Tag: Performance

Bach’s counterpoint – who’s on “top”?

The amazing thing about Bach’s counterpoint is that the musical meaning of a piece does not change if you transpose one or more voices to different octaves or simply rearrange the order of the voices from highest to lowest.  The voice that is originally on top does not have to be on top for the counterpoint to have the same effect.   It is as if each voice in a Bach fugue is transparent to all the others.  No voice, regardless of its pitch range, is opaque in that it blocks or occludes any other voice.   The voices shine through each other.  Figuratively speaking, every voice is on top.

Being the top voice loses its automatic prominence to the ear.  All of them, in effect, are on top, or none of them are.*  You can think of it either way.

It is a mistake to concentrate our attention when playing on the top voice, or to whichever voice currently “has the theme.”

Of almost no other composer is this true.  In a Beethoven sonata, for instance, only rarely can you move the right hand notes into the bass and the left hand notes into the treble and have a result that ‘works’ as well musically, that aesthetically resembles the un-switched version.   In Beethoven, depending on which octave a note sounds in, it creates an impression of sound that is different than the same note sounding in a different octave.  Each octave range has its own sound-personality.

One of the ways of determining the unique properties of the musical space of a great composer is to experiment subjecting the piece to certain specific types of distortions and then noting the results sound-wise.  One such attempt at distortion is the rearrangement of the voices described above.  Here are some others:

– how much can you change the tempo of the piece without distorting the meaning of the piece or its character?

– how much can you change the rhythm of the melody without changing the essence of a melody?

– how much of the essence of the music is changed by changing the instrument or instruments playing it?

– to what extent will the integrity of the piece be ruined by starting to switch around parts (what comes first, what comes second, etc.)?

The answers to these questions will be obvious to the ear, and will vary from one composer to another.  They reveal to us what some of the basic, “geometric” properties of that composer’s “musical space” are.

These experiments are similar to those in the mathematical subject of topology, where a common question is to ask how much one can distort a shape and still have it retain certain basic properties.

*This is why Bach’s fugues for organ work as fugues even though it is often the case that each voice sounds in several octaves at once because of linking an eight foot stop with a four foot stop and/or a sixteen foot stop.  The same applies to Bach on a harpsichord when ‘couplers’ are used to cause a note to sound in more than one octave at once though we are only pushing down one key.

Leave Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Rubato. How to make the piece “breathe”

P. is an adult student who is playing Brahms.  She asked me, “How do I make my playing sound like the recordings of it I enjoy?”

“I know there is flexibility in the performer’s playing. I sense it, I ‘hear’ it.  It sounds so natural and obvious, yet I don’t know how to reproduce it when I am playing the piece.  I can’t make it sound as musical.  Why?”

My answer: “Welcome to the wonderful world of rubato!”

The following was my attempt to “teach” rubato.

First, I said that I thought the way by which she evaluated her success at playing a piece was ‘measured’, and in inverse proportion, by the number of mistakes.

P: “Shouldn’t I know all the notes solidly before I start playing around with them musically?”

Me: “To develop rubato, you may need to change some long term habits.  Don’t evaluate your success by right or wrong notes.  And don’t wait until you know the notes of the piece before you add musicality as if it were an added ingredient.  It’s the other way around: you learn the notes more quickly if you play musically if, for no other reason, than  you come to understand more quickly why the composer the chose the notes that he chose.”

Michelangelo said that he could envision the finished shape of his sculpture when he was looking at the untouched block of marble, from which the sculpture was to be carved.  He would then cut away the material of what in retrospect would be the “negative space” surrounding the inner form, until the positive space of the inner form was revealed.

His first cuts of the chisel were probably rough and not meant to be specific.  Your first experiments at rubato can also be rough and random.  Pick a spot in the piece, any spot, and try speeding up or slowing down the next group of notes.  As the sculptress of the musical phrase, you then ask yourself, did this change, did this bending of the steady tempo into a curve, make the piece sound better?  Did the change make the music make more sense to the ear?  If not, pick a different random spot and try to distort it again with a change of speed lasting a group of notes.  It will not take much experimentation to discover the proper places and degrees of “bending” in the phrase.

In the above analogy, the un-hewn block of marble stands for the piece played with correct notes but without rubato.  The final form of the marble is what results from the rubato.

A “planar” (and “plainer”) version of the Venus de Milo would be far less inviting to the eye than the curvaceous form we see in the Louvre.  A musical phrase is an elastic thing, it stretches and contracts.  It is a living thing that constantly breathes in and out.

The violin bow must distend the linear shape of the violin string before it will ring true to its pitch.  Harmony is always the result of a reconciliation of conflicts.

In the following more geometric analogy a base,  or unvarying tempo, is like the x-axis of a graph: an unvarying straight, horizontal line.

