Take the chord Cmaj9. Five notes: C, E, G, B, D. Now take away the C.
That is the whole idea, and it sounds like vandalism until you play it. What is left is E, G, B, D, and the ear still hears Cmaj9 — not because you tricked it, but because the third and the seventh are what identify a chord, and the root is what a bass player is already playing while you are deciding what to do with your hands.
Thirteen of the thirty-six pianists on this site play that shape. It is the most widely shared hand position in the whole collection, and the two ends of that list are Art Tatum, born 1909, and Jacob Collier, born 1994. Eighty-five years, and the same four fingers.
What is actually in the hand
The shape has a name, Maj7 Rootless A, and in C the left hand is E3, G3, B3, D4 — the third, fifth, seventh and ninth, stacked from the bottom up. That is the whole grip. The right hand is then free, and what it does is the difference between comping and playing.
The reason it works is a division of labour. In a trio, the bass owns the bottom of the harmony, and every root the piano adds is a note the bass is already playing. Doubling it does not reinforce anything — it thickens the low register, which is the part of the piano that turns to mud fastest, and it costs a finger that could be playing a colour.
Take the root out and the same five fingers can carry the ninth, the eleventh, the thirteenth. The chord gets more specific, not less, while getting one note smaller.
Two forms, one small region of the keyboard
The A form starts on the third. The B form starts on the seventh. That is the entire difference, and it exists for one practical reason: voice leading.
In C, the A form is E3, G3, B3, D4 and the B form is B3, D4, E4, G4. Same four notes, different note on the bottom, and about an octave apart in where the hand sits.
Play a ii-V-I with the A form on every chord and your hand leaps around the keyboard. Alternate the forms and it barely moves.
Here is the actual mechanism, in a transcription rather than a description. The app stores a Bill Evans ii-V from Dm9 to G13, and its left hand goes:
F–A–C–E, then F–A–B–E.
One note moves. The C becomes a B, a single semitone, and the chord has changed from a minor ninth to a dominant thirteenth. Three of the four fingers do not lift. That is not a party trick; that is what people mean when they say voice leading, and it is the entire argument for learning both forms rather than picking a favourite. The forms are not two flavours. They are the mechanism that lets you stay put.
The dominant shapes work the same way with more colour available: Dom7 Rootless A is E3, A3, Bb3, D4 — third, thirteenth, flat seventh, ninth — so the thirteenth is baked into the grip rather than added on top. Its partner is Dom7 Rootless B.
A warning about the fifth. In the maj7 A form the fifth is in there, and in most rootless dominant shapes it is not. That is deliberate. On a dominant chord the fifth is the least informative note in the chord — it says nothing the root has not already said — and the space it frees is what the thirteenth and the ninth occupy. If you find yourself adding a fifth back in because the chord sounds thin, the problem is usually the register rather than the missing note.
Where it sits, and why that is not negotiable
Notice that every grip above starts in the octave from E3 to D4. That is not arbitrary and it is not stylistic.
Below about C3, a CLOSE-POSITION chord turns to rumble: the intervals are narrow relative to the beating between their overtones, and the ear stops separating them. Widely spaced notes down there are fine — Chick Corea has an open spread whose left hand runs C2 to E3 and stays transparent, because the gaps do the work. It is the spacing that fails in the low register, not the note count. Above about C5 you are in the range where the melody lives, and a comping hand parked there is competing with whoever is playing the tune.
The rootless voicing sits in the octave between those two problems. It is one of the reasons the shape is so durable — it was designed to be playable in the only part of the piano where a four-note left hand can be dense and still legible. The app’s own voicing checks lift generated voicings into that region for the same reason, and let documented transcriptions sit lower when the spacing earns it.
Why the shape is a floor and not a ceiling
Here is the thing the shared-shape statistic actually tells you.
Thirteen pianists play it, and not one of them is defined by it. Everyone on this site who reaches for the rootless voicing also reaches for something rarer: Bill Evans has his own B-form variant, McCoy Tyner has his quartal stack, Herbie Hancock has the altered upper structure, and Kenny Barron has the Lydian pad.
That pattern is worth taking personally if you are learning. The rootless voicing is the shape everyone agrees on — the reason it is the first one taught is that it is the one that stops being a decision. What makes a player recognisable is whatever they do beside it.
How to actually get it under your hands
Not by learning twelve keys of it as a table. Three things, in order:
- Play the A form on Cmaj7 and say the four notes out loud while you play them. E, G, B, D. Naming the degrees rather than the letters — third, fifth, seventh, ninth — is what makes it transferable, because the shape is the degrees and the letters are just today’s key.
- Play one ii-V-I, alternating A and B forms, and watch your hand instead of listening. If it is jumping, you have the forms the wrong way round. Fix that before you add a second key.
- Then move it in fourths, not chromatically. C, F, Bb, Eb. That is the cycle standards actually travel in, so the work transfers to a tune rather than to an exercise.
The last step is the one people skip and it is the one that matters. A shape you can play in twelve keys as a drill and cannot find in a tune is a shape you have memorised rather than learned.
Where to hear it
Anything by Bill Evans from 1959 onward. Then, for the distance the same shape can travel, put on a Robert Glasper record straight after. Evans was born in 1929 and Glasper in 1978, and the grip is the same.
The point of listening to them back to back is not to admire the continuity. It is to hear how little of what you are noticing is the voicing. The shape is identical; the touch, the time, the register, what the right hand chooses and how long anybody waits before playing — none of that is in the grip, and all of it is what you are actually hearing.
Which is the honest summary of the whole thing. The rootless voicing is a solved problem, and solving it buys you a starting position rather than a sound.
Questions people actually ask
- Do rootless voicings work without a bass player?
- Solo, they lose their footing: nobody is supplying the root, so the ear has to guess the harmony from the upper structure and it will often guess wrong. Play them solo by adding the root in the left hand an octave or more below, or by walking a bass line under them. The shape was developed inside a trio, and it assumes somebody else is holding the bottom.
- What is the difference between the A form and the B form?
- Which chord tone is on the bottom. The A form starts on the third, the B form starts on the seventh, and between them you can move through a chord progression without your hand jumping around the keyboard. Most players learn one, then learn the other purely so they can stay in one small region.
- Is the root ever actually played?
- Not by the left hand, and that is what the name means: the four-note grip itself has no root in it. The stored shapes in the app do place a C in the right hand, in the upper register, and that is a completely different job from a root in the bass — up there it is one voice among several, down there it is the floor the whole harmony stands on. In performance the bass has the bottom, or your own left hand does in a much lower register.