This is in relation to my last question about chord-scales. I was wondering if secondary dominants enable an application of chord-scale theory that would allow me to expand an existing chord progression by adding new chords to a song. I'm asking about using chord-scale theory to improvise a new (or expanded) chord progression while soloing over a song.

As background, a secondary dominant is the V chord of a target chord. When we select a scale for the secondary dominant, we can think of the secondary dominant as being the V of a new tonal center we temporarily establish. But establishing that new tonal center (or key) sounds to me like a possible theoretical reason to justify other chord-scale selections. So couldn't I add other chords in order to improvise a whole progression in that new scale/key? For example, in addition to the V of the new key, could I add a ii chord, a iii chord, etc. of that new tonal center and effectively extend that tonal center for a while before returning to my original chord progression?

When soloing, do musicians (e.g., in the rhythm section) consider a particular chord-scale selection to determine what additional chords they could improvise?

Edit: I think chord-scales establish modes as the new scales, whereas secondary dominants establish a chord as the new scale.

Edit#2: Secondary dominants are different than chord-scales in that chord-scales are diatonic (and use modes) whereas secondary dominants are chromatic (go outside of the key). I got this a bit mixed up when I asked this question.

  • 4
    I can’t answer because I don’t know much jazz theory, but I will say that, at least generally speaking, if you go much beyond just playing the secondary dominant and its target, that starts to become more of a modulation than a secondary dominant. Some theorists will all but automatically call it a key change as soon as pre-dominant harmonies outside the original key are being used. In fact, I wouldn’t use the term “new tonal center” in regard to a secondary dominant, I’d just call it a brief “tonicization.” To some extent this is just pure semantics, but I think it points to something deeper. – Pat Muchmore Mar 14 '19 at 13:04
  • is there really such a thing as "jazz theory" isn't this all just music theory? I mean the chord-scale that they talk about in jazz sounds relevant to me across all music genres. – user34288 Mar 14 '19 at 13:07
  • 7
    Most definitely there’s such a thing as jazz theory: it’s one of the more well-researched branches of the larger world of music theory. The areas of music theory I’ve studied and teach are tonal theory (which mostly means theory of the Classical era in European music), post-tonal theory (various 20th-century “atonal” styles), and rock theory. You’re right that all of these things are “music theory,” but that is an incredibly large field of study and no one is an expert in all of the sub-disciplines. In particular, jazz has a very different grammar. – Pat Muchmore Mar 14 '19 at 13:12
  • 2
    I think language is a good analogy: Just because I now a lot about English grammar doesn't mean I know much at all about Russian grammar. To be sure, there are general "language theory" things I might know that will help—knowing about nouns and verbs and subjunctive clauses for example—but I still would hesitate to say much about Russian language issues until I'd studied it a lot more. The "chord scale" thing you're talking about is very much not a part of tonal theory, and doesn't generally seem to reflect 18th and 19th century European thinking about music. Sounds cool though! – Pat Muchmore Mar 14 '19 at 13:15
  • 2
    I'm not sure I understand the question. It seems like the question is: "In addition to the V, I could play a ii, iii, etc of that new tonal center for awhile and then return to my original chord progression?" If that's the question, of course the answer is yes, you can play whatever you want. But I think you know that, so perhaps are you asking something about playing a progression like that beyond whether it's possible? – Todd Wilcox Mar 14 '19 at 16:10

In addition to the V, I could play a ii, iii, etc of that new tonal center for awhile and then return to my original chord progression?

Yes. I think the only concern is how to label it.

C: I V | Am: IV6 V6/5 i | C: ii6/5 V7 I


C: I V | IV6/vi V65/vi vi | ii6/5 V7 I

I think the standard idea is this: using the Am: label indicates a change of key, a proper modulation, but using the / slash indicates a secondary relationship, a temporary shift or tonicization.

I asked a similar question a few months ago. As long as the analysis of secondary functions is actually fulfilled in the other tonal area - in this case the /vi submediant region - then using the slash to indicate the secondary relationship is sensible from a functional harmony perspective.

My example harmony is classical in phrasing. But it should apply in jazz for something like a traditional 32 bar song form. Of course jazz has its own chord label system explicitly naming chord roots. I gave Roman numerals, because you used them in your question.

I've been re-reading the question.

An example progression with a secondary dominant could be...

I ii65 V65/V V

With the most basic chord/scale pairings:

  - iim7 = dorian
  - V7   = mixolydian
  - I6   = ionian

...my progression in chord/scale becomes something like: C ionian, D dorian, D mixolydian, G ionian.

...new tonal center sounds to me like the definition of a chord-scale.

The new tonal center has chords, obviously, so there will be chord/scale pairs to apply.

The only problem I can see is when someone recognizes diatonic chord patterns and selects chord-scale pairs to fit the larger diatonic area containing the chords, versus someone who doesn't recognize the diatonic relationship - or rejects them - and selects chord-scale pairs that run contrary to diatonic. Basically ii V I is diatonic and you can use dorian, mixolydian, ionian and play diatonically, or you could do phrygian, lydian dominant, lydian to play more chromatically.

If we go back to my example...

I ii65 V65/V V

...we can recognize two diatonic regions C and G...

