How to actually apply the Modes when writing music/improvising.

[elsmandino]

Hi there,

I completely get the theory behind modes but would be really grateful for some clarification as to how they are applied in practise – this is something that I have found really hard to find anywhere.

For example – say I write a chord-progression in C major, with the following chords:

C, Em, G and Am

My obvious choice would be to pick C Ionian to solo.

However, the chords used just happen to fit the C Lydian scale so I could then choose to use this instead.

Therefore, I could choose either one depending on the effect I want to achieve – some people, for example, feel that Lydian has a sweeter sound.

Here is where I have a slight problem – what would happen if the original chord progression was C, Em, F, Am?

The Lydian scale has a sharpened F whereas the backing has a natural F. Could I still use the Lydian – what I am getting at is whether this deliberate clash of tones is what is actually giving the Lydian mode its “flavour “ in the first place?

Or am I completely wrong – can I only use a scale over a backing tracking when it contains the same notes as chords in the backing?

Just to illustrate my point, if I were to write any chord progression, how do I then work out what scales I could use over the top?

For example, if the key is minor and begins with an Am, could I use A Aeolian, A Phrygian or A Dorian over the top?

Perhaps I am overanalysing this, but your thoughts would be greatly appreciated.

Thanks v much.

[JonR]

I completely get the theory behind modes

Not quite, it seems...

For example – say I write a chord-progression in C major, with the following chords:
C, Em, G and Am

My obvious choice would be to pick C Ionian to solo.

Most of us would just call that "C major scale".

However, the chords used just happen to fit the C Lydian scale so I could then choose to use this instead.

That's true.
Or "G major scale", as traditionalists might call it.

It's only "C lydian" if C is clearly the tonal centre, the key chord, and that's not obvious just from your choice of chords. Given that scale, it could be a IV-iii-I-ii in G major, or a VI-i-III-iv in E minor. If all the chords are the same length, they'll probably sound like the key is G major.

Therefore, I could choose either one depending on the effect I want to achieve

Yes, as long as C is the clear key chord. You'd need to spend a lot of time on the C, and very little on the G.

Here is where I have a slight problem – what would happen if the original chord progression was C, Em, F, Am?

Then it would be an obvious I-iii-IV-vi in C major.
You could still use an F# on the other chords. It's just up to you whether you like the sound (which could be quite interesting). Just don't worry about what its theoretical label is.

The Lydian scale has a sharpened F whereas the backing has a natural F. Could I still use the Lydian – what I am getting at is whether this deliberate clash of tones is what is actually giving the Lydian mode its “flavour “ in the first place?

Not exactly. The lydian "flavour" comes from a raised 4th on a major chord. That's all.
Start to add other chords and the whole thing can get blurred. Quite possibly, other chords may start to point to another key centre (the ear is always trying to identify key centre from a group of chords, that's just cultural habit).

Two questions (to ask yourself):

1. Do you want to write a "lydian" tune? If so, just use ONE chord. A maj7 with a raised 4th in the scale. (A good passing chord is a major triad a whole step up, but best if you keep the root of the first chord under it: eg C and D/C for a C lydian sound.)
(Check out Joe Satriani's "Flying in a Blue Dream" for a classic "rock lydian"
exercise. He begins in C lydian, but the piece moves through some other lydian modes. IOW, the chord changes are not to other chords from C lydian, they're to lydian modes on other roots (Ab, G, F).)

2. Do you want to write a piece of music using a specific chord progression? If so, stop worrying about modes, and just choose whichever passing notes on each chord sound right (or best) to you.

Or am I completely wrong – can I only use a scale over a backing tracking when it contains the same notes as chords in the backing?

That's usually the best policy, yes. For a melody or solo, use the notes that are there in the chords, as a first choice. (A no-brainer, right?)
If that pitch collection is too dull, or doesn't quite hit the spot, you can always add to those notes, eg, any note missing from the chords, or any chromatic note. That's a more creative choice, and should be governed by trial and error, using aural judgement. Don't try to replace the given pitch collection with another one (don't "apply" another scale or mode). Of course, if it's your song, you can always change the chords!
Put the theory books down.

