Keep Me In Mind

[LaughingSkull]

Here is the lead sheet of John Scofield's 'Keep Me In Mind':
http://www.oocities.org/bourbonstree...epMeInMind.pdf

Since I know what I could/should play over each chord I can solo over the progression OK. But ...

I tried to understand to role/function of each chord, but I am clueless in this piece. Any insights?

[Color of Music]

Well, did you determined the key? I have not heard the song, but was able to analyze it.

One thing I notice is that I had to use enharmonic spellings to match the key signature.

The piece is in E major (I knew this from the II-V-I (the first three/four bars.

So, we have:

II-V-I (First four bars)

i-VI-bVI (Next four bars)

Then, we go back to II-V-I-VII (VII due to the enharmonic re-spelling) (Next four bars)

III-bIII-bVII-I ... ii-I (ellipses = same quality and scale degree)

II-(or V/V)-V-I as with the first four bars

i-vi-bVI-IV ... (the last four bars)

(Note: I did not wholeheartedly write any secondary dominants although they are being used). However, I will explain them.

So, now to understanding the functions:

II-V-I: This is a very common progression (ii instead if II though) To explain the II though. Like with the ii, if we were to add the b7 scale degree (F#7 in this case), it still leads to the Bm9 chord (A#-->B; E-->D and F#-->B)

i-VI-bVI: Both Major six chords imply Secondary Dominants - this is because VIs are minor in the major scale; however, they are major in the parallel minor scale; therefore, they work because they are borrowed chords.

In the major scale - the second, third and sixth degrees are minor where they are diminished (ii) major and major in minor keys. (I think I'm wrong on this part)

Anyway, when you see a VI chord - especially when it's a dominant, extended, or altered dominant - secondary dominants are noted.

C-Am-Dm-G7-C (I-vi-ii-V-I)

CMaj7-A7b9-D7-G7b9-C - This is still I-VI-II-V-I; however, to the classical composer/musician, it would be written as such:

I-V7/ii-V7/V-V-I.

I couldn't grasp this at first, but all one has to do is this:

"Look to the right. Look to the left." A V7/ii in C means, what is the V7 chord built on D minor? Run the D minor scale: D-E-F-G-A-Bb-C-D, then, once you find the indicated note, look to the left, then ask yourself, what is the V7 of D? Run the D Major scale: D-E-F#-G-A-B-C#-D. Where do you land? A. Build the dominant seventh from that root. (Modes work in this fashion, believe it or not kind of. G Mixolydian = G major scale + b7 = G7 chord = V7/ii in Bb; Look to the right - Cm; look to the left - V7 from Cm = G7)

That way of remembering is alot easier than how it is explained - though I had to understand it that way first.

From the way it's traditionally explained and demonstrated is going up the major/minor scale with transitional chords in-between. ie: Going from I-ii, and V7/ii is played in the middle.

You get this effect with the melody note:

F (V7/V) E (I) G (V7/ii) F (ii) A (V7/iii) G (iii) V7/iv (Bb) IV (A) V7/V (C) V (B) V7/vi (D) vi vii (V7) (B) I (C)

Here's the harmonic analysis: G7(Bdim)-C-A7-Dm-B7-Em-C7-F-D7-G-E7-Am-Bdim(G7)-C.

Notice from a circle progression standpoint, the use of secondary dominants displays such a motion: (V/vii)-I-VI-ii-VII-iii-I-IV-II-V-III-iv-vii(V7)-I. Not quite circular as with a tune like Autumn Leaves: i-iv-VII-III-VI-ii-V-i, but you get that same "step down" idea with in the harmony (Bb-A, A-G, G-F#) and the melody does this anyway (Eb-D-C-Bb)

Now back to analyzing the song:

The bVI is a scale degree borrowed from the minor scale - you technically could apply a secondary dominant to this as well (bVI = bV7/ii - how frequently this is ued, I don't know. Perhaps someone can clear this up.)

