Can music really be taught? For that matter, can it be learnt by any other way than constant practice? In the first of a series looking at the other bits of musicians' lives John Morrish takes the scholastic route.

The other day I took a few moments to gather up all the books I've collected over the years which claim to instruct me about music.

That leaves aside all the stuff about individual musicians, or about styles, or about musical history. No, I'm referring here to how-to books, like "Bert Weedon's Book Of Feedback" or "Madrigals For Your Fairlight".

It turned out to be quite a collection, from chord books and books of guitar technique ("How To Play Jazz Guitar" or "Mickey Baker's Jazz Guitar"), through books of exercises ("Carcassi's 25 Studies Opus 60"), to books of arrangements ("Bossa Nova", "The Beatles For Classical Guitar"), and then on to more generally theoretical books, like the little red book of Rudiments, or "Learn To Read Music". And then there's a whole string of books with titles like "Introducing Music", covering everything from Gregorian Chant to Stockhausen.

Some of these I have read and understood, more I read once and then put aside, confused, and still more await a proper evaluation, one day when I'm free from more attractive distractions like wallpapering the hall or de-worming the dog.

Now, I'm sure that the collection I've put together is not that unusual, even allowing for the fact that most classical guitarists don't buy books on jazz and vice-versa.

Go into any specialist music bookshop and see the packed shelves, and watch the traffic at the checkouts: it's a big business.

And yet, after all that reading, how many of us feel we have a coherent understanding of what music is all about? Certainly, the rock musician is desperately ill-served by what passes for the instructional literature.

Now for the classical musician, that's not so important. Mostly, the classical student works with a teacher who leads him or her through a logical sequence of practical and theoretical study. But even where the classical student opts to work alone, the material is available to make that a realistic aim. Take, for instance, Frederick Noad's masterly books on solo guitar.

Where is their equivalent for the jazz or rock player? Look in most rock guitar books and you find the same old six-note barre chords and pentatonic scales, usually wrapped up in sufficiently mystifying terms to enable the star name whose face adorns the cover to retain the appropriate air of "mystery".

Even books which aim a little higher, like the Mickey Baker series, offer lots of information about what to play, but precious little about why. A player could learn a great deal about chord substitution, for instance, from these books: but only in terms of "correct" procedure. Mr Baker tells us: "The minor fifth of any dominant chord can be substituted in place of the Dominant itself." But he doesn't tell us why. It's all part of the mystery.

Now it could be argued that none of this matters, that rock music has always been made by people who are ignorant of technicalities. Indeed, some people exalt that ignorance to the point where it comes to be seen as the guarantee of rock's integrity. There's something in that, not least because traditional musical education has insisted on teaching aesthetic "standards" rather than handing over information.

Perhaps in the days when rock was a new, vibrant musical form the ignorance-is-bliss approach would stand up. Now it won't. The music needs all the help it can get if it's going to escape from providing background sounds for a succession of ever-more-narcissistic video performances.

In the first place, a better technical education for the average rock player would ease the terrible communication problems experienced by the average set of musicians when they get together for the first time. At last, they'd have a common language in which to express musical ideas and argue about them. And they wouldn't be limited any longer to people who'd sweated over approximately the same guitar tutors. They could speak to classical players, jazz players, indeed anyone working in Western musics, and they could work together.

And if reading and writing music became universal it would effectively end the days of the one-style player, and good riddance too. The whole written musical literature of the West would become available for performance, study and incorporation in fresh, new musical hybrids.

But this emphasis on reading and writing should not lead us to believe that all the faults are on our side of the great musical divide. An average rock player has a better, though "untrained", ear than a classical contemporary. And if you want to frighten a classical instrumentalist, just invite him or her to solo over a 12-bar. And then there's the prejudice. Perhaps the worst element of conventional musical education is the insistence on the standard repertoire. It is not long since people found playing jazz were expelled from music colleges, and many of the prejudices against the "popular" musics still remain, usually disguised as criticism of the rock or jazz player's "poor" technique or technical "ignorance".

In turn, on our side of the divide, we have never stressed the positive aspects of our methods. Our techniques are different, certainly, but they have evolved to meet musical needs. And the fact that we have no standard repertoire, the fact that we are all essentially self-taught, is a plus point, given that most rock players are also, in some measure, composers.

If we are to reform musical education, we want to be careful not to throw out the baby with the bathwater. We want a common, standard body of theory, of musical language and technical skills. But we don't want to dictate style, or aesthetics. It's the language that's important: what people choose to say with it is their own business.

