Dr. Liu makes a point

My colleague, baritone James Liu (who also happens to be an MD), made the following point after an earlier post:

It is also interesting, how the process is so similar, and yet so different, for those poor schmoes like me who are trying to memorize operas. On the one hand, the process is structured to provide a framework for gradually memorizing more and digging deeper, and you do get coaches and people who can try to help/cue/rescue you. But on the other hand, you have to master all the dynamics, shadings, texts, and cutoffs. You have to be able to listen for your fellow singers and be flexible enough to be ready to inflect your performance based on what they just did. You have to be able to move on a stage, act and react in a vaguely consistent character. And somewhere along the way, you have to tell a very specific story. Memory lapse stories in the opera world are legion, and just as colon-evacuating as those in the instrumentalist world (though with enough repetition, sometimes you dive into the void and muscle memory takes over ...).

 (James wasn't the only one who picked up on my "restless colons" comment in that post.)

James reminds me of my conversation with violinist Rachel Barton Pine, in which she talked about conductors who memorize. She said that, when the performance came around, it was common for those conductors to forget what was discussed in rehearsal. She also said, rightfully so, that in chamber music performance there are so many interpretive changes from ensemble to ensemble that there's really no point in attempting to memorize the music. In both cases, memorization gets in the way of the performance.

From my perspective as an instrumentalist, I would think that operatic staging would make the process somewhat easier. Does the physical stage movement becomes a type of mnemonic? In my limited opera conducting experiences, I came to feel that the stage direction took away from the music. James' comments seem to confirm my feeling, although eliminating opera in the name of better music-making is not a solution!

Singers, I'd appreciate your comments about your either staged or unstaged experiences with memorization. Do the words make the process easier or harder? Does the language make a difference? What are some of your lapse anecdotes? Do you have any singer-specific advice on memorization techniques?

Memorizing Bach and his fugues

Along with some music of the 20th and 21st centuries, there's not much music more difficult to memorize than that of J.S. Bach. As a keyboardist, I write from experience, and I know that other instrumentalists and singers also find Bach to be among the most difficult music to memorize. I don't want to speculate too much on why this is, but, as a case in point, I can honestly write that it takes me about twice as long to memorize a Bach movement than it takes me to memorize, say, Handel or Couperin. And it easily takes me about four times as long to memorize a Bach fugue.

In conservatory, I was taught to memorize Bach fugues by memorizing each voice separately. I remember doing this. It was absolutely nuts. I also vaguely remember bringing two voices together, then three, in all manners of possibilities. This was a long and arduous process, and, today, I really question the musicality and practicality of this approach.

Today, I memorize Bach fugues just as I memorize other works. I use my landmarks, metronome, practice log, mental work away from the instrument--you know, all those good habits. This entire process is intensified by the fact that Bach, unlike so many contrapuntal composers, accounts for every single fugal voice with striking regularity. (Handel's fugues, in contrast, are much more informal, with dropped lines and voices that don't follow through from their starting points.) In this memorization process, one doesn't just memorize notes or chords. For Bach, each note is part of a line that is carried throughout the entire composition; each line is a living entity that is co-dependent upon the other voices in the fugue. That, my friends, is the highest level of contrapuntal writing!

When I performed Bach fugues from score, the learning process did not require me to know what each voice was doing. I intellectually understood that I should be aware of the voice leading, but, since I wasn't memorizing the music, really understanding and committing to the counterpoint wasn't essential. At that time, I played fugues with a certain level of indifference to the compositional process.

Memorization has forced a new respect for Bach's fugues. The number of voices has become paramount to the learning process, something that wasn't so much the case when working from score. Once memorized, there are always new discoveries in the relationships between the voices. This can be a distraction on stage, especially when hearing something differently for the first time...in front of a live audience! For this and for so many other reasons, I consider Bach fugues to be the hardest things to memorize and, once memorized, perform.

Aside from the techniques already described, there's not much advice I can give to learning fugues any quicker than the very slow route. I do advise, however, thinking of the appearance of fugue subjects and answers as landmarks and mnemonics. Along with larger phrases, fugal entrances can make good landmarks. At the very least, the performer should know the placement of each and every one. 

Fugue subjects and answers can also be thought of as mnemonics behind the foreground of other material that's developing. In my earlier Wikipedia quote, mnemonics are used when "the human mind more easily remembers spatial, personal, surprising, physical, sexual, humorous, or otherwise 'relatable' information, rather than more abstract or impersonal forms of information." If we think of a fugue subject or answer as a spatial or physical element, we can concentrate more on the activity around it, thus aiding the memorization process.

I know, these are tough thoughts to consider. But, I assure you, memorizing Bach fugues is much harder! Good luck.

But wait...there's more!

Repetitious practice is not always the best practice, but it can't be ignored. In my worst, most impatient type of practice, I tend to repeat and repeat. I'm just too anxious to get the piece to playing level, and all I want to do is play. 

