Constructing Jazz Scales from Chord Tones
I noticed something the other day about the relationship between jazz chords and scales. I first thought it was a shortcut for thinking about an appropriate scale over any chord on the fly, but I'm thinking it might be more fundamental than that.
To start with, here are the basic chord families found in most jazz. Beside each is the scale with all consonant notes. These aren't necessarily the best or most often-used scales, but it's generally accepted that every note in each of these scales sounds strong and consonant over its chord. It seems that there's a unique scale fitting this role for each chord family. This is probably an oversimplification, but bear with me for the purposes of my discovery.
- maj7 (1 3 5 7) => Lydian (1 2 3 #4 5 6 7)
- m7 (1 b3 5 b7) => Dorian (1 2 b3 4 5 6 b7)
- 7 (1 3 5 b7) => Lydian Dominant (1 2 3 #4 5 6 b7)
- m7b5 (1 b3 b5 b7) => Locrian #2 (1 2 b3 4 b5 b6 b7)
- dim7 (1 b3 b5 bb7) => Whole-Half (1 2 b3 4 b5 #5 6 7)
- aug7 (1 3 #5 7) => Lydian Augmented (1 2 3 #4 #5 6 7)
- m/maj7 (1 b3 5 7) => Melodic Minor (1 2 b3 4 5 6 7)
- 7b5 (1 3 b5 b7) => Whole Tone (1 2 3 b5 #5 b7)
- 7#5 (1 3 #5 b7) => Whole Tone (1 2 3 b5 #5 b7)
- 7b9 (1 3 5 b7 b9) => Half-Whole (1 b2 b3 3 #4 5 6 b7)
- 7#9 (1 3 5 b7 #9) => Super Locrian (1 b2 b3 3 b5 b6 b7)
Here's the pattern in constructing the scales from the chords. Start with the chord tones in the scale. Those will obviously be the most consonant. Now add the note one whole step above each chord tone. Whenever this results in consecutive minor 2nds in the scale, lower the new note by a half step. That's it. That works for every chord above, and it has held up with all sorts of complex extensions as well. See for yourself and let me know if you find any exceptions.
Shortly after discovering this pattern, I found a couple reasons why it works. The first relies on the concept that melody is just harmony in motion. When you play a melody, with or without accompaniment, you're implying different underlying chords or different voicings of the same chord. In a typical jazz setting, the bass player plays the lowest note of the chord, and the rhythm player plays more chord tones together, often omitting the root. The soloist emphasizes certain notes in the melody with their rhythmic placement. These notes, usually falling on the beat, at the end of a phrase, or held for longer than other notes, help define the current harmony along with the bass and rhythm parts. For example, bass plays root, piano plays 3 5 b7 9, and sax emphasizes a 6 in the melody. At that point, the current chord becomes a dominant 13 instead of just a dominant 9 from the rhythm section. It seems like the above scales work because any chords implied by using these scale tones as chordal extensions will remain in the jazz idiom. In other words, the above scale tones are the most common default extensions for their corresponding chords in typical jazz harmony.
The second reason I came up with is more compelling. It relies on the undesired dissonance of the minor 9 interval (octave plus a half step), played harmonically. There's a ton of tasteful dissonance used in jazz for creating tension or strange sounds, but the minor 9 interval seems to have an extra awkwardness about it when applied to harmony, I don't know why. Try it for yourself: play a maj7 chord with 7 in bass and root on top, or a 7#9 chord with #9 in bass and 3 on top, or a m9 chord with 9 in bass and b3 on top, etc. They all sound kind of awkward, especially compared to voicings without a minor 9.
Now, when creating the above scales by adding whole steps above the chord tones, all such minor 9 clashes with the rhythm section (usually an octave or more below the soloist) are avoided. Hurrah! This clears up a few other things for me as well, such as the 1 and 4 being weak points in a major scale over a maj7 chord. They create minor 9 intervals with the 7 and 3, respectively, in the rhythm section. Hurray! I've also found that a b2 over a maj7 chord sometimes sounds cool, albeit really weird and outside. And it's because it's a whole step above the 7. Hurroo!
While there are a number of exceptions to avoiding minor 9 intervals, and to my discovery in general, this still gives me a better grasp of consonance in melody. Besides, exceptions occur everywhere you look in music and music theory anyway.
bob romeo (3 Jun 2008 at 8:57am)
hi,your jazz scales to chords are good,but for a beginner like me not good-you have a maj7=lydian,can you tell me what chord,AtoG,and what lydian.sorry i'm so confused.isee guitarists flying up and down the neck,i don't know what to do,please help me to understand
yours truly(stupid)bob romeo
Joe (3 Jun 2008 at 3:15pm)
Hi Bob, thanks for commenting.
Sorry to confuse; I write most of these articles with a lot of prior knowledge assumed. I would try googling for maj7 or lydian and see what you get. When I started learning jazz, all those "most consonant scales" were among the first things to learn. They were not obvious to me at first. I always thought you just play a major scale over any major chord.