Remember that songs are more traditionally written the other way round: melody first. The chords are then added later, usually built from notes in the melody (the scale the melody uses). And still, chords can sometimes employ other notes, for special effects. You are never limited to one scale.

You CAN of course write songs chords first (many do). But the melody has to rule in the end, and you sometimes need to change the chords if a strong melody demands it. The melody IS the "song", because it's what a singer will "sing". (Nobody can sing chords )

Just to illustrate my point, if I were to write any chord progression, how do I then work out what scales I could use over the top?

See above.
The notes in the chords, plus any others if necessary, or if you like the sound of them. You don't need to identify that bunch of notes as a "scale" of some kind. The label makes no difference.

For example, if the key is minor and begins with an Am, could I use A Aeolian, A Phrygian or A Dorian over the top?

Depends on the other chords.
Am followed by Dm (and eventually returning to Am) = A aeolian
Am followed by D or D7 (but without ever going to G) = A dorian
Am followed by Bb (and returning swiftly to Am) = A phrygian

The traditional "key of A minor" is based on A aeolian, but normally includes an E major chord, including a G# note and signifying "harmonic minor" at that point (but probably not elsewhere).

[Malcolm]

Modes made simple.

If you are going to use a mode - for the modal sound - play it over a modal vamp. Don't waist your time playing a mode over a chord progression. If you have a chord progression pick your melody notes from the chord's pentatonic notes - of course, not in pentatonic order. Running a mode, over a chord progression, and hoping it will sound like anything other than a mode exercise - good luck with that.

Here is a post from a bass forum that seemed to hit it on the head:
Originally Posted by miltslackford

To me this is where part of the problem lies.

The idea that a mode has a mood is one thing.

But you have to stay in that mode for a long time. When modes really became famous was with modal jazz. The point of this music was that up til that point, jazz had been all about changes, so players were so busy figuring out how to play over the complex changes they couldn't really stay with one scale for very long.

So modal jazz was about saying, we are going to keep the same scale available for much longer, or have a set of changes that allow people to use one scale only, and we'll explore what we can do with melody when we don't have to continually dart round corners. So people could then say 'the phrygian sounds like x' because they played it for a while.

Now what's happened since then is people have started teaching things backwards. They start off talking about modes for people who are learning to play over changes which aren't for modal jazz. So if you have an opportunity to play a particular mode for a single bar or half a bar it makes absolutely no sense to talk about the 'mood' of that mode.

For playing over changes, I would recommend learning how to identify the key centres in a piece and get really really good at major and minor ii Vs. Look at chord tones and keys rather than scales. Look at how chromatic chords in a key centre introduce accidentals which you need to observe.

You will find people who have very isolated examples which work for them about how to play over a particular set of changes using a particular mode on a certain chord. This is really different to using modes as your main way of approaching changes. These are more like 'tricks' than what you could call a system. Start with the arpeggios, the chord tones and you'll get to where you want to go faster.

I think if you want to explore the 'mood' of a mode you really need to be playing some music that allows you to stay in that mode for a long time, so it needs to be modal music. If you want to explore this, get a drone and play the mode over the top. But I wouldn't try and solve your approach to note choice over harmony with modes.

To view the original post -- http://www.talkbass.com/forum/f22/sl...-modes-895896/ Understand this is a bass forum, keep that in mind as you read the posts.

A little more on modal jazz: http://en.wikipedia.org/wiki/Modal_jazz

..... modal jazz compositions or accompaniments may only or additionally make use of the following techniques:
1.slow-moving harmonic rhythm, where single chords may last four to sixteen or more measures
2.pedal points and drones
3.absent or suppressed standard functional chord progressions
4.quartal harmonies or melodies

Number 3 and 4 I'm not familiar with, however number 1 and 2 hit it on the head. Perhaps Jon can talk about number 3 and 4.

Have fun.