The A Major scale: A-B-C#-D-E-F#-G#-A; the bVI comes from the three A minor scales: - A-B-C-D-E-F-G-A (Natural minor) A-B-C-D-E-F-G#-A (Harmonic minor) and (melodic minor going down which looks exactly like A natural minor)

As noted above, with the Eb chord being called a VII - Eb is the enharmonic of D#. (There is no rule that says to stick woth the key signature; however, it helps to at least interpret note names as such to make analysis easier.) D# is the VII scale degree of E. (Eb in E would be a bI. Using this naming would be rare)

Phrase Four:

III (Enharmonic spelled - G# instead of Ab = bIV), then, the bIII which like the bVI and bVII is borrowed from the parallel minor.

III-bIII-bVII-I-ii-i (G#-G-D-E) - The ii(7)-i part of this phrase: the ii7 is functioning as the V (a V9sus = B9sus), so that is why the ii-i in this phrase is "cadential." (F#m7 = B11 without the B or A/B (A chord in the right hand. B chord in the left) B9sus-Em is V-I, but the ii7 (with B in the bass) works because it's a substitute.

Now, with the last phrase: i-vi-bVI-IV.

First off, the i comes from again, the parallel minor. It would appear that he is tonicizing this key. This means staying here temporarily.

The bVI has already been explained. (Borrowed chord)

The vi comes from the relative minor of the home key (E major). This is the parallel-tonic relationship. Due to this, we can sub the vi for the I at any time.

Think of Parallel Tonic as being similar to Major and relative Minor keys. (C/Am, Eb/Cm. F/Dm, Cb/Ab, etc.)

The IV: It has two routes: it either goes to the V chord (in C; F-G) or the I chord (F-C), given you a Plagal cadence. This chord while major in the appropriate scale can also be minor (thus making it a borrowed chord, too. iv-bIII-bVI-bVII). In this particular song though, it is major.

Does this help?

[LaughingSkull]

Thank you. I will reply further when I manage to digest the post.
I guess it will take me a while to go through it ...
In the meantime, here is the link to the original version by John Scofield:

http://www.youtube.com/watch?v=a8wb-HfFaXw

[LaughingSkull]

II-V-I (First four bars)

Ok, I get that. I feel a fool, but I thought that it could be ii V I or ii V i. It's the first time that i see it as II v I. But obviously it works.

i-VI-bVI (Next four bars)

Ok, I understand C69 is borrowed from Em, but VI should be C#m and here we have C#m7b5.
Why is this substitution possible? Borrowed from D major? Why?
Perhaps it would make more sense to me if I spell it as Em6/C#.

III-bIII-bVII-I ... ii-I (ellipses = same quality and scale degree)

I don't understand this part. (Ab7 Gmaj7 ...) If G#/Ab is borrowed from the minor where there is G ...? Why?
But i followed the trail of secondary dominants (hopefully correct) and this (Ab7 Gmaj7 ...) might be spelled as V/vi and V/bVI?

Does this help?

I think I might have bite more than I can chew at the moment. But it's the only way to learn ....
Anyway, it is a beautiful piece and I love playing it. I hoped I might learn a bit more theory while I do it.

[JonR]

I'm going to offer a (slightly) alternative view.

Firsly, this is a post-tonal piece of jazz: informed as much by modal practices as key-based ones. As such, many of the chords may not make much sense in functional terms. However...

F#/A# - Bm9 = V-i in B minor (pretty clear, IMO);
Bm9 - Esus/E = ii-V in A major - or at sounds like that, IMO.

Em11 - C#m7b5 - C69. This is an intriguing sequence. The A melody note on C#m7b5 is going to make it sound like A9, which suggests a ii-V in D major (or ii-vii if you like), but of course the C69 is a deceptive cadence. The clue may be in the voice leading: D (on Em11) > C# > C, or D > C# > D (9th of C). Otherwise all 3 chords share a lot of notes (E, G, potentially B, and A if we include the melody).
Another idea is that C69 has a lot in common with Am (Am11); but really it's a quartal chord, with a modal flavour (IMO).

