Where to start?

[twang]

How far did your theory class go? Probably it stopped just short of where it would be really useful. Understanding harmony & voice leading and being able to analyze a piece goes a long way to getting you unstuck after you've noodled yourself into a corner, or when you find that riff you want to expand into a theme.

Just after writing the above I noticed this discussion:
viewtopic.php?f=4&t=82692

[charles.patrick]

Pfffff...... awesome link to that John Hall site. Why have I never seen that before?

Nice one.

[simonm]

Pfffff...... awesome link to that John Hall site. Why have I never seen that before?

Nice one.

Interesting to notice the jazz link here. The ABRSM syllabus for guitar requires theory from grade 6 up.

"PREREQUISITE FOR ENTRY:Grade 5 (or above) in Theory of Music, Practical Musicianship or any solo Jazz
subject."

However, it effectively equates theory with playing jazz.

[charles.patrick]

Ok on closer inspection, that John Hall site looks like it has some interesting analysis of Sor, Carcassi, etc..... it's also probably too much jazz theory for what I was looking for. Pretty much anywhere I see the word 'jazz' I'm instantly disinterested.

"The tritone sub Ab13 replaces the Dm7 (ii) functioning as a V/V substitute" - NOPE!

Anyways, back to where we were before, basic counterpoint. There's a book called The Study of Counterpoint by Alfred Mann. Looks like a good starting point to me. Anyone read it or got any better suggestions before I buy it?

[Max Karios]

The Alfred Mann book basically is Fux' "Gradus ad Parnassum", the standard text on counterpoint that all composers of the 19th century learned from. Don't expect too much from it, though. The so-called "species counterpoint" presented in the book is a set of technical exercises. You are encouraged to avoid recognisable patterns, which would be required for musical coherence. It basically separates the technique from the art. There is another volume by Alfred Mann on imitative counterpoint called "The study of fugue". This contains the last chapters of Fux' book and texts from other authors.

Apart from that you could take a look into the books by Arnold Schönberg. They are not about twelve tone techniques, as one might expect, but about tonal music. "Preliminary Exercises on Counterpoint" contains the same material as Fux, but it is much more detailed. "Fundamentals of Musical Composition" is about musical form. Highly recommendable and probably quite unique. "Structural Functions of Harmony" goes very deep and can be overwhelming.

One aspect of composition where you hardly find good texts is modulation. I know only one book that treats this subject in detail, but it is available only in German. (Heinz Acker: "Modulationslehre"). There is a tiny booklet by Max Reger simply called "Modulation", but this is not really a theory book. Instead, it contains the shortest modulation paths from any key to any other in four-part counterpoint examples. They are not always playable on guitar, and stripping it down to three parts causes new problems, because three parts often turn out to have a weak balance (i.e. it is impossible to find solutions where one voice goes up, one goes down, and the third remains stationary.)

Now, after all these recommendations I must add that simonm is perfectly right when he points out that any amount of books will not help you in the end. I would suggest to use an iterative approach, where you go through various topics repeatedly and try to improve what you are writing. Improvisation is indispensible in this process. It doesn't have to be stage-perfect improvisation, but you can get a good grip on counterpoint only by playing around with what you learned from the books. You will find out that it is not too difficult to create things that sound like genuine baroque or classical music. Switching between third and sixth parallels in patterns or in unpredictable ways is a big part of it already. Holding one voice back to avoid running into (hidden) parallel fifths is another trick that works out well.

If you are getting good ideas noodling around then it is only a matter of diligence to get the rest done. When the Muse comes around she should find you working.

[Christopher Freitag]

There are lots of approaches to composition. It's sad but true that there aren't a lot of good books on the subject; music theory textbooks (which I publish in my other life) cover the basics or harmony, phrase construction, and so on, but from an analytical rather than creative perspective.

The music theory textbook that I used many many years ago was by Leo Kraft and was called Gradus. One element of that was exercises that had us composing short works "in the style of" other composers. A minuet like Haydn. A dance like Schubert. I think this is an excellent approach for one just starting out. Take some very short guitar pieces - little dances by Sor, for example - and study them. Then try to create something on the same formal pattern, but with your own musical ideas.

For this to work, it helps to know at least the basics of theory...particularly harmony. You should understand how chords are constructed and the basics of triads, inversions, 7th chords. You should be able to take a simple melody, harmonize it with basic chords, and then learn what chords you can substitute. As a poster above mentioned, counterpoint is also a good thing to explore. Take the same simple melody and try writing a second voice - a bass line is a good place to start.

One other suggestion I can make. If you know the company Coursera (an education company that does MOOCs - massive open online courses) they have a course starting in January called How to Compose like Mozart. Like all Coursera courses, it's free.Could be a useful place to start.

