practice another passage—this way or “normally”—or have supper or go for a walk. When you next play the passage that you have worked on this way, it may feel or sound a bit different—more vivid, or more relaxed, or both. You don’t have to scrutinize it or analyze it. If you happen to notice a difference, that is wonderful; if not, no harm done. The extra attention that you have paid to the passage will help solidify the learning of it in any case.

In the case of this Bach fugue subject, another step is putting the theme back together with the rest of the texture of the piece, or, to put it another way, continuing to use this technique as other voices come in. This immediately highlights the relationship between the fugue subject and the various eighth-note countersubjects. This piece is full to the brim of passages that can be approached this way—for example, the measures immediately following the second entrance of the fugue subject (Example 5).

In this passage, leaving out the off-the-beat sixteenth notes in the lower voice—both with and without also playing the upper voice—is a good way to explore the rhythmic relationship between the voices. It should clarify the interaction between the implied detached eighth notes of the lower voice and the actual eighth notes in the upper voice.

In the opening of the famous Widor Toccata there are all sorts of possibilities for leaving notes out (Example 6). Any of the following might be illuminating:

1) Play the first three sixteenth notes of each beat in the right hand and leave out everything else; play the left hand as is.

2) Play the notes found on the first and fourth eighth notes of each half-note beat, in both hands, and leave out everything else. (These are the places where the pedal plays when it comes in.)

3) Play the right hand as is, and in the left hand only play the chords that fall on half-note beats.

4) An idea that is quite specialized and geared to this passage: play the right hand as is on a loud sound, and in the left hand—on a softer sound—play the chord on each half-note beat, and hold it for the entire half note. When you restore the written rhythm and texture, this will perhaps guide you towards hearing all of the off-the-beat left-hand chords as growing out of the reverberation or bloom of the on-the-beat chords.

There are probably many more possibilities as well. The process of figuring out what scheme of omitting notes might make sense is itself a good learning opportunity.

In a passage whose rhythm is less regular, or in which the hierarchy of beats is less clear, it might take more analysis to discover what pattern of omitting notes makes sense, if any. It is also probably true that the less clear it is what notes might be omitted to create an exercise of this sort, the less compellingly useful the exercise will be. However, it never hurts to try it out. 

In the Franck Choral in A Minor, for example, there are several unusual ways of applying this idea. I like to hear each half of the opening measure as having a diminuendo to it, or, perhaps more accurately, I like to hear the last three sixteenth notes of each half measure as being an “after-beat” to the second or fourth quarter note. Therefore I would play the opening a few times like Example 7.

I like to think of the first note of measure 7 as being quiet: a diminuendo coming off the suspension over the bar line (Example 8). I would consider trying this phrase leaving the C#’s out. This would be jarring and unsatisfying harmonically, but might train the ears to play the notes “quietly.” In various passages that have writing like that in the right hand part here (Example 9) I would try leaving out the off-the-beat sixteenth notes. And in this passage (Example 10) I would try leaving out the (double) pedal sixteenth notes. In this case, it is important to release the octave E’s, which have temporarily become repeated notes, smoothly and in plenty of time, so that when you add the D#’s back in, your feet will not have grown accustomed to being stuck on the E’s.

There are, in any piece, many possibilities for leaving out notes that you think of as being quieter. As I am suggesting, there is nothing wrong with simply trying such things out. Even if a particular practice pattern ends up seeming not to have made much difference, the process of working on it still has not been wasted. As I alluded to above (and elsewhere), anything that makes you notice more of what is going on in a piece will contribute to the learning of the piece and to its becoming increasingly solid. Students—and teachers—can play around with the idea.

It is worth mentioning one more time the importance of light touch in this particular context. In order for the magic to work—or for the subconscious to guide the hands and feet to do subtle things, too subtle to describe analytically—it is important that there be no tension. Tension or tightness will make it much harder for the subtleties heard by the ear to express themselves in the fingers—perhaps impossible.

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

Additional practice techniques

This month and next month I will discuss a number of practice techniques that involve changing something in the music while practicing. This includes practicing in rhythms other than the true rhythm of the passage being practiced, and practicing passages with some of the notes actually omitted. (The former is the subject this month, and the latter next month.) Some of the techniques are generally quite familiar, although I may have my own twist on them. I have at least briefly alluded to some of them elsewhere, but I think that it is useful to bring them together here and to zero in on them in a bit more detail.