Bridge:
F#/A# - Bm7 - Esus - Eb7#9. As first line, but the Eb7#9 is a cool harmonisation of the F# (I might steal that idea ). IMO, functionally it works a tritone sub for A13(b5), making the Esus-A13 a II-V in D major (again ).
But of course it doesn't resolve to D; the Eb7 becomes V of the following chord (Ab7) - which is itself a tritone sub for D7, resolving to Gmaj7.

Esus-E. This doesn't sound like a resolution to E major to me, more like a dominant in A major (particularly as it's preceded by D/F#). The climb at the end of that line (F#m-E/G#) also leads the ear to expect A, but instead it goes back to F#/A#.

In short, this tune is not in any specific key. The chords tease the ear with functional suggestions, but don't resolve anywhere convincing. Even the opening V-i in B minor is subverted by the 7th in the Bm.
The Gmaj7 sounds briefly like a (secondary?) tonic, but that's abandoned straight away. And the frequent Esus-E motifs suggest an A key centre to me.
Oddly enough, playing it through, the C69 sounds most like a tonic chord to me. Even though the cadence is unorthodox, my ears accept it - helped perhaps by the fact that it occurs exactly where we'd expect to find the tonic chord: bar 7 of the A-sections.

Beautiful tune, btw! Reminds me a little of Steve Swallow's "Falling Grace".

[JonR]

[I]Ok, I understand C69 is borrowed from Em, but VI should be C#m and here we have C#m7b5.

C#m7b5 is diatonic to E melodic minor (if you want to think E key centre, which is by no means certain).

Why is this substitution possible? Borrowed from D major? Why?

Why not?

Perhaps it would make more sense to me if I spell it as Em6/C#.

Yes. Or A9/C#, given the melody note.

Anyway, it is a beautiful piece and I love playing it. I hoped I might learn a bit more theory while I do it.

IMO, the theory involved is centred around the voice-leading and what might best be called deceptive cadences.
In particular, think tritone sub (as I mentioned above) for the Eb7 and Ab7.
But don't get tied up with classical functional theory. This tune is beyond that.

[LaughingSkull]

Thank you Jon. That was illuminating.
I guess I tried to over-think it.
On the other hand, this piece demonstrated to me, how much cooler it is if you compose melody first and the harmonize it instead of vice versa.

[Color of Music]

Thank you Jon. That was illuminating.
I guess I tried to over-think it.
On the other hand, this piece demonstrated to me, how much cooler it is if you compose melody first and the harmonize it instead of vice versa.

Not that you can't compose from harmony first.

It's something that seems to be "taboo." Why? I don't know because you can sing a melody over chords, just as easy. The trick is finding where your melody note is in the chord - and this generally at the top.

C-G7-G7-C-C7-F-C-G7-C.

Do you recognize this song?

I realize it could be any song, but what's the most obvious?

Nowadays, with songs that have the same progressions, how melody flows is about the only way to distinguish them (apart of the lyrics); however, you can compose from a harmonic perspective still. And you've got many more melodies to choose from:

If you're harmony is triads, three notes - Melody from the root, third or fifth. If a seventh, the melody could come from either of those four. Five choices if it's a ninth.

It's still Happy Birthday, if I write:

D-D#-E-E-F#-F / F-F-G-F#-F-E / E-E-G-G-G#-G-F / G#-G#-G-F#-F-E

The melody is next up from the bottom, but it is still a melodic line.

Gm-Eb-Bb-F

Love The Way You Lie - G-G-F-D-G-F-D-G-F / Eb-Eb-Eb-D-D-Eb-Eb-D-D-C-C

Love The Way You Lie II - F#-G-A-Bb-Bb-F-Bb-A-G / Eb-G-Gb-F-F-Gb-G-D-F-G-A

Again, the first string is the actual melody, but it doesn't have to be. What I did was pick out less obvious notes. The second string still comes from those chords.

And notice how Lie II has more melodic movement.

It's not so cool as it is so easy.

