Piano World Home Page Forums Our Most Popular Forums Piano Teachers Forum I know how mistakes happen!

Really? They have to be within a certain vicinity perhaps, but a forest need to not replicate a specific pre-existing pattern. If you want to learn a piece, you have to put the notes in the right places. It's not about recreating any old forest but a specific forest. That's the product of individual trees. It simply cannot be created without attention to detail.








Sure. And if you do not know where any of the trees are you will not be able to play it. So both are equally important. Who says that looking at details means you're not going to be able to simultaneously realise what a minuet entails? Sorry, but we're not talking about a mutually exclusive situation. We're talking about a balance between the two elements.

However, it's notable that few pianists improve their accuracy by thinking more about the whole. Sorry, but nobody ever fixed or prevented an unexpected finger slip by consideration of large scale sonata form. Such things are fixed by isolating the details of how far you have to travel between a small series of notes. There is no end to the number of inaccurate, sloppy pianists who went on to improve their accuracy by starting to pay attention more to the individual components, instead of skimming across them with the bigger picture in mind (myself included). Far more inaccuracy is caused by simply not knowing the details, than by excessive concentration on them. When details are learned, you can forget them without losing them. When they have not been learned, you cannot afford to forget them or the whole suffers as greatly as anything from the inevitable technical failings. Any whole is a product of a series of individual details. No details, no whole. The only sensible way to work (for all but geniuses) is to concentrate on details, with their role as part of the whole in mind. It's a constant unity of conception. However, if we're talking about the way to achieve accuracy, the balance always needs to be geared slightly more towards the detail.

When a student persistently misses a particular interval within a piece that is otherwise secure, do you think that it's their grasp of the whole piece that is flawed? Or is it simply that they have never stopped to figure out how to connect the two notes and fingers that are causing the problem? In my own experience, such problems can frequently be fixed exceedingly efficiently by literally just repeating the single interval that is failing to connect. Accuracy comes above all from such details. The only time thinking about the whole prevents slips is when you have repetitive passages that lead off in different directions. However, even in those, it is extremely useful to isolate the exact moment where the music changes course. Detail always matters.

Your's is a very pragmatic view, Nyiregyhazy. Your focus is on piano as execution.

What else is one supposed to say after reading such a craptastic thread? The truth hurts, don't it?

You are totally mistaken. My focus on preparation is about execution. The very reason this is the case is because I wasted years thinking solely about the general impression I wished to create, without making terribly much improvement. I did scarcely anything that could be counted as 'practise' because I was only concerned with aiming for the bigger picture within that particular run-through (however sloppy it might be in the details), rather than with stopping to isolate the components that would fit together. I still think exactly the same way, in terms of how I intend to play. It's simply that I realised that the best whole is possible, when you have taken the time to prepare yourself first. That involves a wealth of different procedures.

When I'm actually performing, 'execution' could scarcely be further from my mind. Indeed, the better I have prepared the details, the easier it is to have a chance of forgetting about technical procedures and concentrate on the bigger picture (without disaster ensuing).

When somebody simply has little idea as to what they are doing, the more relaxed they are, the better they will perform. However, to follow up by suggesting that it's generally best not to worry about trying to understand what you're actually doing, because that will help you relax more, is going to result in a rapid brick wall. Sustainable progress only comes by understanding what you are doing more, not by aiming to stop thinking about that which you have yet to grasp. Those who feel 'pressured' when worrying about getting details right ultimately need to devote MORE practise to thinking actively (and above all going slowly enough to fulfill what they are capable of), if they ever want to break through the wall and start making real progress. Turning their brains off (or thinking about the whole rather than the details that they have not taken the time to understand or master) is futile. Detail really is everything, in the formative stages.

+1

Or we haven't yet defined the topic sufficiently operationally to be useful.

To dismiss it entirely I think overlooks a significant movement in performance history, including a major controversy.

That movement is based on the idea that much artistic expression is natural if it is not blocked, and the blockages are due to personality factors. Hence the key to performance mastery was some form of personal growth towards psychological healthiness. At one time anybody having technical difficulties was considered to be in need of psychoanalysis before he would make progress.

This theory is no longer in vogue, as far as I know, but the Inner Game of Tennis intuitive approaches may have grown out of it.

Within the theory I think belongs the controversy about whether art involves the freedom to express the inner individual, or the ability to convey the concept. As a simple example, suppose you are an actor required to portray a serial killer. That's easy if you are in fact an as yet undetected serial killer! however that happy state of affairs would be unlikely. Does the actor reach inside himself for the serial killer that lurks within? Or is he so professional an artist that he can portray an evil he does not himself feel? And, of course, if he can't do either it is because of one of those blockages curable by Freudian approaches.