[mwtzzz]

Hi there,

I completely get the theory behind modes but would be really grateful for some clarification as to how they are applied in practise – this is something that I have found really hard to find anywhere.

Actually there is a common misconception about the use of modes.

Modes only have meaning if you are aware of which tones need to be avoided for each mode. There is a specific avoid tone or tones for each one. If you don't know which ones to avoid, you are not making any clear statement about harmony.

Here is where I have a slight problem – what would happen if the original chord progression was C, Em, F, Am?

On a most basic level, there are things you need to know when working with chord progressions, you need to know the tonal center, the temporary tonic chords, and the avoid tones. If you don't know these things, you're just shooting in the dark.

Just to illustrate my point, if I were to write any chord progression, how do I then work out what scales I could use over the top?

See above comment on the "three things" you must know.

For example, if the key is minor and begins with an Am, could I use A Aeolian, A Phrygian or A Dorian over the top?

Depends on the surrounding chords. If key is natural C major, and starting chord is A minor, and next chords are diatonic to C, then obviously A Aeolian. On the other hand, if in the song you have a key modulation or rehamonized sequence of chords which includes Am, then another mode would be the correct choice.

Message me offline if you want details.

-Michael

[Ken Valentino]

Any note that is not in play (meaning hasn't been played), and is a tritone connection to boot, is usually tough to put in, but not impossible. If you can get the F# to connect more with the G (as its 7) or the D (as its 3), then it could possibly sound intended. If that bright extended sound is what you're after, then I say go for it.
Oh it also matters how strong that F root is. If the F is really tense then this whole strategy is alot easier.

[mwtzzz]

Not quite, it seems...
Most of us would just call that "C major scale".
That's true.
Or "G major scale", as traditionalists might call it.

lol

Start to add other chords and the whole thing can get blurred. Quite possibly, other chords may start to point to another key centre (the ear is always trying to identify key centre

ie. "tonality"

For a melody or solo, use the notes that are there in the chords, as a first choice. (A no-brainer, right?)
If that pitch collection is too dull, or doesn't quite hit the spot, you can always add to those notes, eg, any note missing from the chords, or any chromatic note. That's a more creative choice, and should be governed by trial and error, using aural judgement.

Uh ... I liked your post up to this point, but it doesn't have to be trial and error, there are specific avoid tones and specific do's and dont's for adding tones to an improvisation

Put the theory books down.

But not your understanding of harmony.

Remember that songs are more traditionally written the other way round: melody first. The chords are then added later, usually built from notes in the melody (the scale the melody uses).

]

Melody is linear harmony, that's all

[JonR]

Uh ... I liked your post up to this point, but it doesn't have to be trial and error, there are specific avoid tones and specific do's and dont's for adding tones to an improvisation

Kind of. I wouldn't say "do's and don'ts" myself, but there is a hierarchy of notes from "good" to "awkward". There is never any note that can't be used (no note that needs to be avoided altogether), but there are always one or two that are hard to handle, and - at the other end - a few that will always fit, no matter how you play them.
My point was that you can usually tell these notes by ear. All of us know when we hear a bad note, we don't need any music education at all.
But of course - and I guess this is your point - a little education in the rules ("common practices") can always help.
(Ie, without any education, we can't say why the bad note is bad, or how to make it good, or whether it's always a bad note.)

But not your understanding of harmony.

Well, when I say put the books down, I mean that one has to understand these things by ear in the end, whether or not you use books to help along the way.

An "understanding of harmony" - at least as employed in popular music of all kinds - can be got well enough from listening to (and copying) enough music, totally by ear. That's how the vast majority of the greatest pop songwriters (since the Beatles at least) acquired their craft.
It's only like learning a foreign language by ear - by living in that country for a while. I wouldn't deny, of course, that a few books can be of assistance .

Melody is linear harmony, that's all

Well, for me that's blurring the terminology.

The reverse (almost) is more true: harmony (as in a sequence of chords) always contains melodies, as the voices move from tone to tone.

There is music that contains plenty of melody, but no harmony - such as Indian raga. You can choose to hear implications of harmony if you like (using your western aural habits), but there is none actually there.

