Piano World Home Page Forums Our Most Popular Forums Pianist Corner why is sight-reading hard?

In particular, why is sight-reading so much harder than learning a natural language?

The recent sight-reading posts remind me of this question I've always thought about. Sight-reading is certainly something you can work on, and maybe get good at. But very few of us will ever be able to sight-read a Rachmaninoff etude-tableau at tempo. Why is that? Why doesn't our brain eventually learn to process notes as fast as it processes letters?

To put it in perspective: you can, in your 40s, start studying a language with thousands of characters, like Chinese. Now, it might take years of serious study, but eventually, you can achieve adult literacy, and basically just sit down and read anything. So why can't I, who have been reading musical notes since I was 6, just sit down and play anything?

I'm not a terrible sight-reader; maybe I can read a Chopin Mazurka at half-speed, with stops and mistakes. But imagine if whenever you tried to read a newspaper article, you had to read it first at half-speed! What's the difference? I have a couple hypotheses, but I'd like to hear what other people think.

Finally, for inspiration, here's a link to a great sight-reader. This is how it should be...

http://www.youtube.com/watch?v=JZMroQOtS_U&feature=related

I think the weakness in your argument is equating reading music with reading a piece of text. When one reads music, one is not simnply reading one word at a time or even one phrase at a time in a horizontal, linear fashion. One is also reading, and trying to absorb, vertically.

One is often
- reading a number of notes for the right hand and a number of notes for the left hand, simultaneously
- reading dynamics and accidentals
- reading pedal indications

Then, one has to transfer from the written page, through the brain and through body movements to the instrument, one's interpretation of all these symbols. It seems to me that other instrumentalists and vocalists, most of whom read only one note at a time, could appear to be better sight readers than pianists. Consider, however, their challenges compared to that of the pianist : two staves, two hands, multiple notes on each stave and even (sometimes) two feet on three pedals.

It seems to me, therefore, albeit from a somewhat simplistic, non-technical point of view, that sight-reading a piece of music is a much more complex activity than reading a language from a printed page.

I have never found sight reading difficult. Perhaps that is because I don't expect that, whatever my training, I should be able to read a Liszt Etude or a Bach four-part Fugue, up to speed, at sight.

Regards,

Why is sight-reading hard?

Are you still beating your wife?

Who is the King of France?



Steven

The average American adult reads prose text at about 300 words per minute. The average professional typist can type about 70 words per minute. There is more than a factor of four difference between the speed with which the brain processes symbols on a printed page and the speed with which the motor system can translate the processed symbols into the production of text at a keyboard. The same is true for sight reading, or more accurately, “sight playing” at the piano. Your brain can take in the music symbols quite rapidly but you cannot expect your fingers to follow that rapidly. It simply takes time to activate the playing mechanism of the arms, hands, fingers, and feet. The other limitation is that of technique. If you cannot play practiced scales at 160 bpm how on earth would you expect to “sight play” a scalar passage at an indicated tempo of 160 bpm?

Very insightful posts, BruceD and jazzyprof!

I have never found sight-reading hard. Memorizing, now, that's another matter...

Besides, it all happens in different parts of the brain... playing piano in my brain seems to be located in the same area as numbers are... mighty strange

if you go for reading, you might want to think about texts, that are not familiar to you... foreign words or strange grammar or unknown use of phrases... you inevitably have to slow down... you're still able to make sense of it, just not as fast as if it was, let's say, a child's book...

Very good question! I've wondered the same quite a few times. Interesting answers, too!

I've nothing to do with biology, but I feel that language is pretty different from music in terms of the way our brain handles them. Words evoke an immediate response (they make sense at once) from our brain, whereas music isn't something immediate. We naturally take time to understand and enjoy the music.

I recently found out that I cannot do math while playing. I tried to take the metronome speed of 78 and calculate what 79 would be if the time were doubled, i.e. 78 x 2 vs. 79 x 2. (I was playing with the click on every other beat).

It was almost impossible to do, although I can do that in my head while not playing without too much stress.

Since then, I have been forcing myself to do simple math as I play something complex, in the hope that it will expand my playing ability, because the same part of the brain is apparently involved in both.

I completely agree. In addition to piano I also play cello and I sing. I have no problem sight reading with those instruments, but on the piano it's a whole different story. Because of polyphony; it's a lot of information to process at once.

Besides the good points already made, there is also the fact that reading words is related to the language skills that virtually all of us develop automatically and use everyday of our lives, whereas playing the piano is a decidedly "artificial" thing to do.

I think a large part of it is the way many of us are taught to read music, using a mnemonic device such as "every good boy does fine" or "all cows eat grass". This works great when you are taking a test in music class, but completely destroys any hope you have of reading notes at speed. I wish there was a way to completely obliterate those phrases from my mind, to force me to just memorize notes by location and not falling back on that crutch. (And I've been trying flash cards, along with the on-line tools - getting better but still not there yet).

The other problem is you are required to do too many things at once (more than one thing, that is). First, try typing some text using only one hand. Now interleave the lines from two different texts -- type the top line with your right hand on one keyboard, and type the second line with your left hand on the second keyboard, and make sure that both hands finish the line at the same time. That is almost what it takes to go from sheet music to a piano keyboard.

This also reminds me of a talent I saw on one of those 70's era variety shows (I think it was "Real People" or something like that). They had a guy that could whistle and hum at the same time. He'd whistle one tune while humming a harmonizing tune at the same time. Never could figure that one out.

That is a bit of an overstatement...I have plenty of students who learned those mnemonics (or others) who can sightread.

I don't think the rag time sight reader in the OP was that impressive, although personally I couldn't sight read that as well. The chord progressions in those old rags are very predictable and if one plays a lot of ragtime the jumps are pretty easy.