Any and all of the techniques that I will describe here are meant only to supplement the basic, solid way of practicing that I try never to miss a chance to describe: slow, careful repetition—with or without separating hands and feet, depending on the exact circumstances—with correct notes and correct rhythms, speeding up only gradually, in a way that is determined by the flow of the practicing itself, rather than by any predetermined schedule. This discussion of, among other things, practicing “wrong” rhythms comes the month after I talked about various ways of thinking about counting and the art of learning correct rhythms. Therefore it is important to remember that practicing in alternate rhythms is a technique that should have a well-focused purpose and that should be kept conceptually separate from playing the piece or passage as such. I will talk some about how to maintain this separation.

Alternate rhythms: dotted values

For me, the purpose of practicing in alternate rhythms is quite specific: it is a halfway point between a slow practice tempo and a faster tempo, which may be the performance tempo or may be even faster than that. When a rhythm is altered, some notes become faster and some slower. The faster notes are being practiced in isolation at a faster tempo than they would have otherwise; the slower notes, because they are slower than they would otherwise be, provide an opportunity to rest and regroup between fast notes. The classic form of this sort of practice is, therefore, to create pairs of practice rhythms, in each of which half the notes are fast and the other half slow, and between which all of the notes get a chance to be the ones that are being played (extra) fast. 

In turn, the classic form of that technique is to take a passage in which the note values are uniform and to make those note values dotted. So a passage that looks like that in Example 1 will be played first like Example 2, and then like Example 3.

The purpose of this, again, is technical. The first rhythm offers a chance to practice half of the movements from one note to the next quickly; the second offers the chance to practice the other half equally quickly. Because this kind of practice allows the player to stop and rest, in effect, after each fast gesture, it is usually possible to include and drill those fast gestures sooner in the process of learning the piece than it would be possible to boost the overall tempo to that same speed.

A modification of this technique that I think makes it even more useful is to replace the dotted rhythms (“regular” and “reverse”) with a kind of unmeasured over-dotting. So, in the above rhythmic templates, the dotted eighth-notes would all be replaced with fermata-ed notes to be held as long as you want, and the sixteenth-notes would be replaced with notes as short, quick, and light as you can make them. This approach creates an even more intense drilling of the quick gestures and an even more effective rest between those gestures. It also, by virtue of its being farther from any sort of regular pulse, has less ability to affect or possibly undermine the regular steady rhythm of the passage when the player returns to regular rhythm.

This sort of practicing can be applied very naturally to a line such as the opening of Bach’s Prelude & Fugue in G Major, BWV 541 (Example 4), which is in one voice only and in all or almost all one note-value. Practicing a line such as this in some version of dotted rhythm is most useful if you analyze what it is that you are gaining in each spot in the line. For example, when it is the on-the-beat sixteenth-notes that are held long, then going from the first beat to the second beat of the second measure (marked x) you will be practicing quickly moving the hand position in such a way as to reach the middle D reliably. 

In a passage such as Example 5—from the Vierne Divertissement (Pièces en Style Libre)—if a dotted rhythm is applied to the sixteenth-note line, the quarter-note chords will come along for the ride, and the practicing of those chords will also be affected. If the on-the-beat sixteenth-notes are being held long, then in spots like those marked with x’s the student must be fully ready to play the next chord, as well as the next sixteenth-note, before playing the (very fast) off-the-beat sixteenth-note. The actual gesture of moving from one chord to another will then be fast: probably faster than it will need to be in the piece. However, the student will have time, waiting on the “and of two,” to prepare that very fast gesture. The other half of the exercise—the reverse dotting—applied to this passage probably has less effect on the feel of the playing of the chords. In that case, moments like those marked with y’s will probably constitute the most intense and useful part of the exercise—playing wide intervals very quickly. 

In writing like Example 6—from the Bach Toccata & Fugue in F Major—these sorts of rhythms can be applied to one hand at a time and then to both hands together. Again, the student should analyze and pay attention to what exactly is being practiced at each moment in the passage. For example, going from the second beat to the third beat of the first measure of this excerpt, the left hand has a “stretch” or shift in hand position while the right hand does not; the opposite is true near the end of the fourth measure, or heading into the last measure. The use of the two complementary dotting patterns will highlight some of these technical details.

(Would it be a good idea to practice this passage or one like it using opposite dotting in the two hands? That is, first place the lengthened notes on the beats in the right hand and off the beats in the left hand, then reverse both of these. I have actually never tried that and I can’t recall a student’s doing it. The purely physical practice would be unchanged from the method described above, but the concentration required would be different—and it would probably be harder.)