Of course, in western music (as that is our topic ), melody and harmony are intimately bound up together. It's quite natural for us to hear implied harmonies in (say) an unaccompanied folk song. (I'm not sure that's always a good thing, however...)

In terms of blurring categories, I did like Mark Levine's view in The Jazz Theory Book that a chord is frozen (or stacked) scale, while a scale is a kind of strung-out chord. But chord-scale philosophy is often misapplied, in a way that ignores traditional harmonic practices.

[mwtzzz]

There is music that contains plenty of melody, but no harmony - such as Indian raga.

Or Hal Crook .... hahah ... Actually Hal Crooks' improvisations are sometimes deliberately atonal, and other times he deliberately begins to establish a harmony, but moves on to an unrelated harmony before giving the first one sufficient chance to develop. Which to the ear sounds like it has no tonal center, or a often-moving tonality ....

You can choose to hear implications of harmony if you like (using your western aural habits), but there is none actually there.

Often what we do when we improvise is to "imply" a harmony with a single tone, even though of course a single tone cannot establish harmony. But since we're accustomed to hearing and playing same chord progressions time-and-time again, we know that can simply hit a single tone in those cases and enough for the listener to pick up on the cue.

In terms of blurring categories, I did like Mark Levine's view in The Jazz Theory Book that a chord is frozen (or stacked) scale, while a scale is a kind of strung-out chord. But chord-scale philosophy is often misapplied, in a way that ignores traditional harmonic practices.

My view is that the chord-scale approach is accurate as long as the avoid tones are specified. There are certain tones that can be used on each chord, and other tones that shall be avoided so as not to change the harmony. The sets of tones are different from one chord to the next. As long as you're aware of the tones that can be used, and the tones that cannot, then whether you play the chord (vertical) or the improvisation (linear) it's the same harmony.

[JonR]

My view is that the chord-scale approach is accurate as long as the avoid tones are specified. There are certain tones that can be used on each chord, and other tones that shall be avoided so as not to change the harmony.

True. The "avoid notes" are those which upset or confuse the harmony. That's why they sound bad.

The sets of tones are different from one chord to the next. As long as you're aware of the tones that can be used, and the tones that cannot, then whether you play the chord (vertical) or the improvisation (linear) it's the same harmony.

Yes, I see the point. That would work in functional harmony as well as in modal harmony.

Except I still don't like the idea of tones you "cannot" play. I can see the principle as something simple for a beginner to remember. But as soon as you look at actual melodies (never mind actual solos), you see these "avoid notes" quite often.

(And actually, in pure modal jazz, there aren't really any avoid notes, because chords don't have functions which could be upset by them. There are definitely dissonances in modal chords, but we hear them differently. Dissonance is just "colour" in modal chords.)

My own experience with chord-scale theory (CST) is being inspired by Mark Levine's Jazz Theory Book, about 15 years ago. I'd been playing jazz on and off - among other genres - for around 30 years before I read it (and been taking jazz lessons for at least 5 years), but his idea (or my interpretation of his idea) of a chord being a stacked or frozen scale, and a scale being a kind of strung-out or melted chord - ie, both of them being different expressions of the same thing - seemed lliberating. It was, for me, a new way of looking at music.
In the jazz classes I took (group workshops), it was a standard part of the teaching, although it wasn't heavily pushed.
I went along with it intellectually, feeling I was understanding at a higher or deeper level (Levine's was the only jazz theory book around) - but somehow I never found it easy to apply in practice. I couldn't get out of my old habits of playing from the melody, and using chord arpeggios combined with passing notes, either diatonic or chromatic. (I did frequently get complemented on my solos, but didn't feel I was quite getting it.)

