QWERTY
The layout built to slow us down
Description
The keyboard under almost every pair of hands in the English-speaking world is arranged in a specific, peculiar order. Q-W-E-R-T-Y across the top-left of the letter block. A-S-D-F on the home row. The pattern is so familiar that most people who use it every day couldn't tell you why it's arranged that way. It was designed in 1874, for a physical problem that stopped existing over a hundred years ago.
Several measurably better keyboard layouts have been designed since. Dvorak, published in 1936, places the most-used letters under the strongest fingers and reduces finger travel by roughly two-thirds on common English text. Colemak, released in 2006, preserves many QWERTY shortcuts while improving ergonomics. Workman and a dozen other alternatives exist. None of them has displaced QWERTY. QWERTY controls well over 99% of the English-language keyboard market and has held that share for over a century.
The persistence of QWERTY is a useful case for thinking about how technologies actually get chosen. The commonly told version is that markets converge on the best solution. QWERTY shows that this isn't what markets do. Markets converge on solutions that are good enough, adopted early, and embedded in networks of mutually reinforcing practices. Once that happens, the better solution can't win. The keyboard this article is being typed on is a physical monument to that dynamic.
● The question we're asking: why are we still typing on a layout designed in 1874 to solve a mechanical problem that stopped existing a century ago, and why has every better alternative failed to replace it?
● What we'll see: the jamming problem Sholes was solving, the alternative layouts that are measurably faster, the economic concept of path dependence that keeps QWERTY in place, and what that pattern tells us about how technologies actually get chosen.
Table of contents
01The jamming problem of 1874
The QWERTY layout was designed by Christopher Latham Sholes, an American newspaper editor and part-time inventor working in Milwaukee. Between 1868 and 1874, Sholes developed several prototypes of what would become the first commercially successful typewriter. His early keyboards were organized alphabetically. The alphabetical layout had a physical problem. When two typebars were struck in quick succession, they tended to collide and jam at the printing point.
The problem was most common with frequently paired letters — the combinations that appear often in English, like T-H, E-R, I-N, and S-T. If the typebars for those paired letters were next to each other on the machine, the striking motion of the first bar interfered with the motion of the second. The resulting jams had to be cleared by hand, which slowed typing and wore the machine down. The fix was mechanical but also counterintuitive: arrange the keys so the commonly paired letters were physically far apart on the machine.
02The layouts that should have won
The first serious attempt to design a better layout came in 1936 from August Dvorak, a professor of educational psychology at the University of Washington. Dvorak studied English letter frequencies and finger motion carefully and arrived at a layout built around a different principle. The most-used letters — A, O, E, U, I on the left, D, H, T, N, S on the right — would sit on the home row. The least-used letters would be on the bottom row. The result was a layout where, for typical English text, roughly 70% of keystrokes happened on the home row.
Dvorak's evidence that his layout was faster came mainly from training studies. The United States Navy conducted a well-known experiment during World War II, training typists in Dvorak and comparing their speed to QWERTY typists. The Navy reported a roughly 74% increase in productivity from the Dvorak-trained group. Other studies have reported smaller but still substantial gains — 15% to 35% — along with reduced finger travel, lower error rates, and less typing fatigue.
03Why the switch never happens
The economist Paul David gave the phenomenon its modern name in a 1985 paper titled Clio and the Economics of QWERTY. David argued that QWERTY was the clearest available example of what he called path dependence — a situation in which an early choice locks in a later equilibrium that no individual actor has the incentive to change, even if a better alternative exists. His analysis became one of the most-cited papers in institutional economics and is the standard reference for the concept.
The mechanism is straightforward. Every typist who learns QWERTY produces one more person in a QWERTY-trained labor pool. Every typewriter and every keyboard manufacturer that builds QWERTY hardware reinforces the prevailing format. Every secretarial school that teaches QWERTY maintains the supply of QWERTY-trained workers. Each of these decisions, taken individually, is rational — nobody wants to be the oddball who learned a layout that nobody else uses. In aggregate, they produce an outcome that locks the system in place.
04What the pattern teaches
QWERTY is the canonical example, but the pattern generalizes. The gauge of modern railroads — 4 feet 8.5 inches — descends from the width of horse-drawn carts in Roman Britain, because early railway engineers reused existing wagon-building tools. The gauge was never the optimal choice for rail; it was the adopted choice, and every subsequent rail system had to match to remain interoperable. The current global rail network is shaped by a decision about Roman-era carts.
The QWERTY pattern appears throughout technology adoption. VHS defeated Betamax not because it was a better format but because it won earlier on the adult-film market and captured the video-rental network first. Windows captured corporate desktops in the 1990s despite several technically superior operating systems because IBM signed one deal in 1981. The metric system never captured the United States because the American industrial base was too deeply invested in imperial units by the time the metric system had consolidated globally.
05Conclusion
In 1874, a newspaper editor in Milwaukee arranged the keys of a typewriter in an order designed to prevent mechanical typebars from jamming. A hundred and fifty-two years later, a laptop keyboard in 2026 still has the same arrangement. The mechanical problem Sholes was solving no longer exists. The layout that was a workaround to that problem has outlived the problem by a century and a half, and will likely continue to outlive it for the foreseeable future.