It would be interesting to see how well he could sight read a classical piece. I think many professional pianists could sight read much more difficult material.

I once heard Irina Morozova give a master class where the student played some obscure piece by Szymanowski (which makes the rag look like a piece someone would play in the first few months of study}. Morozova said she had played some Szymanowki but had never heard the piece. She followed the score while the student played it and after hearing it once was able to play pages of it at sight. Not exactly like sight reading the piece, but the piece was so difficult I could never learn to play the notes with years of practice.

Interesting ideas.... Two responses:

1. The analogy between reading music and reading natural language is imperfect, of course. But I don't totally buy that it breaks down because music has many lines to read at once, while language has only one. When we read a sentence, we don't read letter by letter; the brain groups the letters into processable chunks. Similarly, when we sight-read, we don't read note by note; the brain groups vertical sections, chords and melody lines for example, into similar chunks. Are these vertical chunks inherently far more complicated than their language equivalents? Yes, and maybe that's just the difference. Still, I'm surprised I can't sight-read at least simple things more easily.

2. Excellent demonstration, jazzyprof, about the physical factor in all this. Of course I can't read an etude-tableau at speed, when working it up would take weeks or months of training my fingers on the technique!

So, beet31425, what are your hypotheses?

Rocket I read your post yesterday and, inspired by you, was going to start doing likewise until the penny dropped this morning that, no offense intended, I think you are wasting your time.
It has been shown in numerous studies, using a number of different study designs, that if you use your mobile phone whilst driving you are more likely to crash. We all know from personal experience that it is difficult to have a conversation whilst trying to do something else.
Your logic suggests that we would become better drivers if we practiced driving whilst using our mobile phones, or that politicians, broadcasters and others who talk for a living would say more sensible things if they practiced asking the kids how their day was whilst trying to work out the ingredients for the evening supper. Clearly this is nonsense.
There is no part of piano playing that requires you to perform mental arithmetic whilst playing (unless you're a professional trying to work out if you've put enough money in the parking meter to see you through your gig, or if the pay will be enough for that weeks outgoings).
I honestly think Hannon, scales, Czerny or anything else on could think of would be better at improving your playing than practicing mental arithmetic whilst playing, and stick to Soduku and Nintendo brain training to improve your maths!
Best wishes

Rocket
Just listened to those excellent tracks on your home page and you don’t need Hannon or Czerny: you Rock big time already.
But I still think you’re wasting your time practicing a redundant skill.
Bigger best wishes and thanks for posting that great playing on your home page
Wombat

Well, you can spend 40 years learning Chinese and still be clueless reading the ancient texts (for example, any of the Confucian texts). You can probably sound out 90% of the characters, but you'd have no idea what you just read. The grammar, syntax, and different combinations of characters will be near-impossible to decipher without weeks of extensive study.

Similarly, you can spend 40 years learning the piano and still be unable to sight read the difficult pieces like Liszt's Hungarian Rhapsodies.

To me, you are comparing apples and oranges.

Something to consider:

When people compare sight-reading to reading language, I think they fail to realize that most of our written language - newspapers, bestselling paperbacks, and magazines - are written at about a 5th grade level.

But the music they want to sight-read is rarely at a 5th grade level. (Which would be the rough equivalent of simpler intermediate level literature - Clementi sonatinas, etc...)

The literary equivalent of a Beethoven sonata would be more like Umberto Eco. I don't know very many people who can easily skim through a copy of "Foucault's Pendulum." Or what about something like Goethe's Faust? I know a lot of very educated people who would have a hard time getting into Faust because of the unfamiliarity with the style and subject matter.

If you want to sight-read music with the same ease as reading through Time magazine, then you have to be fair and pick music that uses a very common vocabulary and very familiar style.

We also see people occasionally fault our notation system for not making music accessible to the reader. But do we fault the English language for the fact that many of us cannot read technical manuals or research papers?

The English language is not to be faulted for the fact that I cannot understand this:

http://repositories.cdlib.org/cgi/viewcontent.cgi?article=1022&context=lrc

I would just need a lot of training and experience in linguistics.

Good try Kreisler, but it doesn't wash. I think it's more reading Dante on a unicycle.

As above users have mentioned, compared to reading text, sightreading involves more components of cognitive function.

Reading text requires the following steps:

1. Processing a group of letters (26 for English) - and occasionally other characters - encoding them as a word, and assigning them a definition as referenced from memory (and perhaps a more accurate meaning based on local context)
2. Repeating for the entire phrase or sentence
3. Attaching a meaning to the phrase or sentence based on context (immediate and past) and the interpretation thereof
4. Retaining the image/impression as the process repeats for the next phrase or sentence

Sightreading requires the following steps:

1. Processing a vertical line of notes (or rests) - a musical word, so to speak - according to the position of each note (88, 87, etc. possibilities for the respective 1st, 2nd, etc. notes) - here, key signature must be taken into account as a template for transposition
2a. Storing the duration of each note in memory
2b. Assigning each note to a corresponding finger
2c. (optional) Employing dynamics and other elements
3. Coordinating both hands for striking the appropriate notes
4. Sustaining/releasing notes with the appropriate timings

One of the major differences is in the words. Using English as an example, most people's vocabularies number in the tens of thousands. On the other hand, for piano, with up to five keys per hand, an octave range located anywhere on the piano will yield 2379 possible combinations. However, many of the combinations for 4 and 5 keys are heavily dissonant and/or physically implausible, so there are perhaps 1000 possibilities one might encounter in a given vertical line of music. With 76 unique octave ranges, either hand can play 63 octaves (on the left for LH and right for RH) whilst allowing room for the other hand to play without causing a need to cross (for simplicity, we're ignoring the fact that we can cross hands, not to mention that the octave ranges can often intersect or coincide). Thus, with each hand having 63000 possibilities, there are roughly 4 BILLION possibilities for any given vertical line of music.