[LaughingSkull]

Not that you can't compose from harmony first.
It's something that seems to be "taboo." Why? I don't know because you can sing a melody over chords, just as easy. The trick is finding where your melody note is in the chord - and this generally at the top.

Agreed. Most of my compositions starts more or less with harmony first. Then melody screams to be discovered.
I would say that both approaches have strong and weak points.
Since I compose mostly prog rock/metal/fusion, I use hybrid approach (best of both worlds) or different approaches for different parts of composition.
Whatever works best in given situation.

Anyway, jazz improvisation is instantaneous composition over given harmony, isn't it?

[Color of Music]

Agreed. Most of my compositions starts more or less with harmony first. Then melody screams to be discovered.
I would say that both approaches have strong and weak points.
Since I compose mostly prog rock/metal/fusion, I use hybrid approach (best of both worlds) or different approaches for different parts of composition.
Whatever works best in given situation.

Anyway, jazz improvisation is instantaneous composition over given harmony, isn't it?

Oh, certainly. Use a hybrid method. And yes, but the melody is not neglected. Even if you don't write a melody, at least have an idea, so the harmony can flow melodically.

Voice-leading is where such an approach is based. This is also an excellent way to figure out progressions.

Am-Dm-G7-CMaj7-FMaj7-Bm7b5-E7b9-AMaj7

E-F, D-E, C-D, B-C#

i-iv-VII-III-VI-ii-V7-I

[JonR]

Thank you Jon. That was illuminating.
I guess I tried to over-think it.
On the other hand, this piece demonstrated to me, how much cooler it is if you compose melody first and the harmonize it instead of vice versa.

Yes!
Writing chords first is like designing a suit of clothes and then looking for a person to wear them....
(It could work, but you'll likely need to make some alterations, so why not start with the person?)

[JonR]

Not that you can't compose from harmony first.

It's something that seems to be "taboo."

I don't get that. What I see is a lot of recommendations to write melody first, but only in a context where most people seem to write chords first (and then get stuck).

IOW, chords first is not "taboo" - it's just often not very efficient. It doesn't get you very far, because there are always too many options at every change, and no way of deciding.

I don't know because you can sing a melody over chords, just as easy. The trick is finding where your melody note is in the chord - and this generally at the top.

Yes, but that depends how you voice your chords. And on guitar, the top note of most chords is beyond most male voices, so we tend to sing notes from the middle of the chords if we are singing along.
Of course, I'm generalising...

C-G7-G7-C-C7-F-C-G7-C.

Do you recognize this song?

No.

I realize it could be any song, but what's the most obvious?

Nothing is "obvious" to me about that (at least, as there is no guide to metre or timing). It's a stock progression used in dozens, maybe 100s of songs.

Nowadays, with songs that have the same progressions, how melody flows is about the only way to distinguish them (apart of the lyrics);

Yes, and I think that's one of the arguments against composing from chords. One tends to end up with familiar formulas. Eg, that ubiquitous I-V-vi-IV sequence...

But you're right that many different melodies can be composed over the same sequence. Chords-first doesn't have to be a limitation, if you're melodically imaginative.
Sometimes, in fact, a chord sequence is a very useful inhibition on melodic possibilities, if your creative inspiration is too wild and needs reining in.

The only rule, really, is that melody has to govern the process. Melody is primary, and chords support (and enhance) it. You can start with chords, but you have to let any melody that emerges take control, and change the chords if the melody demands it. (And even then, chord changes can inspire melodic moves, like a set of advisors to a political leader...)

It's still Happy Birthday...

Ah, it was Happy Birthday! (Yeah OK, it's obvious NOW....)

[Color of Music]

Yeah, taboo was a bad choice of words, but From The Bottom Up is not often suggested. I think this has to do with - not so much what is easier, but what you're tuned into more. For me, it's harmony.

I understand, it's hard to extract one note when four or perhaps five are being played at once. Again, though, harmony needs a melodic structure to pull out a melody.