...with brackets we can enclose the two regions and the chord-scale pairs: [C ionian, D dorian] [D mixolydian, G ionian].

...we could ignore those diatonic implications are use these chord-scale pairs: C lydian, D phrygian, D lydian dominant, G ionian.

Keep in mind I'm trying to make chord-scale choices that are "OK" according to that system, but clearly trample over the diatonic/keys. Maybe it will sound good, maybe it won't? The chord-scale system seems ambivalent about recognizing a diatonic foundation to harmony. Note that I'm not saying jazz has that ambivalent attitude. When looking at the chords for jazz standards you have to be blind to not see the diatonic patterns. The chord-scale system just doesn't reflect that.

When you say "tonal center" I think you have one foot in the diatonic/key signature mindset, and when you say "chord-scale" I think you have the other foot in a diatonic ambivalent mindset. That's OK. Just be aware of the difference when navigating your way.


The answer is yes! You can use chord-scale theory in this way. Whether or not the secondary dominant qualifies as introducing a new tonal center depends on context.

A simple scenario is one where we have a single secondary dominant, such as: | D7 | G7 | CMaj |. A good candidate scale for D7 would be D mixolydian. But D mixolydian doesn't come from the CMaj parent scale. Yet, despite this, the use of D7 and D mixolydian does not change the tonality or establish a new tonal center, largely because:

  1. D mixolydian still resolves really nicely to G mixolydian
  2. the transition is so quick, and
  3. this is so common and familiar that we're accustomed to hearing the entire progression in CMaj.

So in that case, the secondary dominant uses a different chord-scale but does not establish a new tonal center.

A more complex scenario is one where we have a longer sequence of secondary dominants. This can disrupt the tonal center. For example, let's consider the standard Rhythm Changes bridge. For Rhythm Changes in B♭, the bridge is the first 8 measures shown below:

| D7 | D7 | G7 | G7 |

| C7 | C7 | F7 | F7 |

| B♭ |...

We can't really say this has a tonality of B♭, and it certainly doesn't sound like it's in B♭. It would be hard, though, to identify a single tonal center, because the consequence of having so many second dominants is that we don't hear a conclusive resolution until the B♭ chord. When soloing, you could use D mixolydian, G mixolydian, C mixolydian, and F mixolydian. None of these come from the same parent scale. So in this case, we don't really have a single tonal center, and this is the result of our secondary dominants.

In both the simple scenario and the complex scenario, it's extremely common to hear the ii chord in front of the V7 chord, because ii-V7 progressions are the backbone of jazz:

Example 1: | Amin D7 | G7      | CMaj |
Example 2: | Amin D7 | Dmin G7 | CMaj |


Example 3: 
| Amin | D7 | Dmin | G7 |
| Gmin | C7 | Cmin | F7 |

There's no strict rule, but in general, the more chords you add, the more ambiguity you create around the tonal center. So Example 2 will generally tend to sound more ambiguous than Example 1.

  • Does the chord-scale system say something about a tonal center, ii, I, IV, V, ... any of that? The way I read the Wikipedia description is, the chord-scale system only gives you a list of scales. How to know which of the possible scales to pick, or how the selection might limit or expand your creative possibilities for creating tension and release, moving the perceived tonal center and that sort of stuff, it's up to you to get that knowledge from somewhere else. – piiperi Reinstate Monica Mar 14 '19 at 19:44
  • I agree with you that the chord-scale system doesn't necessarily say anything about tonality. I interpreted the original question as asking what implications the chord-scale theory has on tonality when secondary dominants are introduced. I think the OP is noticing that certain chords can solicit scale choices that are not in the song's main key. Does this scale choice change the tonality? I've tried to answer by explaining that, in some cases, the secondary dominant implies scale choices that do alter the tonality; but other times, those choices don't impact tonality. Does this clarify? – jdjazz Mar 14 '19 at 20:08
  • So if you're playing a song in some key X and there's a C major chord... there's a C major in many other keys as well, and those other keys may have chords that are not in the original key? :) Shouldn't this be self-evident if you've played anything in more than one key... Surely you don't need the chord-scale system to know that secondary dominants may temporarily affect the scale? Would someone actually first use the chord-scale system to get a list of scales, then somehow pick one from the list, and only then notice a secondary dominant in the scale? This feels like backwards logic. :) – piiperi Reinstate Monica Mar 14 '19 at 20:53
  • The bridge of Rhythm Changes is sequential both harmonically and melodically when considering the original tune, sequences lead to some tonal target and would normally be analyzed accordingly Bb: V7/vi V7/iii V7/V V7 I en.wikipedia.org/wiki/Secondary_chord#Mozart_example – Michael Curtis Mar 14 '19 at 21:56
  • This is the strange ambivalence of chord-scale, to not see the obvious diatonic nature of something like Rhythm Changes, especially the vanilla changes. Even most of the harmonic variations I have seen fit pretty neatly into functional analysis. – Michael Curtis Mar 14 '19 at 21:59

As usual, you're over-thinking this :-)

Yes, you can modulate to a new key and do all the usual stuff relative to THAT key for a bit.

Yes, the classic way to modulate to a new key is to introduce its dominant chord. If you want to get to D major, throw in an A7 chord. Doubtless the A7 could be considered a 'secondary dominant' in the original key.