With the above having been established, we can see difference between reading text and sightreading. In the case of the former, to obtain the definition of a word upon processing it, we use our memory. However, in the case of the latter, as we cannot possibly remember the sounds of every musical word (or, provided the skill, synthesize them mentally at a sufficient rate) we instead derive that "definition" from the sound emitted by the piano, which acts as a sort of translator. Because of this relationship, we must interact with the piano, but as we also cannot possibly remember the exact finger positions (and in fact, due to constraints such as timing, legato, slurs, etc., there technically aren't any exact positions) we have to rely on the coordination process outlined above.

Thus, a more fitting analogy would be that reading text is like reading text, whereas sightreading is like reading text, typing it with a keyboard with many more keys, inputting certain groups of letters (words) simultaneously rather than sequentially (often requiring both hands), and holding each note for variable lengths of time - all at once! Quite clearly, compared to reading text, sightreading is an exceedingly complex task. We can always improve how fast we are it, though the speed vs. time curve gets increasingly steep as we approach thresholds for coordination and short-term memory.

Wow Symphony you're so smart! I'm in awe. I bet you sight read well. Stick around, we could do with some extra brains round here.

Nah, I've just done a lot of thinking about the stuff covered in this thread. And my sightreading is nothing spectacular. I was absolutely amazed by the video beet posted. I can only imagine how much more enjoyable piano is when you can basically play pieces from the get-go. Haha...it's almost mouthwatering thinking about sightreading something like an etude.

I've just been playing through some Purcell. I'd almost forgotten the pleasure of playing for pleasure!

Not trying to wash anything, just making some observations.

Bach said, or so it is said, that one ought to be able to play perfectly any score from the first time.

Well, since no one in the history of piano could do this, I guess Bach was wrong.

I have a couple, and they've been touched on here already.

One is that the number of visual units the brain has to be able to identify, analogous to the number of characters in Chinese or the number of standard letter groupings in English, is just so much larger here. As Symphony above has demonstrated.

On the other hand, the brain is really good at recognizing visual patterns. Just think of all the patterns it has to recognize, sometimes from partial data. ("That's a tree. That's a car. That's Bob.")

So I like the second answer a little more: with playing the piano, there's this physical component you just don't have with natural language. You're not just sight-reading; you're sight-playing. And just as hard drive access is slower than internal computer memory, so brain-to-hand signal processing must be slower, and more complicated, than pure visual processing. As was mentioned before: If it takes months to build up the technique behind an etude, why do I expect to sight-read it at speed?

Thx for the discussion!

Thank you for your kind words. I was just wondering if doing math while playing would strengthen my playing, in that I was multitasking with the same part of the brain. I does give me a headache, though!

My 7 cents:
Our brain is not a single giant CPU that processes all inputs similarly and it does not have a specific hard-wired music processing center as is the case for speech for example. For one, it has specialized regions that process specific inputs received via our sensory organs, eg. an olfactory lobe to process smells, an occipital lobe to process visual stimuli and a temporal lobe to process auditory signals. While all stimuli are eventually reduced to electrical impulses, they are processed in different circuitry and are fed to different networks that determine their fate i.e. our reaction to those specific inputs.
Our speech areas for example are usually located in one hemisphere (the left for most) and consist, simplistically, of an area that produces speech (expressive) and an area that processes speech (receptive). The two areas talk to each other and to many parts in the brain such as the memory area and motor areas that control movement of the tongue, jaw etc. Text reading and comprehension are actually processed similarly with the exception of the input originating from the occipital lobe (vision). At the same time, since these circuits are "built-in", text decoding happens via the speech reception area almost simultaneously. Input from the memory area remind us of the meaning of words and provide context etc.. Now music processing, as science stands today, does not have a specialized "in-built" region for it. At least, not in most of us. You have to build these networks for it as you experience music, which is why the plastic brain of a child is best suited for this activity. Of course the adult brain is also capable of plasticity and can set up networks for music processing. But "music" is multiparametric: in addition to individual note names, pitch, time, rhythm, phrasing, etc, you have to add motor control of individual muscles, some of which are not part of the routine movement repertoire of daily life. Thus the task of processing it is quite complex. If you add some genetic traits that allow us facility, or lack thereof, in processing specific parameters, eg. pitch, auditory memory, "dexterity" (???talent??), you can easily imagine why something like sight reading can be hard and can exhibit significant inter-subject variability.

I’ve got much better ways of giving myself a head-ache than that!

Why is sight reading so difficult?
I personally (and I am not alone) find simply reading a score very difficult. Comparisons with Chinese and other hyrogryphical languages are to some extent appropriate, BUT the musical score does not convey any sort of linguistic meaning, instead a different meaning of pitch, harmonics, rhythm and meter etc. In Wombat’s world at least, apart from the musical score there is no other language that conveys these meanings, and it is a far harder language for a relative new comer such as myself to learn than any specialised jargon written in the English alphabet, as posted by Kreisler, or that encountered by the reader of any medical journal, since it is not a new linguistic meaning (as in jargon), but whole new concept of musical meaning that the score conveys, written in an entirely different alphabet (2 staves etc).
Apart from the difficulties inherent in reading the score, of recognizing the contents of the printed page and assigning appropriate musical meaning to it, the process gains a second level of difficulty with the second part of sight reading: the motor task of playing the piano.
One can read a description of a chaotic battle scene on a beach in Normandy 65 years ago without actually having to run around the beach, chucking hand grenades and shooting your rifle at the Germans ahead who are trying to kill you.
Unfortunately when trying to sight read a Beethoven Sonata those motor instructions coded in the largely unintelligible score (unless allowed 30 seconds a bar to decode) have to be followed to the letter.
Comparisons with battle scenes, chaos and carnage are sadly, for me at least for the immediately foreseeable future, entirely appropriate.

