Look down at a laptop keyboard. Its keys no longer pull metal arms towards an inked ribbon, yet the top letter row still begins QWERTY. An arrangement developed for nineteenth-century machinery remains the default for much of the English-language computing world. How did a layout survive when the machine it belonged to did not?
The short answer is that a keyboard is more than a set of keys. It is also a skill people learn, a design manufacturers make, and an expectation shared across workplaces and devices. Once those things line up, replacing the layout becomes a collective project. A better-looking plan on paper is not enough.
The origin story is messier than the myth
The Sholes and Glidden Type Writer, made by Remington and put on sale in 1874, was the first commercially successful typewriter. The machine in the Smithsonian's collection has a recognisable QWERTY arrangement.¹ Early versions of the device went through many revisions, so the familiar layout was the result of development rather than one flash of inspiration.²
You may have heard that its inventor, Christopher Latham Sholes, deliberately scattered letters to slow people down and stop the type bars jamming. The mechanical problem was real, and the Smithsonian's account links the layout to preventing type bars from clashing.¹ But the stronger claim that QWERTY was intended to make typists inefficient is poorly supported. A historical study by Koichi and Motoko Yasuoka traces the arrangement through earlier printing telegraph designs and feedback from Morse operators. Its authors argue that several practical demands and revisions shaped the final result.² Historians do not agree on a single, tidy cause.
There is a useful lesson here: an old design can be a response to the constraints of its time without being an intentional handicap. By the time those constraints disappeared, other reasons to keep it had grown.
The more people use it, the more useful it becomes
Imagine an office buying typewriters in the late nineteenth century. It wants machines that trained typists can use. A typing school wants to teach the layout employers are buying. A typist wants a skill that transfers between jobs. Each choice makes the same choice easier for the next person.
Economist Paul David used QWERTY to illustrate this kind of path dependence in a famous 1985 paper. Early adoption, learning and compatibility can reinforce one another, allowing a standard to persist even after the original reason for it fades.³ When typewriters gave way to electric machines and then computers, retaining familiar keys spared users a fresh course of lessons. The continuity had value of its own.
That does not prove society is trapped with a bad keyboard. Economists Stan Liebowitz and Stephen Margolis later challenged David's QWERTY example. They argued that the evidence did not establish that the leading alternative, Dvorak, was clearly superior or that the market had rejected a large, demonstrable gain.⁴ Their disagreement matters because two questions are often bundled together: Why does a standard persist? and Would another standard make us better off? The first can be answered with habits and coordination; the second needs convincing comparisons.
Why a rival has not swept it away
The Dvorak layout places common English letters differently and was designed with efficient typing in mind. It has loyal users. But comparing layouts fairly is difficult. A lifelong QWERTY typist will be slow on a layout learnt yesterday; someone trained on Dvorak is not an impartial novice. Text, practice time, errors and physical comfort all need measuring, and results for one task need not apply to every typist.
In a series of experiments, psychologists Donald Norman and Diane Fisher found that alphabetically arranged keyboards were no easier for novices than QWERTY. They reported only a modest performance edge for Dvorak in their comparisons, around five per cent, and concluded that switching established typists was unlikely to be worthwhile on speed alone.⁵ Their experiments are not a final ranking of every layout for every person. They do show why the claim that QWERTY must be vastly slower than its rivals deserves scrutiny.
Today the technical barrier to trying something else is small. Macs offer Dvorak layouts, and Windows lets users add or switch keyboard layouts.⁶,⁷ Yet a software setting does not relabel the keys under your fingers or erase years of muscle memory. It also does not change the keyboard on a colleague's desk, a borrowed laptop or the computer you meet in a public place. For most people, any possible gain arrives later, after the immediate inconvenience of learning.
That is QWERTY's durable advantage. It need not be the fastest conceivable arrangement. It has to be familiar enough, good enough, and available almost everywhere its users expect it.
A standard, not a law of nature
QWERTY is not universal. Languages and countries have developed their own physical and software keyboard arrangements; Windows alone lists many regional defaults.⁸ Nor must a modern keyboard obey a nineteenth-century machine: people can choose other layouts, use touchscreens or dictate text. The continued availability of alternatives makes the persistence of QWERTY more revealing. The obstacle is no longer metal type bars. It is the cost of coordinating a change across people, products and habits.
That cost is easy to miss because it is spread around. It sits in the time spent practising, the moment of hesitation at an unfamiliar desk, and the fact that every familiar keyboard is a little invitation to keep using the one you know. QWERTY won its place through history. It keeps it because, for most typists, a proposed replacement still has to beat the whole system built around it.