practicing sight reading

[nyiregyhazi]

Not quite. My second paragraph was to demonstrate that the first and last positions do in fact have a special role.  

The paragraph in the article serves a different purpose. It shows that there are times when it doesn't work. That in some instances, the word remains ambiguous, and the words chosen by them are actually carefully chosen to demonstrate this. A random paragraph converted to this rule will in most cases be legible.

Now, I did not use this as an example of what happens when we sight read. I used it as an example of where the way we actually process information is not as straightforward as might appear at first glance.  In cognitive sciences, the oddities - things that fail (such as optical illusions) and things which surprisingly work (such as the word example) are often useful tools in understanding what's going on under the hood.

Applying that latter point to sight reading might prove a useful starting point. The things we cannot read, or the points where we make reading errors may prove a useful insight into the underlying mechanism. And, being a complex task, the underlying mechanism may work differently at different stages of its development and possibly differently in different people anyway.

In my experience - subjective entirely, here - there are things I can read easily, where I am completely oblivious to what's going on, and then a range of various elements that, when introduced,  increasingly make the task more complex, slower and more conscious.

It also seems to me that part of the sight reading process also involves the brain in forming an auditory expectation, and a mechanical strategy for execution. I suspect that the better the sight reader, the more intertwined all of that is, and that also serves to complicate the picture.

I referred to the end of the article not the end of your paragraph. It shows the supposed formula is wrong.

Sorry, but it's complex isn't good enough against probabilities. You objected in response to a post where I merely said the brain must process the actual notes to get them right. I didn't even say the conscious brain and I'm baffled by your objection to this evident fact, whatever your stance. The only version in which the player somehow gets all individual notes right without the brain processing them all is when a very simple formula is stuck to so rigidly nothing ever surprises even a little or if the player is insanely lucky. When notes are not predictable at all, it's 100 percent sure that they must all be processed for accuracy. Saying "It's complex" doesn't dent that. Under the hood, everyone who is highly accurate is processing details.

Also, you can't bring something to the table and then try to shut down debate on the issues where it is most relevant, just because you only wanted to show that stuff can be complex. Whatever your purpose, the word example you brought is a very easy thing to decipher due to limitations of language. The examples in the article better illustrate complexity in a situation with more possibilities and also reveal the sheer importance of detail. Any piece of moderate complexity has vastly more permutations than even the example at the end of the article. Even if words are read as a whole (which in no way suggests failure of the brain to process details, given that I reliably see small spelling errors) it's easy to build a repertoire of most words and remember them next time. It's not easy or even possible to build a comprehensive repertoire of the wealth of possible musical building blocks. Even if a piece has nothing but tonics and dominants, there are so many different spellings that memory could not help ensure that I get every chord accurate according to every note unless I read in proper detail. Even in such limited circumstances, nothing but mental processing of details results in anything but a very low probability of the actual notated detail being played. There are too many possibilities in even a guaranteed tonic and good readers are as capable of reading new chords as they are of doing very standard or very non standard arrangements of c major. It's chiefly mental processing of detail. Even if memory were a notable issue, it simply pretty useless for the vast majority of what is actually possible. Processing detail into a visualisation of what to do is what yields reliable results. Not hoping to have such a wide repertoire that any combination that comes up will already be familiar. It's like hoping to have already played every game of chess. Good reading is chiefly from processing in the moment, not from having prior familiarity with unique situations or from hoping everything is so badly formulaic that you've seen it and done it already.

[j_menz]

I referred to the end of the article not the end of your paragraph. It shows the supposed formula is wrong.

My second paragraph, quoted above addresses that. It appears you missed it.

Sorry, but it's complex isn't good enough against probabilities.

I'm missing entirely your continued reference to probabilities.

My "it's complex" is a response to your "it's this and only this".

I am aware you are talking about the unconscious level of processing. So am I.

We both, I think, agree that good sight readers reliably and consistently get notes right, and that it is a reliable and fast way of processing the information off the page that allows them to do that. Not a "luckier then most" approach, but one that consistently works.

I would also suggest that for even the best sight readers, there are occasional errors. I would also suggest that errors are informative.

Indeed, sometimes it would be an error to play as written as the score itself is in error. I would suggest that successful corrections of these may also be informative. Further, hypercorrections would provide an additional level of insight.

I am happy to pursue a discussion on any of these avenues, but frankly am not interested in doing so if you propose to wilfully misunderstand what I say and to merely restate your "I think this therefore it must be true" line.

[nyiregyhazi]

My second paragraph, quoted above addresses that. It appears you missed it.

I'm missing entirely your continued reference to probabilities.

My "it's complex" is a response to your "it's this and only this".

I am aware you are talking about the unconscious level of processing. So am I.

We both, I think, agree that good sight readers reliably and consistently get notes right, and that it is a reliable and fast way of processing the information off the page that allows them to do that. Not a "luckier then most" approach, but one that consistently works.

I would also suggest that for even the best sight readers, there are occasional errors. I would also suggest that errors are informative.

Indeed, sometimes it would be an error to play as written as the score itself is in error. I would suggest that successful corrections of these may also be informative. Further, hypercorrections would provide an additional level of insight.

I am happy to pursue a discussion on any of these avenues, but frankly am not interested in doing so if you propose to wilfully misunderstand what I say and to merely restate your "I think this therefore it must be true" line.

All I have ever stated is that you cannot accurately determine a whole except by processing all of its components in full - because if one of those components should diverge from expectation then it is not the whole it was mistaken for. And if one and only one note should diverge, that too can only be known for certainty if every single one of the other notes is processed accurately, so as to reveal for certain that it was only that one note that diverged and that no others did. In any scenario involving accuracy rather than sketching out the general harmonic impression, sightreading is based on thoroughly processing all the notes. This should really not be controversial within a serious discussion. From this necessity, it's clear that it's a case of the brain creating wholes from detail and not one of the brain deriving details from instant wholes in sightreading. Only when detail is being processed is the brain operating under an accurate knowledge of bigger chunks, rather than under speculation. Good sight readers only speculate as an emergency resort, not as part of their normal foundations.

PS regarding mistakes, if we used an internal autocorrect in the same way as we do with language we'd autocorrect a wealth of "mistakes" that are not mistakes. Once again, this example shows how much detail we must process. Once every now and then a mistake in the score really is a mistake. But the vast majority of divergences from expectation are not. We have to be overwhelmingly sure about actually reading details to make sure we get these details right and do not allow false expectation to stick to a forum that fails to apply. Accurate internal corrections are based on processing so much detail that we can then use a broad context to realise that the score really is wrong. It's nothing like being able to read through spelling mistakes or reordering those letters say - as the brain has to go far deeper to either confirm or deny whether the unexpected note is meant to be just that. Things which make reading language quicker would simply generate errors in music, were there not more attention to detail. It's small wonder that a study recently showed that musicians are better at spotting mistakes (outside of music).

[timothy42b]

j_menz,
I'd be interested to know if you think this recent experience is relevant.  I apologize in advance for the length, which regrettably may approach one of n's shorter posts.

I don't sightread well on piano because normally one can only sightread a couple levels below what you can play, and my playing level isn't high enough!  <smiley>

I play trombone in several community level ensembles, and I am consistently one of the better sightreaders.  I also sing in a church choir.  We hire a couple of section leaders, music majors from the local university, and they have much better voices than I but I always out read them.  I mention this because these experiences are the basis for my theories, and I freely admit my lesser ability at piano means they may not apply 100%. 

At any rate, one of my strategies for improvement has been to note where I make an error or struggle, and work specifically on that.  On trombone, it was big band rhythmic figures, something not as common in most of the music we play.  Heavy syncopation, rapid tempos, swung eights, conventions for notating.  So I spent time daily working out of a syncopation book until I could play them correctly by feel, rather than counting.  And this improvement transferred immediately to ensemble. 

Until this week, when I sightread a jazz piece in rehearsal.  The band had worked on it before but I had not seen it.  The arranger knew what he wanted to hear but not how to write it.  (For example, by convention an eighth note is legato and a quarter staccato in most patterns;  he wrote a quarter note anywhere he wanted a note long, and an eighth note eighth rest everywhere he wanted a note short.  Think of the difference between seeing 4 quarters in a measure or 4 eighth note eighth rests where and this is key you don't expect it.)

