What does 'reading music' real mean?

[ascolta]

As the title says; what does reading music really mean?
Is it the ability to simply read the notes? To be able to sight read well?

I think the word 'reading' throws me a little when used for music.
When we read words we say the word internally, you can't help but say it even when you quickly glance at a word. When you read a word the words meaning is understood, and its this understanding of 'meaning' that I view as 'reading'

Music is different and by nature is based on sound. So what does it mean to you to read music? Do you say the note names internally when reading? Do you hear the melody accurately or is it a vague representation of what is written?
Many musicians enjoy time away from their instrument just reading a music score. I'v read some well known pianists used to just read a score for a few weeks before ever attempting to play it at the piano. So again, what is the internal process going on reading or studying the music in this way?

I apologyse if the question is poorly constructed, its hard to put across what I'm trying to say in writing. Your views on the subject though will be really interesting!

[awesom_o]

To a skilled musician, music is akin to speech.

He or she can read the music of different composers with great fluency, understanding the subtleties of style that pertain to the work of each one.  With practice and good training, one can learn to read music from scores as easily as normal people can read aloud from a book. Once you are really good, you will be able to read music BETTER than normal people can read aloud.... you can learn to read music as well as a professional actor could read aloud, as in, with truly beautiful speech and inflection in the words.

A novice musician cannot read music with that level of fluency.

[Bob]

Set the music in front of you, without you having practiced it (or practiced it much).  Then you play through it.  Usually at as high a level of performance as you can.

[outin]

I think the word 'reading' throws me a little when used for music.
When we read words we say the word internally, you can't help but say it even when you quickly glance at a word.

Not necessarily when you are a fast and fluent reader. You do chunking and skipping a lot just as people do when sight reading.

[brogers70]

I really envy and admire those who can look at a score and hear it in their head, just as I can look at a poem and hear it in my head. I have had some ear training, and I'm a decent sight singer, but I am in awe of folks who can read a score in that way with easy fluency.

[iansinclair]

I think the term is really used in two different -- very different -- ways.  First, of course is "sight reading" -- generally taken to mean taking a score more or less without having ever seen it before, and being able to play it at a reasonable tempo with a reasonable approximation to the form -- melody and harmony -- even if without quite all the notes and nuances.

The second way is very different: I would use it to mean taking a musical score -- either for a solo instrument or for a group (very often, in my past career, a choir) and being able to hear what the music will sound like when you get the group (or instrument) rehearsed to the point of performance.  It is an art, but it is at least partly learnable... with a lot of practice.

[fleetfingers]

When referring to a "sight reading level", it's generally understood to mean that the pianist can indeed look at the music and know the meaning of it; and also, be able to play it. In that sense, playing music is more like reading a book aloud to a group than reading that same book in your head. If you can see the music on the page, understand its meaning, AND play it with fluent expression per the markings on the page, then you are reading it the way you'd read a book in front of a group.

If you take a young child, who is learning to read, and watch them struggle through a simple book with words like "cat" and "hat" . . . if they are sounding out the words themselves and eventually getting through the whole book, then you can say that the child is reading. But there is so much more for them to learn. And you wouldn't say that they are reading "fluently". I feel that music is much the same way, and I also think that reading music is not harder than reading words. The difference is that we put children in a classroom and expect them to read words for 30 hours a week and make them practice reading at home on top of that. I understand that reading words is the priority, and that's fine....just pointing out that music could be learned just as well if given as much attention.

As for calling it reading, I think it makes sense because reading is just using symbols to communicate ideas in written form. Notes on a page are the same thing: a person with an idea wanted to record it and share it with others. If you can understand the idea (the music they composed) and reproduce it, then I think it makes sense to say that you are reading it.

[ajspiano]

I think the word 'reading' throws me a little when used for music.
When we read words we say the word internally, you can't help but say it even when you quickly glance at a word. When you read a word the words meaning is understood, and its this understanding of 'meaning' that I view as 'reading'

Music is different and by nature is based on sound. So what does it mean to you to read music? Do you say the note names internally when reading? Do you hear the melody accurately or is it a vague representation of what is written?

