Why Your Keyboard Isn't Alphabetical: The QWERTY Origin Story

Christopher Latham Sholes was not the first person to build a typewriter, but he was the first to build one practical enough to sell commercially. Working in Milwaukee in the early 1870s with colleagues Carlos Glidden and Samuel Soulé, Sholes developed the machine that would eventually be manufactured by the Remington Arms Company and become the foundation of modern typing. His first keyboard layouts were experimental and varied, but the version that Remington standardized — with the letters Q, W, E, R, T, and Y across the top row — became what we now call QWERTY.

The Mechanical Problem QWERTY Solved

Sholes's early typewriters used a mechanism in which each key was attached to a typebar — a metal arm with the letter embossed at its tip. When you pressed a key, the typebar swung up to strike the ribbon against the paper. The problem was that if you pressed two adjacent keys in rapid succession, their typebars could collide and jam, requiring manual intervention to untangle them.

The QWERTY layout was designed, at least in part, to minimize jamming by placing commonly used letter pairs on keys whose typebars were not adjacent in the mechanism. By separating letters that frequently appeared together in English — like 'T' and 'H', or 'E' and 'D' — the layout reduced the probability that a fast typist would trigger the mechanical collision problem.

The historical record is somewhat murky on exactly how much of the QWERTY design was driven by mechanical jamming versus other factors, including the need for telegraph operators to efficiently type Morse code transcriptions and the commercial consideration that Remington salespeople could type "TYPEWRITER" using only the top row of keys (as a demonstration trick). What is clear is that the layout was empirically derived through trial and error rather than systematically optimized for human typing speed.

The Dvorak Alternative and Why It Failed

In 1936, university professor August Dvorak developed an alternative keyboard layout — the Dvorak Simplified Keyboard — that was explicitly optimized for typing efficiency. Dvorak analyzed letter frequency in English and placed the most common letters on the home row (the row where your fingers naturally rest), alternated commonly typed sequences between hands to reduce same-hand fatigue, and pushed rare letters to the outer keys. Studies on the Dvorak layout have generally found that it reduces finger movement and can be faster for touch typists to learn from scratch.

The Dvorak keyboard is still available — most modern operating systems allow you to switch to it in settings — but it has never achieved significant adoption. The explanation is almost entirely economic: by the time Dvorak developed his layout, millions of people had already learned QWERTY. The cost of relearning was real and immediate; the benefits were marginal and long-term. Employers who had already trained typists on QWERTY had no incentive to pay to retrain them. The market had locked in.

Path Dependence and the QWERTY Lesson

The persistence of QWERTY has become one of the canonical examples of path dependence in economics — the phenomenon where the outcome of a process depends not just on current conditions but on the historical path that led to them. Once a technology achieves sufficient adoption, the network effects and switching costs can lock it in even if technically superior alternatives exist.

Economists and historians debate how universal this pattern is — some researchers have argued that QWERTY is not as inferior to Dvorak as commonly claimed, and that its persistence reflects not irrationality but genuine optimization at the system level (standardization benefits outweigh individual efficiency gains). But the broader point holds: the layout of the keyboard in front of you was designed to solve a problem in a mechanical typewriter, a machine that ceased to be the dominant writing technology roughly 40 years ago, and yet it remains the global standard because the cost of changing a universal interface is almost never worth paying.