grasp of structure

[demented cow]

One often reads that a pianist does/doesn’t have a good grasp of the structure of a piece. You can read for instance that Richter, Arrau, Schnabel and other ‘intellectual’ pianists had it, and some people say that Horowitz didn't have it (maybe because they are sick of all the overblown praise and want to say something critical that can’t be easily refuted by Horowitz fans). I only understand some aspects of structure, and would like to know more about what things there are in this domain that I can listen for and strive for when playing. I want to go beyond nice little details.
Some aspects I half understand already, as well as some questions I have are:

-Grasp of structure involves choosing a tempo that can be played without too much fluctuation. E.g. Gould strove to maintain a ‘basic tempo’ throughout the whole Goldberg variations. Do you know if it is common to do this throughout a whole set of variations or a whole sonata/concerto? I probably don’t have enough basic rhythmic sense to be able to tell whether the pianist is doing a good job in this area. Could people name recordings where a pianist (or conductor etc) obviously loses track of the basic tempo? Maybe this might help me to understand what to look for.

-Am I correct in saying that grasp of structure necessarily involves bringing out recurrent melodic material that holds the piece together. Examples could be the G-C-Eb theme in Beethoven’s Pathetique (which he seems to have borrowed from the Bach C-minor partita BTW), the C-Db-C Leitmotiv in the Appassionate or the C-G-G theme in the Waldstein 3rd movt. (which theme superficially appears to be just a bass accompaniment in some passages).

-Grasp of structure involves identification of the climax(es) of a piece and adjustment of dynamics & maybe tempo accordingly, and not focussing on passages whose main purpose is to lead to another more important passage. This seems obvious enough, but I think I could appreciate the point better if you can name recordings where professional musicians obviously stuff up in this area.

What other things am I missing?
In particular, I’d be interested in hearing about whether/how a pianist’s playing can indicate that they have a grasp of the structure of a whole multi-movement work (e.g. sonatas, concertos, suites) and even whole series like the Chopin preludes or Schumann Carnival.
Sorry to offload another long post on you, and thanks for reading if you're still there.

[pianistimo]

very good topic.  i like everything you've pointed out so far.  Chopin's Op. 2 variations might be one exception on the tempo.  The final two variations are two 'contrasting objectives.'  one is in minor mode and slow and the other brilliant and dramatic.  the final variation (reading from roeder's 'a history of th epiano concerto) is an alla polacca finale which implies using strong dance rhythms.

i wonder if an understanding of structure is tantamount to having taken some theory at the same time as piano and actually being able to analyze the pieces (however simply) and growing at this ability as your maturity in piano technique grows.  then, if someone asks you a question about the music, it makes enough sense to you that at least you can give some sort of adequate answer.

like art, you start looking for more details later on.  sometimes it helps to have a teacher just tell you what things tie the piece together (if you can't find them yourself).  some things are like secrets and hidden until someone points them out - and then you highlight those areas.

where i see the most ability to perceive a pianists grasp of structure is when they write their own piano cadenzas for other composers.  Do they follow the basic structure of that composer's own written cadenzas, or do they just get off on their own composition and ignore.  it's basically give and take and not all just 'my interpretation.'  i think one can learn a lot from structure by listening to, talking to, and watching conductors.  how they do their analysis, how they prepare for a performance, and how they interpret a piece during performance.  of course, they have to bring in so many different instruments, but piano does too, in a manner of speaking.  whenever we bring in a new voice, we have to let the other fingers soften to hear that new voice.  whenever a movement is over, we have to know if the composer specified to play the next movement without a break (as in Vaughn Williams Piano Concerto) or to make an appropriate break.  How long is appropriate.  What have we heard in other recordings.  What amount of time gives people the ability to adjust from one type of sounds to another.  Silence and the amount of it becomes almost as important or as important as sound.

[mrchops10]

This is a great, complicated topic too little discussed on pianoforum. My "motto" regarding structure relates to what Neuhaus said of Richter: "His singular ability to grasp the whole and at the same time miss none of the smallest details of a composition suggests a comparison with an eagle who from his great height can see as far as the horizon and yet single out the tiniest detail of the landscape." This is the essence of structure; it is far more vague than than merely holding the same tempo. It means precisely understanding how each detail, each tempo variation, fits into the whole piece, and to describe it generally is impossible. However, take Appassionata, which you mentioned; after the first theme when the left starts playing the triplets, most pianists "take off," so this section seems to have little to do with the rest of the piece. Pianist with little "structure" play pieces in many parts, and the playing is episodic rather than coherent.