Music is different not because it is based on sound, all language is based on sound. We have a desire to communicate something, and we express that to each other through sound. Sounds which we have as a society collectively agreed upon a meaning for.

Written language only exists as a way to communicate when we are not able to speak. It allows us to say what we wish to say, without actually having to directly tell someone out loud.

The difference with music is that the meanings that go with the sound are not logical, and we are not able to associate words to them. The meaning is instead very generalized and can be interpreted in different ways by different people. Also, since it typically speaks on a very emotional plain, words are generally terrible at saying what music says, just as words are terrible at really explaining what we feel like.

Just as written language is a sound, that you hear and understand immediately in your mind, so is music notation. If you can truly read music, you can hear every single note accurately in your mind. It is not vague, it is exact.

Not only that, but for me, reading music is a secondary step to speaking it (through the instrument of course) in that you must be able to think ideas and say them at your instrument, and when you hear music, you must be able to respond with music in a way that makes musical sense.

When you have these skills you are able to start making more sense of simple written music. And, in the same way that reading books expands our vocabulary, allowing us to draw more meaning from our spoken languages both in conversation or when reading, so to does reading new pieces of music. Which leads us to the notion that it would be unreasonable to expect to comprehend and understand music similarly to the way you do English until you have spoken as much music as you have English - but very few of us will ever do that.

[outin]

Which leads us to the notion that it would be unreasonable to expect to comprehend and understand music similarly to the way you do English until you have spoken as much music as you have English - but very few of us will ever do that.

It seems you just forgot the estimately 90% of people who do not speak English at all or only little 

[awesom_o]

I really envy and admire those who can look at a score and hear it in their head, just as I can look at a poem and hear it in my head. I have had some ear training, and I'm a decent sight singer, but I am in awe of folks who can read a score in that way with easy fluency.

This is what being musical is really about!

[nyiregyhazi]

Not necessarily when you are a fast and fluent reader. You do chunking and skipping a lot just as people do when sight reading.

It depends what you mean by "say". When I see a film with subtitles, it's impossible for me not to process the words in my head if I glance anywhere near them. I don't "say" them internally in the sense that I would if trying to slowly memorise a sentence. However, i can't fail to process the words in full if I even glance. I suspect that was the point being made.

Although I'm generally a good reader , I don't have the same issue so consistently with sheet music. Some things I can and see and just process without trying, but I have to be a lot more conscious not to make mistakes. Due to the level of complexity in piano scores, I doubt whether anyone really reads quite as fluently as is possible with words.

[nyiregyhazi]

I think the term is really used in two different -- very different -- ways.  First, of course is "sight reading" -- generally taken to mean taking a score more or less without having ever seen it before, and being able to play it at a reasonable tempo with a reasonable approximation to the form -- melody and harmony -- even if without quite all the notes and nuances.

The second way is very different: I would use it to mean taking a musical score -- either for a solo instrument or for a group (very often, in my past career, a choir) and being able to hear what the music will sound like when you get the group (or instrument) rehearsed to the point of performance.  It is an art, but it is at least partly learnable... with a lot of practice.

Is that really what people normally mean by "reading"? Obviously it's a good skill to have, but I think people normally use alternative terminology if that's what they mean to imply.

[j_menz]

It seems to me that reading music is similar to reading language. It means having the ability to translate a score (or a text) into its language equivalent. With words, that means translating the written word into the spoken word and recognising what that spoken word means.

For example, I am familiar with both the Greek and Cyrillic alphabets, but speak neither Greek nor Russian. I can only read something written in these scripts when I recognise the word through its English equivalency, though I may be able to make a stab at pronouncing what it would sound like without understanding.

With music, reading means being able to translate a written score into the sounds it represents.  Whether that is done by playing it, singing it or internally hearing it.  The latter is, I think, a rarer ability and I'm not sure anyone has it who has not also the ability to play/sing it. In any case, though, it is reproducing the musical sound - including the rhythm, not just (in the case of piano as an example) being able to hit the corresponding key(s).