I could not absorb these notes and play them at sight, though they were well within my comfortable technique AND reading level, in fact easy.  Instead I had to focus on the details n talks about them and actually prima vista sightread them.  It was 10 times as hard.  I can do it, I'm good at counting and reading, but there was an extra layer of mental processing going on that does not exist if you slap In the Mood or String of Pearls (or a Sousa march) on my stand, and that layer interferes somewhat with being stylistically correct and leading the section. 

[nyiregyhazi]

j_menz,
I'd be interested to know if you think this recent experience is relevant.  I apologize in advance for the length, which regrettably may approach one of n's shorter posts.

I don't sightread well on piano because normally one can only sightread a couple levels below what you can play, and my playing level isn't high enough!  <smiley>

I play trombone in several community level ensembles, and I am consistently one of the better sightreaders.  I also sing in a church choir.  We hire a couple of section leaders, music majors from the local university, and they have much better voices than I but I always out read them.  I mention this because these experiences are the basis for my theories, and I freely admit my lesser ability at piano means they may not apply 100%.  

At any rate, one of my strategies for improvement has been to note where I make an error or struggle, and work specifically on that.  On trombone, it was big band rhythmic figures, something not as common in most of the music we play.  Heavy syncopation, rapid tempos, swung eights, conventions for notating.  So I spent time daily working out of a syncopation book until I could play them correctly by feel, rather than counting.  And this improvement transferred immediately to ensemble.  

Until this week, when I sightread a jazz piece in rehearsal.  The band had worked on it before but I had not seen it.  The arranger knew what he wanted to hear but not how to write it.  (For example, by convention an eighth note is legato and a quarter staccato in most patterns;  he wrote a quarter note anywhere he wanted a note long, and an eighth note eighth rest everywhere he wanted a note short.  Think of the difference between seeing 4 quarters in a measure or 4 eighth note eighth rests where and this is key you don't expect it.)

I'm a little surprised that this would throw you. You simply need to think in beat locations and picture every beat in the bar before starting it. The first thing I'd do in that scenario is notice four quavers, which would almost certainly be spread out equally through the bar, visually speaking. Instantly, that tells me I'm just lining four equal notes to four standard beats. Seeing the rests is a confirmation of the fact that the bar adds up correctly in details, not an integral part of how I judge how four visually similar notes must match four beats. In my mind, it's little different to seeing four tenuto crotchets.

I spoke of how utterly implausible it is to accurately read off notes if you depend too much on expectations or memories rather than actually read all the notes in their own right. However, in rhythm, the right mental organisation can make all kinds of logical deductions that simplify the process of aligning notes to beats with certainty. Unlike with note reading, one piece of information can give 100 percent certainty about another detail, regardless of whether you read that detail directly. I should add that, although the situation you describe wouldn't throw me at all, the notation used by singers (where notes are often beamed separately rather than together, in a way that would have clarified which notes fall on a beat) confuses the hell out of me. In that situation, there is less scope to deduce beat locations via logic or common sense.  You start with fewer instant visual references for beats. It takes a more conscious and more mathematical mental process to zone in on exactly which notes align directly to beats of the bar.

[timothy42b]

 I should add that, although the situation you describe wouldn't throw me at all, the notation used by singers (where notes are often beamed separately rather than together, in a way that would have clarified which notes fall on a beat) confuses the hell out of me.

I teach handbells, and the notation does not include all the rests. There will be a stacked column of notes of different values.   In every new piece there are a few places where my ringers do not understand whether the note is on the beat or somewhere else.  I tell them what to play and they scribble it onto the music.

[timothy42b]

You simply need to think in beat locations and picture every beat in the bar before starting it.

No.  In this type of music, before ending it. 

[nyiregyhazi]

No.  In this type of music, before ending it. 

? Why? If you're reading ahead in a single line melody, there's no reason not to have perceived that whole bar in advance before going in- no matter how quick. This is what good pianists aim for, so there's really no excuse when there are just four notes to process in advance.

[nyiregyhazi]

I teach handbells, and the notation does not include all the rests. There will be a stacked column of notes of different values.   In every new piece there are a few places where my ringers do not understand whether the note is on the beat or somewhere else.  I tell them what to play and they scribble it onto the music.

I'm not sure what you mean exactly, but the overview of beats should be instantly visible with or without rests, except when there are unbeamed quavers and semi quavers. Is it that the case here? The only other exception would be with ongoing syncopations. But for the most part, the great thing about notational conventions is quite how easy they make it to pinpoint moments where beats land, without having to add up every note value for the start onwards, in order to be sure. In rhythm you can often use a combination of memory and logic to point to beat locations, without always needing more.

[timothy42b]

I'm not sure what you mean exactly, but the overview of beats should be instantly visible with or without rests,

Believe me, it is not.  It is often the case that my ringers can't tell what is intended and sometimes I have to puzzle over it a bit myself.  You'd have to look at some handbell music to see why.  It's a compromise system, designed for people who are only playing their two notes out of a dense cluster of notes and for the most part are not fluent readers.  It is not uncommon to have a dotted half, 4 quarter notes, and a couple of eighth notes stacked on one beat, with some of them offset if the note heads are adjacent, and then the following notes may be beamed to show the melody line, and it's a real hash. 

[nyiregyhazi]

Believe me, it is not.  It is often the case that my ringers can't tell what is intended and sometimes I have to puzzle over it a bit myself.  You'd have to look at some handbell music to see why.  It's a compromise system, designed for people who are only playing their two notes out of a dense cluster of notes and for the most part are not fluent readers.  It is not uncommon to have a dotted half, 4 quarter notes, and a couple of eighth notes stacked on one beat, with some of them offset if the note heads are adjacent, and then the following notes may be beamed to show the melody line, and it's a real hash.  

I don't mean that it automatically will be. I mean that with practise it should be. If there are a lot of rests, one of the most important things is to zone in on rests which land on beats and feel an imaginary accent on those. Every beat location needs to be clearly organised first and then any syncopations or odd notes can simply be precisely aligned in between clear references. Also, much of the technique for using logic is based on working backward from the end of the bar as much as reading forwards from the beginning. This often reduces the amount of processing required. For example if the bar ends in two quavers then you should instantly know that the first of them is the fourth beat. That might also be more useful at narrowing down where the previous beat falls than counting lots of notes from the beginning. It's very important to get the overview of a whole bar before trying to execute it. Unless the notation in score is plain wrong, it shouldn't take too much work to get used to pinpointing the associations to the beats, if logic is used effectively in the process- rather than just maths.

Fugues work the same way- where you don't count every long note, so much as simply hold on to it until association to beats that are marked by other voices tell you that the next note of the voice is due. Vertical locations of beats must be understood first, for reading rhythm to become natural.

[timothy42b]

For example if the bar ends in two quavers then you should instantly know that the first of them is the fourth beat.

In jazz, the second quaver is effectively the 1st beat of the next measure. 

[nyiregyhazi]

In jazz, the second quaver is effectively the 1st beat of the next measure.  

? Regardless of the musical association to it, it still falls before it in time and the two quavers still tell you precisely where the 4th beat lands- even if you can view the last quaver as a kind of anticipation to the next beat. What I'm saying doesn't contradict that at all. Jazzers still feel where the beats actually fall, no matter how much they might anticipate or delay the notes around the them. In fact, this is the crux of what I'm talking about. You must always know where the beats are, no matter how syncopated the notes you play might be. The square old beats still organise even the funkiest off-beat swing rhythms. It's when people are so lost in when they play that that they lose awareness of the location of the beat itself, that there are problems. At that point, nothing is truly organised around the pulse.

[j_menz]

j_menz,
I'd be interested to know if you think this recent experience is relevant. 

Indeed it is. I'm often surprised that some things seem much harder to read than they should be, and some others work out quite a bit easier.

Sometimes, that's a frame of mind as well. I recently bought a book of "concert transcriptions" of hymns. I completely butchered the first one until I twigged that the arranger was a pop type composer. Then it all fell into place.