This sort of practice can be thoughtless. Even if it isn't thoughtless, it does little else but to assure a tactile foundation. Those musicians who bring a piece to playing level before attempting to memorize it have probably done not much more than repeat passages after, perhaps, marking fingerings and other foundational notes. After a point, it may be possible to play such work from memory, but is this really as far as you want to prepare before going on stage?

As I write this blog post, I'm asking myself, "What are some other early memorization techniques that I can use?" You should be asking yourself the same question, because there is no one way to do this, and you must always be creative with your approach. 

What about form? I've just started working on the slow movement of Bach's D Minor Harpsichord Concerto. I began by putting in landmarks, and that inevitably raised the question of the movement's form. This is a simple form, essentially a series of episodes flanked by a ritornello. Knowing that the ritornello only occurs at the beginning and at the end of the movement has helped me understand where to put some landmarks while also minimizing my work. (That is, if I learn the first ritornello, I don't need to learn the final one, since it's the same thing.) Looking closer at the form, I see that the ritornello functions like an ostinato bass, occurring regularly throughout the entire movement.

What about harmony? With that same concerto movement and now knowing about the ostinato bass, I can use harmony to help my memorization. For instance, the opening ritornello is in g minor. When it's repeated as a bass line with melody (my "A" landmark), we're in g minor, closing to a c minor cadence. In the next section (my "B" landmark), we start in c minor, but the ostinato bass appears a few bars later in d minor, eventually cadencing in B-flat Major. From there, there are some sequences and other harmonic machinations, all eventually bringing us back to g minor. 

None of this formal and harmonic analysis is worth publishing in a scholarly journal. I did this in sketch form for my own purpose: to memorize. With that in mind, you don't have to come up with a hardcore analysis. Simply do what works for you, but make sure that whatever you're doing is going to bring you to a better level of memorization. Think of these small steps as mnemonics towards the greater goal of memorized music.

Sometimes I work and work on a piece, and parts of it just won't stick. One way of dealing with this is to take time off, something I'll discuss later in this blog. By getting away from the piece for a few days or an even longer period of time, I might return and find the piece in a better place than when I left it. It's hard to judge how and if such mental rest will work, so you might not want to count on it. A more reliable technique is to play through your piece as you might have already been doing. Every time you get stuck, write down the location, taking detailed notes of where the problem spots are. In truth, you might find that there are less sticky places than you originally thought! This process in itself leads to better memorization, and it also gives you a practical foundation from which you can fix any problems.

I'll admit that I'm not always this intellectual in my memorization practice. I recently started work on an unmeasured prelude by Larry Bell. Because the movement isn't in my ear, I'm using the fingering I've established to help me memorize the right hand. This is tactile work--though mnemonic in its own right--, but it's with the goal of creating aural memory. The visual memory will follow once I have a better sense of how the piece "goes."

And what about aural memory? And what about visual memory? Both of these are very important to the early memory process, and I'll address them in my next posts.

What are some techniques you use in the early stages of memory? Did I leave something out? Post your comments here!

Some examples

I use repetitious practice to help to establish a sense of pattern. In my previous post, I wrote about how pattern could be manifested in terms of notes, harmony, or rhythm. Below are a few examples from my recent work. In each case, these are excerpts of the whole, and the entire pattern-finding process is greater than the single examples I give for each piece.

Bach: D Major Prelude, Well-Tempered Clavier I
click here for a score
I've been aware of this prelude for a long time, having played it as an undergraduate pianist. I already had a lot of the right hand (the "tune") in my ear, so memorizing the left hand became the challenge.

The left hand begins with a series of repeated notes (in octaves)--six, in fact. Measure 6 (a landmark) starts with four repeated notes, and then there are eight repeated notes starting in m.8. These moments of pedal points continue throughout the prelude.

Bach: C# Major Fugue, Well-Tempered Clavier I
click here for a score
This was a fresh learn for me last fall, having heard others play it for years but never having played it myself. Believe me, it took a long time to memorize!

I came to think of this as an etude on thirds. An obvious example would be the descending gesture in the right hand (m.7). This returns in the left hand in m.12 and many more times throughout.

On a larger scale, Bach takes more complicated patterns and has them descend by thirds: mm.16-18.  Later on, he overlaps the two descending thirds gestures in mm.31-34.

Bach: c# minor Prelude, Well-Tempered Clavier I

click here for a score
In m.5, a pattern emerges in the alto voice. The first pitch appears as a quarter note and is then repeated as a dotted half-note in the same measure. This alto line descends for three measures.

Couperin: Les baricades mistérieuses
click here for a score
In the third couplet, there is a famous pattern of deceptive cadence, repeating a bass line of D-E-F-G nearly four times. In looking at the tenor line, one sees an undulating D-C-D-C pattern. However, in once instance, the bass line is D-E-G-F. In this case, the D-C pattern is D-C-C-C.

Couperin: Le tic-toc-choc
click here for a score
Throughout the piece, the right hand has repeating fifths or arpeggiated patterns. In the rondeau, the arpeggiations conclude with the lowest pitch in the pattern, as opposed to what one would expect: the next pitch in the pattern (e.g., m.3). A hallmark of the third couplet, however, is that the arpeggiations conclude as one would expect: on the next pitch in the pattern.

