The following 2-part question was sent to me by a musician friend - here's the first part:
"I've got a jazz guitar soloing question. Most of the books I've seen and articles I've read advocate a scales-over-chords technique for jazz soloing. So if you know jazz chords with a fifth and sixth string root note, you can quickly find the scales that go with those chords, assuming you know your jazz theory. This however forces you to jump around the neck a lot. I came across an article years ago, that said it was easier to play all your modes out of the basic major scale box position, starting on different scale degrees every time you need a different mode, assuming the key of the moment stays the same. I asked Tom DeMasters about this and he said that while this would work, it was actually a much more difficult way of playing the guitar. What do you think?"
This is a great question and the answer will vary with the player. You can find great players who do it all different kinds of ways. I can only tell you how I do it, and why I do it the way I do.
Think Chords & Arpeggios!
Generally, I don't think in terms of scales or modes – I can, but more often than not, I think in terms of chord shapes and their corresponding arpeggio patterns.
Back when I was first exposed to Jazz, it was a mystery to me how jazz players soloed through changes. I could hear the chord changes in their single-note solos. I tried to mimic that by ear without success for several frustrating years. In the early years of studying with John Elliott I came to the conclusion that if I based my lines on arpeggios rather than scales, I could take everything that John was teaching me about chords and apply it to lines. I began to obsessively practice my seventh chord arpeggios. Arpeggios are more difficult to play than scales and I uncovered and dealt with some technique problems I had as a result of this process. If I was working on II-V-I's with John I would, on my own, also arpeggiate everything that he was giving me. I would randomly pick out tunes from the Real Book and play the changes as arpeggios instead of chords – like a horn player would deal with chords. When I say, “arpeggio,” I have a very narrow definition; the notes of a chord played in numerical order – 1-3-5-7 – ascending and/or descending. From a guitar standpoint, the arpeggios should be drilled through at least two octaves from a 6th-string root and a 5th-string root. BTW - arpeggios also have inversions, just like chords, i.e. Root Position: 1-3-5-7. First Inversion: 3-5-7-1. Second Inversion: 5-7-1-3. Third Inversion: 7-1-3-5.
The 3rd and 7th Define the Chord!
John said, “The third and the seventh are the defining notes of the chord.” So I began to isolate the thirds and the sevenths out of the arpeggios. I would play through the changes of whatever tune using just thirds and sevenths and I would do it on every set of two strings, visualizing how the thirds and 7ths related to the arpeggio pattern and the chord shapes. I began to analyze the melodies of the standards and I discovered there was a great focus on the third of the chord; dozens and dozens of examples come to mind – Autumn Leaves, All the Things You Are, How High the Moon, Solar, Stella by Starlight, etc. The most important note by far in any chord is the third – I call it the “money note” or a "target note." I began to build a vocabulary of licks and patterns based on arpeggios that emphasized the third of the chord. I added notes to arpeggios to create what John called, “melodic fragments.” Any note can be added. No matter how dissonant it may be, it can be resolved, and it always resolves to a chord tone, i.e. an arpeggio note – the third is the strongest resolution for stating or implying the chord; after that, the seventh. After that, the root. Then the fifth – they all work, but the third is the strongest. It seemed to me that starting from arpeggios was the safest starting point; all the notes are chord tones, therefore safe. A typical 7-tone scale or mode is going to have 3 non-chord tones that you need to eliminate to get the chord tones, then 2 more to get to the third and the seventh. It seemed simpler to me to start with the arpeggio and then add whatever to create my lines rather than start with a scale, and have to reduce it down to four chord tones, and then reduce it further to the third and seventh of the chord.
There are several lessons on the “Free Lessons” page of my website that explain and demonstrate this process. Go to the “Advanced Soloing” section of the page and start with the first of 7 lessons under “7th Chords.”
You must be a site member and logged in to view! If you aren't a site member go here to sign up - it's FREE!
There Are No Wrong Notes, Only Wrong Resolutions!
At first, I didn't think about or define what notes I was adding, you don't have to; you just have to know where the resolutions are. I've heard Pat Metheny say, “There are no wrong notes, only wrong resolutions.” I would make note choices based on what I wanted to hear – what feeling I wanted to evoke. An educated listener might hear me play A-C-E-F#-G over an A Minor chord in a solo and conclude, “He's playing in the A-Dorian Mode.” They'd be right, but I wasn't thinking about it that way. I was thinking in terms of an A Minor 7 arpeggio (A-C-E-G) and then adding the F# because that's what I wanted to hear. I could've just as easily added F-natural because that's what I wanted to hear, in which case my educated listener would conclude I'm using the A-Aeolian Mode. But again, that's not the way I was thinking.