Then I began to read criticisms of Levine and CST (not always together). In particular I encountered Ed Byrne on allaboutjazz, whose description of "linear" improvisation exactly matched my old habitual thinking - which of course I'd developed through years of actually listening to old jazz and blues recordings and copying what they did (as well as I could). For me, that way of playing was a no-brainer - it seemed obvious, it was easy and it worked. It was the intellectuality of CST, its authoritative logic, that made me think my way was "amateur".
But now I was starting to see that a lot of experienced jazz musicians and theorists thought the same way that I (instinctively) did.
I saw Joe Henderson's quote about solos "sounding like the index of a book" (jazz graduates playing correctly according to CST, but with no musical interest).

And then there was Hal Galper, whose rant at the beginning of this video almost made me jump out of my seat with excitement:
http://www.youtube.com/watch?v=NehOx1JsuT4
- the veil fell away: dammit the guy is right!

He does go a bit too far the other way (obviously irritated by the popularity of CST). He's thinking from the perspective of functional jazz (even though he presumably includes modal in that "last 100 years"). CST developed out of modal jazz, where it's quite correct to separate chords into individual vertical entities. You can't play modal jazz with the old functional habits. There is often voice-leading between chords (horizontal elements), but there are none of the old key scale relationships. There are no chord functions.

But the trouble with CST is that (as HG says) it's an easy system to plan out and teach; it's an appealingly simple set of rules; and that means it tends to assume an authority beyond its actual importance. The old ways of playing jazz weren't taught in colleges, or written in text books. They were taught by ear and experience; most of the time they were actually learned rather than taught.
Of course that sounds terribly old-fashioned (should we actually close down all college jazz courses??); but the point is that those old ways still sound good, even when applied to modern jazz. And not just because modern jazz still retains functional elements, or because old-fashioned jazz standards are still played. It's because melody still matters, it still has a broad appeal.
You can't learn melody, or melodic playing, by studying CST. The only way to learn it is by playing melodies: over and over, learning them by heart. As many melodies as you can get your hands on.

[jimc8p]

There is music that contains plenty of melody, but no harmony - such as Indian raga. You can choose to hear implications of harmony if you like (using your western aural habits), but there is none actually there.

I think it's our Western aural habits that would cause you to come to that conclusion! I'd say that Indian ragas are purely and unadulteratedly harmonic. Harmony generically refers to the intervallic relationships of tones, like the perfect fifth, as a simple example. Those two tones 'harmonise' together to create an aesthetic effect. This happens most strongly when taking place vertically, so that the texture of the sonic ratio is happening directly...and this can also take place horizontally by a kind of implication. In this sense, I consider "melody" to be a sub category of harmony.

Indian ragas use harmony and melody. The traditional Indian musical set up is designed to create an extremely complex harmonic backdrop, albeit focussed exclusively on the tonic. Just think of a sitar with its sympathetic strings. The chords taking place are naturalistic and complexly based on the melodic content. Physically speaking, these harmonic textures are so much richer and more intricate than those created by Western triads!

I know you may be thinking that I am speaking too generically, and that Harmony involves chord progressions and root movement..but I would disagree. The Western practice of creating movements within harmonic textures is really an adaptation or construction based on the more naturalistic concept. That's why we usually need to modify the term when we talk about "functional harmony". Because, really, the functionality is the only differing aspect between the west-centric dichotomy usually labelled tonal / modal. Otherwise, how do we describe this kind of vertically and horizontally intervallic music? And it's not just Indian raga...most pop tunes nowadays have next to no true functional movements!

[JonR]

I think it's our Western aural habits that would cause you to come to that conclusion! I'd say that Indian ragas are purely and unadulteratedly harmonic. Harmony generically refers to the intervallic relationships of tones, like the perfect fifth, as a simple example. Those two tones 'harmonise' together to create an aesthetic effect. This happens most strongly when taking place vertically, so that the texture of the sonic ratio is happening directly...and this can also take place horizontally by a kind of implication. In this sense, I consider "melody" to be a sub category of harmony.

Indian ragas use harmony and melody. The traditional Indian musical set up is designed to create an extremely complex harmonic backdrop, albeit focussed exclusively on the tonic. Just think of a sitar with its sympathetic strings. The chords taking place are naturalistic and complexly based on the melodic content. Physically speaking, these harmonic textures are so much richer and more intricate than those created by Western triads!