I hope that is helpful.

[charles.patrick]

You will find out that it is not too difficult to create things that sound like genuine baroque or classical music. Switching between third and sixth parallels in patterns or in unpredictable ways is a big part of it already. Holding one voice back to avoid running into (hidden) parallel fifths is another trick that works out well.

If you are getting good ideas noodling around then it is only a matter of diligence to get the rest done. When the Muse comes around she should find you working.

Thank you very much Max. Some excellent information.

I do occasionally make some nice noises from noodling around but I've always felt that it doesn't sound like "classical" music. It doesn't sound like the pieces I enjoy playing. That's really why I wanted something to study, because clearly I have no idea how pieces are composed or how they work.

I suppose I shouldn't worry at all about what it sounds like, I should just write and it is what it is.

But if I don't like my own music then it's hardly worth writing is it.

[charles.patrick]

I hope that is helpful.

Yes it is. Very helpful. Thank you.

I like the idea of learning how to copy other composers to start with. That is essentially how we learn anything isn't it.... by copying experts.

[charles.patrick]

The music theory textbook that I used many many years ago was by Leo Kraft and was called Gradus.

Gradus looks like a decent book actually. I wonder why it doesn't have more reviews on the book selling website?

[Christopher Freitag]

Gradus looks like a decent book actually. I wonder why it doesn't have more reviews on the book selling website?

It has been out of print for some years now. I'm surprised that there are new copies even available.

[Brynmor]

Have you looked at the "Dummies" Guides? There is one for music composition and another for music theory.

[SusanGRas]

I disagree, I find a lot of the time if you just start writing, come up with something meh, tweak it til you like it, etc is the best way to get started


Counterpoint is the standard, so to speak. 90% of our repertoire comes from people who have studied counterpoint, and even most modern music follows it to some degree--jazz is just counterpoint with 7ths & 9ths

I recently watched the special on Stephen Sondheim, Six by Sondheim, and this was his suggestion. Just start writing and then tweak.

[charles.patrick]

I agree that probably a lot of music is composed by improvising some nice idea and then 'tweaking'.

However, these composers surely have some basis for their tweaking. It's unlikely that they just fumble around randomly and then suddenly come up with a well composed piece of classical music.
No doubt they are already very comfortable with theory of classical harmony, counterpoint, form and structure.

Is it not like an artist or a painter? Sure they may be just "fiddling" or "trying out ideas" but they're fiddling within the guidlelines of what they know about colour mixing, composition, drawing, tone, form and so on.

I perhaps should have worded my original post differently. I want to start composing little bits and pieces in a classical style.....is what I meant.

I can already sit and noodle, put chord progressions together, improvise simple melodies. I don't really consider this composing though.

[Max Karios]

In my opinion, form is the crucial point that separates noodling from composition. It is also the most difficult part, because there are no fixed rules. Most people consider harmony and counterpoint difficult, but once you know their rules, you have something to cling to. You break the rules only on purpose to create a specific effect. See for instance the hidden parallel fifth in the first bar of the Aria of the Goldberg Variations. This could easily have been avoided, but Bach actually puts it on a pedestal as if he would be shouting: "Parallel Fifth! See what I'm doing here!"

With form it is exactly the opposite. There are no reliable rules to set you free. How much exact repetition or immediate variation of a theme is enough, how much is too much? The more you try to tweak it, the more you tend to mess it up. If the legend is correct, Chopin wrote many of his pieces as spontaneous improvisations, then he tried to "fix" them for several days, and at the end of his nerves he gave up and went back to the initial version.

Every composer has his own ways to keep pieces together. As an example, look at the Prelude BWV 998. The theme starts on the root note, then it is repeated lower, starting at the fifth, then at the third. While it still is running in "third mode" the piece modulates to the dominant key and everything starts again. Then, at the same spot as in the original key comes a modulation to the dominant (of the dominant), and in the new key the theme is modified for the first time, while the lower voice starts a kind of "walking bass". There is not really that much variation going on, but the piece doesn't get boring at all.

Scarlatti does things completely differently. Almost all of his sonatas have two parts. The first usually is a succession of motives that tend to have nothing in common. He uses repetitive patterns to avoid putting too much material into it. Then these motives are modified and combined with one another in part two, often with surprising key changes. It is actually the B parts that make Scarlatti a great composer.

So here are already two completely different approaches, and both do the trick. I don't think that either of them is easier to emulate. But there is one thing that I'm sure of: what distinguishes a top notch composer from the rest is not how well he knows the rules of the game (they all know the tricks of the trade), but how well someone is doing in places where you cannot find any rules.

[Oliver Newman]

http://alanbelkinmusic.com/ABWritingNAV.html