Alternate rhythms: 

Groupings of notes

Another format for altering rhythms to create effective targeted practice strategies involves speeding up not one note at a time (every other note, as above) but clusters of notes. The classic way of organizing this is to play first all of the notes after each beat very fast, ending on and then holding the next beat, then to play all of the notes starting on each beat very fast, ending with the last off-the-beat note of each grouping. The template for doing this works as follows. For a set of notes written like Example 7, you would first play as shown in Example 8, with the notes under each slur played as fast as possible, and the notes under the fermatas held as long as necessary to feel ready to play the next cluster of fast notes; then Example 9.

In this case, the notes under the slurs should again be played as fast as possible. Then the last note of each grouping can be held until it feels comfortable to execute the next cluster of fast notes, or the note can be released, and the waiting can take place while not actually holding any notes: in effect a fermata in the gap between groups of notes. In the latter case, of course, it is a good idea not to let the hands or feet move too far above or away from the keys.

(As it happens, I myself have recently used the fast-cluster approach myself on this Buxtehude harpsichord passage, from the La Capricciosa Variations, which I have always found extremely hard—harder than I had originally expected [Example 10]. This measure is full of funny changes of direction and unexpected intervals. I was eventually able to get comfortable with it—and to get it to be reliable—and I believe that the fast-cluster practicing was the most important part of the process.)

How to use alternate rhythms

Practicing in “off” rhythms is, as I said above, a technical practice. The purpose is not to learn something about rhythm or any other artistic or interpretive aspect of the piece. It is to drill isolated gestures at a fast tempo in a focused way that does not demand that the fast tempo be kept up for very long. Therefore, it only makes sense to practice a passage this way (that is, in one of these ways) once the fingering and/or pedaling has been worked out and is indeed fairly well learned. If the fingering or pedaling is uncertain, then the fast moments in the rhythmic patterns are not going to work: they will be hesitant or actually fall apart. It also makes sense to practice this way only with a complete texture or with a part of the texture that involves the same fingering and pedaling as the complete texture. That is, it is fine—very useful in fact—to apply different rhythms to one hand at a time or to the feet alone, but not to separate voices extracted from a contrapuntal piece.

Since the technical purpose of this sort of practice is focused on the fast moments, the slow moments, whether they amount to every other note or to one note in a larger grouping, can be held for as long as the player wants. There does not need to be an overall tempo. Since every gesture—moving from one note to the next—takes its turn at being the fast gesture, everything gets practiced effectively regardless of how long—how slow—the held notes are. The less regular the timing of the held notes—the more the student simply waits on those notes until he or she feels ready to play the next fast gesture—the less this kind of practice will have any tendency to interfere with normal rhythm, because it will not be presenting an effective alternate rhythm.

It is easiest to apply this kind of practicing to passages in which rhythm is more or less even, as in all of the examples so far. In a passage such as Example 11, from Buxtehude’s Praeludium in C Major, BuxWV 136, the rhythm is less regular. There are beats in which the surface rhythm is two eighth-notes, in which it is four sixteenth-notes, and in which it is mixed. In the second beat of each of the first two measures of this excerpt, for example, if all of the notes within that beat are being played fast, then it would be conceivable either to try to maintain the rhythmic relationships within the beat, or just to play all of the notes as fast as possible and, therefore, as fast as one another, whether they are in fact eighth-notes or sixteenths. Either approach could be fine, as far as I can tell, and they would both amount to effective practicing. In fact, students can be encouraged to create their own short clusters of notes for very fast practicing, and to figure out how they would like to deal with rhythms within those clusters. As long as the clusters are not very long, and as long as they overlap by a note or two, the practicing should work well. 

Next month I will talk about a sort of practicing that is, in a sense, the opposite of what I have been discussing here, namely playing some notes of a passage while leaving the other notes out. The reasons for doing this are usually not technical, but rather about developing the ears’ relationship to the music in a way that enhances understanding of structure, rhythm, and pulse.

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

Some thoughts about substitution

Substitution in organ playing—both with fingers and with feet—is a technique that is practiced to some extent by almost every organist. This includes both those who plan it out and know that they are doing it and those who don’t plan it and nonetheless do it by chance at the last minute. It has been the subject of heated debate and disagreement—one of those subjects that can sometimes seem almost political in nature. The disagreements are usually about the relationship between substitution and various aspects of authenticity—and of course authenticity is the most political subject in the realm of performance. Indeed, it is quite certain that substitution has been more prevalent at certain times in the long history of organ music than at others, and practiced, or expected, more by some composers than by others. 