Reading a score is reading a type of graph, not reading a language. It may be interesting to compare them, but working from an assumption that they have a lot in common may be a mistake; they may actually have very little in common.

That's a very good point. English is an alphabetic and (somewhat) phonetic script. Music is more of an ideographic system and, as such, is processed very differently in the brain.

I've just started learning a bit about linguistics, and the research being done is rather fascinating!

Yet it seems more and more that they have much in common. Music and language both create a story within a framework of patterns on a canvas of time, using rhythms, pauses and inflections. The meaning of a sentence is held not only in the meaning of words, but the rise and fall of the voice, how long a sound is held, whether and where there are pauses. It is this "musical" aspect that makes foreign accents so hard to understand. One can solve difficulties in language by drawing on music, and one can probably solve some music difficulties by drawing on language. A "knock knock" joke is easy to follow because of the expected form that unfolds over time, and music that moves to a cadence and goes ABA likewise.

When we read music are we really dealing with hundreds of notes and combinations? Are we not dealing with smaller sets of patterns and lots of things that repeat? If you have simple music like a Clementi with an Alberti bass, that bass is so predictable that you could probably play most of it without even looking at the notes. I don't think that we do read every single note. Is it possible that music is more predictable than written words and that we can use that?


This is true. Written language is not graph-like. That might give written music an advantage over written words when we care about how the words sound.

In regards to written words, how about the enunciation and inflection of the stage actor? Do we get closer to the performance of a musician in this?

While it's true that musical notation is more visually oriented than writing, surely that's not the explanation for the relative difficulty in sight-reading: the brain is very good at visual pattern recognition. Think how quickly you can identify a person from their face.

I was thinking of another possible difference that might be relevant: redundancy. Language has a lot of redundancy built into it, so you can get the meaning of a message without parsing every letter. Ths s hw y cn ndrstnd ths sntnc wtht vwls. And think about what a pain it is to specify a letter to someone over the phone ("That's 'M' as in 'Mary'..."). That's because redundancy isn't baked into the names of letters. But we don't have that problem in general, because everyday language evolved with redundancies.

I think this notion of redundancy is largely absent in music, and that could be a reason why sight-reading is harder than language-reading. Imagine every third note were missing in a piece of music. Could you reconstruct it? Maybe in some cases; maybe with some Mozart. But in general I think it would be impossible. "Every note counts," more so than in written language.

I haven't thought through this, but I'm wondering if this isn't an important difference....

Just thinking with no answers:

But is it relatively difficult? Or is it how it's approached?
I'm trying to think that through. Do you have any experience in comparing your own reading ability in languages that have the same script as opposed to different scripts? Among several foreign languages, if they use Latin letters my reading is relatively fluent. If the script is Cyrillic I'm reading more slowly than in sight reading music.

If you put in the same amount of time practising notation as you did reading and writing letters, would you be as fluid? What if it were reversed.

And speaking of reading and writing, what's up with that? We don't learn to hand-write music or write it at all, but when we learn to read words, it's read and write and grammar and spelling. What if they left out "write", "grammar" and "spelling". What is the musical equivalent of grammar and spelling, and would reading become easier if we had this?

It is indeed something to think about.

That's actually not true in sight reading. The most important thing in sight reading is maintaining the rhythm and that is often done even if it means leaving out a bunch of notes.

Also, unless one is playing some new-fangled piece of contemporary classical music, most tonal music follows certain patterns, chord progressions (e.g. ii, V, I) such that there is a huge amount of built-in redundancy. A seasoned sight reader need not read every note. He fills in by using his knowledge of harmony, voice-leading, and scales.

No use comparing music to language.. when the brain does not have a regionalized "home" for music processing, while it is absolutely designed to handle language and dedicates roughly a third of its cortical volume to a specialized speech region.. and this is not including the connecting fibers.

... processes both in such similar fashions. What if?

The functional anatomy of the brain is not exactly a complete mystery..Functional studies definitely show different processing of language and music.. No question about it. And the speech areas were defined in the late 1800s.
The speech production area in fact is named after the man who described it:Paul Broca. (His own brain can be found in a jar on the shelves of the Musee de L'Homme in Paris.)

Whatever studies may show, when we actually work with both music and language ourselves, we use many of the same processes and it appears to be the same areas. I have set these out in a previous post. I am more interested in what makes me succeed by doing certain things, than what anatomical mapping might indicate. However, I'm willing to explore the notion to see what this is about. Do you have a link?

Also does what you are saying preclude the idea of using the attributes of language and music that cross over into each other's area in order to use them to enhance abilities in either?

I don't agree that music necessarily creates a "story". There is some music that does have strong narrative and/or rhetorical qualities, but those qualities are not at all required for music to be music. I find much of Bach's WTC to be fairly devoid of any story-telling qualities, for example.

But anyway, I wasn't talking about a comparison of music as it sounds relative to spoken language, but about scores relative to writing.

Wr, bear with me for a moment. Actually it's times like these that this medium seems so limited for doing more than scratching the surface of anything.

First, I meant "story" in a rather abstract sense. The formal structure of most music tells a kind of "story" - It moves from some place and back toward some place where it concludes, which typically we hear as a cadence. There are motifs, sentences and larger structures of form. The way these are arranged tell stories, and stories within stories. The similarity with language is that we are hearing something unfold over a duration of time, and we are listening within expected frameworks. In language we expect something to be introduced, developed, and concluded: music has the same. In language we have upward inflections (= question), pauses, groupings of sound (k-a-t) which create a unit and the same is true in music.


I did agree with what you wrote about the graph-like quality of music. That is indeed very different, and for directing the production of sound, written music is superior to written words as well as different.