[bernadette60614]

From a student's perspective, each of my teachers has told me my sight reading is excellent.

I do 15 minutes of sight reading a day at the end of my focused practice.  I chose something that is in contrast to what I'm currently working on...so Beethoven for study, jazz for sight reading, Mozart for study, the blues for sightreading.

Practicing and learning is serious work, but I find that I learn more readily if there's a element of fun and surprise.

[kevin69]

I think its a bit simplistic to talk about processing the notes when sightreading without specifying what sort of processing you are talking about, because it seems to me that there are several different types of processing going on.

There is analysis of individual notes: i expect that this is generally relatively slow and will feel 'hard' even when it is reliable.

But i think that for easy, familiar, styles there is also pattern recognition (based on memory) followed up by verification. This still involves processing every note, but verifying a patterm is as expected will be much faster than a full analysis.

In between, the (fast) verification process will quickly show up which notes are surprising (in that they don't match the patter). These notes can then (slowly) be analysed in full. In this way memory and pattern recognition can resolve many notes quickly and then act as a guide for where we need to perform a detailed, slow, analysis.

At least, this is how it feels like to me, and it offers an explanation of why familiar patterns of music can be  sightread more easily than unfamiliar ones: its the number of relevant patterns that we have available to match against that matters for fluid reading because we want to minimise the amount of analysis we need to do.

But you do need to deal with every note, just not delaing with them in the same way.

[nyiregyhazi]

I think its a bit simplistic to talk about processing the notes when sightreading without specifying what sort of processing you are talking about, because it seems to me that there are several different types of processing going on.

There is analysis of individual notes: i expect that this is generally relatively slow and will feel 'hard' even when it is reliable.

But i think that for easy, familiar, styles there is also pattern recognition (based on memory) followed up by verification. This still involves processing every note, but verifying a patterm is as expected will be much faster than a full analysis.

In between, the (fast) verification process will quickly show up which notes are surprising (in that they don't match the patter). These notes can then (slowly) be analysed in full. In this way memory and pattern recognition can resolve many notes quickly and then act as a guide for where we need to perform a detailed, slow, analysis.

At least, this is how it feels like to me, and it offers an explanation of why familiar patterns of music can be  sightread more easily than unfamiliar ones: its the number of relevant patterns that we have available to match against that matters for fluid reading because we want to minimise the amount of analysis we need to do.

But you do need to deal with every note, just not delaing with them in the same way.

That's how it feels to me too, but I believe it's subjective illusion. A jazz musician will be familiar with all kinds of dense harmonies. But will they identify them as quickly as a straight c major by sight merely because they are familiar? Even if they read a lot of fully notated jazz I doubt it, because they have to process more detail in order to accurately compare it memory banks. Nobody sees a dense accidental riddled chord as an instant whole. There are so many possibilities in chromatic harmony that any possible note could be subtly different. I could have played a particular chromatic chord 1000x over but it wouldn't enable me to spot it another atonal work any more easily- because I'd need to process every bit as much information as with every other complex chord to reference it to that memory. I don't think it would be easier to read due to familiarity if just one note was different, either. If anything, I'd be at greater risk of error than with a unique complex chord, due to the desire of the brain to make things fit existing patterns. Unless we are detailed enough in our processing to appreciate that they don't always fit to expectation of memory, we make false assumptions all the time.

Of course, good readers process dense chromatic harmonies better than bad ones. But that's where logic tells us that what they have is better processing power. The only place where memory helps "reading" significantly is in knowing how to execute patterns that have been done before. I do suspect to a degree that the brain is quicker when it has some kind of expectation to confirm or deny than when expectation is non existent. But I think the real issue in difficult music is primarily processing. Look at any standard major or minor chord and there are only thirds and a fourth so that's very easy to process. Only when they are written unusually do they seem difficult. If c major had an a double flat rather than g, it's not about lack of memory but about having more information to process, which would make it a fraction harder to do quickly. On the same principle a chromatic chord involves much more processing of different intervals (rather stacks of thirds) and adjustments to black keys. I only see memory as being an obvious issue after the reading already happened, in terms of whether you remember being there before or have to figure out from scratch how your hand gets to the notes visualised. With the way good readers read, even if they didn't have memory of c major, it would be spectacularly easy due to to the simplicity of processing straight thirds plus the odd fourth. A bad reader might have read c major triads 1000 times over yet still be crap at processing and mentally organising intervals and thus fail to get it in an instant.

You could draw parallels in language. Weird combinations of letters are harder because they don't spell an obvious sound and thus leave the brain getting confused when it tries to integrate them into something bigger. But I could also make a fictional word and read it effortlessly, if there is logic in phonetics. Plab is not a word but it's no harder to read than slab. I don't buy the memory thing at all, because there is a big flaw in the argument about familiar vs unfamiliar. Examples of the familiar are almost unfailing simple to process and examples of the unfamiliar are almost unfailingly difficult to process. If you use the example of dense jazz harmonies, I think it actually turns out that complex familiar chords are just as hard to read with precision- for the simple reason that they are more complex to process from the page.

[dima_76557]

@ nyiregyhazi

Not to argue with you (I think I lack the skills for that ), but since you mentioned Bach's fugues, I'd like to mention that after practising (and transposing to all possible keys!) Cortot's preparatory exercises for polyphonic technique from his "Rational Principles", Bach's works became a lot easier to process (that is: sightread and execute right away). I hope this makes sense, but I have a feeling that good sightreaders don't sightread with their eyes alone.

[nyiregyhazi]

@ nyiregyhazi

Not to argue with you (I think I lack the skills for that ), but since you mentioned Bach's fugues, I'd like to mention that after practising (and transposing to all possible keys!) Cortot's preparatory exercises for polyphonic technique from his "Rational Principles" in all keys, reading Bach became a lot easier to process (that is: sightread and execute right away). I hope this makes sense, but I have a feeling that good sightreading pianists don't sightread with their eyes alone.

I'll have to look into those. Funnily enough, I had just that kind of idea while I was playing fugues the other day. There's a very big physical element in the ability to hold one long note clearly and precisely while simultaneously passing from one finger to another. I was thinking to myself how futile it would be to attempt these these techniques in complex music if you haven't already got down the basics of this via simple exercises. I can execute the five part fugues with proper voice leading, but i do still notice that a slight wobbly or stiff quality on held notes makes it a fraction less simple than it could be.

 Anyway, I agree entirely with the importance about having memories and experiences of these basic building blocks. My point is about quite how little what goes in through the eyes relates to memory and quite how much it relates to processing skill. A fugue cannot be sightread at all without prior technical skills, but even with those skills in place it's one of the ultimate illustrations of how much is down to speed of processing, rather than need for things to match to memories on the page. Even those with excellent physical habits won't do well in fugues without ultra evolved processing.

[lostinidlewonder]

Having learned more than half of the first volume of the wtc, unfamiliar fugues continue to be what I would be least confident at sightreading. It's nothing to do with familiarity.

Maybe you are not familiar enough that is why they continue to be least confident with. If you are familiar with them this means you have had a lot of experience playing counterpoint and patterns you face when sight reading will be easily executed and inferred, you will understand what you hear and anticipate the changes, the reading will be more easy and fluent, the fingerings are well known.

Put an expert in Bach into a totally different style they will most likely fall flat on their face. Try to sight read some Kasputin or perhaps with other rhythmic music which Bach does not explore. It will be terribly difficult. But put a stride pianist in front of some Kasputin and much of what is read might be easily understood because of their experience base.

Familiarity has everything to do with good sight reading.

The idea that memory is the big issue in reading difficult repertoire is just a load of old guff.

Your conscious, muscular and sound memory helps you to sight read a huge amount. Sight reading courses I give require students to read through hundreds and hundreds of works. The more experience you have with pieces and the general movements at the piano the more this will help you ability to sight read these commands and produce the correct fingerings efficiently.

I have no problem with any style, but only with levels of difficulty.

What about a style which is full of procedure you are unfamiliar with? Saying you have no problems with ANY STYLE is rather over confident, and ridiculous.  