For an excellent example of perfect structure without perfectly stable tempo, listen to Bronfman's new recording of Beethoven 3 with Zinman and Zurich Tonhalle. The whole piece works in one direction, never easing the tension. I had a difficult time breathing in the first movement, so inevitable and truly dreadful was it. Schnabel, whom you mentioned correctly as an "intellectual" pianist, definitely varied his tempi. Structure, however, can be dangerous. I miss sometimes in Richter's Chopin the necessary element of "improvisation" within the perfect structure. It certain composers it can seem to single-minded, at least to my taste. I'm glad you brought this up, however, because I so often hear so many talented students who make beautiful music that makes little sense, because, however colorful, it has little of the intellectual rigor necessary for a truly satisfying performance.

BTW, Horowitz was a structural genius in his reportoire. His great ability to perform miniatures by Schumann, Chopin, Scarlatti, etc. testifies to this. The old idea that he was not "intellectual" is pure German snobbery, which holds the three Bs (Bach, Beethoven, and Brahms) on a higher level than anyone else and snubs whoever can't play them satisfyingly. For an example of Horowitz's structural intensity, listen to his Mozart. His transcription of Sousa's "These Stars and Stripes" is also highly structurally sound, moving with ease to greater and greater pitches. Very often, however, his pianistic genius makes the listener forget his intellectual rigor. Structure, however, is a virtue of all music, for Bach to Liszt to Prokofiev.

[pianistimo]

mr. chops, you and demented cow are most interesting to talk with.  thanks for sharing all that you did.  i will listen to bronfman's new recording of beethoven 3 and compare it to schnabel - and listen for richter's lack of improvisation.  these specific  things to listen to are really like the eagle eye you pointed out.

it does seem like truly great performances happen only once in a very long while (listening live).  at least we have cd's to choose from, but it almost takes having teachers *such as yourselves - point out to students what they personally find enjoyable and what they don't.  of course, once you get familiar with the idiosyncracies of tastes and all, you may find that you start being more picky about your music (food). 

i think this is good, actually.  but, for others, it ruins the pleasure.  they don't want to be told that hamelin isn't always perfect and doesn't always interpret every piece he plays with a full idea of the structure. or, perhaps better said, you can hear the structure rather than feel it in some pieces.  in fact, it may be that many concert artists simply struggle to play so much music all the time, that their real achievements are in the pieces that they come back to and rework from time to time.  like a good recipie, refined.  but, of course, who am i to critique the likes of hamelin.  all i know, is when i really like a performance - it is like being shot or something.  it hits me that it was the BEST performance i'd ever heard - and i try to think about what made it that way - and what the conductor, pianist, and/or orchestra did to make it so much to my personal liking.  (i know for many people things that they like are personal preferences).

for me, my personal preferences are to distinctly hear the lines of the music, to hear the piano imitate the orchestra fairly precisely in terms of tempi, tone, dynamics - and to have some sort of give and take between pianist and conductor in a piano concerto.  i hate it when you can tell that neither wants to give up control and both are struggling.  or the conductor takes over and won't be flexible.  or the pianist goes berzerk and takes the tempo too far, and the orchestra struggles to keep up.  you have to be thoughtful of each side.

[rohansahai]

This is how I work on structure:
Draw a horizontal line (representing the length of the piece). now draw vertical lines to divide it into sections each representing the major segement of the piece. Don't divide it into too small units....usually about one unit per minute or two.  Rank each of these units and figure out the 'most important' one or the 'peak' so as to say. Now, try and stay withing the 'scale' of each of those segments. For instance, an 'ff' marking in the top ranked segment will be louder and more distinct than an 'ff' marking in the lower ranked ones. Try and keep this rating in your head throughout and your playing will automatically become 'structured' as to say.
N.B. This applies only to large scale works. The small scale works are to be approached very differently.

