Pedagogical question - playing over non diatonic chords day 1.

[Acci]

This has been covered to an extent on this board, but I was wondering if we could get really specific and practical...

What is a good method to use in this instance:

Say I have a student (guitar) that is familiar with major and minor keys and arpeggios.

When we encounter, say a secondary dominant, what is the first thing to do, you think, to have him be able to improvise over the chord?

There are obviously ten thousand things he could/should do and there are ten thousand things I have personally done. What do you think is first? My own path wasn't direct at all and would not be one I'd advocate to my students.

I feel like a common suggestion I've heard is to think of the parent key but make the necessary alterations for the different chord. But this can be problematic for a few reasons.

For example, say we are in the key of C and have an D7. The new note introduced is F# - how to explain the difference to the student as to why we might raise the F from the key of C, rather than flatting the G? The enharmonics are obvious in this example, but in a different key it might not be so clear cut. We could say the D7 is borrowed from key of G...ok. If we have E7 in the key of C we don't say it's borrowed from the key of A major, we'd probably say it's from A (harmonic) minor...then there are still holes in the explanation.

I suppose part of my question is, how much ear training and harmonic theory does a student need to have before he can, in the context of a lesson or a classroom, be taught how to improvise over secondary dominants or other non diatonic chords in a key?

Thanks in advance...

[jazz oud]

Hey Acci,

The first thing I would say is that, enharmonics aside, either raising the F to F# or flattening the G to Gb would work just fine. In fact:

D E F F# A B C

is a pretty interesting pitch set and has no strong dissonances. Also, having an 8-note scale is not a problem, either: D E F F# G A B C

That said, I am having a hard time thinking of an instance where a secondary dominant where the original chord is minor doesn't obviously involve sharping the 3rd rather than flatting the 4th (in the most direct diatonic appraoch). It doesn't really matter what the key signature is, especially on guitar, where if you know your chord forms, it is pretty obvious where the m3 is and that it is the note that will be raised. But on III7, flatting the fourth gives a great alternative pitch set (1 b2 b3 3 5 b6 b7).

The only case that might be confusing is a "secondary" dominant that is two keys away, like F#7 in C major is really from B minor, which would be from the key of G, making it the V/iii/V, requiring altering more than one note.


Theory aside, I think a good approach is to focus on specific examples and on targeting chord tones. Rather than focusing on arpeggios (which lead to patterns that students often don't understand the contents of), focusing on individual chord tones avoids a lot of worrying about keys and scales.

I've thought quite a bit about this, and in the simplest terms, you can ascertain a logical pitch collection over any functional chord as follows:

Starting with 1357
If the chord is maj7: add 2, 4, 6
If the chord is m7: add 2, 4, m6
If the chord is 7: add 2, 4, 6 if the target chord is major, or b2, 4, b6 if the target chord is minor. For more advanced players, #2, and #4 can be incorporated in various combinations to create altered or diminished fragments.
If the chord is ø7: add b2, 4, b6 (2 is an alternative, especially when targeting major keys)
º7 is the only chord that gets tricky, and the easiest approach is to learn the standard situations one by one (iii biiiº7 ii, IV #ivº7 I, I #iº7 ii, ii #iiº7 iii, V #vº7 vi) rather than trying to theorize a general approach.

Now, these will not necessarily be the most diatonic pitch collections (differentiating between vi, iii, and ii would be a good start, as well as using #4 on IV), but they will work fine especially if the focus is on targeting the chord tones.

Regarding the notion of secondary dominants being borrowed from keys, I think most students can understand the following:

If the target chord is major, the secondary dominant is borrowed from that major key (IV and V)
If the target chord is minor, the secondary dominant is borrowed from that minor key.

If the target chord is ambiguous (say, IIImaj7, or a string of secondary dominants), then the default is what the native chord would have been, but the alternative can work also . . . trust your ears and experiment.

The problem with all of this with most guitar students is that it requires knowing the notes on the instrument pretty well, a weakness for most beginning and intermediate guitarists.

[engelbach]

Originally Posted by Acci

When we encounter, say a secondary dominant, what is the first thing to do, you think, to have him be able to improvise over the chord?

