Most of the people who email us saying their Japanese knife "doesn't feel as sharp as expected" are holding it wrong and cutting with it wrong, and they've been doing both for so long with a German knife that it feels natural. The knife is fine. The habits are the problem. I know this because I had the same habits for about a decade, and because last spring I spent a weekend un-teaching them to my brother-in-law, who'd just bought a gyuto and was convinced he'd been sold a dud. By Sunday night he was dicing onions faster than me. So here's the weekend, more or less, in the order it happened.

I'll say up front that none of this is hard. It's just different, and the difference matters more on a thin, hard blade than it ever did on a thick, soft one.

Saturday morning: why his knife kept "going dull"

The first thing I did was watch him cut an onion, and the first thing I saw was the rock. Tip planted on the board, heel going up and down, the blade rolling through the onion the way you see on television. That's a perfectly good technique for the knife it was invented for. A Western chef's knife has a deep curve along the edge, a thick spine, and soft steel, and all three of those are basically an invitation to rock: the curve makes it smooth, the thickness means you can lean on it, and when you twist a bit at the bottom of the stroke the soft steel just shrugs and bends back.

His new gyuto was none of those things. Flatter along the edge, far thinner behind the apex, and hard enough that it doesn't bend at all, it just chips. A santoku or a nakiri is flatter again. So when he rocked it, the flat edge slapped the board instead of rolling, and the little sideways twist at the bottom of every stroke was landing on an edge that was never built to take it. I held the knife up to the window and there they were, a row of tiny bright spots along the edge where it had micro-chipped. Three weeks old. Not the steel's fault. I've had that exact email from strangers a dozen times and the window test always shows the same thing.

What I told him, and what I'd tell you, is that the motion has to change but it changes less than you'd fear. You stop rolling through the food and start moving the blade forward or back while it goes down, so the edge slices instead of crushing. A thin, properly sharp edge barely needs any downward force to do that. The rest of the weekend was really just getting his hands to believe it.

Saturday, still morning: the grip

Before I'd let him cut anything else I fixed his hand, because he was holding the knife like a hammer, all four fingers wrapped round the handle. Everyone does. It feels secure. It also puts the pivot point a long way from the edge, so every cut gets steered from the wrist and the tip wanders off on its own.

The pinch grip is the fix and it's older than any knife we're talking about. Thumb and the side of the index finger go on the blade itself, just in front of the handle, and the other three fingers wrap loosely round the handle. You're holding steel, not wood. He hated it for about an hour. By lunch he'd stopped noticing, and by Sunday he said the knife felt like part of his hand, which is exactly the phrase everyone uses, because the pivot is now right where the cutting happens and the tip goes where you point it.

On a wa-handled knife with no bolster the pinch is especially natural, since there's nothing in the way. On a Western-handled one like his you pinch just forward of the bolster. The one thing I kept repeating was: lighter. If your knuckles are white you're holding it like it might get away, and all that tension comes out the other end as wobbly cuts.

The other hand, which nobody had ever shown him

Then the hand holding the food, which it turned out he'd never thought about at all. Fingertips curled under so the first knuckles face the blade and the tips are tucked behind them. The flat of the blade rests lightly against those knuckles and uses them as a guide, and with each cut the guiding hand walks back a few millimetres and the blade follows.

Two things happen once that's in place. Your fingertips are out of the way, obviously, which he appreciated after the second close call. And your slices start coming out the same thickness without you trying, because the blade can only go where the knuckles let it. Every uniform julienne you've ever admired was cut by someone whose knuckles were doing half the work, whether they knew it or not. I learned the claw embarrassingly late myself and it's the thing I'd go back and teach twenty-year-old me first.

Saturday afternoon: the push cut

With both hands sorted we did the stroke, and I made him do it slowly on a carrot about forty times. The knife starts with the tip a little down and the heel raised, and as the blade comes down you also push it forward, so the edge travels through the food on a diagonal rather than straight down. By the time the edge reaches the board it's moved forward a few centimetres.

That forward movement is the entire difference between a cut and a chop, and he felt it on about the fifth carrot. A sharp edge moving sideways through a carrot parts it with almost no pressure. The same edge pushed straight down has to wedge its way through, and you can hear the difference as much as feel it, a clean whisper against a sort of crack. Onions, carrots, celery, cabbage, potatoes, nearly everything that lands on a board gets the push cut. The how to use a gyuto page has it photographed, which helped him more than my demonstrating did, to be honest.

Late afternoon: the pull cut, on a tomato

The reverse stroke. Heel of the blade on the far side of the food, and you draw the knife toward you while bringing it down, so the edge travels backward through whatever you're cutting. I showed him on a tomato because that's where it makes sense: soft, delicate things that would tear or squash under a push. Tomatoes, fish, cooked meat, ripe fruit, herbs you don't want bruised. It's also the whole basis of single-bevel knife work and the reason a yanagiba is so long, since sashimi gets cut in one pull from heel to tip without stopping.