It is also a technique that can be carried out in ways that are natural and easy or in ways that are awkward and difficult. In this column I want to discuss, just briefly, the history and theory of substitution and also to suggest ways of thinking about applying the technique itself.

Before the eighteenth century

The usual succinct way of describing the history of substitution in keyboard playing is this: that it was unknown or at least very uncommon before the late Baroque, and that by the nineteenth century it had become extremely common—though more so in organ playing than in piano playing. As far as the historical record is concerned, there are no sources from before the early eighteenth century that explicitly discuss substitution or that direct a player to use it, but there are many from the nineteenth century and beyond. 

The first surviving printed or written reference to finger substitution comes from François Couperin, writing in his L’Art de toucher le clavecin—a harpsichord method published in 1716. He speaks of it as something unusual if not downright new, contrasting it to “l’ancienne manière”—“the older way [of fingering].” Couperin’s prose is concise or even cryptic, and he says very little about the thinking behind his employment of substitution, but it seems to be intended some of the time to make it easier to achieve some sort of legato, as in Example 1, or to cope with suspensions or other complicated textures, while maintaining legato as in Example 2. (Note that there are no fingerings given for the upper left-hand notes. The G in the second measure would be played with the thumb—after the substitution on the lower note is complete. But what then? A substitution to preserve strict legato in the inner voice or not?) 

There are two significant pieces of evidence that substitution was not common before the time of Couperin. They are, first of all, Couperin’s own attitude about substitution: that it was something new. Of course, he did not know everything about musical practices over the whole world and for all of the decades before his own time, but he probably kept himself rather well informed. The second piece of evidence is simply that no one mentions it prior to 1716. During the seventeenth century there was by no means as much written about keyboard playing and pedagogy as there was in later centuries. However, enough such writing has survived that it seems significant that substitution is not even hinted at in any of it. This is of course consistent with much of what else is known about keyboard playing, whether on organ or on harpsichord or clavichord: in particular, that full legato was the exception, especially over long chains of notes such as the ornamented melody in the first example above.

I however have always felt cautious about assuming that no one in the time before Couperin ever used substitution, either with fingers or with feet. My principal reason for this is that it seems to me to be a natural human thing to do—not necessarily as a result of planning or artistic decision-making, but as a tool for coping with situations that might arise. I have seen many students—beginners, with no preconceptions about fingering, to whom I had certainly not (yet) said a word about substitution—in effect re-invent the technique because they found themselves at some sort of fingering dead end. Furthermore when they do this, they usually do it—because it is not self-conscious—in a relaxed and natural way that constitutes good technique. Also, substitution is a natural thing to do by analogy with other human activities. For example, arriving at the front door recently from a trip to the grocery store, I shifted the bag of groceries from one arm to the other so that I could reach my keys. This is conceptually the same thing. If a player is holding a note with a particular finger and finds that it is inconvenient to be holding it with that finger, slipping a different finger onto the note is a natural, spontaneous human response. I put it this way only partly to make the case that players prior to the eighteenth century may well have used substitution willy-nilly, so to speak, even though it was clearly not a prominently taught technique. I also want to suggest that because it is a natural and physically obvious thing to do, when we in fact want to do it, we should remember to do it in a physically natural and comfortable way. Also, perhaps, that we should avoid it whenever it cannot feel natural and comfortable, unless there is a very strict reason for its being necessary. 

The choice about whether to use substitution in playing music that was written in the era when it was, at a minimum, clearly not being taught as a core technique is of course one for each individual player to make. It will inevitably stem in part from that player’s overall approach to questions of authenticity, and also from other things about technique and habit. Every teacher should frame this particular issue to students in whatever way is consistent with the teacher’s and students’ interactions over matters of authenticity in general. To me the bedrock caveat or concern about substitution in what we might categorize as “pre-Couperin” music is this: that if you actually need substitution to make a particular articulation or phrasing happen, then that articulation or phrasing is almost certainly not anything that the composer specifically had in mind. 