[pianistimo]

being too careful doesn't work either.  i think it is assuming the role of 'conductor' if you are soloing, or interchanging the role if playing a PC.  taking charge (not worrying what people will think or say).  i'd rather hear someone take charge than play it carefully (too slow of tempo) and correctly.  (yet, i see what you say about analyzing pieces and not having too many high points)

there is a decay to sound that one of my prof's pointed out.  so when you play a slow movement too slow, you start hearing decay before the next note.  partly, it's knowing what your instrument is CAPABLE of.  then, structure becomes more than just an awareness of dynamics, but of shadings, of pairing of notes (even if it is a simple run), of distinguishing a slight hold here, or a faster pace there.  a sort of blowing with the wind.  all really great performances, to me, have an element of improvisation (that someone else brought out).  it is a 'live' performance (not performed the same way before or after).  even though the composer wrote down the notes, they are alive and made to be unique by this particular performer.

structure and taste probably go somewhat together.  i mean, if you feel something in your heart, it is your gut instinct.  but, if the conductor or your teacher *or you find out by studying that something should be a certain way - then you try to accomodate your feelings to the most correct playing of the passage possible.  maybe in mozart it would be the distinction of one player playing ABSOLUTE staccato notes vs. a 'melted version' where they are somewhat marcato yet still light.  this small distinction starts giving people who are critics something to go by in the first few measures of a performance.  structure might mean simply understanding what tricks are in the bag (pallette of colors) and how they are spread throughout the piece.

maybe kind of like a first reading of a novel.  you take everything literally.  you read it with a sort of innocence of what it's about (especially if you don't know much about the topic).  then, the second time you read it, you can mark the highest point (as rohansahai pointed out), then the third and fourth time if you are reading it to someone else - you can make certain words mean this and that by how you say them - and it all leads to the intricacies of the plot (speeding up the reading when very intense - slowing down in suspense).  to me, it's like a curtain being drawn up, and you (the pianist) start a literal drama with yourself.  you can be the protagonist and antagonist and the conductor all at once.

[pianistimo]

has anyone ever made up a pallette of their own sounds.  i know they are already categorized on a digital piano, but what if we catalogued our own 'tricks' and tried to remember how and when to use them in various places.  i'm learning that's what my teacher does.  i can hear when he's trying to be a human voice, when a violinist, when a percussionist.  it's amazing what the piano can sound like by the various kinds of touches.  the combination of actual touch and intention.  things you can do at the precise moment the hammer hits the keys.  in those few seconds, you can have the pedal at various moments of depression, you can leave your finger on the keyboard, spring it off, variations in between. 

a good performance is a lot of great variation.

[rohansahai]

being too careful doesn't work either. 

Yes. It simply depends on what you want to project in your performance. The method I described above gives you an almost "perfectly structured" performance. But perfectly structured performances do not mean that they will be good performances. You do have to deviate from it and add certain 'surprises' so as not to bore the audience. The more you deviate , the less structured the performance becomes, so you just have to find the right balance for your taste. Richter was often accused of being the "too structured" type and horowitz was was more of the 'surprise' type. That's why Richter was very good at the large works whereas Horowitz did the shorter works more justice. It is important to figure out your taste. But it always helps if you have a basic framework of the structure in your head. i.e. start with the method i suggested and then add all the required 'spices' and 'exceptions to the rule' s till you arrive at your "ideal" performance.

[alzado]

Here is an interesting exercise in structure.  At least, I can share one of my experiences.

Some of you may know Erik Satie's compositions of the 1890s.  Satie left out all bar lines, all phrase markings, and if he put any dynamics or tempo guidance -- he often did not -- his notations were intentionally absurd.

Playing some of the longer and more complex pieces of Satie's from this period, I found myself penciling in structural demarcators and other notes. 

For example, penciling in some phrase lines, etc.  Marking some dynamics.

I started doing this after I had played some of these pieces for several days, and realized that I was swimming in jello.  I needed to develop a sense of the phrases, and of the respective sections of the pieces.  Where separate statements within the compositions began and ended.

After playing and thinking enough, I did begin to get a sense of what this composer was trying to do.  By pencil-marking the pieces, I was trying to capture and retain my insights.

My piano teacher and I had some interesting and, I think, rewarding discussions over the playing of these pieces.

Try it yourself.

Just a thought . . . .

[prometheus]

Don't write this off as an excuse to attack pianists without you having a chance to defend them.