However, when I look at what happens in reading word and reading music, at least for me, some deeper things emerge. Both written words and written music are an encoding/decoding of the audible phenomenon that I have described above. I don't read either without being aware of these sound patterns, or that these patterns lead somewhere ("story" in the abstract sense). On that level, the process of reading music feels similar to the process of reading words. I cannot divorce what words or music are about in the sound-world that they represent.

At this point I don't know if other people read music the same way that I do. On some level, as you read and play music, are you hearing and anticipating what you will hear? Can you feel a cadence coming up as well as seeing it, or hear and see melodic patterns or chord patterns developing? You see, most of what I have been reading describes a mechanical process of moving fingers about on keys according to symbols on a page. It is described in overwhelming terms which I'm sure you've read. I've been reading these descriptions and thinking that I don't process written music like that: it's a small part of it. It's the other things that go into my reading music that make it a whole lot less daunting, and a lot those seem to involve similar things to processing language (for me). Ultimately that leads me to wonder whether the way we approach reading music causes it to be easier or harder.

I ask the same thing about language since I learn and play music, and work in languages and am also learning another one. Here too it seems as though approach or mindset plays a large role. When I come to either, then I find aspects of music make the language task easier, and aspects of language make the music task either. That, in turn, makes me look at the connection rather than dissimilarities between language in music - including decoding and encoding the symbols of what is an oral, meaningful process.

This has probably become one long gobbledygook in the telling.

I realized in the middle of writing that I don't know how to. I'm frustrated knowing that it's come out as nonsense. I have had the feeling that something is missing as the posts on reading music piled up, and therefore it's probably less complicated. I've also had an increasing feeling about the link between language and music (regardless of what neurologists might find). But I can't seem to know how to write it and feel a bit foolish for trying. (sigh)

I have got to read this entire thing!

ks,

I read your longer message before I read the short one that acted as a disclaimer.

When you used the term "reading" (as applied to music), I wasn't sure whether you meant score-reading generally or sight-reading. And that made me think of another question that's perhaps relevant: considering the similarities between reading music (in any sense) and reading text, what role does familiarity play (as in foreknowledge of structures, patterns, etc.)? Is the sense of anticipation and "expected frameworks" different when we read music or language prima vista?

Steven

"If you have simple music like a Clementi with an Alberti bass, that bass is so predictable that you could probably play most of it without even looking at the notes. I don't think that we do read every single note. Is it possible that music is more predictable than written words and that we can use that?"

The above from keystring resonates with me, specificaly (I don't think we read every single note.) As one gains in music reading ability, some chunks or 'words' become recognizable and help speed the reading process, just as with text.

Just some thoughts that the discussion sparks in my mind.

James

Isn't there an anecdote that Grieg brought his concerto to Liszt, who then sight-read it at speed on the spot while giving a simultaneous running commentary? I've also read that Saint-Saens was among one of the best sight-readers of his generation.

In any case, if Bach said what he said about being able to read a score at sight (notwithstanding that baroque music maybe isn't as "vertically" saturated as later music), maybe he was able to? I doubt someone like Bach would talk out of his ass like that. Pardon the crudeness.

Anyhow, that Liszt story seems to be widely well known. Maybe there's some truth to it, after all.

Steven, sometimes (more and more) I wonder if I know anything and that is on the rise, and also what is normal. I've worked on my first Bach Inventions. You know that he wrote these in order to teach his sons the components or patterns of compositions. The structures are so in your face that the problem is to linger on the notes long enough to not be caught by some surprise. You can almost anticipate whole sections and play them without having read them, for example when LH and RH trade statements, or something is the mirror image of what was going on before. Now, if you recognize these patterns, prima vista reading will be a much different experience than if you are following the notes as seems to be described here. I've been thinking about this.


That's exactly the lines I was thinking along. Anticipation can also be a trap because we can read what we expect to be there. When I proofread a translation, I walk around the house reading a printout out loud with my finger running along the words. If you read "He saw dog" you might think you see the word "the" when it's not there. In music I might expect it to go a certain direction and always see the notes doing that even though they are doing something else.

I'll leave it at that to keep it relatively brief.

I also remember vaguely that one or two teachers said that there are things we learn outside of reading practice, that then go into our reading practice.

Assuming Liszt could sight read the Grieg Concerto, that's hardly difficult compared to numerous other works. If Liszt (or anyone)could play everything perfectly at sight as Bach suggested, then he wouldn't have had to practice.

Also, even if Liszt could play anything perfectly at sight, one would be talking about one of the greatest musicians of all time. Surely Bach didn't think everyone, even professional musicians, should be able to perform at that level.

As I understand, sight reading is for getting at the notes, and practising is for developing interpretation and applying technique to bring the music to performance level. So someone who can sight read flawlessly also practising doesn't seem contradictory.

Isn't the consensus that Liszt was one of the greatest of all time? Also there is that argument that one can't do two things at once. Surely playing the Grieg while giving a running commentary at the same time would have been extremely difficult, but possible if it was actually done. Second, I've seen video footage of Glenn Gould giving running commentary while playing Bach flawlessly. Now, he's not sight reading but my point remains that it is possible to do two things at once. Whether or not it's possible to do by most people I really don't know; Liszt and Gould were towering musicians. One can only wonder what kinds of connections their brains were able to make.

There's also a less popular and less founded rumor that Liszt sight-read Islamey. It is a very difficult work, physically, but I've found that it contains a lot of easily recognizable and very pianistic patterns. Back to the Grieg-- it is a challenging work, and how 'difficult' it is is subjective. Surely most people think it is a challenging piece, and I doubt many of us would be able to sight-read the whole work flawlessly the first time as Liszt purportedly did.