Fair enough, but what I meant was there is no theoretical system of harmony in raga, as we understand it. In short, they don't use chords or chord progressions . The harmony that there is is of what we would call the modal kind, non-functional, colouristic.
In a sense, in place of harmonic development (which takes a central place in western music) they have melodic and rhythmic development. The melodic and rhythmic elements of raga are a lot more sophisticated than those in the west. Our melodic and rhythmic sense suffers by being constrained by the demands of our harmonic system (those damn equal tempered triads!).

I know you may be thinking that I am speaking too generically, and that Harmony involves chord progressions and root movement..but I would disagree. The Western practice of creating movements within harmonic textures is really an adaptation or construction based on the more naturalistic concept. That's why we usually need to modify the term when we talk about "functional harmony". Because, really, the functionality is the only differing aspect between the west-centric dichotomy usually labelled tonal / modal. Otherwise, how do we describe this kind of vertically and horizontally intervallic music? And it's not just Indian raga...most pop tunes nowadays have next to no true functional movements!

Well yes, the concept of "harmony" maybe needs definition before we start discussing it!

Obviously, when more than one note occurs at the same time - or even consecutively in melodies, implying a logical relationship between them - we can say some kind of "harmony" occurs. That doesn't get us very far.

The main question is: is that harmony incidental (as it is in most non-western music) or is it carefully controlled according to some theoretical system?
Obviously the western system is just one possible system. It has some basis in natural acoustics, but not enough to claim any universality of relevance.
So the next question is: what (if any) other harmonic systems exist in other cultures?
Are they simply modal - melodic intervals with a drone root note - or are there other kinds? Comparable - or not - to the European classical system?
AFAIK, raga doesn't have this kind of theoretical underpinning to its harmonic effects. (I'm happy to be corrected if you have some good reference links.)

There are some aspects of music that we tend to regard as incidental (not worth theorising about) in western music, such as timbre or instrumental tone. Classically, we tend to aim for "pure" tones, according to some vague aesthetic principles, but there's not really any system to it, or much attention paid - other than trying to "make things sound nice". But timbre is an important expressive element in other cultures, such as Africa and the Far East. (I'm not sure about the buzz of the sitar, whether that's a consciously thought-about and controlled thing, or as incidental as timbres are in the west.)

In rock music too, however, timbre is fundamental - or we wouldn't regard guitar distortion as so important - but there is no theoretical system supporting it.
We (or some of us!) still feebly try to analyse rock according to conventional harmonic theory. And sometimes it works, up to a point - because rock still does instinctively follow some tonal principles (mixed up with equally instinctive modal principles). But there is obviously a lot of other stuff going on which is not explainable in those terms, even though we take it all for granted. Eg, why does rock have to be so loud? Why are the drums always played the same way? Why are the male vocals typically so high register?
These are obvious stylistic rules, but without any theory behind them. (Of course I'm not saying all music MUST be supported by a theoretical system. Only that it's interesting that there isn't one; no one yet - AFAIK - seems to have thought it worth developing one.)

[jimc8p]

Fair enough, but what I meant was there is no theoretical system of harmony in raga, as we understand it. In short, they don't use chords or chord progressions .

Have you ever tried following a tuning for the sympathetic strings on a sitar? It's an absolute nightmare! These strings certainly aren't for melody, but are there to provide the harmonic texture of the raag. Surely all of the formalised raags and tunings would class as a 'theoretical system'? It might not be a system of functional harmony, but it is a system of harmony, or at least modal harmony..

is that harmony incidental (as it is in most non-western music) or is it carefully controlled according to some theoretical system?[...]AFAIK, raga doesn't have this kind of theoretical underpinning to its harmonic effects.[...](I'm not sure about the buzz of the sitar, whether that's a consciously thought-about and controlled thing, or as incidental as timbres are in the west.)

I don't think you could describe the harmony in Indian music as 'incidental'. It is clearly designed and controlled to a high degree. The exactness of the systems used for sympathetic strings is probably the most explicit example. The textures of vertical harmony are consciously utilised in this way, and it's also worth remembering that the drone outwardly played contains notes other than the tonic.