[outin]

It depends what you mean by "say". When I see a film with subtitles, it's impossible for me not to process the words in my head if I glance anywhere near them. I don't "say" them internally in the sense that I would if trying to slowly memorise a sentence. However, i can't fail to process the words in full if I even glance.

Speed reading is based on the ability to not do that for every word. Concentrating on individual words will often hinder full understanding. For me it's rather easy to avoid that when I am really interested in what I am reading, otherwise it's really hard and needs practice.

I feel there's at least some analogy for music reading. There are things in the score you don't need to process at all (many rests for example). I assume when you get more competent you will have more of those?

Due to the level of complexity in piano scores, I doubt whether anyone really reads quite as fluently as is possible with words.

Probably so.

[ajspiano]

It seems you just forgot the estimately 90% of people who do not speak English at all or only little 

..or your native spoken language, obviously.

[outin]

..or your native spoken language, obviously.

Not all languages are written down as text in the same way we are used to, so makes one wonder if people develope different mechanisms and skills for reading.

[nyiregyhazi]

Speed reading is based on the ability to not do that for every word. Concentrating on individual words will often hinder full understanding. For me it's rather easy to avoid that when I am really interested in what I am reading, otherwise it's really hard and needs practice.

I feel there's at least some analogy for music reading. There are things in the score you don't need to process at all (many rests for example). I assume when you get more competent you will have more of those?
Probably so.

There's actually a big limit to what doesn't need to processed unless using memory. Even rests after a note must register, to time release correctly. Vastly less assumptions are possible in music than in reading. The difference between an implied perfect cadence and a surprise interrupted can be just one note of 4 or 5. Music needs much more verification of even strong expectation to stay accurate. The most competent logically must just process quicker and more accurately.

I don't call it speed reading unless processing the words, personally. I don't "say" them in an internal voice but my eyes run across each word. I can read very fast indeed this way. I'd call it skim reading, if the eyes are merely looking for something of interest. At that point, there's so much information on the page that many words can neither be said internally nor even processed. But speed reading doesn't mean failing to process words outright, to me.

[outin]

There's actually a big limit to what doesn't need to processed unless using memory. Even rests after a note must register, to time release correctly.

Sure there's a limit.

But one does use the memory anyway for recognition of forms, so how would sight-reading without taking advantace of memory be even possible?

I guess sometimes one has to register a rest after a note. But there are often several rests in one hand before the next note is to be played. No need to register them all (and which ones were used to fill the measures on one staff), only the next note to play in that hand.

Vastly less assumptions are possible in music than in reading. The difference between an implied perfect cadence and a surprise interrupted can be just one note of 4 or 5. Music needs much more verification of even strong expectation to stay accurate. The most competent logically must just process quicker and more accurately.

But there's still plenty of music with little surprise and an internal logic that helps with reading. There's also lot of repetition in music and patterns can may be easily recognized when they reappear without closer study. The interaction between the two staffs is not random. One can often concentrate more on one staff while ignoring things on the other. Aren't these the things one needs to take advantage of to be able to sight read new material at a reasonable tempo?

It's really the same with text, if it's complicated or foreign language, one has to pay much more attention to single words.

I don't call it speed reading unless processing the words, personally. I don't "say" them in an internal voice but my eyes run across each word. I can read very fast indeed this way. I'd call it skim reading, if the eyes are merely looking for something of interest. At that point, there's so much information on the page that many words can neither be said internally nor even processed. But speed reading doesn't mean failing to process words outright, to me.

Written text is littered with words unnecessary for understanding the sentences and even sentences unnecessary for undertsanding the text. One part of fluent reading is to not waste time on them. It's not the same as only skimming through test to look for something, which can be done much faster, but very little is actually understood and registered.

[j_menz]

I guess sometimes one has to register a rest after a note. But there are often several rests in one hand before the next note is to be played. No need to register them all (and which ones were used to fill the measures on one staff), only the next note to play in that hand.

Nope.

Good sight readers process larger chunks of information as one item, not more smaller items faster.  Those rests form part of a chunk.  They aren't registered individually, but then neither are the individual notes.  The "next note to play" is part of a much larger picture incorporated in a chunk and isn't processed individually.

