An Argument for Dvorak By hjhornbeck in Science Fri Aug 29, 2003 at 05:39:22 PM EST Tags: Science (all tags)
	I use the Dvorak keyboard layout for my computer. I learnt about it by chance, through a Discover Magazine article I laid my eyes on. I did some more research, and the arguments in favour Dvorak were strong enough to cause me to switch. I also turned up a few counter-claims, that QWERTY is a better layout than Dvorak. Most of those pointed to one or two studies that seemed to prove as much. Pro-Dvorak supporters would counter by pointing out flaws in those studies, and very quickly both sides would try to bury each other in scientific studies. I think many of them miss the point.

The arguments in favour of Dvorak have a lot to do with human motion, and are simple enough to be evaluated on their own terms. Scientific studies are overkill; I think I can make a really good argument with only a short essay. Most of my argument rests on the specifics of touch typing, so I'll take a paragraph to describe it. By default, all your fingers rest on what's known as the "home row", the middle row of letters on a keyboard. They will move the "upper row" above it, or the "lower row" below, as other keys need to be pressed. Your thumbs rest on the space-bar. You may have noticed that two letter keys have raised markers or dots on them; these are where your index fingers are supposed to sit. In the QWERTY layout, they are the F and J keys. This has been studied heavily by several researchers, one of which was August Dvorak. He spent a good decade discovering the mechanics of touch typing, and used this work to design his keyboard layout. Some of his discoveries were: It takes less effort to press a key on the home row than on the upper row. It takes less effort to press a key on the upper row than on the lower row. The digits on your hand vary in strength. The strength of your digits roughly decreases from thumb to pinkie finger. It is easier if the hands alternate when typing as much as possible. You do not need a full study to test these for yourself. It takes more force to curl your fingers than stretch them out, for instance, so reaching the upper row is easier than reaching the lower row. If your hands alternate when typing, one can move itself into position for the next keystroke while the other is typing, balancing the amount of work both hands have to do. It avoids long sequences of typing with just one hand, and overuse can lead to injury. In addition to his keyboard studies, Dvorak also studied the English language to make sure that previously discovered statistics were still true. They were, and over half a century later they have changed little: The most common letters are E, T, O, A, I, and N, though the exact order of the last four can vary depending on the text measured. About 1/8th of all the letters typed in English will be E, and over 1/2 will be one of those six. Consonants and vowels are commonly found in pairs. You can test these facts quite easily, though it may take some time unless you can get a computer to do all the tallying. If you believe the above, we can make some predictions about what an ideal keyboard layout would look like. By combining facts #1, #2, and #6,you can deduce that an ideal layout would have those six letters on the home row. Facts #5 and #7 imply that the vowels are best grouped on one side of the keyboard, under one hand; this is the easiest way to get the alternation desired by #5. Adding on facts #3 and #4 suggest that E, T, O, and N should be under your index and middle fingers. You should not be surprised that the Dvorak layout follows this very closely. Those six letters can be found on the home row. All vowelsare located on the home row, under the left hand, and Y is in the upper row by the left index finger. Curiously, the Dvorak layout places E and O under the middle and ring fingers respectively, not quite what we expected. Fortunately there's a reason for this difference; it is to take some load off the index finger, which controls twice as many letters as the others. T and N are under the right middle and ring fingers, for the same reason. You should also have guessed that the QWERTY layout differs greatly from this. Only A is on the home row; E, T, O, and I are all on the upper row; and N is on the bottom. The vowels are spread across both hands. E and T are above the left middle and index fingers, while O is above the right ring finger, and N is below the right index. Clearly, the Dvorak layout fits our ideal much closer than QWERTY. But is that because our facts are skewed? It's possible, but they are pretty simple. You can test the first four easily if you have a keyboard in front of you. Fact #5 takes a bit more faith, but even if you ignore it my arguments are still quite strong. The last two are verified easily by an Internet search. So long as you accept the majority of these, Dvorak must be better than QWERTY for typing English text. No studies necessary. Naturally, you don't have to take my word for it. Here are some references: The Fable of the Keys, by SJ Liebowitz and Stephen E Margolis; the most commonly cited article by Dvorak dissenters The Curse Of QWERTY, by Jared Diamond; the Discover Magazine article that inspired me a listing of letter frequencies, including a bar graph more letter frequency pages, courtesy of Google (note that while few of them agree with my letter order, nonerank S,D,K, or L highly either) (edited by arca)

Poll

What keyboard layout do you use? QWERTY 70% Dvorak 17% I don't type in English! 8% D. None of the above 3% Votes: 131 Results | Other Polls

Related Links

Google Dvorak keyboard layout The Fable of the Keys, by SJ Liebowitz and Stephen E Margolis The Curse Of QWERTY, by Jared Diamond a listing of letter frequencies more letter frequency pages Also by hjhornbeck