Bach: Concerto in d minor, 3rd movement
click here for a score
One of the hard things about memorizing this movement is keeping the cadences straight. The cadential formula set forth in the opening ritornello (mm.8-12) never seems to come back the same way twice. The trick here is to establish how each ritornello cadence is different from each other and then find a mnemonic to keep you on track. Otherwise, it would be very easy to play any d minor ritornello cadence exactly as it appears in mm.8-12!

These are just a few micro-examples from my work. Obviously, the memorization of an entire piece will be made up of many of these types of examples. And, in the end, the performer may not even remember the process that led to full memorization--especially if this slow process results in very good aural and visual memory. Regardless of how one ultimately retains a piece, these patterns create essential mnemonics for the early stages of memorization.

A discernible regularity

Now that I've gotten us through the fact that not everyone agrees on memory types, I will admit that I'm not always sure when a mnemonic is memory and when memory is a mnemonic. I'm not a strong advocate for memorizing music that's already in one's fingers, and, as a general rule, I never listen to recordings in order to get to know the music that I'm working on.

That all said, let's presume that, in the early stages of memorization, just about everything is a mnemonic. The big advantage to this is that you won't develop tactile memory early on. This advantage will play out later when you might experience self-doubt during a performance; you'll have your mnemonics to rely upon.

Landmarks are mnemonics on a grand scale. I make them a part of my daily routine, and I find that they are essential to a successful memorized performance. But landmarks don't come right away. What about memorizing a few measures a day?

Repetition plays an extremely significant role in the memorization of even the simplest passage. You might find yourself repeating the same measures over and over again. The goal should not be to create a tactile memory experience, however. It can be used to create aural memory, but that's a hard thing to maintain when a piece is being learned from scratch. 

My repetitious practice creates a comprehension of pattern. I like the Wikipedia definition of pattern: "a discernible regularity in the world or in a manmade design." In musical terms, that pattern could be a simple regularity of notes, and it could also be the regularity of harmony or rhythm. It can be any combination of those three, and possibly even more elements. 


Presuming that you've already created a phrase-influenced landmark in a piece, repetition of that phrase should help you discern the patterns therein. But be careful: don't let your repetition ingrain tactile memory. This is where a timer and metronome become important. Limit your time with repetitious practice while also keeping your metronome slow so as to avoid tactile memory.

You say memory, I say mnemonic

About a month before I started this blog, an audience member approached me after a recital with the usual memorization question: "How do you do it?"

My response was, "Well, it's not really memorization. I use a bunch of mnemonics to create the experience--but if I didn't have those tools, the piece wouldn't be memorized."

"But you just played a recital from memory!" he replied.

Silly me, he was right. I was, in fact, playing from memory, but my perception was from the business end. I was doing hard work while this audience member was experiencing something that appeared effortless.

Look at the Wikipedia definition of mnemonic:

A mnemonic, or mnemonic device, is any learning technique that aids information retention. Mnemonics aim to translate information into a form that the human brain can retain better than its original form. Even the process of merely learning this conversion might already aid in the transfer of information to long-term memory. Commonly encountered mnemonics are often used for lists and in auditory form, such as short poems, acronyms, or memorable phrases, but mnemonics can also be used for other types of information and in visual or kinesthetic forms. Their use is based on the observation that the human mind more easily remembers spatial, personal, surprising, physical, sexual, humorous, or otherwise 'relatable' information, rather than more abstract or impersonal forms of information.

The word mnemonic is derived from the Ancient Greek word μνημονικός (mnēmonikos), meaning "of memory" and is related to Mnemosyne ("remembrance"), the name of the goddess of memory in Greek mythology. Both of these words are derived from μνήμη (mnēmē), "remembrance, memory". Mnemonics in antiquity were most often considered in the context of what is today known as the Art of memory.

Ancient Greeks and Romans distinguished between two types of memory: the 'natural' memory and the 'artificial' memory. The former is inborn, and is the one that everyone uses automatically and without thinking. The artificial memory in contrast has to be trained and developed through the learning and practicing of a variety of mnemonic techniques.

Mnemonic systems are special techniques or strategies consciously used to improve memory, it helps employ information already stored in long-term memory to make memorization an easier task.

Is there every any case in musical memorization when one isn't using mnemonics? Sure,
there are ingrained tunes that require no memory device (e.g., Twinkle, twinkle, little star), but, for the purpose of memorizing music to play in concert, a large majority of the work is based in mnemonic memory. 

A while ago, I wrote about three types of memory: tactile, aural, and visual. Because mnemonics play such an important role in the musical memory process, it almost seems as if they should be considered a fourth type of memory. But mnemonics are tools, not actual memory. Of course, this can get really confusing when we start needing mnemonics to remember our mnemonics!

I use mnemonics to create all three types of memory. From landmarks to form, from harmony to visualization of chord shapes, mnemonics guide the memorization process from the inception to the final stages.