Hand Independence Is a Lie

"Practice hands separately until each hand is fluent, then put them together" is the default instruction in every method book, and it reliably produces players who can play each hand alone and still fall apart the moment the hands meet. That isn't a coordination deficit. It's that hands-separate and hands-together are solving two different problems, and drilling one does not buy you the other.

Geography is not timing

Hands-separate (HS) practice solves geography: where the fingers go, in what order, at what force, on which keys. It answers "what is the physical shape of this passage." Nothing about HS practice, however long you run it, touches the second problem, which only exists once two hands move at once: timing — the relationship between what the left hand is doing and what the right hand is doing, at every instant. Timing isn't a property of one hand. It has no meaning for a single hand alone, the same way "distance between" has no meaning for a single point. You cannot practice a relationship by yourself, twice.

LH + RH ≠ HT

Drill the right hand alone until it's automatic and you build a self-sufficient internal representation of that hand's timing — a clock tuned to that hand's rhythm, with no reference to the other hand, because there is no other hand in the room. Drill the left hand alone and you build a second, equally self-sufficient clock. Put the hands together and the brain doesn't merge those two clocks. It discards both and builds a third: a single joint representation of the whole texture, from nothing. That third thing is what "hands together" practice is actually practicing, and hands-separate work never touches it, no matter how fluent each hand gets first. This is why note-perfect HS playing can collapse completely the first time the hands meet — the geography was never the bottleneck.

The word is backwards

"Hand independence" is the standard term for what good players have, and it points at exactly the wrong thing. What separates a player who can hold two rhythms together from one who can't isn't that their hands are independent — independent hands drift, because nothing is holding them to the same clock. What they actually have is closer to the opposite: a hard lock. One motor plan generating two simultaneous outputs, the way a single score generates two staves. The hands don't negotiate with each other in real time; they don't have to, because they were never two separate plans that needed reconciling. Training toward "independence" as a goal trains the two hands to run further apart. What you want is for there to functionally be one plan.

Build the composite, not the halves

If timing is a single relationship rather than two things to reconcile, the fix is to practice the composite rhythm directly instead of assembling it from HS parts. Take the combined rhythm of both hands — every attack point, from both hands, on one timeline — and learn to produce that one pattern with one intention, before assigning which hand plays which note. Clap it. Tap it on a table, both hands landing wherever the pattern puts them, treated as one rhythm with two colors rather than two rhythms that happen to overlap. Once the composite is one gesture in your body, distributing it onto the actual pitches on two hands is a small step — you're relabeling an already-unified pattern, not assembling coordination from scratch.

This is normally taught as a specialized fix for one narrow case — polyrhythms, 3-against-2 and its relatives — with a chapter to itself and then dropped. There's nothing about the technique that's specific to polyrhythm. Every hands-together passage is a composite rhythm; polyrhythm is just the case irregular enough that nobody can fake it by feel. Generalize the technique past that one chapter and it becomes the default way to build hands-together timing, not the emergency fallback for the hard bars.

The second half of the method is a loop: improvise a simple hands-together pattern — a riff, a broken chord against a bassline, anything — and keep going until it falls apart. The stall is the point. It's the exact moment the joint representation runs out and the two hands revert to being separately managed — the seam you're trying to eliminate. Note where it happened, drill the composite rhythm at that seam in isolation, then improvise back through it. Repeat somewhere new. This builds the lock under conditions where there's no score to fall back on and no way to fake it by remembering two separate hand parts.

How to tell if you actually have a lock

A hard lock and a good guess look identical at performance tempo on a piece drilled to death, which is why the failure hides for years. Three tests that don't let it hide:

Change tempo suddenly, mid-passage, without warning. A lock holds the relationship at any speed; a memorized combination degrades or breaks, because you were pattern-matching a specific timing, not producing a relationship.
Drop one hand out for a bar, silently, then bring it back. If the other hand hesitates, drifts, or restarts from the top of the bar, it was never running off a joint plan — it was listening to the hand you removed.
Play against a metronome set to double or half your usual subdivision. If the relationship survives the recount, you have it. If you have to re-derive the whole passage, you had memorized coordinates, not a rhythm.