The chord that follows determines how one should play over the secondary dominant.

I'd tell a beginner to think about targeting the next chord through the leading tone.

So in C major, on an A7 chord target D7 or Dm7 through the C#.

That doesn't preclude the playing of a D note over A7 (as part of the C major scale), as long as the end of A7 clearly leads to a D chord through C#.

[Acci]

Thanks guys for the fast responses!

That all makes a lot of sense.

I think the key ingredient that Jazz Oud mentioned, that takes away my potentially confusing point, is that we are turning what normally is minor into a major. So a student can clearly see that the diatonic chord is minor, the adjustment is raising the b3 of the chord back up the 3. And yeah, including the b3 and turning the 4 into natural 3 works great as well...I can see it potentially getting confusing to a 'beginner.'

I guess after hearing those considerations, next time I'm in the situation I'll refer to that point. This also points out the answer to another question in the original post: to understand secondary dominants well, the student has to understand the diatonic chords to the key very well.

I suppose another simple way of looking at it is that a secondary dominant is the V7 of either a major or a minor, so the most conventional choice points to either considering the target's major scale or harmonic minor scale.

I've thought quite a bit about this, and in the simplest terms, you can ascertain a logical pitch collection over any functional chord as follows:

Starting with 1357
If the chord is maj7: add 2, 4, 6
If the chord is m7: add 2, 4, m6
If the chord is 7: add 2, 4, 6 if the target chord is major, or b2, 4, b6 if the target chord is minor. For more advanced players, #2, and #4 can be incorporated in various combinations to create altered or diminished fragments.
If the chord is ø7: add b2, 4, b6 (2 is an alternative, especially when targeting major keys)
º7 is the only chord that gets tricky, and the easiest approach is to learn the standard situations one by one (iii biiiº7 ii, IV #ivº7 I, I #iº7 ii, ii #iiº7 iii, V #vº7 vi) rather than trying to theorize a general approach.

Is this different from basic chord scale assignment just in the language that is used to express it? Encouraging the student to think of chord tones and then the holes in the pitch collection, rather than an entire pitch collection matching the chord?

--

I've been teaching guitar for about fifteen years and have only had a very small handful of students who were even interested in scratching the surface of playing over changes. Like I wrote before, my own path was pretty haphazard but I got there. It's certainly not a good model, at all.

Seems there are so many considerations, so many things to be aware of, leaves me not totally sure of the best place to start, especially if the student doesn't have good enough ears to help guide them through. I suppose maybe there isn't really a 'best' way to start...there are a lot of educational materials and even college curriculums that tackle the issue in different ways.

Maybe another thing that could be primary is, if we're talking about secondary dominants or borrowed chords, hopefully it is because the student is listening to some music that has this kind of harmony...we can analyze lines from the recordings and go from there.

[jazz oud]

Originally Posted by Acci

Is this different from basic chord scale assignment just in the language that is used to express it? Encouraging the student to think of chord tones and then the holes in the pitch collection, rather than an entire pitch collection matching the chord?

Yes and no . . . noting the functional relationship to the chord tones suggests that aeolian is a better "safe" choice over minor than dorian, and ionian a better choice than lydian. Because the idea is to target chord tones rather than having a scale in which all the tensions can also function as extensions, you end up with a little different results.

It's not an "either or"; but a sort of quick way of getting a sense of functional lines with minimal theory or thinking.

[JonR]

Originally Posted by Acci

This has been covered to an extent on this board, but I was wondering if we could get really specific and practical...

What is a good method to use in this instance:

Say I have a student (guitar) that is familiar with major and minor keys and arpeggios.

When we encounter, say a secondary dominant, what is the first thing to do, you think, to have him be able to improvise over the chord?

There are obviously ten thousand things he could/should do and there are ten thousand things I have personally done. What do you think is first? My own path wasn't direct at all and would not be one I'd advocate to my students.

I feel like a common suggestion I've heard is to think of the parent key but make the necessary alterations for the different chord. But this can be problematic for a few reasons.

