Piano World Home Page Forums Our Most Popular Forums Pianist Corner Learning pieces backwards

I did an experiment learning Bach's Invention #9 in F major: I learned it backwards. Last page, then second page, then first page. I know I'm always so much better at the beginning that it gets boring to play after a while. I thought doing this would give me "more time" with the end and more difficult part of the invention.

It had some interesting results. I googled a little about playing backwards and it looks like there are different approaches. I didn't think I was original, I just wanted to see how it worked out for me before I did some learning. (When I was growing up playing, we didn't have the internet and the ability to google learning. All I had were my teachers and whatever I could hammer out myself. Playing piano in the 21st century is interesting.)

How do you all feel about playing things backwards? Are there pitfalls I should avoid doing this? I'm thinking of bringing the approach to my white whale Chopin Scherzo in C# minor. I'm so bored with playing the beginning slowly, I think the end might give me a bit more energy.

The advantage to this method, as you may have realized, is that it's not only the first section that gets practiced the most. Many who always start at the beginning are much more secure in the first section of the piece but less so at the end because they have not practiced the last section(s) as much, having run out of time with the conventional from beginning-to-end method.

I often have used this method, and I've used another method that I've devised for longer pieces to avoid the boredom you refer to - though I must add that I've never found practicing boring.

I will divide a (longer) work into sections, A, B, C, D, etc. for as many sections as I think makes sense. Then I write the letters on separate slips of paper and, each day, randomly pick from the shuffled slips of paper and practice the section that is indicated on the slip. Each session I pick another slip from the pile until the slips have all been used up. Then, I re-shuffle the slips of paper and start over again. That way, I have a different section to work on each day, without going in printed order.

Caution: I make sure that I occasionally play through the sections in order (entire piece, perhaps) and I also practice the transitions from one section to another.

Regards,

edit: I see that Bruce posted a reply while I was typing, and I see that he already said essentially what I'm saying. But since I already typed it anyway.....


I 'sort of' do it, sometimes. I'd never heard of it or thought of it until I saw a friend do it -- a pianist who was visiting New York and did some of his practicing in my apartment. He didn't explain it; it was more like, I asked him if it was because otherwise we wind up playing the beginning too much and the end not enough, and he just nodded yes, and it made instant sense to me. While I don't exactly do it, I utilize this idea a lot; I make sure not to be deficient in how much I play and practice the ends of pieces, and I make extra sure that I know the ends very securely, since I know that if we don't give attention to that, we can easily be short-shrifting it.

As to any pitfalls, the only thing I can think of is that it could possibly get in the way of our having a good mental picture of the piece -- i.e. a good picture of the 'road map' of the piece. The way to guard against this is, well, I guess, just to guard against it.
Just make sure to also play through the piece 'the normal way' a great deal -- which we usually do anyway, more than enough. And, when we're learning and practicing the end, never lose sight of it being the end.
(I think this is essentially what Bruce was emphasizing, in the "Caution" part.)

BTW, what I usually work on first is whatever seems to be the hardest part of the piece. But then, I never wait very long before working on the ending. Usually it comes second, after the 'hardest' part. When I was learning the last really new piece that I worked on -- Scriabin's 10th Sonata -- I didn't have any hard decision, because the ending is (to me) the hardest part.

When I want to learn a piece, I sight-read through it and take note of the tricky sections - and learn them first. In the Chopin Scherzi (and Ballades), they will be the codas, and that's what I'll learn first. In fact, in pieces where there are easy sections, I don't even bother to practice them, but concentrate on the difficult parts.

I'm attracted by technical difficulties, and simple, slow music tends to bore me (hardly any of my memorized rep is slow), so practicing the difficult parts suits me just fine. If a piece (or section of a piece) is easy enough for me to sight-read easily, I don't bother to practice it, unless I'm wanting to commit the piece to memory. But practicing to memorize is different from practicing to master technical difficulties.

I think the "learning the piece backwards" has quite a few very big flaws:

1. If the idea is to avoid knowing the first part of the piece better than the last, then learning the piece backwards will clearly result in the reverse since the parts at the end will be played more. There is false logic to the approach.

2. Many people will start practicing at the beginning much of the time because they have the pleasure of playing through what they've learned so far. So not doing this takes away some of the pleasure of learning the music.

3. A reasonably serious and intelligent student will understand that the harder parts need the most work and proceed accordingly. This idea is not rocket science.

Your comment makes me want to share an approach I'm doing right now out of curiosity to see how it works: I don't want to memorize until I've got the piece "up to tempo." I always memorized so early when I was young because we had the "school year" to learn my repertoire. We figured that at a certain point it's just as easy to memorize it than to continue to play it with the music if it's in my fingers almost enough.

Right now I'm working on Chopin's G-flat major waltz, which my teacher wanted me to play in high school but I insisted on the F major (no idea). The right-hand is all over the place in the upper registers. I almost want to memorize it NOW so I can try to hit those G-flats and A-flats without fumbling.

What is your approach? I agree, I like fast pieces in my repertoire. I'm working on some slow-ish transcripted pieces from an anime I liked. The melodies are pretty and I think they're ripe for transposition practice. But I don't think that's memorization.

Do you have any thoughts?