[nyiregyhazi]

Nope.

Good sight readers process larger chunks of information as one item, not more smaller items faster.  Those rests form part of a chunk.  They aren't registered individually, but then neither are the individual notes.  The "next note to play" is part of a much larger picture incorporated in a chunk and isn't processed individually.

That's a common view but highly improbable. Expectation is too limited to be reliable without verification of detail. Words are much more predictable than music. Any one of many details could vary subtly compared to expectation. A perfect cadence vs interrupted is but one note different. Expectations cannot be trusted without an eye for detail. It's the eye for detail that identifies the chunk accurately, or there would be endless false assumptions. It would be like if there were infinite combinations of letters in words. Combinations of letters are more limited and thus require less detailed verification for certainty.

Imagine if any existing word could have any single letter changed to another random letter yet mean something else. . We'd need to be much more careful not to make mistakes. That is the level of scrutiny involved in good score reading. We see chunks effectively if we have a rapid eye for detail, but not without that.

[j_menz]

That's a common view but highly improbable. Expectation is too limited to be reliable without verification of detail. Words are much more predictable than music. Any one of many details could vary subtly compared to expectation. A perfect cadence vs interrupted is but one note different. Expectations cannot be trusted without an eye for detail. It's the eye for detail that identifies the chunk accurately, or there would be endless false assumptions. It would be like if there were infinite combinations of letters in words. Combinations of letters are more limited and thus require less detailed verification for certainty.

There was nothing in what I said that implied any level of expectation, which I take you to mean some level of educated guess. The chunk is read accurately and in full detail.

[nyiregyhazi]

There was nothing in what I said that implied any level of expectation, which I take you to mean some level of educated guess. The chunk is read accurately and in full detail.

How do you read a chunk without seeing it's components- if any small change would make it something else? It's a subjective experience assembled via rapid processing of details, surely? What else could be possible?  The problem is that although experienced this way, it's impossible without grounding in both comparative intervallic reading and reading of individual notes.

The significance for a good reader is nil, but for a poor reader it's necessary to assemble chunks by  developing rapid reading of details or the concept of chunks is inaccessible.

[j_menz]

How do you read a chunk without seeing it's components- if any small change would make it something else? It's a subjective experience assembled via rapid processing of details, surely? What else could be possible?  The problem is that although experienced this way, it's impossible without grounding in both comparative intervallic reading and reading of individual notes.

The significance for a good reader is nil, but for a poor reader it's necessary to assemble chunks by rapid reading of details or the concept of chunks is inaccessible.

I'm probably not a good person to ask how it's done, it's just something I do.  Part of it is experience - I read a lot of music - and so my collection of separately identifiable chunks is probably pretty extensive.  It's not universal, of course, and there are certainly instances where I have to construct the chunks in the way you describe. I don't dispute the need for solid grounding in intervallic reading and reading of individual notes, it would seem absolutely essential.

Poor readers clearly need to do the groundwork, and it's no good to point them at chunking as a way to improve, it's something that happens, but I don't think it can be forced and I'm not sure there are any shortcuts.

[nyiregyhazi]

I'm probably not a good person to ask how it's done, it's just something I do.  Part of it is experience - I read a lot of music - and so my collection of separately identifiable chunks is probably pretty extensive.  It's not universal, of course, and there are certainly instances where I have to construct the chunks in the way you describe. I don't dispute the need for solid grounding in intervallic reading and reading of individual notes, it would seem absolutely essential.

Poor readers clearly need to do the groundwork, and it's no good to point them at chunking as a way to improve, it's something that happens, but I don't think it can be forced and I'm not sure there are any shortcuts.

I think my point is that all chunks are constructed- regardless of perception. It's just that sometimes we need to perceive the assembly and sometimes we don't. Expectation does play a big role but good readers are better at verifying or discarding assumptions rather than running blind.

