Check the Handwriting, Not the Brain

— And you wouldn't have. But let's start with what you said about the hand. "Wooden" is your hypothesis, not your observation. Pick up a cup with your left hand. Catch a ball. Scratch your nose, do up a button. All fine?

— So coordination is intact, strength is intact, sensitivity is intact. The hand is healthy and precise. Exactly one thing doesn't work — and it's the one thing that was trained separately, for 20 years, on the other hand.

— It does. Writing is stored on two floors, and confusing them is the main mistake in this conversation. Upstairs: which letter this is and what shape it has. That sits in the left hemisphere, and when a right-hander writes with the left hand, exactly the same left-hemisphere writing areas are working as when writing with the right; writing with the left just adds the right dorsal premotor cortex on top. Potgieser and de Jong measured that in 2015. And downstairs: in what order and with what force to pull the muscles. That lower floor is tied to a specific hand, and the left one doesn't have it.

— It has the letters. It doesn't have a performer. You're handing it the score, and no orchestra has been assembled for that score.

— It doesn't transfer even within one hand. Karni and colleagues, in 1995, trained people daily for 4 weeks on a single sequence of finger movements. By the fourth week the patch of primary motor cortex serving that learned sequence had grown larger than the one serving an unlearned sequence — meaning there was rewiring, real rewiring. And then they gave them a second sequence: same fingers, the very same movements, just a different order. Speed dropped back to day-one level.

— The same. Nothing carried over. And to the other hand, of course, nothing either. It isn't the muscle that gets trained, and it isn't the hand. What gets trained is the order.

— Here's where it gets unpleasantly vivid. Roux and colleagues, in 2009, worked with 12 patients undergoing awake brain surgery. The surgeon runs an electrode across the cortex while the person talks and writes. There's a spot just in front of the hand area: you send current through — the person goes on talking calmly, the hand moves normally, and the letters under the pen fall apart into something unreadable. Take the current away and the handwriting comes back. That's how it went in 6 of the 12.

— Separate from "move the hand" and separate from "speak." And that address isn't in the right hemisphere, the one that controls the left hand. Look at what happens after a left-hemisphere stroke: learned complex actions fall apart in both hands, including the healthy left one. After a right-hemisphere stroke that doesn't happen. And it isn't rare: Donkervoort and colleagues examined 492 people after a first left-hemisphere stroke — apraxia turned up in 28% in rehabilitation centres and 37% in nursing homes.

— It has taken its programs from the left hemisphere its whole life. It's just that nobody ever issued it the program for writing.

— Imagine you've walked home by one route for 20 years and know it well enough to arrive while thinking about something else. Now you're dropped off on the same street blindfolded and asked to walk the same path while naming every step out loud: 42 steps straight, turn right, 3 stairs. The map in your head is the same. The city is the same. But walking by the names of the steps is a completely different job, and you'll crawl, stumble and recount.

— Every step has to be named. The legs that walked this route a thousand times — the left hand doesn't have those.

— 28 days. Walker and colleagues, in 2007, took 21 adults aged 20 to 56 and for a month made them write 2 standard phrases daily with the non-dominant hand. After 28 days the handwriting was of good quality and people were already comfortable with it. Age had no effect on the result.

— The right hand didn't take 20 years. It took roughly the same amount, and the other 19-plus it simply spent using what it had. There's no 20-year debt on the left hand — that's an illusion that comes from counting seniority instead of work.

— There's a neat study by Sawamura, 2019: 32 right-handers, half of them spending 6 weeks learning to eat with chopsticks left-handed, 30 minutes a day, 5 days a week, the rest as controls. By the end, activity in the left dorsolateral prefrontal cortex had dropped, and in both premotor areas had risen. Nothing "got stronger." The work was handed over from the planning department to the shop floor.

— You've got it right. It isn't that you're talentless — but no easy road appeared out of that either. You don't need 20 years; you need the few dozen hours over which the program gets built. You've done about two of them.

— That's Raibert, 1977: he wrote the same phrase 5 ways — normally, with the wrist immobilized, with the left hand, with the pen between his toes and with the pen in his teeth. The lines are crooked, and the handwriting in every version is recognizably his. A lovely story, people love it. But in 1990 Wright recounted it more carefully and showed that the similarity between different effectors is noticeably weaker than the textbook examples suggest.

— It's a historical experiment and a matter of dispute. And I'll tell you honestly where we stand: it seems there's a quickly acquired representation of what to draw — and a slowly accruing one of how exactly to pull. The second doesn't transfer. And where exactly the seam between them runs, and why different tasks give different percentages of transfer — nobody can say yet.