Memory of other pieces doesn't make works of great difficulty any easier in anything but a slender handful of respects....

Whatever "slender handful of respects" you mean you have obviously skipped over with no explanation. If I have memorized one thousand pieces you think this will not help reading difficult works? Works become less difficult to play the more known procedure you read from them. If you have a large experience base of a lot of technique and fingering at the piano then when you are faced with technically demanding pieces you can understand what it requires of you without having to recreate the entire fingering wheel of understanding to produce it.

The more pieces you know the better you will be at sight reading because what you read will be procedure at the piano your conscious, muscular and sound memory has a close affinity with.  Your fingers know what to play you don't have to write in the fingering because it is procedure you have done many times. When you stop and keep trying to work out the fingering then you are playing music with content you have little experience with, you can only produce the fingerings instantly without having to stop and decode it beforehand because you are drawing from experiences from many many pieces and the procedure is well known because of that!

Becoming more adept at figuring out what the score asks for (and more adept at executing whatever you can visualise at once, no matter how hard) are what makes sightreading easier.

You haven't explained yourself at all. What does more adept at figuring out what the score asks for.......visualise at once" mean? It is meaningless without any elaboration. Explain this with zero connection with piece memory or experience because you have already said they do not contribute a great deal at all to sight reading. Go on, please tell us what this "figuring out" and "visualise" is all about.

A good sightreader simply decodes scores better and sends intentions to their hands better.

But how do they decode it better? Where do they draw their experience from? Smokes and mirrors!!!

[nyiregyhazi]

"Maybe you are not familiar enough that is why they continue to be least confident with. If you are familiar with them this means you have had a lot of experience playing counterpoint and patterns you face when sight reading will be easily executed and inferred, you will understand what you hear and anticipate the changes, the reading will be more easy and fluent, the fingerings are well known."

I've learned more than half of book one of wtc to a standard of accuracy. I suspect that plenty of pianists who have done it all would still struggle to do fugues of the same standard. By your logic, having learned the complete transcendental etudes should make la campanella feasible sightreading. It becomes easier to learn fast when you know the techniques. But not to process enough information to do it straight off first time, simply because you processed a lot of information before at your own leisure. Sight reading starts with reading information rapidly and accurately. All too many pianists forget that. The reading side doesn't automatically become great either by ploughing through loads of material once or by learning loads of repertoire in your own time- unless you are doing both in the right kind of way. It's no different to the fact that some pianists who practise hours per day are still rubbish. Quantity doesn't always make for quality.

"Put an expert in Bach into a totally different style they will most likely fall flat on their face. Try to sight read some Kasputin or perhaps with other rhythmic music which Bach does not explore. It will be terribly difficult. But put a stride pianist in front of some Kasputin and much of what is read might be easily understood because of their experience base."

So anyone who has played enough kapustin should be sightreading anything else by him? Does that also work with Tatum transcriptions? Learn a few and then breeze through the rest at sight? Sorry but the processing power required to know what is being asked for does not automatically improve by learning complex things under no pressure. It may but it's not guaranteed. Familiarity with riffs and patterns improved ease of execution but not ease of accurate reading, if the demands take a lot of visual processing.

"What about a style which is full of procedure you are unfamiliar with? Saying you have no problems with ANY STYLE is rather over confident, and ridiculous."

Why? I have played very little jazz. I read through numerous Keith jarrett transcriptions fine at first sight. However, I cannot do so easily ones involving elaborate passage work. Not because I don't know the style (because I sure don't have experience of the harmonically dense and chromatic style in the ones I read fine) but because it's too much information to process to rattle it off right there. I have friends who play jazz properly and are classically trained. I doubt whether many of them can actually sightread dictations of elaborate solos notably better than I can and they may even be worse at first sight. They'd be better equipped to improvise around, to fake it. But unless they process more information than I can at greater speed, they'll not rattle them off accurately first time merely down to their knowledge of the style.

Difficulty is what limits people with a broad skill set, not style. Good tools don't stop working on specific musical styles.


"Whatever "slender handful of respects" you mean you have obviously skipped over with no explanation. If I have memorized one thousand pieces you think this will not help reading difficult works?"

That depends whether you got better at processing information by sight better.

"When you stop and keep trying to work out the fingering then you are playing music with content you have little experience with, you can only produce the fingerings instantly without having to stop and decode it beforehand because you are drawing from experiences from many many pieces and the procedure is well known because of that!
"

Actually doing just that is why I sightread fugues much better these days. If you're not good at finding effective fingerings with time to think, there's slender chance under pressure. I learned the processes of how to finger counterpoint better in freer time. But it's not memories in reading so much as transferable skills in executions. Of course skills help execution of what you have correctly read, in sightreading. My point is about the actual reading process. Memory is scarcely an issue there. Even overall, it's general reading skill plus general ability to turn visualised counterpoint into an execution. Clearly there's a difference in the skills to execute counterpoint and homophonic music, but once we start getting much further than that there's no need to know really specific styles. When you have skills in general plus good visual processing, you can be sightreading French or Russian or jazz or whatever. The only limiter is innate difficulty, not style.

"You haven't explained yourself at all. What does more adept at figuring out what the score asks for.......visualise at once" mean? It is meaningless without any elaboration. Explain this with zero connection with piece memory or experience because you have already said they do not contribute a great deal at all to sight reading. Go on, please tell us what this "figuring out" and "visualise" is all about."

I suggest you read the Karl Leimer book that Gieseking was associated with. I didn't think this was exactly an unknown concept in pianism. If it is to you, it's well worth investigating. Gieseking practised little other than visualisation most of the time. It's having an internal conception of what you are actually going to do. All good sightread are ultra advanced in processing information and creating a visualisation from that. Unfortunately, no amount of memory automatically turns densely packed symbols into an instant visualisation of the piece, unless there is extremely rapid ability to actually read and convert those symbols.

[j_menz]

Put an expert in Bach into a totally different style they will most likely fall flat on their face. Try to sight read some Kasputin or perhaps with other rhythmic music which Bach does not explore. It will be terribly difficult. But put a stride pianist in front of some Kasputin and much of what is read might be easily understood because of their experience base.

You forgot the Kapustin Preludes and Fugues. 

[kevin69]

So anyone who has played enough kapustin should be sightreading anything else by him?

I don't think its much to do about having played enough, its about recognising patterns. If you don't understand the style of the composer enough to recognise their characteristic patterns, then you will have to analyse every note, rather than recognize the pattern and verify that the notes do indeed match the pattern.

Unfortunately, pattern recognition seems quite individual and variable, so some people find it hard to recognise patterns in some styles of music, and i don't know of any reliable methods of improving that kind of pattern recognition.

If you understand Kapustin well enough, you can sight read him.
If you can sight read Kapustin, then you understand him well.

[nyiregyhazi]

"I don't think its much to do about having played enough, its about recognising patterns. If you don't understand the style of the composer enough to recognise their characteristic patterns, then you will have to analyse every note, rather than recognize the pattern and verify that the notes do indeed match the pattern. "

But you're starting on the assumption that these are mutually exclusive acts- ie using circular logic. In chromatic music, how can you recognise a pattern unless processing its components to put them against memory? Logic does not suggest memory reference demands less any less actual analysis. It just suggests that it's done more instinctively - ie with more evolved processing skills. Also, in chromatic music there are just so many permutations. It's like if any letter could go next to any other letter in words. Hoping to use memory in anything that is not simple is like trying to build a repertoire of all five letter combinations for recognition. Expectation would generally be low percentage gamble unless confirmed by processing details- which is simply inescapable when there are so many possibilities, memory or not.

"Unfortunately, pattern recognition seems quite individual and variable, so some people find it hard to recognise patterns in some styles of music, and i don't know of any reliable methods of improving that kind of pattern recognition."

You simply need to practise plenty of chromatic music, with a precise awareness of intervals and clarity in terms of how accidental move a specific distance from the reference pitch. Then it scarcely matters how familiar or unfamiliar something is. You calculate first, no matter how slow and then make sure you always get it right first time. Then the process gradually starts to be attempted under more pressure.

"If you understand Kapustin well enough, you can sight read him.
If you can sight read Kapustin, then you understand him well."