Later on, I began to be more cognizant of the notes I was adding, and how they related to the chord over which I was imposing them. So, while soloing, I began to be aware of things like, “This is the Dorian 6” or “this is the Aeolian 6” or “this is a flat 9, i.e. a Phrygian 2” - I had been playing all those things but not necessarily defining them, or rather, defining them in terms of the feelling, sound &/or color they evoked. I began to define the notes I was adding using modal or scalar terms - “this is the Harmonic Minor 7” etc. In this way I began to layer modal thinking on top of my chordal thinking on top of my right-brain visceral thinking and kind of backdoored into the modes that way. I have a way of thinking about and teaching modes that I've not seen elsewhere - it avoids using the parent scale as the source for fingering (which the article you referred to seems to suggest doing) - but that's another post.
Ultimately, it all integrates and mooshes together. John Mclaughlin said, "Scales are linear chords and chords are vertical scales." It's not entirely accurate to say I think or visualize chords/arpeggios as opposed to scales/modes when I solo – I do, mostly – but at the same time, I'm aware of the modal dimension that's going on with the notes I add to that arpeggio-based vocabulary.
At some point in my studies with John, he gave me some linear exercises to play. In hindsight, I realize now that they were modal exercises through II-V-I progressions. I really couldn't see the point at the time. I was already well into my arpeggio-based approach – I was doing that all on my own; I never told John about it – and my lukewarm response probably caused him to discontinue that direction. We returned to the “all about chords” material. My approach was affirmed by John (unbeknownst to him) towards the very end of my time with him when we delved into what he titled, “Melodic Bitonals” - soloing over bitonal chords where you use the upper-extension/bitonal triad as your tonal center for lines. He said, “Add seconds or fourths to the bitonal triad arpeggios to create melodic fragments.” I took to that immediately – it's the same kind of thing I had been doing on my own all along with 7th arpeggios.
So, most of the time, I solo through each chord, whether they are diatonic (related by key) or not, using arpeggio-based vocabulary.
However, there are situations where I think primarily in terms of scales/modes and that brings us to the second part of your question:
"I also question the idea of changing scales for each chord. This works if the chords aren't changing very quickly, but if you have a fast piece and a case where there are 3 chords in one measure, surely you don't think 3 scales. It seems like it would be much more practical in that case to just use one scale for the measure-whatever key the measure is in-a major scale maybe, etc. Otherwise you might only be using 1 note from each of 3 scales, which seems ludicrous."
The limitation I've run up against with my "vertical" approach has to do with sheer speed. Arpeggios are more difficult to play; they have wider intervals (that's one of the things I like about them; wider intervals are more interesting and surprising to my ear), and so it happens sometimes that I want to play them faster than I can. That would probably happen no matter what approach one takes – desire for speed can be a trap, an endless treadmill. No matter how fast you get, you always run into a situation where you wish you could play just a little bit faster...I gave up pursuing speed for speed's sake decades ago. Still...I wish I could play faster, dammit!
One obvious solution is sweep-picking arpeggios. Frankly, I've never liked them - even when played cleanly and precisely, which is kind of rare. I've only heard a couple of guys use them in a way that caught my ear. For one thing, I rarely play a straight arpeggio, I'm always adding notes to it here and there for melodic interest. The sweep-picking thing – 2 octaves up and down, up and down, up and down – boring! Boring in spite of the awesome speed, the appeal of which wears off after about a minute. If I could do them, which I can't, I would use them very sparingly.
For me, the solution has been to switch my primary way of thinking from arpeggiation to a more scalar/modal paradigm. This is because I can play scales way faster than I can play arpeggios.
When I'm dealing with a solo this way it's usually because the tempo is really quick! Faster than I am capable of arpeggiating eighth note phrases. At a high tempo, I don't change the scale with every chord. I reduce the form of the song into tonal centers, as you suggested. Some would call them key centers. They are diatonic progressions like II-V-I's or I-VI-II-V-I's that make up the building blocks of most standards. So I switch scales when the key or tonal center changes.
Scales/Modes - Still Chordal!
Even when playing scalar, I can visualize the “money note” of each chord in the progression in relation to its position in the scale or mode I'm working in. So, in a sense, even though my hand is “thinking” in terms of a scale pattern, my mind is still thinking vertically, that is, chordal.
And in either case, scales or arpeggios, there are multiple fingerings for everything that, if known, can enable you to play through an entire song form without shifting more than a couple of frets one way or the other – so you don't have to “jump around like a frog on acid.” That's a quote from a previous blog where I dealt with this issue. BTW - there are legitimate reasons for playing things in a non-efficient manner - jumping around like a frog on acid - but lack of knowledge about fingering possibilities is not one of them.
(Thanks to Brad Allen for the thought-provoking question.)