This is obviously a different conception of 'chords' from what we are used to, but the point I wanted to make is that Indian music is not just single notes, unisons, melody or counterpoint. It is very intentionally using the physical textures of many tones interacting both horizontally and vertically. There are roots, fifths, thirds, sevenths, seconds etc sounding together - not just a drone and a melody.

Classically, we tend to aim for "pure" tones, according to some vague aesthetic principles, but there's not really any system to it, or much attention paid - other than trying to "make things sound nice". But timbre is an important expressive element in other cultures, such as Africa and the Far East.

I don't know, I'd say timbre or 'tone' is very widely considered and even formalised in classical music. That's really the main reason why there are so many instruments - to give a timbral palette.

[JonR]

Have you ever tried following a tuning for the sympathetic strings on a sitar? It's an absolute nightmare! These strings certainly aren't for melody, but are there to provide the harmonic texture of the raag. Surely all of the formalised raags and tunings would class as a 'theoretical system'? It might not be a system of functional harmony, but it is a system of harmony, or at least modal harmony..

OK, you know more about this than me, and that sounds completely correct.
I didn't mean raga has no theoretical system - only that (to the extent of my sadly limited knowledge) its theoretical system is concerned primarily with tunings, scales/modes and rhythmic "tal" patterns. I'm not aware of any aspect of that theory which addresses the mix of pitches that sympathetic strings produce. Except I can see that it's probably a mistake to try to separate the melodic aspects of a mode from its harmonic content...

Let me retreat to my "but like, they don't play no chords, dude!" position...

I don't think you could describe the harmony in Indian music as 'incidental'. It is clearly designed and controlled to a high degree. The exactness of the systems used for sympathetic strings is probably the most explicit example. The textures of vertical harmony are consciously utilised in this way, and it's also worth remembering that the drone outwardly played contains notes other than the tonic.

Well, I know it contains the 5th, which merely supports the root. What else (typically)?

This is obviously a different conception of 'chords' from what we are used to, but the point I wanted to make is that Indian music is not just single notes, unisons, melody or counterpoint. It is very intentionally using the physical textures of many tones interacting both horizontally and vertically. There are roots, fifths, thirds, sevenths, seconds etc sounding together - not just a drone and a melody.

Point taken.
All I would say - and it's not an argument - is that the harmony doesn't change. It's a background of various tones, that might resonate differently in sympathy with different melody notes, but as a "colour wash", not as a clear sequence of different harmonic entities, different selections of notes.
While I stand corrected that there IS harmony in raga, I would still see it as a melody-led system, kind of trailing harmonic echoes and modal colours behind it.
I accept that the fact that "they don't play chords" does not mean that they are not concerned with harmony in the broad sense (mutual resonances of different pitches).

I don't know, I'd say timbre or 'tone' is very widely considered and even formalised in classical music. That's really the main reason why there are so many instruments - to give a timbral palette.

Yes, but specification of timbre is mostly a matter of orchestration, of choosing instruments; just occasionally of indicating timbral variation (such as "pont." for string instruments).
My point was that our "music theory" has no interest in this, no terminology for it; it only concerns itself with harmony, rhythm and form, and to a lesser extent with melody.

[mwtzzz]

Except I still don't like the idea of tones you "cannot" play.

Yep, actually avoid tones can be played with certain qualifications. But for learning how to improvise (I think that's the original post?) it's best to just "avoid" the avoid tones.

And then there was Hal Galper, whose rant at the beginning of this video almost made me jump out of my seat with excitement:
http://www.youtube.com/watch?v=NehOx1JsuT4
- the veil fell away: dammit the guy is right!

When you hear improvisation, there's basically three different kinds. There's rhythmic improvisation with chords, a few jazz musicians only do it like this. There's patterns using whatever pattern machine the improvisor has decided. And there's the construction of an alternate melody using the fulcrum points of the original melody.

