Colemak, Explained

Seventeen keys change position. The home row becomes A R S T D on the left and H N E I O on the right, which puts the most frequent letters in English under the strongest fingers.

What stays put is the more interesting list. Z, X, C and V remain in the bottom left, which means undo, cut, copy and paste are unchanged — and those are pressed constantly by everyone, in every application, regardless of what they are typing. Q and W stay. The punctuation keys on the right stay. A, the single most-used key on the left little finger, does not move.

The practical effect is that a QWERTY typist switching to Colemak keeps a substantial fraction of their existing muscle memory rather than discarding all of it. That is the entire design thesis, and it explains most of the differences from Dvorak, which moved nearly everything.

The design target was reducing same-finger bigrams — sequences where one finger has to type two letters in a row.

Those are expensive for a structural reason: one finger cannot overlap two presses, so it must fully lift and re-place, while a sequence on two different fingers can begin the second press before the first has finished. On QWERTY, common pairs like 'ed' and 'de' land on the same finger, and there is no technique that removes the penalty — only mitigation, which is what the same-finger bigram drill is about.

Colemak's arrangement reduces how often that situation arises in English text. It also increases the proportion of typing done on the home row, which shortens average finger travel.

Those are real properties of the arrangement and they are measurable from letter-frequency data. What they translate to in speed for an individual typist is a much weaker claim, and it is the point at which layout discussions usually overreach.

Dvorak's central idea is hand alternation: arrange the letters so that consecutive characters tend to fall on opposite hands, because the second hand can move while the first is finishing.

Colemak's central idea is same-finger reduction and home-row concentration. These are not the same optimisation and they do not always agree — a layout can score well on alternation and poorly on same-finger repeats, or the reverse.

The visible consequence is in the vowels. Dvorak puts all five vowels together under the left hand, which forces alternation because English words interleave vowels and consonants constantly. Colemak spreads them, keeping A on the left and putting E, I and O on the right home row.

Neither approach is obviously correct, which is a large part of why the alternative-layout world has not converged on one answer. A survey of the wider family and what each was reacting against is in Workman, Norman and the rest.

A widely used variant moves two more keys — D and H — down from the home row into the lower row, on the grounds that the index finger's reach inward along the home row is less comfortable than its reach downward.

The argument is about which movements are actually cheap. The original arrangement treats a sideways stretch of the index finger as low-cost because it stays on the home row; the variant argues that lateral stretching is worse than a short downward curl, because a curl uses the finger's own range while a stretch pulls the whole hand.

This is a real disagreement about hand mechanics rather than about letter frequency, and it is the same question that separates staggered from ortholinear keyboards, discussed in ortholinear and split keyboards.

For anyone considering the switch, the practical note is that the variant is common enough that choosing between them is a decision you have to make up front — switching between them later means relearning a second time.

Colemak reduces same-finger bigrams and finger travel relative to QWERTY. That much is arithmetic on letter frequencies and is not in dispute.

Whether it makes a given person faster is a much weaker claim, for a reason that has nothing to do with the layout: the largest determinants of typing speed are how much you have practised, whether you look at the keys, your error rate and your rhythm. Those swamp the effect of key arrangement for almost everyone below the top end.

The honest summary is that a typist who switches, relearns thoroughly and practises consistently will probably end up somewhat faster than they were — and that the same person doing the same amount of consistent practice on QWERTY would also have ended up faster, by an amount that is hard to separate from the layout's contribution.

Which means the case for switching is rarely a speed case. It is a comfort case, or a curiosity case, or a case about finger travel. Those are legitimate; the speed claim is the one to be sceptical about, and the full accounting is in the real cost of switching layouts.

Expect several weeks below your previous speed, with the worst of it in the first fortnight. The floor is lower than people anticipate — dropping to a small fraction of your QWERTY speed initially is normal, because you are not learning key positions from scratch but competing against an existing set of highly practised ones.

The complication specific to Colemak is that partial overlap with QWERTY cuts both ways. Keeping Z, X, C and V reduces what you must relearn, and it also means your hands are running two overlapping maps at once, which produces a particular kind of error: reaching for a letter and getting its QWERTY neighbour, especially under pressure or fatigue.

The practical advice most switchers converge on is not to hedge. Learning both properly is harder than learning one, and alternating between them during the transition extends it considerably. Either commit for a couple of months or do not start.