Db-Bb9sus-Bb7b9-GbMaj7-Gbm7-Fm7-Bb7b9

Ebm7-Ab7b9-Fm7-Bb7b9-Ab9sus-Ab7b9-Db6

Fm7-Fdim7-Ebm9-Ebm7-Eb9-Ab9sus-Ab7

Fm7b5-Bb7b9-Ebm9-Eb7-Eb13-Eb9#5-Ab9sus-Ab7b9

DbMaj9-Db6-Abm9-Db13b9-Gm7b5-C7b9-Fm7b5-Bb7b9

B13-B9/C-DbMaj9-Ddim-Ebm7-Ab7b9-Db

What's this?

Yeah, you're right on the last part. Well, everything you said really; however, you necessarily don't need it if you understand how progressions work.

It's easy to say:

Given the above example:

My melody starts on Db, so write that chord, the next note is Ab, so write that one, etc. Don't get me wrong, I did this, but I had the tune in my head which I why I was able to and do so mentally.

But I could have easily extracted that note - which in effect - you have to do that - to see if it fits.

It just dawned on you what the last sequence was (haha) Seriously, often times, pieces are found out from melodies - but bits and pieces of it. (E-C-E - with just three notes, it could be anything, but put Am9-Am-E7#5#9-Am9-Am - and more than likely you'll say it's Summertime)

The familiar formula thing has to do with the limitation of there being twelve tones/chords and chord movements. (ie: the I564, example you posted, but also 1625, 1645 (in Major Keys) 1451 + 1637 (in minor keys) And of course, they can get longer until eventually you end up going around the circle. Often times, music goes halfway.

I'm not disagreeing that you need an idea though and do agree with the formulatic chord sequences, but again, chords only have so few places to go and that's why - this is not even with the melody considered. The melody in a way restricts the harmony.

[Don't tell me to go to F! I think I know since I'm on a C7!] It doesn't have to go there, but it certainly desires too. And given that it's a V-I ...

["The melody won't tolerate it, so I can't do it." Um, it's a two-way street or at least I hope it is.] That's another reason for the formulatic sequences. Of course, the lament is coming from an arranging standpoint, but it applies to original compositions also.

[And what's this thing called Voice-leading? Never heard of it.]

Brackets is sarcasm.

But wait, you can easily end up with the same tried and true sequences if/when the melody demands it.

Dm-Bb-F-C (vi-IV-I-V)

F-F-F-D-F-E-D-C-F-E-E

D-D-D-Bb-D-C-Bb-A-G-G

Same chord sequence (and yes, I know I'm harmonizing the melody, but that isn't the point).

The upside as I said before, is that you have more options for melodic notes; you aren't limited to only the top note. Barring cadential guidelines regarding voice-leading [Again, I've never heard of this concept]

And even while using substitutes, it can still seem formulatic - due to the movement limitation.

D7-Gm7-C13sus-A7 (V7/ii-ii-V-V7/vi; VI-ii-VII-III)

See? I could only move the same way - though I changed chords when asked to.

I gave the example of Love The Way You Lie and Apologize can fit, too.

Both in a minor key with this harmonic move: i-VI-III-VII.

Sure, the melody dictates it, but again, this is where they like to move. (Of course, this just one way)

Here's Love Song (Sara Bareilles)

Gm-C7-F-Bb2-D7 (not played, but implied by the bass). ii-V7-I-IV-VI(V7/ii)

This is why functional subs exists (Tritone, Tonic-Parallel, Inversion, Common Tone, Color/Added Tones. Alterations, Extensions), so you won't get the same overused progressions; this doesn't mean that the functional subs get rid of them - they just hide them.

This is why arrangements make songs sound better. (That's subjective, of course)

The intent is not to bury the melody, but lots of techniques are used to spice it up as well.

And sometimes there's lots of dead space between the melody (and/or harmony), so fill it in (singers do this very often as well as improvisers).

Makes you wonder when you look at classical music with tons of 32 notes ... (that is not a slight on such music), makes you wonder which method they choose.

So, yeah, either way is fine, but sometimes trying both methods really help.