On a gyuto I use the pull for tomatoes and for slicing a roast, and the push for nearly everything else. A sujihiki is basically a knife built for nothing but pulls. He didn't need to know that on Saturday but he asked, so.

The straight chop, and the one time it's fine

Straight down, no forward or back. I let him keep this one for scallions and herbs and the pile of already-sliced onion he was turning into a dice, because on things that thin there's nothing to wedge. A nakiri does it all day; the flat edge lands along its whole length at once so there's no accordion of uncut bits left hanging. Where it goes wrong is on anything hard and tall. Straight down through a dense carrot and the blade wedges, and when the blade wedges people twist to free it, and that twist is the chip. Which brought us back to Saturday morning.

Why rocking is the habit to break

I'll be fair to the rocking chop: on a German knife, with its curved edge and soft steel, it works and it's fast. The problem is transplanting it to a Japanese blade.

The flat profile means the rock doesn't flow; the edge thumps down and bounces. The thin edge means the sideways load at the bottom of each rock is a chip waiting to happen. And the tip-on-board pivot means you're grinding the thinnest, most fragile bit of the knife against the board a hundred times a minute, which is why rockers' gyutos always have a rounded-off tip after a year.

If you must rock, do it on herbs only, lightly, with the tip barely touching. And consider a bunka or a santoku, which have less tip to worry about. But the honest answer is to replace the rock with the push cut. A week of feeling clumsy and you won't go back.

What not to do with the edge

Scraping. Using the edge to sweep cut food off the board into a pan. Everyone does this and it rolls or chips the apex every time. Flip the knife and scrape with the spine, or use a bench scraper.

Twisting. The one that chips knives. If the blade is stuck in something, stop, pull it straight out, and come at it again. Never lever.

Bones, frozen things, hard squash, pits. A Japanese edge is for slicing food, not forcing through it. Keep a cheap thick knife for those jobs. I've written about squash separately and the rule there applies to everything hard.

Cutting on glass, stone, ceramic, or bamboo. The edge is harder than any of those is kind to. End-grain wood or a soft plastic board, nothing else. The cutting board guide goes through the options.

An onion, start to finish, the Japanese way

Since the onion is where everyone's habits show. I halve it through the root with a single push cut, peel it, and lay the half flat. The guiding hand goes into the claw with the root end under my fingers, because the root holds the layers together. Then horizontal cuts: tip of the blade into the onion parallel to the board, two or three pull strokes, stopping short of the root. Then vertical cuts, tip first, push cut, the knuckles walking back a few millimetres each stroke. Then across, the same push cut, and the dice falls apart on its own. No rocking anywhere. The root end goes in the bin.

First time it takes twice as long as your old way. By the tenth onion it's faster, and the dice is more even than it ever was.

Board height and standing

One thing nobody mentions until a chef points it out. If the board is too high, your elbow is up and the push cut turns into a stab; if it's too low, you hunch and the stroke gets short. The board surface should sit a little below your elbow with your arm relaxed at your side. Most kitchen counters are a touch high for shorter cooks, and a thinner board or a stool fixes it. Stand square to the board, feet apart, and let the stroke come from the shoulder rather than the wrist. It sounds fussy. It's the difference between cutting for an hour comfortably and cutting for twenty minutes and wondering why your forearm aches.

Speed comes last

The people you see on video going through an onion in six seconds learned slow. The claw and the pinch grip and the push cut all have to be automatic before you can add speed, and if you try to be fast before they're automatic you'll cut yourself or the knife. Give it a month at a comfortable pace. Speed turns up on its own, and it turns up safely.

A carrot, since it's the other test

Carrots show up wedging. Peel it, cut a thin strip off one side so it sits flat, and then with the claw guiding, push cut rounds or planks. If the blade sticks halfway through a thick carrot and you feel the urge to lean on it, that's the moment to notice two things: you've stopped moving the blade forward, so it's wedging instead of slicing, and the knife may be a bit thick behind the edge for the job. Keep the forward motion through the whole stroke and the carrot parts without the crack-and-shatter that leaves pieces on the floor. If it still fights you with good technique, the blade probably wants thinning, which is a separate conversation.

Left-handed, briefly

Everything above mirrors. A double-bevel Japanese knife is symmetrical and doesn't care which hand holds it. Single-bevel knives do care, and come in left-handed versions; the single vs double bevel page explains why.

The thing I'd leave you with

The knife is sharper than you think, and it wants less force than you're giving it. Nearly every bad habit on this page comes from trying to make the knife work harder, when the whole design of a Japanese blade is that it shouldn't have to. Lighten the grip, slow the stroke, let the edge do the slicing, and the knife you thought was disappointing turns out to be the one you've been reading about.

Andrew Kuzmin · Editor-in-Chief, Okami Blades