Now that I've veered somewhat from mistakes, back to them.

The comment about a hesitation being better than a mistake horrified me. Sorry, I meant no offense, but to me a hesitation is far, far worse than a clumsily fingered wrong note. The hesitation allows the anxiety level to drop, but that's counterproductive when learning. The anxiety level shoots up as the time for the note approaches, generating more noise in the machine - which also generates the raw material for both creativity and technical improvement. And, significantly, produces the reward for correct play. The reduction of anxiety is enormously reinforcing, the technical term is negative reinforcement. Slow play reduces anxiety to unuseful levels. That's one of the reasons it is so comfortable and intuitive. We don't want anxiety so high it paralyzes us, but we need some to generate the noise.

I would point out also we've done no discussion of Skinner, et al, and the various stimulus chaining and reinforcement schedule approaches. Frankly I think they are more applicable to mistakes than psychoanalytic theorizing. Sorry kbk, I think that's the track you were on, but I think you've been led astray by assigning consciousness and motivation to a simple set of stimulus-response relationships. I think they are simple, just not visible.

Love the word "craptastic!" Will have to remember that one.

And if you do not have time either to be certain of which note you are about to play next or to sense the finger that is going to be playing it? Precisely how does launching into a random note simply for the sake of not stopping help either creativity or technical improvement? I can't see any way in which that could possibly make sense. If a student is working at learning a piece, personally I always stress that it's better to take time and get it right- after which you can go back and get the rhythm, once you know how to get to the correct notes. I really don't see how guessing is going to help in any fashion- UNLESS you are specifically practising the skills of playing a piece through at sight, rather than intending to learn it. Those students who start by going quickly with various wrong notes are rarely those who go on to achieve the most consistent performances. Any teacher can bear witness to that.

As is so often said, it's easier if you've never played anything inccorrectly in the first place. That's why slow practise is vital. It seems to fit perfectly with everything that is known about how the brain is able to reproduce movements that have been repeated. The point about going slow is that you make it possible to repeat things accurately. Launch in rapidly (particularly if you prefer to guess rather than take time to check uncertainties) and there is no realistic way of getting to a point of consistency. Habits are acquired through consistent reproduction of movement patterns. Not through making different mistakes each time you run through something that is not familiar at speed, simply for the sake of keeping the anxiety levels up. Your theory about anxiety sounds interesting, but do you have any sources that would demonstrate why removing the anxiety levels by practising slowly might be detrimental? I've never encountered such a concept before and frankly I'm intensely skeptical. Why should anxiety be desirable during the process of learning the notes? Quite how would that help the procedure of programming a series of movements into the brain? It's anxiety that results in the ugly 'stabbing' movements that students so often employ- when they panic and seize at the next note, rather than find a comfortable way of reaching it.

From my own experience, the slower I practise and the less anxiety I feel when practising, the better I play when I play at speed. That comes after years of only practising quickly, without bothering to stop and learn the details first. I really don't see how you can honestly suggest that going slow is bad. Going too slow rarely introduces problems (provided the student remebers to plan ahead in groups of notes, rather than merely think about one note at a time). Conversely, going faster than you are ready for ALWAYS introduces problems. It's clear which is more dangerous. Anything that isn't slow enough for the student to get things right (when starting off) is too fast. There are alternate ways that involve working on small units at quicker speeds that can also be useful, but the basic principle is simple- anything that involves guessing or imprecision is counterproductive to learning. It's not exactly hard to see why.

You have described two very different approaches to learning music, athletics, etc. These are the two standard approaches - the intuitive vs the analytical, the goal oriented vs the process oriented, the dissociator vs the hyperassociator, many ways to describe this.

You started with the wrong one - for you. And you eventually figured out the right one - for you. I hope you have not leapt to the conclusion that this is therefore the right one for everybody.

I strongly believe the learning approach to be neurologically hardwired, not within our control. We can use one or the other, but not choose. And I think that accounts for much of the frustration when a student fails to make progress using the approach that worked so very very well for his teacher.

The issue is why your idea for alternative might work and particularly how? I don't believe for a moment that it would work for anyone but a true genius. You ignored my questions about the 'anxiety' claims. What is your basis for those? And why is it better go completely wrong for the sake of not stopping? I'm really struggling to see your point.