In practise, that just doesn't work. I don't understand Keith jarrett harmony one bit. However I sightread transcriptions of his music very accurately first time- by processing what I was being asked to do and then doing just that. I can simply process chords pretty quick, memory or not.

A player who has more advanced skills than myself could do the same with a greater number of notes in kapustin, whether they played him before or not. Another player could learn plenty of the music and still lack the processing skills to deal with the difficulty at first sight. You might as well be arguing that a player who has learned 23 Chopin etudes should thus be able to do the winter wind at first sight. It doesn't work that way. Knowing Chopin in general doesn't either mean you can process that much information at once, or figure the necessary physical issues to execute it without preparation. To put style before innate difficulty of reading and execution is missing the point. Those with the most evolved skills won't need to know the style. Those who know the style but lack the skills won't do as well.

Stylistic awareness might scrape you through some tight corners where you've pushed your ability past its limit and have to scrape your way through with educated guesses or outright fakery. But the primary factor in sightreading is the quality of the fundamental skill set. If difficulty falls comfortably within your skillset, style matters nothing except in terms of the interpretation.

[kevin69]

"I don't think its much to do about having played enough, its about recognising patterns. If you don't understand the style of the composer enough to recognise their characteristic patterns, then you will have to analyse every note, rather than recognize the pattern and verify that the notes do indeed match the pattern. "

But you're starting on the assumption that these are mutually exclusive acts- ie using circular logic.

Where is the circular logic? i don't see it.
You can acquire patterns through slow learning of pieces (ie not sight reading), and then use those patterns for quick verifications during sight reading. During sight reading, the known patterns will help process many notes quickly leaving a few to analyse slowly. The experience of learning pieces slowly or sight-reading may lead you to add extra patterns to your repetoire.

I did not say that you didn't need to process every note.
I said that you don't need to analyse every note, because you can verify that the note fits a pattern much faster.

[nyiregyhazi]

Where is the circular logic? i don't see it.
You can acquire patterns through slow learning of pieces (ie not sight reading), and then use those patterns for quick verifications during sight reading. During sight reading, the known patterns will help process many notes quickly leaving a few to analyse slowly. The experience of learning pieces slowly or sight-reading may lead you to add extra patterns to your repetoire.

I did not say that you didn't need to process every note.
I said that you don't need to analyse every note, because you can verify that the note fits a pattern much faster.

All you've done is repeat the unproven assumption above and build the rest on top. You can't use something to prove itself. The fact that patterns are recognised as familiar does not prove that the brain processes them more easily specifically because of memory. It is equally possible (and more logically probable based on other issues) that the brain processes components just the same and then matches what it has processed against memories to see if it fits or not. In other words memory conceivably aids execution after identification of a stock technique, but has little effect on actual processing of the instructions.

If it has to fit a pattern, forget ever learning to sightread thick chromatic jazz harmonies or atonal music. I don't notably differentiate between processing and analysis.  However I certainly didn't consciously "analyse" when I read through totally unfamiliar patterns in the jarrett transcription, if that's what you presume would have to happen. If you think you need to know the style in order to avoid conscious analysis, you're wrong. I just processed the information in the same way internal way I process a much easier c major chord and my fingers found the keys. I have methods for very consciously analysing as a training process but I didn't consciously use any of those processes. It was all internal processing based on the acquired skills and direct familiarity was simply not necessary whatsoever. Good enough processing skills are why I could do it. Good sight readers don't need anything in the memory bank other than the basic reading skills. I don't care if something fits a memory. If I can decode the information quickly, I can execute it whether that involves matching to a familiar pattern or whether that involves seeing a brand new pattern thanks to the notes I have read. I don't need anything to match up to any memory, as long as I've read it correctly in time. I'd be a poor sightreader I was limited to executing only chords that I've done before, without a method for new ones.

You're speaking as if it's crazy to think that someone can actually read a comprehensive checklist and use precisely that to get the result comprehensively correct. How on earth could it be more rational to assume that note perfect executions come more from hoping that things fit memories that they might not, than to assume that perfect execution comes from the skill to simply process what it actually says to do, both quickly and precisely enough?

[kevin69]

Sorry, i am completely failing to get my point across here.
I don't know how to explain my viewpoint in a way that you will understand, so i'll stop trying.

[flashyfingers]

A good goal is to be able to sight read away from the piano. Sing the melody, practice rhythm. That way, if you can sightsing the melody correctly, you can already kind of figure out how the piece is going to sound, and no modulation shall throw you off. Also, since in piano you are reading 2 parts, make sure when you look ahead, you can memorize one of the hand's part based on the next beat or few beats, and play the other part, timing both to happen at the next beat, which happens the next time you look at the sheet music.

[j_menz]

timing both to happen at the next beat, which happens the next time you look at the sheet music.

What are you proposing you look at in the interim?

[flashyfingers]

What are you proposing you look at in the interim?

your hands! if need be

thank you, jmenz 

[j_menz]

your hands! if need be

thank you, jmenz 

Part of becoming a good sight reader is to reduce the need to look at your hands. Ideally, eliminate it entirely, but in the interim reduce the occasions where it is needed and reduce the time it takes to do so.

Two reasons for this: first it gives you more time with the score to take in and analyse information, secondly, it stops you losing your place and repeating or missing whole bits.

[lostinidlewonder]

If you don't understand the style of the composer enough to recognise their characteristic patterns, then you will have to analyse every note, rather than recognize the pattern and verify that the notes do indeed match the pattern.

That is very clearly written and I agree with it very much so!!

[lostinidlewonder]

(L)"Maybe you are not familiar enough that is why they continue to be least confident with. If you are familiar with them this means you have had a lot of experience playing counterpoint and patterns you face when sight reading will be easily executed and inferred, you will understand what you hear and anticipate the changes, the reading will be more easy and fluent, the fingerings are well known."

(N)I've learned more than half of book one of wtc to a standard of accuracy. I suspect that plenty of pianists who have done it all would still struggle to do fugues of the same standard.

(L)I play all 48 preludes and fugues and they became easier and easier to sight read the more of them I learned and the more of Bachs music that I played as a whole. Put a new fugue in front of me and I can play all the notes with good fingerings no problems, if it is a rapid tempo then of course you can't rattle it off at speed instantly but that is irrelevant because from slow and controlled playing comes any speed.


(N)By your logic, having learned the complete transcendental etudes should make la campanella feasible sightreading.

(L) You can sight read first go it with the correct fingerings perhaps at half tempo no worries.


(N) It becomes easier to learn fast when you know the techniques. But not to process enough information to do it straight off first time, simply because you processed a lot of information before at your own leisure. Sight reading starts with reading information rapidly and accurately.

(L) You are misunderstanding how sight reading is used. We don't just read it 1 time and expect to play it at tempo perfectly. But I can sight read it 100 times in a day and get it solved, rather than a memoriser who has to spend weeks analyzing their fingerings, notes etc etc. Sight reading, do you really know what is used for


(N)All too many pianists forget that. The reading side doesn't automatically become great either by ploughing through loads of material once or by learning loads of repertoire in your own time- unless you are doing both in the right kind of way.

(L) If you actually did read through a lot of music you would realize that it helps you to read the generali procedures there is in musical writing. There is no magical "right kind of way" you are trying to say again with ZERO clarification as to what this right kind of way is lol. You will simply be able to read better and faster and more accurate the more you read. There are of course structured ways to improve your sight reading, but simply reading a lot of music will help AUTOMATICALLY.


(N)It's no different to the fact that some pianists who practise hours per day are still rubbish. Quantity doesn't always make for quality.

(L)If everyone who practices does so in a wrong fashion then of course they are not going to get better. But the reality is that most people who practice will see improvement, there is only a small portion of people who practice completely wrong and this can be easily remedied with a good teacher. No practice is worse than bad practice.


(L)Put an expert in Bach into a totally different style they will most likely fall flat on their face. Try to sight read some Kasputin or perhaps with other rhythmic music which Bach does not explore. It will be terribly difficult. But put a stride pianist in front of some Kasputin and much of what is read might be easily understood because of their experience base."

