In a piano trio, the bass player handles the root. Doubling that root in the left hand makes the band sound thick and the voicing sound dull. Bill Evans, more than anyone, made the choice to leave the root out and stack the upper four chord tones into a clean, four-note shape. The result — rootless voicings — is the left-hand vocabulary of every working jazz pianist since 1958.
A rootless voicing is a four-note left-hand chord that omits the root and instead stacks the 3rd, 5th, 7th, and 9th of the chord (with the 5th sometimes replaced by the 13th, or the 9th by other tensions). The bass player — or the listener’s ear — supplies the root.
The two canonical shapes are Type A (3rd on the bottom) and Type B (7th on the bottom). They alternate in a ii-V-I so the inner voices barely move — this is the whole point. The smoother the voice leading, the more the harmony feels like one organism instead of three separate chords.
On the page, this looks like accounting. At the keyboard, it sounds like the entire 1960s.
Type A stacks the chord tones 3 – 5 – 7 – 9 from the bottom up. The bottom note is always the 3rd of the chord. For minor 7 chords (the ii in a ii-V-I), this is the most idiomatic shape.
Dm9 (the ii of a ii-V-I in C)
F – A – C – E (3rd, 5th, ♭7, 9 of D minor)
Dm9 Type A — F4, A4, C5, E5
G13 (the V) — Type A with 13 in place of 5
B – E – F – A (3rd, 13, ♭7, 9 of G dominant)
G13 Type A — B4, E5, F5, A5
Cmaj9 (the I)
E – G – B – D (3rd, 5th, maj7, 9 of C major)
Cmaj9 Type A — E4, G4, B4, D5
Look at the three voicings together. Three of the four notes barely move — or move by a single semitone — from chord to chord. That economy is the whole reason the technique exists.
Type B inverts Type A: the 7th sits at the bottom, then the 9, the 3, and the 5 on top. It’s the same four notes, just rotated.
Dm9 (Type B): C – E – F – A (♭7, 9, 3, 5 of D minor)
G13 (Type B): F – A – B – E (♭7, 9, 3, 13 of G dominant)
Cmaj9 (Type B): B – D – E – G (maj7, 9, 3, 5 of C major)
The voicing pitches sit slightly higher than Type A and have a brighter, more “up-top” flavor. They’re also useful when the previous chord ends with the 7th already in the upper register — using Type B keeps voice motion small.
The trick that makes ii-V-I’s sound seamless: if the ii is Type A, make the V Type B (or vice versa). Alternating prevents the voicing from sitting in exactly the same range chord after chord, which would sound static.
| Chord | Type A (3-5-7-9) | Type B (7-9-3-5) |
|---|---|---|
| Dm9 | F – A – C – E | C – E – F – A |
| G13 | B – E – F – A | F – A – B – E |
| Cmaj9 | E – G – B – D | B – D – E – G |
| Fmaj9 | A – C – E – G | E – G – A – C |
| Bø7 | D – F – A – C | A – C – D – F |
| E7♭9 | G♯ – D – F – A | D – F – G♯ – A |
Rule of thumb: if the previous voicing’s top note was high, switch to the other type. If it was low, stay where you are.
Rootless voicings are left-hand shapes. The right hand is free to play melody, improvise, voice extensions on top, or comp upper-structure triads. Bill Evans famously used rootless voicings in the LH and played sparse, lyrical lines in the RH, often weaving between chord tones and the 9th, 11th, and 13th of whatever the LH was holding.
When playing solo (no bassist), the LH rootless voicing is incomplete — you need to imply the root somehow. Some pianists hint at it on beat 1 with a short root note (a stride feel), then drop the root once the voicing is established. Others rely on the listener’s ear filling in the missing root, which works surprisingly well after the chord is named in context.
Week 1: Type A on every minor 7 chord around the cycle of 4ths. Dm9 → Gm9 → Cm9 → Fm9 → ... 12 keys, one shape, both hands sing the chord internally.
Week 2: Type A on every dominant 9/13 around the cycle. Same drill, dominant flavor.
Week 3: ii-V in 12 keys, alternating types. Dm9 (A) → G13 (B). Then Cm9 → F13. Then B♭m9 → E♭13. Etc.
Week 4: Full ii-V-I in 12 keys. Layer in a melody on top with the right hand. This is the ground floor of jazz piano comping.
They are used most heavily in jazz and post-bop, but the same principle — let someone else handle the root — appears in funk, neo-soul, gospel, and any context with a dedicated bass instrument. The Bill Evans shapes are the canonical version, but the underlying idea is universal.
Yes, but you need to imply the root somehow. Solo pianists often add a brief root on beat 1 before settling into the rootless voicing, or play a walking bass line in the LH while the RH handles the rootless shape. The rootless voicing alone, with no implied root, sounds suspended in air.
On dominant chords, the natural 11 clashes with the 3rd (it's the avoid note). On major chords, same problem. So the canonical Type A/B leave the 11 out and use 13 instead when the 5th is replaced. The minor 11 IS used — but typically as a separate shape (3-5-7-9-11), not the canonical 4-note Type A/B.
Type A and B both adapt: replace the 5 with ♭13 or ♯5, and the 9 with ♭9 or ♯9. So Galt becomes B – E♭ – F – A♭ (3, ♭13, ♭7, ♭9 of G) — same Type A shape, just two altered tensions instead of natural 5 and 9.
Some pedagogies (Mark Levine, Hal Galper) use a fuller alphabet: A, B, plus extra inversions for sus chords or altered shapes. Levine's "Type V" voicings layer a 4th on top of a Type A. Start with A and B; the rest are extensions of the same principle.
Continue learning