The application of rubato creates a curve that is sometimes above the x-axis and sometimes below.  If the y-coordinates of this curve are summed up over an entire cycle or phrase then, the positives added to the negatives, in a true rubato, will result in the values canceling each other out by the end of the phrase.  This leaves us arithmetically a net value of zero, but at the expense of loosing any aesthetic shape.

This balancing out can happen in many ways. Only rarely will the speed-ings up (positive y values) and the slow-ings down (negative y values) exhibit the same values as each other, although this can happen in principle.  More likely there will be an unequal division of the musical phrase through time in terms of when accelerando is taking place and when a ritard is taking place.

Here is one of many possibilities: a few large positive y values (a few notes dramatically speeding up) can be balanced out by a large number of small negative y values (a greater number of notes slowing down but little by little).

The result of this final balancing out by the end of the phrase is that the next phrase tends to begin at the same time on the clock that it would have begun if the phrase had been played without rubato, but in a steady tempo.  But in the interim instead of a static phrase we have one that exhibits a dynamic equilibrium of opposing forces.

P’s first attempts at rubato resulted in her playing a measure or two in one steady tempo followed by another phrase at a different, but still steady, tempo.  Rubato probes more deeply into the evolving phrase.  The changes occur more frequently and dramatically.  Slight ritards and accelerandos may not always be enough to carve out a shape to a phrase.

A last word about “emotion.”

P. said, “I try to feel the emotion in the Brahms but it doesn’t seem to emerge. ”

Joe: “You don’t put emotion “in,” the emotion comes “out” as a byproduct of the bending and shaping of the phrase.  Then the phrase resonates with emotion, both pianist and listener experiencing it for the first time.  When properly shaped the phrase ‘rings’ like a bell when struck.”

Leave Comment

Leave a Reply

Your email address will not be published. Required fields are marked *


  1. This is brilliant. From a listener’s point of view, maybe you don’t hear the rubato, but the curves in the phase that cause you to emote. What an insight you have here! Genius! I guess this is why that I always hear music when you play, and not just a composition.

  2. I agree that one can create rubato by trying things out and manipulating the tempo and push and pull of the notes to see what sounds or feels best. However, I think that the true rubato that makes us want to play comes from emotion. So I don’t know if I agree that you don’t put emotion in, that it comes out. I think it only comes out naturally if you have it first – you don’t ‘put’ it in, but you feel it, you hold it. I think that, like the Michaelangelo statue, an emotion or thought triggers how the rubato happens. Otherwise it sounds too calculated and logical – unnatural. Or, even if it’s beautiful, it’s lacking that something – the feeling and trueness that we only get when we can hear it or imagine it how we want it before. That hearing it or imagining it isn’t always thought out of course, most of the time it happens naturally in a performance state. But we can practice it the same way. I do agree that every time we play something, we need the musicality and emotion from the beginning, but that happens when we have an idea of how we want it at the beginning. The same way that when we listen to our favorite performances, we feel something, and when we think about a certain piece, we feel something (or lots of things!). We can always change it (and should, every time because no feeling is alike). I would approach it by having someone have a story or a feeling to go with a chunk (meaning they know how they want that chunk to go), then inner sing or play a small chunk in that way, then play it right afterward. Trying this in different forms I think can trigger the rubato naturally, and do what music is supposed to do – communicate. I guess the main difference in method then is more imagination or feeling on the front end rather than playing and seeing how it makes you feel.

Being in the groove

“Dear Matt, thank you for sending me the New York Times article written by Ben Zucker about the violinist Saul Lipshutz (a.k.a. Saul Chandler)”

Last night I was writing a blog about what it feels like to me to be “in the groove” while playing.  I described it as being by the banks of a river and watching the water flow by, I.E. I observe the notes flowing by, sometimes slower sometimes faster, but I am myself standing still, just observing and not making any particular effort to create the sounds.  Late in the Times article, Chandler is quoted as saying,  “Nobody wrote music… They heard something. Except Mozart. He wrote the same thing over and over.”

I believe both of those things.  It put me back to being a teen and thinking that Mahler (whom I loved) wrote the same symphony nine times, but did so only to remain constant and loyal throughout his life to that unchanging truth which he always reached out towards. For in a world where time brings about constant change, the only way to remain faithful to something is to take into account the changing perspective of time.  Whatever the “local” distortion of time is, we have to take it honestly into account, but still see through it to the unchanging truth.

Leave Comment

Leave a Reply

Your email address will not be published. Required fields are marked *

If a computer can play more evenly than a human pianist, is the difference worth it?

My answer requires that we think ‘through’ time as it evolves or passes in the performance.  A phrase of music, especially in the hands of a master composer, is something that ends up with a balance of opposing forces, but does not start that way.  It is like a seesaw (with the fulcrum at the mid-point).  Person “B” (for beginning of the phrase) sits down on one end and the horizontal line that the see saw would have if in a balanced state is replaced by one end touching the ground and the other end twice as far from the ground as it would be if things were level.