Also, yes we can choose. I 'chose' to start making an effort to go slowly and accurately. I improved immeasuarbly as a result, compared to when I used to just spend all day sightreading pieces that I had never worked on the details for. You have to consider the reasons why such an approach offers benefits, before you start to consider whether alternatives might offer equal benefits. Generally they don't and for good reason. Going fast early on is dangerous because it increases both the likelihood of physical errors and mental errors. It's not just some old wives tale that it's better to practise slow. There's a wealth of evidence for it. Slow practise can't offer absolutely everything (I'm also extremely keen on practising small overlapping units at speed), but unless you have the talent of a Volodos, it's simply lunacy to expect to make optimal progress without it.

Specifically, no, I don't have references to music. Neither did I pull it out of thin air. Originally it came from classes in learning theory taken during an MS for Clinical Psychology - later I reversed direction and became a mechanical engineer. It also comes from discussions with jazz improviser about how they come up with novel riffs or licks.

Skepticism is good. From past discussions I suspect you sometimes overdo it and miss the possibility.





Hesitation is not just ugly to me, it is painful. This is a form of synthesthesia that makes it difficult for me to listen to or play with a group with rhythm problems. Intonation problems offend me, but they don't hurt physically. So I admit I am somewhat hypersensitive on this point. But hesitating rather than playing a wrong note teaches hesitation, WHICH IS A WRONG NOTE!. There is no more consistent bad habit in the beginner. Even the lay audience understands this. I have played for many church services, and my stumbles on the keys went largely unnoticed provided I kept the beat smoothly. In the advanced player, hesitation doesn't disappear, it morphs into unwanted rubato - and is oft praised as expression when it is not.

I am not recommending ugly stabbing movements, of course. We stop short of blind panic. (well, I've had the occasional one in performance!) The anxiety I'm talking about is not that extreme. It produces a forced choice response, there is no shortage of literature on that if you want to look it up. That increases randomization. Remember that a beginner is in a far different position from you. He has to learn movements that are vague and undescribable. To learn them he must produce them the first time, be rewarded, and increase the likelihood of producing them the next time. Much of good teaching is of subtle nonverbal reinforcement of the precursors to these motions. (and some of course is a skilled teachers direct teaching of correct motion) With no randomization, he produces the same motion repeatedly, never hitting the tiny invisible difference that makes it correct. As the anxiety increases, randomization increases, and the likelihood of both error and correct responses increase. But also, the reward for correct response increases as the anxiety reduces at a set time.

The identical process occurs with jazz improvisation. The anxiety produces randomization, noise in the machine. Then two filters are applied: is it good? is it novel?

There are other ways of producing more raw material, more randomization, than holding one to a strict time standard. Drugs (standard approach for jazz musicians of a certain period), sleep deprivation, actual mental illness, etc. Playing in time is one of the more healthy ones.

Popular press equates negative reinforcement with punishment. That's wrong. Reinforcement is anything that increases the likelihood of a future behavior. Negative reinforcement does so by reducing a stress, positive reinforcement does so by increasing a reward.

No, it's cr*p.

Slow practise works very very well when the skill is available but the piece needs to be learned.

It is worthless for learning a new skill.

There's a wealth of evidence for the speed walls produced. And if you raise the question, somebody with 20 years of experience will immediately sit down and show you how he can learn a piece with slow incremental practice.

Try the same thing with a beginner. Different story.

"Specifically, no, I don't have references to music. Neither did I pull it out of thin air. Originally it came from classes in learning theory taken during an MS for Clinical Psychology - later I reversed direction and became a mechanical engineer. It also comes from discussions with jazz improviser about how they come up with novel riffs or licks."

Totally different issue though. Improvisers must constantly adapt and respond. Pianists must learn how to play the same notes in the correct order, whatever should occur. A burst of adrenaline could help a jazz pianist (and indeed a classical one, in performance, interpretively speaking). But how does this benefit someone who has yet to learn how to play a series of notes in the right order? Sorry, but I'm not seeing it.


"Hesitation is not just ugly to me, it is painful. This is a form of synthesthesia that makes it difficult for me to listen to or play with a group with rhythm problems. Intonation problems offend me, but they don't hurt physically. So I admit I am somewhat hypersensitive on this point. But hesitating rather than playing a wrong note teaches hesitation, WHICH IS A WRONG NOTE!. There is no more consistent bad habit in the beginner. Even the lay audience understands this. I have played for many church services, and my stumbles on the keys went largely unnoticed provided I kept the beat smoothly. In the advanced player, hesitation doesn't disappear, it morphs into unwanted rubato - and is oft praised as expression when it is not."