There is also that little rumor that Liszt had one of the best, if not the best, piano techniques of the 19th century. Surely this, plus his sight-reading ability probably meant that he indeed could play anything written up to that time without practicing it. Whether he would have been able to sight-read something as strangely notated and unorthodox like Sorabji or Ligeti would be an invalid proposal, as nothing remotely resembling that had been written at that time. But he's not here, we can't just ask. We can only speculate. My point is that if there were well-attested and well-documented anecdotes of what Liszt could do (and there are many stories), by respectable musicians and knowledgeable people of the time, then I don't doubt that he did what he did.

Maybe it IS just a matter of practice, as there are also stories of him practicing 8-12 hours a day.

But who knows, I guess I'll only ever be sure if I quit my day job and emulate him.

I find this entire discussion extremely interesting. One of the things that is helping me with sight reading is following the proceedure outlined by Geiseking and Leimer in their book "Piano Technique." They emphasize the necessity of training by "means of reflection (systematic logical thinking."

What is particularly interesting is that when they approach a composition they actually verbalize what is going on in the music. For example, in analyzing what is going on in a compostion---in this case, Sonata in F minor of Beethoven,they say "the sonata opens with the broken chord of F minor, from middle C to A flat above the treble Clef, followed by the notes of a turn on F in the second masure." Measure by measure they make this verbal connection between what is going on in the music and the names of the notes and chords involved.

Which leads me to believe that there is a strong connection between the notes, chords, phrases on the page of music, and the verbal centers of the brain, just as there is between letters, words,phrases and stentences in language.

Their method may not work for everyone, but it is certainly helping me. Gaby Tu

First, if we grant that sight-reading has something to do with reading a language, we might notice that it is very hard to read anything in a language one doesn't know, even if one memorizes the letters, or characters. My strong impression is that great sight readers make a musical interpretation of the score well in advance of playing the notes. The best sight reader I know is the accompanist of a chorus and even on a challenging score like Mahler's 8th Symphony, he has an amazing facility for making his way through the orchestral reduction at sight. In addition, he can mix in whatever bits of the 8 vocal parts he or the conductor feels will help the singers the most. It seems clear that he is not just reading notes, but is reproducing a musical idea based on the notes.
One rehearsal, when I marvelled at this, he shrugged and said, "Well, I've been accompanying choruses since I was 11." He's now in his mid 40s.
Be that as it may, there is some evidence that reading music and reading language don't happen in the same bit of the brain or the same way. Rocket88's troubles with doing math while sight-reading and the widely observed ability of good pianists to chat with bystanders while playing - even while reading - non-simple music, would seem to indicate that the music and math regions are related, so that one interferes with the other and that the language and music parts are not so that both can operate simultaneously. This would be supported by the studies cited in Musicophilia on people who have lost their ability to speak, but not their ability to sing....
I think learning to sight-read is hard because we didn't learn it when we were kids. And like everything else in adult life it involves a race to develop competence before arthritis or worse sets in.
Still, it is about as fun a thing to work on as there is. Bach, Haydn, Mozart, Purcell - there is a whole literature out there that is wonderfully rewarding even at my customary Allargandissimo.

It (mostly) made sense to me, although I don't know if I understood exactly everything you were saying, and that could be my fault rather than your writing. And I appreciate it when you and others try to write things that are difficult or elusive; they can be the most interesting posts. I often post something and immediately wonder if it makes any sense to anyone. Sometimes I chalk it up to starting out with a language other than English when I was a kid (but that's a poor excuse, since that early non-English phase didn't last long).

Okay - I was probably interpreting how you used "story" too literally. I understand and agree with what you are saying about these similarities, but think it is easy to get so caught up with the similarities with language that they begin to count for too much. For me, the connection to spoken language is fascinating, but the connections to other parts of our experience are also important.

One of the most obvious connections that music has outside of the one to spoken language is the connection to dance, gesture, and kinetic sensations. I am guessing that I probably make those kind of connections in some cases where you are making language connections.

Music also can mimic or allude to various things that are out in our physical environment, from the movement of a fish, to the noises chickens make, to birdsong, to light on water, to wind blowing through gravestones, and much besides.

The way a piece is put together may resemble the way a story is put together in language. Of course, the convention of talking about some kinds of music as being made of words, sentences, and paragraphs, together with the idea of musical "argument", all reinforce the musical connection to spoken language. On the other hand, there are also ideas about music resembling architecture or mathematical logic, too. Or natural phenomena and processes or even other man-made shapes.



This is interesting. For me, reading words and music have only vague similarities, but really drastic differences. Words seem to be primarily concerned with pragmatic meaning, even though they can be used for artistic expression. Music has no obvious pragmatic meaning at all.

I read words with their sound being completely contained in my imagination and with no need for them to be actually made audible (unless I'm really unsure of the pronunciation), but I have nothing like that kind of highly detailed internal representation of sound for reading music, nor can I extract much of the "meaning". I do have some sense of how music might sound simply by looking at the score, but it is vague, and I can't trust it much - I have to hear it as I play it for it to really become "music".



No, it's not gobbledygook. And sure, I generally am aware of the various musical patterns when I am reading music (if it's not in some unfamiliar idiom) and that makes all the difference in ease of reading it. I am not convinced of a strong connection to how language works, though, because that sort of pattern recognition is part of expertise in almost anything you can name.

Joining the hoary debate(we’ve got mid-winter on the other side of the globe) ... knowing all too well that the subject has grown whiskers ...

Any newcomer to playing the piano (once they’ve grasped the fundamentals) and discovered that keyboard sight-reading is not on a par with learning how to read a book ... and face the dreary reality of a lengthy internship (say 3 years) before coming to grips with the beginnings of quality music ... and then at the cost of infinite practice before achieving a modestly quality rendition ... will naturally ask the question ... WHY?