A drill ladder

Clap the composite rhythm alone — every attack from both hands, as one pattern, no pitches.
Tap the composite on two surfaces, the correct hand doing the correct attacks, still no pitches — the pattern moving from an abstraction into two sides of your body.
Play the composite at the keyboard on a single note per hand. Geography is trivial here, so any failure is purely the joint plan breaking.
Add real geography to one hand while the other stays on a single note.
Add real geography to both hands, slow, metronome on the offbeat so you can't lean on it for the downbeat.
Remove the metronome and improvise variations on the passage until it stalls; drill the stall; resume.

None of this replaces HS practice for actual geography — a passage with an awkward stretch or a fast run still needs the fingers to learn where to go, and that stays HS's job. The ladder above is the missing rung: the one that builds the relationship, which nothing in the standard hands-separate-then-together sequence ever targets on purpose.

This isn't new — it's just never been named as the mechanism

Chuan C. Chang comes closest to naming the tax without naming the mechanism. In Fundamentals of Piano Practice, he writes: "With HS practice, you acquire finger/hand technique; then with HT practice you only need to learn how to coordinate the two hands" — treating the two as genuinely separable stages, the position this piece is arguing against. But his own numbers are the best evidence that they aren't separable. His readiness rule holds that you're prepared for hands-together only once you can play each hand at 110–150% of the target speed alone — and even then, "the maximum HT speed you can play is 50% to 90% of the slower HS speed." That gap, up to half your speed gone no matter how overprepared each hand is, is the missing lock, priced. If HS technique simply combined into HT technique, the tax wouldn't exist. It exists because HT is a third thing HS never built.

Abby Whiteside got closer to the actual mechanism seventy years earlier and it never made the method books. In Indispensables of Piano Playing, her whole technique is organized around what she called "basic rhythm" — a single physical continuity meant to originate the movement of both hands from one gesture — and she was openly hostile to counting individual note values, warning that "no more devastating habit can be formed than that of associating the rhythm of playing with the action for producing single tones." She named the failure this produces "note-wise procedure": treating music as a sequence of unconnected pitches instead of one shape. That's a direct ancestor of the composite-rhythm approach above. What's missing from her account, and what this piece adds, is generalizing it past a philosophy of touch into a specific, drillable technique — plus the improvise-until-you-stall loop, which turns "have a unified basic rhythm" from an aspiration into something you can practice by finding exactly where it breaks.

Drummers will recognize the composite-rhythm approach immediately: Jim Chapin's Coordinated Independence, the standard reference for four-limb jazz drumming, is built entirely on practicing combined patterns rather than layering separately-drilled limbs. But it's worth being precise about what that similarity is and isn't. A drummer's four limbs are inherently unpitched, built for exactly this kind of layered attack-point thinking, so composite practice there is closer to the native approach than a late correction. Pianists have the identical word problem — build the relationship, not the halves — dressed in a different instrument. It's the same fix, arrived at independently, not evidence that pianists are behind.

Hands-separate is not the problem

None of this argues against hands-separate practice. HS is the correct, close to the only, tool for what it's actually for: getting the fingers to know where to go, at what force, in what order, without the other hand's demands competing for attention. A passage with an awkward fingering, a wide stretch, or an unfamiliar hand shape needs isolated repetition to become physical knowledge, and trying to build that while also building the joint timing relationship is genuinely harder, not more efficient — one drill solving two unrelated problems at once. The failure isn't hands-separate practice. It's treating hands-separate practice as stage one of a single pipeline that ends in hands-together fluency, as if hands-together were just hands-separate plus assembly. It's a different workshop, solving a different problem, and it needs its own drill.

Two clocks don't make one relationship, no matter how well each one is set. Independence was always the wrong goal — you're not trying to make the hands need each other less, you're trying to make them, for the length of the piece, need each other completely.

If the geography side is where you're actually stuck, Real Sight Reader drills reading and rhythm on real repertoire, so hands-separate practice stops being the bottleneck it usually is.