(N)So anyone who has played enough kapustin should be sightreading anything else by him?

(L) Yes, but someone who plays a lot of Kasputin would play a lot of other styles which are similar from other composers. There is a network of knowledge that will allow you to read a particular style fluently. Playing 5 of his pieces is not a lot.... Sightreading through all his works and you will see his overall style. Read through other composers as well and you will see similiarities and contrasts between different people.



(L)"What about a style which is full of procedure you are unfamiliar with? Saying you have no problems with ANY STYLE is rather over confident, and ridiculous."

(N)Why? I have played very little jazz. I read through numerous Keith jarrett transcriptions fine at first sight. However, I cannot do so easily ones involving elaborate passage work. Not because I don't know the style (because I sure don't have experience of the harmonically dense and chromatic style in the ones I read fine) but because it's too much information to process to rattle it off right there.

(L) Maybe not on your first read but what about those that follow? How fast does it improve? 


(N) Difficulty is what limits people with a broad skill set, not style. Good tools don't stop working on specific musical styles.

(L) A style can be difficult because of its style.... If you haven't read the procedure enough then you are not familiar enough with it. difficulty may slow the tempo of your playing while sight reading but it is irrelevant because from controlled playing comes any tempo.


(L)"Whatever "slender handful of respects" you mean you have obviously skipped over with no explanation. If I have memorized one thousand pieces you think this will not help reading difficult works?"

(N)That depends whether you got better at processing information by sight better.

(L) Learning thousands of pieces does nothing at all? You think there could be even a chance that it does nothing? AHAHAHAHAHAH



(L)"When you stop and keep trying to work out the fingering then you are playing music with content you have little experience with, you can only produce the fingerings instantly without having to stop and decode it beforehand because you are drawing from experiences from many many pieces and the procedure is well known because of that!
"

(N)If you're not good at finding effective fingerings with time to think, there's slender chance under pressure. I learned the processes of how to finger counterpoint better in freer time.

(L) This means you can't sight read the works because you are still studying them to become familiar with them. The fingering is not automatically understood because you haven't gone through the procedure countless times.
 

(N) But it's not memories in reading so much as transferable skills in executions.

(L) If there is no memory in the reading then how do you see a group of notes and react to it immediately without having seen it before? You must remember the pattern, at first the memory is raw conscious observation, later it becomes an instant muscular reaction in the fingers to what the eyes see.


(L)"You haven't explained yourself at all. What does more adept at figuring out what the score asks for.......visualise at once" mean? It is meaningless without any elaboration. Explain this with zero connection with piece memory or experience because you have already said they do not contribute a great deal at all to sight reading. Go on, please tell us what this "figuring out" and "visualise" is all about."

(N)I suggest you read the Karl Leimer book that Gieseking was associated with.

(L) AHHAHAHAHAAHAH... wait? HAHAHA... please. If you cant explain yourself don't throw boooks at me.

[j_menz]

LNLNLNLNLNLNLNLNLNLN

Once wasn't enough?

BTW, it's Kapustin, not Kasputin.

Somewhere in there though you hint (or more)  that there may be different purposes for which we may be sight reading a piece, and I think that's an important point.

Not the sight-reading vs just reading thing, but what we are reading for. For example, if one is reading to accompany a choir or other instrumentalists, the approach needs to be one of maintaining tempo and reasonable accuracy, missing what's not doable. If you're reading as a sort of private performance - this is what it goes like - then something like that works too, but maybe a bit more flexible. If you're reading to get a feel for a piece, the tempo can be a bit more fluid and off the pace, you can afford more mistakes (not a lot) and so forth. If you're reading to see what you would need to do to learn the piece, even slower, less steady and mistake ridden is possible. Horses for courses - so long as it works for what you're hoping to achieve with the exercise.

[nyiregyhazi]

(L)I play all 48 preludes and fugues and they became easier and easier to sight read the more of them I learned and the more of Bachs music that I played as a whole. Put a new fugue in front of me and I can play all the notes with good fingerings no problems, if it is a rapid tempo then of course you can't rattle it off at speed instantly but that is irrelevant because from slow and controlled playing comes any speed.

That isn't the point, if we're talking about advanced sightreading. As an experiment I just read through the B minor fugue of book 1 which is already slow. I had to a go a little slower than performance tempo but I could do it with a high standard of accuracy when I allowed a tiny bit of added space. I'm not in my comfort zone in the same way as in less contrapuntal styles but I can do it with focus. Then I tried the A minor fugue and fell flat on my arse unless I slowed down to around half the performance tempo.

You don't hear a whole lot of legendary stories about Liszt or Richter performing the remarkable feat of going slowly enough to get through a difficult piece first time and everyone being impressed because that's how you should learn pieces anyway. Less so still if a pianist agrees to accompany a Brahms sonata they don't know for an examination, due to an emergency. In relation to all this memory business it takes away from your line of argument- because going slow means it's more feasible to simply process all the information at that slow tempo. If the memory thing is so good WHY do you have to go slow- if not for the simple reason that you cannot process information quickly enough to actually decode all the written instructions (for reference against memories) except when going slow enough to do more processing? It's not a strong case. You can't do it faster because you simply can't process enough information to do it quickly. Truly great sightreaders can.

(L) You can sight read first go it with the correct fingerings perhaps at half tempo no worries.

Again, hardly a truly remarkable feat. Yet, equally, many skilled performers would struggle. Some people don't process the written score quickly enough unless they are allowed to take pauses to stop and have a good think. Learning is not a speed contest.

(L) You are misunderstanding how sight reading is used. We don't just read it 1 time and expect to play it at tempo perfectly. But I can sight read it 100 times in a day and get it solved, rather than a memoriser who has to spend weeks analyzing their fingerings, notes etc etc. Sight reading, do you really know what is used for

Actually, great sightreaders do. I'll not stop considering what creates the skills unless I can do just that. Being able to do something really slowly as the first step of learning a piece is not a remarkable ability but merely a pretty decent one. Also, you make a crazy false polarisation there. Optimal learners would typically neither "sightead 100 times" nor spend weeks due to analysis. They'd have a good hard look and think about a passage to mentally clarify what they want and use that deep understanding to have it down fast- a la Gieseking, who typically expected to get things right first time due to mental clarity. 100 times? Seriously? Is it not better to actually think where the target is rather than just fire off 100 shots?

(L) If you actually did read through a lot of music you would realize that it helps you to read the generali procedures there is in musical writing. There is no magical "right kind of way" you are trying to say again with ZERO clarification as to what this right kind of way is lol. You will simply be able to read better and faster and more accurate the more you read. There are of course structured ways to improve your sight reading, but simply reading a lot of music will help AUTOMATICALLY.

All the time I've been playing I ploughed through tonnes of music. I was a reasonable sightreader but I faked everything. The throw time at it method is very poor without organisation. You don't improve your limits unless you work at processing the information in fully first time without fakery. When I took this attitude (allowing tempo stretching if need be) I also improved at sightreading in strict tempo- and dramatically so. You need a variety of methods, not simply enough time on reading new stuff.

It's like in maths. If someone spends enough time on long sums, they will get better at processing any sum quickly via the base memories such as 2+2= 4 - not "memorise" all additions of two five digit numbers for memory recall. But only if they learn good methods while doing so. The children who are trained to use abacuses for rapid mental arithmetic will develop far quicker than those who use the standard western method taught in school. It's not an issue of memory and time only aids speed of processing notably if  it is focussed to develop the right experience and skill- just like musical reading. That's why neither a chord nor a specfic sum needs to have been met before to be either read or executed quickly, by a person of skill.

(L)If everyone who practices does so in a wrong fashion then of course they are not going to get better. But the reality is that most people who practice will see improvement, there is only a small portion of people who practice completely wrong and this can be easily remedied with a good teacher. No practice is worse than bad practice.

That's quite a novel argument about the small percentage. Unfortunately, if we're talking about advanced rather than notably subpar sighteading (and I thought we were, no?), the most important thing is that bad practise is far worse than good practise and that good practise doesn't find itself- regardless of the statistic you invented about how only a small percentage practise wrong. Only a small percentage practise completely right is the reality- as evidenced by the fact that truly great sightreaders are rare, not the norm.