If a complex chord with odd intervals and many accidentals were to recur in many pieces of music, sometimes in the exact same form sometimes with a small chnge, you'd have to take tremendous care not to make errors. You couldn't simply use memory or you'd miss changes. It would take as much Care to genuinely verify the familiar weird chord (without mistaken positives) as to correctly read the one with a single different note. It's not about familiarity but the complexity involved in a true verification. Memory could start to assume based on incomplete information but it couldn't verify the chord without seeing each individual component. Chunks that we read at a glance are more basic.

Logically, it all points to good readers not so much using memory to read familiar chunks as using rapid processing of detail to to assemble a picture and then compare that against memories. It's not quite the same as identifying something quick because of memory.

[j_menz]

I think my point is that all chunks are constructed- regardless of perception. It's just that sometimes we need to perceive the assembly and sometimes we don't. Expectation does play a big role but good readers are better at verifying or discarding assumptions rather than running blind.

If a complex chord with odd intervals and many accidentals were to recur in many pieces of music, sometimes in the exact same form sometimes with a small chnge, you'd have to take tremendous care not to make errors. You couldn't simply use memory or you'd miss changes. It would take as much Care to genuinely verify the familiar weird chord (without mistaken positives) as to correctly read the one with a single different note. It's not about familiarity but the complexity involved in a true verification. Memory could start to assume based on incomplete information but it couldn't verify the chord without seeing each individual component. Chunks that we read at a glance are more basic.

Logically, it all points to good readers not so much using memory to read familiar chunks as using rapid processing of detail to to assemble a picture and then compare that against memories. It's not quite the same as identifying something quick because of memory.

I think to some extent what I do, at least, is read the differences as much as the familiar. So for me, that complex chord would be a template, and any difference - one note, even - would be what I consciously saw - as strongly as the chunkised chord. It's not a matter of constructing, or reconstructing the familiar from scratch, however quickly, as recognising how what is there on the page differs from an inbuilt pattern.

[nyiregyhazi]

I think to some extent what I do, at least, is read the differences as much as the familiar. So for me, that complex chord would be a template, and any difference - one note, even - would be what I consciously saw - as strongly as the chunkised chord. It's not a matter of constructing, or reconstructing the familiar from scratch, however quickly, as recognising how what is there on the page differs from an inbuilt pattern.

I'm not so convinced. If it were a standard tonal chord, I'd expect to recognise an alteration as a glaring issue. If it were in Scriabin or Ives, I couldn't honestly say that a single alteration would stand out enough for me to be certain. A change in a visible interval might. But a single missing or different accidental among a variety of sharps naturals and flats probably wouldn't be something that would come out to me. I'd have to run through every detail quickly in order to avoid any risk of false positives- even when the chord was indeed the same. That suggests that reading details primarily triggers a memory of an assimilated chunk and compares it after processing detail- not that recollection of a chunk is how the reading side is carried out. In atonal chords, alterations don't necessarily stand out. Even in more straightforward chords, there's no reason to assume that a good reader doesn't quickly verify details rather than literally see a C major chunk due to memory of it. After all its very easy there. More likely, the brain just puts together things so quickly and then compares to memory that it seems like reading due to memory. Then any less expected details go on to stand out after the whole picture is assembled. The issue of whether false positives could be avoided in a complex atonal harmony casts significant doubt that details no longer need to be individually assessed, merely because a chord looks very familiar. When it's complex and dense, it's no easier to be certain (rather than casually assumptive) of a positive than to correctly read a minutely altered chord that merely misses a single accidental- unless details are being processed. You might save time by running them against expectation (not that you can exactly build a particularly comprehensive repertoire of an atonal chords and use it expectation widely in the real world), but you still need to verify in full. You can only be sure of registering all changes if you are also verifying non-changes too. So merely to repeat a complex chord that really is the same requires as much detail to be processed as the first one, otherwise it's the product of assumption rather than certainty due to memory recall. Logically that would suggest that the brain zones in on the alteration quite simply because it's done with verifying every other detail- not because it ignored them.

[j_menz]

I'm not so convinced.