The Classic period and beyond

Substitution is referred to in printed sources only a little bit through the middle of the eighteenth century, but references to it in keyboard methods and elsewhere proliferate in the Classic era. It is interesting to note that although we organists (rightly) think of substitution as being most at home in a certain branch of organ technique, it was in early piano playing and teaching that it first caught on. This was in the era when the damper ‘pedal’ on pianos was usually either a hand stop or a knee lever but, in any case, was awkward to operate. Legato lines were by and large achieved through fingering. During the nineteenth century two developments shifted the emphasis on substitution from piano to organ: first, the invention and quick universal acceptance of the (real) damper pedal, and second, the use of a more legato style in  organ technique.

It is worth remembering that even for the relatively well-documented nineteenth century we do not in fact know how everyone did everything. Franck, for example, left no substitutions among the few fingerings that he provided for his own organ music. There are clearly many places indeed where substitution is required, especially if the goal is to create true unbroken legato. In this passage from the Prière (Example 3), there is a need for substitution in, probably, a majority of the transitions from one moment to the next—if, again, true legato is to be maintained. (It is possible to play these notes without any substitution if the full-fledged legato is abandoned.) The section of the piece from which this passage comes is under the marking “très soutenu,” which suggests legato. However, I myself cannot devise any way at all to play the last three eighth-note left-hand chords of either the second or the fourth measure truly legato, with or without substitution. There are other spots throughout the section of the piece about which I would say the same thing. Does this tell us anything about articulation? We know that Franck had large hands—much larger than mine or than those of other players whom he might have expected to play his music. We also know that most European churches, certainly including Sainte-Clotilde, have spacious acoustics in which listeners can experience the effect of legato even through subtle breaks. 

The point here is not to resolve anything in general about the articulation practices of Franck. It is rather this: that we should be ready to use substitution where it makes sense, but be cautious about assuming that legato is necessary in nineteenth or twentieth century music, simply because the possibility of substitution exists. There is a kind of circular logic that says: “This passage must be played with a lot of substitution because it has to be legato, and we know that it has to be legato because players at that time used a lot of substitution.” This doesn’t make sense. (I have certainly caught myself thinking that way, as well as students and people who have written about these things.) The more appropriate way of thinking about it something like this: “If I want this passage to be legato and that requires substitution—even a lot of substitution—that is indeed completely consistent with what the composer might have intended or expected.” 

This is the fingering, just as an example, that I myself would use for part of the left hand in the excerpt from the Prière (Example 4). The function of most of the double substitutions is clear and normal: to create the possibility of smooth legato. There are a few specific things to say about these fingerings. The 3–4 substitution on the fifth eighth-note of the first measure is really just for comfort—better hand position reaching that chord and moving away from it. That is specific to my hands: another player might play the lower note initially with 4 or hold it with 3. Holding that D# with 3 might take away the need for the lower part of the next substitution (Example 5).

I would execute these substitutions quickly and in an unmeasured fashion (more on that below), with the possible exception of the one on the chord that spans the first bar line. That one I might divide between the two beats in a measured way. The 3–4 in the second measure is—again in relation to my particular hand—an attempt to make the large jumps at that spot more comfortable and closer to legato, although, as I said above, I cannot quite reach them fully legato.

As I wrote in my column on repeated notes (January 2009), there is a relationship between the practice of playing repeated notes with different fingers and substitution, but with a difference in articulation. In playing two notes in a row with different fingers, you release the first finger before playing the second; in substitution you play the second finger before releasing the first. (It is interesting, by the way, that Couperin was an advocate of, and wrote about, both of these practices.) There is also this difference: with repeated notes you have the opportunity to get the first finger out of the way early, but with a substitution the two fingers must be able to share the key at least briefly. This can dictate the details of how the fingerings are carried out—which finger goes above or below which, and whether the finger being released is released up, sideways, or down. 

Playing substitutions

The most important thing in preparing a substitution to be comfortable and natural is working out the details of the direction in which the fingers move—the choreography of the fingering event. In the first Couperin example, the first few substitutions are all 4–3. In these cases the fourth finger must be released down (towards the player) and slightly to the right. In the 4–5 substitution at the beginning of the second line, the fourth finger must be released up (away from the player) and probably straight or slightly to the left. These shapes will prevent the fingers from interfering with one another. In double substitutions the most important part of the choreography is the order in which the chain of substitutions is carried out. There is always an order that allows the hand to contract during the process and an order that causes the hand to stretch. It is never hard to tell which is which. The former is always better; it is always possible to tell which is which: it is important to do so. 