Surely this is very subtle, complex and it does contain the essense of european classical music. I am not saying Horowitz isn't very good at this. Maybe he is actually quite good at this, I don't know. But surely there are differences between pianists.

[rc]

then, structure becomes more than just an awareness of dynamics, but of shadings, of pairing of notes (even if it is a simple run), of distinguishing a slight hold here, or a faster pace there.  a sort of blowing with the wind.  all really great performances, to me, have an element of improvisation (that someone else brought out).  it is a 'live' performance (not performed the same way before or after).  even though the composer wrote down the notes, they are alive and made to be unique by this particular performer.

It means precisely understanding how each detail, each tempo variation, fits into the whole piece, and to describe it generally is impossible.

After playing and thinking enough, I did begin to get a sense of what this composer was trying to do.  By pencil-marking the pieces, I was trying to capture and retain my insights.

I like these.

I'd say structure is to know the piece well enough that you feel you understand what the composer was doing, you get why everything was put where it was. Intellectually and emotionally (I think it was 'affekt' in baroque, I like that term). This is the point where I begin to feel free/secure in the piece and it takes on that improvised quality.

A little less abstract, I like rohansahai's idea of figuring out where the climaxes are and building around that and the idea of having one main climax (more than one detracts from all). I've read some pianists calling it "the point" of the piece, which is a good way of looking at it.

Another thing that occurs to me is to pay attention to harmonic and melodic cadences, that's a big hint of how the composer was thinking structurally. Tension/release as well, you can sometimes get right into a building tension and make for a more powerful release (as opposed to playing through without any intent).

There are countless little things to be aware of, but I still see it as ultimately being able to see the piece the way the composer did.

[pianistimo]

yes.  i agree that rohansai is right!  just was adding to it a few things more.  and, i also like the idea about cadences becoming tenser or resolving.  i suppose that is what i was lacking the first time around in the mozart fantasy.  just buzzing right through some important cadences.  i thought my teacher intentionally hurt my feelings by telling me i didn't have any at that point, but what he meant was to linger over a particular cadence.  when you haven't heard a piece before, sometimes you just gloss over very beautiful parts .  now i try to linger more on parts that seem to have some importance in either resolving or possibly speeding up a little when building tension.

also, he asked me to do a surprising thing, and go past what i thought was the climax of the piece (fast section) and keep the dynamics and drive up.  has anyone else done this in any other pieces - where you don't suddenly diminish, but practice keeping the elements up - so that when you perform it doesn't have a 'minor' climax, but a much more 'dramatic' one?  i don't know if that made sense, but i keep the speed and dynamics up past the point that i think the actual climax was.  maybe it helps to practice this way, but not necessarily perform it that much past the desired point.

about marc-andre hamelin...i have the greatest respect and i didn't mean to use the thread to be critical.  basically, all pianists have their best pieces and then their so-so pieces.  there's really no distinction between him and other concert pianists - and even richter and horowitz.  obviously they are all doing something right to have such great careers.  and, they must understand structure and taste very well beyond me.  i just know, when i listen to a piece - if it moved me.  if i would play it similarly.  or, if the piece 'made sense.'  and, as pointed out, if it would make sense to the composer.  these are all sort of dependent on your level of piano.  i am far under the level of my teacher, but everytime something is pointed out - i try to remember it.  i am in love with the sounds that a good piano is capable of.  sometimes all i need to dislike a performance is a piano that sounds too brilliant, or plays too softly.  i tend to like huge variations in dynamics - and some the of the very easy to play pianos only give you piano, mezzo piano, and forte (and double forte).  to me, that's kinda boring.  i want to hear the triple pianissimo as well as the triple forte.  i mean, that's why we bother to play on grand pianos. 