You're delving into the realm of subconscious processing. You may well be right, though I doesn't wholly ring true to what it feels like for me - that feeling may well mask a whole lot of processing along the lines you suggest nonetheless.   Not sure there is any benefit to the analysis, though, in terms of practical improvement.

[outin]

Nope.

Good sight readers process larger chunks of information as one item, not more smaller items faster.  Those rests form part of a chunk.  They aren't registered individually, but then neither are the individual notes.  The "next note to play" is part of a much larger picture incorporated in a chunk and isn't processed individually.

That's what I actually meant, but you said it better. Same with text, at some level of consciousness you do probably register the unnecessary things as well, just not to the similar detail. Just like when you sight read, when I read text really efficiently and fast I am nor fully aware of what I do, it's just natural.

[ajspiano]

That's what I actually meant, but you said it better. Same with text, at some level of consciousness you do probably register the unnecessary things as well, just not to the similar detail. Just like when you sight read, when I read text really efficiently and fast I am nor fully aware of what I do, it's just natural.

I suspect you could say its to do with having a certain familiarity with the syntax of the language. We know how to predict the structure of sentences and paragraphs and so long as small details are done correctly we brush over them as they were at least partially known in advance.

This is true of music, once we have a familiarity with a genre or style, common chord progressions and scales, the ability to recognise themes/motifs and developments of those etc. Well written music also often has leading tones that tell us roughly what the next development will be before it happens.

[thorn]

In my other life as a language teacher, there are various *types* of reading that we need to help people to develop. I think a lot of these are applicable to reading music...

The ability to scan and extract specific information (e.g. you never consciously read an entire menu, you just locate things you might like to eat). Example in music, picking up a new piece and flicking through it to see if there's anything that might make learning it a bit difficult. The more experienced you are, the more accurate this will be. For example an amateur will literally look at how much black there is on the page where a professional will be able to scan and relate it to their own technique at the same time.

The ability to skim for gist. On a lower skill level this is quickly gathering the who/where/when etc, on a higher level subconsciously predicting what is going to come next. In music reading I personally relate this to a musician's ability to work on the score *away* from the piano. From basic things like picking out repeated passages to higher things like getting an idea of fingers/hand positions.

Most language examinations (and music sight reading tests) test ability to combine these things. In controlled sight reading tests we usually literally have time to gauge the structure and pin point anything that looks tricky and might need working out before they examiner asks us to play it. (Comparison, in reading text I remember when I took the Japanese proficiency test there would be about a page and a half followed by a "what does the writer thing about x and y, and what is the main point they want to make".)

I think any more detailed kind of reading can take place because of the above three skills.

I also think that "reading", "reading aloud/playing from score or memory" and "speaking/composing/improvising" are entirely different.

I can read a piece of music away from the piano and hear it in my head, I can't necessarily then go and play it exactly like that and I couldn't necessarily write/improvise anything like it. It tends to work in reverse when I've taught English; people can hold a conversation, struggle reading alone and HATE reading aloud.

[nyiregyhazi]

I suspect you could say its to do with having a certain familiarity with the syntax of the language. We know how to predict the structure of sentences and paragraphs and so long as small details are done correctly we brush over them as they were at least partially known in advance.

This is true of music, once we have a familiarity with a genre or style, common chord progressions and scales, the ability to recognise themes/motifs and developments of those etc. Well written music also often has leading tones that tell us roughly what the next development will be before it happens.

Internally speaking, it may be quicker for the brain to confirm an expectation than to process a failure to meet expectation. However, it's really quite impossible for expectation to ever serve as a replacement for observation. Otherwise the permutations involved would mean that those who work that way would be going wrong left right and centre.

To be honest, I see the ill effects of casual but unverified expectation all over the place. Subtle differences in passages that are otherwise identical are routinely missed by students. In my opinion, the greatest sightreaders are so good precisely because they don't need to be too reliant on guesswork, no matter how educated. The hallmark of a good sightreader is that they verify well enough never to be fooled by expectation. The only possible explanation I can logically see is that it's easier to process confirmation of expectation than it is to read something unexpected. That would account for why so much is perceived as being about expectation, without situations where expectation creates errors.