The QWERTY Keyboard

Look at your keyboard. The top row of letters starts Q, W, E, R, T, Y. This is called the QWERTY layout. Why are the letters in this order and not A, B, C?

The layout comes from early typewriters in the 1870s, made by a man named Christopher Sholes. Old typewriters had metal arms with letters on the end. When you pressed a key, an arm hit the paper. If you typed two nearby letters very fast, the arms could hit each other and stick together.

The common story is that Sholes put common letter pairs far apart. This slowed the typist a little, so the machine did not jam. Later, one company, Remington, sold many machines with this layout, and it became the standard.

Better layouts were invented later. But people had already learned QWERTY. Teachers taught it. Factories made it. Changing everything was too hard and too expensive. So a 150-year-old design is still on almost every keyboard and phone today.

The arrangement of letters on almost every keyboard and touchscreen traces back to a mechanical problem solved in the 1870s. Early typewriters used pivoting metal bars, each carrying a letter, that swung up to strike the paper. Type two adjacent bars in quick succession and they could collide and jam.

Christopher Latham Sholes and his collaborators reorganised the keys over several iterations. The traditional explanation is that they deliberately separated frequently paired letters so that the corresponding bars were less likely to clash — accepting a slightly slower typing speed in exchange for fewer jams. Historians have complicated this: the layout may also have been shaped by the needs of telegraph operators transcribing Morse code, and by commercial decisions once the Remington company began mass-producing the machines.

Whatever the exact mix of reasons, the layout spread, and then it stuck. Typing schools taught it, manufacturers built it, and every new typist added to the pool of people whose skills were tied to it.

Alternative layouts designed for speed and comfort, such as Dvorak in the 1930s, never displaced it. The problem was not that QWERTY was good but that switching required millions of people to relearn a skill simultaneously. Economists call this "path dependence": an early choice becomes locked in, not because it is best, but because the cost of coordinating a change is too high.

The QWERTY keyboard is the textbook illustration of path dependence — the idea that history, not just present merit, determines which technologies we end up with. Its layout was fixed by the constraints of a specific machine, the 1870s type-bar typewriter, in which letters sat on bars that swung to a common printing point and fouled one another if neighbouring bars fired in quick succession.

The received account is that Sholes and his associates arranged the keys to space out common digraphs, trading a marginal loss of speed for mechanical reliability. Revisionist historians have muddied that story, pointing to the influence of telegraphers who used the machines to transcribe incoming Morse, and to path-shaping commercial choices after Remington took the design to market. The precise causal weighting is probably unrecoverable. What is clear is that the layout was contingent — a product of transient engineering and business circumstances rather than any analysis of the hand.

By the time the type-bar problem had been engineered away, QWERTY was entrenched. A generation of typists had invested in the skill; a training industry had standardised on it; manufacturers competed within the layout rather than over it. When August Dvorak patented a demonstrably more ergonomic arrangement in 1936, it faced not a technical barrier but a coordination one: any individual switcher pays the full retraining cost immediately while the collective benefit only materialises if nearly everyone switches together.

Whether Dvorak is meaningfully faster remains disputed — some influential comparisons were methodologically weak, and the modern consensus is that the gap, if it exists, is small. But that ambiguity is almost beside the point. The interesting lesson is structural: once a standard achieves critical mass, the market stops selecting for quality and starts selecting for compatibility, and improvements have to clear a bar that has nothing to do with how good they are.