Person “E” (for end of the phrase) sits down near the center of the see saw.  Although of the same weight as person B, from E’s initial position the see saw is still well out of balance.  But gradually through time (as the phrase evolves), E begins to move along the see saw until she reaches the opposite end from B.  The seesaw is now in balance.  While the seesaw can remain in this state indefinitely if E and B don’t move from their positions, this state exists only transiently in music; at the moment of the completion of the phrase, just as another phrase begins, and there is another general progress towards a balance of musical forces.

I grant that this description is over simplistic, but it is true to the basic principles of good playing.

However if the human rendition is too uneven, and duration through time doesn’t eventuate into a dynamic balance or equilibrium, then it is preferable to listen to the computer play its notes, where each quarter note is exactly equal in duration to every other.

I had an early experience of this.  In the 1980’s, when computers were still rare, so were recorded performances of Bach’s 215 cantatas.  There were few choices.  One of the first recorded performances of most of the cantatas was conducted by Nikolaus Harnoncourt.  The playing was so bad in general, that just hearing the computer sound the correct pitches at the correct times, in the correct rhythm, was sufficient to reveal the greatness of Bach’s music.  In this case, the computer’s rendition was vastly preferable to the human rendition.


Comment from Roy Doughty sent by email on March 30:

The see-saw analogy of a musical phrase in this blog and in ‘Introducing the Beginner to the Keyboard‘ is brilliant.  And the insight into clarifying playing at octave intervals in ‘Melodies in Octaves’: also brilliant.

Leave Comment

Leave a Reply

Your email address will not be published. Required fields are marked *


  1. Perhaps the question could be rephrased: “Which would you rather listen to, a poor human player playing sublimely composed music or a computer rigidly reproducing a score?”, Joe. Your example at the end points this out.
    When extending this to composers other than Bach (and there was a reason that ‘Switched on Bach’ wasn’t ‘Switched on Chopin’) I think the question of whether a computer, even now, can interpret a composer’s intentions better than a poor or even mediocre human player is not settled. Even if it were to be settled in the computer’s favor (does that even make sense?) the point is whether you, as a human, would rather listen to and watch a human relating a human composer’s work to you, another human, or listen to (and watch?) a computer get in the middle of things. To take the discussion a bit wider, would robotic baseball be more fun to watch than the simply human?
    Psychology and even errors are part of the game, I think, whether we talk about baseball or music.

My least favorite Brahms Symphony, up till now.

At last I’ve heard a decent performance of Brahms’ Second Symphony!

I am a Brahms junkie. I live and breathe Brahms. The air he breathes is the air I breathe. His emotions are my emotions. When I was growing up I felt many new emotions for the first time by listening to Brahms rather than from experiences in real life.

Given this enthusiasm, please understand that when I say that I have a “least” favorite Brahms Symphony, it is all relative. My preferences, in descending order, is: #1, #4, #3, and #2. The second had never satisfied me. Every time I hear a new performance of it, I raise my hopes, but they are dashed in short order.

In retrospect I realize that how I esteem a piece has often been dependent more on the quality of the first performance I hear of it, rather than what I come to know of the piece itself as my experiences with the piece grow in number. Let us just say that tonight, I heard for the first performance of the Second Symphony that reached all the standards I impute to Brahms. It was Martin Alsop conducting the London Philharmonic. I missed the first two thirds of the first movement. It took a few minutes to accustom myself to the interpretation, but after that I was sold:

1) Meaningful and intelligent parsing and phrasing (not prefabricated phrasing that approximates but never custom fits the notes, or the orchestration or the harmony). Each phrase came out of the unique meaning of a passage.

2) Pellucid orchestration: clear, harmonious and well balanced as is true of all the great composers when writing for orchestra.

3) Harmonically well-informed playing. So many interpretations skim over, merely summarize the obvious steps in Brahms’s masterful harmonic progressions.

4) Every detail of the harmony had its proper aesthetic sense. No passing tone, no tone of embellishment at all, that didn’t gain meaning from the governing chords.

5) A masterful realization of Brahms’ mastery of counterpoint that made no attempt to avoid difficulties due to texture or orchestration.

I was engaged; I was not bored by what previously had seemed like trite overindulgences in the more superficial emotional aspects of the piece – mixed with ignorance of all that came in between.

So, although my order of preference among the Brahms symphonies remains the same, the Second Symphony is now not as far away from as it was from the firrst symphony.

A caveat to myself. I’ve heard other Brahms performances conducted by Martin Alsop that have engaged me the first time I heard but whose appeal fell off with subsequent hearings. I’m optimistic this time will be different.

Leave Comment

Leave a Reply

Your email address will not be published. Required fields are marked *