The gesture of substitution can usually be carried out either in a way that is measured—the new finger moves onto the note at a definable time, the old finger moves away at a definable time—or in a way that is unmeasured, with the new finger simply sweeping onto the note as promptly, quickly and lightly as possible, while the old finger is swept away. The first of these is, in a way, analogous to a measured appoggiatura in the Baroque style, and the second to a quick grace note or an acciaccatura. Furthermore, whether the substitution itself is measured or quick, it can, on a longer note at least, be positioned either right at the beginning of the note or anywhere else in the lifespan of the note before the moment when the logistics require the new finger to be in place. 

On the whole, I prefer to execute suspensions quickly—the acciaccatura model—and as close to the beginning of the note as possible. I believe that I do this because I want them to have no rhythmic weight of their own. Not that they would have audible rhythmic weight since by definition they are silent. In fact, that is the point. As much as possible I like to reduce the feeling—which substitutions by their nature are inclined to create—that there is something going on physically in the playing that is not reflected by anything audible to the listener. All else being equal, I believe that a one-to-one correlation between physical gestures that we feel ourselves making and sounds that we hear our instrument producing helps to intensify our focus on the rhythmic aspect of the music and to make it more likely that we will effectively project rhythm and pulse to the listeners. In my fingering example above, I might play the substitution on the note that crosses the first bar line at the downbeat of the second measure rather than at the beginning of the note. This is because there is an implied rhythmic event there—the strong beat across which that chord is suspended. I would probably still make the gesture of the substitution a quick one, since there is only one rhythmic event going on with which the gesture might be correlated. 

The use of substitution

There are three ways in which substitution is used: 1) as part of a well worked-out fingering plan, with some specific goal in mind, usually related to achieving legato, but sometimes for comfort, reliability, or good hand position; 2) as part of an approach to fingering, even when it is not explicitly worked out in advance, but still with goals in mind, again usually having to do with creating legato; and 3) as a way of scrambling around to get notes at the last minute in a passage that has not been adequately prepared. For an observer, including a player observing him- or herself and also including a teacher observing students, these last two can be hard to tell apart. The third of these is on the one hand a useful fallback if it is needed: substitution can be a powerful way of crawling around the keys and getting notes in an emergency. However, any tendency to rely on it for that, except in the occasional emergency, has to be resisted ferociously. If a teacher believes that a student is using a facility with substitution to avoid having to think about the best fingerings, listen for what fingering does for interpretation and performance, and practice enough and well, then the teacher must step in and ask the student to pull back from that and restore substitution to its place amongst the legitimate technical tools that we have at our disposal.

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

Continuo, part I

The musical practice known as continuo playing was an integral part of ensemble music from about 1600 until about 1750—the dates that we assign to the “Baroque Period.” Indeed, it makes a lot of sense to define the Baroque specifically as the era in music history when continuo playing was the norm. During that period, almost every work of music that was not a solo keyboard or lute piece included a continuo part. (Exceptions, such as pieces for unaccompanied violin, or lute songs, probably amount to no more than five percent or so of the repertoire.) This includes sonatas, trio sonatas, works for larger instrumental ensembles, songs, cantatas, Masses, operas, and oratorios—arias, choruses, recitative, and so on. The practice of writing continuo parts certainly persisted into the second half of the eighteenth century—the “Classical” era—but became less common, less mainstream, less central to what was going on in the world of musical performance. Continuo died out early in the nineteenth century. (Mendelssohn, however, still included continuo parts in some of his sacred music in the 1830s.)

What is continuo? 

This month’s column will begin to answer that question, or, really, to address it in ways that I hope will be helpful to students. (Of course if any of us as organ teachers have students who have already studied continuo fairly deeply or who have specialized in it, then those students will already know or understand more than I am going to write about here. So this is, at least directly, for everyone else.) Next month I will outline in a fairly basic form my own approach to teaching the nitty-gritty practical side of creating and performing continuo parts at the keyboard, starting with how to read the notation, and I will discuss how to deal with the artistic choices that creating such a part entails.

I actually have a memory—a distant memory by now—of my own first encounters with the word “continuo.” These happened at a number of chamber music concerts on the Yale campus that I heard in the late 1960s, as a youngster just getting interested in music. There were pieces described as “Sonata for violin and continuo” or “Trio Sonata for violin, oboe, and b.c.” or other such phrases. (The word “continuo” and the expression “b.c.” are both abbreviations for basso continuo.) I noticed that some of these pieces turned out to have the wrong number of players, that is, a trio sonata might have four people playing. I still remember a sort of “Twilight Zone” feeling that I got looking at descriptions in the programs that seemed not to be written in any normal language that I could discern and that seemed not to correspond to what I was seeing on the stage. I vaguely remember asking someone (my father?) what it all meant and his not knowing either. I believe that we considered the possibility that it might be some sort of misprint.