To spell out an example, a Cmaj7 chord (C E G B) would use the C Lydian scale (C D E F# G A B). You're free to use plain major (C D E F G A B) or whatever else you want, but landing on the F# and holding it tends to sound better than F.
Garrett (29 Jul 2008 at 4:27pm)
Hey, I remember seeing a chart similar to the one you made in an cheap "learn jazz" book (made of paper, bound with staples) that came with an Abersold play-a-long that I bought two years ago. It's a pretty great idea. Something else that I like to do to create interesting chord tones is to play around with the modes of the harmonic and jazz melodic minor scales. That's where most of the weird dominant scales (phyrgian dominant, super locrian, etc.) come from anyway.
I dig the blog! I'm spending this summer practicing a lot too (guitar and piano) and it's great to see a kindred spirit at work.
Julian (10 Apr 2009 at 5:58pm)
Hey I'm studying jazz music theory and incorporating it into my playing. When constructing a major seventh chord (i.e Gmaj7), isn't the mode being utilized Ionian, not Lydian. Could it be used for both, but the different notes of the scale would correspond to different modal scales? Is it more effective with regards to the Lydian mode or the Ionian major mode???
Joe (13 Apr 2009 at 2:16pm)
Hi Julian,
Conventional jazz wisdom says the natural 4 is an "avoid note" over a maj7 chord, partly because it clashes with the 3 and partly because the 1 and 4 establish a tonality for the IV chord, which is not the current chord. Plus, the 4 to the 3 is probably the most important voice movement in a V7-I cadence.
See for yourself. Try improvising over a maj7 vamp using Ionian vs. Lydian. Focus on the difference between the 4 and the #4.
When emphasized over a maj7 chord, the 4 tends to sound awkward, while the #4 sounds more intriguing, albeit more outside than 3, 5, or 7.
Steven (21 Apr 2009 at 6:12pm)
I'm looking at your 7#9 scale- super locrian- and wondering where the 5 is... you have it listed as 1 b2 b3 3 b5 b6 b7 . As a guitarist, I seldom play the 5 with a 7#9, but sometimes it calls for it... for that scale, why not have a 1 b2 b3 3 5 6 7? You have both a b5 and b6 listed, which when combined with the 5 of the chord would create a dissonance issue.
The scale you're using implies an altered chord, which is pretty similar to a 7#9, but I feel like slapping an altered scale to a 7#9 is a bit risky. The altered scale implies a #5 and or b5 which is not written on the 7#9 chord symbol. I don't read too often, though, so is an altered 5 simply implied with that chord?
What do you think?
Bob (3 Aug 2010 at 10:05pm)
I am interested in this also. AMAZING site!
Gerald Nkhoma (5 Apr 2011 at 1:36am)
Hi Joe,
while googling jazz scales so that i could have a better understanding & a more in-depth grasp of their usage i serendipitously stumbled upon your website and i have to admit that in terms of content both practical & theoretical, it eclipses the rest of the field by a country mile.
However, i am at a loss regarding two aspects, namely how you arrived at the super locrian scale for the 7#9 & how to resolve minor 9 dissonances.With regard to the 7#9,my calculation using your formula makes the relevant scale a half-whole scale ditto for the 7b9;1 b2 b3 3 b5 5 6 b7.
I'll definitely be recommending your site 1000%.
What a treasure trove of knowledge!
Many thanks & kind regards
Gerald
Joe (25 Apr 2011 at 1:26am)
Hi Gerald,
I was probably mistaken when I was dealing with the 7#9 chord. Back when I wrote this, I wasn't making the distinction between a simple 7#9 (1 3 5 b7 #9) and a 7#5#9, or 7alt (1 3 #5 b7 #9, w/ b5 b9 also implied). They are different sounds, but I was treating them as the same thing in jazz. I still often convert a 7#9 to a 7alt if I see it on a lead sheet and I think it's more appropriate.
So you are right, a 7#9 should lead to Symmetrical Dominant (half-whole diminished). And any time I see an altered 5 and an altered 9 in the same (dominant) chord symbol, I assume both altered 5s and 9s are in the chord. That spells out all seven notes of Super Locrian, no formulas necessary.
Thanks!
Joe
WBass (24 Aug 2012 at 8:42am)
Hi Joe,
Can you explain why
aug7 (1 3 #5 7) => 1 2 3 #4 #5 6 7, in stead of #6 7.
I think it has to do with:"Whenever this results in consecutive minor 2nds in the scale, lower the new note by a half step". Maybe it's because English is not my native language or I lack music theory. What exactly do you mean with "consecutive minor 2nds? Is it because when u use #6 you'll get:
#6-7-8 = 2 half steps = consecutive minors?
Greetings and thanks for this lesson :), WBass
Joe (27 Aug 2012 at 2:29pm)
You got it. A minor 2nd is the same as a half step, so two half steps in a row (as in #6-7-8) would be consecutive minor 2nds.
Brian Barrett (15 Sep 2013 at 1:46pm)
This one should be a sticky "How to start handling Jazz improvisation". Anything that simplifies the construction of stuff. Gets close to the CAGED pent thinking.