And all the traditionalists will carp the usual myopic drivel, always based on their own lengthy service before the mast ... about paying one’s dues and working hard, etc ... but not for a moment cottoning on to the crux of the problem ... namely the INSCRUTABILITY of the antiquated system of natation which we have inherited.

The analogy with literature (books) is valid ... from first beginnings we move from letters of the alphabet to register complete words and slowly build up a vocabulary ... but the fact is that we learn to take a mental snapshot of the word ... the word can be visualized (how often we look at our spelling of a word and say "that doesn’t look right" ... and then correct to match what we’ve pigeon-holed in our brain-box) ... and so we constantly expand our vocabulary.

But not keyboard music ... WHY?

Simply because, from viewing the score, we are not able to instantly envisage total note patterns (equivalent to words) ... and so we never build an ongoing vocabulary, and are totally subject to the labourious process of

1. Identifying the notes
2. Establishing the best fingering
3. Practising like mad
4. Gaining the support of aural and muscle memory
5. Eventually realising a reasonable rendition by memorizing

And yet with literature... from early days of learning ... "the cat sat on the mat" ... we graduate, to delight in freely reading Shakespeare, John Keats or the local rag without needing to change gears.

The fault lies in the antiquated notation which needs upgrading to meet our Computer lifestyle ... no scientist would accept a flawed notation format which does not visually portray the exact pitch and duration of notes (x/y axis) ... and thereby lead to linking notes into patterns and then sentences ... to portray the breathtaking SHAPE OF MUSIC.

At this end we’ve now got clear blue winter skies .... the mercury has risen pleasnantly from 2-17C.

Hi, btb!

You are surprisingly late to the party, especially considering your perennial interest in this topic.

Though I know it's impossible to convince you, we "traditionalists" defend our "antiquated system of notation" precisely because we do find in it what you do not, viz. a "notation format which ... portray[s] the exact pitch and duration of notes (x/y axis) ... and thereby lead[s] to linking notes into patterns and then sentences ... to portray the breathtaking SHAPE OF MUSIC."

Whether or not such perception comes from "lengthy service before the mast," we find the present system transparent rather than inscrutable—and that's the crux of the problem for any proposed alternative: what ain't broke doesn't need fixing. Alternative notation is interesting, but I've yet to be persuaded that any given alternate system is more accurate, palpable, efficient or flexible than the traditional one.

Hey, I'm still awaiting my award of virtual doubloons for recognizing your MIDI graph of the beginning of Beethoven's Op. 73! (Or perhaps you hadn't noticed because I was such a latecomer to that thread. )

Steven

Sight reading is hardest on pieces that confuse the key that its written in. Sometimes the extreme difficulty of the technical aspects can also make sight reading difficult but if one knows their key signatures and recognizes chords the sight reading becomes much easier.

Do you know people who can actually read Shakespeare and Keats? It's not the medium, it's the messenger.

LJC is correct. Sight reading depends largely on how quickly one knows the key signatures, and recognizes the chords in their various inversions. For example, once a person recognizes that the cord is a G minor, or a D 7, one's fingers go automatically to the correct piano keys. Also, a knowledge of chord progressions and harmony is essential. Gaby tu

Lots of practice will help. Before I started to work on classical piano again and take classes at the college, my sight reading was horrendous. But I just slogged through more and more challenging pieces...did a little sight reading practice, but mainly just reading a lot. So now, a couple of years later, I picked up some blues/jazz stuff I set aside years ago (since I don't have a decent piano right now, it's hard to play classical), and things that were hard to read are so easy now!

Assuming it is written in a key at all...

I've only briefly read through this thread, but you have nailed it.

What is the mystery of sight reading? To me it is all in a day's work. Why is that such a problem?

As a church organist, I get hit with this all the time. I don't consider myself any genius, but after a while one learns to recognize repeating patterns: the organ parts in a Stanford anthem are often found in the later music of Howells. Differences by degree, that is all.

You are indeed a savage klutz if you don’t delight in reading Shakespeare (scrub round the messenger bit) ... these days, even our South African indigenous peoples pass matric with a reading of Macbeth, Hamlet, or Romeo and Juliet ... in recently returning a bad (sticking) DVD disc of Julius Caesar (dim me hadn’t checked a clean surface), I was amazed to discover that the young black salesman had completed his school days cum big Julie ... and could quote the last words of Caesar "et tu Brute" (after being stabbed in the Senate by his closest and most trusted friend Marcus Brutus..)

Your virtual doubloons are in the post sv.

Imagine a traditionalist being able to recognize from a MIDI diagram the opening of Beethoven’s Emperor ... must treat you with more respect.

Brave words to your second para ... sadly without proof ... the x/y axis format demands a visually accurate representation of the two directions ... traditional notation is framed on an
alphabetic 7-note stave for pitch and antiquated neumes for note duration (with the format of the SHORTEST notes carrying MOST flags ... absurd) .. It’s only when you make a truly accurate MIDI diagram of keyboard music that you will be able (like Alice) to enter Wonderland ... but traditionalists must stick with what they’ve got ... and be content to plod on (like fan) "with lengthy service before the mast"... for my part, I want to play everything prima vista.

Here’s my Grand Stave (space for all notes) ... it’s just a matter of "joining the dots" to produce a musical tapestry ... and see the SHAPE OF MUSIC ... thanks for the chat.

Read as in read.

I don't see the point in trying to play a piece right away from the score. Could anyone play anything other than beginner or intermediate pieces on an acceptable level right away?

Even if you immediately recognize what has to be played...isn't it still physically impossible to play a lot of music right away? Doesn't most (advanced) piano music simply have to be practiced a lot until it's imprinted in 'muscle memory'?

This whole sight-reading thing just boggles my mind.