What is worse? Experiencing the feeling of progressing from one finger to a random one, unprepared one (typically accompanied by an awkward seizing movement, as the student panics- something that happens when having to press on unprepared, whether recommended by a teacher or not)? Or stopping, feeling the correct connection between two correct fingers and then going back to put that connection into a rhythmic context? Stopping should never be applauded. The student ought to go slowly enough not to need to stop, until they know the passage well enough to go faster with equal comfort. However, it leaves far less to fix compared to screwing up- and then having to go back without yet having had ANY experience at the physical connection.

The ability to fake for performance is important but that should never be confused with the art of learning pieces. Sight reading is one skill. Learning to play a piece to a high standard is another. Are you trying to say that nobody should do anything in practise that they wouldn't do in a performance? The two really don't overlap that much. The ability to avoid stopping when you screw up should be part of any pianists range of skills. However, if that's the primary goal in both practise and rehearsal, just don't expect to play Chopin studies to any standard. There's no room for faking in those.

"Remember that a beginner is in a far different position from you. He has to learn movements that are vague and undescribable. To learn them he must produce them the first time, be rewarded, and increase the likelihood of producing them the next time."

Sure. It's for this reason that I advocate lifting the fingers a little before each note, for beginners. It helps to have already sensed the finger. Another way of preventing mistakes before they occur. The last thing you want is a rash movement. The more you sense what is happening, the sooner you get to the stage when you can make instantaneous movements with absolute certainty. Even beginners don't need to learn from errors. If they begin by lifting the fingers a fraction, they soon get to the point of having full sensory awareness without needing to move the finger first. Randomly stabbing at a finger and seeing if it turns out to be the right one is simply not going to compete with learning what it feels like to prepare the finger.

"As the anxiety increases, randomization increases, and the likelihood of both error and correct responses increase. But also, the reward for correct response increases as the anxiety reduces at a set time."

And the penalties for a randomly struck note become all the more undesirable. If the student hasn't had time to think, the odds aren't good for them. Also, I believe it's well established that it takes far more correct executions to 'cancel out' the memory of just a single incorrect one.

"The identical process occurs with jazz improvisation. The anxiety produces randomization, noise in the machine. Then two filters are applied: is it good? is it novel?"

If you're looking to make new material, that's great. If not, why on earth should randomization be desirable? What could be less welcome?

Contrary to the obvious fact that it's harder to be consistent and accurate when starting fast and the wealth of evidence about how repeated movements can be more easily reproduced, it's 'crap'? Yet, you gladly take the word of some jazz pianist who once told you that anxiety should always be present in practise?

So you don't think that the pianist who has been playing for 20 year might possibly have learned something during those twenty years? That couldn't be the issue?

Tell me, are you suggesting that the beginner should sit down and play this piece fast in the first place? Perhaps that would be better than them starting slow and having to take time before they have learned the control required to go faster? Where exactly are you going with this comparison? You're criticsing slow practise, but you're not offering any credible alternatives. Accuracy doesn't come by magic, if a beginner starts by playing fast. It becomes all the more elusive. Sure, a beginner doesn't become a concert-pianist in a month simply by practising slowly. But are you honestly harboring the delusion that if they started by practising quickly, they might do better? Why? On the opinion of some unnamed jazz pianist you once chatted to?

I'd really like to see you put this radical new theory into practise by teaching a number of youngsters to play the piano at high speeds (on the grounds that slow practise is worthless). I wouldn't urge you to hope for terribly much from the results. Slow practise breeds comfort. Comfort is what permits fast playing.

What I am trying to explain is that the process of coming up with a novel jazz improvisation and the process of learning technique are precisely the same.

Both are trial and error processes.

So both require error. And error comes from randomization.

I'm not talking about playing a wrong note as error. I'm talking about the complex, almost mathematically undescribably geometric complex motion of getting the finger to the right note at speed. I'm talking about the entire sequence of weight shift, shoulder motion, elbow motion, wrist motion, etc.

Some of this is known by good teachers and they can simply instruct how to do it. Just like all of the notes are on the page - there is no mystery what to play next.

But much of it is a mystery until experienced. And unless done perfectly the first time, it doesn't happen without error. You must do it wrong many times to do it right. Then you must recognize right, and be rewarded sufficiently to do it more often.

Slow does not have to be unconnected to time, that is just an unhappy accident of the way we teach piano. Slow practice with a metronome could conceivably convey some of the benefits of practice at tempo. But slow practice without strict time is worse than useless, and beginners have flexible time.

The process of learning a technique unconnected to time, then relearning it at slow tempo, then relearning it at fast tempo, is simply WRONG.

Guitar players, even beginners, learn at tempo by playing along with records, the radio, CDs, and their peers. This is the reason they pretty much all succeed.