[demented cow]

Thanks to the people who responded. Here are a few return remarks.
-re Horowitz: I read somewhere that one of the German Beethoven players (Serkin or Backhaus?) heard Vladmir doing the Hammerklavier somewhere and found it totally mindblowing, which would have been unlikely if he were a structural dimwit.
-I'm still waiting for an answer to the following: what (if anything) can performers do to unify a work consisting of several separate movements?
-Some of the above comments (unintentionally) confirm my suspicion that many aspects of the sensitivity for the overall structure of a piece (whether in performing it or assessing someone else’s performance) are pretty much inborn: you either have or you don’t. Knowing/analysing the music seems only a necessary, but not a sufficient condition for this. Take the Appassionata example again. I know every note of it, having played it (badly) from memory. I can theoretically grasp the point made by mrchops10 about people running away with the triplets bit, but I doubt I would notice this in a performance by a professional pianist. And frankly, I don’t think I would give a flying duck if I did. Ok, so somebody radically alters the overall tempo when the triplets start, but does this NECESSARILY make the performance unstructured or otherwise weaken it? After all, all music breaks down into sections, subsections etc. with different melodic material, rhythmic characteristics, textures, etc., so is there anything inherently evil about fanfaring the arrival of a new section by a different, even radically different, basic tempo (unless the performance is meant to be danced to). Mind you, maybe this idea is a residue of years of listening to myself changing tempos to accommodate my bad technique...
Moreover, I could go through the score with you and point out lots of important structural things like the first and second subjects in the 1st movt and what he does with them in the development, how these themes relate to themes in the other movements, and so forth, but at the moment I can’t see how having discovered these things (ingenious, beautiful though they may be) is going to affect my assessment the relative merits of performances of the piece by Arrau, Brendel, Kempf, Horowitz, the Naxos dude or whomever else. (BTW, I don’t mean that assessing their relative merits is an end in itself, I am just mentioning this as an example of a way of thinking about what true artistry consists of, which will hopefully lead to a greater enjoyment of my cds and better playing.)
Maybe I am beyond help, and will go through life only enjoying more superficial things like someone’s velvet touch, hair-raising octaves or the odd beautiful phrase. If that’s the case, there are scores of cd’s I might as well sell, because I simply can't see the good things in them that other people see.

[rc]

-I'm still waiting for an answer to the following: what (if anything) can performers do to unify a work consisting of several separate movements?

I'd say a lot of the unity between movements is the composers doing, contrast is as important as unity between movements. Something that comes up describing single movements as well as multiple movements is the general idea of having a starting point, deviation and returning... Like a journey and coming back home. Movements ending in the same key they began, the final movement being in the same key as the first...

-Some of the above comments (unintentionally) confirm my suspicion that many aspects of the sensitivity for the overall structure of a piece (whether in performing it or assessing someone else’s performance) are pretty much inborn: you either have or you don’t. Knowing/analysing the music seems only a necessary, but not a sufficient condition for this. Take the Appassionata example again. I know every note of it, having played it (badly) from memory. I can theoretically grasp the point made by mrchops10 about people running away with the triplets bit, but I doubt I would notice this in a performance by a professional pianist. And frankly, I don’t think I would give a flying duck if I did. Ok, so somebody radically alters the overall tempo when the triplets start, but does this NECESSARILY make the performance unstructured or otherwise weaken it? After all, all music breaks down into sections, subsections etc. with different melodic material, rhythmic characteristics, textures, etc., so is there anything inherently evil about fanfaring the arrival of a new section by a different, even radically different, basic tempo (unless the performance is meant to be danced to). Mind you, maybe this idea is a residue of years of listening to myself changing tempos to accommodate my bad technique...

I think the idea mrchops is getting at is that if you vary the sections too much it begins to sound like they're all seperate, as opposed to each idea leading smoothly into the next which would be more coherent for the listener.

Tempo fluctuations is getting into matters of taste. I personally prefer to keep the beat fairly stable throughout a piece, some accelerandos, ritardandos and very subtle rubato throughout the piece, but I like to reach the end at pretty much the same pulse as it began. When the beat fluctuates too much it seems too chaotic to me, hard to follow and distasteful (to me). Others like to be more free with their tempos. I figure so long as you're in control it's up to your personal tastes. Technical limitation is no way to choose tempo .

Moreover, I could go through the score with you and point out lots of important structural things like the first and second subjects in the 1st movt and what he does with them in the development, how these themes relate to themes in the other movements, and so forth, but at the moment I can’t see how having discovered these things (ingenious, beautiful though they may be) is going to affect my assessment the relative merits of performances of the piece by Arrau, Brendel, Kempf, Horowitz, the Naxos dude or whomever else. (BTW, I don’t mean that assessing their relative merits is an end in itself, I am just mentioning this as an example of a way of thinking about what true artistry consists of, which will hopefully lead to a greater enjoyment of my cds and better playing.)
Maybe I am beyond help, and will go through life only enjoying more superficial things like someone’s velvet touch, hair-raising octaves or the odd beautiful phrase. If that’s the case, there are scores of cd’s I might as well sell, because I simply can't see the good things in them that other people see.