There's also a fourth kind, which is not really improvisation, but the untrained ear and the novice musician thinks is improvisation: doodling. This is basically random stuff that sometimes sounds okay because it just by chance happens to hit upon something.

in all cases, you need to know which tones work and which tones don't work.

What Hal is talking about is the importance of constructing alternate melodies. he's absolutely right that if you cannot improvise slowly, you have no business trying to improvise quickly.

Who was a master of this on piano? John Bunch was one.

[JonR]

Yep, actually avoid tones can be played with certain qualifications. But for learning how to improvise (I think that's the original post?) it's best to just "avoid" the avoid tones.

Agreed - slightly reluctantly, but it's a fair point!

When you hear improvisation, there's basically three different kinds. There's rhythmic improvisation with chords, a few jazz musicians only do it like this. There's patterns using whatever pattern machine the improvisor has decided. And there's the construction of an alternate melody using the fulcrum points of the original melody.

There's also a fourth kind, which is not really improvisation, but the untrained ear and the novice musician thinks is improvisation: doodling. This is basically random stuff that sometimes sounds okay because it just by chance happens to hit upon something.

Right, that's a nice way of looking at it.
Of course, ideally, the first three blend together, they're just different aspects of the same thing (IMO).

Might be worth mentioning the UK Trinity improvisation sections of the grade exams (optional supporting tests). They come in two kinds (at least in lower grades), roughly aligning with the 1st and 3rd kinds you list: there is "rhythmic", where the candidate is given a simple chord progression; and there is "melodic", where they are given a handful of notes (not a whole scale) and have to invent a few bars of melody using them.
This is not for jazz candidates, mind; it's an option for the classical guitar grades, right from grade 1, which I think is interesting. (Good to see it being encouraged, no matter what genre you're learning.)
In fact, even in the classical theory grades, you are expected to be able to improvise answering phrases to a given couple of bars of melody.

Although the fourth kind seems dumb in comparison, it's still a very valuable way of playing. It's arguably not "improvisation" at all (if we define that as "based on given material", which is the traditional jazz definition), but it's a useful journey of discovery, and is the way many composers work.
When jazz musicians play that way, they call it "free", or "free improv" - although those who play that way all the time just call it "improvised music".

in all cases, you need to know which tones work and which tones don't work.

Well, not in the fourth kind, because you discover that by trial and error, based on your aural awareness of musical grammar (which we all have, assuming we've heard enough music in our lives). You don't know which notes are wrong or right until you hear them, but when you hear them you know. The "theory" is then a system developed to make logical sense of it (as far as possible). Eg, "OK, that note sounds bad, and I see it's the "4th" on a "maj7" chord... therefore I'd best avoid 4ths on maj7 chords..."

And there's a certain amount of trial and error in the first three as well (with the arguable exception of the 2nd kind, the prepared patterns), although of course there's plenty of given material to guide you.

I could also argue that the concept of what "works" can be quite a personal one. It means "sound good", and there can be elements of personal taste in that. (Maybe we sometimes like playing the 4th on a maj7 chord...)

Then again, we are always working within particular genres when improvising, and there are stylistic rules governing what sounds "right" and "wrong" for that genre. What "works" in rock may not "work" in jazz, and vice versa. (The maj7 chord itself is standard in jazz, hardly thought about, but a strange and very specific sound in rock.)

What Hal is talking about is the importance of constructing alternate melodies. he's absolutely right that if you cannot improvise slowly, you have no business trying to improvise quickly.

Indeed.
It's always possible to play faster than your inventive skills allow, but you produce garbage that way.

In fact, that's an inherent problem with what you call the "pattern machine" method, if I understand that right. If that's used exclusively - whether it means tried-and-tested licks, or chord-scale orthodoxy - simple practice of the patterns over and over means you can end playing faster than you can think (if you even think at all). So your lines may always work (if you've done your homework properly), but it's hardly "improvisation"; it's more a demonstration of instrumental skill. It makes improvisation like an assault course, a challenge one has to get through unscathed. In the frantic effort to avoid mistakes, the "music" kind of disappears...