Both are trial and error processes.

So both require error. And error comes from randomization.

I'm not talking about playing a wrong note as error. I'm talking about the complex, almost mathematically undescribably geometric complex motion of getting the finger to the right note at speed. I'm talking about the entire sequence of weight shift, shoulder motion, elbow motion, wrist motion, etc.

You may not be talking about it, but if you're not practising slowly, it's happening vastly more than with someone who begins at a comfortable pace. Guaranteed. And if you're not even taking the time to ensure that you play the notes correctly (nevermind to balance in between them), how much value do think can be found within the feedback your arms will receive? Any?

In any case, no trial and error process 'requires error'. That's grossly false logic. To arrive at a correct answer merely requires having reached it. Having got it wrong a few times doesn't change the final result (at least not positively). To claim that you have to screw something up before you can do it properly is totally missing the point. And, for that matter, learning the notes for a piece is not supposed to be viewed as a 'trial and error' process. That's simply sloppy, lazy thinking. Rapid learners don't 'have a go'. They take the time to ensure they have read every note properly and then seek to play them properly. They don't have a few guesses. You're really wildly off the mark here. If you're not able to get something right first time (without resorting to 'trial and error'), it shows just how flawed your method is. There is absolutely no place for thoughtless errors, if you're hoping for serious progress.


Some of this is known by good teachers and they can simply instruct how to do it. Just like all of the notes are on the page - there is no mystery what to play next.

Unless, of course, you both insist on going quickly at once and never taking any time between notes. In which case there frequently is a complete mystery as to what to play next and you hence go wrong.

But much of it is a mystery until experienced. And unless done perfectly the first time, it doesn't happen without error. You must do it wrong many times to do it right. Then you must recognize right, and be rewarded sufficiently to do it more often.

This is fundamentally at odds with all science about repetive movements. Out of interest, what level do you play to? Are you playing Chopin Studies or Bach fugues with this approach? If you're playing things wrong many times, I sincerely doubt that you'll ever even approach getting such difficulties right. I'm sorry, but I really can't believe that you're defending this approach. I'd like to hear of ANY pianist who plays advanced classical repetoire to a high level by going wrong repeatedly, rather than by ensuring that things are done comfortably and correctly. Recognize right? Have you never consider having the patience to read the score properly and take the time to do it right the first place? You seriously think this would be more harmful than pissing around playing something completely wrong, time after time? Is it me you're trying to persuade of this ridiculous theory or yourself? How much time do you waste playing things incorrectly exactly? You really don't think it might be quicker to get it right the first time?

The process of learning a technique unconnected to time, then relearning it at slow tempo, then relearning it at fast tempo, is simply WRONG.

Who said anything about being uncconnected to time? Looks like the strawman is being pulled out. Pause once before playing a note and you can go back and fix it. Screw up and you have far more work on your hands. It's better to pause than to go wrong- provided that you go back and reinstate the rhythm straight after.

1) There is a vast amount of literature out there on the non-conscious (or un, pre, but not necessarily sub. Prenoetic is a good one). The problem is that it's very much multi-disciplined and dense. What goes on in the mind will always be idle speculation for those who only look at the surface. Some of the posters here have obviously gone deeper into it and I'm valuing their contribution.

2) That cognition is going on away from consciousness is quite a revelation to many, nonsensical to most but quite revealing to a few (the cognitive therapy folks can't abide the idea). We lived under Decartes' shadow for too long. I for one, am observing a different 'set up'. So yes, for the vast majority this topic is 'as pertinent as...' but I would like to keep on topic. We can and do have threads on how to prevent mistakes all the time.

As for posting in the Teachers Forum, you're welcome to (its those that post here just to slag off teachers that are annoying).

That's a good observation and why I think teachers should always be looking 'under the hood', both theirs and their students'. All the time, and I'm sure it's non-conscious, that I'm teaching my mind is saying 'Why did he/she get this?' 'Why didn't he/she get that?' - a constant 100%-of-the-time probing. I think that's what sets us apart from other primates.

Your twitty comments hurt? That's a laugher!

The inner game books have come up as they often tend to with approaches to the non-conscious. My quibble there is that they give equal weight and treatment to both self one and self two and always, correct me if I'm wrong, on a cognitive level. I think the cognitive differences are of kind rather than degree or identity.

For those who are frustrated at the lack of self-help in this topic, look at it this way. If you're trying to get across a river sometimes it's worth climbing a tree. Though climbing is not exactly an activity you'd think of as relevant to crossing a river, who knows what you'll see from there?