(L) Yes, but someone who plays a lot of Kasputin would play a lot of other styles which are similar from other composers.

Indeed. Processing other densely chromatic music in which few details are simplistically predictable would prepare them. Which is exactly why you need not have played a single work by him if you sharpen your tools elsewhere. Especially if half tempo counts as amazing sightreading, in your book.

(L) Maybe not on your first read but what about those that follow? How fast does it improve?

What? That comment makes no sense on any level. Who ever said memory wouldn't be an issue when you play exactly the same piece over again? I'm talking about the slender relevance of memory when decoding brand new material for the first time.

(L) A style can be difficult because of its style....

Yes, if there is a large amount of information to process in order to decode the score. Which is exactly why your best effort has to be in half-tempo, no matter how well you know the style, when it's that complex (and why ANY style is easy when there's less information to process) It's still hard because you need to go slow enough to allow your processing skills to do the job of identification. But a truly great sightreader with better processing skills can do it faster at once whether they know the style or not.

(L) Learning thousands of pieces does nothing at all? You think there could be even a chance that it does nothing? AHAHAHAHAHAH

Drop the strawman. It's a cheap way to argue. I said it doesn't guarantee excellence. Do you think there might perhaps be another category that falls between zero improvement and achievement of excellence? I can certainly see the possibility for somewhere between those two extremes. Can you not? You subscribe to the idea that anyone who has not achieved rare excellence must therefore have achieve "nothing"?  

As I said, I did tonnes of sightreading all the time. I improved in a truly notable way, only when I made changes to the METHODS which I used while reading material- which improved the quality of my learning procedures in general and which also spilled into how effectively I process new information under tempo pressure.

(L) If there is no memory in the reading then how do you see a group of notes and react to it immediately without having seen it before? You must remember the pattern, at first the memory is raw conscious observation, later it becomes an instant muscular reaction in the fingers to what the eyes see.

Why is it a mystery to everyone that a clear and comprehensive checklist of information can actually be read off and executed without recourse to anything but itself? I already told you, I had no problem reading a Jarrett transcription without stylistic knowledge of jazz harmony. Why do I need any? The score covers that for me by telling me what to play. Is it seriously so hard to believe that I can simply read what a score says and then do just that, for no other reason than because that's what it told me too?

If this is controversial (and would demand the assumption that I secretly know jazz harmony and got my accuracy by taking educated guesses based on memory, rather than by simply observing what it said) the mind simply boggles. It's one thing to believe that memory helps to some degree in some areas but it's plain irrational to make such a silly claim, and speaks of desire to prove a fixed outcome at any cost, rather than an open and self-critical mind. I sure as hell have no experience of popular Chinese tunes either, or composers like Einaudi but it certainly didn't make it any harder to read them off at once when students have brought them. After looking at broad evidence, I dropped the beliefs I used to hold about how reading occurs well "because" of memory and switched to more credible view that any memory recollections are caused by good reading practises- and that you can do just as well where memory is not an issue, depending on how much information must be processed.

If something is difficult for the brain to process, it's difficult to sightread. If something is easy to process, it's easy.

(L) AHHAHAHAHAAHAH... wait? HAHAHA... please. If you cant explain yourself don't throw boooks at me.

If you can't be arsed to research the very widely known concept of visualisation, I'm not going to fill in the gaps for you. You might as well insist that I have to explain how C major fingering works before I'm allowed to speak of it. Visualisation is not some rare or obscure concept in pianism so I suggest you fill in the gaps in your knowledge by doing some research- rather than heckle a method that was used by Gieseking, Fiorentino, Hofmann and also Volodos today.

[flashyfingers]

Part of becoming a good sight reader is to reduce the need to look at your hands. Ideally, eliminate it entirely, but in the interim reduce the occasions where it is needed and reduce the time it takes to do so.

Two reasons for this: first it gives you more time with the score to take in and analyse information, secondly, it stops you losing your place and repeating or missing whole bits.

Looking at your hands shouldn't throw you off from your place in the score, if you completely synchronized where you are in your left hand part and where you are in your right hand part, and where your hands are on the piano and everything else that is going on with the score and t he performance.

In my opinion, you shouldn't need much time to look at the score and play the right thing, if you are a good sight reader. But while you are learning, you should be able to look away, look at your hands, analyze that situation and adjust immediately, and be able to switch what the hands are doing to where you were in the score, except to the next place of musical instance.  

What it seems to me, you're saying, is about what a good sight reader does. I am trying to explain how sight reading works for me. There isn't always a prelude or fugue to be read. There are crazy contemporary pieces where notes aren't practical skips or mostly by step/close together. And yes you would practice those impractical skips to be reliable without looking, but what if you are reading on the spot?

It is important to look ahead, always. And be aware of what your hands are doing, not only by having a sense of touch and awareness of how that works for you in sight reading, but also of how looking at your hands can help you be more precise sight-reader.

If you have the chance to learn a piece purely by sight-reading and not looking at your hands, then do it, and yes, you would want to mostly or altogether eliminate the looking part.  Although, this has benefits and downfalls, because some people memorize only one way and not the other. Your strengths should be many, not few. In my opinion

[timothy42b]

Looking at your hands shouldn't throw you off from your place in the score, if you completely synchronized where you are in your left hand part and where you are in your right hand part, and where your hands are on the piano and everything else that is going on with the score and t he performance.

Well, maybe it shouldn't.  But it sometimes does for me, and more as I age.  Where I notice it is when playing in a wind ensemble, and while sometimes I can follow the conductor with peripheral vision, at times I need to memorize a couple bars and actually focus on him.  Then it isn't always easy to quickly get my eye back on the right spot.  Older eyes don't change focus as quickly.

[j_menz]

There isn't always a prelude or fugue to be read. There are crazy contemporary pieces where notes aren't practical skips or mostly by step/close together. And yes you would practice those impractical skips to be reliable without looking, but what if you are reading on the spot?

No note is ever more than 87 from the last one, and the piano mostly stays still.

There are, I should point out, crazy contemporary preludes and fugues that employ large distances between notes too, but I wasn't limiting myself to that genre anyway.

If you're reading on the spot, you need to be able to manage those larger gaps with minimal looking. None if you can.

You appear to be confounding reading (off the sheet a piece you know) with sight-reading (a piece you've never played or heard before).  In the former, losing your place is less likely, but in the latter it's entirely possible - especially where bars look alike.

[nyiregyhazi]

No note is ever more than 87 from the last one, and the piano mostly stays still.

There are, I should point out, crazy contemporary preludes and fugues that employ large distances between notes too, but I wasn't limiting myself to that genre anyway.

If you're reading on the spot, you need to be able to manage those larger gaps with minimal looking. None if you can.

You appear to be confounding reading (off the sheet a piece you know) with sight-reading (a piece you've never played or heard before).  In the former, losing your place is less likely, but in the latter it's entirely possible - especially where bars look alike.

Nothing is banned outright. Good sightreading skills are based on breadth of experience. You're missing the fact that good sightreaders process information so quick that they should not always have to keep their eyes on the score. If they do have to stay glued forwards, they're the limited one- not the sightreader who can process and remember a chunk well enough to free up their vision for a bit.

One of the most important exercises is to look at a bar or half bar or whatever it takes and deliberately look away from the score as you execute it. I usually look at a wall or something rather than at my hands. But it's a powerful exercise in visualisation. Even in pieces you know, you can look at a chunk and then try to make sure the hand are executing on command of reading and visualising- to try to avoid either coasting on physical habit alone or reading on a note at a time basis.  If I read the score continuously, it simply doesn't test my visualisation or processing speed in the same way. It doesn't force you to make bigger mental organisations or plan in full This little trick both aids my sighreading practise and how I work on learned pieces. It also helps build the skills that allow some to do meaningful practise entirely in their head.

 If a good sightreader has the skill to glance and memorise a large chunk at once, so they can sometimes move their vision to their hands for a bit without any problem, all credit to them. Options are a positive, not a negative.

[j_menz]

Options are a positive, not a negative.