When someone says he never has much trouble sight-reading, but finds memorizing much harder, I get confused.
I'm a slow reader and memorizing is for me basicly the essence of piano playing. I couldn't improve on my playing if I still had to look at the score all the time.

I have that exact problem. I'm quite a fast learner, (grade 5 in a year and a half), but my sight reading just hasn't kept up! And also, if I'm given a piece of music, and told to sightread it, I can usually do it, but if I'm given a piece of music specifically for sight reading, I'm rubbish!!! I think its down to our brains just thinking, sight-read, sight-read, sight-read, whereas if we don't think about it, it works. I also agree with some people who have said about recognizing chords - that helps a lot.
I don't have a problem recognizing the key signatures though. I just look at how many sharps/flats there are, play the scale, and I can do it from there.

That dude on youtube is amazing!

It could be for entertainment, for the ability to explore unfamiliar music at the keyboard, or for sharpening and maintaining your reading skills … to name but a few reasons.


Probably not, but that's the nature of the distinction between what one can sight-read and what one can play after weeks of study. If we could sight-read material of the same difficulty that we can eventually learn to play, there would be little need to practice.


It seems that you understand how it works but not the point of sight-reading or its benefits.


If you improve your sight-reading skills, your ability to read will improve, too (as in the context of learning new pieces). Slow readers generally do rely on memorization; I read fluently and must rely on the score instead (unless or until I can play something from memory, which is a challenge).

It would be great to be equally adept at reading and memorization, and obviously many people are. Others may always be better at (and/or more comfortable with) one than the other.

Steven

Thanks. It makes more sense now.

I couldn't play anything if I couldn't memorize... unless it's an absolute beginner level piece or a very simple passage.
Fortunately I don't have to make an extra effort to remember things, it goes hand in hand with practice. I divide a piece in sections and by the time I can play a section correctly, it's memorized -> stored in musicle memory.

You say you must rely on the score. So even if you've studied hard on an advanced piece, you still need the score. In that case I assume it's only to keep track of where you are... because surely you are actually playing a lot of it from 'muscle memory', especially the harder passages that required a lot of practice.

That's exactly right, except I wouldn't say I need the score only to keep track of where I am. Much of the time I only need to give fleeting glances at the score, but sometimes I do need to look more closely to play the correct notes.

I've always been lazy about memorizing because there's never been a compelling need to do it (other than the occasional frustration of having to turn a page at an awkward spot). I am taking memorization much more seriously at present, though, primarily because of advancing age. Many pianists say that their ability to memorize declines as they approach their senior years, and I want to push myself to do it now while I'm able. Maybe it will help keep the grey matter intact and in shape, too.

Steven

Assuming it is written in a key at all...

Yes, or more than one key at the same time! (ex:Ligati 1st etude)
But how many recognize that there are 5 sharps (B Major) so that the pentatonic scale in the bass is easily notated? I also quickly note that there are no notes outside that pentatonic scale.

There is lots of wisdom here. I love contemplating this topic though I don't have the answers. I do think, as some suggest, there is a degree of similarity with reading, but not a perfect correlation. As others have pointed out, if you can digest the music in chunks, like a seasoned book reader would do with a text, you'll go faster. But there is musculature involved too, as others noted. Have you tried sightreading music with chords written in, like hymn books with the guitar chords printed above each measure? Not that those are always "correct", in that often, passing chords get assigned their own chord name and make things much more complicated than necessary...but I find it helps sooo much to have those; as you dive into each measure you already anticipate some of the notes you will encounter. Same for classical (Mozartish stuff)...there's enough "logic" in where the chords go, often, that you can do some predicting. But modern stuff, sometimes I literally go back to reading note by note, sometimes (if it's very dense with accidentals) literally spelling it out like a kindergartner on the first day. *sigh*

When I was in college, I did some lesson accompaniment work, and it was not unusual to have a teacher bring out something like a Mozart violin sonata (which are as much piano sonatas as for the violin) or a Brahms clarinet sonata during the instrumentalist's lesson. I'd have to sight read the piano part, not at performance level, to be sure, but well enough keep the music going. Although not at the level of a Liszt Transcendental Etude or Ravel's Gaspard, they were not exactly what you would call easy to play. And too, the piano typically is going to have many more notes and they are in more complex combinations than what the instrumentalist or singer has to deal with (this is so not fair ).

Also while in college, I was once called upon the day before the performance to fill in for a pianist who canceled out of doing the accompaniment for a chamber opera (the entire "orchestra" was piano and percussion). Basically I had time to sight-read through the score once, try to figure out a few of the trickiest moments, go through dress rehearsal, and then do the performance.

I once attended a summer chamber music school, and sight-reading was the first thing we did, which gave the faculty an idea of how to divide up the players into groups and choose repertoire. After that initial session, we still had weekly sight-reading sessions and all of the music was professional level stuff. And the public performances we gave were all of music we had only had to work on for a week or two.

Many orchestral keyboard players don't get their parts very far in advance, and certainly don't have time to get the music into muscle memory before having to play it. And some of those parts are pretty difficult.

So those are some examples of actually needing to be able to sight-read quite fluently. Besides, it's just plain old fun. And speaking of sight-reading fun, sight-reading music with other musicians purely for pleasure can be pretty wonderful. Piano duet sight-reading with someone who is really into it can be especially great (and even more so at parties).

Those who won't or are unable to make the effort to become proficient at sight reading don't realize all the musical fun they are missing.

Regards,

It is similar for me. I need the score in front of me, otherwise I am lost. However, my eye is almost always on the score; I only give fleeting glances at the keys. I do feel that this stands me in good stead if I try to sight-read.

I don't find memorisaton easy, and without a score I am lost. But I tend to feel that such time as I am able to spend at the piano is better spent on training the fingers than on memorisation.

If I have to sight read with other musicians, I find it terrifying.