Perhaps a poor choice of words, but also an indicator of some of my immaturity I'm trying to overcome. Mea culpa.

I love your idea of dividing the piece up into sections and working on a different one each day. I think I'm going to start that immediately with the Chopin waltz, I already broke up that piece into the sections anyway. The scherzo might have A-Z, lol, but I think it'll be nice too. And I'm glad to you know you work on the scherzi codas too at the beginning.

I definitely see the need for stringing it together with the proper transitions. I'm looking forward to this! Although I think I'm going to use an rng to decide my current part

Thank you!

What you and Bruce said I can't believe I hadn't thought of before: Take the hardest parts and start with those. It's so obvious that I missed it for so long. I'm excited to work on the coda for the Chopin C# Scherzo. It's so gorgeous.

The Op.70/1 is one of the few pieces where I memorize one hand (RH) fairly early on just so that it can be trotted out automatically - including those 1 - 5 leaps to hit the high notes - while I get the long LH leaps right. I look at the LH when playing the outer sections.

Go for it! Better than pages, divide it by sections. Then work a little harder on the sections which are giving you difficulty. A lot of students, particularly the younger ones, get upset by the idea of not beginning at the first page and going on from there. It's a mental block.

With the longer pieces, I've found it almost essential to divide the score up a bit otherwise, as you say, one finds the beginning is always a little "ahead" in terms of fluency. I did run across a pianist who carried this to an extreme where she was playing backwards, in tiny chunks of a bar or two. Now that is miserable. You lose a sense of the music and it becomes a tedious exercise. Make sure your portions are divided musically into sections which are somewhat complete in themselves. The Invention #9 is a lovely one with many places to divide the portions. I'd suggest bar 17 ... if you're dividing into two sections Or if four, then I'd choose bars 9, 17 and 24. But the fact that it's written in a "fugue" style gives you many options.

My rule for teaching ... whether it's piano or quantum theory ... is to make things easy. I'm outraged by the tedious blathering which accompanies so much teaching "techniques". Keep it simple and ENJOY your practice sessions.

lol, I'm always so all or nothing. Thanks for reminding me to be flexible. I'll go with our instincts here and start committing that to memory.

I finally made it through the entire piece front to end today. Then I went to listen to Rubinstein. Not a great idea, but inspirational.

Exactly the main thing to guard against!
(Maybe the only.)

Interestingly, this page talks about the chord-by-measure approach to backwards learning as positive: http://musicmotionblog.com/2009/07/the-fine-art-of-practicing-backwards/

I agree that would get tedious. I like this idea of dividing it into sections and moving around to give variety too. I may have to go jump on that Chopin coda!

There was a teacher I tuned for who apparently could play Beethoven's 32 Variations in reverse order. He knew that piece backwards and forwards!

I would agree that any section, practiced more than any other section, is going to result in that section being more thoroughly learned than any other. A little common sense is all that is needed, and anyone suggesting "learning backwards" exclusively, from end to beginning is surely not using common sense.

It is not, as you say, rocket science. Practice most what needs the most practice.

Regards,

.....but I would say there's a good argument that for most pieces, the learning is more secure if we sort-of learn them "backwards." It may not be necessary for most people to do it that way, and the lesser convenience may not be worth it, but.....I think there's a good argument that for most pieces the learning would be more secure.

In other words: that it's an approach worth considering.

So what is the argument for learning a piece backwards? Why will it be more secure?

I think I've already shown the obvious logical flaw in the argument that says this approach avoids over learning the beginning at the expense of the end. And this seems to be the only reason given so far for the backwards approach.

I'd say the answer is implied in Bruce's first sentence, and also dealt with in my first post.

The beginning is the part that we play the most anyway, even if we don't try. Probably even if we try not to. Plus it's generally the most familiar part, the part by which we most identify the piece in our minds, the part that we 'think of' whenever the piece pops into our minds. I think that except for someone who uses the "backwards" technique to an extreme and to a fault, the beginning is the part we know best anyway. The backwards approach stands to balance it out -- somewhat but probably not entirely, and I doubt there's much danger of reversing it.

One either uses the backwards technique or not. And if one does use it, then the last parts will be by definition over practiced compared to the beginning parts. Unless you can explain the flaw in this argument, then I don't see any logic to the backwards approach. It will not be true that "The beginning is the part that we play the most anyway, even if we don't try."

In addition, the BA assumes a student is not intelligent enough to know which parts require the most or additional practice. Those parts could be anywhere in the piece, including the beginning. I'm sure someone at your level or even a lower level would know if the end(or any part)of a piece needed more work and/or was under practiced.

I think it can be a good thing to vary the beginning and ending points of the sections you practice by a beat or a measure (or two or three) each time you practice, just to make sure you aren't practicing any sharp breaks.

There have been a few times when I've heard some pretty famous concert pianists give performances that told the listener exactly where they had broken a piece into sections for practice. There were discernible gear shifts at points between sections. Sometimes, I think it was because they had practiced one difficult part MUCH more than an adjacent easier part, and they always had started practice at exactly the same place. I felt embarrassed for them - most likely they didn't even realize it could be heard. Moving the start and stop points around just a bit when practicing might have prevented that sort of gaffe.