Now musical education has an exceptionally long history, at least for the wealthy, but it's difficult to believe that what is achieved today is any more successful than what was done 100, 200 or even 400 years ago. In Elizabethan England, it was expected that every young aristocrat would learn to sing at sight madrigals and part-songs of quite horrendous complexity. How many people leave school able to sight-read today? More leave with a violent distaste for anything resembling "serious music" and a horror of having to sing or play a musical instrument in public.

The woeful lack of success in school music teaching which, I suspect, most of us experienced or are experiencing, is odd considering that music has long been considered an essential in every school.

Sadly, even where it is successful, school music seems misguided, obsessed with nurturing the instrumental talents of the few, especially the few who want to climb the ladder of "grades", a certain route to the kind of academic isolation that has traditionally marked the classical player.

Interestingly, the intensely technical musical training provided by the Royal College of Music exams has come under attack from its own side, for producing "mechanical" players, capable of playing from a score or memorising, but with no true musicianship about them. That traditional type of training sees the musician translating written notation directly into a muscular movement, and producing a sound which is completely unanticipated.

An increasing number of classical musicians have come to the conclusion that that was the wrong way round. The playing should come first, then the reading and writing and associated theory.

A complete musical education needs to start with singing in tune, then to work up through a simple melodic instrument to the reading and writing of simple tunes, including taking down simple tunes by dictation, and then moving on to basic harmony study on a keyboard instrument or even the guitar.

If you started at five years old, and stuck with it right through school, think what you might achieve by 16. Of course, in many ways school is the least suitable place to learn about music. Too many distractions. Adults have a greater motivation, but practical circumstances make it more difficult to get them into a classroom, if indeed that's the best place to learn.

In practice, most adults tend to study on their own. Oh, they may go to a teacher or a class, but the bulk of their study will be on their own. And monitoring your own performance honestly is one of the most difficult parts of learning music on your own.

Lately, I've become interested in the possibilities of the computer as a device for teaching certain skills. For instance, there are all the "typing tutor" programs, which display a series of letters, measure how quickly you hit the right key, tot up your errors, display them on a graph, then produce a new exercise to help you improve your performance on the particular letters you got wrong.

In reading music. I've always found the rhythm the difficult part. If I'm trying to work something out you can see me tapping my pencil on the desk, or slapping my thigh until it is raw. Then when I've got it, I set about putting the notes in. Of course, if I don't know the piece. I'm never 100 per cent sure that what I've worked out is actually correct. Now, why couldn't you have a device with a simple display and one key, which presents you with a series of rhythm patterns and invites you to tap them out, correcting you when you're wrong? Alternatively, you could program it with the particular rhythm you want, and let it play it back to you. I'm sure the technology to do this already exists, even if such a device has never been put on the market.

But why stop there? A slightly more elaborate machine could display full music notation and invite you to hit the right keys on its keyboard as the symbols appeared. As each exercise finished, it could tell you your score and give you a different test, to improve your weak fingers.

So that the thing does not become too mechanical, it could offer dictation, by producing a string of sounds and inviting you to match them up by playing the keyboard. Or it could display a musical phrase and ask you to sing it into its microphone, to ensure development of your melodic sense.

Later, it could offer a phrase and invite you to produce a bass-line and harmonies. Or it could present a bassline and harmonies and ask you for a melody.

Like all computers, it would work at your own pace. It would repeat an exercise as many times as you wanted, it would tell you where you were making mistakes, and it would never get bored or tell you about its days on the road with the London Philharmonic or Ritchie Blackmore's Rainbow.

Best of all, it would not dictate your repertoire, or your taste. It would take the mechanical part of music, and invite you to practice it mechanically.

So painless would the training be (and it would be just that, not education) that "listeners" and "non-musicians" would be rushing to take part, while serious instrumentalists would have plenty of time for human instruction in the technique of their chosen weapon, and for expression and musicianship.

Having briefly played with the Yamaha CX5M music computer, I'm quite sure that the technology exists to do all the functions I've mentioned above. Perhaps the best bet would be for the thing to have no pretensions to being a computer, or a musical instrument, but for it to be, bluntly, a teaching machine.

But whatever it is called, such a machine could be at the heart of a renaissance of real musical activity, based on a common language. At a time when schools are falling over themselves to teach children computer languages, it would be good to see the most widely-used symbolic code of all, the language of Western music, come into its rightful prominence.