August 27, 2007

The Dvorak Keyboard Controversy

Chasing QWERTY

When I was first learning to type, many years ago, I asked the same question everyone else asks: why are the keys arranged so stupidly? Why aren’t they laid out in a more logical order, as in, just to take one random example, alphabetically? The answer I’ve heard countless times is that the first typewriter keyboards were arranged alphabetically, but that caused mechanical problems—once typists became reasonably proficient, the keys jammed frequently because the hammers corresponding to certain frequently used letter sequences were too close together. As a result, so the story goes, the QWERTY layout was designed to prevent jamming by moving those letters farther apart, thus slowing down the typists to a speed the machine could handle. Meanwhile, a more sensible and efficient layout called the Dvorak Simplified Keyboard has been around for a long time, but never became very popular because QWERTY simply had too much momentum in the marketplace.

That story has frequently been used (even by me) as an example of how an inferior, inefficient design came to be the standard—and remained so, long after the original reasons for its success became irrelevant. But the truth is more complicated and surprisingly controversial.

Faster than a Speeding Typist
The first issue is whether the switch to the QWERTY layout was truly intended to slow down typists—whether it was a matter of deliberately inconveniencing people for the sake of machines or whether it was simply expedient engineering. I have never seen any statistics as to how fast anyone could type using the original, alphabetical keyboard layout, but I think it’s fair to imagine that it would be faster only for people who have to look at the keys while they type. For well-trained touch-typists, I very much doubt that an alphabetical layout would yield any speed increase over QWERTY, and there is some evidence to suggest exactly the opposite. QWERTY may have its faults, but it seems to me that one need not get upset (as many have) that we’re all using a layout that’s substantially worse than the original.

But is there some layout that’s demonstrably much better than QWERTY? A lot of people think the Dvorak Simplified Keyboard meets that description. Dr. August Dvorak, a professor at the University of Washington in Seattle, designed this alternative layout in 1932 and patented it in 1936. Dvorak’s goal was to reduce typing fatigue by minimizing finger movement, so he put the most commonly used letters (including all the vowels) on the “home” row, while placing letters like Q and Z and certain punctuation characters in spots that are harder to reach. The Dvorak layout also favors the right hand, on the grounds that the majority of people are right-handed. Dvorak naturally claimed that his design was much better than QWERTY, but we need not take his word for it. A study performed by the U.S. Navy in 1944 showed speed improvements of as much as 75% when people who had previously learned the QWERTY layout were retrained on Dvorak keyboards. And that, say Dvorak supporters, should be that.

Dark Days for Dvorak
However, another study—this one performed in 1956 by the U.S. General Service Administration (GSA)—failed to confirm the results of the Navy test. It found, in a nutshell, that QWERTY was at least as efficient as Dvorak, and possibly more so. Further research conducted from the 1950s through the 1970s showed little or no advantage for Dvorak.

Two economists, Stan Liebowitz and Stephen E. Margolis, have written extensively about the Dvorak-versus-QWERTY debate. They mention the research suggesting QWERTY’s superiority and point out a number of significant flaws in the 1944 Navy study, not the least of which is the fact that Dvorak himself apparently oversaw that research in some capacity. In short, they say, the Dvorak layout isn’t and never was any better than QWERTY, and the only thing the pro-Dvorak studies really prove is that anyone who is retrained on any keyboard layout will get better. Therefore, no one should use QWERTY as an example of the free market “choosing” an inferior technology.

Rejoinder to the Rebuttal
But wait! Dvorak’s not out yet. Supporters of Dvorak claim that Liebowitz and Margolis have an axe to grind, and that in the very process of showing how earlier research was fudged, they fudged facts themselves. Specifically, pro-Dvorak folks say, the economists failed to mention that Earle Strong, who was in charge of the 1956 GSA study, had a personal grudge against Dvorak and had made public statements before that study was even performed voicing his opposition to any alternative keyboard layout. Randy C. Cassingham, author of the 1986 book Dvorak Keyboard: The Ergonomically Designed Keyboard, Now an American Standard (an unbiased title if ever I heard one), attempted to debunk Liebowitz and Margolis’s findings soon after they were originally published, but his work has been little noticed—except by other die-hard Dvorak fans looking to bolster their position.

The ferocity with which both pro- and anti-Dvorak views are evangelized in some circles rivals that of a religious or political cause. Both sides selectively downplay or emphasize whichever facts suit them best, and there’s precious little research on the subject that’s both truly objective and modern enough to have been performed using computers rather than typewriters. Anecdotally, Dvorak users frequently cite greater comfort as one reason for preferring it, and some claim that because Dvorak involves less finger movement, it’s less likely to contribute to repetitive stress injuries. Opponents counter that if you truly can type faster with Dvorak, then the increased number of movements will offset the ergonomic gains made by the decreased range of motion. And the debate goes on and on

The Best Test
All modern computers—Macs, PCs, and Linux machines—include the capability of switching into a Dvorak layout if that’s what you prefer (though the physical keys won’t match the characters they type unless you perform some minor surgery on your keyboard or put stickers over the existing letters). So if you want to try out Dvorak yourself, you need only consult your computer’s Help to find out how to change that setting. Chances are you’ll find that it takes a few weeks or so to retrain yourself to the point where you’re about as fast as you were using QWERTY. If you can tolerate the temporary loss in productivity, you may find the experiment useful.

I tried learning Dvorak myself, years ago, but gave up before I became proficient—I had to get my work done. On the other hand, I can already type as fast as I should ever need to using QWERTY. For me, the bottleneck is usually how fast I can think, not how fast I can type, so I suspect Dvorak would not make my life meaningfully better. If I spent all day, every day, doing straight transcription or had a quota to meet, though, I might relish any chance for a potential speed increase. Better yet, I might start looking for a job that didn’t involve typing at all. —Joe Kissell

More Information about The Dvorak Keyboard Controversy…

Anti-Dvorak articles by Liebowitz and Margolis include:

Pro-Dvorak articles include:

Other articles about the debate include:

Randy C. Cassingham’s book Dvorak Keyboard: The Ergonomically Designed Keyboard, Now an American Standard is now out of print but may be available used, and may or may not shed any light on the situation.

The Wikipedia has articles on the Dvorak Simplified Keyboard and the QWERTY keyboard.