I still get the very basic question—“What does ‘continuo’ mean?”—both from audience members at concerts and from (new) students. The basic answer is this: a continuo part is a line of music, mostly in the range of the bass clef, that forms the lowest part of the texture of a piece, that is meant to be played by one or more instruments in unison, and that is meant to be supplemented by notes not written by the composer: chords or bits of melody supplied by one or more of the performers. The choice of instrument or instruments is not, except in rare cases, specified by the composer. The performer’s process of deciding what “notes not written by the composer” to add is called “realizing” the continuo part. In the Baroque period this was almost always done by actual improvisation. Nowadays it is done either by improvisation or by planning and writing a part in advance.

Elements of improvisation

Part of this picture is that Baroque composers—from the most iconic such as Bach or Handel through thousands of others whom most of us have never heard of—expected the actual notes of their pieces to be different from one performance to another, with part of the note picture composed not by the “composer” but by any given performer. This often blows people’s minds: we associate the notion of a performer writing part of the music with certain kinds of twentieth-century experimental art—participatory or aleatory music. The music of the Baroque often seems to embody an opposite principle, one of rigorous form, often expressed through complex counterpoint. 

Sometimes the simple act of becoming aware of the nature of continuo accompaniment can reset a student’s sense of what Baroque music is all about, away from structure and control towards spontaneity and change, and, in a sense, away from the composer towards the performer. Of course, it is also true that a lot of Baroque keyboard and lute repertoire was improvised from scratch. In fact, we assume that something close to all of the keyboard playing that took place in the Baroque era was improvisation. However, in a funny way, improvised repertoire suggests a less radical departure from composer control than continuo accompaniment does, in that with improvised repertoire the performer is the composer.

Of course with continuo accompaniment, the additions to the music put in place by the performer exist within certain well-defined bounds—and we’ll come back to that below. However, it is clear from comparing all of the recordings of just about any piece of Baroque music that the differences between one player’s version of the keyboard continuo part and another’s can make a huge difference in the overall effect of a piece. And, again, this is something to which composers routinely ceded control.

The key to accompaniment

So why did composers give up control over a crucial aspect of their pieces—consistently and over a period of more than 150 years? I believe that the answer lies in the nature of accompaniment and in the nature of the instruments used for accompaniment during those years. There is a lot to say about accompaniment, whether of the continuo variety or of the obbligato variety, as represented by such things as Schubert song accompaniments. Great accompaniment requires all sorts of subtleties and sensitivities. However, one thing is absolutely fundamental, without which accompaniment runs the risk of being not just artistically sub-par but really grotesque: the ability to vary dynamics in a way that tracks what the other instruments or voices are doing. Without this basic ability an accompanist constantly runs the risk either of drowning out the other instruments or voices or of failing to support them adequately. If the keyboard instrument is one on which dynamic variation is inherently possible, say, the piano, then a composer can write accompaniments in which the note picture is fixed once and for all, that is, written by the composer as part of writing the piece. If, however, the accompanying instrument is, like the harpsichord or the Baroque organ, not capable of inherent dynamic flexibility, then it is important that the performer be allowed to change the number of notes being played at any one time in order to change the effective dynamics. A Schubert song piano part played as written on a harpsichord would be an almost pathetically ineffective accompaniment. It would fail to support a singer with a robust or just plain loud voice, it would drown out or at any rate compete too much with a light or delicate singer, and it would fail to reflect or mirror or complement nuances of dynamics executed by any singer. However, it is possible, in a piece with continuo accompaniment, to make the keyboard part of a whole passage louder or softer by choosing to play a thicker or thinner texture of added notes and chords. It is also possible to place an accent on certain notes or beats while allowing other notes or beats to be unaccented, again by actually playing more notes, a thicker texture, on the accented moments and fewer—or no—notes elsewhere. It is possible in the same way to respond appropriately to crescendo, diminuendo, and other dynamic gestures that singers or other players carry out.  

