The anarchy of the keyboard

Every once in a while, one runs into a nerd who rants to no end about how inefficient standards like QWERTY have taken over the world in spite of better solutions like Dvorak being available.

I have been told numerous times that QWERTY is really bad design and an ergonomic abomination. Christopher Latham Sholes invented the QWERTY keyboard in 1868 for typewriters in order to slow down the typists, and prevent the keys from jamming. Occasionally, someone brings up the fact that the first line of the keyboard (QWERTYUP) contains the word "typewriter", as an illustration of the assertion that this keyboard was developed for typewriters, not computers. The rights to this "new and improved" keyboard layout were sold to E. Remington and Sons, who promptly commercialized it.

Since then, so many people have gotten used to the layout that there is a lock-in/ path-dependence due to collective switching costs. The world could be a slightly better place if everyone switches to a more ergonomic layout such as the Dvorak keyboard.

The Dvorak keyboard was created in 1936 by Dr. August Dvorak, who was hugely inspired by Frank Gilbreth, the father of time and motion (efficiency) study. He called it the American Simplified Keyboard (ASK), which was the fruit of his ten years of research on increasing efficiency and reducing strain.

It would be reasonable to assume then, that such a well-researched, ergonomically-designed layout will win people over overnight and boost the efficiency of the world. However, there is one little problem called network effects. You may also call it catch 22. Typists will not train on Dvorak keyboards as they are hard to find in offices. And companies will not equip offices with Dvorak keyboards because typists trained on Dvorak are hard to find. And hence, the world lives on with a self-reinforcing inefficient layout.

Or so the story goes. The inspired anti-establishment enthusiasm of the Dvorak supporters is amusing because it is so misplaced. There is no empirical evidence of the superiority of the Dvorak keyboard. The most popular source cited as proof is a paper dated 1936 written by Dvorak et al. Of course Mr. Dvorak thought his keyboard was better.

Are there any impartial studies done on the subject? Well, it is claimed that the Navy ran an extensive test, though the Navy itself says, "we have no record of and did not conduct such a speed test." Further research reveals that there is, in fact, such a report in existence.

Unfortunately for Dvorak fans, the test was performed on severely under-qualified typists. Further, the experiment was conducted by "Lieutenant Commander August Dvorak" (surprise, surprise).

On the other hand, the case against Dvorak appears to be much more solid. According to Strong (1956), research shows that there is no advantage of retraining on Dvorak vs. retraining on QWERTY. A Miller and Thomas study (1977) found no significantly superior keyboard to QWERTY.

Norman and Rummelhart (1983) demonstrated most emphatically that there is no ergonomic advantage to Dvorak over QWERTY. Both layouts have about equal loads on right and left hand, with the split for Dvorak being 47-53 and that for QWERTY 57-43, making Dvorak only marginally better. Both keyboards maximize the load on the middle row - Dvorak about 67% and QWERTY just over 50%, making QWERTY significantly better. By putting successively typed keys apart, the QWERTY maximizes the frequency of alternating hand sequences and minimizes the frequency of same-finger typing.

These findings, though relatively recent, are hardly surprising. After all, QWERTY was not the first attempt at keyboard layout. No less than 51 inventions predated it. In fact, after Remington commercialized the design, it was met with fierce competition from other designs. Of course, QWERTY prevailed, and won many typing competitions at the end of the 19th century.

It is possible that a win in that era was a win for the typewriter layout, not for a computer keyboard. The QWERTY may today be inefficient, and may have given Carpal Tunnel Syndrome (a repetitive stress injury) to many a typist, but there is not a shred of evidence to suggest that Dvorak's ASK layout is any better. Nor that there exists any other layout better than good old QWERTY. Till such an alternative is invented, rest in peace all you keyboard anarchists.