If it's an option, it's fine. For most struggling sightreaders, it's a necessity to look at the hands, rather than an option. And that necessity will hinder them.

[nyiregyhazi]

If it's an option, it's fine. For most struggling sightreaders, it's a necessity to look at the hands, rather than an option. And that necessity will hinder them.

Sure. But so will only looking at the score. They should practise instant memorisation of bars and deliberately play from visualisation of what they read, rather than allow themself to work note by note, based on small quantities of decontextualised information. I can't think of any safer way to break the habit of reading small, rather than big chunks.

[j_menz]

Sure. But so will only looking at the score. They should practise instant memorisation of bars and deliberately play from visualisation of what they read, rather than allow themself to work note by note, based on small quantities of decontextualised information. I can't think of any safer way to break the habit of reading small, rather than big chunks.

So you're suggesting that instead of reading (once) of the page, they read of the page, memorise it, and then read off the memorised image? And that's better? 

[flashyfingers]

So you're suggesting that instead of reading (once) of the page, they read of the page, memorise it, and then read off the memorised image? And that's better?  

No, you look at it once, memorize it (in a way that works for sight reading), play it. If you are proficient enough in piano, in general, you don't need to look at your hands while you READ music, anyways.

It works great in sight reading, but that's how I learn music, in the end, too...

[nyiregyhazi]

No, you look at it once, memorize it (in a way that works for sight reading), play it. If you are proficient enough in piano, in general, you don't need to look at your hands while you READ music, anyways.

It works great in sight reading, but that's how I learn music, in the end, too...

To clarify, I'd just add the term visualisation. I can sort of "see" the text for a brief time but it's not exactly sightreading from a photographic memory (and I certainly don't retain an internal picture of how the score looked beyond the short term). It's more a case of picturing the meaning of the symbols and having a comprehensive picture of how to execute them, before starting- so no details need to be confirmed by an extra look at the music. If that's impossible, a player must start with smaller chunks and practise making them bigger. Good sight readers have to both process a lot of notes quick and be able to form an accurate visualisation of the execution. This is an excellent reveal of whether you can actually do that, or whether you can only bumble your way around a note at a time. Without seeing a broader context, notes cannot flow smoothly. Someone who doesn't sightread well may be poor at decoding the text, but it's equally possible that they just aren't very good at processing chains of notes into a broader visualisation.

This comes back to the memory illusion too BTW. Memory is an organiser of what you read, not the explanation for how you read it. After reading many notes, a really good sightreader creates a visualisation from whatever they just read-regardless of whether it fits something routine. A less experienced one may be able to visualise a run of notes that fit C major, but they may not yet  have the visualisation power to process 6 chromatic notes that fit no obvious pattern and instantly turn them into a single completed thought, that can then be executed without checking back to the music. A really good sightreader can either use memory of a standard pattern to simplify organisation of what they have read (if it fits), or they can create their own internal organisation to cover notes that fit no routine pattern. A lesser reader is just out of their depth when they have to make their own pictures in an instant, without reliance on existing ones. Obviously the good readers they don't look away much in normal sightreading. But forcing yourself to is exactly how you train and check the speed of visual processing plus the internal ability to organise many details into chunks that are ready for execution. It's way too easy to stay on a small scale (no matter how many times you hear about the need to read ahead) unless you force short term memory of larger chunks to evolve.

PS. the next challenge is where rather than look away, your eyes jump to the next chunk and start processing, while your fingers are doing the one you already processed. Although I'm a fair sightreader, attempting to always be one bar ahead shows me that my eyes are too absorbed in wanting to confirm the chunk that I'm still doing, rather than getting truly stuck in to the next chunk. I don't think small but I've realised that I have room to train myself to think far bigger still. It's really useful to have practise techniques that actually make it possible to bring the advance reading thing about - rather than stock advice that merely says that you're meant to be able to do it.

[j_menz]

attempting to always be one bar ahead

Why pick an arbitrary amount like a bar?

Not all bars are equal, and the amount of information in one an be vastly (vastly!) greater than another. If you aim for "a bar ahead", for any level of processing ability bar the heroic sometimes it's trivial and sometimes it's impossible.

[nyiregyhazi]

Why pick an arbitrary amount like a bar?

Not all bars are equal, and the amount of information in one an be vastly (vastly!) greater than another. If you aim for "a bar ahead", for any level of processing ability bar the heroic sometimes it's trivial and sometimes it's impossible.

It's not a rule. As I said, it could be half a bar. Whatever is reasonable.

[lostinidlewonder]

Once wasn't enough?

BTW, it's Kapustin, not Kasputin.

I could have said Rasputin, you'd still know who i was talking about right? Thanks for correcting me I am not good with names. I still mistake Schubert and Schumann all the time -_-

....there may be different purposes for which we may be sight reading a piece, and I think that's an important point.

Not the sight-reading vs just reading thing, but what we are reading for. For example, if one is reading to accompany a choir or other instrumentalists, the approach needs to be one of maintaining tempo and reasonable accuracy, missing what's not doable. If you're reading as a sort of private performance - this is what it goes like - then something like that works too, but maybe a bit more flexible. If you're reading to get a feel for a piece, the tempo can be a bit more fluid and off the pace, you can afford more mistakes (not a lot) and so forth. If you're reading to see what you would need to do to learn the piece, even slower, less steady and mistake ridden is possible. Horses for courses - so long as it works for what you're hoping to achieve with the exercise.

Sight reading can be used in many ways and you have given a few good examples. It is not necessarily being able to pick up a piece and play it at mastery, but at the same time it is and you should be able to do this for particular pieces. One should also question what is stopping them from being able to sight read other works with mastery, you can certainly discover a lot studying the piano in terms of sight reading, sure puts a different spin on the approach to the study (Eg: reading hundreds of pieces a month opposed to polishing a few pieces to mastery).

[lostinidlewonder]

The madness begins!


Quote from: nyiregyhazi on August 08, 2014, 01:11:55 AM
Memory of other pieces doesn't make works of great difficulty any easier in anything but a slender handful of respects....


L:
Whatever "slender handful of respects" you mean you have obviously skipped over with no explanation. If I have memorized one thousand pieces you think this will not help reading difficult works?

N:
That depends whether you got better at processing information by sight better.

L: Learning thousands of pieces does nothing at all? You think there could be even a chance that it does nothing? AHAHAHAHAHAH

N:Drop the strawman. It's a cheap way to argue. I said it doesn't guarantee excellence. Do you think there might perhaps be another category that falls between zero improvement and achievement of excellence? I can certainly see the possibility for somewhere between those two extremes. Can you not? You subscribe to the idea that anyone who has not achieved rare excellence must therefore have achieve "nothing"? 

As I said, I did tonnes of sightreading all the time. I improved in a truly notable way, only when I made changes to the METHODS which I used while reading material- which improved the quality of my learning procedures in general and which also spilled into how effectively I process new information under tempo pressure.



L:
Wow... I mean... wow, in one mad rattling of keyboard strokes you have managed to super tangent the entire argument into a multi armed monster of randomness.

[nyiregyhazi]

So, out of all the many points that a person with strength of conviction could have responded directly to (including those where I outlined the methodology that you had heckled me for not providing and demanded to know about), your response is to take a small passage that pertains entirely to the topic of sight reading and call it a "tangent"?

If you want to make a topical follow up (rather than attempt a to divert everything into a single tangent, about tangents) , I'll continue. If your only interest is in going off topic and taking ad hominem shots, I'll bow out. You are welcome to a free shot in which you can slag me off outside of the topic, because I will not be replying to anything further unless it contains a topical argument. A free low blow is yours, Sir, so please go ahead and "win" by taking it. Feel free to add more cackling too, so we can see how much you enjoy your uncontested victory (out of a field of yourself and not a single other interested competitor).

As you've left the topic by the wayside, I'll leave you to teach your students on the basis that quantity is the most important, the assumption that most people already know how to practise and the motto that lots of bad practise is better than none at all. (I sincerely wish those points were a ridiculous strawman of the kind you had created to argue against- rather than points you sincerely used to support your stance, in your last post)

[timothy42b]

The madness begins!continues!

Fixed that for you.