(I should mention that years ago I subscribed, without having really consciously thought about it, to the absurd idea that Baroque composers wrote continuo lines rather than obbligato accompaniments because their composing skills were too rudimentary to concoct complex accompaniments. In this story line, the development of “real” keyboard parts for chamber music and songs in the second half of the eighteenth century was a kind of progress, akin to the scientific progress that—genuinely—characterized that era. The notion that composers who wrote the elaborate, complex counterpoint that was routine in the seventeenth century couldn’t have written compositionally successful keyboard parts for their songs and chamber music is indeed absurd. However, I think that some people do fall into the trap of assuming some such thing, as we have a general tendency to believe that the passage of time brings progress. We feel that people of old simply couldn’t do a lot of what became normal or easy later on.) 

Some confirmation of the notion that the continuo texture really did serve the purpose I have described is found in this: when composers in the Baroque era wrote song accompaniments intended to be played on an instrument that had dynamic flexibility—namely the lute—they did write obbligato accompaniments. This gives us the lute song repertoire, with all of the notes of the pieces written by the composers.

Continuo instrumentation

The instrumentation of a continuo part is flexible. This is one of the reasons that the part is given the somewhat abstract name that it has. It is not the “organ” part or the “harpsichord” part. It was customary for a continuo part to be played by at least two instruments: a bass melody instrument playing the continuo line itself and a chordal instrument—keyboard or lute—also playing the written continuo line, but adding the extra notes and chords that we have been referring to. It was also common for more instruments to be involved. Typical combinations include cello and harpsichord; cello and organ; bassoon and organ; gamba, organ and lute; cello, double bass, and harpsichord, and so on. This flexible instrumentation is the source of my old confusion about the number of players on stage. A “solo” sonata can have anything from two players to four or, somewhat atypically, five; a “trio” sonata might indeed have only three players, but more usually will have four, often five or more. A continuo group for a large-scale piece—a cantata or oratorio or orchestral piece—can easily have half a dozen or more players.

Regardless of the exact instrumentation—which, again, is almost always at the discretion of the performers—the structure of the part is the same. The line actually written by the composer, the bass line, which is the foundation of the harmony of the piece, is played in unison by all of the instruments participating. Notes that are added by a keyboard player or lutenist are played only by that one instrument. Thus, most of the time it is the bass line itself that, within the texture of the continuo part, is the most prominent, with the added notes always somewhat in the background. (An organist performing a continuo part without the help of a melodic bass instrument should bear this in mind in planning registrations.)

Figured bass

So, if a keyboard player performing a continuo part is supposed to add notes to the texture, how is the choice of those notes to be made? The first answer is that they must be notes that are consistent with the prevailing harmony, and not in conflict with what is going on in the written parts. The player needs to have a way of knowing what that prevailing harmony is. This can be achieved by ear, for players who are skilled at such things, or by studying the score. However, this is also where the figures that are often written under the musical notes of a continuo part come into play. Those figures are in effect a short score of the harmonic picture of the piece. To some extent they indicate what notes the other instruments and voices are actually producing. Beyond that they indicate what other notes are consistent with the harmony implied by the notes being played or sung or by the harmonic logic of the piece. The system of figures is a system of abbreviations. As mentioned above, I will go into detail about how to read figures next month. The figures—or more accurately the figures in conjunction with the printed notes—never tell the keyboard continuo player what to play. They tell the player what the range of possibilities is for notes to be played, or, to put it another way, they tell the player by implication what notes are not available to be played. In many pieces the abbreviated nature of the figuring is taken to its logical extreme, that is, there are no figures. This in no way implies that the player is not meant to add notes and chords. It is not a situation in which anything different is going on. The player has to rely on other things—the listening and studying mentioned above—to glean the information that figures could have given.

From within the constellation of notes that would be acceptable to play at any given moment, then, how can a player make specific choices? This is both the most difficult part of continuo playing and its artistic/interpretive component. It is actually rare that a keyboard continuo player has to play notes—any notes—for the purpose of providing or filling out the harmony. This is true for two basic reasons. First, in most passages of chamber or vocal music, most of the harmony is provided anyway over the course of a beat or two, amongst all of the instruments or voices. (Clearly the thicker the texture, the closer this will come to being completely true.) Second, there is nothing in the rules or expectations of tonal music that says that every part of the theoretical harmony has to be present at all times. 

Instead, choices about exactly what notes to play (to add) at any given point are based on considerations that have nothing to do with completing the harmony as such. These are considerations of texture, volume, accent, rhythm, pulse, shaping of phrases or sections, and, very practically, both helping and not hindering the other performers. They all stem from the basic fact that adding more notes makes things louder and adding fewer notes or no notes makes things quieter. Thus “thicker chords on accented beats” is a simple but valid guideline, and there are plenty of others. More on this next month.