For example, say we are in the key of C and have an D7. The new note introduced is F# - how to explain the difference to the student as to why we might raise the F from the key of C, rather than flatting the G? The enharmonics are obvious in this example, but in a different key it might not be so clear cut. We could say the D7 is borrowed from key of G...ok. If we have E7 in the key of C we don't say it's borrowed from the key of A major, we'd probably say it's from A (harmonic) minor...then there are still holes in the explanation.

I'm with Jerry here, in that I see the simplest explanation as the correct theoretical one: the purpose of the chord is to provide a leading tone to the root of the following chord.
And any scale notes outisde the chord arpeggio should (in the first instance) be taken from the following chord.

So the reason for a D7 in key of C (assuming it IS a secondary dominant ) is to lead to G via a half-step from F#.
The relevant scale would be G major, although all the passing notes (non-chord tones) will be same as C major notes of course.

You can talk of the F# as an alteration to a diatonic Dm; that's a simple observation; but it doesn't explain why it's there, nor offer any clue to improvisation strategy.
The dynamic idea is to see the chord as leading to G. So whatever you play over the D7 has to resolve on to the G chord in some way.

I certainly don't think it could ever be a good idea to think of D7 with a Gb in it. (I'm aware that once we start invoking the altered scale, then if we take it from 7th mode of Eb melodic minor, then that note is Gb in Eb melodic minor. But that connection is coincidence. In the context of the chord itself, that notes is the major 3rd, not a diminished 4th, and it makes no sense to regard it as the latter.)

Same with an E7. It has a G# in order to lead to A (root of Am). Again, passing notes will be from the C major scale. That amounts to A harmonic minor, but no need to name that scale, IMO.

If it's A7 leading to Dm, then the same process results in D melodic minor (C major with C raised to C#), again with no need to name the scale.

If it's B7 leading to Em, the chord requires two notes to be raised: F# and D#. Add the remaining C major scale notes and you get E harmonic minor.
Again, there seems to be no need to identify the scale as that. The purpose of those notes is to lead to G and E respectively (better than F and D do).

IMO, any beginner improvisation strategy on secondary dominants would be to use the chord arpeggio only (no passing notes), resolving the leading tone to the root of the next chord. That will underline the purpose of the chord, as well as the essential concept of the half-step resolution, which is what most jazz chord alterations are about. (That can be combined with the 7th of the chord descending to the 3rd of the following one - although that will be a whole step if the target chord is minor.)
A second alternative would be to resolve the 3rd of the secondary dominant down (again by half-step) to the 7th of the following chord.

IMO, such moves are way more important than scale choice on a single chord itself. Or rather, the moves on to the next chord (the voice-leading options) are what governs the scale choice.
The observations about half-step voice-leading will prepare the improviser well for more advanced jazz chords like altered dominants, dim7s and lydian dominants.

Originally Posted by Acci

I suppose part of my question is, how much ear training and harmonic theory does a student need to have before he can, in the context of a lesson or a classroom, be taught how to improvise over secondary dominants or other non diatonic chords in a key?

Thanks in advance...

I'd say that playing those leading-tone>root resolutions is how the ear is best trained to understand secondary dominants. The sound of that is what the theory springs from.

With other kinds of chromatic chords, different strategies might apply. Sometimes it really might be simpler to just point out which notes are altered and fill in with rest of the key scale. But most of the time (in functional harmony anyway) there will be a good reason for a chromatic chord, and that will be something to do with the following chord, and perhaps with the chord before too - eg in providing a chromatic line (or more than one) between the chords either side.

IMO, those voice moves are the first thing to grasp, before asking about what other notes might be added as passing notes between those of the chord arpeggio. Quite often, no passing notes are needed, eg if the SD chord has a 9th, meaning the arp contains 5 notes. Why would a beginner improviser need more than that?

[engelbach]

Excellent post, Jon, which eschews complications that would probably confuse a beginner and clearly explains the purpose of the secondary dominant.

[jazz oud]

Originally Posted by JonR

I certainly don't think it could ever be a good idea to think of D7 with a Gb in it. (I'm aware that once we start invoking the altered scale, then if we take it from 7th mode of Eb melodic minor, then that note is Gb in Eb melodic minor. But that connection is coincidence. In the context of the chord itself, that notes is the major 3rd, not a diminished 4th, and it makes no sense to regard it as the latter.)