We may be wading into matters of taste again with this. The value of hearing different interpretations of a piece is how differently one person may play it from the next. If you're familiar with the piece, you can listen carefully and pick out all kinds of little details that each person does differently, which is great to have a few versions to compare the particular piece by, and also to get ideas for your own playing. Maybe it's how a trill is played, how a phrase slows down slightly towards the end or how someone plays a low figuration as a rumble instead of clearly. Pay attention to these little details and you'll have a wealth of ideas to choose from in your own playing. You'll like some things and not others, develop your personal tastes, change them... Analysis comes secondary if you ask me; you analyse after you've heard something in order to find out why you like/don't like it.

I recommend less worrying about how other people hear music, and more listening for yourself. No matter how much you learn about music, don't lose touch with the innocent listening that first drew you in.

[crazy for ivan moravec]

i have a hard time myself trying to figure out how to put details of a musical work together. somehow, i try my best to do so.

however, i know when music is well-structured and "well-connected" when i HEAR one and FEEL one, in recordings (sometimes live performances can be difficult to decipher because the presence of the performer clouds my listening to the music itself, but i appreciate live performances in a different way).

when i deal too much on details in my study of a musical work, especially with large works, i get them all mixed up and find it almost impossible to put them all together (although dealing with details is really helpful in making decisions on interpretation objectively).

but in listening to a recording that i feel is well-structured, one by one, i am able to realize the reasons why the pianist did a particular segment/section/passage this way and not that way; and surprisingly i can hear even the minute details.
 
one recording that i can site as an example of what i think is the best-ever structured interpretation is Ivan Moravec's Chopin Polonaise-fantasie, or, maybe his ballade in fm too.

the tones change in color with subtlety, depending on the flow of his thought. he does not use pianistic effects simply for the sake of surprising his audience (in my opinion, like Horowitz would do) but rather he uses them for bringing out his thought and meaning on a particular musical piece. the peak is always given the throne, but he does not neccesarily tell his audience that "this is the spot where you'll find gold!!!" Rather, he does it in a more subtle way to create an effect more to the FEELINGS of the listener than to the brain. when he colors the music with his rich palette, you will be able to hear the relativity of his dynamics. everything is so well connected that even his rubatos (which some people think is too much) mean very well for the other technical aspects of his interpretation: sound, tones, dynamics, the decrescendos, accelerandos and diminuendos, the nuances, etc! in other words, i actually find it too deep an interpretation. deep, because he deals even with the details of details (i dont know if he's aware that there is an impression of him doing that), carefully connecting each one.

i also think that in dealing with works that are comprised of small pieces (example, Schumann Kinderzenen), the set of pieces should be thought of as if they were the parts of a musical form. it's like the sonata form having the parts: exposition, development and recap. if u took one piece out of the set which follows this kind of interpretation, and play it as an encore, i think the performance will become boring because it is taken out of context. it is so unlike Horowitz's famous traumerei encore- it sounds so interesting because it was interpreted as a single complete work and also performed as one even though it is part of the whole set.

one recording i heard which i thought was so badly structured, is horowitz's chopin barcarolle. this is just my opinion. i thought that it is nothing compared to Moravec's chopin barcarolle. then again, we don't want to criticize the romantics for the reason that they played simply with their heart and instincts, it was the tradition then. and simply, we don't know how it affected the people back  then. maybe they had felt it to be really well-structured, who knows... we think differently now.

but i say that the best way to get a grasp of structure is to hear it, then feel it! if it touches the heart, then it must be the best option.

working around one's own recorded performance of a piece is therefore good practice. the performance or interpretation does not necessarily have to please every piano expert. however, if you are pleased and convinced of it, then do it that way. because the more you want it that way, the more conviction you'll have in playing it. the more conviction you'll have, the more people you will be able to touch and move (although not neccesarily make them agree with your ideas).

all the best.

- clement >crazy for ivan moravec<