Just to clarify, I wasn't suggesting that the student should think of it that way. Since Acci brought it up as a potential confusion for a student who didn't get the theoretical nuances of enharmonics, I was just pointing out that even if they made that mistake, the consequences are not particularly problematic from a practical standpoint. But you are absolutely right that it should be made clear to a student what the difference is and why.

That said, if the student has trouble grasping the distinction at first (as many do), I don't think that it is necessary to suspend the attempts to improvise until they get the theoretical logic.

IMO, any beginner improvisation strategy on secondary dominants would be to use the chord arpeggio only (no passing notes), resolving the leading tone to the root of the next chord. That will underline the purpose of the chord

I respectfully disagree--as I said above, I think that focusing on individual chord tones, rather than arpeggios, is a better starting strategy, and that the tensions are helpful in clarifying the role of the chord tones. But I think we are on the same page, more or less--the amount of information that the beginner is working with should be limited (we just disagree on the definition of the limits) and that scales are not a good starting point.


Your central point, that there should be an emphasis on how the chord tones move from the secondary dominant to the resolution chord and gaining an aural appreciation of the logic therein, is certainly correct IMO.

[JonR]

Originally Posted by jazz oud

Just to clarify, I wasn't suggesting that the student should think of it that way. Since Acci brought it up as a potential confusion for a student who didn't get the theoretical nuances of enharmonics, I was just pointing out that even if they made that mistake, the consequences are not particularly problematic from a practical standpoint. But you are absolutely right that it should be made clear to a student what the difference is and why.

That said, if the student has trouble grasping the distinction at first (as many do), I don't think that it is necessary to suspend the attempts to improvise until they get the theoretical logic.



I respectfully disagree--as I said above, I think that focusing on individual chord tones, rather than arpeggios, is a better starting strategy, and that the tensions are helpful in clarifying the role of the chord tones. But I think we are on the same page, more or less--the amount of information that the beginner is working with should be limited (we just disagree on the definition of the limits) and that scales are not a good starting point.


Your central point, that there should be an emphasis on how the chord tones move from the secondary dominant to the resolution chord and gaining an aural appreciation of the logic therein, is certainly correct IMO.

Thanks j.o. - no disagreement from me.
I'll regard your disagreement as simply a clarification of my poorly worded point - because you said exactly what I meant to say.
(I was simply using "arpeggio" as a collective term for the various chord tones; their individual connections to the following chord is what matters, especially the "tendency tones".)

[michaelsorg]

In my view, secondary dominants often expose the weaknesses even of intermediate improvisers, because these musicians neglect the creation of strong melody. How often do melodies actually incorporate the major third of a secondary dominant? More often than not, the maj 3rd is only an underlying harmonic force. (With the exception of the leading tone in parallel minor.) My first exercise with beginning improvisers is to become aware of non-diatonic chord tones, and to avoid conflicts with them when creating a melodic line. That means, at first, leaving out the third altogether. Excellent with the II7, for instance, is just to use the minor pentachord of the vi chord (what many excellent improvisers do anyway). When they get the hang of that, they can add the major 7 to the pentachord, that is, #11 of the V7/V. Look at melodic lines that do incorporate the maj 3rd. How do they do it in a graceful fashion? Usually from the half step above, the 4 of the chord. Learn licks that work well with the various secondary dominants, emulate some cliches, then the student will be well on his/her way. As far as resolving leading tones (maj 3rds of secondary dominants), often the solution in jazz melodies is to move to the ninth first before moving to the root of the resolution (if the melody goes to the root at all). Think of "I'm a Fool to Want You" or "Nica's Dream", last phrase of the bridge going back to the top...
Certainly, some outlining of chords can be necessary in some settings, and I can see where instrumentalists other than chordal players want to hear all the chord tones when they're practicing. Nevertheless, the awkward use of thirds often bears some resemblence to jumping around to the roots of chords: it may be unnecessary and melodically weak. Just like expert musicians hear the roots of chords without actually playing them, the player needs to be able to hear the third of a non-diatonic chord without always having to play it.